1

Lecture #1: Stagnation Point Heating

 Background 2

• The kinetic energy of an entry vehicle is dissipated by

transformation into thermal energy (heat) as the entry system
decelerates

• The magnitude of this thermal energy is so large that if all of this

energy were transferred to the entry system it would be severely
damaged and likely vaporize
– Harvey Allen - the blunt body concept

• Only a small fraction of this thermal energy is transferred to the

entry system
– The thermal transfer fraction is dependant on vehicle shape, size,
aerodynamic regime and velocity
– Near peak heating, 1% to 5% of the total thermal energy is transferred to the
entry system
– Example: at the peak heating point the freestream energy transfer for
Pathfinder was qÝ  12 V 3 ~ 4,000 W/cm2 but only about 110 W/cm2 (2.7%) was
actually transferred to the surface


 Example 3

E V2
Energy density:   goh
m 2
V E/m
Entry
(km/s) (MJ/kg)

MER 5.6 16
Note that:
Apollo 11.4 66 Water boils @ 2.3 MJ/kg
Carbon vaporizes @ 60.5 MJ/kg
Mars
14.0 98
Return

Galileo 47.4 1130

In each case goh is about 1% of total

Side Note: What Can We Test? 4

Missions
of Interest
Live here
 Blunt Body Rationale 5

• Why is a blunt body used for

planetary entry?
– Slender body: low drag, highly
maneuverable
– Blunt body: high drag, not very
maneuverable

• Blunt bodies generate strong

shock waves
– Efficient energy dissipation. Shock
waves convert kinetic energy to
internal energy. Result is: heating
of the gas, dissociation, ionization
– Most of this energy is convected into the vehicle
wake rather than transported to the surface
– Intuitively, blunter is better (more bluntness equals
stronger shock). Hold that thought; we will come back
to it…
 Blunt Body Rationale (2) 6

Apollo Wake Flow

• Normal shock heats the gas to

many thousands of degrees
• Much of this heat is conducted
into the vehicle wake and
propogated downstream
• Can be tracked as a
“velocity deficit” and persists
long downstream of the
vehicle
 Definitions 7

• Heat Rate (q)

– Instantaneous heat flux at a point on the vehicle (W/cm2)

• Heat Load (Q)

– Integration of heat rate with time over a trajectory (J/cm2)

• Convective Heating
– Heat flux to the vehicle from conduction ( gradT)

• Catalytic Heating
– Heat flux to the vehicle due to surface facilitated chemical reactions
– Commonly lumped with convective heating by convention

• Radiative Heating
– Heat flux to the vehicle from radiation produced by excited atoms and
molecules in the shock layer
 What is Aerothermodynamics? 8

• Accurate and conservative prediction of the heating

environment encountered by an Earth or planetary entry
vehicle
• Aerothermal modeling is coupled and entwined with
Thermal Protection System (TPS) design
• The TPS is designed to withstand the predicted environment with risk-
appropriate margin
• For ablative systems, the flowfield and TPS interact with each other in
non-reversible manner; the physics themselves are coupled

• At its core, aerothermodynamics becomes the study of

an energy balance at the surface of the material

 Heat flux (with pressure & shear) used to select TPS material
 Heat load determines TPS thickness
 Principles of Aerothermal Models 9

Planetary Atmospheres
Mars&Venus: CO2/N2 Thermal Protection
Titan: N2/CH4 System (TPS)
Giants: H2/He Surface Energy
Earth: N2/O2 Balance

Hot Shock Layer

(up to 20000 K)
Thermochemical qrerad
nonequilibrium,
Ionization, Radiation qcond
qc

Boundary Layer qrad qmdot

(2–6000 K)
Transport properties,
Ablation product
mixing, Radiation
blockage
Design Problem: Minimize conduction
V into vehicle to minimize TPS mass/risk
qcond = qc + qrad – qrerad – qmdot
“Cool” Surface
(2–3000 K) Incident Aeroheating
Surface kinetics,
Ablation Material Response
 Current State of the Art : CFD 10

• The current SOA involves the steady solution of the reacting

Navier-Stokes equations via CFD or DSMC methods
• Full 3D simulations possible in hours to days
• Longer time required for the simulation of OML details (steps,
gaps, seals, windows, etc.
 Pushing the Current State of the Art 11

• DES, DNS, LES

• Unsteady RANS (URANS) simulations of Supersonic Retro-
Propulsion flowfields; going on right now…
 NASA CFD Development Strategy 12

LAURA
DPLR
Today
• Structured, Finite Volume, mostly steady-state
• Also coupled to Radiation and Ablation codes

US3D-NASA
FUN3D (LAURA-path) In 2-3 Years
• Unstructured, Finite Volume, low-dissipation schemes,
DES/LES, DNS capability, well-balanced schemes

DG (Discontinuous Galerkin)
CESE (Conservation Element Solution In 5-10 Years
Element)
• Unstructured, higher order, unsteady, beyond finite
volume
 Why Engineering Methods? 13

With present computational abilities, why use engineering methods?

• CFD is a powerful tool, but high-fidelity simulations remain time (and

resource) consuming

• Some applications of simple relationships for calculating non-ablating

convective and radiative heating

– Negligible computation time

– Included in most atmospheric trajectory codes-stag. pt. heating
– Initial estimates of heating rates and loads for use during
conceptual design stage

• But most important:

In this day of commodity supercomputers it is all too easy to run simulations
without truly understanding the physics involved or the trends that are
expected. The fact that it “converged” doesn’t make it right. Engineering
methods are based on sound approximations to theory and provide a valuable
sanity check on CFD results
 Theory of Stag. Pt. Convective Heat Transfer 14

• Pioneering engineering theories were developed in the

1950’s (missile technology)
Lees, L. “Laminar Heat Transfer Over Blunt-Nosed Bodies at Hypersonic
Speeds,” Jet Propulsion, pp. 256-269, Apr. 1956

Fay, J.A. and Riddell, F.R., “Theory of Stagnation Point Heat Transfer in
Dissociated Air,” Journal of Aeronautical Sciences, Feb. 1958

• Extensions to higher velocities were required to account

for chemistry and ionization
• Many extensions and simplifications followed for specific
applications, non-Earth atmospheres
 Theory of Stag. Pt. Convective Heat Transfer (2) 15

• Early correlations for convective heating have the form:

1
   
3
2

qÝs ~ V    
 Rn  
• Why?

• At first cut, one

 might expect heat flux to the3 surface to be
proportional to freestream energy flux ( 12 V )
• From previous discussion one would expect convective
heat flux to decrease as bluntness (Rn) increases, but with
what functionality? 

• (insert brief derivation here)

 Fay-Riddell Method 16

 Convective: derived from boundary

layer and stagnation point theories
w = wall
e = edge
Fay & Riddell (1958):
Boundary layer eqns, similarity transformation

Velocity gradient from mod. Newtonian theory ~(1/Rn)

due 1 2pe  p 

dx R e

Significant advance, but still requires many quantities that are

not readily available to designer

Allows for chemistry effects, non-unity Pr, Le (Prandtl, Lewis

numbers)
 Simplified Methods 17

Chapman Equation (Earth):

1
   2 3   hw  
4 
q s  1.63  10     V  1   
 Rn     h  
T
h   C pTdt  12 V2
0
“hot wall correction” can
frequently be neglected in
 hypersonic flow (hw << h∞)
Sutton Graves: 
1
    2 3
q s  k    V k = 1.7415e-4 (Earth)
 Rn   k = 1.9027e-4 (Mars)
(SI units)

• Calculated

for specific atmosphere (Earth or Mars),
accounting for thermodynamics.
• Above assume a fully catalytic surface; equivalent
expressions for non catalytic wall are available.
 Hot Wall Correction Term 18

• Negligible above about 100 W/cm2 assuming radiative equilibrium

• Actual effect is smaller than this for ablative TPS
0.7

0.6

 
0.5
hw
HWC  1
Enthalpy Ratio

h
0.4

0.3


0.2
Radiative Equilibrium

0.1
Approximate
Ablative Correction
0
1 10 100 1000
q (W/cm2) – log scale
 Generalized Chapman Method 19

C  hw 
qc,0  ( ) (V ) 1 ;
m n

Rn  h 
Earth : m = 0.5, n=3
Mars: m = 0.5, n = 3.04

Cis derived for problem of interest

Powerful design tool - can be used to approximate heating from
a small number of CFD “anchor points” even away from the
stagnation point by letting C, m, and n be curve fit coefficients
Comparison of Data to Correlations 20
 Nuance – Effective Nose Radius 21

• Prior correlations are straightforward and require only

readily available quantities
• However, there is a nuance. All are dependent on the
effective nose radius of the vehicle under investigation
• For a hemisphere, Reff = Rn, but corrections are required for
other vehicle shapes.
• For example, Apollo was a truncated sphere, with an
effective radius almost twice the base radius of the
capsule. MER/MSL use sphere-cones, where the conical
flank increases the effective radius of the nose
• For bodies with a rounded corner, Zoby and Sullivan have
computed tables of effective radius as a function of Rb/Rn
and Rc/Rb:
Zoby, E. and Sullivan E, “Effects of Corner Radius on Stagnation Point Velocity Gradients on
Blunt Axisymmetric Bodies,” Journal of Spacecraft and Rockets, Vol. 3, No. 10, 1966.
 Nuance – Effective Nose Radius (2) 22

When does it matter?

Can the flow “tell” that the nose is finite?

45° Sphere-Cone 60° Sphere-Cone

Supersonic Oblique Shock Subsonic Shock
Reff = Rn Reff > Rn
 Theory of Stag. Pt. Radiative Heat Transfer 23

•Theory is less intuitive, more involved Upper Level (U)

2h
Stimulated
Emission
•Atoms or molecules are excited by h
Spontaneous
Emission
collisions. Excited species can emit a h
photon that carries energy with it
•Photons are emitted isotropically, and
travel effectively instantaneously h = DE = EU-EL
•Radiative heating is the integration of
those photons that hit the surface times
the energy they carry; intuitively should
be proportional to the size of the Absorption
radiating volume Lower Level (L) h

•Partition functions for excited states

imply a near exponential dependence on Ei Ei
 
temperature Ni gi e kT
gi e kT
 
•Radiation is coupled to the fluid N 
Ej
Q
mechanics for two reasons: g e j
kT

• Emitted photons carry energy out of control j

volume (adiabatic cooling)
• Photons can be absorbed in the boundary LTE-Plasma
layer and heat the gas
 Relative Importance of Radiation vs Convection 24

Radiative (neglecting coupling effects)

Radiative (including coupling effects)

Convective

Nose Radius = 4.5m

Altitude = 60 km

Adapted from Anderson, Hypersonic and High Temperature Gas Dynamics, Fig. 18.10
 Theory of Stag. Pt. Radiative Heat Transfer 25

Martin:
q r ~ rn  V
Ý 1.0 1.6 8.5
 Earth

Direct dependence on Rn agrees with

intuitive argument about radiating volume

Tauber-Sutton:
 qr  Ci rn  f i V 
Ý a m
 based on tabulated data,
Earth : a ~ 1, m ~ 1.2 equilibrium shock theory
Mars: a = 0.526, m ~ 1.2
fi are tabulated, near exponential
at moderate velocity


Theory is less intuitive, more involved. Typically relies on table
lookups and has limited range of validity
Fortunately, radiation is not a major issue for many problems of
interest: Mars (moderate velocity), LEO return, Titan
 Importance of Radiative Cooling 26

•The shock layer is cooled by the emission of photons. Clearly this effect will
become more important as a larger fraction of the total shock layer energy is
converted to photons
•Tabular or engineering expressions for stagnation point radiation typically
include the radiative cooling effect
•However it is very important to recognize this phenomenon when computing
radiation from CFD data (inherently uncoupled operation)
•Goulard proposed a non-dimensional parameter that is essentially the ratio of
total energy flux to that lost to radiation:
2q R, unc
 1
2
V 3
•The net radiative heating can then be computed from (Tauber-Wakefield):
q R, unc
q R, coup 
 1   0.7 
•Where  is an atmosphere-specific constant
•  = 2 for Titan
 = 3.45 for Earth
 ~ 3 for Mars/Venus

 Example - Galileo Probe 27

Radiative (no coupling)

Radiative (including coupling)

Convective (no blowing)

Convective (blowing)

Adapted from Anderson, Hypersonic and High Temperature Gas Dynamics, Fig. 18.16
 Wall Temperature Estimation 28

•How hot does the TPS surface get?

•A body radiates heat at a rate proportional to the 4th power of its temperature
•Stefan-Boltzmann Law: q rerad  T
4

• where  is the emissivity of the TPS ( = 1 for a blackbody),  is the Stefan-Boltzmann

constant ( = 5.67e-8 W/m2/K4), and T is the wall temperature (assumes the ambient
temperature is much lower)

•The wall heat flux balance is in general given by the sum of heat into the material
minus reradiation, conduction, and material response. A primary function of TPS
is to minimize conduction (good insulator), and thus, neglecting material
response we can assume that:
q rerad ~ q conv  q R
which can readily be solved for Tw.
•Examples:
 (Tw = 1600 K)
• Orbiter peak heating
• MER peak heating (Tw = 1725 K)
• Orion peak heating (Tw = 3360 K)
– by this point we are overpredicting by ~20% due to material response effects
 Example: Shuttle Orbiter 29

•For the Shuttle-Like entry previously studied, what is the stagnation point
heating rate and the wall temperature at 60 km altitude? Assume a 1m nose
radius and a TPS emissivity of 0.8
–  = 3.1459e-4 kg/m3
– V = 3.535 km/s
– qw = 1.7415e-4*(3.1459e-4/1)0.5*(3535)3 = 13.6 W/cm2 (Sutton-Graves)
– qR = 0 (Tauber-Sutton)
– Tw = [(13.6*1e4)/(0.8*5.67e-8)]0.25 = 1316 K
 30

Further Reading
 31
Engineering Methods
• Tauber, M., “A Review of High-Speed, Convective Heat Transfer Computation Methods,” NASA
TP-2914, Jul. 1989.
• Tauber, M., Bowles, J., and Yang, L., “Use of Atmospheric Braking During Mars Missions,” Journal
of Spacecraft and Rockets, Sept.-Oct. 1990, pp. 514-521.
• Tauber, M., Yang, L. and Paterson, J., “Flat Surface Heat-Transfer Correlations for Martian Entry,”
Journal of Spacecraft and Rockets, March-April 1993, pp.164-169.
• Compton, D. L. and Cooper, D. M., “Free-Flight Measurements of Stagnation Point Convective Heat
Transfer at Velocities to 41,000 ft/sec,” NASA TN D-2871, Jun. 1965.
• Marvin, J. G. and Deiwert, G. S., ”Convective Heat Transfer in Planetary Atmospheres,” NASA TR
R-224, Jul. 1965.
• Kaattari, G. E., “Effects of Mass Addition on Blunt Body Boundary Layer Transition and Heat
Transfer”, NASA TP-1139, 1978.
• Tauber, M. E. and Sutton, K., “Stagnation Point Radiative Heating Relations for Earth and Mars
Entries”, Journal of Spacecraft and Rockets, Jan.-Feb. 1991, pp. 40-42.
• Page, W. A. and Woodward, H. T., “Radiative and Convective Heating during Venus Entry”, AIAA
Journal, Oct. 1972, pp.1379-1381.
• Tauber, M. E., “Some Simple Scaling Relations for Heating of Ballistic Entry Bodies”, Journal of
Spacecraft and Rockets, July 1970, pp. 885-886.
• Chapman, G.T., “Theoretical Laminar Convective Heat Transfer & Boundary Layer Characteristics
on Cones at Speeds to 24 km/s,” NASA TN D-2463, 1964
• Sutton, K. and Graves, R.A., “A General Stagnation Point Convective Heating Equation for Arbitrary
Gas Mixtures,” NASA TR- R-376, 1971
• Fay, J.A, and Riddell, F.R, “Theory of Stagnation Point Heat Transfer in Dissociated Air,” J.
Aeronautical Sciences, 25, 1958, pp. 73-85,121.
 32

Lecture #2: Distributed Heating and

Trajectory Effects
 33
Distributed Heating - Sphere
• It can be shown that the heat transfer rate along the body
varies according to
q
 cos 
q stag

for angles as large as 45° (in theory) and 70° (in practice)
• This expression permits us to integrate the total heat flux

into a spherical nose as
 qdA  q stag  cos dA
dA  2yRn d  2Rn2 sind
 /2
 qdA  2Rn2q stag  sin cos d  Rn2q stag
 0

• For a laminar boundary layer, the heat input to a hemisphere is ~ equal

to the product of stag. point heating times the projected area
 34
Distributed Heating - Sphere (2)
 35
Local Similarity - Flat Faced Cylinder
• Local similarity methods (see e.g. Anderson) can be
extended to other geometries
• Take for example a flat-faced cylinder with a rounded
corner
• For this case, local similarity theory (and more
sophisticated methods) show that the stagnation point is
not the highest heating location; rather heating is higher
on the corner
– Physically, the large favorable pressure gradient causes the boundary
layer to thin. This increases the magnitude of h, which increases heat
transfer per previous arguments. The magnitude of increase is inversely
related to the radius of curvature.
 36
Distributed Heating - FF Cylinder (2)
 37
Distributed Heating - Approximate Methods
• Many other approximate methods have been developed for
the calculation of heating on other geometries, e.g. wings,
attachment lines.
• Detailed assessment is beyond the scope of these
lectures, but the interested student can read further in:

Tauber, M.E., “A Review of High Speed Convective Heat Transfer

Computation Methods,” NASA TP 2914, 1989

which is included as a handout for this course.

 Real World Examples - 38

Laminar Flow

MSL Shape in T5 Predicted Stardust Heating

 39

Trajectory Effects
• The discussion up to now has focused on the calculation
of an instantaneous heat flux (primarily at the stagnation
point).
• However, the heating on the vehicle is obviously coupled
to the trajectory flown, and thus it is important to develop
expressions that quantify the relationship between heating
and trajectory.
• You have already learned two basic trajectory equations
(Allen-Eggers and Equilibrium Glide); lets start with Allen-
Eggers
• For simplicity, lets use the simplest of convective heating
relationships: 1
q s ~  2 V 3
 40

Intuition (1)
• Two identical ballistic vehicles enter the atmosphere. One
is on a steep entry trajectory and one is on a shallow entry
trajectory. Which has the higher peak heat flux? Load?

shallow 

steep 
 41

Intuition (2)
• Two ballistic vehicles enter the atmosphere on an identical
flight path angle. One has a higher ballistic coefficient.
Which has the higher peak heat flux? Load?

high 

low 
 42

Allan-Eggers Trajectory Equation

V  Vatm expCe  h / H   Vatm exp C o 

Vatm = Velocity at atmospheric interface 0
C
 = m/CDA 2sin
Exponential atmosphere assumed
Ballistic entry

• Substitute above for V into approximate heating equation:

  
1
q s ~   Vatm
23
exp 3C o

• Differentiate w.r.t density:

 3C   3
  V    dq s
 12
 
1
1 3
exp 3C 
o       Vatm exp 3C  o
2 

  o   d
2 atm
 43

Allan-Eggers Trajectory Equation (2)

• Looking for a maximum of qs, which should occur when dqs/q = 0:


1  6C 0
o
• So the density of maximum convective heating is:

o  sin
   
*
q max
6C 3H

• For a given atmospheric scale height, the density (altitude) of peak

heating increases with ballistic coefficient and flight path angle

 44

Allan-Eggers Trajectory Equation (3)

• So, in the exponential atmospheric model

 sin
  oe  h*/ H
3H
h*   sin  
  ln  
H  3H o  

• The altitude and velocity of peak heating are given by:

  sin  
 h *
 H ln  
 3H o  
q max

 C    o  
V*
 Vatm exp      Vatme 1 / 6  0.846Vatm

q max
 o  6C  

 45

Allan-Eggers Trajectory Equation (4)

• As in the case of the previously derived expression for the velocity at
peak deceleration, the velocity at peak heating is a function only of the
entry velocity.
• Recall that Vgmax = 0.606Vatm. Therefore, peak heating occurs earlier in
the entry than peak deceleration. In fact, it can be shown that
hq* max  1.1h*g max

• We are now in the position of being able to calculate the peak

stagnation point convective heat rate for a ballistic entry vehicle
• Substitute the 
evaluated expressions for V and  qmaxinto the qmax
Sutton-Graves Equation:
1
  1     sin  2
1
2

q s, max  k        .6055Vatm
3

 Rn     3H  

• In addition to the nose radius dependence shown earlier, we now see

that peak heating rate increases with increasing ballistic coefficient
and flightpath angle
 46

Heat Load
• Stagnation point heat load is just the time integration of the heat flux

k 1
Qs   2
V 3dt
Rn

• How do we convert this to an integral that we now how to evaluate

(redefine dt through change of variables)? Lets borrow some logic
 of Motion:
from the Equations
dh ds
sin   ; V
ds dt

ds dh
dt  
 V
 V sin

• Using the exponential atmosphere model we can write this in terms of

d 
 47

Heat Load (2)

• Exponential atmosphere model

   oe  h / H

d o h / H 
• Differentiate:  e 
 dh H H

Hd
 dt:
• Substitute into dt 
V sin

• Now we can substitute into the heat load integral:

 2
k VatmH  o  12  2C
 
Qs   q sdt    exp  d
Rn sin 0   0  
1

2
    2
Qs ~ kVatm   After some manipulation…
R
 n sin 

 48

Heat Rate vs Heat Load

• Quantitative expression can be derived from approximate evaluation
of the integral: 1
    H  2
k is the Sutton-Graves
Qs  kVatm
2
  
 
 
 o n R
 sin  
constant

• Compare the derived expressions for heat rate and heat load:
 1 1
  1     sin  
2
1
     H  2

  .6055Vatm
2
q s, max  k      3
Qs  kVatm
2
   
 Rn     3H    o  Rn sin  

• Heat rate increases with both  and , while heat load increases with ,
but decreases with 
 
• This leads to a second mission design trade (the first was Rn and its
impact on drag, convective heating, and radiative heating):
• The selection of  becomes a trade between peak heat rate (TPS
material selection), and total heat load (TPS thickness and mass)
 49

So, Did you get it Right? (1)

• Two identical ballistic vehicles enter the atmosphere. One
is on a steep entry trajectory and one is on a shallow entry
trajectory. Which has the higher peak heat flux? Load?

shallow  Higher Peak

Heat Load
steep 
Higher Peak
Heat Flux
 50

So, Did you get it Right? (2)

• Two ballistic vehicles enter the atmosphere on an identical
flight path angle. One has a higher ballistic coefficient.
Which has the higher peak heat flux? Load?

Higher Peak
high  Heat Load

Higher Peak
Heat Flux
low 
 51
Mars Entry Heating - Example
Entry Flight Path Variation
 = 90 kg/m2; Vi = 5.5 km/s

Heat rate falls and heat load grows as FPA decreases

 52
Mars Entry Heating - Example
Ballistic Coefficient Variation
 = -12 deg; Vi = 5.5 km/s

Rising ballistic coefficient raises heat rate and load

 53

Equilibrium Glide Entry

• Can perform the same analysis of an equilibrium glide (lifting) entry
• Details are left as an exercise for the student

Vq*max  2
3 Vc (for Vatm  2
3 Vc)
1
  1      2

q s max  1.94  10      

4

R  n  L / D  
 1 2 
1
      Vatm   2  
 Vatm    
7   L     1 Vatm
2

Qs  2.05  10       sin      1       

R n  D       Vc   Vc     Vc      
  
• Compare to Allen-Eggers; similar dependence on , but a lifting body
(L/D > 1) will have heat flux inversely dependent on L/D and heat load
directly dependent on L/D
 54

Numerical Example: MER

• What is the peak stagnation point heating for the MER example
previously examined (Rn = 0.5Rb)?
• At peak heating:
Vqmax = 0.846*5.45 = 4.61 km/s
Rn = 2.65/4 = 0.6625 m
h = 40.87 km
 = 3.11e-04 kg/m3

• From the Allen-Eggers expressions derived herein:

1 1
    3
2 
4 3.11  10
 4 2

 
3
q s  k    V  1.9027  10    4610  40.4 W/cm2
 Rn     0.6625  
1
 q w     40.4  10 4  
4
Tw          1727K
8  
 
    0.8  5.67  10  

(literature quoted values range from 40-44 W/cm2 based on CFD)

 55

Other Trajectory Effects

• Prior discussion focused on impact of trajectory on stagnation
point heating

• However, trajectory selection has other aerothermal impacts as

well

• Transition to turbulence
- Can dramatically increase heating levels away from stagnation point (4-
6 times laminar levels)
- Governed by Reynolds number (uL/m), therefore exacerbated by large
entry bodies, steeper flight path angle, higher entry velocity, higher
ballistic coefficient

• Heat soak
- Longer trajectory time increases the amount soak of energy into the
TPS, which increases the amount of TPS required to protect the
structure (a given TPS tends to be less efficient as peak heat flux drops
but heat load stays constant)
 56

Orbiter Thermal Imagery

Turbulent flow from

wing BLT protuberance
Stagnation Point
(Laminar)

Turbulent flow from

unknown origin
STS-119
Mach ~ 8.5
Mar 28, 2009
 57
Review: 20 minutes of Peter Parameter
 58

Lecture #3: Advanced Topics

 59
CFD Process for Entry Vehicle Design
 High fidelity CFD tools based on 20-year old Shuttle RCC Repair
Concept Evaluation
methodologies
 Recent advances in parallel computing, efficient implicit
algorithms have enabled rapid turnaround capability for
complex geometries
 Full body three-dimensional CFD is an integral part of
the design of all planetary and Earth entry TPS

Genesis Penetration Analysis

Arc Jet Model Simulation

Test model

Inlet
conditions

Nozzle flow CFD Model CFD Arc jet test

simulation simulation
 Identification of Aerothermal Modeling Needs 60

for Entry Missions

 Needs are both physics and process driven

• process improvements are important for modeling complex geometries -
not covered in this presentation
• physical model improvements are important across the spectrum of NASA
missions

 Gaps are destination and mission specific

• shock layer radiation in particular will dominate aeroheating for some
missions and be unimportant for others
• sensitivity analysis must be performed for each candidate mission

 Gaps can be divided into general categories

• reacting gas physical models
• surface kinetics
• transition and turbulence
• afterbody heating
• shock layer radiation modeling
• coupling between radiation/material response/fluid dynamics/aerodynamics
• unsteady separated flows (wakes, control surface shock-BL interaction)
• geometry effects
 61
Turbulence: The Eyeball…

70o Sphere-Cone:
Hypersonic Flight in Ballistic Range T (K)
Lower P∞ Higher P∞
 62

Orbiter Thermal Imagery

Turbulent flow from

wing BLT protuberance
Stagnation Point
(Laminar)

Turbulent flow from

unknown origin
STS-119
Mach ~ 8.5
Mar 28, 2009
 Transition and Turbulence 63

Status and Remaining Gaps

70o Sphere-Cone:
Hypersonic Flight in Ballistic Range
T (K)
Lower P∞ Higher P∞

 Transition is less of a concern for blunt

capsules
• shorter trajectories, smaller surface area leads to
less heat load augmentation
• single use ablative TPS can withstand heating if
mass penalty not large – design to fully turbulent
 Conclusion: Transition cannot be accurately Transition Front
predicted for most problems of interest. Designs
must rely on testing and conservatism. Mars Science Laboratory
Peak Heating Condition
Laminar Turbulent
 Acreage turbulent heating predictions
generally within 25% for orbital Earth
entries (RANS), but additional
developments are required for chemistry,
blowing, roughness
V
 DNS, LES, DES type models under
development to replace current RANS
Stagnation Point
 Turbulent Heating: 64

Effects of Surface Roughness

 Previous discussion centered on smooth Effects of Roughness on Heating

wall turbulence Smooth Wall Rough Wall
Heating Heating
 However, all ablators develop a roughness
pattern that can augment heating
 Analysis for MSL based on correlations from
WT experiments and DoD RV data
– 1mm roughness  potential for up to 50%
augmentation to baseline smooth wall predictions
– if true, roughness has eaten up entire turbulent
heating uncertainty!
 Roughness can also lead to a positive
feedback loop  vortical structures are from Brown, ARC
generated that augment roughness
 65
Surface Catalysis

 No validated model exists for Mars: Impact of Catalysis Model on Heating

CO + O; O + O; CO + O2
 As a consequence, Mars entry vehicles are
designed assuming a worst case scenario
– so called “supercatalytic” wall
 For MSL there is a factor of four difference
in heating between the various models

Parametric Analysis of Catalytic Heating

Centerline Heating -
Fully Turbulent

 What are the key gaps?

– quantum chemistry to determine reaction rates
(gas phase and gas-surface)
– MD simulations of key GSI processes
– experimental data on TPS materials at relevant
conditions

from Bose, ARC

 66

Afterbody Heating
 Wake flows are much harder to simulate MSL Afterbody Heating
than forebody
– separated, low density, unsteady, nonequilibrium
flowfield
– significant code-to-code differences still exist

 Current uncertainty levels ~50-300%

– primary reason: lack of validation; we have not
quantified how good (or bad) we are

CFD Validation with

AS-202 Flight Data from Edquist, LaRC

 What are the key gaps?

– additional ground test data (including free
flight or stingless models)
– explore advanced methods (DES, LES) for
hypersonic separated flows
– advocate for additional flight data
 67
Singularity Heating

 Now throw OML singularities (such as RCS Thruster Impingement

Heating (No Margin)
RCS thrusters) into the wake flow
– does not make things easier!
 MSL is actively guided; thrusters must fire
during hypersonic entry
– predicted locally high heating rates necessitated
a late change in backshell TPS for MSL (with
significant cost and mass penalty)

MSL RCS
Thruster Design from Dyakonov, LaRC
(Preliminary)

 MSL backshell design requires canted

thrusters for control authority
 Thrusters sticking into the flow; must be
designed to withstand aerothermal
environment
– no validation of our methods for this application
JPL
 68
Shock Layer Radiation
Shock layer radiation is highly non- CN Radiation Model Validation
equilibrium, non-blackbody
– Titan analysis showed order of magnitude
differences between equilibrium&accurate model
Not important for Mars missions to
date, but critical for HMMES
– importance increases with velocity & vehicle size
– primary radiator, CO(4+) emits in UV

EAST Test Data

 What are the key gaps?

– obtain additional shock tube data for Mars entries
– build collisional-radiative models for all atomic and
molecular radiators
– compute excitation rates from QM
– develop medium-fidelity methods for design
– develop models for coupling to fluid dynamics

from Bose, ARC

 69

Flowfield-Radiation-Ablation Coupling

Titan Radiation Coupling

Flowfield-Radiation (adiabatic cooling)

• Engineering approximation
  2q rad /( 12 V3 ) qcoup / qunc  1/(1 0.7 )
• Loose coupling is also possible
• More accurate answer requires simultaneous
solution of the Navier-Stokes and radiative transfer
 equations; not possible except for limiting cases

Flowfield-Ablation
• Blowing reduces heat transfer
• Ablation products mix with boundary layer gases
Stardust Ablation Coupling • Typically solved via loose-coupling approximation
Radiation-Ablation
• Injected ablation products can absorb/emit radiation
Ablation-Trajectory
• Significant ablation can lead to changes in
aerodynamics/trajectory/GN&C
• Primarily a concern for RV’s
 70

TPS – Boundary Layer Interaction

We have already discussed gas-surface Pattern Roughness
on RV Nosetip
and ablation coupling, but other
interactions are important
Ablation induced distributed roughness
• Surface roughness generated on TPS surface as a
consequence of ablation.
• Strong interaction with boundary layer - increased
heating and shear stress result
• Heating augmentation from zero to factor of three
possible over turbulent smooth wall
Protruding Gap Filler in
Discrete roughness Arc Jet Test
• Due to gaps, repairs, geometrics singularities, etc.
• Generate local heating and shear augmentation
factors which must be accommodated

For MSL:
• Distributed roughness adds about 20% to heating
(pattern roughness not expected)
• Discrete roughness adds another 40% locally in
areas of gaps or repairs)
 71

TPS – Boundary Layer Interaction

Melt layer interactions Melt Flow induced by stream wise vortices
• One class of ablators uses a
glassy substrate material
• Energetically favorable; glass
vaporization is highly
endothermic
• Can cause strongly coupled Flow
instabilities in environments
where glass melts but does
not vaporize
• Interactions or instabilities
can range from minor to
catastrophic

What to do?
• Simple solution: don’t fly Research topic: Better models for all
glassy ablators in such
environments
aspects of material / fluid interactions
• Better long term solution:
develop models of the
boundary layer surface
interaction
 72
Shape Optimization
 The primary reason we continue to use 70° Sphere Cone vs. Ellipsoidal Aeroshell
70° sphere cones for Mars entry is
“heritage”
– argument is weak: clear finding of MSL
aerothermal peer review last summer
 Non-optimal from aerothermal perspective
– expansion around nose leads to boundary layer
instabilities, early transition, high heating levels
 Modified ellipsoid aeroshell has significant
advantages with same aerodynamics
– for Mars aerocapture this shape led to 50% lower
heat flux, potential 67% TPS mass savings
HEDS Ellipsled from Brown, ARC

 For large entry masses other shapes

(e.g. ellipsled, biconic, bent biconic)
should be explored as well
 A full shape optimization study should be
part of any future Mars systems analysis
 Validation: 73

AS-202 Flight Data

Afterbody Calorimeter Placement Surface Oilflow
• Problem: Current t= 4900 s,ReD = 7.6105
uncertainty on afterbody
heating predictions is
very high
• Goal: reduce uncertainty
levels by validation with
flight data

“c” Computations
generally agree with
flight data to within
±20% uncertainty at 15
of 19 calorimeter
locations.

Ref: AIAA 2004-2456

 74
Flight Data: MER-B Heatshield

 Unique opportunity to observe in-situ

flight hardware during Opportunity
extended mission
 Multiple images of (inverted) heatshield
made with cameras and micro-imager
 Work ongoing to
compare visualized
material response to
predictions

Image courtesy Christine Szalai, JPL

Flight data are the gold standard for

final model validation
 75
Flight Data: MEDLI
 HQ approval for MSL instrumentation suite!
MSL Heatshield Layout
 High TRL sensors to be installed in seven
locations on heatshield
 Flight data obtained will go a long way toward
validating ARMD-developed tools to drive
down uncertainties discussed herein
 No backshell instrumentation (backshell is on
critical path)

Recession
Sensor
Wound resistive wire

Outer kapton layer (tube or coating)

Hollow kapton tube

Thermocouple
Plug Pressure Sensor

ARC Sensor Lab

 76

Question and Answer Period

 77

Thermal
Protection Systems

John A. Dec
NASA Langley Research Center
 78

Outline
 Background Information
– What is TPS?
– Selecting the Right Material for the Mission
 Ablative TPS Modeling
– Ablator Characteristics
– Surface Recession
– In-Depth Models
 TPS Sizing and Margin
 TPS Testing
 Look to the Future
 Three Kinds of TPS 79

Passive (Reusable)
Rely on reradiation to reject heat, low thermal conductivity to limit penetration
Coatings to increase emissivity, reduce catalycity
Limited by reusable temperatures of common materials
Uses: Shuttle Orbiter, X33, X34

Active (Reusable)
Rely on active cooling for heat rejection
Plumbing systems, active transpiration
Very complex; seldom considered; very low technology readiness

Ablative (Non-Reusable)
Combine reradiation with ablation and pyrolysis for heat rejection
Can be considered passive transpiration cooling
Ideal for high heat flux/load entries, particularly when reusability not required

The focus of today’s lecture is on ablative systems; baseline for all planetary EDL to date
 80

Ablation
• Definition:
– The term ablation is encountered in many
fields of science and engineering
• In the medical field it refers to the surgical removal
of a body part or tissue
• In glaciology it refers to the removal of ice and
snow from the surface of a glacier

– In space physics, ablation is the process of

absorbing energy by removal of surface
material by melting, vaporization, sublimation,
or chemical reaction
 81
Why Ablative Materials?

Courtesy Bernie Laub, NASA Ames

 82
How is TPS Chosen?

• Heat rate, along with pressure and shear, determine

type of TPS to employ
• Material classes have clear performance limits marked by poor
performance/material failure

• Heat load determines overall thickness of TPS

material

• Other design features play a role

• Need for tiles, forebody penetrations, compression pads,
structural loads, etc. can impact material selection and TPS
design
• RF transparency for materials that protect antennae
 83
Ablative TPS:
History of Success, Little Recent Development

Ablative TPS Chronology (forebody)

100000
Galileo (Jupiter)
Peak Heat Flux (W/cm )
2

Pioneer Venus FM 5055 Carbon Phenolic

10000
FM 5055 Carbon Phenolic

Stardust
CEV
Avcoat
1000 Apollo PICA
Genesis
AVCOAT 5026-39/HC-G C-C dual layer

Mars Pathfinder
100 Gemini MSL
SLA-561V MER
PICA
DC-325 Mars Viking SLA-561V

SLA-561V
10
1960 1970 1980 1990 2000 2010
Year
No Human Rated Ablative TPS Available Today!
CEV/Orion is working to develop Avcoat, for a human rated
system - Very Close to Achieving This Goal!
Courtesy Bernie Laub, NASA Ames
 84

Available Materials
Density Limit
Material Name Manufacturer
(kg/m3) (W/cm2)
Not viable for high
SLA-561V Lockheed-Martin 256 ~ 200
shear

FM 5055 Carbon Fibercote (formerly US No source of

1450 > 10,000
Phenolic Polymeric), Hitco Inc. heritage Rayon

Flown on Shuttle
MX4926N Carbon Cytec (pre-preg), ATK, SRM, never as a
1450 > 10,000
Phenolic HITCO heat shield

Applied Research
PhenCarb-20,24,32 320-512 ~ 750 Never flown
Associates (ARA)

PICA (Phenolic
Must be tiled above
Impregnated Carbon Fiber Materials, Inc. (FMI) 265 > 1500
1m diameter
Ablator)

Avcoat 5026 (Apollo) Textron Systems 513 ~1000 Recreated for CEV

Heavy, not readily

ACC Lockheed-Martin 1890 ~ 1500 extendible above 2m
 85
TPS Mass Fraction Requirements

CEV Current MEL Wow! This is going to

be hard w/o a
significant
MSL Apollo Design improvement to the
state of the art

Courtesy Bernie Laub, NASA Ames

 86

What Are They Made Of?

Large Cell Phenolic Honeycomb

Organic Fiber Reinforced Phenolic

 87

Pyrolyzing Ablators
Substrate Material (e.g. fibers, cloth)
• Desire ability to withstand high temperatures (reradiation)
• Carbon is best; glass also good (heat of vaporization)
Organic Resins (e.g. phenolics)
• Pyrolyzing ablators only
• When heated resin generates gas and leaves carbon residue
• What are they good for?
– in-depth and surface transpiration
– endothermic reactions absorb energy
– carbon char for reradiation
Additives (e.g. microballoons, cork)
• Density & thermal conductivity control
Added Reinforcement (e.g. honeycomb)
• Structural integrity, bond verification (adds mass)
 88

How do they Work?

 89

Surface Ablation Mechanisms

• Melting – common ablation mechanism, but doesn’t
absorb much energy

• Vaporization – absorbs significant amount of energy

• Oxidation – exothermic process that adds energy

• Sublimation – Can be significant energy absorber

• Spallation – Mass loss with minimal energy

absorption (Thermostructural Failure – HIGHLY
UNDESIRABLE)
 90

Oxidation

Other exothermic surface chemistry is possible (“nitridation”

and “hydridation”) but these are not typically significant players
Courtesy Bernie Laub, NASA Ames
 91

Other Mechanisms

• Material decomposition …aka pyrolysis

– Endothermic reactions absorb energy
– Convection of pyrolysis gas through the char
• Conduction through the material
– Transfer energy to structure or heat sink
• Re-radiation from the surface
– Largest percentage of energy is dissipated
through this mechanism
 92

Modeling Approach
• In the mid to late 1960’s, Kendall, Rindal, and Bartlett, and Moyer and Rindal
extended the work by Kratsch et. al.
– Included unequal heat and mass transfer coefficients
– Non-unity Lewis and Prandtl numbers
– Corrected in-depth energy equation:
• to account for the energy of the pyrolysis gas convection and generation within
the solid
• to account for grid motion due to a coordinate system that is attached to the
receding surface

T   T   T hg
cp     hg  h   S cp  mg (12)
k
t xS  xS  t x xS xS
 
 eU eCH  H sr  hsw   eU eCM    Zie*  Z iw* hi0  Bchc  Bg hg  Bhw   q*  qrad   qrad
dT
k (13)
dx  i  out in

• If the diffusion coefficients are assumed equal and the Le=Pr=1.0, the surface
energy balance simplifies to

 eU eCH  H sr  hsw  Bchc  Bg hg  Bhw   q*  qrad   qrad

dT
k (14)
dx out in
 93

Example: SLA-561V Failure

 94

Derivation of the Governing

• Differential
1-Dimensional controlEquations
volume

 u2   u2 
 g g
m h  m g  m g gZ  mcv  g g
m h  m  m gZ 
 x
g g
2  2  x dx

mg x Est mg x  dx
qx qx dx
 95

Conservation of Mass
• Pyrolysis gas flows from the pyrolysis zone
through the porous char to the heated surface
– Assume gas flow is 1-D and normal to the  u2 
 mg hg  mg 2  mg gZ 
 x
mcv
 u2 
 mg hg  mg 2  mg gZ 
  x dx

heated surface mg x Est mg x  dx

– Assume dp~0 across the char (neglect the qx qx dx

momentum eqn)
mcv
 m in  m out (15)
t
mg x
Where mcv   Adx (16) mg x  dx  mg x  dx (17)
x
  m g x   mg x
A dx   m g x  dx   m g x 
t x
(18)
  t x

 Determined experimentally and modeled mg x = Mass flow rate per unit area
t with an Arrhenius fit
 96

Conservation of Energy
• Two energies associated with this
 u2   u2 
control volume  mg hg  mg 2  mg gZ 
 x
mcv  mg hg  mg 2  mg gZ 
  x dx

– Pyrolysis gas flow mg x Est mg x dx

qx
– Heat conduction. qx dx

• Pyrolysis gas flow assumptions

• Pyrolysis gas is in thermal equilibrium with the charred material
within the control volume
• Pyrolysis gas residence time within the control volume is small.
• Potential energy of the pyrolysis gas may be neglected since
the change in height across the control volume is negligible.
• The kinetic energy of the pyrolysis gas may be neglected since
it is of small magnitude relative to its enthalpy
 97

Conservation of Energy
 u2   u2 
 mg hg  mg 2  mg gZ  mcv  mg hg  mg 2  mg gZ 

• 1st Law of Thermodynamics

 x   x dx

mg x Est mg x dx
qx qx dx
 Q cv  W cv  m in e  Pv in  m out e  Pv out 
dEcv
(19)
dt
Where e is the total energy per unit mass and
includes kinetic, potential, and internal energy
The internal energy and flow work may be
expressed in terms of the enthalpy by, h  u  Pv

Rewriting equation 19 in a simplified form gives,

dEcv
 E in  E out (20)
dt
 98

Conservation of Energy
• The energy entering and leaving the  u2 
 mg hg  mg 2  mg gZ 
 x
mcv
 u2 
 mg hg  mg 2  mg gZ 
  x dx

control volume can be expressed as mg x Est mg x dx

qx qx dx

E in  q x  m g hg xdx
E out  q xdx  m g hg x
(21)

• Expressing the incremental heat conduction leaving and the

convection of energy by the pyrolysis gas entering the
control volume as Taylor series expansions gives, dropping
H.O.T qx
qx  dx  qx  dx
x
(22)

m g hg x  dx  m g hg x  m g hg x dx
x
 99

Conservation of Energy
• The rate of energy storage within the
 u2   u2 
 mg hg  mg 2  mg gZ  mcv  mg hg  mg 2  mg gZ 
 x   x dx

mg x Est mg x dx
control volume can be expressed in terms qx qx dx
of the density and enthalpy of the solid as
dEcv 
 h Adx (23)
dt t

• Substituting eq 21 into eq 20, and using the

definitions in eqns 22 and 23 gives
  qx
hAdx  qx  m g hg x  m g hg x dx  qx  dx  m g hg x 
  
(24)
t  x   x 
 100

Conservation of Energy
• Canceling like terms, dividing by Adx, and using
Fourier’s law of heat conduction eqn 24 reduces to,

  T 
  h    k x   mgx hg
  (25) • Physically,
t x  x  x
I II III – Term I represents
Where, energy storage
 : density of the solid – Term II represents
h : enthalpy of the solid conduction through the
hg : pyrolysis gas enthalpy material
kx : thermal conductivity in the x-direction
– Term III represents
T : temperature
mg x : local gas flow rate in the x-direction
convection due to
pyrolysis gas flow
x : coordinate direction
 101
Transforming the Governing Equations to a Moving
Coordinate System
Virgin
Original
material
surface

• The control volume is not fixed in

Material begins
S
to recede
x=0, S=0
xS=0 S Char begins to form
space, it is tied to the receding surface xS=0
xS
Char
Char

– Requires transforming eqns 18 and 25 into x

dx=dxS
Virgin

a moving coordinate system Virgin

x=S+xS
– After some elaborate calculus and
algebraic manipulation we arrive at, Backup
Backup
Backup

  
Conservation of mass in a moving S  (26)
coordinate system t xS xS t t x

   T   
Conservation of energy in a moving
t
 h 
xS
 x
k

 


mg x hg  S
xS
  h  (27)
coordinate system xS  xS t x S t t
I II III IV

Where terms I-III are the same as in eqn 25 and term IV is the
convection of energy due to coordinate system movement
 102

Final Form of the Energy Equation

• It is convenient to express the (h) terms in equation
27 in terms of material properties rather than the
thermodynamic quantity of enthalpy
• Performing some algebra and defining a new
quantity, h , the energy equation takes the following
form
T   T   T hg
cp   kx    hg  h   S cp  mg x
t xS  xS  t x xS xS (28)
I III IV V
II

where T
H v  h   c pv dT
0 v : virgin material density
  H  c H c 
v
c : charred material density
h  v v
0

 v  c 
T Hv : total enthalpy of the virgin material
H c  h   c pc dT
0
c Hc : total enthalpy of the charred material
0
 103

Final Form of the Energy Equation

T   T   T hg
cp   kx   g
 h  h  t  S  c  m
t xS  xS xS xS
p gx (28)
 x
I III IV V
II

• Each term in equation 28 has physical significance

– Term I
• rate of sensible energy storage
– Term II
• net conduction through the material
– Term III
• creation of sensible energy due to pyrolysis (ie the heat of
decomposition)
– Term IV
• energy convected due to coordinate system movement
– Term V
• energy convected away due to pyrolysis gas generation at that point
 TPS Sizing Approach 104

• Baseline (zero-margin) sizing computed assuming nominal environments

and response model to hit given bondline temperature limit
• Margin process then applied to account for various sources of uncertainty
• Appropriate factors of safety be applied to trajectory dispersions,
aerothermal loads, initial conditions, and material variabilities
• Primary (thermal) margin is applied directly to the TPS design criterion
(e.g. maximum bondline temperature)
– The impact of this margin on TPS thickness is material-dependent since the sensitivity
of bondline temperature to thickness is material-dependent

• Secondary (recession) margin is also employed

– Bondline is insensitive to excessive recession until it is too late

• Various independent sources of error are RSS’ed to avoid stacked

conservatism
• Additional program imposed thickness factor of safety is recommended
to account for unknown unknowns
• Other factors (e.g. thermal stress, CTE mismatch, adhesive failure) should
also be tracked as possible limiting cases
– Adhesive failure accounted for by maintaining conservative bondline temperature limit
 105

Simplified Approach
• Return to the simplest form of the in-depth energy equation

T   T 
cp  k  (1)
t x  x 

• Here we neglect the affects of decomposition, pyrolysis gas flow and

surface recession.

• Additionally, if it is assumed that the solid extends to infinity in all but

one direction and is characterized by a single identifiable surface, if
a sudden change in conditions is imposed at this surface, transient,
one-dimensional conduction will occur within the solid. This is
known as the semi-infinite solid approximation

• This approach is for illustrative purposes only and should

not be used beyond conceptual design
 106

Simplified Approach
• Using the semi-infinite solid approximation, closed-form
analytical solutions to the in-depth energy equation can
be derived.
• For a thick slab which has a constant surface
temperature at any instant in time, the temperature at a
depth x within the solid at time t is given by,
 x 
T ( x, t )  erf   Ti  Ts   Ts (29)
 2 t 
where,
Ti is the initial temperature k  thermal conductivity
Ts is the surface temperature c p  specific heat
erf is the gaussian error function   density
k
 is the thermal diffusivity 
cp
 107

Simplified Approach

• In this simplified approach, the amount of

material required for insulation and the amount
of material required for recession are
calculated separately
• To calculate the recession in an approximate
way, use the data correlation parameter known
as the heat of ablation (Q*) and solve for
recession rate
 H r  H air
Tw

qcw     Tw
4

s  Hr  (5)
 Q*
 108

Simple Finite Difference Approach

• To increase the fidelity, a finite difference

approximation of equation 1 can be written
incorporating a simplified surface energy
balance
T   T  qconv  qrad  qcond   Tw4  0
cp  k  (1) (30)
t x  x 

 2k Dt  n 1 2k Dt 2Dt 2Dt 2Dt

 1  
2 
T  T n 1
  T n
  q  q   T n4
(31a)
  c p Dxi   c p Dxi 2  c p Dxi  c p Dxi  c p Dxi
i i 1 i rad conv i

kDt  2kDt  n1 kDt

Ti  
n
T n 1
 1  T  Ti n11 (31b)
c p Dxi 2 i 1  c p Dxi 
2 i
c p Dxi 2

 109

Stagnation Point Sizing Example

• Ballistic Earth entry
– Ballistic coefficient = 60 kg/m2 , entry velocity = 12.6 km/s
– 60° sphere cone, 0.8 m diameter, rn= 0.23 m
V2
– At the stagnation point, Hr can be approximated by 2

Stagnation Point Sizing Example

• PICA heat shield
900 4000.00
kg
  265.0
800
Total Heat Rate
3500.00 m3
Radiative equilibrium Temperature
700 W
3000.00
k  1.6
600 mK
Heat Rate (W/cm2)

Temperature (K)
2500.00
J
500
C p  1592.0
2000.00 kg  K
400

1500.00
300

1000.00
• Radiative equilibrium
200
temperature
500.00


100

q 4 
0 0.00 Tw  4  cw  Tsurr 
0.0 50.0 100.0 150.0
Time (sec)
200.0 250.0 300.0
  
 110

Stagnation Point Sizing Example

• Comparing the simplified approach, the simple
FD approach, and the high fidelity code CMA
6.0

5.5 Simplified Approach 1.2

Simple Finite Difference

High Fidelity - CMA 1.0
Nominal Thickness (cm)

5.0
0.8

Recession (cm)
0.6
4.5

0.4

4.0 0.2

0.0
Simplified Approach Simple FD High Fidelity - CMA
3.5

3.0
0 100 200 300 400 500 600
Bondline Temperature (°C)
 111
Typical Approach to Modeling Materials

• Decouple surface recession from in-depth conduction

• Use steady-state surface energy balance expression
• Employ equilibrium thermodynamic tables for surface recession,
corrected as required for finite-rate chemistry, spallation, melt flow
• Validate surface model with arc jet data

• Once recession model is working, develop in-depth

pyrolysis model
• Thermochemical data from materials testing
• Validate model by arc jet data; use first thermocouple as “truth
model” boundary condition
• Tweak char thermal conductivity as required

• Add additional physics as required for the problem

• Multi-dimensional conduction, Darcy’s Law, etc.
 112

Material Modeling
Thermochemical Properties:

1. Conduct Thermogravimetric Analysis (TGA experiments) in inert gas, low

temperature rise rates, usually 10°C/min. Residual mass fraction defines char yield.
Data fits provide decomposition kinetic constants for the Arrhenius equation.

2. Conduct digital scanning calorimeter (DSC experiments) in inert gas, low temperature
rise rates, 10°C/min. Data provides heat of reaction for pyrolysis reactions as
function of temperature.

3. Measure elemental composition of virgin material, by mass spectrometry.

4. Measure heat of combustion of virgin material and derive heat of formation.

5. Derive elemental composition of char from known constituents and char yield data.
Can be problematic to measure thermal conductivity (explained later).

6. Derive heat of formation of char from known constituents and existing data

7. Derive elemental composition of pyrolysis gases. Develop model(s) for pyrolysis gas
enthalpy using combination of thermochemical equilibrium calculations and measured
heat of pyrolysis data.
 113

Material Modeling
8. Measure specific heat of virgin material as function of temperature.

9. Measure thermal conductivity of virgin material as function of

temperature (and orientation, if appropriate).

10. Derive specific heat of char from known (or derived) composition
using method of mixtures.

11. Measure optical properties of virgin material

12. Derive optical properties of char from known composition and

properties of similar materials (or determine experimentally)

13. Measure thermal conductivity of char as function of temperature

(and orientation, if appropriate).

Assertion: the thermal conductivity of the char

cannot be measured in standard lab facilities!
 114

Char Thermal Conductivity

Traditional practice has been to bake the material in an oven and
measure the thermal properties of the resulting “char.” Studies
conducted under the Apollo heat shield program (and re-validated in
other programs) demonstrated that the cellular structure of “oven chars”
was different than the cellular structure of chars formed in ground test
or flight.
 115

Example

Courtesy Bernie Laub, NASA Ames

 116
Margin vs Nominal (Zero Margin)

• The purpose of the TPS thickness margin is to capture two things:

• Uncertainties in operating conditions
• Uncertainties in baseline (nominal) sizing required to meet operational
requirements (including abort)

• As such, the TPS margin captures implicitly the fidelity and level of
uncertainty in the underlying TPS design tools employed to determine the
baseline sizing

• Research is underway to calculate TPS margins probabilistically,

this requires knowledge of the uncertainties in the input parameters
for all analysis codes being used; aerothermal, trajectory, thermal
response
 Aeothermal Margin (RSS)
Threats and4.67
Opportunities to TPS Thickness (Mass) 117
Thermal Margin
4.5
Trajectory Dispersion
0.76
Base Thickness
Excess 4Recession Lien
3.71 3.79
0.2
Localized Application Recession Margins
0.35 0.35 0.35
3.5
Aerothermal Margins
0.17 0.17
0.2
Material Property Uncertainty
Thickness (Inches)

Arc-Jet 3Statistics 0.3Thermal Margins 0.32

Ground-to-Flight Traceability Trajectory
0.17 Dispersions 0.17

2.5 2.30
Excess Recession Lien
Bondline Temperature Limit 0.15
Initial Cold
2 Soak Temperature 0.2
Material Thermal Properties
Roughness Augmentation
3.16
Transitional
1.5 Database
Base 2.72
(Zero Margins) 2.77
Gap and Seam Design
Excess Erosion Behind Penetrations
Radiation 1 Absorption 1.95

Result from
0.5 incomplete knowledge of operating
environment, inability to test in flight environment,
and/or deficiencies
0
in underlying physical models
Courtesy Mike Wright, NASA Ames
 118

Testing

• No ground facility can reproduce all aspects of the flight

environment; every test is a compromise
• Facility classes: arc jets, combustion plasma, lasers,
radiant lamps, the atmosphere of the Earth (flight tests)
• Best facility for a given test depends upon the objectives:
• Materials screening
• Materials characterization and model development
• Performance limit evaluation (failure modes)
• Materials qualification
• Material interface evaluation (gaps, seals, etc.)
• System level testing
 119

Diagnostic Instrumentation
• Instrumentation is critical to the success of the test
• Possible Types:
• Flowfield diagnostics (calorimetry, null points, LIF)
– absolutely essential in arc jets to characterize freestream
• Surface temperature (pyrometry)
– validate recession model, detect local anomalies; global result
• Film or video
– evaluate transient performance, detect failures, recession (PRM)
• In-situ
– thermocouples, both bondline (qualification) and in-depth (material
characterization
– recession sensors
– strain gauges (system level testing)
 120

Arc Jets
• Workhorse facility for TPS testing

• Can put flight-like q,h,p, on sample for long duration (but

usually not more than two at a time)

• Limitations include:
• sample size; subscale testing only
• combined radiative/convective heating (no facility exists)
• non-Earth gas mixture (no domestic facility exists)
• difficult to simulate time-varying (trajectory based) conditions
• freestream characterization (what are we testing in anyway?)
 121
Arc Jets
 122
Sandia Solar Tower

Up to ~200 W/cm2
Concentrated solar radiation

Advantages:
Large models possible
Good for system level testing

Disadvantages:
No flow (other than wind)
Non flight like application of
heat flux (only matching one of
q,h,p,)
Only works on sunny, cloudless
days (but it is in desert!)
Courtesy Bill Congdon, ARA
 123

Sandia Solar Tower

Courtesy Bill Congdon, ARA

 124

Sandia Solar Tower

1 meter aeroshell test (ISP program)

Courtesy Bill Congdon, ARA
 125

LHMEL Laser Facility

Up to 100 kW on user-
defined spot size for up to Small Laser at LHMEL
100 seconds
CO2 radiation

Advantages:
Large models possible
High throughput
Very low uncertainty in
applied heat flux

Disadvantages:
No flow (other than wind)
Non flight like application
of heat flux (only matching
one of q,h,p,)
 126

Current Research
• Current modeling research is geared towards
making improvements
– Multi-dimensional geometry
– Orthotropic material properties
– Loose coupling to CFD codes
– Loose coupling to grid and trajectory codes
– Coupled ablator thermochemistry
– Coupled thermal stress
– Multi-dimensional pyrolysis gas flow
– Non-equilibrium surface thermochemistry
– Probabilistic heat shield sizing
 127

Current Research

Deployable/Inflatable Entry Systems

Trailing ballute concept

(Courtesy Ball Aerospace)

IRVE-3
Vehicle
 128

Flexible Thermal Protection

• Systems
Flexible TPS materials are a mission enabler for
large mechanically deployable or inflatable entry
system aeroshells

• Large aeroshell diameter reduces ballistic

coefficient and therefore peak aerodynamic
heating
 129

Thermal Protection Systems

• Flexible TPS consist of multiple layers of different
materials
• Outer reinforcing fabric
• Inner insulation
• Impermeable gas barrier
 Summary 130

• The current generation of ablative TPS models must be

significantly improved to support the next generation of
complex NASA entry missions
• Advanced modeling and new systems will be a key
component of reducing mass while increasing system
reliability
• Improvements required
– Finite-rate gas-surface interaction capability
– Loose coupling to CFD codes
– Loose coupling to grid and trajectory codes
– True multidimensional analysis, including gaps, seams and other
interfaces
– Coupled ablator thermochemistry
– Built in models for melt flow (glassy ablators), mechanical erosion, etc.
– Robust models for multi-layer ablative systems
 131

Nomenclature
 132

A area, m 2
Nomenclature
Bi pre-exponential factor for the ith resin component
Bc non-dimensional charring rate
Bg non-dimensional pyrolysis gas rate at the surface
B total non-dimensional blowing rate
CH Stanton number for heat transfer
CM Stanton number for mass transfer
Cp solid material specific heat, J/kg-K
C pg pyrolysis gas specific heat, J/kg-K
Eai activation energy for the ith resin component, J/kg-mole
Est rate of energy storage in the control volume, W
Hr recovery enthalpy, J/kg
Hw wall enthalpy, J/kg
Tw
H air enthalpy of air evaluated at the wall temperature, J/kg
Hg pyrolysis gas enthalpy, J/kg
hi0 enthalpy of formation of species i, J/kg
href reference enthalpy at 298K, J/kg
hg enthalpy of pyrolysis gas, J/kg
hc enthalpy of char, J/kg
hw enthalpy of the boundary layer edge gas evaluated at the wall temperature, J/kg
 133

i
Nomenclature
node index, or resin component index (A,B,C)
k thermal conductivity, W/m-K
mg mass flow rate of pyrolysis gas, kg/s
mg mass flux of pyrolysis gas, kg/m 2 -s
mc mass flux of char, kg/m 2 -s
mcv mass stored in the control volume, kg
q source term in the general heat equation
q* condensed phase energy removal, W/m 2
qrad stagnation point radiative heat flux, W/m 2
qconv stagnation point convective heat flux, W/m 2
qcond conductive heat flux, W/m 2
qcw cold wall heat flux, W/m 2
qhw hot wall heat flux, W/m 2
Q* thermochemical heat of ablation, J/kg
also hot wall heat of ablation, J/kg
R universal gas constant, J/kg-mole-°K
s recession rate, m/s
ss steady state
 134

Nomenclature
T temperature, °C or K
Tw wall temperature, °C or K
T0 , Ti initial temperature, °C or K
Ts surface temperature, °C or K
Tsurr surrounding, or ambient temperature, °C or K
t time, sec
ue boundary layer edge gas velocity, m/s
x distance measured from the original surface of the ablating material, m
xS distance measured from the moving surface of the ablating material, m
Z*ie diffusion driving potential at the boundary layer edge
Z*iw diffusion driving potential at the wall
 solar absorptivity, or thermal diffusivity m 2 /s
 emissivity
 transpiration coefficient
 135

DH v
Nomenclature
enthalpy of vaporization, J/kg
DH enthalpy difference, J/kg
DH d heat of decomposition, J/kg
DT temperature difference, °C
 resin volume fraction
r residual density, kg/m3
 , or  s solid material density, kg/m3
e boundary layer edge gas density, kg/m3
 resin density of resin component, kg/m3
 fiber density of fiber reinforcement, kg/m3
(  v) w total mass flux entering the boundary layer, kg/m 2 -s
 Stephan-Boltzman constant, W/m 2 -K 4
i density exponent factor
 transpiration correction factor
 136

References
 137
References - Background
 138
References - Surface Recession
 139
References - Surface Recession
 140
References - Response Models
 141
References - Response Models
 142

References – Numerical Modeling

1. Moyer, C. B., and Rindal, R. A., “An Analysis of the Coupled Chemically Reacting
Boundary Layer and Charring Ablator – Part II. Finite Difference Solution for the
In-Depth Response of Charring Materials Considering Surface Chemical and
Energy Balances”, NASA CR-1061, 1968.
2. Katsikas, C. J., Castle, G. K., and Higgins, J. S., “Ablation Handbook – Entry
Materials Data and Design”, AFML-TR-66-262, September 1966.
3. Kratsch, K. M., Hearne, L. F., and McChesney, H. R., “Thermal Performance of
Heat Shield Composites During Planetary Entry”, Lockheed Missiles and Space,
LMSC-803099, Sunnyvale, CA, October 1963.
4. Munson, T. R., and Spindler, R. J., “Transient Thermal Behavior of Decomposing
Materials. Part I, General Theory and Application to Convective Heating”, AVCO
RAD-TR-61-10, AVCO Corp., Wilmington, MA, May 1961.
5. Curry, D. M., “An Analysis of a Charring Ablation Thermal Protection System”,
NASA TN D-3150, November 1, 1965.
6. Goldstein, H. E., “Kinetics of Nylon and Phenolic Pyrolysis”, Lockheed Missiles
and Space Company, Sunnyvale, CA. LMSC-667876, October 1965.
7. Lees, L., “Convective Heat Transfer With Mass Addition and Chemical Reactions”,
Third AGARD Colloquim on Combustion and Propulsion, Pergamon Press, New
York, 1959.