Heat Transfer Module

User’s Guide
Heat Transfer Module User’s Guide
© 1998–2021 COMSOL
Protected by patents listed on www.comsol.com/patents, or see Help>About COMSOL Multiphysics on the
File menu in the COMSOL Desktop for a less detailed lists of U.S. Patents that may apply. Patents pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement (www.comsol.com/comsol-license-agreement) and may be used or copied only under the terms
of the license agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler,
COMSOL Server, and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see www.comsol.com/trademarks.
Version: COMSOL 6.0

Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general inquiries
or search for an address and phone number. You can also visit the Worldwide Sales Offices
page at www.comsol.com/contact/offices for address and contact information.

If you need to contact Support, an online request form is located at the COMSOL Access
page at www.comsol.com/support/case. Other useful links include:

• Support Center: www.comsol.com/support

• Product Download: www.comsol.com/product-download
• Product Updates: www.comsol.com/support/updates
• COMSOL Blog: www.comsol.com/blogs
• Discussion Forum: www.comsol.com/forum
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/videos
• Support Knowledge Base: www.comsol.com/support/knowledgebase

Part number: CM020801

C o n t e n t s

Chapter 1: Introduction

About the Heat Transfer Module 22

Why Heat Transfer Is Important to Modeling . . . . . . . . . . . . 22
How the Heat Transfer Module Improves Your Modeling . . . . . . . 23
The Heat Transfer Module Physics Interface Guide . . . . . . . . . . 23
Common Physics Interface and Feature Settings and Nodes . . . . . . 31
The Heat Transfer Module Study Capabilities . . . . . . . . . . . . 31
Additional Materials Database . . . . . . . . . . . . . . . . . . 36
Where Do I Access the Documentation and Application Libraries? . . . . 36

Overview of the User’s Guide 40

Chapter 2: Notations

Symbols 46

Chapter 3: Modeling with the Heat Transfer Module

Heat Transfer Variables 62

Predefined Variables . . . . . . . . . . . . . . . . . . . . . 62
Global Variables for Heat and Energy Balance . . . . . . . . . . . . 66
Domain Fluxes. . . . . . . . . . . . . . . . . . . . . . . . 69
Out-of-Plane Domain Fluxes . . . . . . . . . . . . . . . . . . 72
Boundary Fluxes (Heat Transfer Interface) . . . . . . . . . . . . . 73
Internal Boundary Heat Fluxes. . . . . . . . . . . . . . . . . . 74
Internal Boundary Energy Fluxes . . . . . . . . . . . . . . . . . 75
Domain Heat Sources . . . . . . . . . . . . . . . . . . . . . 76
Boundary Heat Sources . . . . . . . . . . . . . . . . . . . . 76
Line and Point Heat Sources . . . . . . . . . . . . . . . . . . 77
Moist Air Variables . . . . . . . . . . . . . . . . . . . . . . 77

CONTENTS |3
 Moisture Transport Variables 80
Predefined Variables . . . . . . . . . . . . . . . . . . . . . 80
Moist Air Properties . . . . . . . . . . . . . . . . . . . . . 82
Domain Moisture Fluxes . . . . . . . . . . . . . . . . . . . . 83
Boundary Moisture Fluxes . . . . . . . . . . . . . . . . . . . 85
Domain Moisture Source. . . . . . . . . . . . . . . . . . . . 86
Mass Balance . . . . . . . . . . . . . . . . . . . . . . . . 86

Using the Boundary Conditions for the Heat Transfer

Interfaces 87
Temperature and Heat Flux Boundary Conditions . . . . . . . . . . 87
Overriding Mechanism for Heat Transfer Boundary Conditions . . . . . 88

Handling Frames in Heat Transfer 91

Heat Transfer Analysis with Moving Frames. . . . . . . . . . . . . 91
Material Density in Features Defined in the Material Frame . . . . . . . 96

Heat Transfer Consistent and Inconsistent Stabilization

Methods 97
Consistent Stabilization . . . . . . . . . . . . . . . . . . . . 97
Inconsistent Stabilization . . . . . . . . . . . . . . . . . . . . 98

Heat Transfer and Fluid Flow Coupling 99

Coupling Based on Model Inputs . . . . . . . . . . . . . . . . . 99
Adding Nonisothermal Flow Coupling in an Existing Model . . . . . . 100
Nonisothermal Flow and Conjugate Heat Transfer Multiphysics
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 100

Boundary Wall Temperature 102

Solver Settings 107

Linearity Property of the Temperature Equation . . . . . . . . . . 107
Linear Solver . . . . . . . . . . . . . . . . . . . . . . . 108
Nonlinear Solver . . . . . . . . . . . . . . . . . . . . . . 109
Time-Dependent Study Step . . . . . . . . . . . . . . . . . 113
Guidelines for Solving Surface-to-Surface Radiation Problems . . . . . 114
Guidelines for Solving Multiphysics Problems . . . . . . . . . . . 115

4 | CONTENTS
Plotting and Evaluating Results in Layered Materials 118
Plotting Along and Through the Layered Material . . . . . . . . . 118
Built-in Operators for Evaluation in the Layered Material . . . . . . 119

Evaluating the Bulk Temperature 123

Bulk Temperature Definition . . . . . . . . . . . . . . . . . 123
Evaluation of the Bulk Temperature in 3D Models . . . . . . . . . 124
Evaluation of the Bulk Temperature in 2D Models . . . . . . . . . 124

Ambient Properties 127

Settings for the Ambient Properties. . . . . . . . . . . . . . . 127
Processing of ASHRAE Data . . . . . . . . . . . . . . . . . 131
Ambient Data Interpolation . . . . . . . . . . . . . . . . . . 132
Ambient Variables and Conditions . . . . . . . . . . . . . . . 133
Use of Ambient Data From the Features . . . . . . . . . . . . . 141

Modeling Heat Transfer by Radiation 143

Interfaces and Features for Radiation in Transparent Media . . . . . . 145
Interfaces and Features for Radiation in Participating Media . . . . . . 147

Specifying the Temperature at a Flow Inlet 150

Using the Temperature Condition . . . . . . . . . . . . . . . 150
Using the Inflow Condition . . . . . . . . . . . . . . . . . . 150
Temperature Contribution to the Inflow Boundary Condition. . . . . 152
Pressure Contribution to the Inflow Boundary Condition . . . . . . 152
Choosing Between the Temperature and the Inflow Conditions . . . . 153

Heat Part Library 154

Heat Part Library Contents . . . . . . . . . . . . . . . . . . 154
Heat Sink — Parameterized Fin Types Part . . . . . . . . . . . . 157
Heat Sink — Dissimilar Border Pins Part . . . . . . . . . . . . . 160
Heat Sink — Pin Fins Part . . . . . . . . . . . . . . . . . . 160
Heat Sink — Straight Fins Part. . . . . . . . . . . . . . . . . 160

CONTENTS |5
 References 161

Chapter 4: Theory for the Heat Transfer Module

Foundations of the General Heat Transfer Equation 165

Thermodynamic Description of Heat Transfer. . . . . . . . . . . 165
The Physical Mechanisms Underlying Heat Transfer. . . . . . . . . 170
The Heat Balance Equation . . . . . . . . . . . . . . . . . . 171
Consistency with Mass and Momentum Conservation Laws. . . . . . 174

Theory for Heat Transfer in Solids 176

Theory for Heat Transfer in Fluids 177

Theory for the Inflow Boundary Condition . . . . . . . . . . . . 178

Theory for Heat Transfer in Moist Air 180

Humidity . . . . . . . . . . . . . . . . . . . . . . . . . 180
Saturation State . . . . . . . . . . . . . . . . . . . . . . 182
Moist Air Properties . . . . . . . . . . . . . . . . . . . . 182

Theory for Bioheat Transfer 187

The Bioheat Equation . . . . . . . . . . . . . . . . . . . . 187
Damaged Tissue . . . . . . . . . . . . . . . . . . . . . . 187

Theory for Heat Transfer in Porous Media 191

When Should Local Thermal Equilibrium and Nonequilibrium Be
Considered? . . . . . . . . . . . . . . . . . . . . . . 192
Local Thermal Equilibrium . . . . . . . . . . . . . . . . . . 193
Local Thermal Nonequilibrium . . . . . . . . . . . . . . . . 195
Packed Bed of Pellets . . . . . . . . . . . . . . . . . . . . 197
Theory for Heat Transfer in Moist Porous Media . . . . . . . . . 200

Theory for Heat Transfer with Phase Change 202

Heat Transfer with Phase Change . . . . . . . . . . . . . . . 202
Arbitrary Lagrangian-Eulerian Method . . . . . . . . . . . . . . 202
Apparent Heat Capacity Method . . . . . . . . . . . . . . . . 204

6 | CONTENTS
Theory for Heat Transfer in Building Materials 208

Theory for Harmonic Heat Transfer 210

Theory for Lumped Isothermal Domain 212

Theory for Heat Transfer in Thin Structures 215

Modeling Layered Materials . . . . . . . . . . . . . . . . . . 215
Theoretical Background of the Different Formulations . . . . . . . 220
Thin Layer . . . . . . . . . . . . . . . . . . . . . . . . 225
Thin Film . . . . . . . . . . . . . . . . . . . . . . . . . 229
Fracture . . . . . . . . . . . . . . . . . . . . . . . . . 231
Thin Rod . . . . . . . . . . . . . . . . . . . . . . . . . 232

Theory for the Lumped Thermal System Interface 234

Temperature Differences and Heat Rates in the Lumped Thermal
System . . . . . . . . . . . . . . . . . . . . . . . . 234
Theory for the Conductive Thermal Resistor Component . . . . . . 236
Theory for the Radiative Thermal Resistor Component . . . . . . . 240
Theory for the Thermal Capacitor and Thermal Mass Components . . 242
Theory for the Heat Pipe Component . . . . . . . . . . . . . . 243
Theory for the Thermoelectric Module Component . . . . . . . . 245

Theory for Surface-to-Surface Radiation 249

Deriving the Radiative Heat Flux for Opaque Surfaces . . . . . . . . 249
Deriving the Radiative Heat Flux for Semitransparent Surfaces . . . . 250
Wavelength Dependence of Surface Properties . . . . . . . . . . 252
Directional Dependence of Surface Properties . . . . . . . . . . 258
The Radiosity Method for Diffuse-Gray Surfaces . . . . . . . . . . 259
The Radiosity Method for Diffuse-Spectral Surfaces. . . . . . . . . 262
View Factor Evaluation . . . . . . . . . . . . . . . . . . . 265

Theory for Radiation in Participating Media 270

Radiation and Participating Media Interactions . . . . . . . . . . . 270
Radiative Transfer Equation . . . . . . . . . . . . . . . . . . 271
Boundary Condition for the Radiative Transfer Equation . . . . . . . 273
Heat Transfer Equation in Participating Media . . . . . . . . . . . 275
Discrete Ordinates Method (DOM) . . . . . . . . . . . . . . 276

CONTENTS |7
 Discrete Ordinates Method Implementation in 2D . . . . . . . . . 278
P1 Approximation Theory . . . . . . . . . . . . . . . . . . 280
Radiation in Absorbing-Scattering Media Theory . . . . . . . . . . 283
Polychromatic Radiation . . . . . . . . . . . . . . . . . . . 285
Radiative Beam in Absorbing Media Theory . . . . . . . . . . . . 285
Rosseland Approximation Theory . . . . . . . . . . . . . . . 286

Theory for Moisture Transport 288

Theory for Moisture Transport in Air . . . . . . . . . . . . . . 288
Theory for Moisture Transport in Porous Media . . . . . . . . . . 289
Theory for Moisture Transport in Building Materials . . . . . . . . 292

Theory for the Heat Transfer Multiphysics Couplings 295

Theory for the Nonisothermal Flow and Conjugate Heat Transfer
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 295
Theory for the Moisture Flow Interface . . . . . . . . . . . . . 303
Theory for the Thermoelectric Effect Interface . . . . . . . . . . 308
Theory for the Building Materials Version of the Heat and Moisture
Transport Interface . . . . . . . . . . . . . . . . . . . 311
Theory for the Moist Air Version of the Heat and Moisture
Transport Interface . . . . . . . . . . . . . . . . . . . 311
Theory for the Moist Porous Media Version of the Heat and
Moisture Transport Interface . . . . . . . . . . . . . . . . 312
Theory for the Heat and Moisture Flow Interfaces . . . . . . . . . 313
Theory for the Electromagnetic Heating Interfaces . . . . . . . . . 313
Theory for the Thermal Stress Interface . . . . . . . . . . . . . 314

Theory for Thermal Contact 315

Theory for the Thermal Contact Feature . . . . . . . . . . . . 315

Out-of-Plane Heat Transfer 320

Equation Formulation . . . . . . . . . . . . . . . . . . . . 320

Convective Heat Transfer Correlations 323

Defining the Heat Transfer Coefficients . . . . . . . . . . . . . 324
Nature of the Flow — The Grashof Number . . . . . . . . . . . 325
Heat Transfer Coefficients — External Natural Convection . . . . . 327
Heat Transfer Coefficients — Internal Natural Convection . . . . . . 334

8 | CONTENTS
 Heat Transfer Coefficients — External Forced Convection . . . . . . 335
Heat Transfer Coefficients — Internal Forced Convection . . . . . . 338
Using the Heat and Mass Transfer Analogy for the Evaluation of
Moisture Transfer Coefficients . . . . . . . . . . . . . . . 339

Nucleate Pool Boiling Correlation 340

Rohsenow’s Correlation for Pool Nucleate Boiling . . . . . . . . . 340
Critical Heat Flux for Nucleate Boiling. . . . . . . . . . . . . . 341

Equivalent Thermal Conductivity Correlations 342

Horizontal Cavity With Bottom Heating . . . . . . . . . . . . . 343
Vertical Cavity With Sidewall Heating . . . . . . . . . . . . . . 343

Temperature Dependence of Surface Tension 345

Heat Flux and Heat Balance 346

Total Heat Flux and Energy Flux . . . . . . . . . . . . . . . . 346
Heat and Energy Balance . . . . . . . . . . . . . . . . . . . 347

Frames for the Heat Transfer Equations 350

Material and Spatial Frames . . . . . . . . . . . . . . . . . . 350
Conversion Between Material and Spatial Frames . . . . . . . . . 351

References 355

Chapter 5: The Heat Transfer Module Interfaces

About the Heat Transfer Interfaces 362

Space Dimensions . . . . . . . . . . . . . . . . . . . . . 362
Study Types . . . . . . . . . . . . . . . . . . . . . . . . 363
Versions of the Heat Transfer Physics Interface . . . . . . . . . . 363
Benefits of the Different Heat Transfer Interfaces . . . . . . . . . 364
Versions of the Heat Transfer in Shells Physics Interface . . . . . . . 364
Benefits of the Different Heat Transfer in Shells Interfaces . . . . . . 365
Layer and Interface Selection Tools . . . . . . . . . . . . . . . 365
Using the Extra Dimension Coordinates . . . . . . . . . . . . . 369

CONTENTS |9
 Versions of the Moisture Transport Physics Interface . . . . . . . . 369
Benefits of the Different Moisture Transport Interfaces . . . . . . . 370
Additional Physics Options . . . . . . . . . . . . . . . . . . 370
Settings for the Heat Transfer Interface . . . . . . . . . . . . . 371
Feature Nodes for the Heat Transfer Interface . . . . . . . . . . 374
Settings for the Heat Transfer in Shells Interface . . . . . . . . . . 377
Settings for the Moisture Transport Interface . . . . . . . . . . . 382
Feature Nodes for the Moisture Transport Interface . . . . . . . . 384

The Heat Transfer in Solids Interface 385

The Heat Transfer in Fluids Interface 386

The Heat Transfer in Solids and Fluids Interface 387

The Heat Transfer in Porous Media Interface 388

The Local Thermal Nonequilibrium Interface 389

The Heat Transfer in Packed Beds Interface 391

The Heat Transfer in Moist Porous Media Interface 393

The Heat Transfer in Building Materials Interface 394

The Heat Transfer in Moist Air Interface 395

The Bioheat Transfer Interface 396

The Heat Transfer in Shells Interface 397

Feature Nodes for the Heat Transfer in Shells Interface . . . . . . . 399

The Heat Transfer in Films Interface 402

Feature Nodes for the Heat Transfer in Films Interface . . . . . . . 404

The Heat Transfer in Fractures Interface 407

Feature Nodes for the Heat Transfer in Fractures Interface . . . . . 409

10 | C O N T E N T S
 The Lumped Thermal System Interface 412
Feature Nodes for the Lumped Thermal System Interface . . . . . . 413

The Surface-to-Surface Radiation Interface 415

Settings for the Surface-to-Surface Radiation Interface. . . . . . . . 415
Feature Nodes for the Surface-to-Surface Radiation Interface . . . . . 422

The Radiation in Participating Media Interface 424

Settings for the Radiation in Participating Media Interface . . . . . . 424
Feature Nodes for the Radiation in Participating Media Interface . . . . 427

The Radiation in Absorbing-Scattering Media Interface 429

Settings for the Radiation in Absorbing-Scattering Media Interface . . . 429
Feature Nodes for the Radiation in Absorbing-Scattering Media
Interface. . . . . . . . . . . . . . . . . . . . . . . . 432

The Radiative Beam in Absorbing Media Interface 435

Settings for the Radiative Beam in Absorbing Media Interface . . . . . 435
Feature Nodes for the Radiative Beam in Absorbing Media Interface . . 436

The Moisture Transport in Air Interface 438

The Moisture Transport in Porous Media Interface 439

The Moisture Transport in Building Materials Interface 440

Chapter 6: The Heat Transfer Features

Domain Features 442

Absorbing Medium (Radiative Beam in Absorbing Medium Interface) . . 444
Absorbing-Scattering Medium (Radiation in Absorbing-Scattering Media
Interface) . . . . . . . . . . . . . . . . . . . . . . . 445
Bioheat . . . . . . . . . . . . . . . . . . . . . . . . . 450
Biological Tissue . . . . . . . . . . . . . . . . . . . . . . 451
Building Material . . . . . . . . . . . . . . . . . . . . . . 454
Convectively Enhanced Conductivity . . . . . . . . . . . . . . 459

CONTENTS | 11
 Cross Section . . . . . . . . . . . . . . . . . . . . . . . 461
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Fluid (Porous Medium) . . . . . . . . . . . . . . . . . . . 467
Geothermal Heating . . . . . . . . . . . . . . . . . . . . 472
Heat Source . . . . . . . . . . . . . . . . . . . . . . . 473
Immobile Fluids (Porous Medium) . . . . . . . . . . . . . . . 477
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 479
Initial Values (Radiative Beam in Absorbing Medium Interface) . . . . . 480
Initial Values (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 481
Irreversible Transformation . . . . . . . . . . . . . . . . . . 483
Isothermal Domain . . . . . . . . . . . . . . . . . . . . . 486
Liquid Water (Moist Porous Medium) . . . . . . . . . . . . . . 488
Moist Air (Heat Transfer Interface) . . . . . . . . . . . . . . . 489
Moist Air (Moist Porous Medium) . . . . . . . . . . . . . . . 492
Moist Porous Medium . . . . . . . . . . . . . . . . . . . . 492
Opacity (Surface-to-Surface Radiation Interface) . . . . . . . . . . 496
Optically Thick Participating Medium . . . . . . . . . . . . . . 497
Out-of-Plane Heat Flux . . . . . . . . . . . . . . . . . . . 499
Out-of-Plane Radiation . . . . . . . . . . . . . . . . . . . 501
Participating Medium (Radiation in Participating Media Interface) . . . . 503
Pellets (Porous Medium) . . . . . . . . . . . . . . . . . . . 509
Pellet-Fluid Interface (Porous Medium) . . . . . . . . . . . . . 511
Phase Change Material. . . . . . . . . . . . . . . . . . . . 512
Porous Matrix (Porous Medium, Moist Porous Medium) . . . . . . . 516
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 520
Pressure Work . . . . . . . . . . . . . . . . . . . . . . 526
Radiative Source . . . . . . . . . . . . . . . . . . . . . . 527
Shape Memory Alloy . . . . . . . . . . . . . . . . . . . . 529
Solid . . . . . . . . . . . . . . . . . . . . . . . . . . 533
Thermal Damage . . . . . . . . . . . . . . . . . . . . . . 537
Thermal Dispersion . . . . . . . . . . . . . . . . . . . . . 541
Thermoelastic Damping . . . . . . . . . . . . . . . . . . . 543
Thickness. . . . . . . . . . . . . . . . . . . . . . . . . 544
Translational Motion . . . . . . . . . . . . . . . . . . . . 545
Viscous Dissipation . . . . . . . . . . . . . . . . . . . . . 546

12 | C O N T E N T S
Boundary Features 548
Boundary Heat Source. . . . . . . . . . . . . . . . . . . . 550
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 553
Continuity (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 554
Continuity on Interior Boundary (Radiation in Participating Media and
Radiation in Absorbing-Scattering Media Interfaces) . . . . . . . 554
Deposited Beam Power . . . . . . . . . . . . . . . . . . . 555
Diffuse Mirror (Surface-to-Surface Radiation Interface) . . . . . . . 557
Diffuse Surface (Surface-to-Surface Radiation Interface) . . . . . . . 558
External Temperature (Thin Layer, Thin Film, Fracture) . . . . . . . 565
Fracture (Heat Transfer Interface) and Porous Medium (Heat Transfer
in Shells Interface) . . . . . . . . . . . . . . . . . . . . 566
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . 571
Heat Flux. . . . . . . . . . . . . . . . . . . . . . . . . 572
Heat Source (Heat Transfer in Shells Interface) . . . . . . . . . . 577
Heat Source (Thin Layer, Thin Film, Fracture) . . . . . . . . . . . 579
Incident Intensity (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 582
Incident Intensity (Radiative Beam in Absorbing Medium Interface) . . . 584
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 586
Initial Values (Thin Layer, Thin Film, Fracture, and Heat Transfer in
Shells Interface) . . . . . . . . . . . . . . . . . . . . . 587
Initial Values (Surface-to-Surface Radiation Interface) . . . . . . . . 590
Isothermal Domain Interface . . . . . . . . . . . . . . . . . 591
Layer Opacity (Surface-to-Surface Radiation Interface) . . . . . . . 595
Line Heat Source on Axis . . . . . . . . . . . . . . . . . . 596
Local Thermal Nonequilibrium Boundary. . . . . . . . . . . . . 597
Lumped System Connector . . . . . . . . . . . . . . . . . . 598
Opaque Surface (Surface-to-Surface Radiation Interface) . . . . . . . 599
Opaque Surface (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 602
Opaque Surface (Radiative Beam in Absorbing Medium Interface) . . . 606
Open Boundary . . . . . . . . . . . . . . . . . . . . . . 607
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 608
Periodic Condition (Heat Transfer Interface) . . . . . . . . . . . 609
Periodic Condition (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 610

CONTENTS | 13
 Phase Change Interface . . . . . . . . . . . . . . . . . . . 611
Phase Change Interface, Exterior . . . . . . . . . . . . . . . . 614
Prescribed Radiosity (Surface-to-Surface Radiation Interface) . . . . . 617
Radiation Group (Surface-to-Surface Radiation Interface) . . . . . . 621
Semitransparent Surface (Radiation in Participating Media and
Radiation in Absorbing-Scattering Media Interfaces) . . . . . . . 623
Semitransparent Surface (Surface-to-Surface Radiation Interface). . . . 628
Surface-to-Ambient Radiation (Heat Transfer Interface) . . . . . . . 632
Symmetry (Heat Transfer Interface). . . . . . . . . . . . . . . 634
Symmetry (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) . . . . . . . . . . . . 635
Temperature . . . . . . . . . . . . . . . . . . . . . . . 635
Thermal Contact . . . . . . . . . . . . . . . . . . . . . . 637
Thermal Insulation . . . . . . . . . . . . . . . . . . . . . 642
Thin Film (Heat Transfer Interface) and Fluid (Heat Transfer in
Shells Interface) . . . . . . . . . . . . . . . . . . . . . 643
Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer in
Shells Interface) . . . . . . . . . . . . . . . . . . . . . 647
Transparent Surface (Radiative Beam in Absorbing Medium Interface) . . 653

Boundary Interface Features 654

Deposited Beam Power, Interface (Heat Transfer in Shells Interface) . . 654
Heat Flux, Interface (Heat Transfer in Shells Interface) . . . . . . . 657
Heat Source, Interface (Heat Transfer in Shells Interface) . . . . . . 660
Lumped System Connector, Interface (Heat Transfer in Shells
Interface) . . . . . . . . . . . . . . . . . . . . . . . 663
Surface-to-Ambient Radiation, Interface (Heat Transfer in Shells
Interface) . . . . . . . . . . . . . . . . . . . . . . . 665
Temperature, Interface (Heat Transfer in Shells Interface) . . . . . . 668
Thermal Contact, Interface (Heat Transfer in Shells Interface). . . . . 670
Thermal Insulation, Interface (Heat Transfer in Shells Interface) . . . . 673

Edge Features 675

Heat Flux (Heat Transfer in Shells Interface) . . . . . . . . . . . 675
Heat Flux (Thin Layer, Thin Film, Fracture) . . . . . . . . . . . . 678
Heat Source (Heat Transfer in Shells Interface) . . . . . . . . . . 681
Line Heat Source . . . . . . . . . . . . . . . . . . . . . . 684
Lumped System Connector (Heat Transfer in Shells Interface) . . . . 686

14 | C O N T E N T S
Shell Continuity (Heat Transfer Interface) and Continuity (Heat
Transfer in Shells Interface) . . . . . . . . . . . . . . . . 688
Surface-to-Ambient Radiation (Thin Layer, Thin Film, Fracture, and
Heat Transfer in Shells Interface) . . . . . . . . . . . . . . 689
Symmetry (Thin Layer, Thin Film, Fracture, and Heat Transfer in
Shells Interface) . . . . . . . . . . . . . . . . . . . . . 692
Temperature (Thin Layer, Thin Film, Fracture, and Heat Transfer in
Shells Interface) . . . . . . . . . . . . . . . . . . . . . 694
Thermal Insulation (Thin Layer, Thin Film, Fracture, and Heat
Transfer in Shells Interface) . . . . . . . . . . . . . . . . 696
Thin Rod . . . . . . . . . . . . . . . . . . . . . . . . . 698

Point Features 700

Point Heat Flux (Thin Rod) . . . . . . . . . . . . . . . . . . 700
Point Heat Source . . . . . . . . . . . . . . . . . . . . . 701
Point Heat Source on Axis . . . . . . . . . . . . . . . . . . 702
Surface-to-Ambient Radiation (Thin Rod) . . . . . . . . . . . . 703
Temperature (Thin Rod) . . . . . . . . . . . . . . . . . . . 705

Global Features 706

Conductive Thermal Resistor . . . . . . . . . . . . . . . . . 706
Convective Thermal Resistor . . . . . . . . . . . . . . . . . 710
External Radiation Source . . . . . . . . . . . . . . . . . . 712
Symmetry for Surface-to-Surface Radiation . . . . . . . . . . . . 718
External Terminal . . . . . . . . . . . . . . . . . . . . . 722
Heat Pipe. . . . . . . . . . . . . . . . . . . . . . . . . 723
Heat Rate . . . . . . . . . . . . . . . . . . . . . . . . 726
Heat Rate Source . . . . . . . . . . . . . . . . . . . . . 729
Nucleate Boiling Heat Rate . . . . . . . . . . . . . . . . . . 730
Radiative Heat Rate . . . . . . . . . . . . . . . . . . . . . 731
Radiative Thermal Resistor . . . . . . . . . . . . . . . . . . 733
Subsystem Definition . . . . . . . . . . . . . . . . . . . . 734
Subsystem Instance . . . . . . . . . . . . . . . . . . . . . 735
Temperature . . . . . . . . . . . . . . . . . . . . . . . 736
Thermal Capacitor . . . . . . . . . . . . . . . . . . . . . 736
Thermal Mass . . . . . . . . . . . . . . . . . . . . . . . 738
Thermoelectric Module . . . . . . . . . . . . . . . . . . . 739

CONTENTS | 15
 Chapter 7: The Moisture Transport Features

Domain Features 746

Building Material . . . . . . . . . . . . . . . . . . . . . . 746
Hygroscopic Porous Medium . . . . . . . . . . . . . . . . . 749
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 752
Liquid Water (Hygroscopic Porous Medium) . . . . . . . . . . . 752
Moist Air (Hygroscopic Porous Medium). . . . . . . . . . . . . 754
Moist Air (Moisture Transport Interface). . . . . . . . . . . . . 755
Moisture Source . . . . . . . . . . . . . . . . . . . . . . 758
Porous Matrix (Hygroscopic Porous Medium) . . . . . . . . . . . 758
Turbulent Mixing . . . . . . . . . . . . . . . . . . . . . . 760

Boundary Features 761

Continuity . . . . . . . . . . . . . . . . . . . . . . . . 761
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Insulation . . . . . . . . . . . . . . . . . . . . . . . . . 764
Moist Surface . . . . . . . . . . . . . . . . . . . . . . . 764
Moisture Content . . . . . . . . . . . . . . . . . . . . . 766
Moisture Flux . . . . . . . . . . . . . . . . . . . . . . . 768
Open Boundary . . . . . . . . . . . . . . . . . . . . . . 770
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 772
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 773
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 774
Thin Moisture Barrier . . . . . . . . . . . . . . . . . . . . 774
Wet Surface . . . . . . . . . . . . . . . . . . . . . . . 776

Global Features 778

Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 778

Chapter 8: Multiphysics Interfaces

The Nonisothermal Flow and Conjugate Heat Transfer

Interfaces 782
Advantages of Using the Multiphysics Interfaces . . . . . . . . . . 782

16 | C O N T E N T S
The Nonisothermal Flow, Laminar Flow and Turbulent Flow
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 783
The Conjugate Heat Transfer, Laminar Flow and Turbulent Flow
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 784
Settings for Physics Interfaces and Coupling Features . . . . . . . . 785
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 786
Physics Interface Features . . . . . . . . . . . . . . . . . . 786
Preset Studies . . . . . . . . . . . . . . . . . . . . . . . 787

The Heat Transfer with Surface-to-Surface Radiation Interface

789
The Heat Transfer with Surface-to-Surface Radiation Multiphysics
Interface. . . . . . . . . . . . . . . . . . . . . . . . 789
Physics Interface Features . . . . . . . . . . . . . . . . . . 791
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 791

The Heat Transfer with Radiation in Participating Media

Interface 792
The Heat Transfer with Radiation in Participating Media Multiphysics
Interface. . . . . . . . . . . . . . . . . . . . . . . . 792
Physics Interface Features . . . . . . . . . . . . . . . . . . 793
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 794

The Heat Transfer with Radiation in Absorbing-Scattering

Media Interface 795
The Heat Transfer with Radiation in Absorbing-Scattering Media
Multiphysics Interface . . . . . . . . . . . . . . . . . . . 795
Physics Interface Features . . . . . . . . . . . . . . . . . . 796
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 797

The Heat Transfer with Radiative Beam in Absorbing Media

Interface 798
The Heat Transfer with Radiative Beam in Absorbing Media
Multiphysics Interface . . . . . . . . . . . . . . . . . . . 798
Physics Interface Features . . . . . . . . . . . . . . . . . . 799
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 799

CONTENTS | 17
 The Thermoelectric Effect Interface 800
About The Thermoelectric Effect Interface . . . . . . . . . . . . 800
Settings for Physics Interfaces and Coupling Features . . . . . . . . 801
Coupling Features . . . . . . . . . . . . . . . . . . . . . 802
Physics Interface Features . . . . . . . . . . . . . . . . . . 803

The Heat and Moisture Transport Interfaces 804

The Heat and Moisture Transport Multiphysics Interfaces . . . . . . 804
The Moist Air Version of the Heat and Moisture Transport
Multiphysics Interface . . . . . . . . . . . . . . . . . . . 805
The Moist Porous Media Version of the Heat and Moisture
Transport Multiphysics Interface. . . . . . . . . . . . . . . 806
The Building Materials Version of the Heat and Moisture Transport
Multiphysics Interface . . . . . . . . . . . . . . . . . . . 808
Physics Interface Features . . . . . . . . . . . . . . . . . . 810
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 810

The Moisture Flow Interfaces 811

The Moisture Flow, Laminar Flow and Turbulent Flow Multiphysics
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 811
Physics Interface Features for the Laminar and Turbulent Flow
Versions . . . . . . . . . . . . . . . . . . . . . . . . 814
The Moisture Flow, Brinkman Equations Multiphysics Interface . . . . 815
Physics Interface Features for the Brinkman Equations Version . . . . 816
Preset Studies . . . . . . . . . . . . . . . . . . . . . . . 816
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 816

The Heat and Moisture Flow Interfaces 817

The Heat and Moisture Flow, Laminar Flow and Turbulent Flow
Multiphysics Interfaces . . . . . . . . . . . . . . . . . . 817
Physics Interface Features for the Laminar and Turbulent Flow
Versions . . . . . . . . . . . . . . . . . . . . . . . . 821
The Heat and Moisture Flow, Brinkman Equations Interface . . . . . 822
Physics Interface Features for the Brinkman Equations Version . . . . 825
Coupling Features . . . . . . . . . . . . . . . . . . . . . 825

The Joule Heating Interface 826

The Joule Heating Interface . . . . . . . . . . . . . . . . . . 826

18 | C O N T E N T S
 Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 826

The Laser Heating Interface 827

The Laser Heating Interface . . . . . . . . . . . . . . . . . . 827
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 827

The Induction Heating Interface 828

The Induction Heating Interface . . . . . . . . . . . . . . . . 828
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 828

The Microwave Heating Interface 829

The Microwave Heating Interface . . . . . . . . . . . . . . . 829
Coupling Feature . . . . . . . . . . . . . . . . . . . . . . 829

Chapter 9: Multiphysics Couplings

Domain Multiphysics Couplings 833

Electromagnetic Heating . . . . . . . . . . . . . . . . . . . 833
Heat and Moisture . . . . . . . . . . . . . . . . . . . . . 834
Heat Transfer with Radiation in Participating Media . . . . . . . . . 837
Heat Transfer with Radiation in Absorbing-Scattering Media . . . . . 839
Heat Transfer with Radiative Beam in Absorbing Media . . . . . . . 840
Moisture Flow . . . . . . . . . . . . . . . . . . . . . . . 841
Nonisothermal Flow . . . . . . . . . . . . . . . . . . . . 845
Thermal Expansion . . . . . . . . . . . . . . . . . . . . . 851
Thermoelectric Effect . . . . . . . . . . . . . . . . . . . . 851

Boundary Multiphysics Couplings 855

Electromagnetic Heating, Layered Shell . . . . . . . . . . . . . 855
Heat Transfer with Surface-to-Surface Radiation . . . . . . . . . . 856
Layered Thermal Expansion . . . . . . . . . . . . . . . . . . 859
Marangoni Effect . . . . . . . . . . . . . . . . . . . . . . 859
Thermoelectric Effect, Layered Shell . . . . . . . . . . . . . . 861

Index 865

CONTENTS | 19
20 | C O N T E N T S
 1

Introduction

This guide describes the Heat Transfer Module, an optional package that extends
the COMSOL Multiphysics® modeling environment with customized physics
interfaces for the analysis of heat transfer.

This chapter introduces you to the capabilities of this module. A summary of the
physics interfaces and where you can find documentation and model examples is
also included. The last section is a brief overview with links to each chapter in this
guide.

• About the Heat Transfer Module

• Overview of the User’s Guide

21
 About the Heat Transfer Module
In this section:

• Why Heat Transfer Is Important to Modeling

• How the Heat Transfer Module Improves Your Modeling
• The Heat Transfer Module Physics Interface Guide
• Common Physics Interface and Feature Settings and Nodes
• The Heat Transfer Module Study Capabilities
• Additional Materials Database
• Where Do I Access the Documentation and Application Libraries?

The Physics Interfaces and Building a COMSOL Multiphysics Model in

the COMSOL Multiphysics Reference Manual.

Why Heat Transfer Is Important to Modeling

The Heat Transfer Module is an optional package that extends the COMSOL
Multiphysics modeling environment with customized physics interfaces and
functionality optimized for the analysis of heat transfer. It is developed for a wide
audience including researchers, developers, teachers, and students. To assist users at all
levels of expertise, this module comes with a library of ready-to-run examples that
appear in the companion Heat Transfer Module Applications Libraries.

Heat transfer is involved in almost every kind of physical process, and can in fact be the
limiting factor for many processes. Therefore, its study is of vital importance, and the
need for powerful heat transfer analysis tools is virtually universal. Furthermore, heat
transfer often appears together with, or as a result of, other physical phenomena.

The modeling of heat transfer effects has become increasingly important in product
design including areas such as electronics, automotive, and medical industries.
Computer simulation has allowed engineers and researchers to optimize process
efficiency and explore new designs, while at the same time reducing the need for costly
experimental trials.

22 | CHAPTER 1: INTRODUCTION
How the Heat Transfer Module Improves Your Modeling
The Heat Transfer Module has been developed to greatly expand upon the base
capabilities available in COMSOL Multiphysics. The module supports all fundamental
mechanisms including conductive, convective, and radiative heat transfer. Using the
physics interfaces in this module along with the inherent multiphysics capabilities of
COMSOL Multiphysics, you can model a temperature field in parallel with other
physics — a versatile combination increasing the accuracy and predicting power of your
models.

This book introduces the basic modeling process. The different physics interfaces are
described and the modeling strategy for various cases is discussed. These sections cover
different combinations of conductive, convective, and radiative heat transfer. This
guide also reviews special modeling techniques for thin layers, thin shells, participating
media, and out-of-plane heat transfer. Throughout the guide the topics and examples
increase in complexity by combining several heat transfer mechanisms and also by
coupling these to physics interfaces describing fluid flow — conjugate heat transfer.

Another source of information is the Heat Transfer Module Applications Libraries, a

set of fully-documented examples that is divided into broadly defined application areas
where heat transfer plays an important role — electronics and power systems,
processing and manufacturing, and medical technology — and includes tutorial and
verification models.

Most of the examples involve multiple heat transfer mechanisms and are often coupled
to other physical phenomena, for example, fluid dynamics, moisture transport, or
electromagnetics. The authors developed several state-of-the art examples by
reproducing examples that have appeared in international scientific journals. See
Where Do I Access the Documentation and Application Libraries?.

The Heat Transfer Module Physics Interface Guide

The table below lists all the physics interfaces specifically available with this module.
Having this module also enhances these COMSOL Multiphysics basic interfaces: Heat
Transfer in Fluids, Heat Transfer in Solids, Joule Heating, and the Single-Phase Flow,
Laminar interface.

ABOUT THE HEAT TRANSFER MODULE | 23

 If you have a Subsurface Flow Module combined with the Heat Transfer Module, this
module also enhances the Heat Transfer in Porous Media interface.

The Nonisothermal Flow, Laminar Flow (nitf) and Nonisothermal Flow,

Turbulent Flow (nitf) interfaces found under the Fluid
Flow>Nonisothermal Flow branch are identical to the Conjugate Heat
Transfer interfaces (Laminar Flow and Turbulent Flow) found under the
Heat Transfer>Conjugate Heat Transfer branch. The difference is that Fluid
is the default domain node for the Nonisothermal Flow interfaces.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• The Physics Interfaces
• For a list of all the core physics interfaces included with a COMSOL
Multiphysics license, see Physics Interface Guide.

PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY

DIMENSION TYPE

Chemical Species Transport

Moisture Transport

Moisture Transport in mt all dimensions stationary; time dependent

Air
Moisture Transport in mt all dimensions stationary; time dependent
Porous Media
Moisture Transport in mt all dimensions stationary; time dependent
Building Materials
Moisture Flow

Laminar Flow(2) — all dimensions stationary; time dependent

Turbulent Flow

24 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Turbulent Flow, — 3D, 2D, 2D stationary with

Algebraic yPlus(2) axisymmetric initialization; transient with
initialization
Turbulent Flow, — 3D, 2D, 2D stationary with
L-VEL(2) axisymmetric initialization; transient with
initialization
Turbulent Flow, k-ε(2) — 3D, 2D, 2D stationary; time dependent
axisymmetric
Turbulent Flow, Low — 3D, 2D, 2D stationary with
Re k-ε(2) axisymmetric initialization; transient with
initialization
Porous Media Flow(2) — all dimensions stationary; time dependent

Fluid Flow

Single-Phase Flow

Laminar Flow(1) spf 3D, 2D, 2D stationary; time dependent

axisymmetric
Turbulent Flow

Turbulent Flow, spf 3D, 2D, 2D stationary with

Algebraic yPlus axisymmetric initialization; transient with
initialization
Turbulent Flow, L-VEL spf 3D, 2D, 2D stationary with
axisymmetric initialization; transient with
initialization
Turbulent Flow, k-ε spf 3D, 2D, 2D stationary; time dependent
axisymmetric
Turbulent Flow, Low spf 3D, 2D, 2D stationary with
Re k-ε axisymmetric initialization; transient with
initialization
Nonisothermal Flow

Laminar Flow(1,2) — 3D, 2D, 2D stationary; time dependent

axisymmetric

ABOUT THE HEAT TRANSFER MODULE | 25

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Turbulent Flow

Turbulent Flow, — 3D, 2D, 2D stationary with

Algebraic yPlus(2) axisymmetric initialization; transient with
initialization
Turbulent Flow, — 3D, 2D, 2D stationary with
L-VEL(2) axisymmetric initialization; transient with
initialization
Turbulent Flow, k-ε(2) — 3D, 2D, 2D stationary; time dependent
axisymmetric
Turbulent Flow, Low — 3D, 2D, 2D stationary with
Re k-ε(2) axisymmetric initialization; transient with
initialization
Heat Transfer

Heat Transfer in Solids(1) ht all dimensions stationary; time dependent;

thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat Transfer in Fluids(1) ht all dimensions stationary; time dependent;
thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat Transfer in Solids and ht all dimensions stationary; time dependent;
Fluids(1) thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Conjugate Heat Transfer

Laminar Flow(1,2) — 3D, 2D, 2D stationary; time dependent;

axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; thermal
perturbation, frequency
domain

26 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Turbulent Flow

Turbulent Flow, — 3D, 2D, 2D stationary with

Algebraic yPlus(2) axisymmetric initialization; transient with
initialization; stationary,
one-way NITF; time
dependent, one-way NITF
Turbulent Flow, — 3D, 2D, 2D stationary with
L-VEL(2) axisymmetric initialization; transient with
initialization; stationary,
one-way NITF; time
dependent, one-way NITF
Turbulent Flow, k-ε(2) — 3D, 2D, 2D stationary; time dependent;
axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; thermal
perturbation, frequency
domain
Turbulent Flow, Low — 3D, 2D, 2D stationary; time dependent;
Re k-ε(2) axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; thermal
perturbation, frequency
domain
Radiation

Heat Transfer with — all dimensions stationary; time dependent;

Surface-to-Surface thermal perturbation,
Radiation(2) frequency domain
Heat Transfer with — 3D, 2D, 2D stationary; time dependent;
Radiation in axisymmetric thermal perturbation,
Participating Media(2) frequency domain
Heat Transfer with — 3D, 2D, 2D stationary; time dependent;
Radiation in axisymmetric thermal perturbation,
Absorbing-Scattering frequency domain
Media(2)
Heat Transfer with — 3D, 2D, 2D stationary; time dependent;
Radiative Beam in axisymmetric thermal perturbation,
Absorbing Media(2) frequency domain

ABOUT THE HEAT TRANSFER MODULE | 27

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Surface-to-Surface rad all dimensions stationary; time dependent

Radiation
Radiation in rpm 3D, 2D, 2D stationary; time dependent
Participating Media axisymmetric
Radiation in rasm 3D, 2D, 2D stationary; time dependent
Absorbing-Scattering axisymmetric
Media
Radiative Beam in rbam 3D, 2D, 2D stationary; time dependent
Absorbing Media axisymmetric
Electromagnetic Heating

Joule Heating(1,2) — all dimensions stationary; time dependent;

thermal perturbation,
frequency domain
Lumped Thermal System lts all dimensions stationary; time dependent

Thin Structures

Heat Transfer in Shells htlsh 3D, 2D, 2D stationary; time dependent;

axisymmetric thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat Transfer in Films htlsh 3D, 2D, 2D stationary; time dependent;
axisymmetric thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat Transfer in htlsh 3D, 2D, 2D stationary; time dependent;
Fractures axisymmetric thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat and Moisture Transport

28 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Moist Air(2) — all dimensions stationary; time dependent;

thermal perturbation,
frequency domain
Moist Porous Media(2) — all dimensions stationary; time dependent;
thermal perturbation,
frequency domain
Building Materials(2) — all dimensions stationary; time dependent;
thermal perturbation,
frequency domain
Heat and Moisture Flow

Laminar Flow(2) — 3D, 2D, 2D stationary; time dependent;

axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; one-way MF; time
dependent, one-way MF;
thermal perturbation,
frequency domain
Turbulent Flow

Turbulent Flow, — 3D, 2D, 2D stationary with

Algebraic yPlus(2) axisymmetric initialization; transient with
initialization; stationary,
one-way NITF; time
dependent, one-way NITF;
one-way MF; time
dependent, one-way MF
Turbulent Flow, — 3D, 2D, 2D stationary with
L-VEL(2) axisymmetric initialization; transient with
initialization; stationary,
one-way NITF; time
dependent, one-way NITF;
one-way MF; time
dependent, one-way MF

ABOUT THE HEAT TRANSFER MODULE | 29

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Turbulent Flow, k-ε(2) — 3D, 2D, 2D stationary; time dependent;

axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; one-way MF; time
dependent, one-way MF;
thermal perturbation,
frequency domain
Turbulent Flow, Low — 3D, 2D, 2D stationary; time dependent;
Re k-ε(2) axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; one-way MF; time
dependent, one-way MF;
thermal perturbation,
frequency domain
Porous Media Flow(2) — 3D, 2D, 2D stationary; time dependent;
axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF; one-way MF; time
dependent, one-way MF;
thermal perturbation,
frequency domain
Porous Media

Heat Transfer in ht all dimensions stationary; time dependent;

Porous Media thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Local Thermal ht all dimensions stationary; time dependent;
Nonequilibrium thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Heat Transfer in ht all dimensions stationary; time dependent;
Packed Beds thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency

30 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Bioheat Transfer ht all dimensions stationary; time dependent;

thermal perturbation,
frequency domain; thermal
perturbation,
eigenfrequency
Thermoelectric Effect(2) — all dimensions stationary; time dependent;
thermal perturbation,
frequency domain
(1)
This physics interface is included with the core COMSOL package but has added
functionality for this module.
(2)
This physics interface is a predefined multiphysics coupling that automatically adds all the
physics interfaces and coupling features required.

Common Physics Interface and Feature Settings and Nodes

There are several common settings and sections available for the physics interfaces and
feature nodes. Some of these sections also have similar settings or are implemented in
the same way no matter the physics interface or feature being used. There are also some
physics feature nodes that display in COMSOL Multiphysics.

In each module’s documentation, only unique or extra information is included;

standard information and procedures are centralized in the COMSOL Multiphysics
Reference Manual.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for

links to common sections and Table 2-5 to common feature nodes.
You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

The Heat Transfer Module Study Capabilities

Table 1-1 lists the Preset Studies available for the physics interfaces most relevant to
this module.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

ABOUT THE HEAT TRANSFER MODULE | 31

 TABLE 1-1: HEAT TRANSFER MODULE DEPENDENT VARIABLES AND PRESET STUDY AVAILABILITY
1
PHYSICS INTERFACE NAME DEPENDENT PRESET STUDIES
VARIABLES

THERMAL PERTURBATION, FREQUENCY

STATIONARY WITH INITIALIZATION

TRANSIENT WITH INITIALIZATION

THERMAL PERTURBATION,
EIGENFREQUENCY
TIME DEPENDENT
STATIONARY

DOMAIN
CHEMICAL SPECIES TRANSPORT>MOISTURE TRANSPORT

Moisture Transport in Air mt φ √ √

Moisture Transport in Porous mt φ √ √
Media
Moisture Transport in Building mt φ √ √
Materials
CHEMICAL SPECIES TRANSPORT>MOISTURE FLOW

Laminar Flow(2) — u, p, φ √ √
Turbulent flow, Algebraic — u, p, G, √ √
yPlus(2) yPlus, φ
Turbulent flow, L-VEL(2) — u, p, G, √ √
uPlus, φ
Turbulent Flow, k-ε(2) — u, p, k, ep, φ √ √
(2)
Turbulent Flow, Low Re k-ε — u, p, k, ep, G, √ √
φ
Porous Media Flow(2) — u, p, φ √ √
FLUID FLOW>SINGLE-PHASE FLOW

Laminar Flow spf u, p √ √

Turbulent flow, Algebraic yPlus spf u, p, G, yPlus √ √
Turbulent flow, L-VEL spf u, p, G, √ √
uPlus
Turbulent Flow, k-ε spf u, p, k, ep √ √

32 | CHAPTER 1: INTRODUCTION
TABLE 1-1: HEAT TRANSFER MODULE DEPENDENT VARIABLES AND PRESET STUDY AVAILABILITY
1
PHYSICS INTERFACE NAME DEPENDENT PRESET STUDIES
VARIABLES

THERMAL PERTURBATION, FREQUENCY

STATIONARY WITH INITIALIZATION

TRANSIENT WITH INITIALIZATION

THERMAL PERTURBATION,
EIGENFREQUENCY
TIME DEPENDENT
STATIONARY

DOMAIN
Turbulent Flow, Low Re k-ε spf u, p, k, ep, G √ √
FLUID FLOW>NONISOTHERMAL FLOW

Laminar Flow(2) — u, p, T √ √
Turbulent flow, Algebraic — u, p, G, √ √
yPlus(2) yPlus, T
Turbulent flow, L-VEL(2) — u, p, G, √ √
uPlus, T
Turbulent Flow, k-ε(2) — u, p, k, ep, T √ √
(2)
Turbulent Flow, Low Re k-ε — u, p, k, ep, G, √ √
T
HEAT TRANSFER

Heat Transfer in Solids ht T √ √ √ √

Heat Transfer in Fluids ht T √ √ √ √
Heat Transfer in Solids and ht T √ √ √ √
Fluids
HEAT TRANSFER>CONJUGATE HEAT TRANSFER

Laminar Flow(2) — u, p, T √ √
Turbulent flow, Algebraic — u, p, G, √ √
yPlus(2) yPlus, T
Turbulent flow, L-VEL(2) — u, p, G, √ √
uPlus, T
Turbulent Flow, k-ε(2) — u, p, k, ep, T √ √
(2)
Turbulent Flow, Low Re k-ε — u, p, k, ep, G, √ √
T

ABOUT THE HEAT TRANSFER MODULE | 33

 TABLE 1-1: HEAT TRANSFER MODULE DEPENDENT VARIABLES AND PRESET STUDY AVAILABILITY
1
PHYSICS INTERFACE NAME DEPENDENT PRESET STUDIES
VARIABLES

THERMAL PERTURBATION, FREQUENCY

STATIONARY WITH INITIALIZATION

TRANSIENT WITH INITIALIZATION

THERMAL PERTURBATION,
EIGENFREQUENCY
TIME DEPENDENT
STATIONARY

DOMAIN
HEAT TRANSFER>RADIATION

Heat Transfer with — T, J √ √ √ √

Surface-to-Surface Radiation(2)
Heat Transfer with Radiation in — T, I (radiative √ √ √ √
Participating Media(2) intensity)
Heat Transfer with Radiation in — T, I (radiative √ √ √ √
Absorbing-Scattering Media(2) intensity)
Heat Transfer with Radiative — T, I (radiative √ √ √ √
Beam in Absorbing Media(2) intensity)
Surface-to-Surface Radiation rad J √ √
Radiation in Participating rpm I (radiative √ √
Media intensity)
Radiation in rasm I (radiative √ √
Absorbing-Scattering Media intensity)
Radiative Beam in Absorbing rbam I (radiative √ √
Media intensity)
HEAT TRANSFER>ELECTROMAGNETIC HEATING

Joule Heating(2) — T, V √ √ √
Lumped Thermal System lts not applicable √ √
HEAT TRANSFER>THIN STRUCURES

Heat Transfer in Shells htlsh T √ √ √ √

Heat Transfer in Films htlsh T √ √ √ √
Heat Transfer in Fractures htlsh T √ √ √ √
HEAT TRANSFER>HEAT AND MOISTURE TRANSPORT

34 | CHAPTER 1: INTRODUCTION
TABLE 1-1: HEAT TRANSFER MODULE DEPENDENT VARIABLES AND PRESET STUDY AVAILABILITY
1
PHYSICS INTERFACE NAME DEPENDENT PRESET STUDIES
VARIABLES

THERMAL PERTURBATION, FREQUENCY

STATIONARY WITH INITIALIZATION

TRANSIENT WITH INITIALIZATION

THERMAL PERTURBATION,
EIGENFREQUENCY
TIME DEPENDENT
STATIONARY

DOMAIN
Moist Air(2) — T, φ √ √
(2)
Moist Porous Media — T, φ √ √
(2)
Building Materials — T, φ √ √
HEAT TRANSFER>HEAT AND MOISTURE TRANSPORT>HEAT AND MOISTURE FLOW

Laminar Flow(2) — u, p, T, φ √ √
Turbulent flow, Algebraic — u, p, G, √ √
yPlus(2) yPlus, T, φ
Turbulent flow, L-VEL(2) — u, p, G, √ √
uPlus, T, φ
Turbulent Flow, k-ε(2) — u, p, k, ep, T, √ √
φ
Turbulent Flow, Low Re k-ε(2) — u, p, k, ep, G, √ √
T, φ
Porous Media Flow(2) — u, p, T, φ √ √
Heat Transfer in Porous Media ht T √ √ √ √
Local Thermal Nonequilibrium ht T √ √ √ √
Heat Transfer in Packed Beds ht T √ √ √ √
Bioheat Transfer ht T √ √ √ √
(2)
Thermoelectric Effect — T √ √ √ √
1
Custom studies are also available based on the physics interface.
2 Multiphysics interfaces.

ABOUT THE HEAT TRANSFER MODULE | 35

 Additional Materials Database
The Heat Transfer Module includes additional material databases:

• The Liquids and Gases material database provides temperature-dependent fluid

dynamic and thermal properties (dynamic viscosity, heat capacity at constant
pressure, density, and thermal conductivity).
• The Bioheat material database contains materials used with the Bioheat Transfer
interface. In addition to thermal properties (heat capacity at constant pressure,
density, and thermal conductivity), properties for thermal damage analysis of
biological tissues are provided.
• The Building material database contains hygroscopic and thermal properties for
materials used with the Moisture Transport interface.

For detailed information, see Liquids and Gases Materials Library,

Bioheat Materials Library, Building Material Library, and Materials in the
COMSOL Multiphysics Reference Manual.

Where Do I Access the Documentation and Application Libraries?

A number of internet resources have more information about COMSOL, including
licensing and technical information. The electronic documentation, topic-based (or
context-based) help, and the application libraries are all accessed through the
COMSOL Desktop.

If you are reading the documentation as a PDF file on your computer,

the blue links do not work to open an application or content
referenced in a different guide. However, if you are using the Help
system in COMSOL Multiphysics, these links work to open other
modules, application examples, and documentation sets.

THE DOCUMENTATION AND ONLINE HELP

The COMSOL Multiphysics Reference Manual describes the core physics interfaces
and functionality included with the COMSOL Multiphysics license. This book also has
instructions about how to use COMSOL Multiphysics and how to access the
electronic Documentation and Help content.

36 | CHAPTER 1: INTRODUCTION
Opening Topic-Based Help
The Help window is useful as it is connected to the features in the COMSOL Desktop.
To learn more about a node in the Model Builder, or a window on the Desktop, click
to highlight a node or window, then press F1 to open the Help window, which then
displays information about that feature (or click a node in the Model Builder followed
by the Help button ( ). This is called topic-based (or context) help.

To open the Help window:

• In the Model Builder, Application Builder, or Physics Builder click a node or

window and then press F1.
• On any toolbar (for example, Home, Definitions, or Geometry), hover the
mouse over a button (for example, Add Physics or Build All) and then
press F1.
• From the File menu, click Help ( ).
• In the upper-right corner of the COMSOL Desktop, click the Help ( )
button.

To open the Help window:

• In the Model Builder or Physics Builder click a node or window and then
press F1.
• On the main toolbar, click the Help ( ) button.
• From the main menu, select Help>Help.

Opening the Documentation Window

To open the Documentation window:

• Press Ctrl+F1.
• From the File menu select Help>Documentation ( ).

ABOUT THE HEAT TRANSFER MODULE | 37

 To open the Documentation window:

• Press Ctrl+F1.
• On the main toolbar, click the Documentation ( ) button.
• From the main menu, select Help>Documentation.

THE APPLICATION LIBRARIES WINDOW

Each model or application includes documentation with the theoretical background
and step-by-step instructions to create a model or application. The models and
applications are available in COMSOL Multiphysics as MPH files that you can open
for further investigation. You can use the step-by-step instructions and the actual
models as templates for your own modeling. In most models, SI units are used to
describe the relevant properties, parameters, and dimensions, but other unit systems
are available.

Once the Application Libraries window is opened, you can search by name or browse
under a module folder name. Click to view a summary of the model or application and
its properties, including options to open it or its associated PDF document.

The Application Libraries Window in the COMSOL Multiphysics

Reference Manual.

Opening the Application Libraries Window

To open the Application Libraries window ( ):

• From the Home toolbar, Windows menu, click ( ) Applications

Libraries.
• From the File menu select Application Libraries.

To include the latest versions of model examples, from the File>Help

menu, select ( ) Update COMSOL Application Library.

Select Application Libraries from the main File> or Windows> menus.

To include the latest versions of model examples, from the Help menu
select ( ) Update COMSOL Application Library.

38 | CHAPTER 1: INTRODUCTION
CONTACTING COMSOL BY EMAIL
For general product information, contact COMSOL at [email protected].

COMSOL ACCESS AND TECHNICAL SUPPORT

To receive technical support from COMSOL for the COMSOL products, please
contact your local COMSOL representative or send your questions to
[email protected]. An automatic notification and a case number are sent to you by
email. You can also access technical support, software updates, license information, and
other resources by registering for a COMSOL Access account.

COMSOL ONLINE RESOURCES

COMSOL website www.comsol.com

Contact COMSOL www.comsol.com/contact
COMSOL Access www.comsol.com/access
Support Center www.comsol.com/support
Product Download www.comsol.com/product-download
Product Updates www.comsol.com/support/updates
COMSOL Blog www.comsol.com/blogs
Discussion Forum www.comsol.com/community
Events www.comsol.com/events
COMSOL Application Gallery www.comsol.com/models
COMSOL Video Gallery www.comsol.com/video
Support Knowledge Base www.comsol.com/support/knowledgebase

ABOUT THE HEAT TRANSFER MODULE | 39

 Overview of the User’s Guide
This Heat Transfer Module User’s Guide gets you started with modeling heat transfer
using COMSOL Multiphysics. The information in this guide is specific to this module.
Instructions on how to use COMSOL in general are included with the COMSOL
Multiphysics Reference Manual.

As detailed in the section Where Do I Access the Documentation and

Application Libraries? this information can also be searched from the
COMSOL Multiphysics Help menu in the COMSOL Desktop.

TABLE OF CONTENTS, NOTATIONS AND INDEX

To help you navigate through this guide, see the Contents, Notations, and Index.

MODELING WITH THE HEAT TRANSFER MODULE

The Modeling with the Heat Transfer Module chapter includes the following topics:

• Heat Transfer Variables

• Moisture Transport Variables
• Using the Boundary Conditions for the Heat Transfer Interfaces
• Handling Frames in Heat Transfer
• Heat Transfer Consistent and Inconsistent Stabilization Methods
• Heat Transfer and Fluid Flow Coupling
• Boundary Wall Temperature
• Solver Settings
• Plotting and Evaluating Results in Layered Materials
• Ambient Properties
• Modeling Heat Transfer by Radiation
• Heat Part Library

THEORY FOR THE HEAT TRANSFER MODULE

The Theory for the Heat Transfer Module chapter includes the theory related to the
heat transfer and moisture transport interfaces and multiphysics interfaces, and also to
some nodes.

40 | CHAPTER 1: INTRODUCTION
After the establishment of the heat balance equation from the energy conservation laws
in Foundations of the General Heat Transfer Equation, the various versions of the heat
equation solved in COMSOL Multiphysics are presented in the following sections:

• Theory for Heat Transfer in Solids

• Theory for Heat Transfer in Fluids
• Theory for Bioheat Transfer
• Theory for Heat Transfer in Porous Media
• Theory for Heat Transfer with Phase Change
• Theory for Heat Transfer in Building Materials
• Theory for Harmonic Heat Transfer
• Theory for Lumped Isothermal Domain
• Theory for Heat Transfer in Thin Structures
• Theory for the Lumped Thermal System Interface
• Theory for Surface-to-Surface Radiation
• Theory for Radiation in Participating Media
• Theory for Moisture Transport

Then the theory related to multiphysics interfaces is described in the section Theory
for the Heat Transfer Multiphysics Couplings.

Finally, topics related to specific features or variables are treated in the sections Theory
for Thermal Contact, Theory for Heat Transfer in Moist Air, Out-of-Plane Heat
Transfer, Convective Heat Transfer Correlations, Equivalent Thermal Conductivity
Correlations, Temperature Dependence of Surface Tension, Heat Flux and Heat
Balance, and Frames for the Heat Transfer Equations.

THE HEAT TRANSFER MODULE INTERFACES

The chapter The Heat Transfer Module Interfaces describes the main Heat Transfer
interface (ht) that forms the backbone for all the fundamental interfaces in this
module, and the other interfaces (Heat Transfer in Shells (htlsh), Radiation in
Participating Media (rpm), Radiation in Absorbing-Scattering Media (rasm), Radiative
Beam in Absorbing Media (rbam), Surface-to-Surface Radiation (rad), and Moisture
Transport (mt)).

The sections The Heat Transfer in Solids Interface, The Heat Transfer in Fluids
Interface, and The Heat Transfer in Solids and Fluids Interface discuss modeling heat
transfer in solids and fluids.

OVERVIEW OF THE USER’S GUIDE | 41

 The sections The Heat Transfer in Porous Media Interface, The Local Thermal
Nonequilibrium Interface, and The Heat Transfer in Moist Porous Media Interface
discuss modeling heat transfer in porous media.

The particular case of heat transfer in moist air and building materials is considered in
the sections The Heat Transfer in Moist Air Interface and The Heat Transfer in
Building Materials Interface.

The section The Bioheat Transfer Interface discusses modeling heat transfer within
biological tissue using the Bioheat Transfer interface.

The sections The Heat Transfer in Shells Interface, The Heat Transfer in Films
Interface, and The Heat Transfer in Fractures Interface describe the physics interfaces
which are suitable for solving thermal conduction, convection, and radiation problems
in layered materials defined on boundaries.

The section The Lumped Thermal System Interface describes the modeling of heat
transfer in a system using a thermal network representation.

The sections The Surface-to-Surface Radiation Interface, The Radiation in

Participating Media Interface, The Radiation in Absorbing-Scattering Media
Interface, and The Radiative Beam in Absorbing Media Interface discuss the modeling
of radiative heat transfer in transparent and participating media.

Finally, the sections The Moisture Transport in Air Interface, The Moisture Transport
in Porous Media Interface, and The Moisture Transport in Building Materials
Interface describe the modeling of moisture transfer in a porous medium through
moisture storage, vapor diffusion and capillary moisture flows; or in air, through
convection and diffusion.

THE HEAT TRANSFER FEATURES

The chapter The Heat Transfer Features describes the Domain Features, Boundary
Features, Edge Features, Point Features, and Global Features available with the Heat
Transfer interfaces.

THE MOISTURE TRANSPORT FEATURES

The Moisture Transport Features chapter describes the Domain Features and
Boundary Features available with the Moisture Transport interface.

THE HEAT TRANSFER MULTIPHYSICS INTERFACES

The Multiphysics Interfaces chapter describes the predefined multiphysics interfaces.

42 | CHAPTER 1: INTRODUCTION
The chapter The Nonisothermal Flow and Conjugate Heat Transfer Interfaces
describes the multiphysics versions of both the Nonisothermal Flow Laminar Flow and
Turbulent Flow interfaces found under the Fluid Flow branch, which are identical to
the Conjugate Heat Transfer interfaces. Each section describes the applicable physics
interfaces in detail and concludes with the underlying theory.

The section The Heat Transfer with Surface-to-Surface Radiation Interface describes
the predefined multiphysics interface used to model heat transfer by conduction,
convection, and radiation in a transparent media.

The section The Heat Transfer with Radiation in Participating Media Interface
describes the predefined multiphysics interface used to model heat transfer by
conduction, convection, and radiation in semitransparent media. The radiative
intensity equations are approximated by the Discrete Ordinates Method or the P1
Approximation. When no emission should be considered, see the section The Heat
Transfer with Radiation in Absorbing-Scattering Media Interface.

The section The Heat Transfer with Radiative Beam in Absorbing Media Interface
describes the predefined multiphysics interface used to model heat transfer by
conduction, convection, and radiation in semitransparent media. The Beer-Lambert
law is used for the approximation of the radiative intensity.

The section The Thermoelectric Effect Interface describes the predefined multiphysics
interface used to model the Peltier-Seebeck-Thomson effect.

The section The Heat and Moisture Transport Interfaces describes the predefined
multiphysics interfaces used to model coupled heat and moisture transport either in
building materials, by taking into account heat and moisture storage, latent heat
effects, and liquid and convective transport of moisture; or in moist air by convection
and diffusion of moisture and heat.

The section The Moisture Flow Interfaces describes the predefined multiphysics
interfaces used to model moisture transport in air by laminar flow (in free and porous
media) and turbulent flows (in free media).

The section The Heat and Moisture Flow Interfaces describes the predefined
multiphysics interfaces used to model heat transfer and moisture transport in air by
laminar and turbulent flows, and in porous media by laminar flows.

OVERVIEW OF THE USER’S GUIDE | 43

 THE HEAT TRANSFER MULTIPHYSICS COUPLINGS
The Multiphysics Couplings chapter describes the Domain Multiphysics Couplings
and the Boundary Multiphysics Couplings available with the predefined multiphysics
interfaces.

44 | CHAPTER 1: INTRODUCTION
 2

Notations

This chapter introduces the notations used in the remaining of the guide. The
notations are listed in alphabetical order and grouped in two tables for Latin and
Greek symbols, respectively.

For each entry the SI unit and a short description are given.

45
 Symbols
LATIN SYMBOLS

NOTATION SI UNIT DESCRIPTION

A 1/s Frequency factor, damage integral analysis

2
A m Total boundaries area
a1,…, a12 dimensionless Legendre coefficients
Ac m2 Cross sectional area of domain
2
Al m Cross sectional area of thin rod
Bi m Spectral band i
b dimensionless Thermal conductivity supplement
3
c mol/m Concentration
3
c kg/m Concentration
2
cl mol/m Liquid water concentration on moist surface
Cp, b J/(kg·K) Specific heat capacity at constant pressure, blood
Cp, l J/(kg·K) Specific heat capacity at constant pressure of thin rod
Cμ dimensionless Turbulence modeling constant
Cp J/(kg·K) Specific heat capacity at constant pressure
Cp, a J/(kg·K) Specific heat capacity at constant pressure, dry air
Cp, A J/(kg·K) Specific heat capacity at constant pressure, Austenite state
Cp, d J/(kg·K) Specific heat capacity at constant pressure, damaged tissue
Cp, f J/(kg·K) Specific heat capacity at constant pressure, fluid phase
Cp, imf J/(kg·K) Specific heat capacity at constant pressure of immobile fluid
in porous media
Cp, imfi J/(kg·K) Specific heat capacity at constant pressure of immobile fluid i
in porous media
Cp, m J/(kg·K) Specific heat capacity at constant pressure, mixture (moist
air)
Cp, M J/(kg·K) Specific heat capacity at constant pressure, Martensite state
Cp, s J/(kg·K) Specific heat capacity at constant pressure, solid phase
Cp, si J/(kg·K) Specific heat capacity at constant pressure of solid material i
in porous media
Cp, v J/(kg·K) Specific heat capacity at constant pressure, water vapor

46 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

Cp, s J/(kg·K) Specific heat capacity at constant pressure of thin layer

Cp, si J/(kg·K) Specific heat capacity at constant pressure of layer i in thin
layer
csat mol/m3 Saturation concentration of water vapor
custr mol/m3 Upstream concentration of water vapor
Cp, w J/(kg·K) Specific heat capacity at constant pressure of water
3
cv mol/m Water vapor concentration
d m Average particle diameter
D m Cylinder diameter (heat transfer coefficient)
D m Cylinder or sphere diameter (heat transfer coefficient)
D W/(m·K) Dispersion tensor
D m Parallel-plate gap average gas particle diameter
D 1/s Strain-rate tensor
2
D m /s Vapor diffusion coefficient in air
2
Deff m /s Vapor diffusion coefficient in moist porous media
Da dimensionless Darcy number
ΔE J/mol Activation energy, damage integral analysis
df m Thickness of thin film
dfr m Thickness of fracture
dpe m Average pellets diameter
DP1 m2/s P1 method diffusion coefficient
ds m Thickness of shell or thin layer
dsi m Thickness of layer i in thin layer
2
DT m /s Turbulent diffusivity
Dw m2/s Moisture diffusivity
dz m Thickness of domain in the out-of-plane direction
DPTamb K Ambient dew point temperature
e m Beam orientation
E J/kg Internal energy
E Pa Young’s modulus
E0 J/kg Total internal energy
EΩ J Internal energy of a body

SYMBOLS | 47
 NOTATION SI UNIT DESCRIPTION

eb, λ(λ, T) W/m2 Blackbody spectral emissive power

eb(T) W/m2 Blackbody total emissive power
Econtact Pa Effective contact interface Young’s modulus
Ed Pa Young’s modulus, down contact surface
Ek J/kg Kinetic energy
ep m2/s3 Turbulent dissipation rate
Ep J/kg Potential energy
Eu Pa Young’s modulus, up contact surface
etot W/m2 Total energy flux
F N/m3 Body force vector
F dimensionless Deformation gradient tensor
Famb dimensionless Ambient view factor
Famb, d dimensionless Ambient view factor, downside
Famb, u dimensionless Ambient view factor, upside
g m/s2 Acceleration of gravity
Gevap kg/m3 Evaporation source
gevap kg/(m2·s) Evaporation flux
G W/m3 Moisture source
G 1/m Reciprocal wall distance
G W/m2 Surface irradiation
Gamb W/m2 Ambient irradiation
Gamb, d W/m2 Ambient irradiation, downside
Gamb, u W/m2 Ambient irradiation, upside
Gd W/m2 Surface irradiation, downside
Gext W/m2 External irradiation
Gext, d W/m2 External irradiation, downside
2
Gext, u W/m External irradiation, upside
2
Gm W/m Mutual surface irradiation
Gm, d W/m2 Mutual surface irradiation, downside
2
Gm, u W/m Mutual surface irradiation, upside
GrL dimensionless Grashof number associated with characteristic length L
2
Gu W/m Surface irradiation, upside

48 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

H m Chimney height (heat transfer coefficient)

H J/kg Enthalpy
h W/(m2·K) Gap conductance (thermal contact)
h W/(m2·K) Heat transfer coefficient
H0 J/kg Total enthalpy
HB Pa Brinell hardness
hc W/(m2·K) Constriction conductance
Hc Pa Microhardness
hd W/(m2·K) Out-of-plane heat transfer coefficient, downside
Hd J/kg Enthalpy, downside
Hext J/kg External enthalpy
hg W/(m2·K) Parallel-plate gap gas conductance
hm m/s Moisture transfer coefficient
hr W/(m2·K) Radiative conductance
Href J/kg Reference enthalpy
hsf W/(m2·K) Interstitial heat transfer coefficient
hth W/K Thermal conductance (isothermal domain interface)
hu W/(m2·K) Out-of-plane heat transfer coefficient, upside
Hu J/kg Enthalpy, upside
hz W/(m2·K) Out-of-plane heat transfer coefficient, 1D
I(Ω) W/(m2·sr) Radiative intensity traveling in direction Ω
Ib(T) W/(m2·sr) Blackbody radiative intensity
Idiff W/m2 Diffuse irradiance
Iext W/(m2·sr) Incident radiative intensity
Ii W/(m2·sr) Radiative intensity traveling in ith discrete direction
is dimensionless Incident radiation direction (external radiation source)
2
Is W/m Solar irradiance
2
Is,amb W/m Ambient solar irradiance
Ish,amb W/m2 Clear sky noon diffuse horizontal irradiance
2
Isn,amb W/m Clear sky noon beam normal irradiance
isx, isy, dimensionless Solar source direction vector components
isz

SYMBOLS | 49
 NOTATION SI UNIT DESCRIPTION

Iwall W/(m2·sr) Boundary radiative intensity

J W/m2 Surface radiosity
Jd W/m2 Surface radiosity, downside
Ju W/m2 Surface radiosity, upside
k W/(m·K) Thermal conductivity
k J/kg Turbulent kinetic energy (turbulent nonisothermal flow)
ka W/(m·K) Dry air thermal conductivity
kA W/(m·K) Thermal conductivity of Austenite state
kB J/K Stefan-Boltzmann constant
kbnd W/(m·K) Thermal conductivity in shell local coordinate system
kcontact W/(m·K) Harmonic mean of contacting surface conductivities
kd W/(m·K) Thermal conductivity, damaged tissue
kdisp W/(m·K) Dispersive thermal conductivity tensor
keff W/(m·K) Effective thermal conductivity
kf W/(m·K) Thermal conductivity, fluid phase
kimf W/(m·K) Thermal conductivity of immobile fluid in porous media
kgap W/(m·K) Parallel-plate gap gas thermal conductivity
kimfi W/(m·K) Thermal conductivity of immobile fluid i in porous media
kl W/(m·K) Thermal conductivity of thin rod
kL W/(m·K) Thermal conductivity of mobile fluid in porous media
kM W/(m·K) Thermal conductivity of Martensite state
kR W/(m·K) Rosseland radiative conductivity
ks W/(m·K) Thermal conductivity, solid phase
ksi W/(m·K) Thermal conductivity of solid material i in porous media and
fractures
ks W/(m·K) Thermal conductivity of thin layer
ksi W/(m·K) Thermal conductivity of layer i in thin layer
kT W/(m·K) Turbulent thermal conductivity
kv W/(m·K) Water vapor thermal conductivity
K m/s Evaporation rate factor
KΩ J Kinetic energy of a body
L J/kg Latent heat

50 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

L m Total edge length

L m Wall height or plate diameter, distance or length (heat
transfer coefficient)
Ld, c J/kg Damage enthalpy change, cryogenic analysis
Ld, h J/kg Damage enthalpy change, hyperthermia analysis
Lit, c J/kg Transformation enthalpy change, overcooling analysis
Lit, h J/kg Transformation enthalpy change, overheating analysis
Lv J/kg Latent heat of evaporation
m kg Mass
ma kg Dry air mass
Ma kg/mol Dry air molar mass
masp m Asperities average slope (surface roughness)
Mg m Parallel-plate gap gas parameter
Mn kg/mol Mean molar mass
mtot kg Total mass (moist air)
mv kg Water vapor mass
Mv kg/mol Water vapor molar mass
n dimensionless Refractive index, transparent media
n dimensionless Normal vector toward exterior
na mol Amount of dry air
nr dimensionless Refractive index, participating media
ntot mol Amount of moist air
Nu dimensionless Nusselt number
NuL dimensionless Nusselt number associated with characteristic length L
nv mol Amount of water vapor
O m Beam origin point
p Pa Contact pressure, pressure
2
P N/m First Piola-Kirchhoff stress tensor
P V Peltier coefficient
P0 W Heat rate
pa Pa Dry air partial pressure
pA Pa Absolute pressure

SYMBOLS | 51
 NOTATION SI UNIT DESCRIPTION

Pb W Heat rate, boundary heat source

Pc m Cross sectional perimeter of domain
pext Pa External absolute pressure
Pext W Power of applied forces
pgap Pa Parallel-plate gap gas pressure
Pindex dimensionless Performance index of the discrete ordinates method
Pl W Heat rate, line heat source, deposited beam power
Pr dimensionless Prandtl number
pamb Pa Ambient absolute pressure
pref Pa Reference pressure
PrT dimensionless Turbulent Prandtl number
Ps W Heat rate, layer heat source
Ps W Source power (external radiation source)
psat Pa Saturation pressure of water vapor
Pstr W Stress power
pustr Pa Upstream absolute pressure
pv Pa Water vapor partial pressure
q W/m2 Conductive heat flux
qf W/m2 Conductive heat flux in fluid phase
qs W/m2 Conductive heat flux in solid phase
Q W/m3 Heat source
q0 W/m2 Inward heat flux
Q0 W/m3 Distributed heat source
q0, d W/m2 Out-of-plane heat flux, downside
q0, s W/m2 Source heat flux (external radiation source)
q0, u W/m2 Out-of-plane heat flux, upside
2
Qb W/m Boundary heat source
2
Qb, tot W/m Total boundary heat source
Qe W/m3 Electromagnetic heat source
3
Qevap W/m Latent heat source in domains
2
qevap W/m Latent heat source on surfaces
Qexch W Exchanged heat source rate

52 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

Qf W/m3 Heat source in fluid phase

qgeo W/kg Radiogenic heating per mass
Qgeo W/m3 Geothermal heat source
QInt W Total heat source over interior boundaries
Qit W/m3 Irreversible transformation heat source
Ql W/m Line heat source
Qmet W/m3 Metabolic heat source
Qp W Point heat source
Qp W/m3 Pressure work
Qr W/m3 Radiative heat source term
qr W/m2 Radiative heat flux
qr, net W/m2 Net radiative heat flux
qr, out W/m2 Radiative heat flux striking the wall
qs W/(m3·K) Production/absorption coefficient
Qs W/m3 Heat source in solid phase
Qs W/m3 Thin layer heat source
qsf W/(m3·K) Interstitial convective heat transfer coefficient
Qsi W/m3 Layer i heat source
Qted W/m3 Thermoelastic damping
qtot W/m2 Total heat flux
Qtot W/m3 Total domain heat source
Qvd W/m3 Viscous dissipation
r m Distance of the irradiated surface from the source
r dimensionless Heat partition coefficient (thermal friction)
R m Heat source radius, beam radius
R J/(mol·K) Universal gas constant
Ra dimensionless Rayleigh number
RaD dimensionless Rayleigh number associated with cylinder diameter D
RaL dimensionless Rayleigh number associated with characteristic length L
Reinf dimensionless Reynolds number at infinity
ReL dimensionless Reynolds number associated with characteristic length L
Rep dimensionless Particle Reynolds number

SYMBOLS | 53
 NOTATION SI UNIT DESCRIPTION

rh m Hydraulic radius
rl m Rod radius
rpe m Average pellets radius
Rs J/(kg·K) Specific gas constant
Rt,s K·m2/W Surface thermal resistance
Rt K/W Thermal resistance
Sb 1/m Specific surface area
sd m Vapor diffusion equivalent air layer thickness
S N/m2 Second Piola-Kirchhoff stress tensor
S V/K Seebeck coefficient
ScT dimensionless Turbulent Schmidt number
Si dimensionless Unit vector of discrete direction in space, i-th component
(angular space discretization)
Sp dimensionless Sparrow number
T K Temperature
T0 K Equilibrium temperature
T’ K Complex amplitude of harmonic perturbation
+
T dimensionless Dimensionless temperature
Tamb K Ambient temperature
Tamb, d K Ambient temperature, downside
Tamb, u K Ambient temperature, upside
Tb K Arterial blood temperature
Td K Temperature, downside
td, c s Damage time, cryogenic analysis
Td, c K Damage temperature, cryogenic analysis
td, h s Damage time, hyperthermia analysis
Td, h K Damage temperature, hyperthermia analysis
Text K External temperature
Text, d K Out-of-plane external temperature, downside
Text, u K Out-of-plane external temperature, upside
Text, z K Out-of-plane external temperature, 1D
Tf K Temperature, fluid phase

54 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

tit, c Transformation time, overcooling analysis

Tit, c Transformation temperature, overcooling analysis
tit, h Transformation time, overheating analysis
Tit, h Transformation temperature, overheating analysis
Tn, c K Temperature of necrosis, cryogenic
Tn, h K Temperature of necrosis, hyperthermia
Tpc K Phase change temperature
Tref K Reference temperature
Tref K Strain reference temperature
Ts K Temperature, solid phase
Tu K Temperature, upside
Tustr K Upstream temperature
Tw K Wall temperature
u m/s Fluid velocity vector
u, v, w m/s Fluid velocity vector’s components
uf m/s Average linear velocity
up m/s Porous velocity field
utrans m/s Translational motion velocity vector
vamb m/s Wind velocity
V V Electric potential
V m3 Total domain volume
W W/m3 Work source
w kg/m3 Moisture storage function
WInt W Work from custom volume forces
Wdiss W Dissipative work from momentum equation
Xa dimensionless Molar fraction of dry air
xpl m Position along the plate (heat transfer coefficient)
xs m Source location (external radiation source)
Xv dimensionless Molar fraction of water vapor
xvap dimensionless Moisture content
xvap,amb dimensionless Ambient moisture content
Y m Parallel-plate gap mean separation thickness

SYMBOLS | 55
 GREEK SYMBOLS

NOTATION SI UNIT DESCRIPTION

(ρCp)eff J/(m3·K) Effective volumetric heat capacity at constant pressure

∂Ω - Geometry domain’s boundaries
∂Ωext - Geometry domain’s exterior boundaries
∂Ωint - Geometry domain’s interior boundaries
∇t - Tangential gradient operator
α 1/K Coefficient of thermal expansion tensor in a solid
α dimensionless Degree of tissue injury (Arrhenius equation)
α dimensionless Irreversible transformation indicator (Arrhenius equation)
α dimensionless Parallel-plate gap gas thermal accommodation parameter
α dimensionless Surface absorptivity
α m2/s Thermal diffusivity
α dimensionless Damaged tissue indicator
αnecr dimensionless Necrosis time indicator
αm dimensionless Vapor mass fraction
αp 1/K Coefficient of thermal expansion in a fluid
β 1/m Extinction coefficient
βp s/m Moisture transfer coefficient
β dimensionless Parallel-plate gap gas property parameter
βR 1/m Rosseland mean extinction coefficient
γ dimensionless Ratio of specific heats
γTeq Pa/K Psychrometer constant
ΔR m Size of transition zone (deposited beam power)
ΔT K Temperature offset for periodic condition
ΔT0 K Temperature perturbation
δ s Vapor permeability of still air
δp s Vapor permeability
δw m Distance between the computational fluid domain and the
wall
εp dimensionless Porosity
ε dimensionless Surface emissivity
εd dimensionless Surface emissivity, downside

56 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

ελ dimensionless Surface spectral emissivity

εu dimensionless Surface emissivity, upside
εz dimensionless Surface emissivity, 1D out-of-plane radiation
θ rad Angle between the normal to the irradiated surface and the
direction of the source
θd dimensionless Volume fraction of necrotic tissue
θimf dimensionless Volume fraction of immobile fluid in porous media
θimfi dimensionless Volume fraction of immobile fluid i in porous media
θit dimensionless Fraction of transformation
θg dimensionless Volume fraction of mobile fluid in porous media
θs dimensionless Volume fraction of solid material in porous media and
fractures
θsi dimensionless Volume fraction of solid material i in porous media
θs rad Zenith angle of the Sun
κ 1/m Absorption coefficient (radiation)
2
κ m Permeability (porous media)
λ W/(m·K) Thermal conductivity (turbulent nonisothermal flow)
λ m Wavelength
Λ m Parallel-plate gap gas mean free path
λi m Wavelength band i endpoint
λijkl m Fourth-order dispersivity tensor’s component
λlo m Longitudinal dispersivity
λtr m Transverse dispersivity
λtr, h m Transverse horizontal dispersivity
λtr, v m Transverse vertical dispersivity
μ Pa·s Dynamic viscosity
μ dimensionless Vapor resistance factor
μ0 dimensionless Scattering angle
μa Pa·s Dry air dynamic viscosity
μf Pa·s Dynamic viscosity
μm Pa·s Mixture dynamic viscosity (moist air)
μT Pa·s Turbulent dynamic viscosity
μTh V/K Thomson coefficient

SYMBOLS | 57
 NOTATION SI UNIT DESCRIPTION

μv Pa·s Water vapor dynamic viscosity

ν dimensionless Poisson ratio
νd dimensionless Poisson ratio, down contact surface
νu dimensionless Poisson ratio, up contact surface
ρ kg/m3 Density
ρ dimensionless Surface reflectivity
ρb kg/m3 Density, blood
ρd kg/m3 Density, damaged tissue
ρf kg/m3 Density, fluid phase
ρimf kg/m3 Density of immobile fluid in porous media
ρgeo kg/m3 Geothermal density
ρimfi kg/m3 Density of immobile fluid i in porous media
ρl kg/m3 Density of thin rod
ρg kg/m3 Mixture density (moist air)
ρs kg/m3 Density of solid material
ρsi kg/m3 Density of solid material i in porous media
ρs kg/m3 Density of thin layer
ρsi kg/m3 Density of layer i in thin layer
σ Pa Cauchy stress tensor
σ m Standard deviation (deposited beam power)
σ W/(m2·K4) Stefan-Boltzmann constant
σ N/m Surface tension coefficient
σasp m Asperities average height (surface roughness)
σs 1/m Scattering coefficient
τ dimensionless Optical thickness
τ Pa Viscous stress tensor
φw dimensionless Relative humidity
φ amb dimensionless Ambient relative humidity
φ rad Tilt angle (heat transfer coefficient)
φ w, ext dimensionless External relative humidity
φ w, d dimensionless Relative humidity, downside
φ w, s dimensionless Surface relative humidity

58 | CHAPTER 2: NOTATIONS
NOTATION SI UNIT DESCRIPTION

φ w, u dimensionless Relative humidity, upside

φ w, ustr dimensionless Upstream relative humidity
Φd → u kg/s Mass flow rate, positive direction
Φu → d kg/s Mass flow rate, negative direction
ϕ ( Ω', Ω ) dimensionless Scattering phase function
ϕs rad Azimuth angle of the Sun
ϕd → u kg/(m2·s) Mass flux, positive direction
ϕu → d kg/(m2·s) Mass flux, negative direction
Ψ J/kg Force potential
Ω dimensionless Unit vector of a direction in space
Ω - Geometry domain
ω dimensionless Specific humidity
ω rad/s Angular frequency
ωb 1/s Blood perfusion rate
ωi W/(m2·K) Discrete incident radiation vector, i-th component
ξ kg/m3 Moisture storage capacity
ξ dimensionless Martensite state volume fraction

SYMBOLS | 59
60 | CHAPTER 2: NOTATIONS
 3

Modeling with the Heat Transfer Module

A variety of modeling techniques are discussed in the following sections:

• Heat Transfer Variables

• Moisture Transport Variables
• Using the Boundary Conditions for the Heat Transfer Interfaces
• Handling Frames in Heat Transfer
• Heat Transfer Consistent and Inconsistent Stabilization Methods
• Heat Transfer and Fluid Flow Coupling
• Boundary Wall Temperature
• Solver Settings
• Plotting and Evaluating Results in Layered Materials
• Evaluating the Bulk Temperature
• Ambient Properties
• Modeling Heat Transfer by Radiation
• Specifying the Temperature at a Flow Inlet
• Heat Part Library

61
 Heat Transfer Variables
In this section:

• Predefined Variables
• Global Variables for Heat and Energy Balance
• Domain Fluxes
• Out-of-Plane Domain Fluxes
• Boundary Fluxes (Heat Transfer Interface)
• Internal Boundary Heat Fluxes
• Domain Heat Sources
• Boundary Heat Sources
• Line and Point Heat Sources
• Moist Air Variables

Predefined Variables
This section lists some predefined variables that are available to evaluate heat fluxes,
sources, and integral quantities used in heat and energy balance. All the variable names
begin with the physics interface name (the prefix). By default the Heat Transfer
interface prefix is ht, and the Heat Transfer in Shells interface prefix is htlsh. As an
example, you can access the variable named tflux using ht.tflux (as long as the
physics interface is named ht).
TABLE 3-1: PREDEFINED HEAT BALANCE VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

Global variables
dEiInt Total Accumulated Heat Rate Global
ntfluxInt Total Net Heat Rate Global
WstrInt Total Stress Power Global
heatBalance Heat Balance Global
Domain variables
tflux Total Heat Flux (Heat Transfer Domains
Interface)

62 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TABLE 3-1: PREDEFINED HEAT BALANCE VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

dflux Conductive Heat Flux (Heat Domains

Transfer Interface)
cflux Convective Heat Flux (Heat Domains
Transfer Interface)
turbflux Turbulent Heat Flux (Heat Transfer Domains
Interface)
tflux_u Total Out-of-Plane Heat Flux (Heat Out-of-plane domains (1D
tflux_d Transfer Interface) and 2D), boundaries
tflux_z
rflux_u Radiative Out-of-Plane Heat Flux Out-of-plane domains (1D
rflux_d (Heat Transfer Interface) and 2D), boundaries
rflux_z
q0_u Out-of-Plane Heat Flux (Heat Out-of-plane domains (1D
q0_d Transfer Interface) and 2D)
q0_z
Boundary variables
ntflux Normal Total Heat Flux (Heat Boundaries
Transfer Interface)
ndflux Normal Conductive Heat Flux Boundaries
(Heat Transfer Interface)
ncflux Normal Convective Heat Flux (Heat Boundaries
Transfer Interface)
rflux Radiative Heat Flux (Heat Transfer Boundaries
Interface)
q0 Inward Heat Flux (Heat Transfer Boundaries
Interface)
ntflux_u Internal normal total heat flux, upside Interior boundaries
ntflux_d and downside (Internal Boundary
Heat Fluxes)
ndflux_u Internal normal conductive heat flux, Interior boundaries
ndflux_d upside and downside (Internal
Boundary Heat Fluxes)
ncflux_u Internal normal convective heat flux, Interior boundaries
ncflux_d upside and downside (Internal
Boundary Heat Fluxes)

HEAT TRANSFER VARIABLES | 63

 TABLE 3-2: PREDEFINED ENERGY BALANCE VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

Global variables
dEi0Int Total Accumulated Energy Rate Global
ntefluxInt Total Net Energy Rate Global
WInt Total Work Source Global
energyBalan Energy Balance Global
ce
Domain variables
teflux Total Energy Flux (Heat Transfer Domains
Interface)
dflux Conductive Heat Flux (Heat Domains
Transfer Interface)
thflux Total Enthalpy Flux (Heat Transfer Domains
Interface)
Boundary variables
nteflux Normal Total Energy Flux (Heat Boundaries
Transfer Interface)
ndflux Normal Conductive Heat Flux Boundaries
(Heat Transfer Interface)
nthflux Normal Total Enthalpy Flux (Heat Boundaries
Transfer Interface)
q0 Inward Heat Flux (Heat Transfer Boundaries
Interface)
nteflux_u Internal normal total energy flux, Interior boundaries
nteflux_d upside and downside (Internal
Boundary Energy Fluxes)
ndflux_u Internal normal conductive heat flux, Interior boundaries
ndflux_d upside and downside (Internal
Boundary Heat Fluxes)
nthflux_u Internal normal enthalpy flux, upside Interior boundaries
nthflux_d and downside (Internal Boundary
Energy Fluxes)
TABLE 3-3: PREDEFINED HEAT SOURCE VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

Global variables
QInt Total Heat Source Global
Domain variables

64 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TABLE 3-3: PREDEFINED HEAT SOURCE VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

Qtot Total heat source (Domain Heat Domains

Sources)
Boundary variables
Qbtot Total boundary heat source Boundaries
(Boundary Heat Sources)
Qltot Total line heat source (Line and Point Edges, Points (2D, 2D axial
Heat Sources) symmetry)
Qptot Point heat source (Line and Point Points
Heat Sources)

TABLE 3-4: OTHER PREDEFINED HEAT TRANSFER VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

Global variables
Tref Reference temperature Global
Domain variables
alpha Damage indicator (see Theory for Domains
Bioheat Transfer)
alphanecr Instant tissue necrosis indicator (see Domains
Theory for Bioheat Transfer)
theta_d Fraction of damage (see Theory for Domains
Bioheat Transfer)
theta_d_sm Fraction of total damage (see Theory Domains
for Bioheat Transfer)
T_dp Dew Point Temperature Domains
T_eq Equivalent Temperature Domains
psat Saturation Pressure Domains
phi Relative Humidity Domains
Lv Latent Heat of Evaporation Domains
Boundary variables
h Heat transfer coefficient (see Exterior boundaries
Convective Heat Transfer
Correlations)

HEAT TRANSFER VARIABLES | 65

 Some of these variables are only available with the Heat Transfer Module
(rflux_u, rflux_d, rflux_z, q0_u, q0_d, and q0_z), or when either the
CFD Module or the Heat Transfer Module is added (rflux and
turbflux).

Global Variables for Heat and Energy Balance

This section describes variables defined by integrals. A concise notation denotes the
different domains of integration: Ω is the geometry domain, ∂Ωext stands for the
exterior boundaries, and ∂Ωint for the interior boundaries.

TOTAL ACCUMULATED HEAT RATE

The total accumulated heat rate variable, dEiInt, is the variation of internal energy per
unit time in the domain:

d
dEiInt =
dt Ω ρEi dω
TOTAL NET HEAT RATE
The total net heat rate, ntfluxInt, is the integral of Total Heat Flux (Heat Transfer
Interface) over all external boundaries. In the case of a fluid domain, it reads:

ntfluxInt = ∂Ω ext

( ρuE i – k∇T + q r ) ⋅ n dσ

This indicates the sum of incoming and outgoing total heat flux through the system.

TOTAL HEAT SOURCE

The total heat source, QInt, accounts for all domain sources, interior boundary, edge
and point sources, and radiative sources at interior boundaries:

QInt = Ω Q dω + ∂Ω int

Q b dω + ∂ Ω int
Q r dω

TOTAL STRESS POWER

The total stress power, WstrInt, corresponds to the work lost by a fluid by
degradation of energy. These works are transmitted to the system through pressure
work and viscous dissipation:

66 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 WstrInt = Ω pA ∇ ⋅ u dω + Ω ( –τ : ∇u ) dω
TOTAL ACCUMULATED ENERGY RATE
The total accumulated energy rate, dEi0Int, is the variation of total internal energy
per unit time in the domain:

d
dEi0Int =
dt Ω ρEi0 dω
where the total internal energy, Ei0, is defined as

u⋅u
E i0 = E i + ------------
2

TOTAL NET ENERGY RATE

The total net energy rate, ntefluxInt, is the integral of Total Energy Flux (Heat
Transfer Interface) over all external boundaries. In the case of a fluid domain, it reads:

ntefluxInt = ∂Ω ext

( ρuH 0 – k∇T – τ u + q r ) ⋅ n dσ

This indicates the sum of incoming and outgoing total energy flux through the system.

TOTAL WORK SOURCE

The total work source, WInt, accounts for volume forces and boundary stresses as
follows:

WInt = Ω ( ρg + F ) ⋅ u dω + ∂Ω ext

K ⋅ u dσ + ∂Ω int
K⋅u ∂Ω int
dσ

where K is the viscous force vector defined as

T 2
K = μ ( ∇u + ( ∇u ) ) – --- μ ( ∇ ⋅ u )I ⋅ n
3

and K ⋅ u ∂Ω int
is the jump of viscous stress at interior boundaries.

HEAT BALANCE
According to Equation 4-200, the following equality between COMSOL Multiphysics
variables holds:

dEiInt + ntfluxInt = QInt - WstrInt

HEAT TRANSFER VARIABLES | 67

 and the heat balance variable, heatBalance, is defined as

heatBalance = dEiInt + ntfluxInt + WstrInt - QInt

This is the most general form that can be used for time-dependent models. At
steady-state the formula is simplified. The accumulated heat rate equals zero, so the
total net heat rate (the sum of incoming and outgoing heat rates) should correspond
to the heat and work sources:

ntfluxInt = QInt - WstrInt

The sign convention used in COMSOL Multiphysics for QInt is positive when energy
is produced (as for a heater) and negative when energy is consumed (as for a cooler).
For WstrInt, the losses that heat up the system are negative and the gains that cool
down the system are positive.

For stationary models with convection by an incompressible flow, the heat balance
becomes:

ntfluxInt = QInt

which corresponds to the conservation of convective and conductive flux as in:

∂Ω ext
ρuE i ⋅ n dσ – ∂Ω ext
k∇T ⋅ n dσ = Q Int

Depending on the radiation discretization method, the contribution to

the heat balance is handled differently. For the definition of ntfluxInt,
the Optically Thick Participating Medium subnode uses the Rosseland
approximation and defines qr, the radiative flux, as an extra contribution
to the conductive heat flux. The P1 approximation and Discrete ordinates
method, however, include the radiative source ∇ ⋅ qr to Q on the domain,
in the variable QInt.

The heat balance can be evaluated assuming constant material properties

using the following variables: heatBalance_cst, dEiInt_cst,
ntfluxInt_cst, WstrInt_cst, QInt_cst. The evaluation of these
variables is faster and accurate for constant material properties or small
temperature gradient.

68 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

ENERGY BALANCE
According to Equation 4-201, the following equality between COMSOL Multiphysics
predefined variables holds:

dEi0Int + ntefluxInt = QInt + WInt

and the energy balance variable, energyBalance, is defined as

energyBalance = dEi0Int + ntefluxInt - WInt - QInt

In stationary models, dEi0Int is zero so the energy balance simplifies into:

ntefluxInt = QInt + WInt

At steady state, and without any additional heat source (QInt equal to zero), the
integral of the net energy flux on all boundaries of the flow domain, ntefluxInt,
vanishes. On the other hand, the corresponding integral of the net heat flux does not,
in general, vanish. It corresponds instead to the losses from mass and momentum
equations, such as WstrInt for pressure work and viscous dissipation in fluids. Hence,
energy is the conserved quantity, not heat.

The heat and energy balance variables such as ht.ntfluxInt or

ht.ntefluxInt do not take into account the flux contributions on thin
structures. For example, if a Thin Layer node is applied on a boundary, the
heat flux on this boundary is not included in ht.ntfluxInt. It is added
instead to the variable ht.sls1.ntfluxInt, defined locally on the thin
structure. Here, sls1 is the tag of the Thin Layer node.

The energy balance can be evaluated assuming constant material

properties using the following variables: energyBalance_cst,
dEi0Int_cst, ntefluxInt_cst, WInt_cst, QInt_cst. The evaluation
of these variables is faster and accurate for constant material properties or
small temperature gradient.

Domain Fluxes
On domains the fluxes are vector quantities. The definition can vary depending on the
active physics nodes and selected properties.

HEAT TRANSFER VARIABLES | 69

 TOTAL HEAT FLUX (HEAT TRANSFER INTERFACE)
In domains, the total heat flux, tflux, is the sum of the conductive and convective
heat fluxes. For accuracy reasons the radiative heat flux is not included.

See Radiative Heat Flux (Heat Transfer Interface) to evaluate the radiative
heat flux.

For solid domains — for example, the solid and biological tissue domains — the total
heat flux is defined as:

tflux = dflux

For fluid domains (for example, Fluid), the total heat flux is defined as:

tflux = cflux + dflux

CONDUCTIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

The conductive heat flux variable, dflux, is evaluated using the temperature gradient
and the effective thermal conductivity:

dflux = – k eff ∇T

In the general case keff is the thermal conductivity, k.

For heat transfer in fluids with turbulent flow, keff = k + kT, where kT is the turbulent
thermal conductivity.

For heat transfer in porous media, keff is the effective conductivity computed from the
solid and fluid conductivities.

For heat transfer in building materials, a latent heat source due to evaporation is
included in the conductive heat flux variable:

dflux = – ( k eff ∇T + L v δ p ∇( φ w p sat ) )

When considering thermoelectric effect, the thermoelectric heat flux is also included
in the diffusive flux, see Equation 4-150.

TURBULENT HEAT FLUX (HEAT TRANSFER INTERFACE)

The turbulent heat flux variable, turbflux, enables access to the part of the
conductive heat flux that is due to turbulence.

70 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 turbflux = – k T ∇T

CONVECTIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

The convective heat flux variable, cflux, is defined using the internal energy, Ei:

cflux = ρuE i

The internal energy, Ei, is defined as:

• Ei = H for solid domains

• Ei = H − p/ρ for fluid domains

where H is the enthalpy defined in Equation 4-5.

The convective heat flux may be oriented in the opposite direction of the velocity
field’s direction, when the internal energy, Ei, has a negative value. This happens when
the sensible enthalpy (variation from reference enthalpy Href) is negative. Href is set to
0 J/kg at pref (1 atm), cref (0 mol/m3), and Tref (293.15 K) in COMSOL
Multiphysics. See Thermodynamic Description of Heat Transfer for details.

RADIATIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

In participating media, the radiative heat flux, qr, is not available for analysis on
domains because it is more accurate to evaluate the radiative heat source Q r = ∇ ⋅ q r .

TOTAL ENTHALPY FLUX (HEAT TRANSFER INTERFACE)

The total enthalpy flux variable, thflux, is defined as:

thflux = ρuH 0

where the total enthalpy, H0, is defined as

u⋅u
H 0 = H + ------------
2

TOTAL ENERGY FLUX (HEAT TRANSFER INTERFACE)

The total energy flux, teflux, is defined when viscous dissipation is enabled:

teflux = thflux + dflux + τu

HEAT TRANSFER VARIABLES | 71

 Out-of-Plane Domain Fluxes

RADIATIVE OUT-OF-PLANE HEAT FLUX (HEAT TRANSFER INTERFACE)

The radiative out-of-plane heat flux, rflux, is generated by the Out-of-Plane
Radiation feature.

• In 2D:
4 4
upside: rflux_u = ε u σ ( T amb, u – T )

4 4
downside: rflux_d = ε d σ ( T amb, d – T )

• In 1D:

4 4
rflux_z = ε z σ ( T amb, z – T )

OUT-OF-PLANE HEAT FLUX (HEAT TRANSFER INTERFACE)

The convective out-of-plane heat flux, q0, is generated by the Out-of-Plane Heat Flux
feature.

• In 2D:
upside: q0_u = h u ( T ext, u – T )

downside: q0_d = h d ( T ext, d – T )

• In 1D:

q0_z = h z ( T ext, z – T )

TOTAL OUT-OF-PLANE HEAT FLUX (HEAT TRANSFER INTERFACE)

The total out-of-plane heat flux, tflux, is the sum of the radiative and convective heat
fluxes:

• In 2D:
upside: tflux_u = q0_u + rflux_u

downside: tflux_d = q0_d + rflux_d

• In 1D:

tflux_z = q0_z + rflux_z

72 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Boundary Fluxes (Heat Transfer Interface)
All the domain fluxes (vector quantity) computed by the Heat Transfer interface are
also available as boundary fluxes. The boundary fluxes are then equal to the mean value
of the fluxes on adjacent domains. In addition, normal boundary heat fluxes (scalar
quantity) are available on boundaries.

NORMAL TOTAL HEAT FLUX (HEAT TRANSFER INTERFACE)

The variable ntflux is defined as:

ntflux = ndflux + ncflux

NORMAL CONDUCTIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

The variable ndflux is defined on exterior boundaries as:

• ndflux = −dflux_spatial(T) if the adjacent domain is on the downside,

• ndflux = −uflux_spatial(T) if the adjacent domain is on the upside,

and, on interior boundaries, as:

ndflux = (uflux_spatial(T) − dflux_spatial(T))/2

Frames for the Heat Transfer Equations for a description of spatial and
material frames.

NORMAL CONVECTIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

The variable ncflux is defined as:

ncflux = mean ( cflux ) ⋅ n

NORMAL TOTAL ENERGY FLUX (HEAT TRANSFER INTERFACE)

The variable nteflux is defined as:

nteflux = mean ( teflux ) ⋅ n – mean ( dflux ) ⋅ n + ndflux

NORMAL TOTAL ENTHALPY FLUX (HEAT TRANSFER INTERFACE)

The variable nthflux is defined as:

nthflux = mean ( thflux ) ⋅ n

HEAT TRANSFER VARIABLES | 73

 INWARD HEAT FLUX (HEAT TRANSFER INTERFACE)
On external boundaries, the inward heat flux, q0, is a scalar quantity defined either in
a general way or as a convective heat flux as follows:

q0 = h ( T ext – T )

where Text is the external temperature defined in the Heat Flux feature.

RADIATIVE HEAT FLUX (HEAT TRANSFER INTERFACE)

On boundaries, the radiative heat flux, rflux, is a scalar quantity defined as:

rflux = ε ( G – e b ( T ) ) + q r, net

where the terms account for surface-to-ambient radiative flux, surface-to-surface

radiative flux, and radiation in participating net radiative flux, respectively.

Internal Boundary Heat Fluxes

The internal normal boundary heat fluxes (scalar quantity) are available on interior
boundaries. They are calculated using the upside and the downside value of heat fluxes
from the adjacent domains.

INTERNAL NORMAL CONDUCTIVE HEAT FLUX, UPSIDE (HEAT TRANSFER

INTERFACE)
The variable ndflux_u is defined as:

ndflux_u = uflux_spatial ( T )

INTERNAL NORMAL CONDUCTIVE HEAT FLUX, DOWNSIDE (HEAT

TRANSFER INTERFACE)
The variable ndflux_d is defined as:

ndflux_d = dflux_spatial ( T )

Frames for the Heat Transfer Equations for a description of spatial and
material frames.

INTERNAL NORMAL CONVECTIVE HEAT FLUX, UPSIDE (HEAT TRANSFER

INTERFACE)
The variable ncflux_u is defined as:

74 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 ncflux_u = up ( cflux ) ⋅ un

INTERNAL NORMAL CONVECTIVE HEAT FLUX, DOWNSIDE (HEAT

TRANSFER INTERFACE)
The variable ncflux_d is defined as:

ncflux_d = down ( cflux ) ⋅ dn

INTERNAL NORMAL TOTAL HEAT FLUX, UPSIDE (HEAT TRANSFER

INTERFACE)
The variable ntflux_u is defined as:

ntflux_u = ndflux_u + ncflux_u

INTERNAL NORMAL TOTAL HEAT FLUX, DOWNSIDE (HEAT TRANSFER

INTERFACE)
The variable ntflux_d is defined as:

ntflux_d = ndflux_d + ncflux_d

Internal Boundary Energy Fluxes

INTERNAL NORMAL TOTAL ENERGY FLUX, UPSIDE (HEAT TRANSFER

INTERFACE)
The variable nteflux_u is defined as:

nteflux_u = up ( teflux ) ⋅ un – up ( dflux ) ⋅ un + ndflux_u

INTERNAL NORMAL TOTAL ENERGY FLUX, DOWNSIDE (HEAT TRANSFER

INTERFACE)
The variable nteflux_d is defined as:

nteflux_d = down ( teflux ) ⋅ dn – down ( dflux ) ⋅ dn + ndflux_d

INTERNAL NORMAL ENTHALPY FLUX, UPSIDE (HEAT TRANSFER

INTERFACE)
The variable nthflux_u is defined as:

nthflux_u = up ( thflux ) ⋅ un

HEAT TRANSFER VARIABLES | 75

 INTERNAL NORMAL ENTHALPY FLUX, DOWNSIDE (HEAT TRANSFER
INTERFACE)
The variable nthflux_d is defined as:

nthflux_d = down ( thflux ) ⋅ dn

Domain Heat Sources

The sum of the domain heat sources added by different physics features is available in
the variable Qtot, which is the sum of:

• Q’s, which are the heat sources added by the Heat Source (described for the Heat
Transfer interface) and Electromagnetic Heating (described for the Joule Heating
interface in the COMSOL Multiphysics Reference Manual) features.
• Qmet, which is the metabolic heat source added by the Bioheat feature.
• Qdmg, which is the cooling source added by the Irreversible Transformation feature.
• Qr, which is the radiative heat source added by the Heat Transfer with Radiation in
Participating Media, Heat Transfer with Radiation in Absorbing-Scattering Media,
and Heat Transfer with Radiative Beam in Absorbing Media multiphysics features.
• Qgeo, which is the geothermal heat source added by the Geothermal Heating
feature.
• Qevap, which is the latent heat source added by the Heat and Moisture coupling
node.

Boundary Heat Sources

The sum of the boundary heat sources added by different boundary conditions is
available in the variable, Qb,tot (SI unit: W/m2). This variable Qbtot is the sum of:

• Qb, which is the boundary heat source added by the Boundary Heat Source
boundary condition.
• Qsh, which is the boundary heat source added by the Electromagnetic Heating
condition (described for the Joule Heating interface in the COMSOL Multiphysics
Reference Manual).
• Qs, which is the boundary heat source added by a Layer Heat Source subfeature of
a thin layer, see Heat Source (Thin Layer, Thin Film, Fracture).

76 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Line and Point Heat Sources
The sum of the line heat sources is available in a variable called Qltot (SI unit: W/m).

The sum of the point heat sources is available in a variable called Qptot (SI unit: W).

Moist Air Variables

The temperature variable solved by the Heat Transfer interfaces corresponds to the dry
bulb temperature. This is the temperature measured by a thermometer with a dry
sensor and screening to prevent from deviation due to external radiation like solar
radiation.

When the presence of water vapor is accounted for in the model, other temperatures
may be considered, depending on vapor pressure.

DEW POINT TEMPERATURE

The dew point temperature of a sample of air with water vapor pressure pv is the
temperature to which it must be cooled to become fully saturated.

The variable T_dp is defined in Ref. 1 by:

p sat ( T_dp ) = p v = phi ⋅ psat

where phi is the Relative Humidity variable. See Saturation State for the definition of
saturation pressure psat as a function of temperature. See also Saturation Pressure for
the definition of the variable psat.

EQUIVALENT TEMPERATURE
The equivalent temperature is obtained by adiabatically condensing all the water vapor
of a sample of air with initial vapor pressure pv. In this process, the latent heat decrease
due to total removal of the vapor is balanced by a increase of the sensible heat and
temperature.

The variable T_eq is approximated in Ref. 1 by:

phi ⋅ psat
T_eq = T + -----------------------------
γ Teq

where phi is the Relative Humidity, and γTeq (SI unit: Pa/K) is the psychrometer
constant, defined in Ref. 1 by:

HEAT TRANSFER VARIABLES | 77

 M a pC p, a
γ Teq = ------------------------
Mv Lv

where p is the total pressure, Cp, a is the heat capacity at constant pressure of dry air at
temperature T, Lv is the latent heat of evaporation at temperature T (see Latent Heat
of Evaporation), and Ma and Mv are the molar mass of dry air and water vapor,
respectively.

See also Saturation Pressure for the definition of the variable psat.

These definitions are illustrated on Figure 3-1.

Figure 3-1: Relation between dry bulb, dew point, equivalent, and wet bulb temperatures.

The wet bulb temperature is also represented on Figure 3-1. It is obtained by

adiabatically reaching saturation state for a sample of air with initial water vapor
pressure pv. In this process, the latent heat increase due to evaporation is balanced by
a decrease of the sensible heat and temperature. It is not available as a predefined
variable, but it can be approximated by solving the following equation:

p sat ( T wb ) phi ⋅ psat

T wb + --------------------------- = T + -----------------------------
γ γ

The psychrometer constant is again evaluated at temperature T.

SATURATION PRESSURE
The variable psat is defined by:

psat = fpsat ( T )

where T is the temperature. See Functions for the definition of the function fpsat.

RELATIVE HUMIDITY
The variable phi is defined by:

78 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 pv
phi = --------------------
p sat ( T )

LATENT HEAT OF EVAPORATION

The variable Lv is defined by:

Lv = lv ( T )

where T is the temperature. See Functions for the definition of the function Lv.

HEAT TRANSFER VARIABLES | 79

 Moisture Transport Variables
Predefined Variables
This section lists the predefined variables that are available with the Moisture Transport
interface. All variable names begin with the physics interface name (the prefix). By
default, the Moisture Transport interface prefix is mt. As an example, you can access
the variable named phi using mt.phi (as long as the physics interface is named mt).
TABLE 3-5: MOISTURE TRANSPORT PREDEFINED VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

dwcInt Total accumulated moisture rate Global

ntfluxInt Total net moisture rate Global
GInt Total moisture source Global
massBalance Mass balance Global
phi Relative humidity Domains
cv Vapor concentration Domains
cl Liquid water concentration Domains
cw Total moisture concentration Domains
ca Dry air concentration Domains
csat Saturation concentration Domains
psat Saturation pressure of water vapor Domains
pv Partial pressure of water vapor Domains
wc_l Liquid water content Domains
wc_v Vapor content Domains
wcVar Moisture content Domains
omega_v Vapor mass fraction Domains
omega_moist Moisture mass fraction Domains
xvap Humidity ratio Domains
Lv Latent heat of evaporation Domains
Xa Molar fraction of dry air Domains
Xv Molar fraction of water vapor Domains
rhoa Density of dry air Domains
rhov Density of water vapor Domains

80 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TABLE 3-5: MOISTURE TRANSPORT PREDEFINED VARIABLES

VARIABLE NAME GEOMETRIC ENTITY LEVEL

rho_moist Density of moist air Domains

mua Viscosity of dry air Domains
muv Viscosity of water vapor Domains
mu_moist Viscosity of moist air Domains
cflux Convective moisture flux Domains
dflux Diffusive moisture flux Domains
tflux Total moisture flux Domains
tfluxMag Total moisture flux magnitude Domains
Gtot Total moisture source Domains
dwc Total accumulated moisture rate Domains
density
h_m Moisture transfer coefficient Boundaries
ncflux Normal convective moisture flux Boundaries
ndflux Normal diffusive moisture flux Boundaries
ntflux Normal total moisture flux Boundaries
g_evap Evaporation flux Boundaries
q_evap Latent heat source Boundaries
q_evaptot Total latent heat source on surfaces Boundaries
gtot Total moisture flux Boundaries
uStefan_u Stefan velocity, upside Boundaries
uStefan_d Stefan velocity, downside Boundaries
ncflux_u Normal convective moisture flux, Interior boundaries
upside
ncflux_d Normal convective moisture flux, Interior boundaries
downside
ndflux_u Normal diffusive moisture flux, upside Interior boundaries
ndflux_d Normal diffusive moisture flux, Interior boundaries
downside
ntflux_u Normal total moisture flux, upside Interior boundaries
ntflux_d Normal total moisture flux, downside Interior boundaries

MOISTURE TRANSPORT VARIABLES | 81

 Moist Air Properties

PARTIAL AND SATURATION PRESSURES OF WATER VAPOR

The partial pressure of vapor pv is the product of the saturation pressure psat and the
relative humidity phi:

pv = phi ⋅ psat

SATURATION CONCENTRATION OF WATER VAPOR

The saturation concentration csat is obtained from the saturation pressure psat and
temperature T by:

psat
csat = ------------
RT

where R is the universal gas constant.

HUMIDITY RATIO
The humidity ratio xvap is obtained from the vapor and air concentrations cv and ca,
the absolute pressure pA and temperature T by:

cv Mv
x vap = ----- ⋅ --------
ca Ma

where R is the universal gas constant, Mv is the water vapor molar mass, and Ma is the
dry air molar mass.

The vapor and dry air concentrations satisfy:

pA
c a = -------- – c v
RT

VAPOR MASS FRACTION

The vapor mass fraction omega_v is obtained by:

cv Mv
ω v = ----------------------------------
ca Ma + cv Mv

LIQUID, VAPOR, AND MOISTURE CONCENTRATIONS

The total moisture concentration cw is the sum of the liquid water and vapor
concentrations cl and cv:

82 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 cw = cv + cl

Note that cl=0 in moist air domains.

CONCENTRATIONS AND CONTENTS

The liquid water and vapor contents wc_l and wc_v are obtained by multiplying the
concentrations cl and cv by the molar mass of water M_v:

wc _ l = M _ v ⋅ cl

wc _ v = M _ v ⋅ cv

MOISTURE CONTENT
The moisture content wcVar is expressed in different ways depending on the domain
type:

• In moist air:
w ( φw ) = ρg ωv

• In hygroscopic porous media:

w ( φw ) = εp ( ρl sl + ρg ωv ( 1 – sl ) )

• In building materials, it is given by the moisture storage function, which may be

given as a material property.

LATENT HEAT OF EVAPORATION

The variable Lv is defined by:

Lv = lv ( T )

where T is the temperature. See Functions for the definition of the function Lv.

Domain Moisture Fluxes

CONVECTIVE MOISTURE FLUX

The definition of the variable cflux depends on the domain model and on the species
formulation:

• In moist air domains, with the diluted species formulation:

cflux = M v u g c v

MOISTURE TRANSPORT VARIABLES | 83

 • In moist air domains, with the concentrated species formulation:
cflux = ρ g u g ω v

• In hygroscopic porous media domains:

cflux = ρ g u g ω v + ρ l u l

• In building materials domains:

cflux = – δ p φ w ∇ ( p sat ( T ) )

In building materials domains, the definition of cflux does not correspond to

convection in the physical sense, but rather to vapor diffusion. However, this term is
numerically handled like a convection flux, and is therefore included in the variable
cflux.

DIFFUSIVE MOISTURE FLUX

The definition of the variable dflux depends on the domain model and on the species
formulation:

• In moist air domains, with the diluted species formulation:

dflux = – M v D∇c v

• In moist air domains, with the concentrated species formulation:

dflux = – ρ g D∇ω v

• In hygroscopic porous media domains:

dflux = g lc – ρ g D eff ∇ω v

• In building materials domains:

dflux = – ξ D w ∇φ w – δ p p sat ( T )∇φ w

In hygroscopic porous media and building materials domains, the first term of the
definition of dflux corresponds to liquid capillary flux. This term is numerically
handled like a diffusion flux, and is therefore included in the variable dflux.

TOTAL MOISTURE FLUX

The variable tflux is defined as:

tflux = cflux+dflux

84 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TOTAL MOISTURE FLUX MAGNITUDE
The variable tfluxMag is defined as:

tfluxMag = norm ( tflux )

Boundary Moisture Fluxes

NORMAL CONVECTIVE MOISTURE FLUX

The variable ncflux is defined on all boundaries as:

ncflux = mean ( cflux ) ⋅ n

INTERNAL NORMAL CONVECTIVE MOISTURE FLUX, UPSIDE

The variable ncflux_u is defined on interior boundaries as:

ncflux_u = up ( cflux ) ⋅ un

INTERNAL NORMAL CONVECTIVE MOISTURE FLUX, DOWNSIDE

The variable ncflux_d is defined on interior boundaries as:

ncflux_d = down ( cflux ) ⋅ dn

NORMAL DIFFUSIVE MOISTURE FLUX

The variable ndflux is defined on all boundaries as:

ndflux = mean ( dflux ) ⋅ n

INTERNAL NORMAL DIFFUSIVE MOISTURE FLUX, UPSIDE

The variable ndflux_u is defined on interior boundaries as:

ndflux_u = up ( dflux ) ⋅ un

INTERNAL NORMAL DIFFUSIVE MOISTURE FLUX, DOWNSIDE

The variable ndflux_d is defined on interior boundaries as:

ndflux_d = down ( dflux ) ⋅ dn

NORMAL TOTAL MOISTURE FLUX

The variable ntflux is defined on all boundaries as:

ntflux = mean ( tflux ) ⋅ n

MOISTURE TRANSPORT VARIABLES | 85

 INTERNAL NORMAL TOTAL MOISTURE FLUX, UPSIDE
The variable ntflux_u is defined on interior boundaries as:

ntflux_u = up ( tflux ) ⋅ un

INTERNAL NORMAL TOTAL MOISTURE FLUX, DOWNSIDE

The variable ntflux_d is defined on interior boundaries as:

ntflux_d = down ( tflux ) ⋅ dn

TOTAL MOISTURE FLUX

The sum of the boundary moisture fluxes added by Moisture Flux features is available
as the variable gtot.

Domain Moisture Source

The sum of the domain moisture sources added by Moisture Source features is
available as the variable Gtot.

Mass Balance
The mass balance variable, massBalance, is defined as

massBalance = dwcInt + ntfluxInt - GInt

where:

• the variation of total moisture content per unit time in the domain, dwcInt, is:
d
dwcInt =
dt  Ω w ( φ w ) dω
• the total net moisture rate, ntfluxInt, is the integral of the total moisture flux over
all external boundaries:

ntfluxInt =
∂Ω ext
tflux ⋅ n dσ

• the total moisture source, GInt, is the integral of the moisture sources over the
domains:

GInt =
Ω Gtot dω

86 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Using the Boundary Conditions for
the Heat Transfer Interfaces
In this section:

• Temperature and Heat Flux Boundary Conditions

• Overriding Mechanism for Heat Transfer Boundary Conditions

Temperature and Heat Flux Boundary Conditions

The heat equation accepts two basic types of boundary conditions: specified
temperature and specified heat flux. The specified condition is of constraint type and
prescribes the temperature on a boundary:

T = T0 on ∂Ω

while the latter specifies the inward heat flux

–n ⋅ q = q0 on ∂Ω

where

• q is the conductive heat flux vector (SI unit: W/m2), q = −k∇T.

• n is the normal vector on the boundary.
• q0 is the inward heat flux (SI unit: W/m2), normal to the boundary.

The inward heat flux, q0, is often a sum of contributions from different heat transfer
processes (for example, radiation and convection). The special case q0 = 0 is called
thermal insulation.

A common type of heat flux boundary conditions is one for which q0 = h·(Text − T),
where Text is the temperature far away from the modeled domain and the heat transfer
coefficient, h, represents all the physics occurring between the boundary and “far
away.” It can include almost anything, but the most common situation is that h

USING THE BOUNDARY CONDITIONS FOR THE HEAT TRANSFER INTERFACES | 87

 represents the effect of an exterior fluid cooling or heating the surface of a solid, a
phenomenon often referred to as convective cooling or heating.

The CFD Module and the Heat Transfer Module contain a set of
correlations for convective heat flux and heating. See Heat Transfer and
Fluid Flow Coupling.

Overriding Mechanism for Heat Transfer Boundary Conditions

Many boundary conditions are available in heat transfer. Some of these can coexist (for
example, Heat Flux and Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer
in Shells Interface)); others cannot (for example, Heat Flux and Thermal Insulation).

Several categories of boundary condition exist in heat transfer. Table 3-6 gives the
overriding rules for these groups.

1 Temperature, Open Boundary, Inflow, Outflow, Phase Change Interface

2 Thermal Insulation, Symmetry (Heat Transfer Interface), Periodic Condition (Heat
Transfer Interface)
3 Heat Flux
4 Boundary Heat Source, Line Heat Source on Axis, Deposited Beam Power
5 Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer in Shells Interface),
Thermal Contact, Thin Film (Heat Transfer Interface) and Fluid (Heat Transfer in
Shells Interface), and Fracture (Heat Transfer Interface) and Porous Medium (Heat
Transfer in Shells Interface)
6 Isothermal Domain Interface

TABLE 3-6: OVERRIDING RULES FOR HEAT TRANSFER BOUNDARY CONDITIONS

A\B 1 2 3 4 5 6

1-Temperature X X X
2-Thermal Insulation X X
3-Heat Flux X X
4-Boundary heat source
5-Thin Layer X X
6-Isothermal Domain Interface X

88 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

When there is a boundary condition A above a boundary condition B in the model
tree and both conditions apply to the same boundary, use Table 3-6 to determine if A
is overridden by B or not:

• Locate the line that corresponds to the A group (see above the definition of the
groups). In the table above only the first member of the group is displayed.
• Locate the column that corresponds to the group of B.
• If the corresponding cell is empty, A and B contribute. If it contains an X, B
overrides A.

Group 3 and group 4 boundary conditions are always contributing. That

means that they never override any other boundary condition. But they
might be overridden.

Example
Consider a boundary where Heat Flux is applied. Then a Symmetry boundary condition
is applied on the same boundary afterward.

• Heat Flux belongs to group 3.

• Symmetry belongs to group 2.
• The cell on the line of group 3 and the column of group 2 contains an X so Heat
Flux is overridden by Symmetry.
This mechanism can be checked on the COMSOL Desktop, in the Override and
Contribution section of each feature, as shown in the following table:

TABLE 3-7: Override and Contribution SECTIONS.

HEAT FLUX SYMMETRY

USING THE BOUNDARY CONDITIONS FOR THE HEAT TRANSFER INTERFACES | 89

 In the example above, if Symmetry followed by Heat Flux is added, the
boundary conditions contribute.

90 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Handling Frames in Heat Transfer
In this section:

• Heat Transfer Analysis with Moving Frames

• Material Density in Features Defined in the Material Frame

Heat Transfer Analysis with Moving Frames

This section discusses heat transfer analysis with moving frames, when spatial and
material frames do not coincide. In this case, the heat transfer interfaces account for
deformation effects on heat transfer properties.

• Material and Spatial Frames

• About Frames in the COMSOL Multiphysics Reference Manual

In the heat transfer interfaces, the entire physics (equations and variables) are defined
in the spatial frame. When a moving mesh is detected, the user inputs for certain
features are defined in the material frame and are converted so that all the
corresponding variables contain the value in the spatial frame.

Conversion Between Material and Spatial Frames

This subsection contains the list of all heat transfer nodes and the corresponding
definition frame:

• Physics Feature Nodes and Definition Frame

• Definition Frame of Domain Nodes
• Definition Frame of Boundary Nodes
• Definition Frame of Edge and Point Nodes

PHYSICS FEATURE NODES AND DEFINITION FRAME

The following explains the different values listed in the definition frame column in
Table 3-8, Table 3-9, and Table 3-10:

HANDLING FRAMES IN HEAT TRANSFER | 91

 Material: The inputs are entered by the user and defined in the material frame.
Because the heat transfer variables and equations are defined in the spatial frame, the
inputs are internally converted to the spatial frame.

Spatial: The inputs are entered by the user and defined in the spatial frame. No
conversion is done.

Material/(Spatial): For these physics nodes, select from the Material type list to decide
if the inputs are defined in the material or spatial frame. The default definition frame
is the material frame, which corresponds to Solid in the Material type list.

(Material)/Spatial: For these physics nodes, select from the Material type list to decide
if the inputs are defined in the material or spatial frame. The default definition frame
is the spatial frame, which corresponds to Nonsolid in the Material type list.

N/A: There is no definition frame for this physics node.

92 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

DEFINITION FRAME OF DOMAIN NODES

TABLE 3-8: DOMAIN PHYSICS NODES FOR FRAMES

NODE NAME DEFINITION FRAME

Absorbing Medium Spatial

Absorbing-Scattering Medium Spatial
Bioheat Material
Biological Tissue Material
Building Material Material
Cross Section Spatial
Thickness Spatial
Fluid Spatial
Geothermal Heating Material
Heat Source Material/(Spatial)
Immobile Fluids Spatial
Infinite Elements Spatial
Initial Values Spatial
Irreversible Transformation Spatial
Isothermal Domain Spatial
Moist Air Spatial
Moist Porous Medium Material (Solid part)
Spatial (Fluid part)
Opacity N/A
Optically Thick Participating Spatial
Medium
Out-of-Plane Heat Flux Spatial
Out-of-Plane Radiation Spatial
Participating Medium Spatial
Phase Change Material Spatial
Porous Medium Material (Solid part)
Spatial (Fluid part)
Pressure Work Spatial
Radiative Source Material/(Spatial)
Shape Memory Alloy Material
Solid Material

HANDLING FRAMES IN HEAT TRANSFER | 93

 TABLE 3-8: DOMAIN PHYSICS NODES FOR FRAMES

NODE NAME DEFINITION FRAME

Thermal Dispersion Spatial

Thermoelastic Damping Spatial
Translational Motion Material
Viscous Dissipation Spatial

94 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

DEFINITION FRAME OF BOUNDARY NODES

TABLE 3-9: BOUNDARY PHYSICS NODES FOR FRAMES

NODE NAME DEFINITION FRAME

Boundary Heat Source Material/(Spatial)

Continuity Spatial
Continuity on Interior Boundary Spatial
Deposited Beam Power Spatial
Diffuse Mirror Spatial
Diffuse Surface Spatial
External Temperature Spatial
Fracture Material (Solid part)
Spatial (Fluid part)
Heat Flux (Material)/Spatial
Heat Source Material/(Spatial)
Incident Intensity Spatial
Inflow Spatial
Irreversible Transformation Material
Isothermal Domain Interface Spatial
Lumped System Connector Spatial
Opaque Surface Spatial
Open Boundary Spatial
Outflow N/A
Periodic Condition Spatial
Prescribed Radiosity Spatial
Radiation Group N/A
Semitransparent Surface Spatial
Surface-to-Ambient Radiation Spatial
Symmetry N/A
Temperature Spatial
Thermal Contact Material
Thermal Damage Material
Thermal Insulation N/A

HANDLING FRAMES IN HEAT TRANSFER | 95

 TABLE 3-9: BOUNDARY PHYSICS NODES FOR FRAMES

NODE NAME DEFINITION FRAME

Thin Film Spatial

Thin Layer Material
The definition frames of the corresponding pair features are identical to the ones of the
standard features.

DEFINITION FRAME OF EDGE AND POINT NODES

TABLE 3-10: EDGE AND POINT NODES FOR FRAMES

NODE NAME DEFINITION FRAME

Heat Flux Spatial/(Material)

Heat Source Material/(Spatial)
Line Heat Source Material/(Spatial)
Point Heat Source Material/(Spatial)
Shell Continuity Spatial
Surface-to-Ambient Radiation Spatial
Temperature Spatial
Thin Rod Material

Material Density in Features Defined in the Material Frame

In a model the features defined in the material frame with an input field for the density
expect the density in the material frame, defined for the reference geometry. Hence this
density is constant in most cases. A non constant density in the material frame means
that there is addition or removal of matter.

As a consequence, when the material is defined from a material library and is

temperature dependent, a constant reference temperature, the volume reference
temperature, is used for the definition of the density in the Material Contents table.

The volume reference temperature defines the density in the reference geometry that
should match with the geometry in the material frame. It is a model input of all the
features defined in the material frame with an input field for the density. Following
Table 3-8, Table 3-9, and Table 3-10, this concerns the following features: Solid,
Porous Medium, Biological Tissue, Building Material, Shape Memory Alloy, Thin
Layer (Heat Transfer Interface) and Solid (Heat Transfer in Shells Interface), Fracture
(Heat Transfer Interface) and Porous Medium (Heat Transfer in Shells Interface), and
Thin Rod.

96 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Heat Transfer Consistent and
Inconsistent Stabilization Methods
The different versions of the Heat Transfer interface include the advanced option to
set stabilization method parameters. This section has information about these options.
To display the stabilization sections, click the Show More Options button ( ) and
select Stabilization in the Show More Options dialog box.

In this section:

• Consistent Stabilization
• Inconsistent Stabilization

Consistent Stabilization
This section contains two consistent stabilization methods: streamline diffusion and
crosswind diffusion. These are consistent stabilization methods, which means that they
do not perturb the original transport equation.

The consistent stabilization methods are active by default. A stabilization method is

active when the corresponding check box is selected.

STREAMLINE DIFFUSION
Streamline diffusion is active by default and should remain active for optimal
performance for heat transfer in fluids or other applications that include a convective
or translational term.

CROSSWIND DIFFUSION
Streamline diffusion introduces artificial diffusion in the streamline direction. This is
often enough to obtain a smooth numerical solution provided that the exact solution
of the heat equation does not contain any discontinuities. At sharp gradients, however,
undershoots and overshoots can occur in the numerical solution. Crosswind diffusion
addresses these spurious oscillations by adding diffusion orthogonal to the streamline
direction — that is, in the crosswind direction.

HEAT TRANSFER CONSISTENT AND INCONSISTENT STABILIZATION METHODS | 97

 Inconsistent Stabilization
This section contains a single stabilization method: isotropic diffusion. Adding
isotropic diffusion is equivalent to adding a term to the physical diffusion coefficient.
This means that the original problem is not solved, which is why isotropic diffusion is
an inconsistent stabilization method. Although the added diffusion definitely
attenuates spurious oscillations, try to minimize the use of isotropic diffusion.

By default there is no isotropic diffusion. To add isotropic diffusion, select the Isotropic
diffusion check box. The field for the tuning parameter δid then becomes available. The
default value is 0.25; increase or decrease the value of δid to increase or decrease the
amount of isotropic diffusion.

In the COMSOL Multiphysics Reference Manual:

• Stabilization Techniques
• Selection Information

98 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Heat Transfer and Fluid Flow
Coupling
COMSOL Multiphysics offers physics interfaces for heat transfer and fluid flow
computations. These interfaces have model inputs that make it possible to couple the
physics. In addition, the Nonisothermal Flow multiphysics coupling node facilitates the
coupling between heat transfer and fluid flow interfaces.

All these options make it possible to build a coupling in different ways. Even if the use
of the predefined multiphysics coupling interfaces — Nonisothermal Flow and Conjugate
Heat Transfer — is the preferred choice, other alternatives can be of interest in
particular cases. This section describes the possibility for coupling heat transfer and
fluid flow interface and lists the advantages and limitations of each approach.

In this section:

• Coupling Based on Model Inputs

• Adding Nonisothermal Flow Coupling in an Existing Model
• Nonisothermal Flow and Conjugate Heat Transfer Multiphysics Interfaces
See The Laminar Flow Interface, The Turbulent Flow, Algebraic yPlus Interface, The
Turbulent Flow, L-VEL Interface, The Turbulent Flow, k-ε Interface, and The
Turbulent Flow, Low Re k-ε Interface in the CFD Module User’s Guide for a
description of the laminar and turbulent single-phase flow interfaces.

See Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow in the CFD
Module User’s Guide for a description of the nodes associated to these interfaces.

Coupling Based on Model Inputs

This option does not use any multiphysics feature. Instead you define the model inputs
in each physics interface. You define the temperature model input in the flow interface
and the velocity and pressure model inputs in the heat transfer interface. This approach
is valid for laminar flow only. Also, the consistent stabilization does not account for the
multiphysics coupling, which can lead to convergence issues when the multiphysics
coupling is strong. Pressure work and viscous dissipation are not handled. In addition
the physics interface settings may not be optimal for the numerical treatment of the
coupling.

HEAT TRANSFER AND FLUID FLOW COUPLING | 99

 One interesting aspect of this method it that the temperature, velocity, or pressure field
does not have to be a dependent variable. You can use an analytical expression instead.

Adding Nonisothermal Flow Coupling in an Existing Model

The Nonisothermal Flow multiphysics feature can be added to a model containing a
single-phase flow and a heat transfer interface, to realize the bidirectional coupling
between the two interfaces.

It is common to start a model with a single physics (for example, fluid flow), then
implement the second one (for example, heat transfer). Then adding the Nonisothermal
Flow multiphysics feature realizes the bidirectional coupling. It also redefines the
consistent stabilization so that the multiphysics coupling effects are accounted for in
the numerical stabilization.

With the Heat Transfer Module or the CFD Module, it accounts for the turbulence in
the coupling. In particular, it modifies the effective thermal conductivity and
implements thermal wall functions if the fluid flow model requires them. Those
modifications affect the implementation of several heat transfer features. It allows to
include work done by pressure changes and viscous dissipation, and Boussinesq
approximation is supported. Finally some physics features are updated when the
Nonisothermal Flow multiphysics feature is active. In particular, the Interior Fan and
Screen fluid-flow features are updated to account for the multiphysics coupling.

Note that the physics interface settings may not be optimal for the numerical treatment
of the coupling when the multiphysics feature is added afterward.

Nonisothermal Flow and Conjugate Heat Transfer Multiphysics

Interfaces
These Nonisothermal Flow and Conjugate Heat Transfer Multiphysics Interfaces
multiphysics interfaces are identical except that they do not have the same default
features. Both contain a single-phase flow interface and a heat transfer interface
coupled with the Nonisothermal Flow multiphysics feature. So all the benefits of this
multiphysics feature (see above) are present when these multiphysics interfaces are
used.

In addition, the heat transfer and fluid flow interfaces are set up with optimal interface
settings: the discretization order of the heat transfer interface is the same as the one
used for the fluid flow interface, and the pseudo time stepping is activated in both
interfaces.

100 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

For these reasons, the use of these multiphysics interfaces is preferred.

Note that you can do a gradual implementation of the model: It is possible to start
from these multiphysics interfaces and to disable the multiphysics feature or one of the
physics in a first step and then reactivate them when the first step is validated.

HEAT TRANSFER AND FLUID FLOW COUPLING | 101

 Boundary Wall Temperature
Depending on the model configuration, a single temperature field per boundary may
not be sufficient to model accurately the temperature. In some cases, different
dependent variables are used to compute the temperature in the wall, at the wall sides,
or in the turbulent boundary layer. This section describes when additional degrees of
freedom are needed and how they are handled.

The boundary temperature variable called ht.Tvar describes the wall temperature.
When the wall has a nonconstant temperature across its thickness, this variable contains
the average value between the temperatures of the two sides of the wall. The actual
definition of ht.Tvar depends on the model configuration.

Some features define a local temperature: ht.feat1.Tvar. For example, when a

boundary heat source is applied on a particular side of the layer, this local variable
ht.bhs1.Tvar contains either the temperature upside or downside the boundary.

The following list includes existing boundary temperature variables that are available
depending on the model configuration:

• T: general temperature variable that coincides with the wall temperature in most
cases
• TWall_u: upside wall temperature defined by a Wall or an Interior Wall feature with
turbulence only if a Fluid feature is defined on the upside of the wall.
• TWall_d: downside wall temperature defined by a Wall or an Interior Wall feature
with turbulence only if a Fluid feature is defined on the downside of the wall.
• Tu: temperature on the upside of the boundary.
• Td: temperature on the downside of the boundary.
• TExtFace: external temperature of an external boundary defined by a thermally
thick boundary condition.
• TuWF: temperature of the fluid in the turbulent boundary layer near the wall only if
the Fluid feature is defined on the upside of the Wall feature.
• TdWF: temperature of the fluid in the turbulent boundary layer near the wall only if
the Fluid feature is defined on the downside of the Wall feature.

102 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

The values of these variables depend on the selections where they are defined. Here are
some illustrated cases:

• Intersection Between a Wall Boundary Feature Selection and Interface Exterior

Boundaries
• Interface Between a Fluid Domain Feature and a Solid Domain Feature Where a
Wall Boundary Feature Is Active
• Interface Between two Fluid Domain Features Where an Interior Wall Boundary
Feature Is Active

Depending on the turbulence model selected for the flow, wall functions are used or
not:

• No turbulence model: no wall functions, and TuWF = Tu, TdWF = Td.

• Turbulence models: wall functions detected by the Single Phase Flow physics
interface. TuWF and TdWF have different definitions and a flux qwf is applied on the
wall boundaries. See Temperature Condition for Automatic Wall Treatment and
Wall functions in the CFD Module User’s Manual for details.

The following sections summarize the definitions of the temperature variables for the
abovementioned configurations.

INTERSECTION BETWEEN A WALL BOUNDARY FEATURE SELECTION AND

INTERFACE EXTERIOR BOUNDARIES
The following figure shows a configuration with a fluid domain, a thermally thick
boundary condition on the left boundary, and a Wall feature on the right boundary.
This example uses wall functions.

Fluid

Tu=TExtFace Td=TWall_d Tu=Td=TWall_d

TdWF=T TdWF=T
Thermally thick layer boundary

BOUNDARY WALL TEMPERATURE | 103

 INTERFACE BETWEEN A FLUID DOMAIN FEATURE AND A SOLID DOMAIN
FEATURE WHERE A WALL BOUNDARY FEATURE IS ACTIVE

Solid in the Domain Downside the Wall

The following figures show configurations with a solid as downside domain and a fluid
as upside domain with either a thermally thick or thermally thin boundary condition
in between. This example uses wall functions.

Solid Fluid

Downside domain Upside domain

of the boundary of the boundary

Td=down(T) Tu=TWall_u Tu=Td=TWall_d

TuWF=up(T) TdWF=T
Thermally thick layer boundary

Solid Fluid

Downside domain Upside domain

of the boundary of the boundary

Tu=Td=down(T) Tu=Td=TWall_d

Thermally thin layer boundary

104 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Solid in the Domain Upside the Wall
The following figures show configurations with a fluid as downside domain and a solid
as upside domain with either a thermally thick or thermally thin boundary condition
in between. This example uses wall functions.

Fluid Solid

Downside domain Upside domain

of the boundary of the boundary

Tu=Td=Twall_d
Td=TWall_d Tu=up(T)
TdWF=down(T)
Thermally thick layer boundary

Fluid Solid

Downside domain Upside domain

of the boundary of the boundary

Tu=Td=Twall_d Tu=Td=up(T)

Thermally thin layer boundary

BOUNDARY WALL TEMPERATURE | 105

 INTERFACE BETWEEN TWO FLUID DOMAIN FEATURES WHERE AN
INTERIOR WALL BOUNDARY FEATURE IS ACTIVE
The following figure shows a configuration with two fluid domains with a thermally
thick boundary condition and an Interior Wall feature in between. This example uses
wall functions.
Fluid Fluid

Downside domain Upside domain

of the boundary of the boundary

Tu=Td=TWall_d Tu=Td=TWall_d

Td=TWall_d Tu=TWall_u
TdWF=down(T) TuWF=up(T)

Thermally thick layer and Interior

Wall boundary

106 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

S o lv e r S e t t i ng s
The information about default solvers given below is specific to the Heat Transfer and
Moisture Transport interfaces when the Stationary and Time-Dependent studies are
used. A comprehensive description of solver settings and corresponding theory is
available in the Study and Study Step Types section of the COMSOL Multiphysics
Reference Manual.

See also Studies and Solvers in the COMSOL Multiphysics Reference

Manual

In this section:

• Linearity Property of the Temperature Equation

• Linear Solver
• Nonlinear Solver
• Time-Dependent Study Step
• Guidelines for Solving Surface-to-Surface Radiation Problems
• Guidelines for Solving Multiphysics Problems

Linearity Property of the Temperature Equation

The Heat Transfer interfaces define an elliptic partial differential equation for the
temperature, T, of the form:

∂T
ρC p +∇⋅q = Q
∂t

q = – k ∇T

with Dirichlet and Neumann boundary conditions at some boundaries:

T = T0

–n ⋅ q = q0

In its basic form, the density, ρ, heat capacity, Cp, thermal conductivity, k, heat sources,
Q, constraint temperatures, T0, and heat fluxes, q0, are all constant, which leads to a

SOLVER SETTINGS | 107

 linear system. Here, linear solvers described in the next paragraphs are completely
suited for the resolution.

However, nonlinearities can appear in the equation in the following cases:

• The material properties, ρ, Cp, and k, have a temperature dependency.

• The heat sources are not linear in temperature.
• The Neumann boundary condition is not linear in temperature, hence
- A convective cooling condition of type −n ⋅ q = h(Text − T) keeps the linearity of
the problem when the heat transfer coefficient, h, is constant.
- A radiative condition of type −n ⋅ q = εσ(Tamb4 − T4) is strongly nonlinear.

Different nonlinear solvers are also provided for these kinds of problems.

Linear Solver

DEFAULT LINEAR SETTINGS FOR HEAT TRANSFER AND MOISTURE

TRANSPORT INTERFACES
The default linear solver is determined based on the number of degrees of freedom and
physics interface settings.

For small number of degrees of freedom, the direct PARDISO solver is used. It is
known to be robust and fast for small-sized problems.

For larger models, the linear iterative GMRES solver with multigrid preconditioner is
used. In most cases, SOR is the presmoother and postsmoother. This solver is memory
effective and fast for large models.

When the model contains settings that lead to a system matrix with 0 on the diagonal
(for example, Lagrange multipliers for weak constraints), SOR cannot be used and is
replaced by Vanka, which is usually slower and uses more memory.

TUNING LINEAR SOLVER

Tuning the linear solver may be considered in case of nonconvergence or low
performance. When convergence fails you should first verify that this is not due to an
ill-posed model, or inappropriate settings in the Time-Dependent study or nonlinear
solver.

Several options are available to tune the linear solver settings. This paragraph focuses
only on the most commonly used ones.

108 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Switch to PARDISO
When the GMRES solver with multigrid preconditioner is set by default, using
PARDISO instead can be considered provided that enough RAM is available. Indeed,
PARDISO usually converges easily but uses much more memory than the default
iterative solver. If PARDISO does not converge, it may indicate that there is an issue
in the model definition or with other solver settings.

Optimize GMRES/Multigrid for Memory

In order to optimize further the memory needed by the iterative solver, the number of
mesh elements on the coarser multigrid level can be reduced by, for instance,
increasing the Mesh coarsening factor or the Number of multigrid levels. The latter
strategy may also increase the resolution time.

Optimize GMRES/Multigrid for Convergence

When the linear solver has difficulties to converge, the following settings can be tuned:

• When the convergence graph of GMRES shows a slow down every 50 iterations, the
Number of iteration before restart parameter (default value of 50) should be
increased — doubled for example. This may also increase the memory consumption.
• Increasing the Number of iteration in the Multigrid settings, and in the presmoother
and postsmoother nodes improves the quality of the preconditioner and
convergence of GMRES.
• Since an excessive difference between two multigrid levels can affect the
convergence, lowering the Mesh coarsening factor in the Multigrid settings can help
convergence.
• Consider creating the multigrid level meshes manually if the automatic coarsening
method fails or leads to poor quality meshes.

Choosing the Right Linear System Solver in the COMSOL Multiphysics

Reference Manual

Nonlinear Solver

DEFAULT NONLINEAR SETTINGS FOR HEAT TRANSFER INTERFACES

Nonlinear solver settings depend on the heat transfer model and on the study type.

SOLVER SETTINGS | 109

 Fully Coupled Solver Attribute
Heat transfer models use a fully coupled nonlinear solver attribute by default. The
Jacobian update is set to On every iteration for Stationary studies, and to On first iteration
for Time-Dependent studies. A Newton nonlinear method is set by default with

• Automatic damping factor computation for stationary studies

• Constant damping factor for time-dependent studies, with a default damping factor
set to 0.9.

Segregated Solver Attribute

The segregated solver attribute is set by default in the following cases:

• Another physics interface is solved together with heat transfer. The dependent
variables of the Heat Transfer interface are placed in a separate segregated group.
• Radiation in participating media using the Discrete ordinates method defines a large
number of dependent variables (up to 168 for a single wavelength), which are placed
in segregated groups. The number of dependent variables per segregated group and
the nonlinear method settings depend on the Performance index parameter available
in the heat transfer interface settings in the Participating Media Settings section.
When multiple wavelength are considered, the variables relative to distinct
wavelengths are not mixed together in the segregated groups.
• The Thermal Damage subfeature (added under Biological Tissue feature) defines an
additional variable alpha that is placed in a dedicated segregated group.

DEFAULT NONLINEAR SETTINGS FOR MOISTURE TRANSPORT

INTERFACES
Nonlinear solver settings depend on the study type.

Fully Coupled Solver Attribute

A fully coupled solver attribute is set by default. A Newton nonlinear method with a
constant damping factor is set by default, with the following settings:

• The Damping factor is set to 0.7.

• The Jacobian update is set to On first iteration for Time-Dependent studies, and to On
every iteration for Stationary studies or when the liquid concentration on surfaces,
cl_evap, is solved.

• The Maximum number of iterations is set to 5 for Time-Dependent studies, and to 50

for Stationary studies.
• The Tolerance factor is set to 0.01 for Time-Dependent studies, and to 0.1 for
Stationary studies.

110 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Segregated Solver Attribute
The segregated solver attribute is set by default when another physics interface is
solved together with moisture transport. The dependent variables of the Moisture
Transport interface are placed in a separate segregated group.

A Newton nonlinear method with a constant damping factor is set by default, with the
following settings:

• The Damping factor is set to 0.7.

• The Jacobian update is set to On first iteration for Time-Dependent studies, and to On
every iteration for Stationary studies or when the liquid concentration on surfaces,
cl_evap, is solved.
• The termination technique is Iterations, with Number of iterations set to 2.

DEFAULT NONLINEAR SETTINGS FOR THE LUMPED THERMAL SYSTEM

INTERFACE
A fully coupled solver attribute is set by default. A Newton nonlinear method with a
constant damping factor is set by default, with the following settings:

• The Damping factor is set to 1.

• The Jacobian update is set to On every iteration.
• The Maximum number of iterations is set to 5 for Time-Dependent studies, and to 50
for Stationary studies.

TUNING THE NONLINEAR SOLVER

Default solver settings are defined to handle efficiently classical configurations. For
particular applications, the default settings may need modifications to improve the
robustness and performance of the solver.

Optimize Nonlinear Solver for Robustness

When the nonlinear solver fails or converges erratically, different options can be
considered:

• Using the Automatic highly nonlinear (Newton) option forces to start the
computation with a very low damping factor and increases it carefully. Alternatively
a low constant damping factor can be used. The damping factor ranges between 0
and 1. A constant damping factor equal to 0.1 is a very low value and should be
robust but slow to converge. For low values of the damping factor, it is thus usually
needed to increase the number of nonlinear iterations. If the nonlinear solver is
unstable with such a damping factor then the automatic option should be used

SOLVER SETTINGS | 111

 because it makes it possible to start with a lower damping factor and gradually
increases it.
• A good initial value, as close as possible from the expected solution and consistent
with the boundary conditions, helps to guide the nonlinear solver to a stable
physical solution. To do that:
- Try to ramp the temperature on the boundary from the initial to the desired value
by using a auxiliary sweep — for stationary problems — or a time-dependent step
function — for time-dependent problems.
- Use results from a simplified problem, for instance with no temperature
dependency, or using a one-way multiphysics coupling, as initial value.
Note that it is sometimes easier to update the boundary conditions than the initial
condition to get consistent initial settings (see the Heat Conduction in a Finite
Slab model).
• When it is not possible to provide a good initial value, the segregated solver
associated with low damping factors in each segregated step helps to achieve
convergence.
• Forcing the Jacobian update at every iteration ensures that the nonlinear solver
iterates using optimal information from the equation system. This is needed when
nonlinearities are due to the temperature itself — for example, in case of strong
temperature dependency of material properties — or to another variable solved in
the same segregated group as the temperature — for example, in natural convection
models.

Optimize Convergence Speed

Low convergence can be improved by following ways:

• Using a constant damping factor equal to 1 for linear problems. The linearity is
determined at the beginning of the resolution and indicated in the Log section of the
solver window.
• Providing a good initial value is an asset for computational speed.
• In the convergence area, the fully coupled solver has a better convergence rate than
the segregated solver.
• Using minimal Jacobian update option avoid to spend time in Jacobian
computation. This is suited for linear models and models with mild nonlinearities.

112 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Time-Dependent Study Step

DEFAULT TIME SETTINGS FOR HEAT TRANSFER INTERFACES

The default time-stepping method for the Heat Transfer interfaces is BDF at second
order. It excludes algebraic variable from the error estimate.

When a Thermal Damage subfeature is present under Biological Tissue feature, particular
settings for the time-dependent solver are used to efficiently compute the damage
indicators:

• The Absolute Tolerance of the scaled damage indicator variable is set to 1, meaning
that these variable are neglected in the error estimate.
• The damaged tissue indicator, α, is solved with an iterative Jacobi method.
• If the Adaptive mesh refinement option is selected in the study settings, the error
indicator is set to ∇θ d, sm ⋅ ∇θ d, sm , where θd, sm is the smoothed indicator of
necrotic tissue (the fraction of necrotic tissue, θd, is discontinuous in general).
• If the Temperature threshold option is used in the Biological Tissue feature, the instant
necrosis indicator, alphanecr, is placed in the Previous Solution step. This setting
avoids wrong detection of irreversible damage due to nonlinear iterations that may
go through a state where the damage criteria is met and then converge to a solution
where the damage criteria is no longer met. It uses a direct linear solver. The default
nonlinear method is the Newton method with constant damping factor.

When the Irreversible Transformation feature is active, similar settings are used:

• The Absolute Tolerance of the scaled irreversible transformation indicator dependent

variable is set to 1.
• The irreversible transformation indicator α is solved with an iterative Jacobi method.

DEFAULT TIME SETTINGS FOR MOISTURE TRANSPORT INTERFACES

The default time-stepping method for the Moisture Transport interfaces is BDF at
second order. It includes algebraic variable from the error estimate.

TUNING THE TIME-DEPENDENT SOLVER

The quality of the time-stepping influences the nonlinear solver convergence. Tiny
time steps usually lead to mildly nonlinear problems at each time step whereas large
time steps can result in (fewer) highly nonlinear problems.

The default solver settings for transient heat transfer and moisture transport define the
maximal number of nonlinear iterations to 5. If this is not sufficient, it is recommended

SOLVER SETTINGS | 113

 to use smaller time steps and to verify if the model definition does not contain
discontinuities in time. If so, consider using smooth step functions to model sharp
variations in time.

There are several ways to control the time step size:

• An implicit way is to define a lower relative tolerance in the study settings. When the
relative tolerance is lowered, the absolute tolerance should be reduced in the same
proportion. The default Relative tolerance is set to 0.01 for Time-Dependent studies,
and to 0.1 for Stationary studies.
• The most explicit way is to define a maximum time step. This is an appropriate
option when the same maximum time step is relevant for the entire simulation.
Otherwise, it is possible to include times of interest in the Times field of the
time-dependent study and to use the Intermediate option in the Time Stepping
settings.
• Lastly you can control the time step by triggering an event when a particular
condition is meet (see the documentation about The Events Interface in the
COMSOL Multiphysics Reference Manual). This advanced method can be
efficient when the other simpler methods are not applicable.

It is also recommended to inspect the solver log and check the default scaling of
dependent variables in case of convergence failure. In case of incorrect automatic
scaling, consider using Manual settings in the Dependent Variable attribute node.

Time-Dependent Solver in the COMSOL Multiphysics Reference

Manual

Guidelines for Solving Surface-to-Surface Radiation Problems

The following guidelines are helpful when selecting solver settings for models that
involve surface-to-surface radiation:

• Surface-to-surface radiation makes the Jacobian matrix of the discrete model partly
filled as opposed to the usual sparse matrix. The additional nonzero elements in the
matrix appear in the rows and columns corresponding to the radiosity degrees of
freedom. It is therefore common practice to keep the element order of the radiosity
variable, J, low. By default, linear Lagrange elements are used irrespective of the
shape-function order specified for the temperature. When you need to increase the

114 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 resolution of your temperature field, it might be worth considering raising the order
of the temperature elements instead of refining the mesh.
• The Assembly block size parameter (found in the Advanced solver feature) can have a
major influence on memory usage during the assembly of problems where
surface-to-surface radiation is enabled. When surface-to-surface is detected, the
solver sets the assembly block size at 100. Using a smaller block size also leads to
more frequent updates of the progress bar.

Introduction to Solvers and Studies and Advanced in the COMSOL

Multiphysics Reference Manual

Guidelines for Solving Multiphysics Problems

MULTIPHYSICS MODELS
Unless the model contains a multiphysics node that defines a coupling between a Heat
Transfer interface and another interface (see Multiphysics Couplings below), each
physics interface defines default solver settings that are merged.

The Heat Transfer interfaces always define a dedicated segregated group that uses a
linear solver optimized for the heat transfer equations. For strongly coupled models, it
may be efficient to merge two (or more) segregated steps. In this case, a unique linear
solver must be chosen for the fully coupled solver or the new segregated group.

Time-dependent settings from different physics interfaces may compete. When the
different settings are merged the strictest one is kept.

MULTIPHYSICS COUPLINGS
When a Heat Transfer interface is coupled with another physics interface through a
multiphysics coupling feature, additional predefined default settings are loaded. The
next two paragraphs describes the subtleties of the Nonisothermal Flow,
Electromagnetic Heating, Heat and Moisture Transport, and Moisture Flow
interfaces.

Nonisothermal Flow
The Nonisothermal Flow multiphysics coupling controls the solver settings for the
flow and the temperature dependent variables.

When it assumes a weak coupling between the flow and the heat interfaces (typically
no gravity force accounted for in the flow interface), the default solver contains

SOLVER SETTINGS | 115

 dedicated segregated groups for heat and flow dependent variables. Each uses the
default linear solver of the corresponding interface, except that the presmoother and
postsmoother in the GMRES iterative solver switches from SOR to SCGS.

When a strong coupling is assumed (gravity force accounted for in the flow interface),
the default solver merges the temperature, pressure, and velocity groups. In this case,
the solver corresponds to the default solver of the heat transfer interface. This is meant
to suits well for nonisothermal flows in which natural convection dominates. See
Default Linear Settings for Heat Transfer and Moisture Transport interfaces, Default
Nonlinear Settings for Heat Transfer Interfaces, and Default Time Settings for Heat
Transfer Interfaces for details.

Electromagnetic Heating
The Electromagnetic Heating multiphysics interfaces (Joule heating, Laser Heating,
Induction Heating, and Microwave Heating) define default settings that solve the
temperature and the electromagnetic fields using a coupled step. It can be the fully
coupled nonlinear solver if there is no additional variable to solve for, otherwise it is a
segregated step containing the temperature and the electromagnetic variables.
However when radiation in participating media or damage variable are solved they are
placed in a separate group as described above.

Heat Transfer with Surface-to-Surface Radiation

The Heat Transfer with Surface-to-Surface Radiation multiphysics coupling assumes a
weak coupling between temperature and radiosity variables solved by a Heat Transfer
interface and a Surface-to-Surface Radiation interface, respectively. The default solver
contains dedicated segregated groups for temperature and radiosity variables.

Heat and Moisture Transport

The Heat and Moisture Transport multiphysics coupling controls the solver settings
for the temperature and relative humidity variables.

The default solver merges the temperature and relative humidity groups. The solver
corresponds to the default solver of the moisture transport interface. See Default
Linear Settings for Heat Transfer and Moisture Transport interfaces, Default
Nonlinear Settings for Moisture Transport Interfaces, and Default Time Settings for
Moisture Transport Interfaces for details.

Moisture Flow
The Moisture Flow multiphysics coupling controls the solver settings for the flow
(velocity and pressure) and the relative humidity variables.

116 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

When it assumes a weak coupling between the flow and the moisture interfaces
(typically no gravity force accounted for in the flow interface), the default solver
contains dedicated segregated groups for moisture and flow dependent variables. Each
uses the default linear solver of the corresponding interface, except that the
presmoother and postsmoother in the GMRES iterative solver switches from SOR to
SCGS.

When a strong coupling is assumed (gravity force accounted for in the flow interface),
the default solver merges the relative humidity, pressure, and velocity groups. In this
case, the solver corresponds to the default solver of the moisture transport interface.
See Default Linear Settings for Heat Transfer and Moisture Transport interfaces,
Default Nonlinear Settings for Moisture Transport Interfaces, and Default Time
Settings for Moisture Transport Interfaces for details.

SOLVER SETTINGS | 117

 P lo t t i ng a nd E v al u at i n g R esu l t s i n
Layered Materials
When modeling heat transfer in thin layers, the thin layer’s thickness is not explicitly
represented in the geometry. However, a 1D extra dimension is attached to the
boundary, via the selection of a layered material, and degrees of freedom are added
through the thickness of the layered material. This concerns the Thin Layer, Thin Film,
and Fracture nodes of the Heat Transfer interface, and all the features of the Heat
Transfer in Shells interface.

Therefore, dedicated tools are available to plot and evaluate the results in the 1D extra
dimension.

Plotting Along and Through the Layered Material

There are predefined plots to display the variables in the 1D extra dimension:

• The Layered Material Slice plot displays the variables along the layered material, for
specified through-thickness locations in the 1D extra dimension.
• The Through Thickness plot displays the variables in the thickness direction, at
specified points on the boundary

See Layered Material, Layered Material Slice, and Through Thickness in

the COMSOL Multiphysics Reference Manual for details about the
Layered Material dataset, and slice and through-thickness plots in layered
materials.

Using Extra Dimensions in the COMSOL Multiphysics Reference

Manual.

118 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Built-in Operators for Evaluation in the Layered Material

EVALUATION ALONG THE LAYERED MATERIAL

The extra dimension temperature T can be evaluated along the layered material by
using the atxd1 operator on the extra dimension, with the expression
xdimTag.atxd1(xd, expr), where:

• xdimTag is the extra dimension tag. For example, it may be slmat1_xdim, when a
single layer material (slmat1) is selected in the Layered material list of the Thin Layer
node. This tag can be deduced from the Selection column of the Equation View
subnode of the node applied on the thin layer, by clicking the Show More Options
button ( ) on the Model Builder tool bar and selecting Equation View.
• xd is the coordinate in the extra dimension. It varies from 0 to slmat1.th, which
is the total thickness of the layered material. By convention, xd=0 corresponds to
the downside of the boundary where the thin layer is defined, whereas
xd=slmat1.th corresponds to its upside. Upside and downside settings can be
visualized by plotting the global normal vector (nx, ny, nz), that always points from
downside to upside. See Tangent and Normal Variables in the COMSOL
Multiphysics Reference Manual. Note that the normal vector (ht.nx, ht.ny,
ht.nz) may be oriented differently.
• expr is the quantity to be evaluated at the point xd. For example, it can be set to T
to evaluate the temperature. There are others postprocessing variables defined on
the extra dimension that can be found in the Equation View subnode of the node
applied on the thin layer.

Extra
dimension

Computational
domain Along
the layer
xd

Thin Layer boundary

Figure 3-2: Schematic representation of a 2D geometry with a thin layer composed of three
layers, with an evaluation of the results along the layer at the coordinate xd.

PLOTTING AND EVALUATING RESULTS IN LAYERED MATERIALS | 119

 This operator can be used to define manually a slice plot in the layered material. In 2D
and 2D axisymmetric geometries, the slice is represented in a line graph under a 1D
plot group, whereas it is a surface plot for 3D geometries.

EVALUATION THROUGH THE LAYERED MATERIAL

The extra dimension temperature T can be evaluated through the layered material, at
a specific point (x0, y0) or (x0, y0,z0), by using the operators atxd1 in 2D and 2D
axisymmetric geometries, and atxd2 in 3D geometries. Note that this time the
operators are used on the component and not on the extra dimension.

In 2D and 2D axisymmetric geometries, use compTag.atxd1(x0, y0, expr), where:

• compTag is the component tag. In most cases, this tag is comp1. It is possible to
check it in the Properties window of the component node (display it by
right-clicking on the node and selecting Properties).
• x0 and y0 are the coordinates of the point in the base geometry that belongs to the
boundary linked with the extra dimension. Note that these are 2D coordinates from
the global coordinate system and not curvilinear coordinates.
• expr is the quantity to be evaluated at the point (x0, y0). For example, it can be set
to T to evaluate the temperature.

In 3D geometries, use compTag.atxd2(x0, y0, z0, expr), where:

• compTag is the component tag. For example, it can be comp1.

• x0, y0, and z0 are the coordinates of the point in the base geometry that belongs
to the boundary linked with the extra dimension.
• expr is the quantity to be evaluated at the point (x0, y0, z0). For example, it can
be set to T to plot the temperature. Note that here too, these are 3D coordinates
from the global coordinate system and not curvilinear coordinates.

120 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 Extra
dimension
Computational
domain

( x0, y0)

Thin Layer boundary

Along the extra dimension

Figure 3-3: Schematic representation of a 2D geometry with a thin layer composed of three
layers, with an evaluation of the results through the layer at the point (x0,y0).

This operator can be used to define manually a through thickness plot in the layered
material. For all dimensions, the section is represented in a line graph under a 1D plot
group. In order to use this, the Dataset selected in the Data section of the 1D plot
group has to select the extra dimension as component. One method is to duplicate the
default Solution node under the Datasets node, set the Component of the new node to
the extra dimension, and use this new dataset into the 1D plot group, with the domains
of the extra dimension selected.

INTEGRATION IN THE LAYERED MATERIAL

The integral of a variable in the layered material defined on a thin layer boundary can
be evaluated by using the xdintopall operator, with the following expression:
featTag.xdintopall(expr), where:

• featTag is the tag of the physics node defining the operator xdintopall. For
example, for a Thin Layer node in the Heat Transfer interface, it may be ht.sls1.
• expr is the quantity to be integrated. For example, it can be set to T to evaluate the
temperature.

LOCALIZATION OF THE EVALUATION IN THE THIN LAYER

For a layerwise computation, the atonly operator should be used to multiply the
expression to be evaluated. This operator, applied to the extra dimension with the
following expression, xdimTag.atonly(dom==i), returns 1 on the i th layer of the
layered material, 0 on the other layers.

PLOTTING AND EVALUATING RESULTS IN LAYERED MATERIALS | 121

 Here, xdimTag is the tag of the extra dimension. For example, it may be llmat1_xdim
when a layered material link (llmat1) is selected in the Layered material list of the Thin
Layer node. This tag can be deduced from the Selection column of the Equation View
subnode of the node applied on the thin layer, by clicking the Show More Options
button ( ) on the Model Builder tool bar and selecting Equation View.

Therefore, by using featTag.xdintopall(xdimTag.atonly(dom==i)*expr), you

evaluate the integral of expr on the i th layer of the layered material.

See atxd, atonly, and noxd in the COMSOL Multiphysics Reference

Manual for details about the operators.

122 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Evaluating the Bulk Temperature
In a nonisothermal flow, the bulk temperature is an average temperature, taking into
account the velocity field in the cross section of the flow. It can be evaluated in post
processing, to estimate the heat transfer coefficient over a wall along the flow main
direction.

This chapter describes how to evaluate the bulk temperature in 3D and 2D models,
using the built-in geometry variables and operators. The nonisothermal flow is
supposed to take place along a tube, with any orientation and curvature, even though
the bulk temperature may be defined in other geometries. It contains the following
sections:

• Bulk Temperature Definition

• Evaluation of the Bulk Temperature in 3D Models
• Evaluation of the Bulk Temperature in 2D Models

Bulk Temperature Definition

The bulk temperature Tb is defined as

 ( u ⋅ τ )T ds
S
--------------------------------
⊥
Tb =
 u ⋅ τ ds
S⊥

where S ⊥ is a cross-sectional surface of the tube where the flow takes place, u is the
flow velocity field, τ is the tangential direction of the flow, and T is the temperature
field.

The bulk temperature can be used to evaluate the heat transfer coefficient h at tube
walls:

qw
h = --------------------
Tw – Tb

where qw is the wall heat flux, and Tw is the wall temperature.

EVALUATING THE BULK TEMPERATURE | 123

 Evaluation of the Bulk Temperature in 3D Models
In 3D models, the bulk temperature may be evaluated along an edge of the tube in the
direction of the flow, by using the built-in diskint(R,expr,tol) operator. When
applied to an edge in 3D models, this operator integrates an expression expr over a
disk of radius R, perpendicular to the selected edge. A numerical tolerance parameter
tol is also available to adjust the numerical precision.

See ball, circle, disk, and sphere in the COMSOL Multiphysics Reference
Manual for details about the operators.

In the evaluation of the bulk temperature, the radius R should be chosen large enough
to encompass the tube cross section, but not too large regarding numerical precision.

If the evaluation disk overlaps other domains, the evaluation may be restricted to the
domain of interest by adding a test on the dom variable.

The evaluation of the tangential velocity over the cross section requires to access the
vector tangent to the edge, in the computational domain. However, the tangent
vectors t1 and t2 are defined on boundaries only. By applying the general extrusion
operator genext1 on the edge, with a mapping to the domain, the tangential vector
genext1(t1) is defined all over the cross section, and can be used in the diskint
operator expression.

See General Extrusion in the COMSOL Multiphysics Reference Manual

for details about the operators.

Finally, the bulk temperature can be expressed as

Tb =
diskint(R,if(dom==i,ut*T,0),tol)/diskint(R,if(dom==i,ut,0),tol)

where ut = u*genext1(t1x)+v*(genext1(t1y)+w*genext1(t1z), i is the domain

entity number, and the tolerance tol should be specified.

Evaluation of the Bulk Temperature in 2D Models

In 2D models, the evaluation disk of the diskint operator is oriented in-plane, so this
operator is not suitable for an integration over the cross section, and the
integrate(expr,var,lower,upper,tol) operator should be used instead. This

124 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

operator computes the integral of expr for the integration variable var over an interval
specified by expressions lower and upper. A relative tolerance tol may be specified as
well.

See integrate in the COMSOL Multiphysics Reference Manual for details

about the operators.

In order to integrate over the cross section of the flow, the cross section line is
parametrized as follows:

S ⊥ = { ( x + sn x, y + sn y ), s ∈ [ 0, 1 ] }

where (nx,ny) is the normal vector to the boundary, and hence the vector tangent to
the cross section.

The at2 operator, which is the spatial evaluation operator in 2D, should be used to
evaluate the expression expr on this parametrization of the cross section.

See at in the COMSOL Multiphysics Reference Manual for details about

the operators.

In practice, the integral is performed over a larger interval [-R,R], where R should be
chosen large enough to encompass the cross section of the tube, but not too large
regarding numerical precision. The isnan operator can be used to ensure that the
solution is defined at the prescribed coordinates, and as in 3D models, the dom variable
may be used to identify the domain of interest.

See isinf and isnan in the COMSOL Multiphysics Reference Manual for
details about the operators.

As in 3D models, the evaluation of the tangential velocity over the cross section
requires to access the vector tangent to the boundary, in the computational domain.
However, the tangent vector t is defined on boundaries only. By applying the general
extrusion operator genext1 on the boundary, with a mapping to the domain, the

EVALUATING THE BULK TEMPERATURE | 125

 tangential vector genext1(t) is defined all over the cross section, and can be used in
the integrate operator expression.

See General Extrusion in the COMSOL Multiphysics Reference Manual

for details about the operators.

Finally, the bulk temperature can be expressed as:

Tb = A/B

where

A = integrate(if(isnan(at2(x+s*nx,y+s*ny,ut*T)),0,at2(x+s*nx,y+s*
ny,ut*T)),s,-R,R,tol)
B = integrate(if(isnan(at2(x+s*nx,y+s*ny,ut*T)),0,at2(x+s*nx,y+s*
ny,ut)),s,-R,R,tol)

and ut = u*genext1(tx)+v*(genext1(ty), i is the domain entity number, and the

tolerance tol should be specified.

126 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Ambient Properties
The ambient data available under the Definitions branch come from the processing of
measured data from ASHRAE Weather Data Viewer 5.0 (©2013 ASHRAE,
www.ashrae.org. Used with permission.) and ASHRAE Weather Data Viewer 6.0
(©2017 ASHRAE, www.ashrae.org. Used with permission.), given as frequencies of
observations, monthly and hourly averaged for several past years of observation. They
provide time-dependent weather conditions for more than 8000 stations worldwide in
terms of the dry bulb temperature, the dew point temperature, the relative humidity,
the wind speed, the absolute pressure, the precipitation rate, and the solar direct and
diffuse irradiance.

In this section:

• Settings for the Ambient Properties

• Processing of ASHRAE Data
• Ambient Data Interpolation
• Ambient Variables and Conditions
• Use of Ambient Data From the Features

When no special mention is added, the term temperature stands for the
dry bulb temperature. See Moist Air Variables for the definition of the dry
bulb temperature, dew point temperature, and relative humidity.

Settings for the Ambient Properties

To add an Ambient Properties node to any Component, on the Physics toolbar click
Ambient Properties or right-click the Definitions node under the Component and
choose Ambient Properties.

In the Ambient Properties node ( ), you can define ambient variables to be available
as inputs from several features: the temperature Tamb, the absolute pressure pamb, the

AMBIENT PROPERTIES | 127

 relative humidity φamb, the wind velocity vamb, and the direct and diffuse solar
irradiances Isn,amb and Ish,amb.

See Heat Flux, Diffuse Surface (Surface-to-Surface Radiation Interface),

Temperature, Initial Values, or External Radiation Source for examples of
features of the Heat Transfer interface that have ambient variables
available as inputs.

See Initial Values, Moisture Content, or Moisture Flux for examples of

features of the Moisture Transport interface that have ambient variables
available as inputs.

Three options are available for the definition of the Ambient data:

• When User defined (the default) is selected, the Temperature Tamb, the Absolute
pressure pamb, the Relative humidity φamb, the Wind velocity vamb, the Precipitation
rate P0,amb, the Clear sky noon beam normal irradiance Isn,amb, and the Clear sky
noon diffuse horizontal irradiance Ish,amb should be specified directly.
• When either Meteorological data (ASHRAE 2013) or Meteorological data (ASHRAE 2017)
is selected, the ambient variables are computed from monthly and hourly averaged
measurements, made over several years at weather stations worldwide. See
Processing of ASHRAE Data for more information. Further settings for the choice
of the location, time, and ambient conditions are needed; and additional input fields
are displayed underneath.

Location
In this section you can set the location by choosing among more than 8000 weather
stations worldwide. Two options are available for the selection of the Weather station:

• When From list is selected, click on the Set Weather Station button to open the
Weather Station browser that allows you to select a Region, a Country, and a Station.

128 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 The station details (World Meteorological Organization reference, location, and
geographical coordinates) are displayed underneath.
• It is also possible to select a station From reference by entering the corresponding
World Meteorological Organization reference, which is a 6-digit number.

A single country may be available for more than one region selection if it
has stations spread over different regions. For example, United States of
America is available in the Country list when either North America,
Eurasia, or Oceania is selected in the Region list.

When a station is selected From list, its World Meteorological

Organization (WMO) reference is displayed in the user interface. The
WMO references can also be retrieved from maps offered by third parties
like the one available as of this product release on
http://ashrae-meteo.info/v2.0/.

Time
The Date and Local time should be set by entering values or expressions in the Day,
Month, Hour, Minute, and Second fields of the two tables.

If On is selected in the Specify year list, a value or expression for the Year should also be
set. As the data are given as averages over several past years, this input is only used for
the detection of leap years, in order to interpolate the data over the months.

For temporal studies, these inputs define the start time of the simulation. By default,
the Update time from solver check box is selected, and the time is then automatically
updated with the time from the solver to evaluate the variables by interpolation of the
measured data. Clear this check box to manually set the time update.

See Ambient Data Interpolation for more information about the

interpolation of data over months and hours.

See Processing of ASHRAE Data for more information about the data.

AMBIENT PROPERTIES | 129

 A time unit suitable for simulations over a day or a year may be set in the
Study Settings section of the Time Dependent node, by using for example
h for hour, d for day, or a for a year. See Using Units in the COMSOL
Multiphysics Reference Manual for more details.

Ambient Conditions
Based on the measured data, several conditions are available for the Temperature, the
Dew point temperature, the Wind speed, and the Precipitation rate. The formula for each
condition is recalled in Table 3-11, Table 3-12, and Table 3-13. The Average
conditions correspond to weighted means of the measured data, whereas the other
conditions are obtained by applying standard or modified deviations (Low, High, and
User defined coefficient for deviation conditions), user defined corrections, or wind
correlations to the average conditions; or by taking the minimum or maximum of the
measured data (Lowest and Highest conditions). More information about these
definitions can be found in Ambient Variables and Conditions.
TABLE 3-11: TEMPERATURE CONDITIONS

CONDITION DEFINITION

Average T amb =  T station

Low T amb =  T station – σ T, station
High T amb =  T station + σ T, station
Lowest T amb = min ( T station )
Highest T amb = max ( T station )
User defined coefficient for deviation T amb =  T station + c σ ⋅ σ T, station
User defined correction T amb =  T station + ΔT
Heating wind correlation(1) T amb =  T station + ΔT wind
(1)
Cooling wind correlation T amb =  T station – ΔT wind
(1)
These correlations are not related to the wind speed conditions described in
Table 3-13.

TABLE 3-12: DEW POINT TEMPERATURE CONDITIONS

CONDITION DEFINITION

Average DPT amb =  DPT station

Low DPT amb =  DPT station – σ DPT, station
High DPT amb =  DPT station + σ DPT, station
Lowest DPT amb = min ( DPT station )
Highest DPT amb = max ( DPT station )

130 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TABLE 3-13: WIND SPEED CONDITIONS

CONDITION DEFINITION

Average v amb =  v station

Low v amb =  v station – σ v, station
High v amb =  v station + σ v, station
Lowest v amb = min ( v station )
Highest v amb = max ( v station )

TABLE 3-14: PRESSURE CONDITIONS

CONDITION DEFINITION

Average P 0, amb =  P 0, station

Low P 0, amb =  P 0, station – σ P
0, station
High P 0, amb =  P 0, station + σ P
0, station
Lowest P 0, amb = min ( P 0, station )
Highest P 0, amb = max ( P 0, station )

The conditions set for Temperature and Dew point temperature should be
consistent in order to keep the temperature larger than the dew point
temperature. However, all settings combinations are available, and the
relative humidity is majored by 1 when necessary.

The sum of the Clear sky noon beam normal irradiance and the Clear sky
noon diffuse horizontal irradiance is available through the postprocessing
variable ht.Is_amb, defined as the Ambient solar irradiance.

Processing of ASHRAE Data

From frequencies of measured values, a weighted mean of the data and a standard
deviation from the weighted mean are computed for each month. For the temperature,
more data are available and the weighted mean and standard deviation are also
computed at each hour. Figure 3-4 shows an example of weighted mean computation

AMBIENT PROPERTIES | 131

 for the diurnal temperature fluctuations from the initial data given as frequencies of
observations.

Figure 3-4: Computation of weighted mean from frequencies of observations for the
diurnal fluctuations of temperature.

These values are used for the definition of different conditions, as detailed in Ambient
Variables and Conditions.

All the monthly averaged observations except the solar irradiance are supposed to be
made at the middle of each month. This time depends on the number of days in the
month:

• Months with 31 days (January, March, May, July, August, October, December):
data at the 16th at noon
• Months with 30 days (April, June, September, November): data at the 16th at
midnight
• Months with 29 days (February, leap years): data at the 15th at noon
• Months with 28 days (February, other years): data at the 15th at midnight

In addition, the temperature observations are supposed to be made at the beginning

of each hour (00:00 a.m. to 11:00 p.m.).

Finally, the solar irradiance observations are made at the 21st of each month at noon.
Depending on the number of days in the month, this date corresponds to 68% (for
months with 31 days), 70% (for months with 30 days), or 75% (for February) of the
month. The leap years are not considered and the 21st of February always corresponds
to 75% of this month.

Ambient Data Interpolation

The temperature is the only variable for which hourly data are available in addition to
monthly averages. So for the temperature a double interpolation is performed to get a

132 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

temperature profile for every time in a year depending on the date and the hour. For
other data the interpolation is based on date only. This interpolation provides:

• The annual fluctuation of the dew point temperature, the relative humidity, the
wind speed, and the direct and diffuse solar irradiances.
• The annual and diurnal fluctuation of the temperature.
In all cases, the interpolation is of second order, with continuous first-order derivative.

Ambient Variables and Conditions

The observed values are processed by computing weighted means, standard deviations,
maximum, and minimum, to define different conditions for the temperature, dew
point temperature, and wind speed.

CONDITIONS OF TEMPERATURE
• Average:
T amb =  T station

• Low:
T amb =  T station – σ T, station

• High:
T amb =  T station + σ T, station

• Lowest:
T amb = min ( T station )

• Highest:

T amb = max ( T station )

• User defined coefficient for deviation:

T amb =  T station + c σ ⋅ σ T, station

• User defined correction:

T amb =  T station + ΔT

AMBIENT PROPERTIES | 133

 where:

• <Tstation> (SI unit: K) is the weighted mean of the observed values of temperature
at the station.
• σT,station (SI unit: K) is the standard deviation of the observed values of
temperature at the station.
• Tstation (SI unit: K) is the set of the observed values of temperature at the station.
• cσ (dimensionless) is a user-defined multiplicative coefficient applied to σT,station.
• ΔT (SI unit: K) is a user-defined additive correction applied to <Tstation>.

All these conditions are illustrated on Figure 3-5 for the variation of temperature over
1 day at New York/John F. Ke, on the 1st of June.

Figure 3-5: Comparison of ambient conditions for the temperature at New York/John F.
Ke, on the 1st of June, with ASHRAE Weather Data Viewer 5.0 (©2013 ASHRAE,
www.ashrae.org. Used with permission.)

Additional conditions are defined from observed couples of temperature and wind
speed and direction values:

• Heating wind correlation:

T amb =  T station + ΔT wind

• Cooling wind correlation:

134 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 T amb =  T station – ΔT wind

where ΔTwind (SI unit: K) is an additive correction applied to <Tstation>, defined as

1
ΔT wind = --- max ( ΔT ws, station, ΔT wd, station )
2

where ΔTws,station (SI unit: K) and ΔTwd,station (SI unit: K) are respectively the
maximal variations of observed values of temperature correlated with a set of wind
speed and direction observed values.

The heating and cooling wind correlations are illustrated on Figure 3-6 for the
variation of temperature over 1 day, at New York/John F. Ke, on the 1st of June.

Figure 3-6: Comparison of heating and cooling wind correlations for the temperature at
New York/John F. Ke, on the 1st of June, with ASHRAE Weather Data Viewer 5.0
(©2013 ASHRAE, www.ashrae.org. Used with permission.)

CONDITIONS OF DEW POINT TEMPERATURE

• Average:
DPT amb =  DPT station

• Low:
DPT amb =  DPT station – σ DPT, station

• High:

AMBIENT PROPERTIES | 135

 DPT amb =  DPT station + σ DPT, station

• Lowest:
DPT amb = min ( DPT station )

• Highest:

DPT amb = max ( DPT station )

where:

• <DPTstation> (SI unit: K) is the weighted mean of the observed values of dew point
temperature at the station.
• σDPT,station (SI unit: K) is the standard deviation of the observed values of dew
point temperature at the station.
• DPTstation (SI unit: K) is the set of the observed values of dew point temperature
at the station.

All these conditions are illustrated on Figure 3-7 for the variation of the dew point
temperature over 1 year at New York/John F. Ke.

Figure 3-7: Comparison of the ambient conditions for the dew point temperature at New
York/John F. Ke, with ASHRAE Weather Data Viewer 5.0 (©2013 ASHRAE,
www.ashrae.org. Used with permission.)

136 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

CONDITIONS OF WIND SPEED
• Average:
v amb =  v station

• Low:
v amb =  v station – σ v, station

• High:
v amb =  v station + σ v, station

• Lowest:
v amb = min ( v station )

• Highest:

v amb = max ( v station )

where:

• <vstation> (SI unit: m/s) is the weighted mean of the observed values of wind
velocity at the station.
• σv,station (SI unit: m/s) is the standard deviation of the observed values of wind
velocity at the station.
• vstation (SI unit: m/s) is the set of the observed values of wind velocity at the
station.

AMBIENT PROPERTIES | 137

 All these conditions are illustrated on Figure 3-8 for the variation of the wind speed
over 1 year at New York/John F. Ke.

Figure 3-8: Comparison of the ambient conditions for the wind speed at New York/John
F. Ke, with ASHRAE Weather Data Viewer 5.0 (©2013 ASHRAE, www.ashrae.org.
Used with permission.)

CONDITIONS OF PRECIPITATION RATE

• Average:
P 0, amb =  P 0, station

• Low:
P 0, amb =  P 0, station – σ P
0, station

• High:
P 0, amb =  P 0, station + σ P
0, station

• Lowest:
P 0, amb = min ( P 0, station )

• Highest:

P 0, amb = max ( P 0, station )

138 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

where:

• <P0,station> (SI unit: m/s) is the weighted mean of the observed values of
precipitation rate at the station.
• σP0,station (SI unit: m/s) is the standard deviation of the observed values of
precipitation rate at the station.
• P0,station (SI unit: m/s) is the set of the observed values of precipitation rate at the
station.

PRESSURE
p amb = p station

where pstation (SI unit: Pa) is the observed value of absolute pressure at the station.
Only a single value is available, so this data does not vary with time.

RELATIVE HUMIDITY
The relative humidity φ amb (dimensionless) is computed from the temperature Tamb
and the dew point temperature DPTamb with the following relation:

p sat ( DPT amb )

φ amb = ---------------------------------------
p sat ( T amb )

where psat(Tamb) is the saturation pressure of vapor at Tamb.

See Relative Humidity for more details.

AMBIENT PROPERTIES | 139

 As the diurnal variation of temperature is available, the diurnal fluctuations of relative
humidity can be computed, as illustrated on Figure 3-9 for New York/John F. Ke, on
the 1st of June, for different ambient conditions.

Figure 3-9: Diurnal fluctuations of relative humidity for different ambient conditions at
New York/John F. Ke, on the 1st of June, with ASHRAE Weather Data Viewer 5.0
(©2013 ASHRAE, www.ashrae.org. Used with permission.)

Several conditions of relative humidity can be obtained from the selection of

conditions for the temperature and dew point temperature. For consistency, the
relative humidity is majored by 1, as shown on the red curve of Figure 3-9.

MOISTURE CONTENT
The moisture content xvap,amb (dimensionless) is computed from the temperature
Tamb, the absolute pressure pamb, and the relative humidity φ amb with the following
relation:

φ amb ⋅ p sat ( T amb ) Mv

x vap, amb = -------------------------------------------------------------------- ⋅ --------
p amb – φ amb ⋅ p sat ( T amb ) M a

where psat(Tamb) is the saturation pressure of vapor at Tamb, and Mv and Ma are the
molar masses of water vapor and dry air.

See Moisture Content for more details.

140 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

DIRECT AND DIFFUSE SOLAR IRRADIANCES
With clear sky conditions, the noon solar irradiance is essentially provided by the beam
normal irradiance, coming directly from the sun. However, the diffuse horizontal
irradiance may be also considered. The sum of direct and diffuse solar irradiances is the
ambient solar irradiance Is, amb (SI unit: W/m3), defined by:

I s, amb = I sn, station + I sh, station

where Isn,station (SI unit: W/m3) and Ish,station (SI unit: W/m3) are respectively the
observed values of the clear sky noon beam normal and horizontal diffuse solar
irradiances.

Figure 3-10 illustrates the evolution of ambient solar irradiance for New York/John F.
Ke, over the year.

Figure 3-10: Decomposition of solar irradiance into normal and horizontal irradiance at
New York/John F. Ke, with ASHRAE Weather Data Viewer 5.0 (©2013 ASHRAE,
www.ashrae.org. Used with permission.)

Use of Ambient Data From the Features

The ambient variables defined in the Ambient Properties node are available as inputs for
several boundary and initial conditions. For example, when defining the heat transfer
coefficient for a Heat Flux boundary condition by using a correlation for external forced

AMBIENT PROPERTIES | 141

 convection, the wind velocity vamb, the ambient absolute pressure pamb, and the
ambient temperature Tamb can be selected as inputs:

Figure 3-11: Use of ambient data from Heat Flux feature

For an example of the use of user-defined ambient data, see:

Temperature Field in a Cooling Flange: Application Library path

Heat_Transfer_Module/Thermal_Processing/cooling_flange

For an example of the use of meteorological ambient data, see:

Condensation Detection in an Electronic Device with Transport and

Diffusion: Application Library path
Heat_Transfer_Module/Power_Electronics_and_Electronic_Cooling/condensati
on_electronic_device_transport_diffusion

Condensation Detection in an Electronic Device: Application Library

path
Heat_Transfer_Module/Power_Electronics_and_Electronic_Cooling/condensati
on_electronic_device

142 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Modeling Heat Transfer by Radiation
This section summarizes the functionalities available within the Heat Transfer Module
for the modeling of heat transfer by radiation.

Radiation is a heat transfer mode that should be considered in addition to conduction

and convection in the following cases:

• In a transparent or semitransparent medium.

• For a surface with a large emissivity (close to 1). Whereas most polished metals have
an emissivity below 0.1, other materials such as concrete, black painting, glass,
plastic, skin, paper, and ceramic have an emissivity over 0.8 for infrared radiation.
• When the temperature difference between a surface and the surrounding surfaces or
the ambient is large.
• When conduction and convection are small, as radiation becomes the predominant
transfer mode.
• At small time scales, for which conduction and convection effects are too slow and
have a less significant effect.

As an example, consider a plate of concrete with a cold bottom wall at temperature T0,
placed in a hot environment with an air flux at temperature Tamb:

MODELING HEAT TRANSFER BY RADIATION | 143

 The figures below illustrate the relative importance of heat transfer by radiation and by
convection (due to the air flux), when the emissivity ε is 0.1 (left) and 0.9 (right).

Figure 3-12: Integral of convective and radiative heat fluxes (W/m) along the top
boundary, for two values of emissivity, as a function of the temperature difference.

Whereas the boundary radiative and convective heat fluxes are of the same order for
temperature difference up to 500K when the emissivity is low (ε=0.1), radiation
becomes the dominant mode of heat transfer even for a small temperature difference
when the emissivity is high (ε=0.9). Note that the convective heat flux decrease
observed for high temperature gradients is related to the fact that the velocity, not the
mass flow rate, is prescribed at the air inlet.

See Heat Transfer Variables for the definition of the variables q0 and
rflux giving access to the convective and radiative heat fluxes on
boundaries.

Two configurations are considered regarding the functionalities available for the
modeling of radiative heat transfer:

• Radiation in transparent media

• Radiation in participating media

144 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Interfaces and Features for Radiation in Transparent Media
The following table recalls the functionalities available for the modeling of radiative
heat transfer in transparent media.
TABLE 3-15: MODELING OF HEAT TRANSFER BY RADIATION IN TRANSPARENT MEDIA

SURFACE-TO-AMBIENT SURFACE-TO-SURFACE

Description

Physical and modeling assumptions

Refractive index of n=1 Any, uniform
domain
Diffuse reflection Yes Yes
Specular reflection No With Ray shooting
Transmission No With Ray shooting
Shadowing effects in No With Hemicube or Ray shooting
view factor
computation
Wavelength Yes (Solar and ambient / Yes (Solar and ambient / Multiple
dependence of surface Multiple spectral bands) spectral bands)
properties No (Constant) No (Constant)
Directional No With Ray shooting
dependence of surface
properties
Dimension of the 3D, 2D axi, 2D, 1D axi, 3D, 2D axi, 2D, 1D axi, 1D
domain 1D
Example in Application Continuous Casting Chip Cooling
Library
License, interfaces and features
License requirement COMSOL Multiphysics / Heat Transfer Module
Heat Transfer Module
Interfaces Heat Transfer (ht) Surface-to-Surface Radiation
Heat Transfer in Shells (rad)
(htlsh)
Domain features Out-of-Plane Radiation Opacity
(1)(2)

MODELING HEAT TRANSFER BY RADIATION | 145

 TABLE 3-15: MODELING OF HEAT TRANSFER BY RADIATION IN TRANSPARENT MEDIA

SURFACE-TO-AMBIENT SURFACE-TO-SURFACE

Boundary features Surface-to-Ambient Diffuse Mirror

Radiation Diffuse Surface
Opaque Surface
Semitransparent Surface
Initial Values
Prescribed Radiosity
Radiation Group
Edges/Point features Surface-to-Ambient N/A
Radiation(1)
Global features N/A External Radiation Source (3)
Symmetry for Surface-to-Surface
Radiation (4)
Multiphysics features N/A Heat Transfer with
Surface-to-Surface Radiation
Numerical aspects
Numerical cost Low Medium to high
Radiation variables None Surface radiosity J (if constant
wavelength), J1,J2 (if solar and
ambient wavelengths),
J1,...,Jn (if multiple wavelengths)
(1)
Only with the Heat Transfer Module
(2)
Only in 2D and 1D components
(3) Only in 3D and 2D components
(4)
Only in 3D, 2D axi, and 2D components
(5)
Only with Rosseland approximation and P1 method

146 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Interfaces and Features for Radiation in Participating Media
The following table recalls the functionalities available for the modeling of radiative
heat transfer in participating media.
TABLE 3-16: MODELING OF HEAT TRANSFER BY RADIATION

ABSORBING MEDIA ABSORBING-SCATTERING PARTICIPATING MEDIA

MEDIA

Description

Physical and modeling assumptions

Refractive Any, uniform Any, uniform Any, uniform
index of
domain
Scattering in N/A Isotropic (P1 method, Isotropic (Rosseland
domain DOM) approximation,P1
Linear anisotropic (P1 method, DOM)
method, DOM) Linear anisotropic (P1
Anisotropic (DOM) method, DOM)
Henyey-Greenstein Anisotropic (DOM)
(DOM) Henyey-Greenstein
(DOM)
Wavelength No Yes (Solar and ambient Yes (Solar and ambient
dependency / Multiple spectral / Multiple spectral
of emissivity bands) bands)
No (Constant)
(5) (5)
Dimension of 3D, 2D axi , 3D, 2D axi , 2D 3D, 2D axi(5), 2D
the domain 2D
Example N/A N/A Glass Plate
model
License, interfaces and features
License Heat Transfer Heat Transfer Module Heat Transfer Module
requirement Module

MODELING HEAT TRANSFER BY RADIATION | 147

 TABLE 3-16: MODELING OF HEAT TRANSFER BY RADIATION

ABSORBING MEDIA ABSORBING-SCATTERING PARTICIPATING MEDIA

MEDIA

Interfaces Radiative Beam Radiation in Radiation in

in Absorbing Absorbing-Scattering Participating Media
Media (rbam) Media (rasm) (rpm)
Heat Transfer (ht):
Rosseland
approximation
Domain Absorbing Absorbing-Scattering Participating Medium,
features Medium, Initial Medium, Initial Values, Initial Values, Radiative
Values Radiative Source Source
Optically Thick
Participating Medium
Boundary Opaque Surface Opaque Surface Opaque Surface
features Incident Incident Intensity Semitransparent
Intensity Continuity of Interior Surface
Transparent Boundary (DOM) Incident Intensity
Surface Periodic Condition Continuity of Interior
Symmetry (P1 Boundary (DOM)
method) Periodic Condition
Symmetry (P1
method)
Multiphysics Heat Transfer Heat Transfer with Heat Transfer with
features with Radiative Radiation in Radiation in
Beam in Absorbing-Scattering Participating Media
Absorbing Media (htrasm) (htrpm)
Media (htrbam)
Numerical aspects
Numerical Medium Medium (P1 method) Low (Rosseland
cost High (DOM) approximation)
Medium (P1 method)
High (DOM)

148 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

TABLE 3-16: MODELING OF HEAT TRANSFER BY RADIATION

ABSORBING MEDIA ABSORBING-SCATTERING PARTICIPATING MEDIA

MEDIA

Radiation Incident Incident radiation G None (Rosseland

variables intensities Ii (P1 method) approximation)
Radiative intensity Incident radiation G
I=I1,..., In, with n up (P1 method)
to 512 (DOM) Radiative intensity
I=I1,..., In, with n up
to 512 (DOM)
(1)
Only with the Heat Transfer Module
(2) Only in 2D and 1D components
(3)
Only in 3D and 2D components
(4)
Only in 3D, 2D axi, and 2D components
(5) Only with Rosseland approximation and P1 method

MODELING HEAT TRANSFER BY RADIATION | 149

 Specifying the Temperature at a Flow
Inlet
When modeling a nonisothermal flow, a Temperature condition can be used at a flow
inlet where the temperature of the fluid is a known quantity. However, there are some
other situations where this is not the case, and an Inflow boundary condition can
improve the model accuracy and reduce the computational cost of the simulation.

In this section:

• Using the Temperature Condition

• Using the Inflow Condition
• Temperature Contribution to the Inflow Boundary Condition
• Pressure Contribution to the Inflow Boundary Condition
• Choosing Between the Temperature and the Inflow Conditions

Using the Temperature Condition

The Temperature condition is available at exterior and interior boundaries of any type
of domain and is equivalent to set the temperature to a known value on the selected
boundary.

Using the Inflow Condition

The Inflow condition is available at boundaries that are exterior to a fluid domain and
is equivalent to having a virtual channel “upstream”. It is used to define a heat flux at
the inlet boundary that brings the same energy to the fluid domain as if you had
modeled the virtual channel as a real CFD domain. The virtual channel can be seen as

150 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

a long insulated channel with given thermal properties at the inlet, and with the same
velocity profile as defined in the settings for the Inflow boundary condition.

Figure 3-13: Representation of the virtual domain corresponding to an Inflow boundary

condition.

This is of particular interest when the geometry contains inlets that are fed by channels
that are not represented in the geometry.

From a mathematical point of view, the boundary condition is formulated as a flux

condition:

– n ⋅ q = ρΔHu ⋅ n

where the enthalpy variation ΔH depends on temperature and pressure difference:

T p
1
ΔH = Tustr Cp dT + pustr --ρ- ( 1 – αp T ) dp = ΔHT + ΔHp (3-1)

where:
T
ΔH T = Tustr Cp dT (3-2)

and
p
1
ΔH p = pustr --ρ- ( 1 – αp T ) dp (3-3)

Equation 3-1 expresses the fact that the normal conductive heat flux at the inflow
boundary is proportional to the flow rate and enthalpy variation between the upstream
conditions and inlet conditions.

SPECIFYING THE TEMPERATURE AT A FLOW INLET | 151

 Temperature Contribution to the Inflow Boundary Condition
As shown in Equation 3-1, the enthalpy variation depends in general both on the
difference in temperature and in pressure. However, the pressure contribution to the
enthalpy, ΔHp, can be neglected when the work due to pressure changes is not
included in the energy equation. This is controlled by the Nonisothermal Flow
multiphysics coupling depending on compressibility assumption.

There is another classical case where this term cancels out: when the fluid is modeled
as an ideal gas. Indeed, in this case,

1
α p = ----
T

When the pressure contribution to the enthalpy is neglected, the boundary condition
reads:

 T 
k∇T ⋅ n = ρ 
 Tustr Cp dT u ⋅ n (3-4)

When advective heat transfer dominates at the inlet (large flow rates), the temperature
gradient, and hence the heat transfer by conduction, in the normal direction to the
inlet boundary is very small. So in this case, Equation 3-4 imposes that the enthalpy
variation is close to zero. As Cp is positive, the Inflow boundary condition requires
T=Tustr to be fulfilled. So, when advective heat transfer dominates at the inlet, the
Inflow boundary condition is almost equivalent to a Dirichlet boundary condition that
prescribes the upstream temperature at the inlet.

Conversely, when the flow rate is low or in the presence of large heat sources or sinks
next to the inlet, the conductive heat flux cannot be neglected. In addition, the inlet
temperature has to be adjusted to balance the energy brought by the flow at the inlet
and the energy transferred by conduction from the interior, as described by
Equation 3-4. This makes it possible to observe a realistic upstream feedback due to
thermal conduction from the inlet surroundings.

Pressure Contribution to the Inflow Boundary Condition

In the general case, the enthalpy variation depends both on the difference in
temperature and in pressure.

152 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Considering that the Inflow boundary condition models a virtual channel feeding the
inlet, pressure losses are expected between the virtual channel inlet and the boundary
where the condition is defined. This explains why the upstream pressure is different
from the inlet pressure. While the fluid flows through the channel, it is subject to
pressure work that results in a temperature change between the virtual channel inlet
and the boundary where the Inflow boundary condition is defined. This is what is
described by the pressure-dependent term in Equation 3-1. Note that the viscous
dissipation in the virtual channel is not accounted for.

Choosing Between the Temperature and the Inflow Conditions

While in many configurations, the Temperature and Inflow features describe similar
conditions and lead to similar simulation results, there are a number of configurations
(especially for slow flow and small dimensions) where the conductive effects are not
dominated by the advective effects and where the Inflow boundary condition usually
leads to a temperature profile that is closer to the reality than a Temperature boundary
condition. In these cases, a Temperature boundary condition may enforce an erroneous
temperature value that induces large heat fluxes that are not realistic.

SPECIFYING THE TEMPERATURE AT A FLOW INLET | 153

 Heat Part Library
The Part Libraries can be used to store and access a set of standard geometric entities.
It is possible to create custom user-defined geometry libraries or to use built-in
collections of geometric entities that are available with many add-on modules for
COMSOL Multiphysics.

Part Libraries in the COMSOL Multiphysics Reference Manual

In this section:

• Heat Part Library Contents

• Heat Sink — Parameterized Fin Types Part
• Heat Sink — Dissimilar Border Pins Part
• Heat Sink — Pin Fins Part
• Heat Sink — Straight Fins Part

Heat Part Library Contents

The part library included with the Heat Transfer Module contains typical heat sink
geometries. When accessing the library, the parts are sorted as in the figure below:

The heat sinks are composed of a rectangular base and an array of pin or straight fins,
as shown on Figure 3-14.

154 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 fins

border offset

base

Figure 3-14: Heat sinks components

All entities are fully parameterized, making them easy to use as parts in industrial
models where heat sinks are used for cooling. For example, you can control the
number, the shape, the dimensions, and the placement of the fins on the base. In
addition, fillet, chamfer, and notch transformations can be applied to the fins, and
parameter checks are applied to ensure that the values set in the Input Parameters
section are valid. Finally, the fins can be defined as solids or as boundaries for
computational efficiency.

HEAT PART LIBRARY | 155

 The library includes four parts. The Heat Sink — Parameterized Fin Types part is the
most general part. It supports multiple parameterization, while a reduced number of
parameters is available in the three other parts, as shown in Table 3-17.
TABLE 3-17: OVERVIEW OF HEAT SINK PARTS

Heat Sink — Parameterized Fin Types

fin_type =1,2,3 (editable)

Heat Sink — Pin Fins Heat Sink — Dissimilar Heat Sink — Straight Fins
Border Pins

fin_type = 1 (not editable) fin_type = 2 (not editable) fin_type = 3 (not editable)

In the Heat Sink — Pin Fins part, all the fins are pins with the same dimension, whereas
the outer and inner fins (in the y direction) can have distinct dimensions in the Heat
Sink — Dissimilar Border Pins part. You may use the Heat Sink — Straight Fins part to
define a heat sink made of only straight fins.

By default, the base of the heat sink is positioned at the origin of the xy-plane. You can
apply a Displacement and a Rotation to this configuration in the Position and Orientation
of Output section.

156 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

Finally, each fin is a separate domain and can be easily deleted, and selections can be
defined, for example for the base bottom face (Boundary Selections section) or the array
of fins (Domain Selections section).

Heat Sink — Parameterized Fin Types Part

The dimensions of the base and the fins can be parameterized, as shown on
Figure 3-15. In addition, an offset from the base border can be added.

n_fins_x=3
n_fins_y=4 Y_fins_top_2
Y_fins_top
X_fins_top

Z_fins

Z_base o_y
o_x
Y_fins_bottom_2

Y_fins_bottom X_fins_bottom

Y_base X_base

Figure 3-15: Fins and base parameters in Heat Sink — Parameterized part

3D AND SHELL FINS

When the thickness of the fins is small in the x direction, the fins can be defined as
boundaries instead of solids in order to reduce the size of the model. Indeed, small
elements are required to mesh the fins across their thickness in this case. It is done by
setting the shell parameter to 1. The default meshes obtained with shell = 0 (left)

HEAT PART LIBRARY | 157

 and shell=1 (right) are shown in Table 3-18 for an example configuration. Setting
shell to 1 significantly reduces the number of mesh elements, as shown below.

TABLE 3-18: DEFAULT TETRAHEDRAL MESH WITH 3D FINS AND SHELL FINS

SHELL=0 SHELL=1

Mesh statistics: Mesh statistics:

8003 tetrahedra 906 tetrahedra
5900 triangles 1058 triangles
1220 edge elements 312 edge elements
104 vertex elements 56 vertex elements

Note that when shell=1, the X_fins_bottom, Y_fins_bottom, X_fins_top, and

Y_fins_top parameters are not used.

STEP
A step can be defined in the x direction at the center on the base. The parameter
step_width specifies the number of filled gaps from middle to border by the step, as
shown on Figure 3-16. This option is not available for shell fins (shell=1).

158 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

 step=1

step_height step_height
step_width=1 step_width=1

(odd number of fins in x direction) (even number of fins in x direction)

Figure 3-16: Step parameters in Heat Sink — Parameterized part

FILLET
Finally, a fillet transformation can be applied at the top and bottom of the fins, as
shown on Figure 3-17. The fillet transformation at the bottom of the fins is not
available for shell fins (shell=1).

fillet_top=1

fillet_bottom=1

Figure 3-17: Fillet parameters in Heat Sink — Parameterized part

HEAT PART LIBRARY | 159

 Heat Sink — Dissimilar Border Pins Part
This part inherits all the functionalities of the Heat Sink — Parameterized Fin Types part,
but different dimensions are set by default for the inner and outer fins in the
y direction:

• Y_fins_bottom = 3 mm (outer fins) and Y_fins_bottom_2 = 6 mm (inner fins)

• Y_fins_top = 3 mm (outer fins) and Y_fins_top_2 = 6 mm (inner fins)

Heat Sink — Pin Fins Part

This part inherits all the functionalities of the Heat Sink — Parameterized Fin Types part,
but identical dimensions are enforced for the inner and outer fins in the y direction:

• Y_fins_bottom = 3 mm (all fins)

• Y_fins_top = 3 mm (all fins)

Heat Sink — Straight Fins Part

This part inherits all the functionalities of the Heat Sink — Parameterized Fin Types part,
but the y direction parameters are enforced to define straight fins.

In addition, notch and chamfer transformations can be applied to the fins, as shown
on Figure 3-18.

notch=1 chamfer=1

notch_height chamfer_height
notch_width chamfer_width

Figure 3-18: Notch and chamfer parameters in Heat Sink — Straight Fins part

160 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE

R e f e r e nc e s
1. J.L. Monteith and M.H. Unsworth, Principles of Environmental Physics, Edward
Arnold, London, 290 pp., 1990.

REFERENCES | 161
162 | CHAPTER 3: MODELING WITH THE HEAT TRANSFER MODULE
 4

Theory for the Heat Transfer Module

This chapter details the theory of the physics interfaces, multiphysics couplings,
and features found under the Heat Transfer branch ( ).

In this chapter:

• Foundations of the General Heat Transfer Equation

• Theory for Heat Transfer in Solids
• Theory for Heat Transfer in Fluids
• Theory for Heat Transfer in Moist Air
• Theory for Bioheat Transfer
• Theory for Heat Transfer in Porous Media
• Theory for Heat Transfer with Phase Change
• Theory for Heat Transfer in Building Materials
• Theory for Harmonic Heat Transfer
• Theory for Lumped Isothermal Domain
• Theory for Heat Transfer in Thin Structures
• Theory for the Lumped Thermal System Interface
• Theory for Surface-to-Surface Radiation

163
 • Theory for Radiation in Participating Media
• Theory for Moisture Transport
• Theory for the Heat Transfer Multiphysics Couplings
• Theory for Thermal Contact
• Out-of-Plane Heat Transfer
• Convective Heat Transfer Correlations
• Nucleate Pool Boiling Correlation
• Equivalent Thermal Conductivity Correlations
• Temperature Dependence of Surface Tension
• Heat Flux and Heat Balance
• Frames for the Heat Transfer Equations
• References

164 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Foundations of the General Heat
T r a ns f e r E qua ti on
This section presents basic results leading to the heat transfer equations solved in
COMSOL Multiphysics. Starting by a presentation of heat as another mode of energy
transfer different from work, the energy conservation laws that apply lead to the
establishment of The Heat Balance Equation. The latter is further derived to give the
different versions of the heat transfer equations in various media.

In this section:

• Thermodynamic Description of Heat Transfer

• The Physical Mechanisms Underlying Heat Transfer
• The Heat Balance Equation
• Consistency with Mass and Momentum Conservation Laws

Thermodynamic Description of Heat Transfer

In continuum mechanics, a domain Ω is submitted to variations of its kinetic energy
due to some external forces according to an equation of motion. The study of such
phenomena is covered by solid mechanics and fluid mechanics and the theories behind
can be found in the Structural Mechanics Module User’s Guide and CFD Module
User’s Guide. From an energy point of view, the aforementioned description is
incomplete because it does not include heat as another form of energy transfer due to
microscopic vibration and interactions of particles. The laws of thermodynamics
introduce several concepts to define heat transfer consistently with mechanical energy.
In the next paragraphs, a concise presentation of the theory adapted to the use of
COMSOL Multiphysics is given. More materials and details are provided in the
references listed in the References section.

EXTENSIVE PARAMETERS CHARACTERIZING A SYSTEM

A homogeneous fluid taking place in a domain Ω is characterized by the knowledge of
three extensive parameters:

• The entropy, SΩ (SI unit: J),

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 165

 • The volume, VΩ (SI unit: m3),
• The mass, MΩ (SI unit: kg).

The internal energy, EΩ (SI unit: J), is an extensive state function of these three
variables. It measures the amount of energy in the system excluding kinetic energy and
potential energy from external applied forces and is the subject of conservation laws
more detailed in The Heat Balance Equation section. To fit with the finite element
method solved by COMSOL Multiphysics, specific quantities per unit mass are
preferred:

SΩ VΩ
S = --------- ν = ---------
MΩ MΩ

The specific internal energy, E (SI unit: J/kg), is then a function of specific entropy,
S, and specific volume, ν, related to EΩ by:

1
E ( S, ν ) = --------- E Ω ( S Ω, V Ω, M Ω )
MΩ

For a solid, the specific internal energy, E(S, F), is a function of entropy and
deformation gradient, F.

Internal energy is related to the enthalpy, H, via the following for a fluid:

p
H = E + ---
ρ

or the following for a solid (7.33 in Ref. 1):

1
H = E – --------------------- P:F
det(F)ρ

Compared to the internal energy, the enthalpy also includes the pressure-volume
potential energy, p ⁄ ρ, necessary for instance in volume expansion after an isobaric
transformation.

FIRST-ORDER PARAMETERS
The variations of internal energy correspond to variations of entropy and volume
according to:

∂E ∂E
dE =   dS +   dν
∂S ν ∂ν S

166 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

First-order parameters are partial derivatives of the specific internal energy. They
correspond to the thermodynamic definitions of temperature and pressure:

∂E ∂E
T =   p = –  (4-1)
 ∂ S ν ∂ν  S

These lead to the fundamental thermodynamic relation:

dE = T dS – p dν

Temperature is the measurable quantity that gives a phenomenological

description of heat transfer. When expressed in kelvin (K), the Second
Law of Thermodynamics ensures that T can only take positive values.

Similar relations as those of Equation 4-1 hold for solids:

∂E ∂E
T =   P = det(F)ρ   (4-2)
∂S F ∂F S

1
dE = T dS + --------------------- P: dF
det(F)ρ

Here, the counterpart of the fluid pressure is the first Piola-Kirchhoff stress tensor, P.

SECOND-ORDER PARAMETERS
Second-order parameters correspond to second partial derivatives of the specific
internal energy and provide a various number of thermodynamic coefficients. These
are usually given as material properties of the domain material. Among them, the heat
capacity at constant pressure and the coefficient of thermal expansion are most often
provided. For a fluid, these are

T 1
C p = --------------- α p = ------------------- (4-3)
 ∂T ∂T
ν  
 ∂ S ν ∂ν S

and for a solid, the definitions become:

–1
– 1 ∂T
α = F  
T
C p = ---------------- (4-4)
∂T ∂F S
 
 ∂ S F

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 167

 Specific heat capacity at constant pressure is the amount of energy required to raise one
unit of mass of a substance by one degree while maintained at constant pressure. This
quantity is also commonly referred to as specific heat or specific heat capacity.

The heat capacity at constant pressure and coefficient of thermal expansion are related
to the enthalpy, seen as a function of T and p (or P), according to:

 ∂H = C  ∂H = ν ( 1 – α T )
∂T p p ∂p  T p

 ∂H = C  ∂H = --------------------

1 -
F ( – I + αT )
 ∂T P p  ∂ P T det(F)ρ

SENSIBLE ENTHALPY
The enthalpy can then be retrieved from Cp and αp (or α) by:

r1
H = H ref + r ∇r H ( r ) ⋅ dr
0
(4-5)

where r is the integration vector variable, containing temperature and pressure or

stress tensor components:

P
 11
P 
 22
P 
 33
r =  
p
or r = P 
 T  12
P 
 23
P 
 13
 T 

The starting point, r0, is the value of r at reference conditions, that is, pref (1 atm) and
Tref (293.15 K) for a fluid. The ending point, r1, is the solution returned after
simulation. In theory any value can be assigned to the enthalpy at reference conditions,
Href (Ref. 2), and COMSOL Multiphysics sets it to 0 J/kg by default. The integral in

168 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Equation 4-5 is sometimes referred to as the sensible enthalpy (Ref. 2) and is evaluated
by numerical integration.

For the evaluation of H to work, it is important that the dependencies of

Cp, ρ, and γ on the temperature are prescribed either via Model Inputs or
as functions of the temperature variable. If Cp, ρ, or γ depends on the
pressure, that dependency must be prescribed either via a model input or
by using the variable pA, which is the variable for the absolute pressure in
COMSOL Multiphysics.

Sensible Enthalpy in Ideal Gas

In an ideal gas,

1
α p = ----
T

therefore

 ∂H = 0
∂p  T

and Equation 4-5 reduces to

T
H = H ref + Tref Cp ⋅ dT (4-6)

Sensible Enthalpy in Moist Air

The ideal gas assumption also holds for moist air, so

 ∂H = 0
∂p  T

In addition, by using a mixture formula, we have

 ∂H = ( C ∂ω v
p, v – C p, a ) ---------
-
∂T p ∂c v

and Equation 4-6 should be modified to include an integral over the vapor
concentration:
cv
T ∂ω v ∂H
H = H ref + Tref ( C p, v – C p, a ) ---------- ⋅ dT +
∂c v cref ∂ cv ⋅ dcv (4-7)

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 169

 The Physical Mechanisms Underlying Heat Transfer
The amount of heat transferred per unit time (heat transfer rate) depends on the
underlying physical mechanisms that define the mode of transfer. These are:

• Conduction — Heat conduction (or diffusion) occurs as a consequence of different

mechanisms in different media. Theoretically, it takes place in a gas through
collisions of molecules; in a fluid through oscillations of each molecule in a “cage”
formed by its nearest neighbors; in metals mainly by electrons carrying heat and in
other solids by molecular motion, which in crystals take the form of lattice vibrations
known as phonons.
In a continuous medium, Fourier’s law of heat conduction states that the
conductive heat flux, q, is proportional to the temperature gradient:
q = – k∇T (4-8)

The coefficient of proportionality, k, is the thermal conductivity (SI unit: W/(m·K))

and takes a positive value meaning that heat flows from regions of high temperature
to low temperature. More generally, the thermal conductivity can take the form of
a symmetric positive-definite second-order tensor (matrix) in anisotropic media
such as composite materials:

k xx k xy k xz
k = k xy k yy k yz
k xz k yz k zz

Thermal conductivity tensors that do not respect the symmetric

positive-definite property lead to unphysical results (Ref. 3).

• Convection — Heat convection (sometimes called heat advection) takes place

through the net displacement of a fluid that transports the heat content with its
velocity. The term convection (especially convective cooling and convective heating)
also refers to the heat dissipation from a solid surface to a fluid, typically described
by a heat transfer coefficient.
• Radiation — Heat transfer by radiation takes place through the transport of
photons. Participating (or semitransparent) media absorb, emit, and scatter
photons. Opaque surfaces absorb or reflect them.

170 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The Heat Balance Equation
The equations of heat transfer in continua are derived from the first law of
thermodynamics, commonly referred to as the principle of conservation of energy. The
present part establishes the heat balance equation in its integral and localized forms
that stand as a root for deriving the different heat transfer equations solved in
COMSOL Multiphysics.

INTEGRAL FORM
The first law of thermodynamics states that the variations of macroscopic kinetic
energy, KΩ, and internal energy, EΩ, of a domain Ω are caused either by the mechanical
power of forces applied to the system, Pext, or by exchanged heat rate, Qexch (2.3.53
in Ref. 4):

dE Ω dK Ω
----------- + ----------- = P ext + Q exch (4-9)
dt dt

Mass and momentum balance are needed to complete the description of the system.
The mechanical laws, either for solids or fluids, generate the following balance
equation between variation of kinetic energy, KΩ, stress power, Pstr, and power of
applied forces, Pext (2.3.64 in Ref. 4):

dK Ω
----------- + P str = P ext (4-10)
dt

This equation involves quantities of the macroscopic level where the variation of the
kinetic energy due to some forces applied to it reflects a sensible displacement. In
COMSOL Multiphysics, the Solid Mechanics or Single-Phase Flow interfaces are
examples of physics interfaces that simulate the macroscopic level described by
Equation 4-10.

Combining Equation 4-9 and Equation 4-10 yields the so-called heat balance
equation (2.3.65 in Ref. 4):

dE Ω
----------- = P str + Q exch (4-11)
dt

This time, the equation involves quantities of the microscopic level (exchanged heat
rate, Qexch, and internal energy, EΩ) more concerned with the atomic vibrations and
similar microscopic phenomena that are felt as heat. The presence of the stress power,
Pstr, in both Equation 4-10 and Equation 4-11 stands for the fact that such power is

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 171

 converted into heat by dissipation. The Heat Transfer interfaces, described in the next
sections, simulate the heat exchanges described by Equation 4-11.

LOCALIZED FORM
In this paragraph, the different terms of Equation 4-11 are more detailed to obtain the
localized form of the heat balance equation.

Variation of Internal Energy

The equations given in the previous paragraph holds for a given macroscopic
continuous domain Ω where the internal energy is defined using the specific internal
energy (per unit mass), E, as:

EΩ = Ω E dm
Note that by conservation of mass, the variation of internal energy in time is:

dE Ω dE dE
----------- =
dt Ω d t dm = Ω ρ d t dv
In these last relations, ρ is the density, and dv denotes an elementary volume of Ω.
Contrary to the constant elementary mass, dm, the elementary volume changes by
expansion or contraction of the domain. Recall that the derivation operator d ⁄ dt
under the integrals is in the material frame (see Time Derivative in the Frames for the
Heat Transfer Equations section).

Stress Power
The stress power, derived from the Continuum Mechanics theory, is defined by
(2.3.59 in Ref. 4):

P str = Ω ( σ:D ) dv
where σ is the Cauchy stress tensor and D is the strain rate tensor. The operation “:”
is a contraction and can in this case be written on the following form:

a:b =   anm bnm

n m

Note that in fluid mechanics, the Cauchy stress tensor is divided into a static part for
the pressure, p, and a symmetric deviatoric part, τ, as in:

172 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 σ = – pI + τ (4-12)

so that Pstr becomes the following sum of pressure-volume work and viscous
dissipation:

P str = Ω p ( ∇ ⋅ u ) dv – Ω ( τ: ∇u ) dv
Equivalently, the stress power can also be expressed as:

-----------------  P:  dv
1 dF
P str = Ω det (F)  d t 

Exchanged Heat
Finally, the exchanged heat rates, Qexch, account for thermal conduction (see Fourier’s
Law at Equation 4-8), radiation and potentially additional heat sources. Joule heating
and exothermic chemical reactions are such examples of domain heat source. The
different kinds of exchanged heat are summarized by the equality below:

Q exch = – ∂Ω ( q ⋅ n ) ds – ∂Ω ( qr ⋅ n ) ds + Ω Q dv

Recall the following notations used above: q for the heat flux by conduction, qr for
the heat flux by radiation, Q for additional heat sources, and n for the external normal
vector to the boundary ∂Ω.

Localized Heat Balance Equation

With all these elements, the heat balance equation (Equation 4-11) becomes:

dE
Ω ρ d t dv + ∂Ω ( q ⋅ n ) ds + ∂Ω ( qr ⋅ n ) ds = Ω ( σ:D ) dv + Ω Q dv (4-13)

which leads to the following localized form in the material frame:

dE
ρ + ∇ ⋅ ( q + q r ) = σ:D + Q (4-14)
dt

or equivalently in the spatial frame:

∂E
ρ + ρu ⋅ ∇E + ∇ ⋅ ( q + q r ) = σ:D + Q (4-15)
∂t

This verbally means that variations of internal energy in time are balanced by
convection of internal energy, thermal conduction, radiation, dissipation of mechanical

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 173

 stress and additional volumetric heat sources. In the next sections, Equation 4-15 will
be derived to obtain the heat transfer equations in different media.

See Frames for the Heat Transfer Equations for more details about the use
of material and spatial frames in the Heat Transfer interfaces.

Consistency with Mass and Momentum Conservation Laws

Although the heat transfer interfaces only solve for the energy equation, the context
leading to Equation 4-15 does account for the three additional conservation laws that
complete the Continuum Mechanics theory:

• Conservation of mass
• Conservation of linear momentum
• Conservation of angular momentum

The equations corresponding to each of them are recalled below in Table 4-1. For
more details about the theory of Solid and Fluid Mechanics, see the Structural
Mechanics Module User’s Guide and CFD Module User’s Guide.
TABLE 4-1: CONSERVATION OF MASS AND MOMENTUM

CONSERVATION LAW MATHEMATICAL MATHEMATICAL EXPRESSION IN SPATIAL

EXPRESSION IN FRAME
MATERIAL FRAME

Conservation of ρ 0 = ρdet(F) ∂ρ
Mass + ∇ ⋅ ( ρu ) = 0
∂t
Conservation of du ∂u
Linear Momentum ρ = ∇ ⋅ σ + Fv ρ + ρ ( u ⋅ ∇ )u = ∇ ⋅ σ + F v
dt ∂t
Conservation of T T
σ = σ σ = σ
Angular Momentum

When modeling a heat transfer problem with one of the Heat Transfer interfaces, the
aforementioned laws needs to be respected. For example, the velocity field, u,
provided in the energy equation and responsible for convection in a fluid, should
satisfy the continuity equation here below in order to avoid unphysical results.

∂ρ
+ ∇ ⋅ ( ρu ) = 0
∂t

174 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

This is especially true when the velocity field is not computed from an additional
physics interface — such as Single-Phase Flow that solves for the continuity equation
— but instead defined by a custom expression in a Heat Transfer interface.

FOUNDATIONS OF THE GENERAL HEAT TRANSFER EQUATION | 175

 Theory for Heat Transfer in Solids
The Heat Transfer in Solids Interface solves for the following equation derived from
Equation 4-15:

∂T
ρC p  ------- + u trans ⋅ ∇ T + ∇ ⋅ ( q + q r ) = – αT: + Q
dS
(4-16)
 ∂t  dt

The different quantities involved here are recalled below:

• ρ is the density (SI unit: kg/m3)

• Cp is the specific heat capacity at constant stress (SI unit: J/(kg·K))
• T is the absolute temperature (SI unit: K)
• utrans is the velocity vector of translational motion (SI unit: m/s)
• q is the heat flux by conduction (SI unit: W/m2)
• qr is the heat flux by radiation (SI unit: W/m2)
• α is the coefficient of thermal expansion (SI unit: 1/K)
• S is the second Piola-Kirchhoff stress tensor (SI unit: Pa)
• Q contains additional heat sources (SI unit: W/m3)

For a steady-state problem the temperature does not change with time and the terms
with time derivatives disappear.

The first term on the right-hand side of Equation 4-16 is the thermoelastic damping
and accounts for thermoelastic effects in solids:

dS
Q ted = – αT: (4-17)
dt

It should be noted that the d ⁄ dt operator is the material derivative, as described in the
Time Derivative subsection of Material and Spatial Frames.

176 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Heat Transfer in Fluids
The Heat Transfer in Fluids Interface solves for the following equation (11.2-5 in
Ref. 5):

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ ( q + q r ) = α p T  + u ⋅ ∇p + τ: ∇u + Q (4-18)
 ∂t  ∂t 

which is derived from Equation 4-15, considering that:

• the Cauchy stress tensor, σ, is split into static and deviatoric parts as in:
σ = – pI + τ

• the dependent variables are the temperature, T, and pressure, p.

The different quantities involved here are recalled below:

• ρ is the density (SI unit: kg/m3)

• Cp is the specific heat capacity at constant pressure (SI unit: J/(kg·K))
• T is the absolute temperature (SI unit: K)
• u is the velocity vector (SI unit: m/s)
• q is the heat flux by conduction (SI unit: W/m2)
• qr is the heat flux by radiation (SI unit: W/m2)
• αp is the coefficient of thermal expansion (SI unit: 1/K):

1 ∂ρ
α p = – ---
ρ∂T

for ideal gases, the thermal expansion coefficient takes the simpler form αp = 1 ⁄ T
• p is the pressure (SI unit: Pa)
• τ is the viscous stress tensor (SI unit: Pa)
• Q contains heat sources other than viscous dissipation (SI unit: W/m3)

For a steady-state problem the temperature does not change with time and the terms
with time derivatives disappear.

The first term of the right-hand side of Equation 4-18 is the work done by pressure
changes and is the result of heating under adiabatic compression as well as some
thermoacoustic effects. It is generally small for low Mach number flows.

THEORY FOR HEAT TRANSFER IN FLUIDS | 177

 ∂p
Q p = α p T  + u ⋅ ∇p (4-19)
∂t

The second term represents viscous dissipation in the fluid:

Q vd = τ: ∇u (4-20)

Theory for the Inflow Boundary Condition

At the inlet boundary of a fluid domain, the Inflow boundary condition defines a heat
flux that accounts for the energy that would normally be brought by the fluid flow if
the channel upstream to the inlet was modeled.

DANCKWERTS CONDITION
The application of the Danckwerts condition on the enthalpy allows to express the
normal conductive heat flux at the inlet boundary as proportional to the flow rate ρu
and the enthalpy variation ΔH between the upstream conditions and inlet conditions:

k ∇T ⋅ n = ρΔHu ⋅ n (4-21)

The enthalpy variation between the upstream conditions and inlet conditions, ΔH,
depends in general both on the difference in temperature and in pressure, and is
defined as:

T pA
1
ΔH = Tupstream Cp dT + pupstream --ρ- ( 1 – αp T ) dp (4-22)

where Tupstream is the upstream temperature, T is the inlet temperature, pustream is

the upstream absolute pressure, pA is the inlet absolute pressure, Cp is the fluid heat
capacity at constant pressure, ρ is the fluid density, and αp is its coefficient of thermal
expansion. See Equation 4-5 for details about the definition of the enthalpy.

In the unexpected case of a velocity field corresponding to an outgoing flow across the
inlet boundary, a zero conductive flux condition is applied to avoid a unphysical
conductive flux condition:

k ∇T ⋅ n = 0, n⋅u≥0

TEMPERATURE CONTRIBUTION TO THE INFLOW BOUNDARY CONDITION

The nature of the temperature contribution depends on the order of magnitude of
advective and conductive heat transfer at the inlet.

178 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

For large flow rates, advective heat transfer dominates over conductive heat transfer at
the inlet. In this case, the left-hand side of Equation 4-21 is small compared to its
right-hand side. As the heat capacity Cp is positive, in the absence of pressure
contribution to the enthalpy, this induces the following constraint on temperature

T – T upstream = 0

which corresponds to a Dirichlet boundary condition that prescribes the upstream

temperature at the inlet.

For low flow rates or in the presence of large heat sources or sinks next to the inlet, the
conductive heat flux cannot be neglected. The first integral in Equation 4-22 has for
effect to adjust the inlet temperature to balance the energy brought by the flow at the
inlet and the energy transferred by conduction from the interior.

PRESSURE CONTRIBUTION TO THE INFLOW BOUNDARY CONDITION

Pressure losses in the virtual channel upstream to the inlet boundary are handled
through the pressure contribution to the enthalpy (the second integral in
Equation 4-22).

In addition to the cases where the upstream and inlet absolute pressures are equal, this
term may be neglected when the work due to pressure changes is not included in the
energy equation, or when the fluid is modeled as an ideal gas (in this case the
coefficient of thermal expansion is the inverse of the temperature).

THEORY FOR HEAT TRANSFER IN FLUIDS | 179

 Theory for Heat Transfer in Moist Air
By making the ideal gas assumption in Equation 4-18, The Heat Transfer in Moist Air
Interface solves the following equation:

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ ( q + q r ) =  + u ⋅ ∇p + τ: ∇u + Q H + Q (4-23)
 ∂t  ∂t 

involving the same variables as in Theory for Heat Transfer in Fluids, and the
additional QH term, accounting for the diffusive flux of thermal enthalpy due to the
rate of change of air and vapor in moist air. It is defined as (Ref. 5):

Q H = – ( C p, v – C p, a )g w ⋅ ∇T

where:

• Cp,v is the specific heat capacity at constant pressure of vapor (SI unit: J/(kg·K))
• Cp,a is the specific heat capacity at constant pressure of air (SI unit: J/(kg·K))
• gw is the vapor flux by diffusion (SI unit: kg/(m2·s))

Humidity
This part defines the different definitions of humidity in the moist air theory.

MOISTURE CONTENT
The moisture content (also called mixing ratio or humidity ratio) is defined as the
ratio of water vapor mass, mv, to dry air mass, ma:

mv pv Mv
x vap = -------- = --------------- (4-24)
ma pa Ma

where pv is the water vapor partial pressure, pa is the dry air partial pressure, and Ma
and Mv are the molar mass of dry air and water vapor, respectively. The moisture
content represents a ratio of mass, and it is thus a dimensionless number.

RELATIVE HUMIDITY
The relative humidity of an air mixture is expressed as follows:

180 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 pv
φ w = ---------- (4-25)
p sat

where pv is the water vapor partial pressure and psat is the saturation pressure of water
vapor.

According to Dalton’s law, the total pressure of a mixture of gases is the sum of all the
partial pressures of each individual gas; that is, p = pv + pa where pa is the dry air
partial pressure.

The relative humidity formulation is often used to quantify humidity. However, for the
same quantity of moisture content, the relative humidity changes with temperature
and pressure, so in order to compare different values of φ w it has to be at the same
temperature and pressure conditions. Then the thermodynamical properties of moist
air can be deduced through the mixture formula described below.

The relative humidity is useful to study the condensation as it defines the boundary
between the liquid phase and the vapor phase. In fact, when the relative humidity φ w
reaches unity, it means that the vapor is saturated and that water vapor may condense.

SPECIFIC HUMIDITY
The specific humidity is defined as the ratio of water vapor, mv, to the total mass,
mtot = mv + ma:

mv
ω = ------------ (4-26)
m tot

When the water vapor only accounts for a few percent in the total mass,
the moisture content and the specific humidity are very close: xvap ≈ ω
(only for low values). For larger values of ω, the two quantities are more
precisely related by:

ω
x vap = -------------
1–ω

CONCENTRATION
The concentration is defined by:

nv
c v = ------ (4-27)
V

THEORY FOR HEAT TRANSFER IN MOIST AIR | 181

 where nv is the amount of water vapor (SI unit: mol) and V is the total volume (SI unit:
m3). According to the ideal gas hypothesis, the saturation concentration is defined as
follows:

p sat ( T )
c sat = -------------------- (4-28)
RT

Saturation State
The saturation state is reached when the relative humidity reaches one. It means that
the partial pressure of the water vapor is equal to the saturation pressure (which also
depends on the temperature).

From Ref. 36, the saturation pressure can be defined using the following expression:

T – 273.15 [ K ]
7.5 --------------------------------------
T – 35.85 [ K ]
p sat ( T ) = 610.7 [ Pa ] ⋅ 10 (4-29)

Temperature and saturation pressure are deduced from this formulation.

Moist Air Properties

The thermodynamical properties of moist air can be found with some mixture laws.
These are defined in this paragraph.

PRELIMINARY DEFINITIONS

Molar Fraction
The molar fraction of dry air, Xa, and the molar fraction of water vapor, Xv, are defined
such as:

na pa p – φ w p sat
X a = ---------- = ------ = --------------------------- (4-30)
n tot p p

nv pv φ w p sat
X v = ---------- = ------ = ------------------ (4-31)
n tot p p

where:

• na is amount of dry air

• nv is amount of water vapor

182 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• ntot is the total amount of moist air in mol
• pa is the partial pressure of dry air
• pv is the partial pressure of water vapor
• p is the pressure
• φ w is the relative humidity, and
• psat is the saturation pressure.

From Equation 4-30 and Equation 4-31, the following relation holds:

Xa + Xv = 1

Relative Humidity, Vapor Concentration, and Moisture Content

Moisture content and relative humidity can be related with the following expression:

c v RT x vap p
φ w = -------------------- = ---------------------------------------------------- (4-32)
p sat ( T ) Mv
p sat ( T )  -------- + x vap
Ma

MIXTURE PROPERTIES
The thermodynamical properties are built through a mixture formula. The expressions
depend on dry air properties and pure steam properties and are balanced by the mass
fraction.

Density
According to the ideal gas law, the mixture density ρm expression is defined as follows:

p
ρ m = -------- ( M a X a + M v X v ) (4-33)
RT

where Ma and Mv are the molar mass of dry air and water vapor, respectively, and Xa
and Xv are the molar fraction of dry air and water vapor, respectively.

The ideal gas assumption sets the compressibility factor and the
enhancement factor to unity. In fact, the accuracy lost by this assumption
is small as the pure steam represents a small fraction.

Heat Capacity at Constant Pressure

According to Ref. 37, the heat capacity at constant pressure of a mixture is:

THEORY FOR HEAT TRANSFER IN MOIST AIR | 183

 Ma Mv
C p, m = ---------- X a C p, a + ---------- X v C p, v (4-34)
Mm Mm

where Mm represents the mixture molar fraction and is defined by

Mm = Xa Ma + Xv Mv

and where Cp, a and Cp, v are the heat capacity at constant pressure of dry air and
steam, respectively.

Dynamic Viscosity
According to Ref. 37 and Ref. 38, the dynamic viscosity is defined as:

Xi μi
μm =  ---------------------------- (4-35)
i = a, v  X j ϕ ij
j a v
where ϕ ij is given by

1 1 2
--- ---
μi 2 Mj 4
1 +  -----  -------
μj Mi
ϕ ij = -----------------------------------------------
1
-
---
Mi 2
8  1 + -------
 M j

Here, μa and μv are the dynamic viscosity of dry air and steam, respectively.

Thermal Conductivity
According to Ref. 38 and Ref. 37, the thermal conductivity of the mixture is defined
similarly:

Xi ki
km =  ---------------------------- (4-36)
i = a, v  X j ϕ ij
j a v
where ka and kv are the thermal conductivity of dry air and steam, respectively.

PURE COMPONENT PROPERTIES

The dry air and steam properties used to define the mixture properties are
temperature-dependent high-order polynomials. The dry air properties are taken from
the Built-in materials database, and the polynomials have been computed according

184 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Ref. 39 for pure steam properties. The steam properties are based on the Industrial
Formulation IAPWS-IF97.

The valid temperature range is 200 K < T < 1200 K for dry air properties and
273.15 K < T < 873.15 K for steam properties.

RESULTS AND ANALYSIS VARIABLES

These variables are provided to display the related quantities:

• Moisture content xvap.

• Vapor mass fraction omega_moist.
• Concentration of water vapor c.
• Relative humidity phi. This variable corresponds to the calculated φ with the system
temperature and pressure.
• Saturation indicator satInd; this indicator is set to 1 if saturation has been detected
( φ w = 1 ), and 0 if not.

FUNCTIONS
The following functions are defined and can be used as feature parameters as well as in
postprocessing. Here, feature stands for fluid or porous, depending on whether the
function is defined in the Fluid or in the Porous Medium feature:

• ht.feature .fc(RH,T, pA), where RH is the relative humidity 0 ≤ φ w ≤ 1 , T is the

temperature (SI unit: K), and pA is the pressure (SI unit: Pa). It returns the
corresponding water vapor concentration (SI unit: mol/m3) by deriving the
following relation from Equation 4-25, Equation 4-28, and Equation 4-32:

x vap p
c v = -----------------------------------------
x Mv
+ -------- RT
 vap M a

The concentration computation assumes that the ideal gas assumption is

valid.

• ht.feature .fxvap(RH, T, pA), where RH is the relative humidity 0 ≤ φ w ≤ 1 , T is

the temperature (SI unit: K) and pA is the pressure (SI unit Pa). It returns the
moisture content (SI unit: 1) by using the following relation:

THEORY FOR HEAT TRANSFER IN MOIST AIR | 185

 φ w p sat Mv
x vap = --------------------------- ⋅ --------
p – φ w p sat M a

• ht.feature.fpsat(T), where T is the temperature (SI unit: K). It returns the

saturation pressure (SI unit: Pa) by using Equation 4-29.
• ht.feature.Lv(T), where T is the temperature (SI unit: K). It returns the latent
heat of evaporation (SI unit: J/kg) as a linear interpolation of the data from Ref. 39,
which provides steam properties based on the Industrial Formulation IAPWS-IF97.
The temperature-dependency is as shown on Figure 4-1.

Figure 4-1: Latent heat of evaporation as a function of temperature.

186 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Bioheat Transfer
The Bioheat Equation
The Bioheat Transfer Interface solves for the bioheat equation using Pennes’
approximation (Ref. 6). This is used to model heat transfer within biological tissue. It
accounts for heat sources from blood perfusion and metabolism in the classical heat
transfer equation:

∂T
ρCp + ∇ ⋅ q = ρ b C p, b ω b ( T b – T ) + Q met (4-37)
∂t

The different quantities involved here are recalled below:

• ρ is the density of the tissue (SI unit: kg/m3)

• Cp is the specific heat capacity at constant pressure of the tissue (SI unit: J/(kg·K))
• T is the absolute temperature of the tissue (SI unit: K)
• q is the heat flux by conduction in the tissue (SI unit: W/m2)
• ρb is the blood density (SI unit: kg/m3)
• Cp, b is the blood specific heat capacity at constant pressure (SI unit: J/(kg·K))
• ωb is the blood perfusion rate (SI unit: 1/s)
• Tb is the arterial blood temperature (SI unit: K)
• Qmet is the metabolic heat source (SI unit: W/m3)

For a steady-state problem the temperature does not change with time and the terms
with time derivatives disappear.

There are specific predefined materials available in the Bioheat material

database. See Materials Overview and Bioheat Materials Library in the
COMSOL Multiphysics Reference Manual.

Damaged Tissue
Add a Thermal Damage subnode under the Biological Tissue node to calculate tissue
damage.

THEORY FOR BIOHEAT TRANSFER | 187

 In hyperthermia and cryogenic processes, tissue necrosis (permanent damage or death
of living tissue) occurs when one of the two following phenomenas happens:

• an excessive thermal energy is absorbed (hyperthermia process) or released

(cryogenic process),
• a critical high (hyperthermia process) or low (cryogenic process) temperature is
exceeded.

Correspondingly, COMSOL Multiphysics has two ways to model energy absorption

— computing the period of time the tissue remained in the necrotic temperature
interval and direct time integration of the energy.

Temperature Threshold
In the first form of damage integral, tissue necrosis occurs in four cases:

• When the temperature exceeds the hyperthermia damage temperature Td, h for
more than a certain time period td, h,
• When the temperature falls below the cryogenic damage temperature Td, c for more
than a certain time period td, c,
• Instantly after the temperature exceeds the hyperthermia necrosis temperature
Tn, h,
• Instantly after the temperature falls below the cryogenic necrosis temperature Tn, c.

For the first two cases, the damaged tissue indicator, α, defined either by

t
1
α = ----------
t d, h  0 ϕ d, h dt
for hyperthermia analysis, or by

t
1
α = ---------
t d, c 0 ϕd, c dt
for cryogenic analysis, with

1 if T > T d, h 1 if T < T d, c
ϕ d, h ( t ) =  ϕ d, c ( t ) = 
0 otherwise 0 otherwise

is the ratio of the period of time when T > Td, h to the time limit td, h, or the ratio of
the period of time when T < Td, c to the time limit td, c. It gives an indication of

188 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

damage state of the tissue. When it reaches 1, the tissue is necrotic. The fraction of
necrotic tissue corresponds to the quantity min(α, 1).

For the last two cases, the necrosis time indicator, αnecr, defined either by

t
α necr = 0 ϕn, h dt
for hyperthermia analysis, or by

t
α necr =  0 ϕ n, c d t
for cryogenic analysis, with

1 if T > T n, h 1 if T < T n, c
ϕ n, h ( t ) =  ϕ n, c ( t ) = 
0 otherwise 0 otherwise

evaluates the period of time when T > Tn, h or the period of time when T < Tn, c. If
αnecr > 0, the tissue is necrotic because it already reached the necrosis temperatures
Tn, h or Tn, c at some time step of the simulation. Hence, the fraction of necrotic tissue
due to immediate necrosis is equal to 1 if αnecr > 0 and 0 otherwise.

Combining all cases, the overall fraction of necrotic tissue, θd, is equal to:

 1 if α necr > 0
θd =  (4-38)
 min (α, 1) otherwise

Arrhenius Kinetics
The second form of damage integral is applicable only for hyperthermia processes and
provides the degree of tissue injury, α, based on the polynomial Arrhenius equation:

– ΔE
∂α -----------
------- = ( 1 – α ) n Ae RT
∂t

Here, A is the frequency factor (SI unit: 1/s), and ΔE is the activation energy for the
irreversible damage reaction (SI unit: J/mol). The parameters A and ΔE are dependent
on the type of tissue and have been characterized for liver tissues by Jacques et others
(Ref. 7) to be A = 7.39 ⋅ 1039 s–1 and ΔE = 2.577 ⋅ 105 J/mol. See Ref. 8, Ref. 9, and
Ref. 10 for the characterization of these parameters for prostate, skin, and fat. See also
Ref. 11 and Ref. 12 for more references on biological tissues material properties.

THEORY FOR BIOHEAT TRANSFER | 189

 The fraction of necrotic tissue is then expressed by:

θ d = min(max ( α, 0 ), 1) (4-39)

Thermal Properties
The material properties of the damaged tissue are redefined to take into account the
influence of tissue injury. If ρd, Cp, d, and kd denote the density, heat capacity at
constant pressure, and thermal conductivity of the necrotic tissue, respectively, then
two effective quantities are defined:

• The effective thermal conductivity, keff = θdkd + (1 − θd)k

• The effective heat capacity at constant pressure, (ρCp)eff = θdρdCp, d + (1 − θd)ρCp

In these equalities, θd takes one of the two definitions given above in Equation 4-38
or Equation 4-39 according to the integral form chosen.

Heat Source
A cooling or heating source is associated with the reaction leading to damage of tissue.
Depending on the damage integral model, this source is expressed as follows:

• Temperature threshold for hyperthermia analysis:

∂θ d
Q = – ρL d, h --------- ( T > T d, h )
∂t

• Temperature threshold for cryogenic analysis:

∂θ d
Q = ρL d, c --------- ( T < T d, c )
∂t

• Energy absorption:
∂θ d
Q = – ρL ----------
∂t

190 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Heat Transfer in Porous
Media
When considering heat transfer in a porous medium at the microscopic scale, two heat
transfer equations can be established, for the solid and fluid phases. For undeformed
immobile solids, Equation 4-16 simplifies into:

∂T s
ρ s C p, s --------- + ∇ ⋅ q s = Q s
∂t

and for a fluid domain where pressure work and viscous dissipation are neglected,
Equation 4-18 becomes:

∂T f
ρ f C p, f --------- + ρ f C p, f u f ⋅ ∇T f + ∇ ⋅ q f = Q f
∂t

At the macroscopic scale, different modeling approaches are available, depending on

the local thermal equilibrium hypothesis, and on the accuracy required on the thermal
effects in the solid phase.

The local thermal equilibrium hypothesis assumes equality of temperature in both fluid
and solid phases:

Tf = Ts = T (4-40)

When this hypothesis can be assumed, the heat transfer equation for porous media is
derived from the mixture rule on energies appearing in solid and fluid heat transfer
equations (see Ref. 13). This rule applies by multiplying the equation of the solid
domain by the solid volume fraction, θs, multiplying the fluid equation by the porosity,
εp, and summing resulting equations.

The theory for this hypothesis is detailed in the Local Thermal Equilibrium section
below. Otherwise, the Local Thermal Nonequilibrium section describes the theory for
modeling heat transfer in porous media using two temperatures.

In the particular case of a packed bed of pellets, and under some assumptions about
thermal conductivities ratio, the local thermal nonequilibrium model can be tuned to
capture more precisely some thermal effects within the pellets. This is done by
replacing the macroscale heat equation for the solid phase by an ordinary differential

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 191

 equation for the microscale variation of temperature in the pellets along the radial
coordinate. See the Packed Bed of Pellets section for more details.

When Should Local Thermal Equilibrium and Nonequilibrium Be

Considered?
The classical local equilibrium hypothesis in modeling heat transfer in porous media
considers pointwise equality of solid and fluid temperatures as said in Equation 4-40.
The Local Thermal Equilibrium section below details the derivation of the energy
equation considering such assumption that remains accurately sufficient for several
applications. Ref. 32 shows for instance that solid and fluid temperatures are equal in
steady conduction problems where only prescribed temperature conditions are
applied. Most slow motion problems can also assume equality of phase temperatures if
volumetric internal heating do not differ in both materials.

In the case of conduction in porous plates, Ref. 33 provides criteria based on the
dimensionless Sparrow number, Sp, to indicate if temperature equilibrium is still valid
or if a nonequilibrium point of view should be preferred. In Ref. 34, the influence of
the Darcy number, Da, and the ratio of phase conductivities is examined for transient
heat transfer in packed beds. The Sparrow and Darcy numbers are defined by:

2
h sf L κ
Sp = --------------- Da = -----2-
k eff r h d

where:

• hsf is the interstitial heat transfer coefficient between solid and fluid phases (SI unit:
W/(m2·K))
• L is the plate layer thickness (SI unit: m)
• keff is the equivalent thermal conductivity of the porous medium (SI unit:
W/(m·K))
• rh is the hydraulic radius (SI unit: m)
• κ is the permeability (SI unit: m2)
• d is the average particle diameter (SI unit: m)

In the situations described in Ref. 33 and Ref. 34, small values of Sp (less than 100 or
500) and large values of Da (from order of magnitude 10-7) indicate discrepancies of
temperature in each phase. However, in general, assessing the validity of local thermal
equilibrium assumption remains not straightforward in specific situations. The Local

192 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Thermal Nonequilibrium approach, described below, makes use of two energy
equations, one for each phase of the porous medium, that solve for two temperature
fields. It numerically doubles the number of freedom to solve but provides a general
framework for heat transfer in porous media where evaluating the validity of the
equilibrium hypothesis is not required anymore.

Local Thermal Equilibrium

By making the local thermal equilibrium hypothesis for solid and fluid phases
(Equation 4-40), The Heat Transfer in Porous Media Interface solves for the
following version of the heat equation (Ref. 17), reformulated using a common
temperature, T:

∂T
( ρC p ) eff ------- + ρ f C p, f u ⋅ ∇T + ∇ ⋅ q = Q (4-41)
∂t

q = – k eff ∇T (4-42)

The different quantities appearing here are:

• ρf is the fluid density.

• Cp,f is the fluid heat capacity at constant pressure.
• (ρCp)eff is the effective volumetric heat capacity at constant pressure, defined by
( ρC p ) eff = θ s ρ s C p, s + ε p ρ f C p, f

• εp is the porosity.
• θs is the solid matrix volume fraction.
• ρs is the solid matrix density.
• Cp,s is the solid matrix heat capacity at constant pressure.
• keff is the effective thermal conductivity (a scalar or a tensor if the thermal
conductivity is anisotropic).
• q is the conductive heat flux.
• u is the velocity field, either an analytic expression or computed from a Fluid Flow
interface. It should be interpreted as the Darcy velocity, that is, the volume flow rate
per unit cross sectional area. The average linear velocity (the velocity within the

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 193

 pores) can be calculated as uf = u/εp, where εp is the fluid’s volume fraction, or
equivalently the porosity.
• Q is the heat source (or sink). Add one or several heat sources as separate physics
features.

For a steady-state problem the temperature does not change with time, and the terms
with time derivatives of Equation 4-41 disappear.

The effective thermal conductivity of the solid-fluid system, keff, is related to the
conductivity of the solid, ks, and to the conductivity of the fluid, kf, and depends in a
complex way on the geometry of the medium. In Ref. 13, Ref. 14, and Ref. 18, the
following models are proposed:

• If the heat conduction occurs in parallel in the solid and the fluid, then the effective
thermal conductivity is the weighted arithmetic mean of the conductivities kf and
ks:
k eff = θ s k s + ε p k f

This volume average model provides an upper bound for the effective thermal
conductivity.
• If the heat conduction takes place in series, with all of the heat flux passing through
both solid and fluid, then the effective thermal conductivity is the weighted
harmonic mean of the conductivities kf and ks:

1- θs εp
-------- = ----- + -----
k eff ks kf

This reciprocal average model provides a lower bound for the effective thermal
conductivity.
• A good estimate is given by the weighted geometric mean of kf and ks, as long as kf
and ks are not too different from each other
θs εp
k eff = k s ⋅ k f

• When the porous medium is composed of solid spherical inclusions in a fluid phase,
the effective thermal conductivity can be expressed as:
2k f + k s – 2 ( k f – k s )θ s
k eff = k f ----------------------------------------------------------
2k f + k s + ( k f – k s )θ s

194 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• On the opposite, when the porous medium is composed of fluid spherical inclusions
in a solid phase, the effective thermal conductivity can be expressed as:
2k s + k f – 2 ( k s – k f )ε p
k eff = k s ----------------------------------------------------------
2k s + k f + ( k s – k f )ε p

• In the case of a wrapped screen, the effective thermal conductivity of the solid-fluid
system can be defined as:
k f + k s – ( k f – k s )θ s
k eff = k f ---------------------------------------------------
k f + k s + ( k f – k s )θ s

• Finally, when the porous medium is made of sintered metal fibers, the effective
thermal conductivity of the solid-fluid system is defined as:

2 4ε p θ s k f k s
k eff = ε p k f + θ 2s k s + --------------------------
kf + ks

Note that when kf and ks are equal, all these models give the same effective thermal
conductivity. In addition, the effective conductivity is equal to kf when the porosity is
1, and to ks when the porosity is 0.

Local Thermal Nonequilibrium

The Local Thermal Nonequilibrium Interface implements heat transfer in porous
media for which the temperatures into the porous matrix and the fluid are not in
equilibrium.

EQUATIONS FOR LOCAL NONEQUILIBRIUM HEAT TRANSFER

Nonequilibrium heat transfer in porous media for binary systems of rigid porous
matrix and fluid phase are governed by a set of two equations. These are the usual heat
equations for solids and fluids, multiplied by the volume fractions θs and εp=1 − θs
respectively, and with an additional source term quantifying exchanged heat between
both phases (2.12 and 2.13 in Ref. 13):

∂T s
θ s ρ s C p, s --------- + ∇ ⋅ q s = q sf ( T f – T s ) + θ s Q s (4-43)
∂t

q s = – θ s k s ∇T s

∂T f
ε p ρ f C p, f -------- + ρ f C p, f u ⋅ ∇T f + ∇ ⋅ q f = q sf ( T s – T f ) + ε p Q f (4-44)
∂t

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 195

 q f = – ε p k f ∇T f

In these expressions:

• θs is the solid volume fraction (dimensionless)

• εp is the porosity (dimensionless)
• ρs and ρf are the solid and fluid densities (SI unit: kg/m3)
• Cp, s and Cp, f are the solid and fluid heat capacities at constant pressure (SI unit:
J/(kg·K))
• qs and qf are the solid and fluid conductive heat fluxes (SI unit: W/m2)
• ks and kf are the solid and fluid thermal conductivities (SI unit: W/(m·K))
• qsf is the interstitial convective heat transfer coefficient (SI unit: W/(m3·K))
• Qs and Qf are the solid and fluid heat sources (SI unit: W/m3)
• u is the fluid velocity vector (SI unit: m/s). It should be interpreted as the Darcy
velocity, that is, the volume flow rate per unit cross sectional area. The average linear
velocity (the velocity within the pores) can be calculated as uf = u/εp, where εp is the
fluid’s volume fraction, or equivalently the porosity.
The Porous Medium feature with Porous medium type set to Local thermal equilibrium
adds the exchanged opposite heat sources qsf(Tf − Ts) and qsf(Ts − Tf) that one phase
receives from or releases to the other when respective temperatures differ. The porous
temperature, T, has the following definition (Ref. 35):

θ s ρ s C p, s T s + ε p ρ f C p, f T f
T = -----------------------------------------------------------------
θ s ρ s C p, s + ε p ρ f C p, f

CORRELATION FOR THE INTERSTITIAL CONVECTIVE HEAT TRANSFER

COEFFICIENT
The Porous Medium feature with Porous medium type set to Local thermal equilibrium
provides a built-in correlation for qsf for a packed bed of spherical pellets (2.14, 2.15,
and 2.16 in Ref. 13):

q sf = S b h sf

The specific surface area, Sb (SI unit: 1/m), for a bed packed with spherical particles
of average diameter dpe is:

6θ s
S b = ---------
d pe

196 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The interstitial heat transfer coefficient, hsf (SI unit: W/(m2·K)), satisfies the relation:

1 d pe d pe
------- = -------------- + ---------
h sf k f Nu βk s

where β = 10 for spherical particles, and Nu is the fluid-to-solid Nusselt number

derived from following correlation (Ref. 15):

1⁄3 0.6
Nu = 2.0 + 1.1Pr Re p

The Prandtl number, Pr, and particle Reynolds number, Rep, are defined by:

μC p, f d pe ρ f u
Pr = -------------- Re p = ------------------------
kf μ

VOLUMETRIC AND SURFACE THERMAL CONDITIONS

Because each energy equation is multiplied by its volume fraction, θs and εp for solid
and fluid phases respectively, a heat source or heat flux defined in one of the phases is
also accounted with that ratio. As shown in Equation 4-43 and Equation 4-44, the
volumetric heat sources θsQs and εpQf are applied to the energy equations when Heat
Source is added as a subnodes of the Porous Matrix or Fluid nodes to specify Qs or Qf.

Packed Bed of Pellets

For the modeling of packed beds of pellets, the The Heat Transfer in Packed Beds
Interface interface implements an hybrid multiscale model of local thermal
nonequilibrium.

As in the standard local thermal nonequilibrium model, two temperatures, one for the
fluid phase and one for the pellets, are solved for. By assuming that the thermal
conductivity of the pellets is much smaller than the one of the fluid phase, heat
conduction among different pellets is neglected, and a 1D microscale equation for
temperature conduction along the radial coordinate in the pellets is defined to replace
the macroscale heat transfer equation. It is coupled to the macroscale heat transfer
equation in the fluid phase, either by assuming temperature continuity, or a convective
heat flux, at the outer surface of the pellets.

This model provides higher accuracy than the standard local thermal nonequilibrium
model for highly nonlinear problems, in particular when combustion occurs within the
pellets.

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 197

 Spheres, cylinders, flakes, or any shape of pellets may be considered, as long as the
specific surface area for heat exchange between the pellets and fluid phase can be
determined. Porosity within the pellets themselves may be considered as well, in
addition to the bed porosity.

EQUATIONS FOR HEAT TRANSFER IN A PACKED BED OF PELLETS

Macroscopic Heat Transfer in the Fluid

For the fluid phase, the heat equation is similar to the one solved in the standard local
thermal nonequilibrium model:

∂T f
ε p ρ f C p, f -------- + ρ f C p, f u ⋅ ∇T f + ∇ ⋅ q f = Q pe, f + ε p Q f (4-45)
∂t

q f = – ε p k f ∇T f

In these expressions:

• εp is the packed bed porosity (dimensionless)

• ρf is the fluid density (SI unit: kg/m3)
• Cp, f is the fluid heat capacity at constant pressure (SI unit: J/(kg·K))
• qf is the fluid conductive heat flux (SI unit: W/m2)
• kf is the fluid thermal conductivity (SI unit: W/(m·K))
• Qpe,f is the pellet-fluid heat exchange term (SI unit: W/(m3·K))
• Qf is any fluid heat source (SI unit: W/m3)
• u is the fluid velocity vector (SI unit: m/s). It should be interpreted as the Darcy
velocity, that is, the volume flow rate per unit cross sectional area. The average linear
velocity (the velocity within the pores) can be calculated as uf = u/εp, where εp is the
fluid’s volume fraction, or equivalently the porosity.

Microscopic Heat Transfer in the Pellets

In the pellets, the heat balance equation at normalized radius r is

∂T pe ∂ ∂T pe
( ρC p ) pe, eff ------------ + ------------------2 -----  – r k pe, eff ------------- = Q pe
1 2
(4-46)
∂t ∂ r  ∂r 
( rr pe )

where

• (ρCp)pe,eff is the effective volumetric heat capacity at constant pressure inside the
pellet filled with fluid (SI unit: J/(kg·K)), defined by

198 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ( ρC p ) pe, eff = ( 1 – ε pe )ρ pe C p, pe + ε pe ρ f C p, f

• ρpe is the density of the porous matrix within the pellet (SI unit: kg/m3)
• Cp,pe is the volumetric heat capacity at constant pressure of the porous matrix within
the pellet (SI unit: J/(kg·K))
• εpe is the porosity within the pellet (dimensionless)
• Tpe is the temperature along the radial coordinate of the pellet (SI unit: K)
• rpe is outer radius of the pellet (SI unit: m)
• kpe,eff is the effective thermal conductivity of the pellet filled with fluid (SI unit:
W/(m·K)), defined by
k pe, eff = ( 1 – ε pe )k pe + ε pe k f

• kpe is the thermal conductivity of the porous matrix within the pellet (SI unit:
W/(m·K))
• Qpe is any heat source applied to the pellet (SI unit: W/m3)

Coupling Condition at Pellet-Fluid Interface

As in the local thermal nonequilibrium model, the interstitial heat exchange between
the fluid and the outer surface of the pellets is accounted for.

The heat flux at the outer surface of the pellet, applied in the microscale pellet
equation, is

q pe r=1 = h pe, f ( T f – T pe r=1

)

with hpe,f the interstitial heat transfer coefficient (SI unit: W/(m2·K)), defined as in
the standard local thermal nonequilibrium model.

The corresponding heat exchange term applied in the macroscale heat transfer
equation of the fluid is:

Q pe, f = – S b q pe r=1

with Sb the specific surface area of the pellets. For a packed bed of spherical pellets, it
is defined as

6 ( 1 – εp )
S b = -----------------------
d pe

with dpe the diameter of the pellet (SI unit: m).

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 199

 When the pellets are not spherical, a correction factor is applied to the radius, to extend
this formula.

When continuity of temperatures at the fluid-pellet interface can be assumed, i.e. when

T pe r=1 = Tf

then the interstitial heat flux can be rewritten as

∂T pe
= – --------  k pe, eff -------------
1
q pe r=1 r pe  ∂r  r = 1

Average Temperature in the Packed Bed

The porous temperature, T, has the following definition (Ref. 35):

Vpe ( ρCp )pe, eff Tpe dV

( 1 – ε p ) ------------------------------------------------------------ + ε p ρ f C p, f T f
V pe
T = --------------------------------------------------------------------------------------------------------------------
Vpe ( ρC p ) pe, eff dV
( 1 – ε p ) -------------------------------------------------- + ε p ρ f C p, f
V pe

Theory for Heat Transfer in Moist Porous Media

A moist porous medium is filled with moist air and liquid water. By making the local
thermal equilibrium hypothesis for all phases (solid matrix, moist air and liquid water),
The Heat Transfer in Moist Porous Media Interface solves for the following heat
equation, obtained from Equation 4-41 by accounting for the liquid water phase in the
definition of the effective material properties and in the convective heat flux:

∂T
( ρC p ) eff ------- + ( ρ g C p, g u g + ρ l C p, l u l ) ⋅ ∇T + ∇ ⋅ q = Q + Q evap (4-47)
∂t

q = – k eff ∇T (4-48)

The different quantities appearing here are:

• ρg and ρl (SI unit: kg/m3) the moist air and liquid water densities.
• Cp,g and Cp,l (SI unit: J/(kg·K)) the moist air and liquid water heat capacities at
constant pressure.

200 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• (ρCp)eff (SI unit: J/(m3·K)) the effective volumetric heat capacity at constant
pressure defined by an averaging model to account for the solid matrix, the moist
air, and the liquid water properties:
( ρC p ) eff = θ s ρ s C p, s + ε p [ ( 1 – s l )ρ g C p, g + s l ρ l C p, l ]

• where sl is the liquid water saturation.

• q the conductive heat flux (SI unit: W/m2).
• ug (SI unit: m/s) the moist air velocity field, that should be interpreted as the Darcy
velocity, that is, the volume flow rate per unit cross sectional area.
• ul (SI unit: m/s) the liquid water velocity field, that should be interpreted as the
Darcy velocity, that is, the volume flow rate per unit cross sectional area.
• keff (SI unit: W/(m·K)) the effective thermal conductivity (a scalar or a tensor if the
thermal conductivity is anisotropic), defined by an averaging model to account for
solid matrix, the moist air, and the liquid water properties. See Moist Porous
Medium for details about the available averaging models.
• Q (SI unit: W/m3), defined as the sum of the diffusive flux of enthalpy and of the
liquid capillary flux:
Q = – [ ( C p, v – C p, a )g w + C p, l g lc ]

See Theory for Moisture Transport in Porous Media for the definition of gw and glc.
• Qevap (SI unit: W/m3) the heat source (or sink) due to water phase change, defined
as:

Q evap = L v G evap

See Theory for Moisture Transport in Porous Media for the definition of Gevap.

THEORY FOR HEAT TRANSFER IN POROUS MEDIA | 201

 Theory for Heat Transfer with Phase
Change
Heat Transfer with Phase Change
Heat transfer and phase change are inter-related processes:

• Matter exists in a certain phase, liquid, solid or gas, at equilibrium under a given
condition of pressure and temperature. The phase transition occurs when the
temperature and pressure of the system come across a critical condition at which the
material changes phase. The critical value of temperature at which the transition
occurs for a given reference pressure can be computed through statistical physics or
tabulated from experiments. In the following, we assume that the phase change is
driven by the temperature, neglecting all other factors that may affect phase change,
as equilibrium time, or purity of the material.
• At the macroscopic and mesoscopic level, phase change costs or releases a certain
amount of energy which is called latent heat of phase transition, which is an input
of the model. Because there is a balance between sensible and latent heat, phase
change induces temperature variations.

The dedicated features for phase change simulation that are available in the Heat
Transfer module handle models were it is assumed that the different phases are non
miscible so that each phase is located in a distinct domain. The simulation predicts the
position of the phase change interface and the associated thermal effects. The
following sections present the two numerical methods available to do the modeling:

• Arbitrary Lagrangian-Eulerian Method

• Apparent Heat Capacity Method

Arbitrary Lagrangian-Eulerian Method

When the phase change interface is tracked using the arbitrary Lagrangian-Eulerian
(ALE) method, the phase change interface velocity is defined explicitly by the Stefan
condition, which expresses the energy balance at the interface.

It applies to the case of phase change between a solid (or immobile fluid) and a fluid
phase, with a sharp transition between the phases, and no topology changes.

202 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The Phase Change Interface boundary condition is applicable on the boundaries
located at the interface between a solid and a fluid domain, two solid domains, two
fluid domains or two porous medium domains.

The Phase Change Interface, Exterior boundary condition is applicable on the exterior
boundaries adjacent to a solid, fluid, or porous medium domain.

The latent heat of phase change is taken into account through the definition of the
phase change velocity.

The Stefan condition defines the phase change interface velocity vn as follows:

q ⋅ n mesh
v n = -------------------------
ρs Ls → f

with L s → f the latent heat of phase change from solid to fluid (SI unit: J/kg), ρs the
solid density (SI unit: kg/m3), and q the conductive heat flux jump across the interface
(SI unit: W/m2), defined as

q = – k s ∇T s + k f ∇T f

with ks the solid thermal conductivity (SI unit: W/(m·K)), kf the fluid thermal
conductivity (SI unit: W/(m·K)), and Ts, Tf the temperatures on each side of the
interface.

When the density changes during the phase change, the volume change is
compensated by a velocity at the interface in the fluid phase. The normal fluid velocity
is expressed as:

ρf – ρs
v f = ---------------- v n
ρf

The phase change velocity vn appearing in Stefan condition is relative to the solid
position. In case of translation of the solid (using Translational Motion subfeature
under Solid feature for instance), as in a continuous casting process, the solid
translation velocity us contributes to vn to describe the interface velocity relative to the
spatial frame dx/dt:

dx-
------ ⋅ n = vn + us ⋅ n
dt

Similarly, the same term is added to the normal fluid velocity definition relative to the
spatial frame:

THEORY FOR HEAT TRANSFER WITH PHASE CHANGE | 203

 ρf – ρs
v f = ---------------- v n + u s ⋅ n
ρf

Apparent Heat Capacity Method

The apparent heat capacity formulation provides an implicit capturing of the phase
change interface, by solving for both phases a single heat transfer equation with
effective material properties. The latent heat of phase change is taken into account by
modifying the heat capacity.

It applies well to materials showing a mushy zone around the phase change interface.
And it allows topology changes of the phase change interface.

The Phase Change Material domain condition should be used on both phases
domains, to solve the heat equation after specifying the properties of the phase change
material according to the apparent heat capacity formulation.

Instead of adding a latent heat L in the energy balance equation exactly when the
material reaches its phase change temperature Tpc, it is assumed that the
transformation occurs in a temperature interval between Tpc − ΔT ⁄ 2 and Tpc + ΔT ⁄ 2.
In this interval, the material phase is modeled by a smoothed function, θ, representing
the fraction of phase before transition, which is equal to 1 before Tpc − ΔT ⁄ 2 and to
0 after Tpc + ΔT ⁄ 2. The density, ρ, and the specific enthalpy, H, are expressed by:

ρ = θρ 1 + ( 1 – θ )ρ 2

1
H = --- ( θρ 1 H 1 + ( 1 – θ )ρ 2 H 2 )
ρ

where the indices 1 and 2 indicate a material in phase 1 or in phase 2, respectively.

Differentiating with respect to temperature, this equality provides the following
formula for the specific heat capacity:

∂H
Cp =
∂T

which becomes, after some formal transformations:

1 dα m
C p = --- ( θ 1 ρ 1 C p, 1 + θ 2 ρ 2 C p, 2 ) + ( H 2 – H 1 ) -----------
ρ dT

204 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Here, θ1 and θ2 are equal to θ and 1−θ, respectively. See Figure 4-2 for details.

Figure 4-2: Phase indicators, phase change temperature, and transition interval.

The mass fraction, αm, is defined from ρ1, ρ2 and θ according to:

1 θ2 ρ2 – θ1 ρ1
α m = --- ------------------------------
2 ρ

It is equal to −1 ⁄ 2 before transformation and 1 ⁄ 2 after transformation. The specific

heat capacity is the sum of an equivalent heat capacity Ceq:

1
C eq = --- ( θ 1 ρ 1 C p, 1 + θ 2 ρ 2 C p, 2 )
ρ

and the distribution of latent heat CL:

dα m
C L ( T ) = ( H 2 – H 1 ) -----------
dT

In the ideal case, when 1 − θ is the Heaviside function (equal to 0 before

Tpc and to 1 after Tpc), dαm ⁄ dT is the Dirac pulse.

Therefore, CL is the enthalpy jump, L, at temperature Tpc that is added

when you have a pure substance.

The latent heat distribution CL is approximated by

dα m
C L ( T ) = L -----------
dT

THEORY FOR HEAT TRANSFER WITH PHASE CHANGE | 205

 so that the total heat per unit volume released during the phase transformation
coincides with the latent heat:

ΔT ΔT
T pc + -------- T pc + -------- dα
 
2 2 m
C ( T ) dT = L ----------- dT = L
ΔT L ΔT
T pc – -------- T pc – -------- dT
2 2

The latent heat, L, can depend on the absolute pressure but should not
depend on the temperature.

Finally, the apparent heat capacity, Cp, used in the heat equation, is given by:

1
C p = --- ( θ 1 ρ 1 C p, 1 + θ 2 ρ 2 C p, 2 ) + C L
ρ

and the ratio of specific heats, γ, is:

Cp θ 1 ρ 1 C p, 1 + θ 2 ρ 2 C p, 2
γ = ------- = --------------------------------------------------------
Cv C p, 1 C p, 2
θ 1 ρ 1 ------------ + θ 2 ρ 2 ------------
γ1 γ2

The effective thermal conductivity reduces to:

k = θ1 k1 + θ2 k2

and the effective density is:

ρ = θ1 ρ1 + θ2 ρ2

To satisfy energy and mass conservation in phase change models, particular attention
should be paid to the density in time simulations. When the fluid density is not
constant over time, for example, dependent on the temperature, the transport velocity
field and the density must be defined so that mass is conserved locally.

The Moving Mesh Interface (described in the COMSOL Multiphysics

Reference Manual) can be used to account for model deformation.

206 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

PHASE CHANGE IN SOLID MATERIALS
When phase change is considered in solid materials, the density is defined on the
material frame. Therefore a single density should be defined for the different phases to
ensure mass conservation on the material frame:

ρ = ρ1 = ρ2

The expression of the specific enthalpy, H, simplifies to:

H = θH 1 + ( 1 – θ )H 2

The apparent heat capacity, Cp, used in the heat equation, is given by:

dα m
C p = ( θ 1 C p, 1 + θ 2 C p, 2 ) + L -----------
dT

where the mass fraction is:

θ2 – θ1
α m = -----------------
2

THEORY FOR HEAT TRANSFER WITH PHASE CHANGE | 207

 Theory for Heat Transfer in Building
Materials
The Heat Transfer in Building Materials Interface solves for the following equations
derived from Ref. 16:

∂T
( ρC p ) eff ------- + ∇ ⋅ q = Q (4-49)
∂t

q = – ( k eff ∇T + L v δ p ∇( φ w p sat ) ) (4-50)

which is derived from Equation 4-15, considering the building material as a porous
medium in local thermal equilibrium in which the following mixing rules apply:

• (ρCp)eff (SI unit: J/(m3·K)) is the effective volumetric heat capacity at constant
pressure, defined to account for both solid matrix and moisture properties:
( ρC p ) eff = ρ s C p, s + w ( φ w )C p, w

where ρs (SI unit: kg/m3) is the dry solid density, Cp,s (SI unit: J/(kg·K)) is the dry
solid specific heat capacity, w ( φ w ) (SI unit: kg/m3) is the water content given by a
moisture storage function, and Cp,w (SI unit: J/(kg·K)) is the water heat capacity at
constant pressure.
• keff (SI unit: W/(m·K)) is the effective thermal conductivity, defined as a function
of the solid matrix and moisture properties:

bw ( φ w )
k eff = k s  1 + --------------------
 ρs 

where ks (SI unit: W/(m·K)) is the dry solid thermal conductivity and b
(dimensionless) is the thermal conductivity supplement.
This definition neglects the contribution due to the volume fraction change of the
moist air.

The heat source due to moisture content variation is expressed as the vapor diffusion
flow multiplied by latent heat of evaporation:

L v δ p ∇( φ w p sat )

208 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

where Lv (SI unit: J/kg) is the latent heat of evaporation, δp (SI unit: s) is the vapor
permeability, φ w (dimensionless) is the relative humidity, and psat (SI unit: Pa) is the
vapor saturation pressure.

THEORY FOR HEAT TRANSFER IN BUILDING MATERIALS | 209

 Theory for Harmonic Heat Transfer
When submitted to periodic sinusoidal heat loads at a given frequency, the temperature
response of a body can, in some cases, be assumed periodic, sinusoidal, and of same
frequency around an equilibrium temperature. If the temperature changes are small
enough or if the thermal properties are constant, this time-dependent periodic
problem may be replaced by an equivalent linear steady-state problem in the frequency
domain, which is far less computationally expensive. The Heat Transfer interfaces
support such frequency domain studies via the Thermal Perturbation, Frequency Domain
study sequence.

In complex notations, the aforementioned temperature decomposition is expressed as:

jωt
T ( t ) = T 0 + T'e (4-51)

where T0 is the equilibrium temperature that verifies the steady-state heat transfer
equation and may come from the solution of a previous study, T′ is the complex
amplitude of the harmonic perturbation around T0, and ω is the angular frequency,
related to the ordinary frequency, f, according to

ω = 2πf

Note: The amplitude, T′, is complex-valued since it includes the phase term ejϕ.

From the temperature decomposition in Equation 4-51, and according to the heat
transfer equation in Equation 4-16, heat transfer by conduction in solids is then
governed by:

jωt jωt jωt

jωρC p T'e + ∇ ⋅ ( – k∇T 0 ) + ∇ ⋅ ( – k∇T' )e = Q + Q'e

where Q′ejωt is the harmonic perturbation in domain around an average heat source,
Q. Removing the terms of the steady-state heat transfer equation satisfied by T0, and
simplifying by ejωt, this reduces to:

jωρC p T' + ∇ ⋅ ( – k∇T' ) = Q'

210 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

which is the governing equation for T′. For constant material properties, the linearized
form reads:

jωρ 0 C p, 0 T' + ∇ ⋅ ( – k 0 ∇T' ) = Q' (4-52)

and for nonlinear material properties:

∂k 
jωρ 0 C p, 0 T' + ∇ ⋅  – k 0 ∇T' – T'  ∇T  = Q' (4-53)
∂ T T 0 0

Here, ρ0, Cp, 0, and k0 denote the density, heat capacity at constant pressure, and
thermal conductivity, evaluated at T0, that is: ρ(T0), Cp(T0), and k(T0), respectively.

When the linearized heat transfer equation, such as Equation 4-52 or Equation 4-53,
can still describe the model accurately, the problem becomes steady-state in the
frequency domain, therefore computationally less expensive than a time-dependent
simulation. An automatic linearization process is performed by COMSOL
Multiphysics so that no additional action is needed from the user to get these
equations, even in the presence of temperature-dependent coefficients, in domains and
boundaries. Only the expressions of the material properties and other parameters, as
functions of the temperature, are required as for usual nonlinear modeling.

Harmonic loads can be prescribed through temperature harmonic constraints on

boundaries or heat sources on domains and boundaries.

THEORY FOR HARMONIC HEAT TRANSFER | 211

 Theory for Lumped Isothermal
Domain
The Isothermal Domain feature considers the temperature to be homogeneous in
space but not necessarily in time. This is an approximation adapted to situations where
a domain is nearly at the average temperature and with small fluctuations, for instance,
solid objects made of conductive material immersed in water, or global temperature of
a heated and well insulated room adjacent to a cold environment.

Recalling Equation 4-18 given previously in the Theory for Heat Transfer in Fluids
section, without pressure-volume work and viscous dissipation, the equation to be
solved reduces to:

dT
ρC p +∇⋅q = Q
dt

Integrating this equation over the domain leads to:

dT
mC p
dt
+ S ( n ⋅ q ) ds = V Q dv (4-54)

where the domain mass and the heat capacity at constant pressure are

1
m = V ρ dv C p = -----
m V ρCp dv

212 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The exterior boundaries of each Isothermal Domain need the heat exchange to be
specified.

Isothermal domain 3

Isothermal domain 2

Isothermal domain 1

Isothermal domain interfaces

Figure 4-3: Interfaces at exterior boundaries of each Isothermal Domain.

COMSOL Multiphysics provides several types of interfaces: thermal insulation,

continuity, ventilation, convective heat flux, and thermal contact.

THERMAL INSULATION
The Thermal insulation condition prevents any heat transfer between both adjacent
domains.

CONTINUITY
The Continuity condition ensures equal temperature at both sides of the interface.

VENTILATION
The Ventilation condition is used for cases when an isothermal domain is considered
fluid and has an adjacent domain containing the same fluid. An opening lets the fluid
going from one domain to another with a determined mass flux, denoted by ϕ d → u
or ϕ u → d , respectively, along or opposite to the geometrical normal vector. The
Ventilation condition is written

THEORY FOR LUMPED ISOTHERMAL DOMAIN | 213

 –nd ⋅ qd = ϕd → u Hd – ϕu → d Hu (4-55)

CONVECTIVE HEAT FLUX

The Convective heat flux condition is adapted to cases when an isothermal domain is
considered solid and is adjacent to a fluid. Convection occurs at the interface with a
specified heat transfer coefficient, h. The interface condition reads

–nd ⋅ qd = –h ( Tu – Td ) (4-56)

THERMAL CONTACT
When an isothermal domain is considered solid and is adjacent to another solid,
thermal contact occurs and is characterized by a given thermal resistance, Rt. At the
interface, the condition Thermal contact reads

Tu – Td
– n d ⋅ q d = – -------------------- (4-57)
R t, s

214 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Heat Transfer in Thin
Structures
In COMSOL Multiphysics, thin domains of solid, fluid or porous layered materials
have dedicated tools to model them with boundaries instead of full domains. This way,
simulations directly benefit from a reduced number of mesh elements. This section
presents the hypotheses needed for such approximations and the resulting heat transfer
equations that hold.

In this section:

• Modeling Layered Materials

• Theoretical Background of the Different Formulations
• Thin Layer
• Thin Film
• Fracture
• Thin Rod

Modeling Layered Materials

The Heat Transfer interfaces contain several lumped conditions for modeling heat
transfer in layered materials: Thin Layer, Thin Film, Fracture, and Thin Rod.

In addition, standalone physics interfaces are available for the modeling of heat transfer
by conduction, convection and radiation in thin structures:

• The Heat Transfer in Shells Interface

• The Heat Transfer in Films Interface
• The Heat Transfer in Fractures Interface

Either the Solid, Fluid, or Porous Medium feature is available by default in each of these
interfaces.

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 215

 The features mentioned above are the counterparts of domain features for the
modeling of heat transfer in solid, fluid, and porous thin structures that can be
represented as boundaries or edges, as described in Table 4-2.
TABLE 4-2: EQUIVALENT DOMAIN AND THIN STRUCTURES FEATURES

TYPE OF MEDIUM DOMAIN FEATURE BOUNDARY FEATURE EDGE FEATURE

Solid Solid Thin Layer (ht) Thin Rod (ht)

Solid (htlsh)
Fluid Fluid Thin Film (ht) —
Fluid (htlsh)
Porous Porous Medium Fracture (ht) —
Porous Medium
(htlsh)

All these functionalities have in common the fact that the thin domains they model are
lumped into boundaries (for Thin Layer, Thin Film and Fracture) or 3D edges (for
Thin Rod).

REDUCED MESH ELEMENT NUMBER

A significant benefit is that a thin structure can be represented as a boundary instead
of a domain and a rod can be represented as a 3D edge. This simplifies the geometry
and reduces the required number of mesh elements. Figure 4-4 shows an example
where a thin structure significantly reduces the mesh density.

Copper wire modeled

with a mesh

Copper wire represented

as a thin structure

Figure 4-4: Modeling a copper wire as a domain (top) requires a denser mesh compared to
modeling it as a boundary with a conductive layer (bottom).

216 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

THIN STRUCTURE AS AN EXTRA DIMENSION
To model heat transfer through the thickness of a thin structure, or multiple
sandwiched layers with different material properties and thicknesses, COMSOL
Multiphysics gives the possibility to create a product space between the dimensions of
the boundary and an additional dimension. This is realized by the Extra Dimension
tool through either the General option of the Thin Layer (Heat Transfer Interface) and
Solid (Heat Transfer in Shells Interface), Thin Film (Heat Transfer Interface) and
Fluid (Heat Transfer in Shells Interface), or Fracture (Heat Transfer Interface) and
Porous Medium (Heat Transfer in Shells Interface) features.

Adding Extra Dimensions to a Model and Using Extra Dimensions in the

COMSOL Multiphysics Reference Manual.

Plotting and Evaluating Results in Layered Materials

An additional 1D segmented line represents the thickness of the thin structure. The
number of mesh points for each interval of the extra dimension is set to 2 by default.

TANGENTIAL AND NORMAL GRADIENTS

In thin structures, the tangential gradient and the normal gradient can be more
appropriate to express the governing equations.

The normal gradient is the projection of the gradient operator onto the normal
vector, n, of the boundary representing the thin structure. This is mathematically
expressed for any scalar field T as:

∇ n T = ( ∇T ⋅ n )n

The tangential gradient removes the normal component from the gradient
operation, so that only tangential components remain. This is mathematically
expressed for any scalar field T as:

∇ t T = ∇T – ( ∇T ⋅ n )n

The gradient operator is then split into a tangential part and a normal part:

∇T = ∇ t T + ∇ n T (4-58)

This relation simplifies to:

∇T = ∇ t T or ∇T = ∇ n T

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 217

 when tangential heat transfer is dominant or negligible. These results will be useful in
the next sections describing heat transfer in the different thin structures.

Equation 4-58 is valid for flat layered shells. For the curved ones, the gradient
expression should account for the surface area and the scaling of each layer. The
gradient in the product geometry of a curved layered shell with variable thickness can
be written as:

∂X –1 T
∇T =  --------- ⋅ ( ∇t T + ∇n T )
 ∂X r

with

∂X-
= I +  z + --- ( – 1 + z off ) ∇tn
d
--------
∂X r  2 

where

• Xr are the reference surface coordinates

• z is the extra-dimension thickness coordinate (varying between 0 and d)
• zoff is the relative midplane offset
• d is the layered material thickness

It should be noted that when an extra dimension is used, the equations

are written from the point of view of the extra dimension. In particular,
the dtang() operator would correspond to ∇n since it performs the
derivation along the extra line. In the thin structure boundary, dtang()
would correspond to ∇t.

In some applications, it is required to model variable thickness layers. This is achieved

by scaling the constant thickness of the layer, d, using a thickness factor, lsc, which can
be a function of the surface coordinates. The deformation gradient in a scaled product
geometry of a curved layered shell can be written as

∂X – 1 T ∇n T
∇T =  --------- ⋅  ∇ t T + ------------
∂X r lsc

∂X-
= I +  z + --- ( – 1 + z off ) ∇t( lscn )
d
--------
∂X r  2 

218 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

AREA SCALE FACTOR
For a curved layered shell, the change in surface area of each layer should be accounted
for while integrating the governing equations. The area scale factor for each layer of
the layered material can be defined as

∂X
ASF = --------
-
∂X r

THERMAL CONDUCTIVITY TENSOR IN LOCAL BOUNDARY SYSTEMS

The thermal conductivity k describes the relationship between the heat flux vector q
and the temperature gradient ∇T as in

q = – k ∇T

which is Fourier’s law of heat conduction (see also The Physical Mechanisms
Underlying Heat Transfer).

The tensor components can be specified in the local coordinate system of the
boundary, which is defined from the geometric tangent and normal vectors. The local
x direction, ex, loc, is the surface tangent vector t1, and the local z direction, ez, loc, is
the normal vector n. Their cross product defines the third orthogonal direction such
that:

x, loc
e = t1
y, loc
e = n × t1
z, loc
e = n

From this, a transformation matrix between the local coordinate system and the global
coordinate system can be constructed in the following way:

x, loc y, loc z, loc

ex ex ex
A = e x, loc e y, loc e z, loc
y y y
x, loc y, loc z, loc
ez ez ez

The thermal conductivity tensor in the local coordinate system, kbnd, is then expressed
as

k bnd = AkA T

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 219

 Theoretical Background of the Different Formulations
Three formulations are available for the modeling of heat transfer in thin structures
defined as boundaries:

• The general formulation, using the Extra Dimension tool to solve the equations
into the boundaries and through the thin structure’s thickness
• The thermally thin approximation, a lumped formulation assuming that heat
transfer mainly follows the tangential direction of the thin structure
• The thermally thick approximation, a lumped formulation assuming that heat
transfer is dominant in the direction normal to the thin structure

They all derive from the energy equation established in Equation 4-15, and recalled
here below:

∂E
ρ + ρu ⋅ ∇E + ∇ ⋅ ( q + q r ) = – ( σ:D ) + Q
∂t

where E is the variable for the internal energy.

GENERAL FORMULATION
The general formulation uses the Extra Dimension tool to solve the equations through
the thin structure’s thickness. The thin structure has its domain represented by the
product space between the lumped boundary and the additional dimension for the
thickness. Applying the split of the gradient operator given earlier at Equation 4-58,
the energy equation becomes

∂E s
ρ + ρu ⋅ ( ∇ t E s + ∇ n E s ) + ∇ ⋅ ( q + q r ) = – ( σ:D ) + Q (4-59)
∂t

The ∇t operator is the tangential derivative in the thin structure boundary, and the ∇n
operator is the derivation operator along the extra dimension which is normal to the
thin structure (see Tangential and Normal Gradients). The subscript s appended on E
(and T in the following) is here to recall that this variable lives in the product space of
the thin structure.

Equation 4-59 comes along with Fourier’s law of conduction:

q = –k ( ∇t Ts + ∇n Ts ) (4-60)

and constraints on the temperature at the extremities of the extra dimension:

220 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 Td = ( Ts )L = 0 Tu = ( Ts )L = d (4-61)
s

Here, ds is the length of the extra dimension, or equivalently the thickness of the thin
structure, and Tu and Td are the temperature at the upside and the downside of the
thin structure.

THERMALLY THIN APPROXIMATION

This formulation applies to a thin structure where heat transfer mainly follows the
tangential direction. The gradient operator is then simplified to

∇T = ∇ t T

This assumption is often valid for thin structures that are good thermal conductors
compared to the adjacent domains, and/or with fast convection along the tangential
direction.

With these assumptions, Equation 4-15 becomes:

∂E
ds ρ + d s ρu ⋅ ∇ t E + ∇ t ⋅ ( q s + q r ) = – d s ( σ:D ) + d s Q + q 0 (4-62)
∂t

q s = – d s k∇ t T (4-63)

where ds is the layer thickness (SI unit: m). The heat source Q is a density distributed
in the layer while q0 is the received out-of-plane heat flux.

In 2D, Equation 4-62 and Equation 4-63 have an additional factor, dz,
to account for the out-of-plane thickness.

When Equation 4-62 is solved in a boundary adjacent to a domain modeling heat

transfer, the two entities exchange a certain amount of heat flux according to:

q0 = n ⋅ q

In this coupling relation, the outgoing heat flux n ⋅ q leaves the domain and is received
in the source term q0 by the adjacent thin layer modeled as a boundary. From the point
of view of the domain, and neglecting thermoelastic effects, the following heat source
is received from the thin structure:

∂E
–n ⋅ q = ds Qs – ds ρ – ( d s ρu ⋅ ∇ t E ) – ∇ t ⋅ ( q s + q r ) (4-64)
∂t

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 221

 Equations for all supported types of medium are presented in the next sections, Thin
Layer, Thin Film, Fracture, and Thin Rod.

THERMALLY THICK APPROXIMATION

This formulation applies to a thin structure where heat transfer mainly follows the
normal direction. The gradient operator is then simplified to

∇T = ∇ n T

This assumption is often valid for thin structures that are thermally resistive compared
to the adjacent domains.

With these assumptions, Equation 4-15 becomes:

∂E
ds ρ + d s ρu ⋅ ∇ n E + ∇ n ⋅ ( q s + q r ) = – d s ( σ:D ) + d s Q + q 0 (4-65)
∂t

q s = – d s k∇ n T (4-66)

where ds is the layer thickness (SI unit: m). The heat source Q is a density distributed
in the layer while q0 is the received out-of-plane heat flux.

In 2D, Equation 4-65 and Equation 4-66 have an additional factor, dz,
to account for the out-of-plane thickness.

When Equation 4-65 is solved in a boundary adjacent to a domain modeling heat

transfer, the two entities exchange a certain amount of heat flux according to:

q0 = n ⋅ q

In this coupling relation, the outgoing heat flux n ⋅ q leaves the domain and is received
in the source term q0 by the adjacent thin layer modeled as a boundary. From the point
of view of the domain, and neglecting thermoelastic effects, the following heat source
is received from the thin structure:

∂E
–n ⋅ q = ds Qs – ds ρ – ( d s ρu ⋅ ∇ n E ) – ∇ n ⋅ ( q s + q r ) (4-67)
∂t

To evaluate the normal gradient operation, ∇n, temperatures Tu and Td are

introduced for the upside and downside of the thin structure boundary. They are
defined from the heat flux across the thin resistive structure. At the middle of the

222 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

thickness, the temperature, T1 ⁄ 2, is approximated by (1 ⁄ 2)(Tu + Td). The term
∇n ⋅ (−dsk∇nT) is then given by:

T d – 2T 1 ⁄ 2 + T u
∇ n ⋅ ( – d s k∇ n T ) ≈ – k s ------------------------------------------
ds

which can be seen as the sum of two contributive sources on the upside and on the
downside of the boundary that compensate:

Tu – Td Td – Tu
– k s -------------------- – k s --------------------
ds ds

Similarly, the time-dependent term can be expressed using Tu and Td by:

∂T ∂T 1 ⁄ 2 1 ∂T u 1 ∂T d
ρC p ------- ≈ ρC p --------------- = --- ρC p ---------- + --- ρC p ----------
∂t ∂t 2 ∂t 2 ∂t

leading to:

ds Q d s ∂T d Tu – Td
– n d ⋅ q d = ----------- – ρC p  ------ ---------- + u d ⋅ n d T d –  – k s -------------------- – ( q r, d ⋅ n d ) (4-68)
2 2 ∂t ds

ds Q d s ∂T u Td – Tu
– n u ⋅ q u = ----------- – ρC p  ------ ---------- + u u ⋅ n u T u –  – k s -------------------- – ( q r, u ⋅ n u ) (4-69)
2  2 ∂t   ds 

Equations for all supported types of medium are presented in the next sections, Thin
Layer, Thin Film, Fracture, and Thin Rod.

UPSIDE, DOWNSIDE, AND EXTERIOR TEMPERATURES

This formulation is provided by the Thermally thick approximation option of the Thin
Layer (Heat Transfer Interface) and Solid (Heat Transfer in Shells Interface) feature.

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 223

 Figure 4-5 shows how Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer
in Shells Interface) splits the temperature into Tu and Td on interior boundaries:

Downside domain Upside domain

of the boundary of the boundary

Td Tu
Thin Layer boundary

Figure 4-5: Upside and downside temperatures at a thin layer applied on an interior
boundary. The thin layer is represented by the gray domain.

On exterior boundaries, it introduces a new degree of freedom represented by the

variable TextFace. Depending on whether the heat domain is on the upside or the
downside of the boundary, TextFace, is equal to Tu or Td and the same thing goes for
the dependent variable T. An example is illustrated in the figure below:

Downside domain
of the boundary

T = Td
Tu = TextFace

Thin Layer boundary

Figure 4-6: Upside and downside temperatures at a thin layer applied on an exterior
boundary.

224 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

FORMULATIONS AVAILABLE WITHIN THE FEATURES
Table 4-3 summarizes the formulations available within the thin structure features of
the Heat Transfer (ht) and Heat Transfer in Shells (htlsh) interfaces.
TABLE 4-3: FORMULATIONS AVAILABLE WITH THE THIN STRUCTURES FEATURES

FEATURE GENERAL THERMALLY THIN THERMALLY THICK

APPROXIMATION APPROXIMATION

Thin layer (ht) Yes Yes Yes

Solid (htlsh) Yes Yes No
Thin film (ht) Yes Yes No
Fluid (htlsh) Yes Yes No
Fracture (ht) Yes Yes No
Porous Medium (htlsh) Yes Yes No

Thin Layer
Thin layers of solid materials can be considered as boundaries when their thickness is
significantly smaller than the typical lengths of the adjacent domains.

GENERAL FORMULATION
With this formulation, multiple sandwiched layers with different material properties
and thicknesses can be modeled. An additional 1D segmented line represents the
multiple layers in the thin structure. In this extra dimension, the governing equation
is derived from Equation 4-59 to give:

∂T s
ρ si C p, si --------- + ∇ t ⋅ q si = Q si (4-70)
∂t

q si = – k si ( ∇tT s + ∇nT s ) (4-71)

Td = ( Ts )L = 0

Tu = ( Ts )L = d
s

where Ts is an auxiliary dependent variable defined on the product space. The

remaining quantities are recalled below:

• ρsi is the density of layer i (SI unit: kg/m3)

• Cp, si is the heat capacity of layer i (SI unit: J/(kg·K))
• ksi is the thermal conductivity of layer i (SI unit: W/(m·K))

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 225

 • Qsi is the heat source applied to layer i (SI unit: W/m3)
• ds is the shell thickness (SI unit: m)

The constraint T = Ts is specified on each side of the extra dimension to connect T to

Ts.

See Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer in
Shells Interface) with Layer type set as General or more information about
the boundary feature solving Equation 4-70 and Equation 4-71.

THERMALLY THIN APPROXIMATION

The Heat Transfer Module supports heat transfer in thermally thin structures in 3D,
2D, and 2D axisymmetry. The material in the thin structure might be a good thermal
conductor for this approximation to be valid. For example, in a printed circuit with
copper traces, where the traces are often good thermal conductors compared to the
board’s substrate material.

The thermally thin approximation is derived from Equation 4-62 to Equation 4-64.
Inside the thin layer, the heat equation becomes:

∂T
d s ρC p, s + ∇t ⋅ qs = ds Q s + q0 (4-72)
∂t

q s = – d s k ∇t T (4-73)

where ds is the layer thickness (SI unit: m).

The heat source Qs is a density distributed in the layer while q0 is the received
out-of-plane heat flux.

In 2D, Equation 4-72 and Equation 4-73 have an additional factor, dz,
to account for the out-of-plane thickness.

From the point of view of the domain, the following heat source, derived from
Equation 4-64, is received from the layer:

∂T
– n ⋅ q = d s Q s – d s ρC p, s ------- – ∇ t ⋅ q s (4-74)
∂t

226 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

This formulation may be used to model efficiently a shell with multiple sandwiched
layers with different material properties and thicknesses, by approximating the
homogenized material properties as weighted averages of constant properties within
each layer:

• The homogenized density is defined as

 ρ i d s, i
i
ρ = ----------------------
-

d s, i
i
• The homogenized heat capacity is defined as

 C p , i d s, i
i
C p = ----------------------------
-

d s, i
i
• The homogenized thermal conductivity is defined as

 ( roti ⋅ A )
T
⋅ k i ⋅ ( rot i ⋅ A )d s, i
i
--------------------------------------------------------------------------------------
-
k =
 d s, i
i
with roti the rotation tensor of layer i:

cos ( θ ) sin ( θ ) 0
rot i = – sin ( θ ) – cos ( θ ) 0
0 0 1

and A the transformation matrix between the local and global coordinates systems:

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 227

 t 1, x t 2, x n x
A = t 1, y t 2, y n y
t 1, z t 2, z n z

See Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer in
Shells Interface) with Layer type set as Thermally thin approximation for
more information about the boundary feature solving Equation 4-74. See
The Heat Transfer in Shells Interface for more information about the
physics interface solving Equation 4-72.

• Heat Transfer in a Surface-Mount Package for a Silicon Chip:

Application Library path
Heat_Transfer_Module/Power_Electronics_and_Electronic_Cooling/surface
_mount_package
• Silica Glass Block Coated with a Copper Layer: Application Library
path Heat_Transfer_Module/Tutorials,_Thin_Structure/copper_layer

THERMALLY THICK APPROXIMATION

When a thin layer is formed of one or more thermally resistive materials, it can be
defined through its thermal resistance:

ds
R s = ------
ks

The heat flux across the thermally thick structure is derived from Equation 4-67 and
gives

1 ∂T d Tu – Td 1
– n d ⋅ q d = – --- d s ρ s C p, s ---------- – k s -------------------- + --- d s Q s (4-75)
2 ∂t ds 2

1 ∂T u Td – Tu 1
– n u ⋅ q u = – --- d s ρ s C p, s ---------- – k s -------------------- + --- d s Q s (4-76)
2 ∂t ds 2

where the u and d subscripts refer to the upside and downside of the layer, respectively.

228 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

When the material has a multilayer structure, ks and ds in the expressions above are
replaced by dtot and ktot, which are defined according to Equation 4-77 and
Equation 4-78:

 dsj
d tot = (4-77)

j=1

d tot
k tot = -----------------
n
- (4-78)
d sj
 -------
k sj
j 1
where n is the number of layers.

See Thin Layer (Heat Transfer Interface) and Solid (Heat Transfer in
Shells Interface) with Layer type set as Thermally thick approximation for
more information about the boundary feature solving Equation 4-75 and
Equation 4-76.

Thin Film
Thin films of fluid can be considered as boundaries of thickness significantly smaller
than the typical lengths of the overall model.

GENERAL FORMULATION
With this formulation, heat transfer is modeled in the whole film, including its
thickness. An additional 1D segmented line represents the thickness in the thin film.
In this extra dimension, the governing equation is derived from Equation 4-59 to give:

∂T s
ρC p --------- + ρC p u ⋅ ( ∇tT s + ∇nT s ) + ∇ t ⋅ q f = Q f (4-79)
∂t

q f = – k ( ∇t T s + ∇n T s ) (4-80)

Td = ( Ts )L = 0

Tu = ( Ts )L = d
f

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 229

 where Ts is an auxiliary dependent variable defined on the product space. The
remaining quantities are recalled below:

• ρ is the density (SI unit: kg/m3)

• Cp is the heat capacity (SI unit: J/(kg·K))
• k is the thermal conductivity (SI unit: W/(m·K))
• Qf is the heat source applied to the film (SI unit: W/m3)
• df is the film thickness (SI unit: m)

The constraint T = Ts is specified on each side of the extra dimension to connect T to

Ts.

See Thin Film (Heat Transfer Interface) and Fluid (Heat Transfer in
Shells Interface) with Thin film model set as General for more information
about the boundary feature solving Equation 4-79 and Equation 4-80.

THERMALLY THIN APPROXIMATION

The thermally thin approximation is derived from Equation 4-62 to Equation 4-64.
Inside the thin layer, the heat equation becomes:

∂T
d f ρC p  ------- + u ⋅ ∇ t T + ∇ t ⋅ q f = d f Q f + q 0 (4-81)
∂t

q f = – d f k∇ t T (4-82)

where df is the film thickness (SI unit: m). The heat source Qf is a density distributed
in the layer while q0 is the received out-of-plane heat flux.

In 2D, Equation 4-72 and Equation 4-73 have an additional factor, dz,
to account for the out-of-plane thickness.

From the point of view of the domain, the following heat source, derived from
Equation 4-64, is received from the layer:

230 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ∂T
– n ⋅ q = d f Q f – d f ρC p ------- – d f ρC p u ⋅ ∇ t T – ∇ t ⋅ q f (4-83)
∂t

See Thin Film (Heat Transfer Interface) and Fluid (Heat Transfer in
Shells Interface) with Thin film model set as Thermally thin approximation
for more information about the boundary feature solving Equation 4-81.
See The Heat Transfer in Films Interface for more information about the
physics interface solving Equation 4-83.

Fracture
When fractures occur in porous media, fluid flow tends to move faster than in the bulk
medium. The transport of heat occurs faster in the fractures that in the surrounding
medium, so in this sense, heat transfer in fractures filled with fluids is more similar to
a highly conductive layer than to a thin thermally resistive layer.

The mass transport in fractures can be modeled as Darcy’s law in a thin sheet of porous
medium:

κ
u = --- ∇ t p
μ

where u is the tangential Darcy’s velocity (SI unit: m/s), κ is the fracture permeability
(SI unit: m2), μ the fluid’s dynamic viscosity (SI unit: Pa⋅s), and ∇tp the tangential
gradient of the fluid’s pressure.

Typically, Darcy’s Law with tangential derivatives is solved to compute mass transport,
so in addition to the fluid properties, the fracture should define its own permeability
(or hydraulic conductivity in case the fluid is water), porosity, and fracture thickness.

For heat transfer in fractures, the fracture also needs to define the density of the porous
sheet, heat capacity, and thermal conductivity. The effective thermal conductivity of
the fracture must be adjusted to the fracture porosity and thermal conductivity of the
fluid. In rocks and geological formations, the fracture might also contain highly
conductive material, different than the bulk porous matrix.

The equation to solve for computing heat transfer in fractures is derived from
Equation 4-62 to Equation 4-64 and using the procedure detailed in Theory for Heat
Transfer in Porous Media to apply the mixture rule on solid and fluid internal energies.
The resulting equations are:

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 231

 ∂T
d fr ( ρC p ) eff + d fr ρC p u ⋅ ∇ t T + ∇ t ⋅ q fr = d fr Q + q 0 (4-84)
∂t

q fr = – d fr k eff ∇ t T (4-85)

Here (ρCp)eff is the effective heat capacity at constant pressure of the fracture-fluid
volume, ρ is the fluid’s density, Cp is the fluid’s heat capacity at constant pressure, qfr
is the conductive heat flux in the fracture-fluid volume, keff is the effective thermal
conductivity of the fluid-fracture mixture, and Q is a possible heat source.

From the point of view of the domain, the following heat source, derived from
Equation 4-64, is received from the fracture:

∂T
– n ⋅ q = d fr Q 0 – d fr ( ρC p ) eff – d fr ρC p u ⋅ ∇ t T – ∇ t ⋅ q fr (4-86)
∂t

See Fracture (Heat Transfer Interface) and Porous Medium (Heat

Transfer in Shells Interface) for more information about the boundary
feature solving Equation 4-86. See The Heat Transfer in Fractures
Interface for more information about the physics interface solving
Equation 4-84.

Thin Rod
The Thin Rod feature is similar to Thin Layer (Heat Transfer Interface) and Solid
(Heat Transfer in Shells Interface) with Layer type set as Thermally thin approximation.
It provides a lumped heat transfer model to model thermally thin rods as edges.

The edge condition reads:

∂T
S ( R ) Q ds = Al Ql – Al ρl Cp, l + ∂ t – ∇t ⋅ ql (4-87)

ql = – Al kl ∇ t T (4-88)

with

2
A l = πr l

232 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

See Thin Rod for node information.

THEORY FOR HEAT TRANSFER IN THIN STRUCTURES | 233

 Theory for the Lumped Thermal
System Interface
The Lumped Thermal System Interface uses a discrete network representation of
thermal systems to model heat transfer by analogy with electrical circuits. The domain
and boundary conditions for heat transfer are idealized by components joined by a
network of perfectly thermally conductive wires.

The analogy with electrical circuits is made by considering the temperature difference
across the network components as the effort variable, and the heat rate through the
network component as the flow variable.

By solving for heat balance within the circuit, the temperatures at the nodes in the
circuit can be determined.

This section presents the underlying theory of the Lumped Thermal System interface,
and the resulting heat transfer equations that hold.

In this section:

• Temperature Differences and Heat Rates in the Lumped Thermal System

• Theory for the Conductive Thermal Resistor Component
• Theory for the Radiative Thermal Resistor Component
• Theory for the Thermal Capacitor and Thermal Mass Components
• Theory for the Heat Pipe Component
• Theory for the Thermoelectric Module Component

Temperature Differences and Heat Rates in the Lumped Thermal

System
The Lumped Thermal System interface uses components to idealize heat transfer in
domains and on boundaries.

For each two-port component, the temperature difference across the component, ΔT,
is expressed as:

ΔT = T p2 – T p1

234 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Note that the temperature difference is attached to a component, and has therefore the
same value at its two ports p1 and p2. It is available for postprocessing, for example
with the lts.R1_DT variable, when considering a Conductive Thermal Resistor
component with name R1.

The heat rate through a two-port component, P, is expressed in different ways

depending on the represented device. Then, the heat rates at ports p1 and p2, Pp1 and
Pp2, are defined from P. The different cases are summarized in Table 4-4.
TABLE 4-4: EXPRESSIONS OF THE TWO-PORT COMPONENTS HEAT RATES IN THE LUMPED THERMAL SYSTEM
INTERFACE

HEAT RATE THROUGH HEAT RATE AT HEAT RATE AT

THE COMPONENT PORT P1 PORT P2

Name P Pp1 Pp2

Expression in thermal resistors and heat pipe
Conductive Thermal ΔT -P P
Resistor – --------
R
Convective Thermal ΔT -P P
Resistor – ---------------
R conv

Radiative Thermal ΔT -P P
Resistor – ------------
R rad

Heat Pipe ΔT -P P
– --------
R
Expression in the other two-port components
Thermal Capacitor ∂ΔT -P P
– C -----------
∂t
Heat Rate Source Psrc 0 P
Thermoelectric P1,src+P2,src P1,src P2,src
Module

Note that for thermal resistors components, the heat rate at port p1 has the same sign
as the temperature difference ΔT across the component. In practice, when the
temperature at port p2 is higher than the temperature at port p1 (ΔT>0), applying
conductive heat transfer between these two ports is equivalent to apply a heat source
(positive heat rate Pp1) at port p1.

For the heat pipe component, the port p1 should correspond to the evaporator side
(hot side), and the port p2 should correspond to the condenser side (cold side).
Therefore, the temperature difference ΔT across the component should be negative in

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 235

 normal operating conditions. With this setting, Pp1 is always negative as well, meaning
that cooling is applied on the hot side.

For the heat rate source component, the heat rate is fully applied on port p2.

For the thermoelectric module component, the heat rates P1,src and P2,src can be
expressed in different ways. See Theory for the Thermoelectric Module Component
for more details.

Theory for the Conductive Thermal Resistor Component

This section presents the underlying theory of the Conductive Thermal Resistor
two-port component.

THERMAL RESISTANCE ANALOGY FOR STEADY CONDUCTION

By analogy with Ohm’s law for electric current, the heat rate P (SI unit: W) in a
material due to the temperature difference ΔT (SI unit: K) is

ΔT
P = – -------- (4-89)
R

where R (SI unit: K/W) is the thermal resistance.

The expression of the thermal resistance used in Equation 4-89 depends on the
geometric configuration.

Plane Shell
When considering steady conduction through a plane shell of surface area A (SI unit:
m²), thickness L (SI unit: m), and constant thermal conductivity k (SI unit: W/(m·K)),
the thermal resistance R is:

L
R = -------
kA

Cylindrical Shell
When considering steady conduction through a cylindrical shell of inner radius ri and
outer radius ro (SI unit: m), height H (SI unit: m), and constant thermal conductivity
k (SI unit: W/(m·K)), the thermal resistance R is:

ro
R = ---------------- ln  -----
1
2πkH  r i 

236 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Spherical Shell
When considering steady conduction through a spherical shell of inner radius ri and
outer radius ro (SI unit: m), and constant thermal conductivity k (SI unit: W/(m·K)),
the thermal resistance R is:

R = ----------  ---- – -----

1 1 1
4πk  r i r o

STEADY CONDUCTION THROUGH A COMPOSITE WALL

When considering several materials, Equation 4-89 can be updated to get a
relationship between the heat transfer rate P and the overall temperature difference
ΔToverall (SI unit: K):

ΔT overall
P = – -------------------------
R tot

where Rtot (SI unit: K/W) is the total thermal resistance.

Thermal Resistances in Series

When considering steady conduction through a composite wall made of n layers of
thermal resistances R1,...,Rn, see Figure 4-7, the total resistance is defined by:
n

R tot =  Ri
i=1

with

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 237

 Li
R i = ---------
ki A

Figure 4-7: Steady conduction through a composite wall made of n layers

This corresponds to the following serial thermal circuit representation:

where ΔToverall = Tn+1 − T1. Note that by analogy with electrical circuits, the elements
placed in series in the circuit share the heat rate (flow variable).

Thermal Resistances in Parallel

When considering steady conduction through a composite wall made of n slabs of
thermal resistances R1,…,Rn, see Figure 4-8, the total resistance is defined by:
–1
 n 
= 
1- 
R tot
  ---
R i
i = 1 

with

238 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 L
R i = -----------
ki Ai

Figure 4-8: Steady conduction through a composite wall made of n slabs.

This corresponds to the following parallel thermal circuit representation:

where ΔToverall = T2 − T1. Note that by analogy with electrical circuits, the elements
placed in parallel in the circuit share the temperature difference (effort variable).

CONVECTIVELY ENHANCED CONDUCTIVITY

See Equivalent Thermal Conductivity Correlations for details about the correlations
used to modify the thermal conductivity to account for convective heat flux.

RADIATION IN OPTICALLY THICK PARTICIPATING MEDIUM

See Rosseland Approximation Theory for details about how to modify the thermal
conductivity to account for radiative heat flux when the optical thickness of the
medium is large.

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 239

 Theory for the Radiative Thermal Resistor Component
Let’s consider a set of opaque diffuse gray surfaces. The net radiative heat rate from
the surface of index i is expressed in function of its radiosity and irradiation:

E bi – J i
P i = --------------------
Ri

where the surface radiative resistance Ri is defined as:

1 – εi
R i = -------------
Ai εi

with Ai the surface area (SI unit: m²) and εi the surface emissivity (dimensionless). See
Theory for Surface-to-Surface Radiation for more details.

This allows a network representation of the radiative heat rate with Ebi and Ji as nodes.
The surface resistance tends toward 0 for a large surface or a surface with large
emissivity, for which the blackbody approximation holds.

When considering the surface of index i as part of an enclosure, the net radiative heat
rate is expressed as:
n n
E bi ( J i – J j )
Pi =  A i F ij ( J i – J j ) =  ------------------------------
R ij
-
j=1 j=1

where the space resistance is

1
R ij = -------------
A i F ij

where Fij is the view factor for surfaces i and j.

TWO-SURFACE ENCLOSURE
In the network representation of the radiative heat transfer in a two-surface enclosure,
the surface resistance and the space resistance are connected in a serial way. Therefore
the total radiative resistance is expressed as follows:

1 – ε1 1 1 – ε2
R 12 = -------------- + ---------------- + --------------
A 1 ε 1 A 1 F 12 A 2 ε 2

And the net radiation exchange is expressed as:

240 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 4 4
σ ( T1 – T2 )
P 12 = ----------------------------
R 12

For specific configurations, the surface areas can be evaluated from the geometric
dimensions and the view factors are equal to 1. Therefore the expression above
simplifies (see Ref. 21 for details).

Concentric Spheres
The surface areas A1 and A2 verify

A1 r 2
------- =  ----1-
A2  r 2

And the net radiation exchange is

4 4
A1 σ ( T1 – T2 )
P 12 = ----------------------------------------2-
1- 1 – ε2 r1
---- + --------------  -----
ε1 ε2 r2

Large (infinite) Parallel Planes

The surface areas A1 and A2 verify

A1 = A2

And the net radiation exchange is

4 4
A1 σ ( T1 – T2 )
P 12 = ------------------------------------
1- ----
---- 1
+ -–1
ε1 ε2

Long (infinite) Concentric Cylinders

The surface areas A1 and A2 verify

A1 r
------- = ----1-
A2 r2

And the net radiation exchange is

4 4
A1 σ ( T1 – T2 )
P 12 = --------------------------------------
1- 1 – ε2 r1
---- + --------------  -----
ε1 ε2 r2

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 241

 These expressions are predefined options in the Radiative Thermal Resistor feature.
Further simplifications can be made on the net radiation exchange when the blackbody
assumption holds (ε1=ε2=1).

Theory for the Thermal Capacitor and Thermal Mass Components

This section presents the underlying theory of the Thermal Capacitor two-port
component and the Thermal Mass one-port component.

THERMAL CAPACITANCE
The thermal capacitance C (SI unit: J/K) is a measure of how much heat a body can
store. It is defined as:

C = VρC p = mC p

with V the volume (SI unit: m3), ρ the density (SI unit: kg/m3), Cp the heat capacity
at constant volume (SI unit: J/(kg·K)), and m the mass (SI unit: kg).

LUMPED THERMAL CAPACITANCE MODEL

Under certain conditions, the temperature T of a solid body subjected to convective
heat transfer on its boundaries may be considered as homogeneous, depending only
on time, using the thermal capacitance:

hA
T – T∞ – -------- t
- = e C
-------------------
T0 – T∞

where:

• T0 is the initial body temperature

• T ∞ is the surrounding fluid temperature
• h is the convective heat transfer coefficient (SI unit: W/(m²·K))
• A is the surface of convective heat transfer (SI unit: m²)

This approximation, referred as the lumped thermal capacitance model (see Ref. 41),
holds when h, C, and A are constant and the gradients of temperature within the body
are expected to be smaller than the gradients of temperature between the body and the
surrounding. It happens for example when the thermal contact between the solid and
the fluid is poor, or when the solid is a good thermal conductor.

This is characterized by the Biot number Bi (dimensionless):

242 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 hL
Bi = --------
k

where L is a relevant length scale of the body, and k is its thermal conductivity.

The lumped thermal capacitance model is usually assumed to be valid when Bi<0.1.

Regarding the network representation of thermal systems, the Biot number may be
used to determine how many nodes should be included, assuming an homogeneous
temperature distribution in the corresponding domains.

Theory for the Heat Pipe Component

Heat pipes are devices which use the latent heat of evaporation and condensation of a
fluid to transfer heat from a source to a sink with high efficiency.

DEVICE DESCRIPTION
Although various geometric configurations are available, a heat pipe includes a vapor
channel delimited by a solid wall, with a porous wick in between, see Figure 4-9 below.

Figure 4-9: Heat pipe components.

The working fluid flows from the hot side to the cold side of the heat pipe under vapor
state in the core channel, and under liquid state on the way back through the porous
wick by capillary action. It evaporates when leaving the wick to the core channel on
the hot side, named evaporator side, and condensates when entering the wick on the
cold side, named condenser side.

The heat is transferred by conduction through the wall, and by conduction and
convection in the wick and in the vapor channel. The latent heat absorbed and released
by the evaporation and condensation of the working fluid makes heat pipes very
effective heat transfer devices, with large effective thermal conductivities.

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 243

 NETWORK REPRESENTATION
Whereas the full modeling of heat transfer in a heat pipe necessitates the modeling of
several processes (vapor transport in the channel, liquid transport in the porous wick,
and phase change at the interface between the wick and the vapor channel), a network
representation allows to estimate the temperature drop between the evaporator and
the condenser sides by considering a set of relevant thermal resistances.

By making the analogy with electrical circuits, the total thermal resistance R (SI unit:
W/K) is expressed as:

R = R wall, e + R wick, e + R lv, e + R a + R lv, c + R wick, c + R wall, c

where

1
R a = ------------------------------------------------------------------
1 1 1
------------ + -------------------- + -------------------
R v, a R wick, a R wall, a

with

• Rwall,e the thermal resistance of the wall on evaporator side

• Rwick,e the thermal resistance of the wick on evaporator side
• Rlv,e the thermal resistance of the liquid-vapor interface on evaporator side
• Rlv,c the thermal resistance of the liquid-vapor interface on condenser side
• Rwick,c the thermal resistance of the wick on condenser side
• Rwall,c the thermal resistance of the wall on condenser side
• Rv,a the thermal resistance of the vapor channel in adiabatic section
• Rwick,a the thermal resistance of the wick in adiabatic section
• Rwall,a the thermal resistance of the wall in adiabatic section

By making assumptions on the relative magnitude of the thermal resistances, the

expression of R reduces to:

R = R wall, e + R wick, e + R wick, c + R wall, c

And the temperature drop ΔT = Tcondenser − Tevaporator is then expressed as

P
ΔT = – ----
R

where P (SI unit: W) is the operating power of the heat pipe.

244 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

This corresponds to the following network representation:

Figure 4-10: Thermal network representation of a heat pipe.

See Ref. 18 for details about the network representation of heat pipes and the
expressions of effective thermal conductivity of the wick used to express its thermal
resistance.

Theory for the Thermoelectric Module Component

Thermoelectric modules are devices which use thermoelectricity (Peltier effect) to
transfer heat from a source to a sink.

DEVICE DESCRIPTION
They are composed of thermocouples, each consisting of p-type and n-type
semiconductors, connected electrically in series and thermally in parallel and
sandwiched between two high thermally conductive but low electrically conductive
ceramic plates, as described on Figure 4-11 below.

Figure 4-11: Thermoelectric module schematic description.

See Theory for the Thermoelectric Effect Interface for details about the Peltier effect.

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 245

 HEAT RATE
In heating and cooling operating modes of the thermoelectric module (that is, no heat
generation), the heat rates at the cold and hot junctions of each thermocouple are due
to Peltier effect, Joule heating, thermal conduction, and heat storage:

• Hot junction:
2
Re I ΔT ∂T hot
P hot = – SIT hot + ------------ + -------- – C hot --------------- (4-90)
2 R ∂t

• Cold junction:

2
Re I ΔT ∂T cold
P cold = SIT cold + ------------ – -------- – C cold ----------------- (4-91)
2 R ∂t

where

• S (SI unit: V/K) is the Seebeck coefficient of the thermocouple

• I (SI unit: A) is the electric current operating in the module
• Re (SI unit: Ω) is the electrical resistance of the thermocouple
• ΔT (SI unit: K) is the temperature difference between the two sides of the module
• R (SI unit: K/W) is the thermal resistance of the thermocouple
• Chot and Ccold (SI unit: J/K) are the thermal capacitances on each side of the
module

For a steady-state problem the temperature does not change with time and the heat
storage terms disappear.

The Peltier effect is a cooling effect at the hot junction, and a heating effect at the cold
junction.

In Equation 4-90 and Equation 4-91 the assumption of a symmetric distribution of

Joule heating at the hot and cold junctions has been made.

THERMOCOUPLE ELECTRICAL AND THERMAL RESISTANCES

As the p-type and n-type semiconductors are connected electrically in series, the
electrical resistance Re of each thermocouple is expressed as:

R e = R e, p + R e, n

where Re,p and Re,n (SI unit: Ω) are the electrical resistances of the p-type and n-type
semiconductors.

246 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

And as they are connected thermally in parallel, the thermal resistance R of each
thermocouple is expressed as:

1- 1 1
--- = ---- + ----
R Rp Rn

where Rp and Rn (SI unit: K/W) are the thermal resistances of the p-type and n-type
semiconductors.

THERMOCOUPLE SEEBECK COEFFICIENT

Finally, the Seebeck coefficient S of each thermocouple is expressed as:

S = Sp – Sn

TOTAL HEAT RATE

The total heat rate through the module is obtained by summing the heat rates in each
thermocouples.

NETWORK REPRESENTATION
The following network representation corresponds to Equation 4-90 and
Equation 4-91:

Figure 4-12: Thermal network representation of a thermoelectric module.

See Ref. 18 and Ref. 19 for details about the network representation of thermoelectric
modules.

PERFORMANCE GRAPHS FOR THERMOELECTRIC COOLERS

It is usual to characterize thermoelectric coolers through graphs showing the heat
removed from the cold side of the module, Q, in function of the difference of
temperature ΔT between the hot and cold sides, the input intensity I, and the ambient
temperature.

THEORY FOR THE LUMPED THERMAL SYSTEM INTERFACE | 247

 Linearized Model at Given Intensity and Ambient Temperature
This model establishes a linear relation between the removed heat and the temperature
difference, as shown on the figure below:

Removed
Heat, Q

Qmax

Temperature Difference, ΔT

ΔTmax

Qmax is the maximum amount of heat that the thermoelectric cooler can remove,
when there is no temperature difference between the two sides of the module.

The linear curves from (ΔT=0, Q=Qmax) to (ΔT=ΔTmax, Q=0) define the removed
heat Q as:

ΔT
Q = Q max –  Q max ⋅ ------------------
ΔT max

Note that this relation is valid when the temperature difference satisfies:

0 ≤ ΔT ≤ ΔT max

Then the powers applied at the cold and hot sides are respectively:

• P=-Q (removed heat)

• P=Q + ReI2 (waste heat, including Joule heating)

For cases where the current through the device varies, the performance parameters
should be made a function of the current.

General Model
The removed heat may be defined as a more general function of the temperature
difference and other parameters such as the temperature at the hot side.

248 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Surface-to-Surface
Radiation
In addition to conduction and convection, the third mechanism for heat transfer is
radiation. Consider an environment with fully transparent or fully opaque objects.
Thermal radiation denotes the stream of electromagnetic waves emitted from a body
at a certain temperature.

The Surface-to-Surface Radiation Interface theory is described in this section:

• Deriving the Radiative Heat Flux for Opaque Surfaces

• Deriving the Radiative Heat Flux for Semitransparent Surfaces
• Wavelength Dependence of Surface Properties
• Directional Dependence of Surface Properties
• The Radiosity Method for Diffuse-Gray Surfaces
• The Radiosity Method for Diffuse-Spectral Surfaces
• View Factor Evaluation

Deriving the Radiative Heat Flux for Opaque Surfaces

In Figure 4-13, consider a point P located on a surface that has an emissivity ε, diffuse
reflectivity ρd, specular reflectivity ρs, absorptivity α, refractive index n, and
temperature T. The body is assumed opaque, which means that no radiation is
transmitted through the body. This is true for most solid bodies.

G ρsG J = ρdG + εn2σT4

εn2σT4

ρdG

P P

Figure 4-13: Incoming irradiation (left), outgoing radiosity (right).

The total incoming radiative flux at P is called irradiation and denoted G. The total
diffuse outgoing radiative flux at P is called radiosity and denoted J. This radiosity is
the sum of diffusively reflected and emitted radiation:

THEORY FOR SURFACE-TO-SURFACE RADIATION | 249

 J = ρ d G + εe b ( T ) (4-92)

According to the Stefan-Boltzmann law, eb(T) is the power radiated across all
wavelengths and depends on the forth power of the temperature:

e b ( T ) = n 2 σT 4

The net inward radiative heat flux, q, is then given by the difference between the
irradiation and the outgoing radiation (radiosity and specular reflected radiation):

q = G – ( J + ρs G )

which can be also written as

q = ( 1 – ρ s )G – J (4-93)

Using Equation 4-92 and Equation 4-93, J can be eliminated and a general expression
is obtained for the net inward heat flux into the opaque body based on G and T.

q = ( 1 – ( ρ d + ρ s ) )G – εe b ( T ) (4-94)

Most opaque bodies also behave as ideal gray bodies, meaning that the absorptivity and
emissivity are equal, and the reflectivity ρd+ρs is therefore obtained from the following
relation:

α = ε = 1 – ( ρd + ρs ) (4-95)

Thus, for ideal gray bodies, q is given by:

q = ε ( G – eb ( T ) ) (4-96)

This is the expression used for the radiative boundary condition.

Deriving the Radiative Heat Flux for Semitransparent Surfaces

In Figure 4-13, consider a point P located on a semitransparent surface that has an
emissivity εu, diffuse reflectivity ρd,u, specular reflectivity ρs,u, refractive index nu, and
temperature Tu on the upside, and an emissivity εd, diffuse reflectivity ρd,d, specular

250 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

reflectivity ρs,d, refractive index nd, and temperature Td on the downside. As the
surface is assumed semitransparent, some radiation is transmitted through the body.

Ju = ρd,uGu + εunu2σT4
εunu2σT4
Gu ρs,uGu
ρd,uGu

τdGd

P P
Gd

Figure 4-14: Upside and downside incoming irradiation (left), upside outgoing radiosity
(right). The downside outgoing radiosity is defined in a similar way.

The total incoming radiative flux at P is called irradiation, and is denoted Gu on the
upside and Gd on the downside. The total diffuse outgoing radiative flux at P is called
radiosity and denoted Ju on the upside and Jd on the downside. This radiosity is the
sum of diffusively reflected radiation, emitted radiation and transmitted radiation
coming from the other side of the semitransparent layer:

J u = ρ d, u G u + ε u e b, u ( T u ) (4-97)

J d = ρ d, d G d + ε d e b, d ( T d ) (4-98)

The net inward radiative heat fluxes on the upside and downside, qu and qd, are then
given by the difference between the irradiation and the radiosity:

q u = ( 1 – ρ s, u – τ u )G u – J u (4-99)

q d = ( 1 – ρ s, d – τ d )G d – J d (4-100)

Bodies are considered to behave as ideal gray bodies, meaning that the absorptivity and
emissivity are equal, and the reflectivity ρs is therefore obtained from the following
relation:

ε u + ρ d , u = 1 – ρ s, u – τ u (4-101)

ε d + ρ d, d = 1 – ρ s, d – τ d (4-102)

THEORY FOR SURFACE-TO-SURFACE RADIATION | 251

 Using Equation 4-97 to Equation 4-102, Ju and Jd can be eliminated and a general
expression is obtained for the net inward heat fluxes into the semitransparent body
based on Gu, Gd, Tu and Td:

q u = ε u ( G u – e b, u ( T u ) ) (4-103)

q d = ε d ( G d – e b, d ( T d ) ) (4-104)

Thus, for ideal gray bodies, q is given by:

q = ε u ( G u – e b, u ( T u ) ) + ε d ( G d – e b, d ( T d ) ) (4-105)

This is the expression used for the radiative boundary condition.

Incident rays which angle of incidence (measured between the ray and the normal to
the surface) is higher than the critical angle are not transmitted, regardless the
transmittance of the surface. They contribute to total reflection instead. Hence the
directional transmissivity coefficient can be defined as


 τ if θ ≤ θ c
τ(θ) = 
 0 if θ > θ c


where θ c is the critical angle. Using the following relation

ρs ( θ ) + τ ( θ ) = 1 – ( ε + ρd )

we can establish


 ρ s if θ ≤ θ c
ρs ( θ ) = 
 ρ s + τ if θ > θ c


Wavelength Dependence of Surface Properties

The surface properties for radiation, the emissivity, specular reflectivity and specular
transmissivity can be dependent on the incident angle, the surface temperature, or the
radiation wavelength. The emissivity, specular reflectivity and specular transmissivity
are defined in Ref. 20.

252 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The Surface-to-Surface Radiation interface in the Heat Transfer module implements
the radiosity method that enables arbitrary temperature and wavelength dependent
surface properties. The dependence of the surface properties on the incident angle can
be considered by using the ray shooting method (Directional Dependence of Surface
Properties).

PLANCK SPECTRAL DISTRIBUTION

The Planck’s distribution of emissive power for a blackbody in vacuum is given as a
function of surface temperature and wavelength.

The blackbody hemispherical emissive power (SI unit: W/(m3·sr)), is denoted

eb, λ(λ, T), and defined as (1-37 in Ref. 20):

2πn 2 C 1
e b, λ ( λ, T ) = ------------------------------ (4-106)
C2
 ------- 
λ e – 1
5  λT
 
 

where:

• the two constants C1 (SI unit: W·m2/sr) and C2 (SI unit: m·K) are given by

2 hc 0
C 1 = hc 0 C 2 = ---------
kB

• h is the Planck constant (SI unit: J·s)

• kB is the Boltzmann constant (SI unit: J/K)
• c0 is the speed of the light in vacuum (SI unit: m/s)
• λ is the wavelength in vacuum (SI unit: m)
• n is the refractive index of the media (SI unit: 1), equal to 1 in vacuum

THEORY FOR SURFACE-TO-SURFACE RADIATION | 253

 Figure 4-15 and Figure 4-16 show the hemispherical spectral emissive power for a
blackbody at 5780 K (the Sun’s blackbody temperature) and for a blackbody at 300 K.
The dotted vertical lines delimit the visible spectrum (from 0.4 µm to 0.7 µm).

Figure 4-15: Planck distribution of a blackbody at 5780 K.

Figure 4-16: Planck distribution of a blackbody at 300 K.

The integral of eb, λ(λ, T) over a spectral band represents the power radiated on the
spectral band and is defined by

λ2 ∞
λ eb, λ ( λ, T ) dλ
1
= F λ1 T → λ2 T 0 eb, λ ( λ, T ) dλ
where F λ1 T → λ 2 T is the fractional blackbody emissive power,

254 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 λ2
λ eb, λ ( λ, T ) dλ
F λ 1 T → λ2 T = ----------------------------------------
∞
1
-

0 eb, λ ( λ, T ) dλ
Recall the Stefan-Boltzmann law that computes the power radiated across all
wavelengths:

∞
0 eb, λ ( λ, T ) dλ = e b ( T ) = n 2 σT 4

where n is the refractive index, and σ is the Stefan-Boltzmann constant equal to

5.67 ⋅ 10-8 W/(m2·K4). The power radiated in the spectral band [λ1, λ2] becomes:

λ2
 λ e b , λ ( λ , T ) dλ = F λ T → λ T e b ( T )
1
1 2

The function eb(T) is available as a predefined function via rad.feb(T)

in the Surface-to-Surface Radiation interface.

Notice that:

F λ 1 T → λ2 T = F 0 → λ2 T – F 0 → λ 1 T and F 0 → ∞ = 1

The figure below shows the value of F 0 → λT for different values of λT.

THEORY FOR SURFACE-TO-SURFACE RADIATION | 255

 GRAY SURFACES
Gray surfaces correspond to the hypothesis that surface properties are independent of
the radiation wavelength. This corresponds to Constant wavelength dependence in the
Surface-to-Surface Radiation interface.

The assumption that the surface emissivity is independent of the radiation wavelength
is often valid when most of the radiative power is concentrated on a relatively narrow
spectral band. This is likely the case when the radiation is emitted by a surface at
temperatures in limited range.

This setting is rarely applicable if there is solar radiation.

SOLAR AND AMBIENT SPECTRAL BANDS

When solar radiation is part of the model, it is possible to enhance a diffuse-gray
surface model by considering two spectral bands: one for short wavelengths and one
for large wavelengths. This corresponds to Solar and Ambient wavelength dependence
in the Surface-to-Surface Radiation interface.

It is interesting to notice that about 97% of the radiated power from a blackbody at
5800 K is at wavelengths of 2.5 µm or shorter, and 97% of the radiated power from a
blackbody at 700 K is at wavelengths of 2.5 µm or longer (see Figure 4-17).

Figure 4-17: Normalized Planck distribution of blackbodies at 700 K and 5800 K.

Many problems have a solar load, but the peak temperatures are below 700 K.

256 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

In such cases, it is appropriate to use a two-band approach with

• A solar band for wavelengths shorter than 2.5 µm

• An ambient band for wavelengths above 2.5 µm

For each surface, properties are then described in terms of a solar absorptivity and an
emissivity.

Solar irradiation,
λ < 2.5 µm

Reradiation to
surroundings,

λ > 2.5 µm

Figure 4-18: Absorption of solar radiation and emission to the surroundings.

By splitting the bands at the default of 2.5 μm, the fraction of absorbed solar radiation
on each surface is defined primarily by the solar absorptivity.

THEORY FOR SURFACE-TO-SURFACE RADIATION | 257

 The reradiation at longer wavelengths (objects below ~700 K) and the reabsorption of
this radiation is defined primarily via the emissivity.

Emissivity

Wavelength

Figure 4-19: Solar and ambient spectral band approximation of the surface emissivity by
a constant per band emissivity.

GENERAL SPECTRAL SURFACES

Spectral surfaces correspond to the hypothesis that surface properties are wavelength
dependent. This corresponds to Multiple spectral bands wavelength dependence in the
Surface-to-Surface Radiation interface.

The Heat Transfer Module supports constant surface properties per spectral bands and
to adjust spectral intervals endpoints.

Emissivity

λ1 λ2 λ3 Wavelength

The multiple spectral bands approach is used in cases when the surface properties vary
significantly over the bands of interest.

Directional Dependence of Surface Properties

The Surface-to-Surface Radiation interface in the Heat Transfer module implements
the ray shooting method, that enables to set the dependence of the emissivity, of the
specular reflectivity and of the specular transmissivity on the incident angle.

258 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The polar and azimuthal angles, θ∈[0,π/2] and ϕ∈[−π,π], are defined below:

e3

θ
e2
ϕ
e1

Figure 4-20: Definition of polar and azimuthal angles

The implementation supports directional surface properties that can be decomposed

as the sum of two expressions. The first is defined as a function of θ and ϕ only, and
the other as independent of θ and ϕ. The latter may depend on any other model
quantity like the temperature, or the spatial coordinates.

• Emissivity
ε tot ( θ, ϕ, T, x ) = f ε ( θ, ϕ ) + ε ( T, x )

• Transmissivity
τ tot ( θ, ϕ, T, x ) = f τ ( θ, ϕ ) + τ ( T, x )

• Specular reflectivity
ρ s, tot ( θ, ϕ, T, x ) = 1 – ρ d – ε tot ( θ, ϕ, T, x ) – τ tot ( θ, ϕ, T, x )

Note that no dependence is defined for the diffuse reflectivity, ρd.

The Radiosity Method for Diffuse-Gray Surfaces

The heat transfer by radiation is combined with convective and conductive heat
transfer through a source term added to the heat equation along with the other
contributions from the heat flux and boundary heat source boundary conditions.
Recalling Equation 4-93, this source account for the difference between incident
radiation, or irradiation, G, and radiation leaving the surface, or radiosity, J:

q = ( 1 – ρ s )G – J

THEORY FOR SURFACE-TO-SURFACE RADIATION | 259

 The radiosity, J, is given in Equation 4-92. It is the sum of diffusely reflected and
emitted radiation. For diffuse-gray surfaces, J is defined by:

J = ρ d G + εe b ( T )

Here

• G is the incoming radiative heat flux, or irradiation (SI unit: W/m2)

• ε is the surface emissivity (SI unit: 1), a dimensionless number in the range 0 ≤ ε ≤ 1.
The diffuse-gray surface hypothesis corresponds to surfaces where ε is independent
of the radiation wavelength.
• eb(T) is the blackbody hemispherical total emissive power (SI unit: W/m2).
• T is the surface temperature (SI unit: K).

The irradiation, G, at a given point is split into three contributions according to:

G = G m + G ext + G amb (4-107)

where:

• Gm is the mutual irradiation, coming from other boundaries in the model

(SI unit: W/m2).
• Gext is the irradiation from external radiation sources (SI unit: W/m2). It is the sum
of the products, for each external source, of the external heat sources view factor
Fext by the corresponding source radiosity:

G ext =  Fext Ps +  Fext q0, s

The first term of the sum gathers radiation sources located on a point. The second
term stands for directional radiative sources.
• Gamb is the ambient irradiation (SI unit: W/m2), defined as:
G amb = ε amb F amb e b ( T amb )

• εamb is the ambient emissivity; (SI unit: 1), a dimensionless number in the range
0 ≤ εamb ≤ 1. εamb < 1 means that part of the energy coming from radiative bodies
is not absorbed by the ambient air.

260 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• Famb is an ambient view factor; its value is equal to the fraction of the field of view
that is not covered by other boundaries. Therefore, by definition, 0 ≤ Famb ≤ 1 at
all points.
• Tamb is the assumed far-away temperature (SI unit: K) in the directions included in
Famb.

The Surface-to-Surface Radiation Interface includes the following radiation types:

• Diffuse Surface (Surface-to-Surface Radiation Interface) is the default radiation

type. It requires accurate evaluation of the mutual irradiation, Gm. The incident
radiation at one point on the boundary is a function of the radiosity, J, at every
other point in view. The radiosity, in turn, is a function of Gm, which leads to an
implicit radiation balance:
J = ρ d ( G m ( J ) + G ext + G amb ) + εe b ( T ) (4-108)

• Diffuse Mirror (Surface-to-Surface Radiation Interface) is a variant of the Diffuse

Surface radiation type with ε = 0. Reradiation surfaces are common as an
approximation of a surface that is well insulated on one side and for which
convection effects can be neglected on the opposite (radiating) side (see Ref. 21).
It resembles a mirror that absorbs all irradiation and then radiates it back in all
directions.
• Prescribed Radiosity (Surface-to-Surface Radiation Interface) makes it possible to
specify graybody radiation. The radiosity expression is then εeb(T). A user-defined
surface radiosity expression can also be defined.

• Opaque Surface (Surface-to-Surface Radiation Interface) is available when the

surface-to-surface radiation method is Ray shooting. It accounts for specular
reflection. The conservation equation reads

ε + ρd + ρs = 1 ,

and the radiosity reads as in Equation 4-108.

• Semitransparent Surface (Surface-to-Surface Radiation Interface) is available when

the surface-to-surface radiation method is Ray shooting. It accounts for reflection,
transmission and the conservation equation reads

ε + ρd + ρs + τ = 1 ,

and the radiosities read

THEORY FOR SURFACE-TO-SURFACE RADIATION | 261

 J u = ( ρ d,u ( G m,u ( J u ) + G ext,u + G amb,u ) + ε u e b ( T ) )
(4-109)
+ τ d ( G m,d ( J d ) + G ext,d + G amb,d )

J d = ( ρ d,d ( G m,d ( J d ) + G ext,d + G amb,d ) + ε d e b ( T ) )

(4-110)
+ τ d ( G m,u ( J u ) + G ext,u + G amb,u )

The Surface-to-Surface Radiation interface handles the radiosity J as a shape function

unless J is prescribed.

Whereas diffuse and specular reflectivities are handled by the different

features of the Surface-to-Surface Radiation interface, only diffuse
emissivity is considered.

The Radiosity Method for Diffuse-Spectral Surfaces

For a general diffuse-spectral surface:

J =  ( ρd ( λ, T )G ( λ ) + ε ( λ, T )eb, λ ( λ, T ) ) dλ
0

where

• ε(λ, T) and ρd(λ, T) are is the hemispherical spectral surface emissivity and diffuse
reflectivity, dimensionless quantities in the range [0,1]. Diffuse-spectral surface
corresponds to a surface properties are dependent on the radiation wavelength and
surface temperature.
• T is the surface temperature (SI unit: K).
• eb, λ(λ, T) is the blackbody hemispherical emissive power (SI unit: W/(m3·sr))
defined in Equation 4-106.

The Surface-to-Surface Radiation Interface assumes that the surface emissivity and
opacity properties are constant per spectral band. It defines N spectral bands (N = 2
when solar and ambient radiation model is used),

262 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

  B i = [λ i – 1,λ i] for 1 ≤ i ≤ N

 λ0 = 0

 λN = ∞

so that the radiosity has a custom definition in each interval:

N

J =  Ji
i=1
J i = ρ d, i G i + ε i e b ( T )

The surface properties can then be defined per spectral band:

• Surface emissivity on Bi: εi(T) = ε(λ, T) for λ in the interval Bi

• Surface diffuse reflectivity on Bi: ρd,i(T) = ρd(λ, T) for λ in the interval Bi
• Ambient irradiation on Bi, assuming that the ambient fractional emissive power
corresponds to the one of a blackbody at temperature Tamb:
λi
G amb, i = λ = λ i–1
G amb ( λ ) dλ = F λi – 1 T → λi T F amb, i ε amb, i e b ( T amb )

• External radiation sources on Bi with q0, s, i and Ps, i the external radiation source
heat flux and heat rate, respectively, over Bi:
λi
G ext, i = λ = λ i–1
G ext ( λ ) dλ = F ext, i ( i s )q 0, s, i

or
λi
G ext, i = λ = λ i–1
G ext ( λ ) dλ = F ext, i ( i s )P s, i

When the external source fractional emissive power corresponds to the one of a
blackbody at Text, external radiation sources on Bi can be defined from the external
radiation source heat flux, q0, s, and heat rate, Ps, over all wavelengths:
G ext, i = F ext, i F λi – 1 T → λi T ( i s )q 0, s

or
G ext, i = F ext, i F λi – 1 T → λi T ( i s )P s

THEORY FOR SURFACE-TO-SURFACE RADIATION | 263

 The Surface-to-Surface Radiation Interface includes the following radiation types:

• Diffuse Surface (Surface-to-Surface Radiation Interface) is the default radiation

type. The incident radiation over the Bi spectral band at one point of the boundary
is a function of the radiosity, Ji (SI unit: W/m2), at every other point in view. The
radiosity, in turn, is a function of Gm, i, which leads to an implicit radiation balance:
J i = ρ d, i ( G m, i ( J i ) + G ext, i + G amb, i ) + ε i e b ( T ) (4-111)

• Diffuse Mirror (Surface-to-Surface Radiation Interface) is a variant of the Diffuse

Surface radiation type with εi = 0. Reradiation surfaces are common as an
approximation of a surface that is well insulated on one side and for which
convection effects can be neglected on the opposite (radiating) side (see Ref. 21).
It resembles a mirror that absorbs all irradiation and then radiates it back in all
directions.
• Prescribed Radiosity (Surface-to-Surface Radiation Interface) makes it possible to
specify the surface radiation for each spectral band. Using the graybody radiation
definition, the radiosity is then F λ1 T → λ2 T e b ( T ) . A user-defined surface radiosity
expression can also be defined.

• Opaque Surface (Surface-to-Surface Radiation Interface) is available with Ray

shooting as Surface-to-surface radiation method. It accounts for specular reflection.
The conservation equation reads

ε i + ρ d , i + ρ s, i = 1

and the radiosity reads as in Equation 4-111.

• Semitransparent Surface (Surface-to-Surface Radiation Interface) is available with

Ray shooting as Surface-to-surface radiation method. It accounts for reflection,
transmission and the conservation equation reads

ε i + ρ d , i + ρ s, i + τ i = 1

and the radiosities read

J u, i = ρ d, u, i ( G mu, i ( J u, i ) + G ext, u, i + G amb, u, i ) + ε u, i e b ( T )

(4-112)
+ τ d, i ( G md, i ( J d, i ) + G ext, d, i + G amb, d, i )

J d, i = ρ d, u, i ( G md, i ( J d, i ) + G ext, d, i + G amb, d, i ) + ε d, i e b ( T )

(4-113)
+ τ u, i ( G mu, i ( J u, i ) + G ext, u, i + G amb, u, i )

264 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The Surface-to-Surface Radiation interface handles the radiosity Ji as a shape function
unless Ji is prescribed.

Whereas diffuse and specular reflectivities are handled by the different

features of the Surface-to-Surface Radiation interface, only diffuse
emissivity is considered.

View Factor Evaluation

The strategy for evaluating view factors is central to any radiation simulation. Loosely
speaking, a view factor is a measure of how much influence the radiosity at a given part
of the boundary has on the irradiation at some other part.

The quantities Gm and Famb in Equation 4-108 are not strictly view factors in the
traditional sense. Instead, Famb is the view factor of the ambient portion of the field
of view, which is considered to be a single boundary with constant radiosity

J amb = e b ( T amb )

On the other hand, Gm is the integral over all visible points of a differential view factor,
multiplied by the radiosity of the corresponding source point. In the discrete model,
think of it as the product of a view factor matrix and a radiosity vector. This is, however,
not necessarily the way the calculation is performed.

Consider a point P on a surface as in Figure 4-21. It can be seen by points on other

surfaces such as S′ in the figure, as well as the ambient surrounding, Samb. Assume

THEORY FOR SURFACE-TO-SURFACE RADIATION | 265

 that the points on S′ have a local radiosity, J′, while the ambient surrounding has a
constant temperature, Tamb.

Samb

S′

J′
P′

r
n′ n

Jamb
Samb

Figure 4-21: Example geometry for surface-to-surface radiation.

The mutual irradiation at point P is given by the following surface integral:

( – n′ ⋅ r ) ( n ⋅ r )
Gm = S′ -------------------------------------
πr
4
- J′ ds

The heat flux that arrives from P′ depends on the local radiosity J′ projected onto P.
The projection is computed using the normal vectors n and n′ along with the vector
r, which points from P to P′.

The ambient view factor, Famb, is determined from the integral of the surrounding
surfaces S′, here denoted as F′:

( – n′ ⋅ r ) ( n ⋅ r )
F amb = 1 – F ′ = 1 – S′ -------------------------------------
πr
4
- ds

The two last equations plug into Equation 4-107 to yield the final equation for
irradiative flux.

The equations used so far apply to the general 3D case. 2D geometries result in simpler
integrals. For the 2D case, the resulting equations for the mutual irradiation and
ambient view factor are

266 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ( – n′ ⋅ r ⊥ ) ( n ⋅ r ⊥ )
Gm = S ′ --------------------------------------------
2r
3
- J′ ds (4-114)
⊥
⊥

( – n′ ⋅ r ⊥ ) ( n ⋅ r ⊥ )
F amb = 1 – S ' --------------------------------------------
2r
3
- ds
⊥
⊥

where the integral over S⊥′ denotes the line integral along the boundaries of the 2D
geometry.

In axisymmetric geometries or when a symmetry plane is defined, the irradiation and

ambient view factor cannot be computed directly from a closed-form expression.
Instead, a virtual geometry must be constructed, and the view factors evaluated
according to Equation 4-114. For cases with specularly reflective surfaces, specular
view factors depends also on specular reflectivities of surrounding surfaces, see Ref. 23.

A separate evaluation is performed for each unique point where Gm or Famb is

requested, typically for each quadrature point during solution. Differential view factors
are normally computed only once, the first time they are needed, and then stored in
memory until next time the model definition or the mesh is changed.

The Heat Transfer Module supports the following surface-to-surface radiation

methods, which are selected in the Radiation Settings section in a Heat Transfer
interface:

• Hemicube
• Direct area integration
• Ray shooting

View factors are always calculated directly from the mesh, which is a
polygonal representation of the geometry. To improve the accuracy of the
radiative heat transfer simulation, the mesh must be refined rather than
raising the element order.

VIEW FACTOR FOR EXTERNAL RADIATION SOURCES

In 3D, the view factor for a point at finite distance is given by

cos θ
------------
2
4πr

THEORY FOR SURFACE-TO-SURFACE RADIATION | 267

 where θ is the angle between the normal to the irradiated surface and the direction of
the source, and r is the distance from the source. For a source at infinity, the view factor
is given by cos θ.

In 2D the view factor for a point at finite distance is given by

cos θ-
-----------
2πr

and the view factor for a source at infinity is cos θ.

SOLAR POSITION
The Sun is the most common example of an external radiation source. The position of
the Sun is necessary to determine the direction of the corresponding external radiation
source. The direction of sunlight (zenith angle and the solar elevation) is automatically
computed from the latitude, longitude, time zone, date, and time using similar a
method as described in Ref. 20. The estimated solar position is accurate for a date
between year 2000 and 2199, due to an approximation used in the Julian Day calendar
calculation.

The zenith angle, θs, and azimuth angle, ϕ s , of the Sun are converted into a direction
vector is = (isx, isy, isz) in Cartesian coordinates assuming that the north, the west, and
the up directions correspond to the x, y, and z directions, respectively, in the model.
The relation between θs, ϕ s , and is is given by:

i sx = – cos ( ϕ s ) sin ( θ s )
i sy = sin ( ϕ s ) sin ( θ s )
i sz = – cos ( θ s )

RADIATION IN AXISYMMETRIC GEOMETRIES

For an axisymmetric geometry, Gm and Famb must be evaluated in a corresponding
3D geometry obtained by revolving the 2D boundaries around the axis. COMSOL
Multiphysics creates this virtual 3D geometry by revolving the 2D boundary mesh into
a 3D mesh. The resolution can be controlled in the azimuthal direction by setting the
number of azimuthal sectors, which is the same as the number of elements to a full
revolution. Try to balance this number against the mesh resolution in the rz-plane.
This number, the azimuthal sectors, is accessible from the Radiation Settings section
in physics interfaces for heat transfer.

Select between the hemicube and the direct area integration methods also in axial
symmetry. Their settings work the same way as in 3D.

268 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

While Gm and Famb are in fact evaluated in a full 3D, the number of
points where they are requested is limited to the quadrature points on the
boundary of a 2D geometry. The savings compared to a full 3D
simulation are therefore substantial despite the full 3D view factor code
being used.

THEORY FOR SURFACE-TO-SURFACE RADIATION | 269

 Theory for Radiation in Participating
Media
In this section:

• Radiation and Participating Media Interactions

• Radiative Transfer Equation
• Boundary Condition for the Radiative Transfer Equation
• Heat Transfer Equation in Participating Media
• Discrete Ordinates Method (DOM)
• Discrete Ordinates Method Implementation in 2D
• P1 Approximation Theory
• Radiation in Absorbing-Scattering Media Theory
• Polychromatic Radiation
• Radiative Beam in Absorbing Media Theory
• Rosseland Approximation Theory

Radiation and Participating Media Interactions

Figure 4-22: Example of interactions between participating media and radiation.

270 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

In some applications the medium is not completely transparent and the radiation rays
interact with the medium.

Let I(Ω) denote the radiative intensity traveling in a given direction, Ω. Different kinds
of interactions are observed:

• Absorption: The medium absorbs a fraction of the incident radiation. The amount
of absorbed radiation is κI(Ω), where κ is the absorption coefficient.
• Emission: The medium emits radiation in all directions. The amount of emitted
radiative intensity is equal to κIb, where Ib is the blackbody radiation intensity.
• Scattering: Part of the radiation coming from a given direction is scattered in other
directions. The scattering properties of the medium are described by the scattering
phase function φ ( Ω′, Ω ) , which gives the probability that a ray coming from one
direction Ω′ is scattered into the direction Ω. The phase function φ ( Ω′, Ω ) satisfies:
1
------
4π 4π φ ( Ω′, Ω ) dΩ′ = 1

Radiative intensity in a given direction is attenuated and augmented by scattering:

- It is attenuated because a part of the incident radiation in this direction is
scattered into other directions. The amount of radiation attenuated by scattering
is σsI(Ω).
- It is augmented because a part of the radiative intensity coming from other
directions is scattered in all direction, including the direction we are looking at.
The amount of radiation augmented by scattering is obtained by integrating
scattering coming from all directions Ω′:
σ
------s
4π 4π I ( Ω′ )φ ( Ω′, Ω ) dΩ′

Radiative Transfer Equation

The balance of the radiative intensity including all contributions (propagation,
emission, absorption, and scattering) can now be formulated. The general radiative
transfer equation can be written as (see Ref. 23):

σs
Ω ⋅ ∇I ( Ω ) = κI b ( T ) – βI ( Ω ) + ------
4π 4π I ( Ω′ )φ ( Ω′, Ω ) dΩ′ (4-115)

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 271

 where

• I(Ω) is the radiative intensity at a given position following the Ω direction (SI unit:
W/(m2·sr))
• Ib(T) is the blackbody radiative intensity (SI unit: W/(m2·sr)), defined as
2 4
n r σT
I b ( T ) = ----------------- (4-116)
π

The quantity Ib(T) is available as a predefined function, ht.fIb(T), in

heat transfer interfaces.

• nr is the refractive index (SI unit: 1)

• σ is the Stefan-Boltzmann constant (SI unit: W/(m2·K4))
• κ, β, σs are absorption, extinction, and scattering coefficients, respectively (SI unit:
1/m) and are related by:
β = κ + σs

• φ ( Ω′, Ω ) is the scattering phase function (SI unit: 1)

• T is the temperature (SI unit: K)

SCATTERING PHASE FUNCTION

The phase function, φ ( Ω′, Ω ) , gives the probability that a ray from the Ω′ direction is
scattered into the Ω direction. The phase function’s definition is material dependent
and its definition can be complicated. It is common to use approximate scattering
phase functions that are defined using the cosine of the scattering angle, μ0. The
current implementation handles:

• Isotropic phase functions:

φ ( Ω′, Ω ) = φ ( μ 0 ) = 1

• Linear anisotropic phase functions:

φ ( μ0 ) = 1 + a1 μ0

• Polynomial anisotropic up to the 12th order:

272 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 12

φ ( μ0 ) = 1 +  an Pn ( μ0 )
n=1

where Pn are the nth order Legendre polynomials.

Legendre polynomials can be defined by the Rodriguez formula:

1 dk 2 k
P k ( x ) = -----------
k
((x – 1) )
k
2 k! d x

• Henyey–Greenstein phase function:

2
1 1–η
φ ( μ 0 ) = ---- ⋅ ------------------------------------------------
3⁄2
-
K 2
( 1 + η – 2ημ 0 )

where – 1 < η < 1 is the anisotropy parameter and K is defined as follows to produce
a normalized phase function:
2
1 1–η
K = ------ ⋅
4π  ------------------------------------------------
( 1 + η
2
– 2ημ )
3⁄2
- dΩ
4π 0

INCIDENT RADIATION
A quantity of interest is the incident radiation, denoted G (SI unit: W/m2), and
defined by

G = 4π I ( Ω ) dΩ

Boundary Condition for the Radiative Transfer Equation

For gray walls, corresponding to opaque surfaces reflecting diffusively and emitting,
the radiative intensity I(Ω) entering participating media along the Ω direction is

1–ε
I ( Ω ) = εI b ( T ) + ----------- q r, out for all Ω such that n ⋅ Ω < 0
π

where

2 4
n r σT
I b ( T ) = -------------------- (4-117)
π

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 273

 • Equation 4-117 is the blackbody radiation intensity and nr is the refractive index
• ε is the surface emissivity, which is in the range [0, 1]
• 1 − ε is the diffusive reflectivity
• n is the outward normal vector
• qr,out is the heat flux coming from the domain and striking the wall:

q r, out =  n ⋅ Ω > 0 I ( Ω ) ( n ⋅ Ω ) dΩ
For black walls ε = 1. Thus I(Ω) = Ib(T).

The net heat flux corresponding to the balance between the energy received by the
surface and emitted by the surface is defined by

q r, net = q r, in – q r, out

This net heat flux accounts for the radiation coming from the domain where the RTE
is solved for, the radiation coming from the exterior is not modeled and hence not
accounted for.

If the surface is semitransparent, the radiative intensity transmitted diffusively or

specularly. For an external boundary the external intensity, Iext, is transmitted in the
domain. The radiative intensity entering participating media along the Ω direction is

ρd τd
I ( Ω ) = εI b ( T ) + ------ q r, out + ----- q r, in, ext + τ s I ext
π π

for all Ω such that n ⋅ Ω < 0 ,

where

• ρd is the surface diffuse reflectivity,

• τd is the surface diffuse transmissivity,
• τs is the surface specular transmissivity,
• Iext is the external radiation intensity,
• qr,in is the heat flux striking the wall from exterior:

q r, in, ext = n ⋅ Ω ≤ 0 Iext ( Ω ) ( n ⋅ Ω ) dΩ

The emissivity, diffuse reflectivity, diffuse transmissivity, and specular transmissivity
satisfy the following relation:

274 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ε + ρd + τd + τs = 1

The net heat flux corresponding to the balance between the energy received by the
surface and emitted by the surface is defined by

q r, net = q r, in – q r, out + ε ext q r, in, net – ε ext πεI b ( T )

For semitransparent surfaces, the net heat flux accounts for the radiation coming from
the domain and from the exterior.

The definition of the radiative intensity on internal boundaries is similar to the

definition for external boundaries except that the external radiation is replaced by the
radiation exiting the adjacent domain.

Heat Transfer Equation in Participating Media

Heat flux in gray media is defined by

qr = 4π I ( Ω )Ω dΩ
Heat flux divergence can be defined as a function of G and T (see Ref. 23):

Q r = ∇ ⋅ q r = κ ( G – 4πI b ( T ) )

In order to couple radiation in participating media, radiative heat flux is taken into
account in addition to conductive heat flux. Recalling Equation 4-18, the heat transfer
equation reads:

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ ( q + q r ) = α p T  + u ⋅ ∇p + τ: ∇u + Q
 ∂t  ∂t 

and is implemented using following form:

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ q = κ ( G – 4nσT 4 ) + α p T  + u ⋅ ∇p + τ: ∇u + Q
 ∂t  ∂t 

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 275

 Discrete Ordinates Method (DOM)

The discrete ordinates method is implemented for 3D and 2D

geometries.

Radiative intensity is defined for any direction Ω, because the angular space is
continuous. In order to handle the radiative intensity equation numerically, the angular
space is discretized.

The discrete ordinates method provides a discretization of angular space. The

quadrature sets Level Symmetric Even, Level Symmetric Hybrid and Equal Weight
Odd are designed using moment-matching conditions (see Ref. 24) and yield SN
approximations with N(N + 2) directions in 3D or N(N + 2)/2 in 2D (see Ref. 23).
The quasi-uniform quadrature set yields TN approximations with 8N2 directions in 3D
or 4N2 in 2D (Ref. 46, Ref. 47). These approximations are not designed using
moment-matching conditions but allow for higher-order discretization compared with
SN approximations.

Thanks to angular space discretization, integrals over directions are replaced by

numerical quadratures of discrete directions:

4π I ( Ω ) dΩ ≈  wj Ij
j=1

Depending on the value of N, a set of n dependent variables has to be defined and

solved for I1, I2, …, In.

RADIATION IN PARTICIPATING MEDIA

Each dependent variable satisfies the equation

n
σs
S i ⋅ ∇I i = κI b ( T ) – βI i + ------
4π  w j I j φ ( S j, S i )
j=1

subject to appropriate boundary conditions, where Si is the ith discrete direction.

276 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

OPAQUE SURFACE
The node Opaque Surface (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) defines the default boundary condition for
radiative intensities I1, I2, …, In:

1–ε
I i, bnd = εI b ( T ) + ----------- q r, out for all S i such that n ⋅ S i < 0 (4-118)
π

with

q r, out =  wj Ij ( n ⋅ Sj )
n ⋅ Sj > 0

In Equation 4-118, the first term in the right hand side is the emitted radiative
intensity, while the second term represents the reflected radiative intensity.

SEMITRANSPARENT SURFACE
The node Semitransparent Surface (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) defines the following boundary condition for
radiative intensities I1, I2, …, In:

ρd τd
I i, bnd = εI b ( T ) + ------ q r, out + ----- q r, out, ext + τ s I i, ext (4-119)
π π

for all Si such that n ⋅ S i < 0 , with

q r, out =  wj Ij ( n ⋅ Sj )
n ⋅ Sj > 0

and

q r, out, ext = –  w j I j, ext ( n ⋅ S j )

n ⋅ Sj ≤ 0

In Equation 4-119, the right hand side is composed of the emitted radiative intensity,
the reflected radiative intensity, and the radiative intensity transmitted in a diffuse and
specular way.

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 277

 INCIDENT INTENSITY
The Incident Intensity (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) node defines a boundary that receives incident
radiative intensity Iext and that is transparent for outgoing intensity:

I i, bnd = I ext for all S i such that n ⋅ S i < 0

RADIATIVE SOURCE
The node Radiative Source accounts for a directional power density Ii in the radiative
transfer equation:

n
σs
S i ⋅ ∇I i = κI b ( T ) – βI i + ------
4π  w j I j φ ( S j, S i ) + q i .
j=1

Discrete Ordinates Method Implementation in 2D

For a given index i, define two indices, i+ and i−, so that

• Ω, Si+, and Si- have the same components in the xy–plane

• and Si+ and Si- have opposite components in the z direction.

Assuming that a model is invariant in the z direction, the radiative transfer equation in
two directions, Si+ and Si-, for the discrete ordinates method (DOM) reads:

n
σs
S i + ⋅ ∇I i + = κI b ( T ) – βI i+ + ------
4π  w j I j φ ( S j, S i )
+

j=1

n
σs
S i- ⋅ ∇I i- = κI b ( T ) – βI i- + ------
4π  wj Ij φ ( Sj, Si ) -

j=1

Halving the sum of the two equations above and using I i+ = I i- = I˜i ( I i- = I i+ in
2D) yields
n
σs
S˜ i ⋅ ∇I˜i = κI b ( T ) – βI˜i + ------  wj Ij ( φ ( Sj, Si ) + φ ( Sj, Si ) )
+ -
8π
j=1

which may recast

278 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 n⁄2
σs
S˜ i ⋅ ∇I˜i = κI b ( T ) – βI˜i + ------  wj Ij ( φ ( Sj , Si ) + φ ( Sj , Si ) )
+ + + + + -
8π
j=1
n⁄2
σs
+ ------
8π  wj Ij ( φ ( Sj , Si ) + φ ( Sj , Si ) )
- - - + - -

j=1

or

n⁄2
σs
S˜ i ⋅ ∇I˜i = κI b ( T ) – βI˜i + ------  wj Ij φ ( Sj , Si ) + wj Ij φ ( Sj , Si )
+ + + + - - - -
8π
j=1
n⁄2
σs
+ ------
8π  wj Ij φ ( Sj , Si ) + wj Ij φ ( Sj , Si )
+ + + - - - - +

j=1

In addition if φ ( S i, S j ) can be rewritten as a function of Si ⋅ Sj, as it is in the

COMSOL Multiphysics implementation, then

φ ( S j+, S i + ) = φ ( S j-, S i - ) and φ ( S j-, S i+ ) = φ ( S j+, S i- )

In addition

I j- φ ( S j-, S i+ ) + I j+ φ ( S j+, S i- ) = 2I˜j φ ( S j+, S i- ) = 2I˜j φ ( S j -, S i+ )

so the above equation can be simplified:

n⁄2
σs
S˜ i ⋅ ∇I˜i = κI b ( T ) – βI˜i + ------ ˜ ˜
 wj Ij φ ( Sj, Si )
˜ ˜ (4-120)
4π
j=1

with

S i, 1
S˜ i = S
i, 2
0

since the third component of ∇I˜i is zero in 2D.

Also notice that

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 279

 n n⁄2 n⁄2
˜ ˜
4π I ( Ω ) dΩ ≈  wj Ij  wj Ij + wj Ij  wi Ii
= - + = (4-121)
- +

j=1 j=1 j=1

with w˜ i = 2w i .

Using results from Equation 4-120 and Equation 4-121 the DOM is formulated in
2D using only radiative intensities, I˜i , on half of the 3D DOM directions, S˜ i , except
for the scattering term. In other expressions than the scattering term, the z component
of the radiative intensities Ii and of the discrete directions Ωi can be ignored (or set to
zero) and the weight wi, multiplied by 2.

P1 Approximation Theory
The P1 approximation is available as a radiation discretization method in The Radiation
in Participating Media Interface.

The P1 approximation is the simplest approximation provided by the method of

spherical harmonics method (PN-method). This approximation provides additional
accuracy compared to a Rosseland approximation even if it remains a very simple
method. The P1 method relies on the following hypotheses:

• The media is optically thick media: τ >>1, where τ is the optical thickness defined by
the integral of absorption coefficient, κ, along a typical optical path:
s
τ =  0 κ ds
• The scattering is linear isotropic.

From a computational point of view this approximation has a limited impact because
it introduces only one additional degree of freedom for the incident radiation G (SI
unit: W/m2), which is a scalar quantity and adds a heat source or sink to the
temperature equation to account for radiative heat transfer contributions. This
method, however, fails to accurately represent cases where the radiative intensity
propagation dominates over its diffusivity or where the scattering effects cannot be
described by a linear isotropic phase function.

The P1 approximation accounts for the radiation transfer equation

σs
Ω ⋅ ∇I ( Ω ) = κI b ( T ) – βI ( Ω ) + ------
4π 4π I ( Ω′ )φ ( Ω′, Ω ) dΩ′

280 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

by solving following equation for G =
4π I ( Ω ) dΩ (Ref. 23):
∇ ⋅ ( – D P1 ∇G ) = – Q r (4-122)

where

• DP1 is the P1 diffusion coefficient, defined as

1
D P1 = ----------------------------------------
3κ + σ s ( 3 – a 1 )

• a1 is the linear Legendre coefficient of the scattering phase function

• Qr is the radiative heat source:

Q r = κ ( G – 4πI b ) (4-123)

When scattering is modeled as isotropic, a1=0 and the P1 diffusion coefficient reduces
to

1
D P1 = ----------------------
3κ + 3σ s

The following boundary condition applies (Ref. 23):

n ⋅ ( – D P1 ∇G ) = – q r, net

where qr, net is the net radiative heat flux at the boundary.

RADIATION IN PARTICIPATING MEDIA

For the Participating Medium (Radiation in Participating Media Interface) feature
node, the equation Equation 4-122 is implemented.

In addition Qr, defined by Equation 4-123, is added as an heat source in the heat
transfer equation:

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ q = Q r + α p T  + u ⋅ ∇p + τ: ∇u + Q
 ∂t  ∂t

OPAQUE SURFACE
The Opaque Surface (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) boundary condition defines a boundary
opaque to radiation and defines the incident intensity on a boundary:

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 281

 n ⋅ ( – D P1 ∇G ) = q r, net

The Opaque Surface feature accounts for the net radiative heat flux, qr, net, in the heat
balance.

Two cases are considered, depending on surface emissivity value:

• Gray wall: ε is a number between 0 and 1

• Black wall: ε = 1

Gray Wall
The radiative heat flux at the boundary depends on the surface emissivity, ε:

ε
q r, net = -------------------- ( 4πI b, w – G )
2(2 – ε)

with

n2 σT4
I b, w = I b = -----------------
π

Black Wall
The radiative heat flux at the boundary expression simplifies to

1
n ⋅ ( – D P1 ∇G ) = – --- ( 4πI b, w – G ) ,
2

with

n 2 σT 4
I b, w = I b = -----------------
π

SEMITRANSPARENT SURFACE
The Semitransparent Surface (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) boundary condition defines a boundary that
emits radiation, and reflects one part of the incident intensity while the remaining is
transmitted diffusively.

The net radiative heat flux, qr, net, accounted for in the heat balance, is defined as

2 1 – ρd
q r, net = --------------- ( πεI b, w + τ d G ) – ------------------------ G
1 + ρd 2 ( 1 + ρd )

where ρd is the surface diffuse reflectivity, and τd is the surface diffuse transmissivity.

282 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

INCIDENT INTENSITY
The Incident Intensity (Radiation in Participating Media and Radiation in
Absorbing-Scattering Media Interfaces) node defines a boundary that receives incident
radiative intensity Iext and that is transparent for outgoing intensity. On these
boundaries, the relation between G (incident radiation), qr, net (net radiative heat flux)
and Iext (incident radiative intensity) is

G + 2q r, net = 4 Ω ⋅ n < 0 Iext ( Ω )Ω ⋅ n dΩ

by defining

I ext = Ω ⋅ n < 0 Iext ( Ω )Ω ⋅ n dΩ

there is

1
q r, net = --- ( 4πI ext – G )
2

which defines the heat radiative heat flux and also contributes to G boundary
condition:

n ⋅ ( – D P1 ∇G ) = – q r, net

RADIATIVE SOURCE
The node Radiative Source accounts for a power density Q in the P1 approximation:

∇ ⋅ ( – D P1 ∇G ) = – Q r + Q .

Radiation in Absorbing-Scattering Media Theory

The Radiation in Absorbing-Scattering Media Interface is available in 2D, 2D
axisymmetric, and 3D components to model the propagation, attenuation, and
scattering of an incident light within a semitransparent material considered to be
non-emitting.

The radiative intensity I(Ω) at a given position following the Ω direction is the solution
of the radiative transfer equation with no emission term (see Ref. 23):

σs
Ω ⋅ ∇I ( Ω ) = – βI ( Ω ) + ------
4π 4π I ( Ω′ )φ ( Ω′, Ω ) dΩ′ (4-124)

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 283

 where

• κ, β, σs are absorption, extinction, and scattering coefficients, respectively (SI unit:

1/m) and are related by:
β = κ + σs

• φ ( Ω′, Ω ) is the scattering phase function (SI unit: 1)

See Radiative Transfer Equation for details about the phase function, φ ( Ω′, Ω ) .

The incident radiation, denoted G (SI unit: W/m2), is defined by

G = 4π I ( Ω ) dΩ
If the Discrete Ordinates Method (DOM) is used for the approximation of
Equation 4-124, G is computed as

G=  ωi Ii
i=1

and

N
σs
S i ⋅ ∇I i = – βI i + ------
4π  ω j I j φ ( S j, S i )
j=1

where

• Si is the ith discrete ordinate.

• Ii is the ith component of the radiative intensity.
• ωj is the ith quadrature weight.
If the P1 Approximation Theory is used instead, G is the solution of the following
equation,

∇ ⋅ ( – D P1 ∇G ) = – κ G , (4-125)

with the following boundary condition for an opaque surface,

ε
n ⋅ ( – D P1 ∇G ) = -------------------- G ,
2(2 – ε)

where DP1 is the P1 diffusion coefficient and ε the surface emissivity.

284 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Polychromatic Radiation
As in the Surface-to-Surface Radiation Interface, polychromatic radiation is modeled
using the stepwise-gray model (Ref. 23) to account for wavelength dependence of
emission, absorption and scattering through material properties, see the section
Wavelength Dependence of Surface Properties.

In the Discrete Ordinates Method (DOM), the Radiative Transfer Equation is solved
for each spectral band:

N
σ s, k
S i ⋅ ∇I i, k = κ k I b ( T )FEP k ( T ) – β k I i, k + ----------
4π  ω j I j, k φ k ( S j, S i ) ,
j=1

where FEP is the fractional emissive power, also presented in the section Wavelength
Dependence of Surface Properties.

If the P1 Approximation Theory is used instead, the incident radiation Gk for the
spectral band k is solution of the following equation,

∇ ⋅ ( – D P1, k ∇G k ) = – Q r, k , (4-126)

where

• DP1,k is the P1 diffusion coefficient, defined as

1
D P1, k = ----------------------------------------------------
3κ k + σ s, k ( 3 – a 1, k )

• a1,k is the linear Legendre coefficient of the scattering phase function

• Qr,k is the radiative heat source:

Q r, k = κ k ( G k – 4πI b FEP k ( T ) ) . (4-127)

Radiative Beam in Absorbing Media Theory

The Radiative Beam in Absorbing Media Interface is available in 2D, 2D axisymmetric,
and 3D components to model the propagation and the attenuation of an incident light
within a semitransparent material. The radiative intensity of the beam is described by

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 285

 the Beer-Lambert Law, which corresponds to the radiative transport equation (RTE)
under the following hypothesis:

• The radiative beam in the absorbing medium is collimated and each beam
propagates always in the same direction.
• The light experiences no refraction, reflection, or scattering within the material
itself.
• There is no significant emission of the material in the wavelength range of the
incident light. This applies well to laser beams, whose wavelength is in general much
shorter than the one of the radiation emitted by the medium.
In these conditions, the radiative intensity Ii (SI unit: W/m2) of the ith beam through
the material decreases as the beam propagates and is absorbed by the medium. This is
described by the Beer-Lambert Law equation:

ei
--------- ⋅ ∇I i = – κI i
ei

where ei is the orientation of the ith beam, and κ is the absorption coefficient (SI unit:
m–1) of the medium.

The radiative heat source Qr (SI unit: W/m3), corresponding to the energy deposited
by the radiative beam, is defined by:

 κIi
Qr =
.
i

Rosseland Approximation Theory

The Rosseland approximation is available as a radiation discretization method as part
of the Optically Thick Participating Medium subnode.

Rosseland approximation relies on the hypotheses that the participating medium is

optically thick — that is, τ >>1 — where τ is the optical thickness defined by the integral
of absorption coefficient, κ, along a typical optical path:

s
τ =  0 κ ds
From a computational cost point of view this approximation has a limited impact
because it does not introduce any extra degree of freedom to the heat equation.

286 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Instead it adds nonlinear contribution to the thermal conductivity. This is why this
method is popular for some applications where the optical thickness is large.
Nevertheless, because it gives a simple approximation of heat transfer by radiation in a
participating media, it should be carefully validated.

In this case, the radiative heat flux can be evaluated by (Ref. 23):

4π
q r, λ = – ------ ∇i b, λ
βλ

For a gray media it leads to

4σ
q r = – ---------- ∇ ( n 2 T 4 )
3β R

Assuming a constant refractive index, this can be rewritten as qr = − kR ∇T with

16n 2 σT 3
k R = ------------------------
3β R

and

16n 2 σT 3
q r = – ------------------------ ∇T
3β R

Rosseland mean extinction coefficient, βR, can be defined from the absorption and
scattering coefficients of the media from βR=κ+σs.

THEORY FOR RADIATION IN PARTICIPATING MEDIA | 287

 Theory for Moisture Transport
Theory for Moisture Transport in Air
In moist air, the transport of moisture is done by convection and binary diffusion of
vapor in air.

Two formulations are available, depending on the magnitude of moist air density
variations with moisture content.

In both formulations, the moisture is supposed to be constituted only of vapor when

considering moist air. In other terms the liquid concentration is null, except on
boundaries where condensation may take place.

DILUTED SPECIES FORMULATION

With this formulation, it is supposed that the spatial and temporal variations of the
vapor concentration are small enough not to induce significant variations of the moist
air density.

The Moisture Transport in Air Interface solves for the following equation, in which
the moisture content variation is expressed through the transport of vapor
concentration, cv, by convection and binary diffusion in air:

∂c v
M v -------- + M v u ⋅ ∇c v + ∇ ⋅ g w = G (4-128)
∂t

g w = – M v D ∇c v

c v = φ w c sat

with the following material properties, fields, and source:

• Μv (SI unit: kg/mol) is the molar mass of water vapor.

• φ w (dimensionless) is the relative humidity.
• csat (SI unit: mol/m3) is the vapor saturation concentration.
• D (SI unit: m2/s) is the vapor diffusion coefficient in air.
• u (SI unit: m/s) is the air velocity field.
• gw (SI unit: kg/(m2⋅s)) is the vapor diffusive flux.
• G (SI unit: kg/(m3⋅s)) is the moisture source (or sink).

288 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

CONCENTRATED SPECIES FORMULATION
When the spatial and temporal variations of the vapor concentration induce significant
variations of the moist air density, the moisture content variation is expressed through
the transport of vapor mass fraction, ωv, defined as:

Mv cv
ω v = --------------
ρg

where ρg is the moist air density (SI unit: kg/m3).

The Heat Transfer in Moist Air Interface solves for the following equation:

∂ω v
ρ g ---------- + ρ g u ⋅ ∇ω v + ∇ ⋅ g w = G (4-129)
∂t

g w = – ρ g D ∇ω v

This is equivalent to solve the following equations on the vapor concentration:

∂ ( cv ⁄ ρg )
M v ρ g ----------------------- + M v ρ g u ⋅ ∇( c v ⁄ ρ g ) + ∇ ⋅ g w = G
∂t

g w = – M v ρ g D ∇( c v ⁄ ρ g )

By neglecting the spatial and temporal variations of ρg in the equations above, the
equation of the diluted species formulation, Equation 4-128, is obtained.

Theory for Moisture Transport in Porous Media

In a porous material, moisture may be present under vapor and liquid state, and
moisture transport occurs through convection and diffusion, liquid water being fixed
on the pores walls mainly in two ways:

• Adsorption: the vapor is fixed under molecular films on the pores walls. Equilibrium
vapor pressure is reached in the pores under thermodynamic conditions that are
significantly different than those necessary in a free medium. Thus saturation
conditions are changed, and the relative humidity is significantly changed as well,
becoming a relevant variable for the characterization of moisture content.
• Capillarity: when the porous matrix is wetted by liquid water, there is a pressure
difference between the liquid and gas (moist air) phases, due to the curvature of the
wetting interfaces in each pore. The pressure difference, called capillary pressure,

THEORY FOR MOISTURE TRANSPORT | 289

 becomes the driving potential for a further filling of the pores by the water. Note
that in some applications, the capillary pressure is replaced by the suction.

The predominance of adsorption or capillarity depends both on the amount of liquid

water present in the porous medium, and on the size and shape of the pores, and
should be defined in function of the material and process under study. Two moisture
regions can be distinguished:

• Hygroscopic region: adsorption is the predominant phenomenon, and the

equilibrium moisture content is better characterized from the relative humidity,
through the specification of a sorption curve. For relative humidity values below
0.95, the slope of the sorption curve is relatively small, and the moisture content
increases in a nearly linear way with relative humidity. Building materials are most of
the time in this moisture region. Above 0.95, a very small variation of relative
humidity induces a large variation of the moisture content, and the capillary pressure
is best suited to characterize the moisture content.
• Capillary water region: capillarity is the predominant phenomenon, and the
equilibrium moisture content is better characterized from the capillary pressure,
through the specification of a retention curve (the van Genuchten and Brooks &
Corey retention models for example).

The relative humidity and capillary pressure are called state variables: they describe the
thermodynamic state of water in the pore space under two phases in equilibrium. They
are related through Kelvin’s law, which is a relation of thermodynamical origin. As long
as equilibrium is satisfied, the state variables are continuous between two media (at the
interface between two porous media with different porosities, or between a porous and
a free medium), and Kelvin’s law is valid on both sides of the interface. In opposition,
the liquid saturation or the moisture content depend on the porous matrix.

Note that capillary transport of liquid water should be considered in both regions.

The Moisture Transport in Porous Media Interface is dedicated to the modeling of

porous media in the hygroscopic moisture region. The total moisture content is thus
defined by a sorption curve as

εp sl ρl + εp ρ ωv ( 1 – sl ) = w ( φw )
g

where the liquid saturation, sl, describes the amount of liquid water within the pores,
and εp (dimensionless) is the porosity.

290 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

By summing the mass conservation equations for liquid water and vapor, a single
equation for the transport of the total moisture can be written, in which the
evaporation and condensation source terms of each equation cancel out:

∂w ( φ w )
-------------------
- + ρ g u g ⋅ ∇ω v + ∇ ⋅ g w + u l ⋅ ∇ρ l + ∇ ⋅ g lc = G (4-130)
∂t

with the following material properties, fields, and source:

• w ( φ w ) (SI unit: kg/m3) is the total moisture content.

• φ w (dimensionless) is the relative humidity.
• ρg (SI unit: kg/m3) is the moist air density, defined from the dry air and vapor
densities, in function of the amount of vapor.
• ug (SI unit: m/s) is the moist air velocity field, that should be interpreted as the
Darcy velocity, that is, the volume flow rate per unit cross sectional area.
• ωv (dimensionless) is the vapor mass fraction in moist air, defined as

M v φ w c sat ( T )
ω v = -----------------------------------
ρg

• Μv (SI unit: kg/mol) is the molar mass of water vapor.

• csat (SI unit: mol/m3) is the vapor saturation concentration.
• T (SI unit: K) is the temperature.
• gw is the moisture diffusive flux is defined as:
g w = – ρ g D eff ∇ω v

• Deff (SI unit: m2/s) is the effective vapor diffusion coefficient in the porous
medium, computed from the diffusion coefficient in a free medium, and accounting
for the porosity and tortuosity of the porous medium.
• ul (SI unit: m/s) is the liquid water velocity field, defined from the absolute
pressure gradient by the Darcy’s law as:
κ rl κ
u l = – ----------- ∇p A
μl

• κrl (dimensionless) is the relative liquid water permeability, that may be defined as
a function of the liquid saturation.
• κ (SI unit: m2) is the porous medium permeability.

THEORY FOR MOISTURE TRANSPORT | 291

 • pA (SI unit: Pa) is the absolute pressure.
• ρl (SI unit: kg/m3) is the liquid water density, defined as a function of the
temperature.
• glc is the liquid water capillary flux, defined either from the capillary pressure
gradient by a Darcy’s law:
κ rl κ
g lc = ρ l ----------- ∇p c
μl

or by a diffusion equation using the relative humidity:

∂w ( φ w )
g lc = – D w -------------------- ∇φ w
∂φ w

• Dw (SI unit: m2/s) is the moisture diffusivity.

• pc is the capillary pressure, that can be related to the relative humidity using Kelvin’s
law:
RTρ l
p c = – -------------- ln ( φ w )
Mv

• G (SI unit: kg/(m3⋅s)) is a moisture source (or sink). See the Moisture Source node.

The moisture source due to evaporation may be obtained from the transport equation
for vapor:

∂ [ εp ρg ωv ( 1 – sl ) ]
G evap = ---------------------------------------------- + ρ g u g ⋅ ∇ω v + ∇ ⋅ g w
∂t

The corresponding latent heat source is defined as:

Q evap = L v G evap

where Lv is the latent heat of evaporation.

For a steady-state problem, the relative humidity does not change with time and the
first term in the moisture transport equation disappears.

Theory for Moisture Transport in Building Materials

By considering that building materials are in general surrounded by an environment of
constant atmospheric pressure and under a certain range of temperature and moisture

292 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

conditions, specific equations are proposed in Ref. 16 for moisture transport in this
kind of materials.

Starting from the equation for the transport of moisture in porous media (using the
diffusion model for capillary flux):

∂w ( φ w ) ∂w ( φ w )
- + ρ g u g ⋅ ∇ω v + ∇ ⋅ g w + u l ⋅ ∇ρ l + ∇ ⋅  – D w -------------------
------------------- - ∇φ w = G
∂t  ∂φ w 

let’s make the assumption that there is no variation of the ambient total pressure
(pA=1 atm). In this case, by defining the gas and liquid velocities ug and ul as Darcy
velocities, the convective fluxes for vapor and liquid water vanish:

∂w ( φ w ) ∂wφ w
- + ∇ ⋅  g w – D --------------
------------------- - ∇φ w = G
∂t  w ∂φ w 

By introducing the moisture storage capacity, ξ, defined as

∂w ( φ w )
ξ = --------------------
∂φ w

the equation reads

∂φ w
ξ ---------- + ∇ ⋅ ( g w – D w ξ ∇φ w ) = G
∂t

Finally, the diffusive flux for vapor can be written as:

Ma Mv pv
g w = – ρ g D eff ∇ω v = – ρ g D eff ∇ ------------------------------------------2- ------
( xa Ma + xv Mv ) A p

By making the approximation that xaMa+xvMv is constant, and that moist air is an
ideal gas, we have:

M v D eff Ma
g w = – ------------------- ------------------------------------------2- ∇p v
RH2 O ( x M + x M )
a a v v

The vapor permeability coefficient is then defined as

M v D eff Ma
δ p = ------------------- ------------------------------------------2-
RH2 O ( x M + x M )
a a v v

THEORY FOR MOISTURE TRANSPORT | 293

 The Moisture Transport in Building Materials Interface solves for the following
equation derived from Ref. 16:

∂φ w
ξ ---------- + ∇ ⋅ ( – ξ D w ∇φ w – δ p ∇ ( φ w p sat ( T ) ) ) = G (4-131)
∂t

This equation models the moisture transfer as the sum of the capillary moisture flux:

∂w
– D w ∇ ( w ( φ w ) ) = – D w ---------- ∇φ w = – ξD w ∇φ w
∂φ w

and the vapor diffusion flux:

δ p ∇p v ( T ) = δ p ∇ ( φ w p sat ( T ) )

with the following material properties, fields, and source:

• ξ (SI unit: kg/m3) is the moisture storage capacity.

• δp (SI unit: s) is the vapor permeability.
• φ w (dimensionless) is the relative humidity.
• psat (SI unit: Pa) is the vapor saturation pressure.
• T (SI unit: K) is the temperature.
• Dw (SI unit: m2/s) is the moisture diffusivity.
• G (SI unit: kg/m3·s) is the moisture source.

294 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for the Heat Transfer
Multiphysics Couplings
In this section:

• Theory for the Nonisothermal Flow and Conjugate Heat Transfer Interfaces
• Theory for the Moisture Flow Interface
• Theory for the Thermoelectric Effect Interface
• Theory for the Building Materials Version of the Heat and Moisture Transport
Interface
• Theory for the Moist Air Version of the Heat and Moisture Transport Interface
• Theory for the Moist Porous Media Version of the Heat and Moisture Transport
Interface
• Theory for the Heat and Moisture Flow Interfaces
• Theory for the Electromagnetic Heating Interfaces
• Theory for the Thermal Stress Interface

Theory for the Nonisothermal Flow and Conjugate Heat Transfer

Interfaces
The following points of the theory of Nonisothermal Flow and Conjugate Heat
Transfer are discussed in this part:

• The Nonisothermal Flow and Conjugate Heat Transfer Equations

• Turbulent Nonisothermal Flow Theory
• Theory for the Nonisothermal Screen Boundary Condition
• Theory for the Interior Fan Boundary Condition

See Theory for the Single-Phase Flow Interfaces and Theory for the Turbulent Flow
Interfaces in the CFD Module User’s Guide for a description of the theory related to
laminar and turbulent single-phase flow interfaces.

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 295

 THE NONISOTHERMAL FLOW AND CONJUGATE HEAT TRANSFER
EQUATIONS
In industrial applications it is common that the density of a process fluid varies. These
variations can have a number of different sources but the most common one is the
presence of an inhomogeneous temperature field. This module includes the
Nonisothermal Flow predefined multiphysics coupling to simulate systems in which
the density varies with temperature.

Other situations where the density might vary includes chemical reactions, for instance
where reactants associate or dissociate.

The Nonisothermal Flow and Conjugate Heat Transfer interfaces contain the fully
compressible formulation of the continuity and momentum equations:

∂ρ
------ + ∇ ⋅ ( ρu ) = 0
∂t
(4-132)
∂u
ρ ------- + ρu ⋅ ∇u = – ∇p + ∇ ⋅ τ + F
∂t

where

• ρ is the density (SI unit: kg/m3)

• u is the velocity vector (SI unit: m/s)
• p is the pressure (SI unit: Pa)
• τ is the viscous stress tensor (SI unit: Pa), equal for a compressible fluid to:
2
τ = μ ( ∇u + ( ∇u ) T ) – --- μ ( ∇ ⋅ u )I
3

• μ is the dynamic viscosity (SI unit: Pa·s)

• F is the body force vector (SI unit: N/m3)

It also solves the heat equation, which for a fluid is given in Equation 4-18 by

∂T ∂p
ρC p  ------- + u ⋅ ∇ T + ∇ ⋅ ( q + q r ) = α p T  + u ⋅ ∇p + τ: ∇u + Q
 ∂t  ∂t

where in addition to the quantities above

• Cp is the specific heat capacity at constant pressure (SI unit: J/(kg·K))

• T is the absolute temperature (SI unit: K)
• q is the heat flux by conduction (SI unit: W/m2)

296 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• qr is the heat flux by radiation (SI unit: W/m2)
• αp is the coefficient of thermal expansion (SI unit: 1/K):

1 ∂ρ
α p = – ---
ρ∂T

• Q contains heat sources other than viscous heating (SI unit: W/m3)

The work done by pressure changes term

∂p
Q p = α p T  ------ + u ⋅ ∇ p
 ∂t 

and the viscous heating term

Q vd = τ: ∇u

are not included by default because they are usually negligible. These terms can,
however, be added by selecting corresponding check boxes in the Nonisothermal Flow
feature.

The physics interface also supports heat transfer in solids (Equation 4-16):

∂T
ρC p  ------- + u trans ⋅ ∇ T + ∇ ⋅ ( q + q r ) = Q ted + Q
∂t

where Qted is the thermoelastic damping heat source (SI unit: W/(m3)). This term is
not included by default but must be added by selecting the corresponding check box.

• The Heat Balance Equation

• Turbulent Nonisothermal Flow Theory

TURBULENT NONISOTHERMAL FLOW THEORY

Turbulent energy transport is conceptually more complicated than energy transport in
laminar flows because the turbulence is also a form of energy.

Equations for compressible turbulence are derived using the Favre average. The Favre
˜
average of a variable T is denoted T and is defined by

˜ ρT
T = -------
ρ
where the bar denotes the usual Reynolds average. The full field is then decomposed as

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 297

 ˜
T = T + T″

With this notation the energy balance equation becomes

˜ ˜ ˜ u˜ 
∂  ρE
˜ u i u i ρu i ″u i ″ ∂  ρu˜  ˜ + u i i ˜ ρu i ″u i ″
  + ----------- + -------------------- +  j  H ----------- + u j -------------------- = (4-133)
∂t  2  2  ∂ xj  2  2 

∂  – q – ρu ″H″ τ u ″ ρu j ″u i ″u i ″ ∂ ( u˜ ( – ρu ″u ″ ) )
+ ij i – ----------------------------- + τ
∂ xj j j 2 ∂ x j i ij i j

where H is the enthalpy. The vector

∂T
q j = – λ ------- (4-134)
∂x j

is the laminar conductive heat flux and

∂u i ∂u j 2 ∂u k
τ ij = μ  -------- + -------- – --- μ --------- δ ij
 ∂x j ∂x i  3 ∂x k

is the laminar, viscous stress tensor. Notice that the thermal conductivity is denoted λ.

The modeling assumptions are in large part analogous to those for incompressible
turbulence modeling. The stress tensor

– ρu i''u'' j

is modeled using the Boussinesq approximation:

˜
T  ∂u˜i ∂u˜ j 2 ∂u k 2
– ρu i ″u j ″ = ρτ ij = μ T  -------- + -------- – --- μ --------- δ ij – --- ρkδ ij (4-135)
 ∂x j ∂x i  3 ∂x k 3

where k is the turbulent kinetic energy, which in turn is defined by

1
ρk = --- ρu i ″u i ″ (4-136)
2

The correlation between uj″ and H″ in Equation 4-133 is the turbulent transport of
heat. It is modeled analogously to the laminar conductive heat flux

˜ μ T C p ∂T ˜
T ∂T
ρu j ″H″ = q j = – λ T ------- = – -------------- ------- (4-137)
∂x j Pr T ∂x j

298 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

The molecular diffusion term,

τ ij u i ″

and turbulent transport term,

ρu j ″u i ″u i ″
----------------------------
-
2

are modeled by a generalization of the molecular diffusion and turbulent transport

terms found in the incompressible k equation

ρu j ″u i ″u i ″ μ T ∂k
τ ij u i ″ – ----------------------------- =  μ + ------ ------- (4-138)
2  σ k ∂x j

Inserting Equation 4-134, Equation 4-135, Equation 4-136, Equation 4-137 and
Equation 4-138 into Equation 4-133 gives

˜ ˜ ˜ u˜
∂ ρE ˜ ui ui  ∂  ρu˜  ˜ + u i i 
  + ----------- + k  +  j  H ----------- + k  = (4-139)
∂ t  2   ∂ xj  2 
∂  q μ T ∂k ∂ u˜
– j – q j +  μ + ------ ------- +
T T
( ( τ + ρτ ij ) )
∂ xj   σ k ∂x j ∂ x j i ij

The Favre average can also be applied to the momentum equation, which, using
Equation 4-135, can be written

∂ ρu˜ ∂ ρu˜ ˜ ∂p ∂ τ T
( i) + ( j u i ) = – ------- + ( + ρτ ij ) (4-140)
∂t ∂ xj ∂x j ∂ x j ij

Taking the inner product between u˜ i and Equation 4-140 results in an equation for
the resolved kinetic energy, which can be subtracted from Equation 4-139 with the
following result:

∂ (ρ(E ˜ ˜
+ k ) ) + ∂ ( ρu˜ j ( E + k ) ) = (4-141)
∂t ∂ xj
∂ũ j ∂  μ T ∂k
– q j – q j +  μ + ------ ------- + ∂ ( u˜i ( τ ij + ρτ ij ) )
T T
–p + 
∂ xj ∂ xj  σ k ∂x j ∂ xj

where the relation

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 299

 ˜ ˜ p
H = E + ---
ρ
has been used.

According to Wilcox (Ref. 27), it is usually a good approximation to neglect the

contributions of k for flows with Mach numbers up to the supersonic range. This gives
the following approximation of Equation 4-141:

˜
∂ ( ρE ˜ ∂ũ j ∂
) + ∂ ( ρu˜ j E ) = – p
T ∂ u˜ T
+ ( – qj – qj ) + ( ( τ + ρτ ij ) ) (4-142)
∂t ∂ xj ∂ xj ∂ xj ∂ x j i ij

Larsson (Ref. 28) suggests to make the split

τ ij = τ˜ij + τ ij ″

Since

τ˜ij » τ ij ″

for all applications of engineering interest, it follows that

τ ij ≈ τ˜ij

and consequently

˜
˜
∂ ( ρE ˜ ∂ũ j ∂  ∂T 
) + ∂ ( ρu˜ j E ) = – p ∂ ( u˜ tot )
+  ( λ + λ T ) ------- + τ̃ (4-143)
∂t ∂ xj ∂ xj ∂ xj  ∂x j ∂ x j i ij

where

˜
tot  ∂u˜ i ∂u˜ j 2 ∂u k 
τ̃ ij = ( μ + μ T )  -------- + -------- – --- --------- δ ij
 ∂x j ∂x i 3 ∂x k 

Equation 4-143 is completely analogous to the laminar energy equation of

Equation 4-15 and can be expanded using the same theory to get the temperature
equation similar to Equation 4-18 (see for example Ref. 28):

˜ ˜ ˜ ˜
∂T ∂T ∂  λ λ ------
∂T  1  ∂ρ  ˜ ∂p ˜ ∂p ˜ ∂u i
ρC p ------- + ρC p u˜ j ------- +  – ( + T ) - = – ---  ------˜- T  ------ + u j ------- + τ ij --------
∂t ∂x j ∂ x j  ∂x j ρ  ∂T ∂t ∂x j ∂x j
p

300 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

which is the temperature equation solved in the turbulent Nonisothermal Flow and
Conjugate Heat Transfer interfaces.

Turbulent Conductivity

Kays-Crawford This is a relatively exact model for PrT, while still quite simple. In
Ref. 29, it is compared to other models for PrT and found to be a good approximation
for most kinds of turbulent wall bounded flows except for turbulent flow of liquid
metals. The model is given by

 1 0.3C p μ T 0.3C p μ T 2 λ  –1
Pr T =  ----------------- + ----------------------- –  -----------------------  1 – exp  – ------------------------------------------   (4-144)
 2Pr T∞ λ Pr T∞  λ    0.3C μ Pr   
p T T∞

where the Prandtl number at infinity is PrT∞ = 0.85 and λ is the conductivity.

Extended Kays-Crawford Weigand and others (Ref. 30) suggested an extension of

Equation 4-144 to liquid metals by introducing

100λ
Pr T∞ = 0.85 + -------------------------------
-
C p μRe ∞0.888

where Re∞, the Reynolds number at infinity must be provided either as a constant or
as a function of the flow field. This is entered in the Model Inputs section of the Fluid
feature.

Temperature Wall Functions

Analogous to the single-phase flow wall functions (see Wall Functions described for the
Wall boundary condition), there is a theoretical gap between the solid wall and the
computational domain for the fluid and temperature fields. This gap is often ignored
when the computational geometry is drawn.

The heat flux between the fluid with temperature Tf and a wall with temperature Tw, is:

ρC p u τ ( T w – T f )
q wf = -----------------------------------------
+
-
T

where ρ is the fluid density, Cp is the fluid heat capacity, and uτ is the friction velocity.
T+ is the dimensionless temperature and is given by (Ref. 31):

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 301

  + + +
 Prδ w for δ w < δ w1

 2 / 3 – 500 + + +
---------- for δ w1 ≤ δ w < δ w2
T =  15Pr
+
δw +2

 Pr T
 ---------- ln δ +w + β +
for δ w2 ≤ δ w
+
 κ

where in turn

+ δ w ρ C μ1 / 2 k + 10
δ w = ------------------------------- δ w1 = --------------
μ Pr 1 / 3
+ κ Cp μ
δ w2 = 10 10 ---------- Pr = -----------
Pr T λ
Pr T κ
β = 15Pr 2 / 3 – ----------  1 + ln  1000 ---------- 
2κ   Pr T 

λ is the thermal conductivity, and κ is the von Karman constant equal to 0.41.

The distance between the computational fluid domain and the wall, δw, is always hw/2
for automatic wall treatment where hw is the height of the mesh cell adjacent to the
wall. hw/2 is almost always very small compared to any geometrical quantity of interest,
at least if a boundary layer mesh is used. For wall function, δw is at least hw/2 and can
be bigger if necessary to keep δw+ higher than 11.06. The computational results should
be checked so that the distance between the computational fluid domain and the wall,
δw, is everywhere small compared to any geometrical quantity of interest. The distance
δw is available for evaluation on boundaries.

THEORY FOR THE NONISOTHERMAL SCREEN BOUNDARY CONDITION

When the Nonisothermal Flow multiphysics coupling feature is active, the conditions
that apply across a screen in isothermal flow are complemented by:

+
[ H0 ]- = 0 (4-145)

where H0 is the total enthalpy.

• See Screen for the feature node details.

• Also see Screen boundary condition described for the single-phase flow
interfaces.

302 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

THEORY FOR THE INTERIOR FAN BOUNDARY CONDITION
When the Nonisothermal Flow multiphysics coupling feature is active, the conditions
that apply across an interior fan are complemented by:

• If direction is Along normal vector, the outlet temperature Tavg is defined by:

 

 down ( u ⋅ nρC p ) dS T avg =
   down ( u ⋅ nρCp T ) dS
Γ Γ

• If the direction is opposite to normal vector, the outlet temperature Tavg is defined
by:

 

 up ( u ⋅ nρC p ) dS T avg =
   up ( u ⋅ nρCp T ) dS
Γ Γ

Theory for the Moisture Flow Interface

The following points of the theory of Moisture Flow are discussed in this part:

• The Moisture Flow Equations

• Stefan Flow at Walls
• Turbulent Moisture Flow Theory
• Theory for the Screen Boundary Condition in Moisture Flow
• Theory for the Interior Fan Boundary Condition in Moisture Flow

See Theory for the Single-Phase Flow Interfaces and Theory for the Turbulent Flow
Interfaces in the CFD Module User’s Guide for a description of the theory related to
laminar and turbulent single-phase flow interfaces.

THE MOISTURE FLOW EQUATIONS

This module includes the Moisture Flow predefined multiphysics coupling to simulate
systems in which the airflow depends on moisture content through its density and
viscosity. It can be used in free and hygroscopic porous media containing moist air.

The density and viscosity of air might also vary due to temperature variations. See
Theory for the Nonisothermal Flow and Conjugate Heat Transfer Interfaces for details
on this other kind of dependency.

The Moisture Flow interface contains the fully compressible formulation of the
continuity and momentum equations. For laminar flow they read:

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 303

 ∂ρ
--------g- + ∇ ⋅ ( ρ g u ) = 0
∂t
(4-146)
∂u
ρ g ------- + ρ g u ⋅ ∇u = – ∇p + ∇ ⋅ τ + F
∂t

where

• ρg is the moist air density (SI unit: kg/m3)

• u is the velocity vector (SI unit: m/s)
• p is the pressure (SI unit: Pa)
• τ is the viscous stress tensor (SI unit: Pa), equal for a compressible fluid to:
2
τ = μ g ( ∇u + ( ∇u ) T ) – --- μ ( ∇ ⋅ u )I
3

• μg is the moist air dynamic viscosity (SI unit: Pa·s)

• F is the body force vector (SI unit: N/m3)

It also solves the equation for moisture transport in air, given in Equation 4-128 by

∂c v
M v --------- + M v u ⋅ ∇c v + ∇ ⋅ g w = G
∂t

where

• Mv is the molar mass of water vapor (SI unit: kg/mol)

• cv is the vapor concentration (SI unit: mol/m3)
• gw is the moisture flux by diffusion (SI unit: kg/(m2·s)):
g w = – M v D ∇c v

• D (SI unit: m2/s) is the vapor diffusion coefficient in air.

• G contains moisture sources (SI unit: kg/(m3·s))

When the spatial and temporal variations of the vapor concentration induce significant
variations of the moist air density, the moisture content variation is expressed through
the transport of vapor mass fraction, ωv, defined as:

Mv cv
ω v = --------------
ρg

The equation for moisture transport in air is then expressed as follows:

304 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ∂ω v
ρ g ---------- + ρ g u ⋅ ∇ω v + ∇ ⋅ g w = G
∂t

with

g w = – ρ g D ∇ω v

STEFAN FLOW AT WALLS

Surface reactions like evaporation and condensation result in a net moisture flux gw
between the boundary and the domain. The resulting effective velocity at the
boundary is the Stefan velocity uStefan, defined as

ρ g u Stefan = n ⋅ g w

The Moisture Flow interface allows to use the Stefan velocity as a leakage velocity at
walls in the fluid flow equations.

TURBULENT MOISTURE FLOW THEORY

The Favre average c˜v of the vapor concentration cv is defined by

ρg cv
c˜v = -----------
ρg

where the bar denotes the usual Reynolds average. The full field can be decomposed
as the sum of the Favre average and the Favre fluctuation:

c v = c˜v + c v ″

By applying this decomposition and averaging the moisture transport equations,

unclosed terms are introduced in the equations. Among them, the turbulent moisture
transport flux contains the velocity and vapor concentration fluctuations:

∂c v ″
u j ″ -----------
∂x j

This flux is modeled using a gradient based assumption, where the additional transport
is related to the turbulent viscosity νT through the turbulent Schmidt number ScT.
The turbulent diffusivity DT is defined by

νT
D T = ----------
Sc T

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 305

 The following moisture transport equation is solved:

∂c˜v ∂c˜v ∂  ∂c˜v

M v -------- + M v u˜ j -------- + M v  – ( D + D T ) -------- = G (4-147)
∂t ∂x j ∂ xj  ∂x j 

Kays-Crawford Model for Turbulent Diffusivity

By analogy with turbulent heat transfer, the turbulent Schmidt number ScT is given by
(Ref. 29)

 1 0.3μ T Sc μ T Sc 2 μ  –1
Sc T =  ----------------- + ----------------------- –  0.3 -------------  1 – exp  – -----------------------------------------   (4-148)
 2Sc T∞ μ Sc T∞  μ    0.3μ Sc Sc   
T T∞

where the Schmidt number at infinity is ScT∞ = 0.85 and the Schmidt number is
defined as

ν
Sc = ----
D

where ν is the kinematic viscosity.

Moisture Transport Wall Functions

Analogous to the single-phase flow wall functions (see Wall Functions described for the
Wall boundary condition), there is a theoretical gap between the solid wall and the
computational domain for the fluid and relative humidity fields. This gap is often
ignored when the computational geometry is drawn.

Assuming that the turbulent heat and moisture transfer in the near-wall region are
analogous, the same type of wall functions used for the temperature (Ref. 31) is also
applicable for the moisture transport. The moisture transfer wall function is formulated
as a function of the turbulent Schmidt number, instead of the corresponding Prandtl
number.

The moisture flux at the lift-off position between the air with vapor mass fraction ωv,a
and a wall with vapor mass fraction ωv,w, is:

ρ g u τ ( ω v, w – ω v, a )
g wf = -----------------------------------------------
-
ωw +

where uτ is the friction velocity, and ω w

+ is the dimensionless mass fraction given by

(Ref. 31):

306 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

  + + +
 Scδ w for δ w < δ w1

 2 / 3 – 500 + + +
---------- for δ w1 ≤ δ w < δ w2
φw
+ =  15Sc δw +2

 Sc T
 ---------- ln δ +w + β +
for δ w2 ≤ δ w
+
 κ

where in turn

+ δ w ρ C μ1 / 2 k + 10 -
δ w = ------------------------------- δ w1 = -------------
μ Sc 1 / 3
+ κ μ
δ w2 = 10 10 ---------- Sc = --------
Sc T ρD
Sc T κ
β = 15Sc 2 / 3 – ----------  1 + ln  1000 ---------- 
2κ   Sc T 

where κ is the von Karman constant equal to 0.41, Cμ is a turbulence modeling

constant, and k is the turbulent kinetic energy.

The distance between the computational fluid domain and the wall, δw, is always hw/2
for automatic wall treatment where hw is the height of the mesh cell adjacent to the
wall. hw/2 is almost always very small compared to any geometrical quantity of interest,
at least if a boundary layer mesh is used. For wall function, δw is at least hw/2 and can
be bigger if necessary to keep δw+ higher than 11.06. The computational results should
be checked so that the distance between the computational fluid domain and the wall,
δw, is everywhere small compared to any geometrical quantity of interest. The distance
δw is available for evaluation on boundaries.

THEORY FOR THE SCREEN BOUNDARY CONDITION IN MOISTURE FLOW

When the Moisture Flow multiphysics coupling feature is active, the conditions that
apply across a screen in moisture flow are complemented by:

+
[ cv ]- = 0 (4-149)

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 307

 to ensure mass conservation.

• See Screen for the feature node details.

• Also see Screen boundary condition described for the single-phase flow
interfaces.

THEORY FOR THE INTERIOR FAN BOUNDARY CONDITION IN MOISTURE

FLOW
When the Moisture Flow multiphysics coupling feature is active, the conditions that
apply across an interior fan are complemented by:

• If direction is Along normal vector, the outlet vapor concentration cavg is defined
by:
 

 down ( ρ g u ⋅ n ) dS c avg =
   down ( ρg u ⋅ ncv ) dS
Γ Γ

• If the direction is opposite to normal vector, the outlet vapor concentration cavg is
defined by:
 

 up ( ρ g u ⋅ n ) dS c avg =
   up ( ρg u ⋅ ncv ) dS
Γ Γ

Theory for the Thermoelectric Effect Interface

The Thermoelectric Effect Interface implements thermoelectric effect, which is the
direct conversion of temperature differences to electric voltage or vice versa. Devices
such as thermoelectric coolers for electronic cooling or portable refrigerators rely on
this effect. While Joule heating (resistive heating) is an irreversible phenomenon, the
thermoelectric effect is, in principle, reversible.

Historically, the thermoelectric effect is known by three different names, reflecting its
discovery in experiments by Seebeck, Peltier, and Thomson. The Seebeck effect is the
conversion of temperature differences into electricity, the Peltier effect is the
conversion of electricity to temperature differences, and the Thomson effect is heat
produced by the product of current density and temperature gradients. These effects
are thermodynamically related by the Thomson relations:

P = ST

308 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 dS
μ Th = T
dT

where P is the Peltier coefficient (SI unit: V), S is the Seebeck coefficient
(SI unit: V/K), T is the temperature (SI unit: K), and μTh is the Thomson coefficient
(SI unit: V/K). These relations show that all coefficients can be considered different
descriptions of one and the same quantity. The COMSOL formulation primarily uses
the Seebeck coefficient. The Peltier coefficient is also used as an intermediate variable,
but the Thomson coefficient is not used.

When simulating the thermoelectric effect, the following fluxes are the quantities of
interest:

• Conductive heat flux q, defined by

q = – k∇T + PJ (4-150)

• Electric current density J, defined by

J = – σ ( ∇V + S∇T ) (4-151)

Thermoelectric efficiency is measured by the figure of merit Z (SI unit: 1/K), defined
as:

S2 σ
Z = ----------
k

where σ is the electrical conductivity and k the thermal conductivity.

Some other quantities of relevance are the electric field E and the Joule heat source Q:

E = – ∇V

Q = J⋅E

From these definitions, conservation of heat energy and electrical current in an

immobile solid reads:

∂T
ρC p +∇⋅q = Q
∂t

∇ ⋅ J = Qj

where ρ is the density, Cp the heat capacity, and Qj is the current source.

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 309

 HOW THE SEEBECK, PELTIER, AND THOMSON EFFECTS ARE INCLUDED
IN THE GENERAL FORMULATION
The general formulation of thermoelectric effect redefines the heat flux and the electric
current according to Equation 4-150 and Equation 4-151, respectively. This
formulation does not necessarily correspond to the formulation used when only a
particular aspect of thermoelectric effect is considered: Seebeck, Peltier, or Thomson.
This paragraph describes how these separated effects can be recognized in the general
formulation.

Seebeck Effect
The Seebeck effect is described as the conversion of temperature gradient into electric
current. The contribution of the Seebeck effect is defined as a current contribution

J Se = – σS∇T

This formulation corresponds directly to Equation 4-151 used in the general

formulation.

Peltier Effect
The Peltier effect is described as the conversion of t electric current in heat source or
sink. It is defined as an heat source contribution

Q Pe = – P∇ ⋅ J

This contribution is obtained by developing the divergence of q term in the heat

equation when q is defined following Equation 4-150.

Thomson Effect
The Thomson effect defines the heat source induced by a current in presence of a
temperature gradient in thermoelectric material. The heat source is defined by

Q Th = – μ Th J ⋅ ∇T

This contribution is obtained again by developing the divergence of the q term in the
heat equation when q is defined following Equation 4-150. This time consider the
term −TJ ⋅ ∇S. Assuming that S is function of T, then:

dS
– TJ ⋅ ∇S = – T J ⋅ ∇T = – μ Th J ⋅ ∇T
dT

310 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for the Building Materials Version of the Heat and Moisture
Transport Interface
The Building Materials version of the Heat and Moisture multiphysics coupling
implements the following equations for heat and moisture transport, derived from
Ref. 16:

∂T
( ρC p ) eff ------- + ∇ ⋅ ( – k eff ∇T – L v δ p ∇( φ w p sat ) ) = Q (4-152)
∂t

∂φ w
ξ ---------- + ∇ ⋅ ( – ξ D w ∇φ w – δ p ∇( φ w p sat ) ) = G (4-153)
∂t

See Theory for Heat Transfer in Building Materials and Theory for Moisture Transport
in Building Materials for more details about the notations.

The heat transfer and moisture transport equations are coupled through:

• the definition of the effective thermal properties accounting for moisture content,
• the definition of the latent heat source from the evaporation source in domains,
• the definition of the vapor saturation in function of the temperature.

Theory for the Moist Air Version of the Heat and Moisture Transport
Interface
The Moist Air version of the Heat and Moisture multiphysics coupling implements the
following equations in domains for heat and moisture transport:

∂T
ρ g C p ------- + ρ g C p u ⋅ ∇T + ∇ ⋅ q = Q (4-154)
∂t

q = – k∇T (4-155)

Q = – ( C p, v – C p, a )g w (4-156)

∂c v
M v -------- + M v u ⋅ ∇c v + ∇ ⋅ g w = G (4-157)
∂t

g w = – M v D ∇c v (4-158)

c v = φ w c sat (4-159)

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 311

 See Theory for Heat Transfer in Moist Air and Theory for Moisture Transport in Air
for more details about the notations.

In addition, on boundaries where evaporation or condensation takes place, the

following equation is solved:

q evap = L v g evap (4-160)

The heat transfer and moisture transport equations are coupled through:

• the definition of the effective thermal properties accounting for vapor concentration
in moist air,
• the definition of the diffusive flux of enthalpy (Equation 4-156),
• the definition of the latent heat source from the evaporation source on surfaces
(Equation 4-160),
• the definition of the vapor saturation in function of the temperature
(Equation 4-159).

Theory for the Moist Porous Media Version of the Heat and Moisture
Transport Interface
The Moist Porous Media version of the Heat and Moisture multiphysics coupling
implements the following equations in domains for heat and moisture transport:

∂T
( ρC p ) eff ------- + ( ρ g C p, g u g + ρ l C p, l u l ) ⋅ ∇T + ∇ ⋅ q = Q + Q evap (4-161)
∂t

q = – k eff ∇T (4-162)

Q = – [ ( C p, v – C p, a )g w + C p, l g lc ] (4-163)

Q evap = L v G evap (4-164)

∂w ( φ w )
-------------------- + ρ g u g ⋅ ∇ω v + ∇ ⋅ g w + u l ⋅ ∇ρ l + ∇ ⋅ g lc = G (4-165)
∂t

g w = – ρ g D eff ∇ω v (4-166)

∂ [ εp ρg ωv ( 1 – sl ) ]
G evap = ---------------------------------------------- + ρ g u g ⋅ ∇ω v + ∇ ⋅ g w (4-167)
∂t

312 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

See Theory for Heat Transfer in Moist Porous Media and Theory for Moisture
Transport in Porous Media for more details about the notations.

The heat transfer and moisture transport equations are coupled through:

• the definition of the effective thermal properties accounting for moisture content in
liquid and gas phases,
• the liquid water saturation and velocity field used to defined the convective heat
flux,
• the diffusive flux of enthalpy in moist air and the liquid capillary flux
(Equation 4-163)
• the definition of the latent heat source from the evaporation source in domains
(Equation 4-164),
• the definition of the vapor saturation in function of the temperature.

Theory for the Heat and Moisture Flow Interfaces

The theory for the equations of heat transfer and moisture transport by laminar and
turbulent air flows can be found in Theory for the Moisture Flow Interface and Theory
for the Moist Air Version of the Heat and Moisture Transport Interface.

Theory for the Electromagnetic Heating Interfaces

COMSOL Multiphysics provides four multiphysics interfaces for handling
electromagnetic heating in the Heat Transfer interfaces:

• The Joule Heating Interface

• The Laser Heating Interface (requires the Wave Optics Module)
• The Induction Heating Interface (requires the AC/DC Module for 3D models)
• The Microwave Heating Interface (requires the RF Module)

They all have in common the multiphysics coupling feature Electromagnetic Heating,
which adds weak contributions due to resistive losses in the domains and boundaries,
and shares the temperature variable with the electromagnetics interfaces. The
underlying theory can be found in the AC/DC Module User’s Guide, RF Module
User’s Guide, and Wave Optics Module User’s Guide.

THEORY FOR THE HEAT TRANSFER MULTIPHYSICS COUPLINGS | 313

 Theory for the Thermal Stress Interface
In the Structural Mechanics interfaces, two multiphysics interfaces handle thermal
stress:

• The Thermal Stress, Solid Interface

• The Joule Heating and Thermal Expansion Interface

Both require the Structural Mechanics Module. They have in common the use of the
Thermal Expansion multiphysics coupling that models temperature dependence of the
strain tensor and thermoelastic damping. For more details about the underlying
theory, see the Structural Mechanics Module User’s Guide.

314 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Theory for Thermal Contact
Theory for the Thermal Contact Feature
The Thermal Contact feature node has correlations to evaluate the joint conductance
at two contacting surfaces.

The heat fluxes at the upside and downside boundaries depend on the temperature
difference according to the relations:

– n d ⋅ ( – k d ∇T d ) = – h ( T u – T d ) + rQ b

– n u ⋅ ( – k u ∇T u ) = – h ( T d – T u ) + ( 1 – r )Q b

At a microscopic level, contact is made at a finite number of spots as in Figure 4-23.

masp,u asp,u
Y

Figure 4-23: Contacting surfaces at the microscopic level.

The joint conductance h has three contributions: the constriction conductance, hc,
from the contact spots, the gap conductance, hg, due to the fluid at the interstitial
space, and the radiative conductance, hr:

h = hc + hg + hr

SURFACE ASPERITIES
The microscopic surface asperities are characterized by the average height σu, asp and
σd, asp and the average slope mu, asp and md, asp. The RMS values σasp and masp are
(4.16 in Ref. 1):

σ asp = σ u2, asp + σ d2, asp m asp = m u2, asp + m d2, asp

THEORY FOR THERMAL CONTACT | 315

 CONSTRICTION CONDUCTANCE

Cooper-Mikic-Yovanovich (CMY) Correlation

The Cooper-Mikic-Yovanovich (CMY) correlation is valid for isotropic rough surfaces
and has been formulated using a model assuming plastic deformation of the surface
asperities. However, this model does not compute nor store the plastic deformations
of the asperities. It means that, despite that a plastic deformation of the asperities is
assumed, this contact model has no memory. For example, if a load is applied twice the
thermal contact is identical in both cases. The Cooper-Mikic-Yovanovich (CMY)
correlation relates hc to the asperities and pressure load at the contact interface:

m asp p 0.95
h c = 1.25k contact -------------  -------
σ asp H c

Here, Hc is the microhardness of the softer material, p is the contact pressure, and
kcontact is the harmonic mean of the contacting surface conductivities:

2k u k d
k contact = -------------------
ku + kd

When ku (resp. kd) is not isotropic, it is replaced by its normal

conductivity nTkun (resp. nTkdn).

The relative pressure p ⁄ Hc can be evaluated by specifying Hc directly or using the

following relation (4.16.1 in Ref. 1) for the relative pressure using c1 and c2, the
Vickers correlation coefficient and size index:

  -----------------------------------
1
p-  p  ( 1 + 0.071c 2 )
------ =  -----------------------------------------------------
-
Hc   σ asp  
c2
c
 1 1.62 -----------
- m
σ 0 asp 

The coefficients c1 and c2 are the Vickers correlation coefficient and size index,
respectively, and σ0 is equal to 1 µm. For materials with Brinell hardness between 1.30
and 7.60 GPa, c1 and c2 are given by the correlation below (4.16.1 in Ref. 1):

c1 HB HB 2 HB 3
------- = 4.0 – 5.77 -------- + 4.0  -------- – 0.61  --------
H0 H0  H0   H0 

316 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 HB
c 2 = – 0.37 + 0.442 --------
c1

The Brinell hardness is denoted by HB, and H0 is equal to 3.178 GPa.

Mikic Elastic Correlation

The Mikic correlation is valid for isotropic rough surfaces and assumes elastic
deformations of surface asperities. It gives hc by the following relation:

m asp 0.94
h c = 1.54k contact -------------  ------------------------
2p
σ asp mE contact

Here, Econtact is an effective Young’s modulus for the contact interface, satisfying
(4.16.3 in Ref. 1):

1 - 1 – ν u2 1 – ν d2
------------------ = --------------- + ---------------
E contact Eu Ed

where Eu and Ed are the Young’s moduli of the two contacting surfaces and νu and νd
are the Poisson’s ratios.

GAP CONDUCTANCE
The gap conductance due to interstitial fluid cannot be neglected for high fluid
thermal conductivity or high contact pressure. The parallel-plate gap gas correlation
assumes that the interstitial fluid is a gas and defines hg by:

kg
h g = ------------------
Y + Mg

Here kg is the gas conductivity, Y denotes the mean separation thickness (see
Figure 4-23), and Mg is the gas parameter equal to:

kB Tg
M g = αβΛ Λ = ------------------------
2
-
2πD p g

In these relations, α is the contact thermal accommodation parameter, β is a gas

property parameter (equal to 1.7 for air), Λ is the gas mean free path, kB is the
Boltzmann constant, D is the average gas particle diameter, pg is the gas pressure
(often the atmospheric pressure), and Tg is the gap temperature equal to:

THEORY FOR THERMAL CONTACT | 317

 Tu + Td
T g = --------------------
2

The mean separation thickness, Y, is a function of the contact pressure, p. For low
values of p near 0 Pa, Y goes to infinity since no contact occur. For high values of p —
greater than Hc ⁄2 in the Cooper-Mikic-Yovanovich model and greater than Hc ⁄4 in
the Mikic elastic model — Y reduces to 0 meaning that the contact is considered as
perfect.

RADIATIVE CONDUCTANCE
At high temperatures, above 600°C, radiative conductance needs to be considered.
The gray-diffuse parallel plate model provides the following formula for hr:

εu εd
h r = ----------------------------------- σ ( T u3 + T u2 T d + T u T d2 + T d3 )
εu + εd – εu εd

which implies that:

εu εd
h r ( T u – T d ) = ----------------------------------- σ ( T u4 – T d4 )
εu + εd – εu εd

εu εd
h r ( T d – T u ) = ----------------------------------- σ ( T d4 – T u4 )
εu + εd – εu εd

THERMAL FRICTION
The friction heat, Qb, is partitioned into rQb and (1 − r)Qb at the contact interface. If
the two bodies are identical, r and (1 − r) would be 0.5 so that half of the friction heat
goes to each surface. However, in the general case where the two bodies are made of
different materials, the partition rate might not be 0.5. The Charron’s relation (Ref. 2)
defines r as:

1 ρ u C p, u k u
r = --------------- ξd = -------------------------
1 + ξd ρ d C p, d k d

and symmetrically, (1 − r) is:

1 ρ d C p, d k d
( 1 – r ) = --------------- ξu = -------------------------
1 + ξu ρ u C p, u k u

318 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

For anisotropic conductivities, nTkdn (resp. nTkun) replaces kd (resp. ku).

Thermal Contact

Contact Switch: Application Library path

Heat_Transfer_Module/Thermal_Contact_and_Friction/contact_switch

REFERENCES
1. A. Bejan and others, Heat Transfer Handbook, John Wiley & Sons, 2003.

2. F. Charron, “Partage de la chaleur entre deux corps frottants,” Publication

Scientifique et Technique du Ministère de l’Air, no. 182, 1943.

THEORY FOR THERMAL CONTACT | 319

 Out-of-Plane Heat Transfer
When the object to model in COMSOL Multiphysics is thin or slender enough along
one of its geometry dimensions, there is usually only a small variation in temperature
along the object’s thickness or cross section. For such objects, it is computationally
more efficient to reduce the model geometry to 2D or even 1D and to use the
out-of-plane heat transfer mechanism. Figure 4-24 shows examples of possible
situations in which this type of geometry reduction can be applied.

qup

qdown

Figure 4-24: Geometry reduction from 3D to 1D (top) and from 3D to 2D (bottom).

The reduced geometry does not include all the boundaries of the original 3D
geometry. For example, the reduced geometry does not represent the upside and
downside surfaces of the plate in Figure 4-24 as boundaries.

Out-of-Plane Radiation and Out-of-Plane Heat Flux

Equation Formulation

2D GEOMETRIES
In 2D geometries, the temperature is assumed to be constant in the out-of-plane
direction (z direction with default spatial coordinate names). The equation for heat
transfer in solids, Equation 4-16, and in fluids, Equation 4-18, are replaced by:

320 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 ∂T
d z ρC p ------- + ∇ ⋅ q = d z Q + q 0 (4-168)
∂t

∂T
d z ρC p ------- + ρC p d z u ⋅ ∇T + ∇ ⋅ q = d z Q + q 0 (4-169)
∂t

Here dz is the thickness of the domain in the out-of-plane direction. Here, the
conductive heat flux, q, becomes

q = – d z k∇T

1D AXISYMMETRIC GEOMETRIES
In 1D axisymmetric geometries, the temperature is assumed to be constant in the
out-of-plane direction (z direction with default spatial coordinate names) in addition
to the axisymmetry ( φ -coordinate with default spatial coordinate names). The
equation for heat transfer in solids, Equation 6-15 is replaced by

∂T
( 2πrd z )ρC p ------- + ∇ ⋅ q = ( 2πrd z )Q + q 0 (4-170)
∂t

where dz is the thickness of the domain in the z direction. The equation for heat
transfer in fluids, Equation 6-5, is replaced by

∂T
( 2πrd z )ρC p ------- + ( 2πrd z )ρC p u ⋅ ∇T + ∇ ⋅ q = ( 2πrd z )Q + q 0 (4-171)
∂t

Here, the conductive heat flux, q, becomes

q = – ( 2πrd z )k∇T

1D GEOMETRIES
In 1D geometries, the temperature is assumed to be constant in the radial direction.
The equation for heat transfer in solids, Equation 6-15 is replaced by

∂T
A c ρC p ------- + ∇ ⋅ q = A c Q + q 0 (4-172)
∂t

where Ac is the cross section of the domain in the plane perpendicular to the 1D
geometry. The equation for heat transfer in fluids, Equation 6-5, is replaced by

∂T
A c ρC p ------- + A c ρC p u ⋅ ∇T + ∇ ⋅ q = A c Q + q 0 (4-173)
∂t

Here, the conductive heat flux, q, becomes

OUT-OF-PLANE HEAT TRANSFER | 321

 q = – A c k∇T

Out-of-plane flux conditions would apply to the exterior boundaries of the domain if
the 1D geometry was seen as a cylinder. With the geometry reduction process, this
heat flux condition is mathematically expressed using the cross section perimeter, Pc,
as in:

q 0 = P c q 0, z

where q0, z is the heat flux density distributed along the cross section perimeter.

322 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Convective Heat Transfer
Correlations
One of the most common boundary conditions when modeling heat transfer is
convective cooling or heating whereby a fluid cools or heats a surface by natural or
forced convection. In principle, it is possible to model this process in two ways:

• Using a heat transfer coefficient on the surfaces

• Extending the model to describe the flow and heat transfer in the surrounding fluid

The second approach is the most accurate if the geometry or the external flow is
complicated. The Heat Transfer Module includes the Conjugate Heat Transfer
predefined multiphysics coupling and the CFD Module includes the Nonisothermal
Flow predefined multiphysics coupling for this purpose. However, such a simulation
can become costly, both in terms of computational time and memory requirement.

The first method is simple, yet powerful and efficient. The convective heat flux on the
boundaries in contact with the fluid is then modeled as being proportional to the
temperature difference across a fictitious thermal boundary layer. Mathematically, the
heat flux is described by the equation

– n ⋅ q = h ( T ext – T )

where h is a heat transfer coefficient and Text the temperature of the external fluid far
from the boundary.

The main difficulty in using heat transfer coefficients is in calculating or specifying the
appropriate value of the h coefficient. That coefficient depends on the fluid’s material
properties, and the surface temperature — and, for forced convection, also on the
fluid’s flow rate. In addition, the geometrical configuration affects the coefficient. The
Heat Transfer interface has built-in functions for the heat transfer coefficients. For
most engineering purposes, the use of such coefficients is an accurate and numerically
efficient modeling approach.

In this section:

• Defining the Heat Transfer Coefficients

• Nature of the Flow — The Grashof Number
• Heat Transfer Coefficients — External Natural Convection

CONVECTIVE HEAT TRANSFER CORRELATIONS | 323

 • Heat Transfer Coefficients — Internal Natural Convection
• Heat Transfer Coefficients — External Forced Convection
• Heat Transfer Coefficients — Internal Forced Convection
• Using the Heat and Mass Transfer Analogy for the Evaluation of Moisture Transfer
Coefficients

Defining the Heat Transfer Coefficients

It is possible to divide the convective heat flux into four main categories depending on
the type of convection condition (natural or forced) and on the type of geometry
(internal or external flow). In addition, these cases can all experience either laminar or
turbulent flow conditions, resulting in eight types of convection, as in Figure 4-25.
Natural Forced

External

Internal

Laminar Flow

Turbulent Flow

Figure 4-25: The eight categories of convective heat flux.

The difference between natural and forced convection is that in the forced convection
an external force such as a fan creates the flow. In natural convection, buoyancy forces
induced by temperature differences together with the thermal expansion of the fluid
drive the flow.

Heat transfer books generally contain a large set of empirical and theoretical
correlations for h coefficients. This module includes a subset of them. The expressions
are based on the following set of dimensionless numbers:

• The Nusselt number, NuL = hL ⁄ k

324 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

• The Reynolds number, ReL = ρUL ⁄ μ
• The Prandtl number, Pr = μCp ⁄ k
• The Rayleigh number, RaL = GrLPr

where

• h is the heat transfer coefficient (SI unit: W/(m2·K))

• L is the characteristic length (SI unit: m)
• ΔT is the temperature difference between the surface and the external fluid bulk
(SI unit: K)
• g is the acceleration of gravity (SI unit: m/s2)
• k is the thermal conductivity of the fluid (SI unit: W/(m·K))
• ρ is the fluid density (SI unit: kg/m3)
• U is the bulk velocity (SI unit: m/s)
• μ is the dynamic viscosity (SI unit: Pa·s)
• Cp is the heat capacity at constant pressure of the fluid (SI unit: J/(kg·K))

Further, GrL refers to the Grashof number, which is the squared ratio of the viscous
time scale to the buoyancy time scale multiplied by the Reynolds number.

Nature of the Flow — The Grashof Number

In cases of externally driven flow, such as forced convection, the nature of the flow is
characterized by the Reynolds number, Re, which describes the ratio of the inertial
forces to the viscous forces. However, the velocity scale is initially unknown for
internally driven flows such as natural convection. In such cases the Grashof number,
Gr, characterizes the flow. It describes the ratio of the time scales for viscous diffusion
in the fluid and the internal driving force (the buoyancy force). Like the Reynolds
number it requires the definition of a length scale, the fluid’s physical properties, and
the density scale (densities difference).

The Grashof number, Gr, is defined as:

3
gρ ( ρ ext – ρ s )L
Gr L = -----------------------------------------
2
μ

where g is the acceleration of gravity, ρs denotes the density of the hot surface, ρext
equals the free stream density, L is the length scale, μ represents the fluid’s dynamic
viscosity, and ρ its density.

CONVECTIVE HEAT TRANSFER CORRELATIONS | 325

 This expression of the Grashof number accounts for the all the density variations in the
fluid. This applies to moist air, for which the density depends on temperature and
vapor concentration.

When the density depends on temperature only, in dry air for example, the Grashof
number is well approximated by using the fluid’s coefficient of thermal expansion αp
(SI unit: 1/K):

3
gα p ( T s – T ext )L
Gr L = ----------------------------------------------
2
(μ ⁄ ρ)

where g is the acceleration of gravity, Ts denotes the temperature of the hot surface,
Text equals the free stream temperature, L is the length scale, μ represents the fluid’s
dynamic viscosity, and ρ its density.

In general, the coefficient of thermal expansion αp is given by

1 ∂ρ
α p = – ---  -------
ρ  ∂T p

which for an ideal gas reduces to

1
α p = ----
T

The transition from laminar to turbulent flow occurs at a Gr value of 109; the flow is
turbulent for larger values.

The Rayleigh number, Ra, is another indicator of the regime. It is similar to the
Grashof number except that it accounts for the thermal diffusivity: Ra = Pr Gr. A small
value of the Ra number indicates that the conduction dominates. It such case using
heat transfer coefficients to model convective heat transfer is not relevant. Instead,
modeling the fluid as immobile is likely to be accurate.

326 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Heat Transfer Coefficients — External Natural Convection

VERTICAL WALL

Figure 4-26: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on a vertical wall.

The following correlations correspond to equations 9.26 and 9.27 in Ref. 21:


  1⁄4 
 --- k-  0.67Ra L 
 L 0.68 + ---------------------------------------------------------
-  if Ra L ≤ 10 9
9 / 16 4 / 9
   1 +  -------------------
0.492k  
    μC p   
h =  (4-174)
  1⁄6 2
 k 0.387Ra L 
 ----  0.825 + ------------------------------------------------------------
9 / 16 8 / 27
- if Ra L > 10 9
 
L  1 +  -------------------
0.492k  
    μC p   


where the height of the wall, L, is a correlation input.

The Rayleigh number is defined as:

2
gα p ρ C p T – T ext L 3
Ra L = -------------------------------------------------------- (4-175)
kμ

if the density depends on temperature only, or as:

gρC p ρ ext – ρ s L 3
Ra L = ----------------------------------------------- (4-176)
kμ

if the density depends on other parameters.

All material properties are evaluated at (T + Text) ⁄ 2, except ρs, which is evaluated at
the wall temperature, T, and g is the acceleration of gravity equal to 9.81 m/s2. This

CONVECTIVE HEAT TRANSFER CORRELATIONS | 327

 correlation is valid for 104≤ RaL ≤ 1013. The laminar-turbulent transition at RaL = 109
is handled by the use of a smoothed Heaviside function with a transition of size 108,
which corresponds to 10% of the threshold value.

INCLINED WALL

Figure 4-27: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on an inclined wall.

The following correlations correspond to equations 9.26 and 9.27 in Ref. 21 (the
same as for a vertical wall):


  
 --- k-  0.67 ( cos ϕRa L ) 1 / 4 
 L  0.68 +  ---------------------------------------------------------- if Ra L ≤ 10 9
9 / 16 4 / 9
  1 +  0.492k
-------------------   
    μC p   
h =  (4-177)
  1⁄6 2
 k 0.387Ra L 
 ----  0.825 + ------------------------------------------------------------
/ /
- if Ra L > 10 9
 L  1 +  0.492k -------------------
9 16
 8 27

    μC p   


where the length of the wall, L, is a correlation input and ϕ is the tilt angle (the angle
between the wall and the vertical direction; ϕ = 0 for vertical walls). These
correlations are valid for −60° < ϕ < 60° and 104≤ RaL ≤ 1013.

The definition of the Raleigh number, RaL, is analogous to the one for vertical walls
and is given by the following:

2
gα p ρ C p T – T ext L 3
Ra L = -------------------------------------------------------- (4-178)
kμ

if the density depends on temperature only, or as:

328 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

 gρC p ρ ext – ρ s L 3
Ra L = ----------------------------------------------- (4-179)
kμ

if the density depends on other parameters.

g denotes the gravitational acceleration, equal to 9.81 m/s2.

For turbulent flow, 1 is used instead of cos ϕ in the expression for h, because this gives
better accuracy (see Ref. 41).

According to Ref. 21, correlations for inclined walls are only satisfactory
for the top side of a cold plate or the down face of a hot plate. Hence,
these correlations are not recommended for the bottom side of a cold face
and for the top side of a hot plate.

The laminar-turbulent transition depends on ϕ (see Ref. 41). Unfortunately, little

data is available about transition. There is some data available in Ref. 41 but this data
is only approximate, according to the authors. In addition, data is only provided for
water (Pr around 6). For this reason, the flow is defined as turbulent, independently
of the ϕ value, when

Ra L > 10 9

All material properties are evaluated at (T + Text) ⁄ 2, except ρswhich is evaluated at the
wall temperature, T. The laminar-turbulent transition at RaL = 109 is handled by the
use of a smoothed Heaviside function with a transition of size 108, which corresponds
to 10% of the threshold value.

HORIZONTAL PLATE, UPSIDE

Figure 4-28: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on the top surface of an horizontal
plate.

CONVECTIVE HEAT TRANSFER CORRELATIONS | 329

 The following correlations correspond to equations 9.30–9.32 in Ref. 21 but can also
be found as equations 7.77 and 7.78 in Ref. 41.

If ρ < ρext, then


 k-
---
1⁄4
0.54Ra L if 10 4 ≤ Ra L ≤ 10 7
 L
h =  (4-180)
 k-
---
1⁄3
0.15Ra L if 10 7 ≤ Ra L ≤ 10 11
 L


while if ρ ≥ ρext, then

k 1⁄4
h = ---- 0.27Ra L if 10 5 ≤ Ra L ≤ 10 10 (4-181)
L

RaL is given by Equation 4-175 or Equation 4-176, and L, the characteristic length
(defined as area/perimeter, see Ref. 41) is a correlation input. The material data are
evaluated at (T + Text) ⁄ 2, except ρswhich is evaluated at the wall temperature, T.

When the density depends only on temperature, the conditions ρ < ρext and ρ ≥ ρext
can be replaced by T > Text and T ≤ Text respectively.

The laminar-turbulent transition at RaL = 107 is handled by the use of a smoothed

Heaviside function with a transition of size 106, which corresponds to 10% of the
threshold value.

HORIZONTAL PLATE, DOWNSIDE

Figure 4-29: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on the bottom surface of an
horizontal plate.

330 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

Equation 4-180 is used when ρ ≥ ρext (or T ≤ Text) and Equation 4-181 is used when
ρ < ρext (or T > Text). Otherwise it is the same implementation as for Horizontal Plate,
Upside.

LONG HORIZONTAL CYLINDER

Figure 4-30: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on a long horizontal cylinder.

The following correlations correspond to equations 9.34 in Ref. 21. It is validated for
RaD ≤ 1012.

 1⁄6 2
k  0.387Ra D 
h = ----  0.6 + ---------------------------------------------------------
⁄ ⁄
- (4-182)
D  1 +  0.559
9 16 8 27

 --------------- 

  Pr  

Here D is the cylinder diameter and RaD is given by

2
gα p ρ C p T – T ext D 3
Ra D = ---------------------------------------------------------
kμ

if the density depends on temperature only, or as:

gρC p ρ ext – ρ s D 3
Ra D = ------------------------------------------------
kμ

if the density depends on other parameters.

The material data are evaluated at (T + Text) ⁄ 2, except ρswhich is evaluated at the wall
temperature, T.

CONVECTIVE HEAT TRANSFER CORRELATIONS | 331

 SPHERE

Figure 4-31: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on a sphere.

The following correlations correspond to equation 9.35 in Ref. 21. It is validated for
RaD ≤ 1011 and Pr ≥ 0.7.

 1⁄4 
k  0.589Ra D 
h = ----  2 + ------------------------------------------------------
⁄ ⁄
- (4-183)
D  1 +  0.469
9 16 4 9

 --------------- 

  Pr  

Here D is the cylinder diameter and RaD is given by

2
gα p ρ C p T – T ext D 3
Ra D = ---------------------------------------------------------
kμ

if the density depends on temperature only, or as:

gρC p ρ ext – ρ s D 3
Ra D = ------------------------------------------------
kμ

if the density depends on other parameters.

The material data are evaluated at (T + Text) ⁄ 2, except ρswhich is evaluated at the wall
temperature, T.

332 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

VERTICAL THIN CYLINDER

Figure 4-32: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection on a vertical thin cylinder.

The following correlation corresponds to equation 7.83 in Ref. 41. It evaluates the
heat transfer coefficient on the lateral surface of the thin cylinder. The horizontal disks
(top and bottom) should be treated as horizontal plates.

For the lateral surface, the heat transfer coefficient reads:

7Ra H Pr 1⁄4
4 ( 272 + 315Pr )H
h = -----  ---  ----------------------------------- + -----------------------------------------------
k 4
 
H 3 5 ( 20 + 21Pr )  35 ( 64 + 63Pr )D 

where D is the cylinder diameter, H is the cylinder height, and RaH is given by

3
gα p T – T ext H
Ra H = -------------------------------------------
kμ

if the density depends on temperature only, or as:

gρC p ρ ext – ρ s H 3
Ra H = ------------------------------------------------
kμ

if the density depends on other parameters.

The material data are evaluated at (T + Text) ⁄ 2, except ρswhich is evaluated at the wall
temperature, T.

If the thermal boundary layer thickness δT is much smaller than the cylinder diameter
D, the curvature does not play any significant role, and vertical wall correlations should
be used. In practice, the (δT < D) criterion requires

1
– ---
D- 4
---- > Ra H
H

CONVECTIVE HEAT TRANSFER CORRELATIONS | 333

 Heat Transfer Coefficients — Internal Natural Convection

NARROW CHIMNEY, PARALLEL PLATES

Figure 4-33: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection in a parallel-plate narrow chimney.

If RaL < H ⁄ L, then

k 1
h = ----- ------ Ra L (4-184)
H 24

where the plate distance, L, and the chimney height, H, are correlation inputs
(equation 7.96 in Ref. 41). RaL is given by Equation 4-175 or Equation 4-176. The
material data are evaluated at (T + Text) ⁄ 2.

NARROW CHIMNEY, CIRCULAR TUBE

Figure 4-34: Schematic representation of geometry and parameters for the heat transfer
coefficient correlation applied to natural convection in a circular narrow chimney.

If RaD < H ⁄ D, then

k 1
h = ----- ---------- Ra D
H 128

334 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

where the tube diameter, D, and the chimney height, H, are correlation inputs (table
7.2 in Ref. 41 with Dh = D). RaD is given by Equation 4-175 or Equation 4-176 with
L replaced by D. The material data are evaluated at (T + Text) ⁄ 2.

Heat Transfer Coefficients — External Forced Convection

PLATE, AVERAGED TRANSFER COEFFICIENT

Figure 4-35: Schematic representation of geometry and parameters for the averaged heat
transfer coefficient correlation applied to forced convection on an horizontal plate.

This correlation is a combination of equations 7.33 and 7.41 in Ref. 21:

 1⁄2
 0.3387Pr 1 / 3 Re L
k- ------------------------------------------------------
 2 --- - if Re L ≤ 5 ⋅ 10 5

L 0.0468- 2 / 3 1 / 4
 ------------------
h = 1 + (4-185)
   Pr  

 2 ---- Pr 1 / 3 ( 0.037Re 4 ⁄ 5 – 871 ) if Re > 5 ⋅ 10 5
k
 L L L

where Pr = μCp ⁄ k and ReL = ρUL ⁄ μ. The plate length, L, and the exterior velocity,
U, are correlation inputs. The material data are evaluated at (T + Text) ⁄ 2.

The laminar-turbulent transition at ReL = 5.105 is handled by the use of a smoothed

Heaviside function with a transition of size 5.104, which corresponds to 10% of the
threshold value.

CONVECTIVE HEAT TRANSFER CORRELATIONS | 335

 PLATE, LOCAL TRANSFER COEFFICIENT

Figure 4-36: Schematic representation of geometry and parameters for the local heat
transfer coefficient correlation applied to forced convection on an horizontal plate.

This correlation corresponds to equations 5.79b and 5.131’ in Ref. 41:


 k 1⁄2
--- 0.332Pr 1 / 3 Re x if Re x ≤ 5 ⋅ 10 5
 x
h =  (4-186)
 k 4⁄5
--- 0.0296Pr 1 / 3 Re x if Re x > 5 ⋅ 10 5
 x


where Pr = μCp ⁄ k and Rex = ρUx ⁄ μ. The correlation inputs are x, the position along
the plate, and U, the exterior velocity. The material data are evaluated at (T + Text) ⁄ 2.
To avoid division by zero when the position along the plate is located at the origin
point (x = 0), the implementation replaces k ⁄ x by k ⁄ max(x, √ε) where ε is the
floating-point relative accuracy.

The laminar-turbulent transition at Rex = 5.105 is handled by the use of a smoothed

Heaviside function with a transition of size 5.104, which corresponds to 10% of the
threshold value.

336 | CHAPTER 4: THEORY FOR THE HEAT TRANSFER MODULE

CYLINDER IN CROSS-FLOW

Figure 4-37: Schematic representation of geometry and parameters for the averaged heat
transfer coefficient correlation applied to forced convection on a cylinder in cross-flow.

This correlation is given by equation 7.54 in Ref. 21:

 1⁄2 
k  0.62Re D Pr 1 / 3   Re D  5 ⁄ 8 4 ⁄ 5
2 ⁄ 3 1 ⁄ 4
- 1 + --------------------
h = ----  0.3 + ----------------------------------------------  282000   if Re D Pr ≥ 0.2 (4-187)
D  1 +  0.4  
 --------
- 
  Pr  

where Pr = μCp ⁄ k and ReD = ρUD ⁄ μ. The cylinder diameter, D, and the exterior
velocity, U, are correlation inputs. The material data are evaluated at (T + Text) ⁄ 2.

SPHERE

Figure 4-38: Schematic representation of geometry and parameters for the averaged heat
transfer coefficient correlation applied to forced convection on a sphere.

This correlation corresponds to equation 7.56 in Ref. 21:

0.4 μ 1 ⁄ 4
h = ----  2 + ( 0.4Re D + 0.06Re D )Pr  ----- 
k 1⁄2 2⁄3
(4-188)
D μs

CONVECTIVE HEAT TRANSFER CORRELATIONS | 337

 
 3.5 ≤ Re D ≤ 7.6 ⋅ 10
4

if  0.71 ≤ P r ≤ 380
 μ
 1 ≤ ------------- ≤ 3.2
 μ ( T)

where Pr = μCp ⁄ k and ReD = ρUD ⁄ μ. The sphere diameter, D, and the exterior
velocity, U, are correlation inputs. All material data are evaluated at Text except μs,
which is evaluated at the wall temperature, T.

Heat Transfer Coefficients — Internal Forced Convection