{"@attributes":{"version":"2.0"},"channel":{"title":"Advanced Design Technology Blog","link":"https:\/\/blog.adtechnology.com","description":"Discover all the latest information about turbomachinery design software and services in the Advanced Design Technology blog.","language":"en-gb","pubDate":"Thu, 30 Jul 2026 14:16:55 GMT","item":[{"title":"Designing a Micro Gas Turbine for Standby Emergency Power Generation","link":"https:\/\/blog.adtechnology.com\/designing-micro-gas-turbine-standby-emergency-power-generation","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/designing-micro-gas-turbine-standby-emergency-power-generation\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/HUBDB\/webinar-hubdb-micro-turbine-siemen.jpg\" alt=\"micro gas turbine\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>ADT and Siemens have partnered to demonstrate a rapid, blank-page-to-concept design workflow for an emergency <strong>Micro Gas Turbine (MGT)<\/strong>.<\/p> \n<p>By coupling <strong>Siemens Simcenter Amesim<\/strong> with <strong>TURBOdesign Suite<\/strong>, the joint workflow dramatically accelerates development:<\/p> \n<ul> \n <li><strong>System Sizing:<\/strong> Amesim defines the mission profile and passes requirements to <strong>TURBOdesign Pre<\/strong> to instantly generate matched meanline designs and performance maps.<\/li> \n <li><strong>3D Blade Design:<\/strong> <strong>TURBOdesign1<\/strong> turns these specs into optimized, 3D compressor and turbine blade geometries in just a few hours.<\/li> \n <li><strong>Multi-Domain Co-Simulation:<\/strong> The matching fluid and solid models are seamlessly exported to Simcenter for automated aerodynamic, structural, and rotor-dynamic analysis.<\/li> \n<\/ul> \n<p>This highly integrated, end-to-end loop slashes complex turbomachinery development timelines from months to days.<\/p> \n<p>&nbsp;<\/p>  \n<p style=\"font-size: 17px;\">&nbsp;<\/p> \n<ul> \n <li> <p><a href=\"#The-challenge\">The Challenge: Emergency Standby Power for Critical Infrastructure<\/a><\/p> <\/li> \n <li> <p><a href=\"#the-solution\"><span style=\"background-color: transparent;\">The Solution: A Seamless, Blank-Page Concept Workflow<\/span><\/a><\/p> <\/li> \n <li> <p><a href=\"#system-level-simulations\"><span style=\"background-color: transparent;\">System-Level Simulation &amp; Cycle Matching in Simcenter Amesim<\/span><\/a><\/p> <\/li> \n <li> <p><a href=\"#matching-radial-inflow\"><span style=\"background-color: transparent;\">Matching the Radial Inflow Turbine Stage<\/span><\/a><\/p> <\/li> \n <li> <p><a href=\"#meanline-to-3d-gemometrty\"><span style=\"background-color: transparent;\">From Meanline to 3D Geometry: Inverse Design &amp; Co-Simulation<\/span><\/a><\/p> <\/li> \n <li><span style=\"background-color: transparent;\"><a href=\"#takeaway\"><span style=\"background-color: transparent;\">Key Takeaway<\/span><\/a><\/span>&nbsp;<\/li> \n<\/ul>","category":["Pumps","Machine Learning"],"pubDate":"Thu, 30 Jul 2026 14:07:31 GMT","guid":"https:\/\/blog.adtechnology.com\/designing-micro-gas-turbine-standby-emergency-power-generation"},{"title":"What is Physics Enhanced Machine Learning (PEML)?","link":"https:\/\/blog.adtechnology.com\/what-is-physics-enhanced-machine-learning-peml","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/what-is-physics-enhanced-machine-learning-peml\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/What%20is%20Physics-Enhanced%20Machine%20Learning\/Small-Training-Dataset.jpg\" alt=\"Physics-Enhanced Machine Learning\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>Physics Enhanced Machine Learning is the only way to create an expert system for the task of detailed, workable and realizable engineering design, based on CAE and simulation. Whilst at the same time restricting the upfront training effort to something manageable that does not require a super-computer and tens thousands of individual simulation datapoints.<\/p>","category":["TURBOdesign Suite","Machine Learning"],"pubDate":"Thu, 30 Jul 2026 13:08:43 GMT","guid":"https:\/\/blog.adtechnology.com\/what-is-physics-enhanced-machine-learning-peml"},{"title":"TURBOdesign Suite 2026.1 Release","link":"https:\/\/blog.adtechnology.com\/turbodesign-suite-2026.1-release","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/turbodesign-suite-2026.1-release\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/Webinars\/2026\/TURBOdesign%20Suite%202026.1%20Webinar\/TURBOdesign-Suite-2026.1.jpg\" alt=\"TURBOdesign Suite 2026.1\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p><span style=\"color: #1f1f1f;\">The 2026.1 release advances <strong>Physics-Enhanced Machine Learning (PEML)<\/strong> and <strong>universal CAE integration<\/strong>, allowing engineers to build highly accurate optimization models with smaller datasets.<\/span><\/p> \n<ul> \n <li> <p><span style=\"font-weight: bold;\">TURBOdesign Suite 2026.1 is the Large Physics Model (LPM) Accelerator<\/span>.<\/p> <\/li> \n <li> <p><span style=\"background-color: transparent;\"><span style=\"font-weight: bold;\">Expanded Ansys Fluent Integration<\/span>: Direct automatic integration from inside TURBOdesign1, supporting all machine types, mesh formats, boundary conditions, and fluid states.<\/span><\/p> <\/li> \n <li> <p><span style=\"background-color: transparent;\"><span style=\"font-weight: bold;\">Open Machine Learning Framework: <\/span>The Reactive Response Surface (RRS) machine learning engine now connects with any external platform, including custom in-house codes and open-source packages.<\/span><\/p> <\/li> \n <li> <p><span style=\"background-color: transparent;\"><span style=\"font-weight: bold;\">HPC Scalability:<\/span> Native remote execution settings allow parallelized optimization jobs to be offloaded directly to remote Linux HPC clusters.<\/span><\/p> <\/li> \n <li> <p><span style=\"background-color: transparent;\"><span style=\"font-weight: bold;\">Advanced Modeling:<\/span> Features new Inlet Guide Vane (IGV) compressor models, automated axial thrust calculations, and two-phase flow functionality.<\/span><\/p> <\/li> \n <li> <p><span style=\"background-color: transparent;\"><span style=\"font-weight: bold;\">Meanline Design Exploration<\/span>: Introduces N-dimensional design space sampling (Latin Hypercube) to map variables before setting detailed 3D geometry<br><br><\/span><\/p> <\/li> \n<\/ul>  \n<p>&nbsp;<\/p> \n<ul> \n <li> <p><a href=\"#key-new-features\">Key New Features and Advancements<\/a><\/p> <\/li> \n <li> <p><a href=\"#overcome-data-bottlenecks\">How to overcome data bottlenecks in AI and machine learning for optimization?<\/a><\/p> <\/li> \n <li> <p><a href=\"#Artificial-Neural-Network\">What is Artificial Neural Network (ANN)?<\/a><\/p> <\/li> \n <li> <p><a href=\"#Physics-Informed-Neural-Network\">Physics Informed Neural Network (PINN)<\/a><\/p> <\/li> \n <li> <p><a href=\"#Physics-Enhanced-Machine-Learning\">Physics Enhanced Machine Learning (PEML)<\/a><\/p> <\/li> \n <li> <p><a href=\"#new-features-2026.1\">What are the New Features in TURBOdesign Suite 2026.1?<\/a><\/p> \n  <ul> \n   <li> <p><a href=\"#new-features-2026.1\">RRS with Customer Own CAE<\/a><\/p> <\/li> \n   <li> <p><a href=\"#TDPre-exploration\">TURBOdesign Pre Design Exploration<\/a><\/p> <\/li> \n   <li><a href=\"#other-improvements\">Other Improvements in TURBOdesign Suite 2026.1<\/a><\/li> \n  <\/ul> <\/li> \n<\/ul>","category":["TURBOdesign Suite","Machine Learning"],"pubDate":"Thu, 11 Jun 2026 10:31:27 GMT","guid":"https:\/\/blog.adtechnology.com\/turbodesign-suite-2026.1-release"},{"title":"Blade Loading, Inverse Design and Shape Parameterization","link":"https:\/\/blog.adtechnology.com\/blade-loading-inverse-design-shape-parameterization","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/blade-loading-inverse-design-shape-parameterization\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/Blade%20loading,%20Inverse%20design%20and%20shape%20parameterization\/Blade-loading-is-pressure-difference-from-one-side-of-the-blade-to-the-other.jpg\" alt=\"blade loading\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p><span style=\"font-weight: bold;\">Blade Loading Defined:<\/span> Fundamentally - the pressure difference between the pressure and suction sides of a blade along a meridional pathline. It directly controls the amount of work done or extracted and reveals problematic flow regions like cavitation or shockwaves.<\/p> \n<p><strong>Spanwise Work Control<\/strong>: The non-dimensionalized angular momentum (rVtheta*), and its distribution from hub to shroud (\u2018spanwise\u2019). It is an input parameter to the 3D Inverse design method and rooted directly in turbomachinery fluid dynamics.<\/p> \n<p><strong>Streamwise Loading Control<\/strong>: The other key input control in 3d Inverse Design. This defines the distribution of work from leading to trailing edge at the hub and shroud . A blade can be \"fore-loaded,\" \"aft-loaded,\" or \"mid-loaded,\" (or some combination across the span) which manages surface pressure\/velocity distribution and mitigates cross-passage gradients and other loss mechanisms..<\/p> \n<p><strong>The Inverse Solver<\/strong>: An iterative process that starts with a flow field and evolves a blade shape that will enforce it.<\/p> \n<p><strong>Shape Parameterization<\/strong>: Because designs are defined by aerodynamic controls rather than multiple geometry controls, the design space is low-dimensional and physics-led, making it highly compatible with <strong>Machine Learning<\/strong> through <strong>Reactive Response Surface<\/strong> optimization.<\/p> \n<p>&nbsp;<\/p> \n<ul> \n <li> <p><a href=\"#why_invest_wastewater_pump\">What is Blade Loading?<\/a><\/p> <\/li> \n <li> <p><a href=\"#Direct_design_vs_Inverse_Design\">What is the difference between Direct Design and 3D Inverse Design?<\/a><\/p> <\/li> \n <li> <p><a href=\"#control_spanwise_work_distribution\">How do we control the spanwise distribution of work?<\/a><\/p> <\/li> \n <li> <p><a href=\"#control_streamwise_work_distribution\">How do we control the streamwise blade loading?<\/a><\/p> <\/li> \n <li> <p><a href=\"#inverse_solver_method\">How does the 3D Inverse Design work?<\/a><\/p> <\/li> \n <li> <p><a href=\"#inverse_design_shape_parameterization\">How does Inverse Design enable shape parameterization and efficient optimization?<\/a><\/p> <\/li> \n <li> <p><a href=\"#examples_Ml\">What examples of Machine Learning for turbomachinery have used 3D Inverse Design?<\/a><\/p> <\/li> \n <li> <p><a href=\"#further_reading\">What is recommended further reading on 3D Inverse Design?<\/a><\/p> <\/li> \n<\/ul>","category":["TURBOdesign Suite","Machine Learning"],"pubDate":"Fri, 15 May 2026 11:04:37 GMT","guid":"https:\/\/blog.adtechnology.com\/blade-loading-inverse-design-shape-parameterization"},{"title":"Machine Learning for Wastewater Pump Design","link":"https:\/\/blog.adtechnology.com\/machine-learning-wastewater-pump-design","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/machine-learning-wastewater-pump-design\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/Machine%20Learning%20for%20Wastewater%20Pumps\/machine-learning-wastewater-pump3.jpg\" alt=\"Machine Learning for Wastewater Pump Designs\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>This blog describes how tools from Advanced Design Technology (ADT) use a combination of 3D Inverse Design and Machine Learning (specifically \"Reactive Response Surface + CAE\") to rapidly design high-performance wastewater pumps. The primary challenge in wastewater pump design is balancing conflicting requirements: high hydraulic efficiency, the ability to pass large solids, and preventing cavitation. By using Machine Learning, engineers can explore vast design spaces to find optimal solutions that meet strict non-clogging constraints in a matter of hours.<\/p> \n<p>&nbsp;<\/p> \n<ul> \n <li> <p><a href=\"#why_invest_wastewater_pump\">Why invest in wastewater pump design optimization?<\/a><\/p> <\/li> \n <li> <p><a href=\"#TURBOdesign-Pre_provides_starting_point\">TURBOdesign Pre provides the starting point for stage design<\/a><\/p> <\/li> \n <li> <p><a href=\"#Inverse_Design_enabling_technology\">3D Inverse Design - the enabling technology for Machine Learning<\/a><\/p> <\/li> \n <li> <p><a href=\"#volute_performance_prediction\">Including the volute in the performance prediction<\/a><\/p> <\/li> \n <li> <p><a href=\"#creating_seed_design\">Creating a seed design and specifications for running CFD within the ML training<\/a><\/p> <\/li> \n <li> <p><a href=\"#pump_optimization\">Pump optimization via Machine Learning<\/a><\/p> <\/li> \n <li> <p><a href=\"#TD1_creates_design_choices\">TURBOdesign1 creates design choices from the Machine Learning solution<\/a><\/p> <\/li> \n <li> <p><a href=\"#training_dataset\">Re-using the training dataset to optimize the pump design for a new set of objectives<\/a><\/p> <\/li> \n <li> <p><a href=\"#conclusion\">Conclusions - Machine Learning for turbomachinery design &nbsp;is now a reality<\/a><\/p> <\/li> \n<\/ul>","category":["Pumps","Machine Learning"],"pubDate":"Mon, 23 Mar 2026 15:30:31 GMT","guid":"https:\/\/blog.adtechnology.com\/machine-learning-wastewater-pump-design"},{"title":"Maximizing Refrigeration Performance with ML & 3D Inverse Design","link":"https:\/\/blog.adtechnology.com\/maximizing-refrigeration-performance-ml-3d-inverse-design","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/maximizing-refrigeration-performance-ml-3d-inverse-design\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/Maximizing%20Refrigeration%20Efficiency%20with%20ML%20Inverse%20Design\/First-stage-of-multistage-chiller-compressor.jpg\" alt=\"First-stage-of-multistage-chiller-compressor\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>Explore how 3D Inverse Design and Machine Learning are replacing conventional workflows to deliver a new benchmark in refrigeration cycle efficiency and aerodynamic precision..<\/p> \n<ul> \n <li> <p><strong>The Challenge:<\/strong> Conventional&nbsp;methods are not&nbsp;sufficient to meet modern energy efficiency demands in refrigeration applications.<\/p> <\/li> \n <li> <p><strong>The Solution:<\/strong> Conventional&nbsp;methods are not&nbsp;sufficient to meet modern energy efficiency demands in refrigeration applications.<\/p> <\/li> \n <li> <p><strong>Rapid Cycle Setup:<\/strong> Engineers can quickly define single or multi-stage cycles, utilize real gas properties (like CO2 or R134a), and instantly generate performance maps and automatic 1D sizing.<\/p> <\/li> \n <li> <p><strong>3D Inverse Design:<\/strong> Instead of leading with &nbsp;blade shapes, you specify the optimum flow field, and the software computes the exact 3D geometry needed for impellers, diffusers, and volutes.<\/p> <\/li> \n <li> <p><strong> Machine Learning:<\/strong> A Reactive Response Surface (RRS) algorithm rapidly evaluates the design space to find the optimal multi-point, multi-objective performance in a fraction of the time.<\/p> <\/li> \n <li> <p><strong> The Result:<\/strong> Faster design times, breakthrough aerodynamic performance, and versatility across all types of refrigerants.<\/p> <\/li> \n<\/ul> \n<p>&nbsp;<\/p>  \n<p style=\"font-size: 17px;\">&nbsp;<\/p> \n<p style=\"font-size: 17px;\">The demand for energy efficiency and sustainability for refrigeration applications has never been higher. Engineers are no longer just looking for \"good enough\" designs; they need a competitive edge that streamlines development while pushing the boundaries of what is possible.<\/p> \n<p style=\"font-size: 17px;\">This is where <a href=\"https:\/\/www.adtechnology.com\/products\">TURBOdesign Suite<\/a> changes the game. By moving beyond traditional trial-and-error methods, it offers an integrated, physics-based approach to the entire refrigeration cycle.&nbsp;<\/p> \n<p style=\"font-size: 17px;\">Here is how TURBOdesign Suite empowers engineers to design superior components and why its 3D Inverse Design technology is now the benchmark for design capability.<\/p> \n<ul> \n <li style=\"font-size: 17px;\"> <p><a href=\"#Axial-Fans-Hidden-Complexity\">Complete Control of Specifications<\/a><\/p> <\/li> \n <li style=\"font-size: 17px;\"> <p><a href=\"#refrigeration_cycle\" style=\"background-color: transparent;\">Complete Refrigeration Cycle Calculations<\/a><\/p> <\/li> \n <li style=\"font-size: 17px;\"> <p><a href=\"#Inverse_Design\" style=\"background-color: transparent;\">3D Inverse Design: Designing by Intent<\/a><\/p> <\/li> \n <li style=\"font-size: 17px;\"> <p><a href=\"#full_cycle\" style=\"background-color: transparent;\">Full-Cycle Component Control<\/a><\/p> <\/li> \n <li style=\"font-size: 17px;\"> <p><a href=\"#machine_learning\" style=\"background-color: transparent;\">Machine Learning for Design Optimization<\/a><\/p> <\/li> \n <li style=\"font-size: 17px;\"> <p><a href=\"#conclusion\">Conclusion - World Leading Refrigeration Cycle Design<\/a>&nbsp;<\/p><\/li> \n<\/ul>","category":["Refrigeration","Machine Learning"],"pubDate":"Thu, 26 Feb 2026 15:32:25 GMT","guid":"https:\/\/blog.adtechnology.com\/maximizing-refrigeration-performance-ml-3d-inverse-design"},{"title":"TURBOdesign Suite Release 2025.2","link":"https:\/\/blog.adtechnology.com\/turbodesign-suite-release-2025.2","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/turbodesign-suite-release-2025.2\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/Webinars\/2025\/TURBOdesign%20Suite%202025.2%20-%20Software%20Release\/TURBOdesign-2025.2-main.jpg\" alt=\"TURBOdesign Suite Release 2025.2\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>The 2025.2 release is focused on speed, automation, and manufacturing readiness for turbomachinery design.<\/p> \n<ul> \n <li> <p><strong>Faster Optimization:<\/strong> Instantly re-run the <strong>Reactive Response Surface (RRS)<\/strong> surrogate model for new optimization objectives\/constraints without repeating costly CFD simulations.<\/p> <\/li> \n <li> <p><strong>Maximum Automation:<\/strong> Introduced a new <strong>Python API<\/strong> for scripted control over all TURBOdesign1 settings, enabling greater flexibility and integration into custom workflows.<\/p> <\/li> \n <li> <p><strong>Expanded CAE:<\/strong> Added support for <strong>Cadence FineTurbo<\/strong> and <strong>ANSYS Fluent<\/strong> integration, alongside new full-stage analysis workflows (Rotor + Volute).<\/p> <\/li> \n <li> <p><strong>Production-Ready:<\/strong> Includes a new feature to generate <strong>flank millable geometry<\/strong> using user-specified tool paths, ensuring manufacturability upfront.<\/p> <\/li> \n <li> <p><strong>New Applications:<\/strong> Enhanced concept prediction tools (<strong>TD-Pre<\/strong>) with new design modules for <strong>waste-water pumps<\/strong> and <strong>mixed-flow fans<\/strong>.<\/p> <\/li> \n<\/ul> \n<p>&nbsp;<\/p>  \n<p style=\"font-size: 17px;\">&nbsp;<\/p> \n<p style=\"font-size: 17px;\">The new release of TURBOdesign Suite brings a set of wide-ranging and application specific improvements to enable designers to swiftly create, assess, and improve turbomachinery stages, and to run the design workflow with more automatization, efficiency and clarity.<\/p> \n<ul> \n <li><a href=\"#faster-optimization\">Faster and more flexible design optimization via Machine Learning<\/a><\/li> \n <li><a href=\"#python-api\">A new Python API and scripting improvements<\/a><\/li> \n <li><a href=\"#full-analysis\">Full stage analysis and more CAE integration options<\/a><\/li> \n <li><a href=\"#enhanced-features-TDPre\">Enhanced Features in TD-Pre and TURBOdesign Volute<\/a><\/li> \n<\/ul> \n<p style=\"text-align: center; padding-left: 36pt;\"><em>TURBOdesign Suite integrates end-to-end concept, meanline design, 3D design, volutes and other additional stage components and preparation for manufacture<\/em><\/p>","category":["TURBOdesign Suite","Machine Learning"],"pubDate":"Mon, 01 Dec 2025 14:50:03 GMT","guid":"https:\/\/blog.adtechnology.com\/turbodesign-suite-release-2025.2"},{"title":"What is Reactive Response Surface (RRS )?","link":"https:\/\/blog.adtechnology.com\/what-is-reactive-response-surface","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/what-is-reactive-response-surface\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/What%20is%20RRS\/reactive-response-surface-thumb.jpg\" alt=\"what is reactive response surface - RRS\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p style=\"line-height: 1.75;\">The <strong>Reactive Response Surface (RRS) <\/strong>optimizer is a <strong>Machine Learning <\/strong>method that efficiently finds optimal solutions by intelligently selecting data points for evaluation.<span style=\"background-color: transparent;\">&nbsp;Instead of relying on a large, pre-determined dataset, RRS <\/span><strong style=\"background-color: transparent;\">iteratively learns<\/strong><span style=\"background-color: transparent;\"> from existing data to decide where to focus the search next, minimizing the cost and time of complex optimization problems requiring high-fidelity simulation. The core components are a probabilistic <\/span><strong style=\"background-color: transparent;\">Response Surface Model (RSM)<\/strong><span style=\"background-color: transparent;\">, a <\/span><strong style=\"background-color: transparent;\">Multi-Objective Genetic Algorithm (MOGA)<\/strong><span style=\"background-color: transparent;\">, and a <\/span><strong style=\"background-color: transparent;\">Reactive Function<\/strong><span style=\"background-color: transparent;\"> that balances exploring the design space with exploiting known good areas. The final verified results show the RRS surrogate model retains the accuracy of high-fidelity simulations with close to identical results (within rounding) between both methods on key performance metrics.<br><br><\/span><\/p>  \n<p style=\"font-size: 17px;\">&nbsp;<\/p> \n<p style=\"font-size: 17px;\">\u25cf&nbsp; &nbsp; <a href=\"#introduction-RRS\">An Introduction to Reactive Response Surface (RRS)<\/a><br>\u25cf &nbsp; &nbsp;<a href=\"#Response-Surface-Model\">What is Response Surface Model (RSM)?<\/a><br>\u25cf &nbsp; &nbsp;<a href=\"#MOGA\">What is a Multi-Objective Genetic Algorithm (MOGA)?<\/a><br>\u25cf&nbsp; &nbsp; <a href=\"#Reactive-Function\">What is the Reactive Function?<\/a><br>\u25cf &nbsp; &nbsp;<a href=\"#Verification\">How is RRS&nbsp;Verified?<\/a><br><a href=\"#Practical-design-cases\"><span style=\"color: #000000;\">\u25cf<\/span>&nbsp;<\/a>&nbsp; &nbsp;<a href=\"#Key-Takeway\">Key Takeaway<\/a><\/p>","category":["TURBOdesign Suite","Machine Learning"],"pubDate":"Fri, 28 Nov 2025 14:22:04 GMT","guid":"https:\/\/blog.adtechnology.com\/what-is-reactive-response-surface"},{"title":"Machine Learning for Hydraulic Francis Runner Design Optimization","link":"https:\/\/blog.adtechnology.com\/machine-learning-hydraulic-turbine-francis-runner-design-optimization","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/machine-learning-hydraulic-turbine-francis-runner-design-optimization\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/Webinars\/2025\/Machine%20Learning%20Optimization%20of%20a%20Francis%20Runner\/Francis-Turbine-images-landing-page-images-1.jpg\" alt=\"machine learning design optimization hydraulic turbine francis runner\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p style=\"line-height: 1.75;\">A new methodology uses <strong>3D Inverse Design <\/strong>technology coupled with <strong>Reactive Response Surface (RRS) Machine Learning<\/strong> to rapidly optimize Francis hydraulic turbine runners. This approach requires only <strong>10 input parameters <\/strong>to explore a vast design space and, in just a few hours, discovered optimized designs that showed significant performance gains, including <strong>5-9 percentage points higher efficiency<\/strong> and an <strong>8-28% increase in shaft power<\/strong> over the baseline model. The RRS+CAE method is validated as an efficient and accurate way to find optimal multi-objective, multi-point solutions.<\/p> \n<p>&nbsp;<\/p>  \n<p style=\"font-size: 17px; text-align: left;\">&nbsp;<\/p> \n<p style=\"font-size: 17px; text-align: left;\">Electricity generation through hydropower is a key option in the portfolio of clean, renewable energy, and at the heart of many of these systems lies the Francis runner. This mixed flow rotor is the most widely used type of hydraulic turbine today, being extremely versatile, making it the ideal choice for sites with a medium head (the vertical distance the water falls, typically from 30 to 600 meters) and a very wide range flow rates from less than 1, to 100s of cubic metres per second. Its widespread adoption is a testament to the broader advantages of hydropower itself: a reliable energy source that provides grid stability that now plays a crucial role in the global transition to sustainable carbon-free power.<br><br>However, a large share of the current installed hydropower was built many years ago, before the appearance of modern digital tools. The average age of the existing hydropower installations in <a href=\"https:\/\/www.rehydro.eu\/\">Europe is 46 years<\/a> and <a href=\"https:\/\/www.eia.gov\/\">50 years in the United States<\/a>. So the turbomachinery itself will not be optimised for fully efficient performance across a range of operating conditions. There is an opportunity then to re-evaluate what it means to have an efficient Francis runner design that can fully deliver all the advantages of hydropower.<br><br>In this blog we look at how ADT\u2019s <a href=\"https:\/\/blog.adtechnology.com\/what-is-reactive-response-surface\">Reactive Response Surface<\/a> + CAE technology (RRS+CAE) is driving better hydraulic turbine design through Machine Learning.<a href=\"#Axial-Fans-Hidden-Complexity\">&nbsp;<\/a><\/p> \n<h2><span style=\"font-size: 18px;\"><a href=\"#francis-runner-performance-challenge\">\u2022 The Francis Runner performance challenge - and the solution<\/a><br><\/span><span style=\"font-size: 18px;\"><a href=\"#starting-point-meanline-design\">\u2022 Where to start - Generate a&nbsp;meanline Francis runner design<\/a><br><\/span><a href=\"#enabling-technology-ML\" style=\"font-size: 18px; background-color: transparent;\">\u2022 3D Inverse Design is&nbsp;the enabling technology for Machine Learning<\/a><br><a href=\"#establishing-baseline\" style=\"font-size: 18px; background-color: transparent;\">\u2022 How to establish&nbsp;a baseline for turbine performance<\/a><br><a href=\"#optimization-ML\" style=\"font-size: 18px; background-color: transparent;\">\u2022 Optimization of a Francis runner via Machine Learning<\/a><br><a href=\"#design-choices-performance-gains\" style=\"font-size: 18px; background-color: transparent;\">\u2022 RRS gives design choices and performance gains<\/a><br><span style=\"font-size: 18px;\">\u2022<\/span> <a href=\"#final-validation-ML-solution\" style=\"font-size: 18px; background-color: transparent;\">Final validation of the Machine Learning solution<\/a><br><a href=\"#conclusion\" style=\"background-color: transparent;\"><span style=\"font-size: 18px;\">\u2022 Conclusions<\/span><\/a><\/h2>","category":["Francis Turbines","Machine Learning"],"pubDate":"Wed, 19 Nov 2025 11:15:44 GMT","guid":"https:\/\/blog.adtechnology.com\/machine-learning-hydraulic-turbine-francis-runner-design-optimization"},{"title":"Machine Learning for Centrifugal Compressor Design","link":"https:\/\/blog.adtechnology.com\/machine-learning-centrifugal-compressor-design","description":"<div class=\"hs-featured-image-wrapper\"> \n <a href=\"https:\/\/blog.adtechnology.com\/machine-learning-centrifugal-compressor-design\" title=\"\" class=\"hs-featured-image-link\"> <img src=\"https:\/\/blog.adtechnology.com\/hubfs\/ADT%20Blog\/Machine%20Learning%20for%20Centrifugal%20Compressor%20Optimization\/machine-learning-centrifugal-compressor-design.jpg\" alt=\"machine learning centrifugal compressor design\" class=\"hs-featured-image\" style=\"width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;\"> <\/a> \n<\/div> \n<p>The centrifugal compressor design process can be significantly accelerated and enhanced using <strong>3D Inverse Design<\/strong> paired with a <strong>Reactive Response Surface<\/strong> (RRS) <strong>Machine Learning<\/strong> optimizer. By using just <strong>8 input parameters<\/strong> to define the blade shape, the system efficiently explored a multi-point, multi-objective design space in only <strong>30 hours<\/strong>. The optimized design resulted in a <strong>2\u20134 percentage point gain in efficiency and extended map width<\/strong> (increased choke margin) over the baseline model, while also improving mechanical properties (reduced stresses). This validation proves the RRS with CAE Machine Learning approach is an accurate and cost-effective method for complex turbomachinery optimization.<\/p> \n<p>&nbsp;<\/p>  \n<p>&nbsp;<\/p> \n<p style=\"font-size: 17px;\">A huge amount of engineering and computational resources goes into the design of turbomachinery especially to meet multi-point, multi-objective, multi-disciplinary problems. This is especially true for centrifugal compressors, which, in almost all circumstances are range operating machines, where the ability to perform consistently at low to high speeds and low to high pressure ratios is as important as peak efficiency at the nominal \u2018design point\u2019.<br><br>The ongoing quest for centrifugal compressor map-width has, until now, been something of an empirical affair, especially when it comes to the design of the rotor itself, i.e. ignoring treatments that can be applied to the casing, diffuser and inlet. Rotor map-width and peak efficiency tend (but not always) to be a trade-off, meaning that you can have one but not the other.<\/p> \n<p style=\"font-size: 17px;\">Ideally then, we would want to optimize compressor rotor designs to give the very best efficiency and map-width possible for a given shaft and tip-speed (so - multi-point optimization). The traditional obstacle to this aim has been the sheer amount of high-fidelity simulation that has to be done in order to sufficiently explore a very complex and high dimensional design space, at multiple operating points. What if we could leverage a small dataset of high-fidelity data to accurately drive an optimization algorithm using Machine Learning methods?<br><br>At ADT we have considered very carefully how one might leverage the power of Machine Learning to the specific and highly complex problem of turbomachinery blade design.<\/p> \n<p style=\"font-size: 17px;\">\u25cf &nbsp; <a href=\"#Inverse-Design\">3D Inverse Design - the enabling technology for Machine Learning<\/a><br>\u25cf &nbsp; <a href=\"#Example-Legacy-Compressor\">An example of how Machine Learning can optimize a legacy compressor design<\/a><br>\u25cf &nbsp; <a href=\"#Optimization-ML\">TURBOdesign1 performs optimization via Machine Learning<\/a>&nbsp;<br>\u25cf &nbsp; <a href=\"#Design-Choices\">Design choices and centrifugal compressor&nbsp;performance gains<\/a><br>\u25cf &nbsp; <a href=\"#Validation\">Validation of the optimized design gives confidence in the Machine Learning solution<\/a>&nbsp;<br>\u25cf &nbsp; <a href=\"#Conclusion\">Conclusions - Machine Learning for turbomachinery is now a reality<\/a><\/p>","category":["Compressors","Machine Learning"],"pubDate":"Wed, 08 Oct 2025 10:25:16 GMT","guid":"https:\/\/blog.adtechnology.com\/machine-learning-centrifugal-compressor-design"}]}}