Appli plications cations of f CFD and d Desi sign gn Exp xplorat loration ion in the Energy rgy & Power r industr dustry Jim m Rya yan
Des esig ign Expl plorati ration on wit ith CFD FD – in in E Ener ergy gy & P & Power er in indu dustry Design Exploration: key concepts and examples A “Maturity Model” for Engineering Simulation Gas Turbines • Turbine blade cooling with Conjugate Heat Transfer (CHT) Combustor liner cooling • with Conjugate Heat Transfer (CHT)* • Combustor flows, temperatures, and emissions* Centrifugal Pumps & Hydro Turbines *This topic is beyond the scope of this presentation 2
Energy & Power Simulation Solutions Gas Turbines Steam Turbines Compressors Combustion Heat Exchangers Renewables Nuclear Pumps & Fans Balance of Plant (Wind, Solar) Hydro Turbines (Ducting, SCRs, etc.) 3
Des esig ign Expl plorati ration on wit ith STAR-CCM+ CCM+ # of Time Designs Set Up Solve & Import Mesh Geometry Physics Visualize Design #N+1 Design #N Change Design (geometry and physics) 4
A Maturit ity y Mode del for r Engi ginee eerin ing g Sim imula lati tion on Ultimate Goal: Discover Better Designs Faster = Feasible = Infeasible Explore digitally, Optimize Confirm physically Explore Objective 2 Predict Objective 1 Troubleshoot Critical inversion point (from reactive to proactive engineering) Validate 5
Des esig ign Expl plorati ration on Concepts epts: : Hea eat Excha hange ger r exampl mple 6
Des esig ign Expl plorati ration on Concepts epts: : Hea eat Excha hange ger r exampl mple Heat Exchanger Objectives: 1) Maximize Heat Transfer (Temperature Change) 2) Minimize Pressure Drop Baseline Design Pressure Drop (Pascals) = a Design iteration = a Best Design = Design improvement (i.e., Better designs) Pareto Front of Best Designs Temperature Change (degrees) 7
SHERPA Benchmark Example Des esig ign Expl plorati ration on Concepts epts: : Hea eat Excha hange ger r exampl mple STAR-CCM+ Change design Responses NOTE: Single Objective History Plots shown here for visualization purposes variables SHERPA / Optimate+ 8
Des esig ign Expl plorati ration on wit ith HEEDS Com ompone ponents nts Modeler Multi-disciplinary process automation • • Scalable high performance computing • Efficient exploration (optimization, DOE) • Sensitivity & robustness assessment Step eps: • Drag and drop process definition • Assignment of compute resources (HPC) Explorer • Define design variables, ranges, constraints • Define responses of interest • Explore, optimize, process results • Assess sensitivity & robustness MDX = Multi-disciplinary Design eXploration 9
2 Sim imil ilar r but Dif iffer eren ent t Envir ironm onmen ents ts for Des esig ign Expl plorati ration on HEE EEDS DS Optim imat ate+ for General CAE and MDX for Design Exploration within STAR-CCM+ DIFFERENT HEEDS Solver HEEDS Solver IDENTICAL IDENTICAL 10
Duct ct Fl Flow w Des esig ign Expl plorat oration ion 1.5 m Challenge: With flow through a duct, rapidly assess changes in 1.0 m pressure due to different turning-vane configurations N = 4 N = 4 N = 4 0.5 m 2.0 m R = 0.15 R = 0.30 R = 0.45 0.5 m Solution: Automated design exploration using STAR-CCM+ with parameterized 3D-CAD and 0.5 m Optimate Turning Vanes: 1.0 m N = 7 N = 7 N = 7 • Uniform, finite thickness R = 0.15 R = 0.30 R = 0.45 • 0.10m < Radius < 0.50m Impact: • Vane count: 1 to 10 • Find better designs • Speed-up time-to-results by as much as 10X N = 10 N = 10 N = 10 • Accelerate time-to-market R = 0.15 R = 0.30 R = 0.45 11
Cen entrif ifuga ugal l Pump mp Des esig ign Expl plorat oration ion Challenge: 1) Modify impeller to increase pump efficiency; minimize power required STAR-CCM+ STAR-CCM+ 2) Obtain set of lowest-power pump designs for set of outlet pressures Solution: • Parametric blade design (3rd-party) Requirements Performance • Flow simulation (STAR-CCM+) • Process automation (HEEDS) Baseline CFturbo CFturbo STAR-CCM+ CFturbo • Optimization (HEEDS) Design Optimized Design Baseline Design Optimization High Power Flow rate = 400 m 3 /h Flow rate = 400 m 3 /h Impact: Pareto Front Required Pressure head = 30 m Pressure head = 30 m SHERPA • Power reduced by 6% Power required = 36.0 kW Power required = 38.4 kW Optimal Design • Found 33 improved designs; SHERPA “ I can now obtain better pump designs faster SHERPA 6% improvement! not just 1 that is “good enough” by spending more time on engineering decision-making, Violates • Scalable platform for optimization and less time on model setup & data transfer .” Constraint and multi-disciplinary simulations – Ed Bennett, VP of Fluids Engineering, Mechanical Solutions Inc. (MSI) 12
GT Blade de Cooli ling g through ough CHT Fluid/Solid mesh considerations for increased solution fidelity: Geometry capturing Conformal interfaces Prism layers Solid Fluid 13
GT Blade de Cooli ling g through ough CHT Fluid/Solid mesh considerations Polyhedral cells allow for accurate representation of complex geometry Solid Fluid 14
GT Blade de Cooli ling g through ough CHT Fluid/Solid mesh considerations Conformal meshes along the entire Fluid/Solid interface yields increased accuracy Solid 3 Prism Layers Fluid 15
Engi ginee eerin ing g Sim imula lati tion on Maturi rity ty Mode del Ultimate Goal: Discover Better Designs Faster = Feasible = Infeasible Optimize Explore Objective 2 Predict Objective 1 Troubleshoot Validate 16
B&B &B-AGEM GEMA: A: Gas Turbin bine e Blade de Cooli ling Turbulence Models Before Upgrade NASA Mark II test vane Challenge: Reduce cost & effort to develop and upgrade gas turbine engines while ensuring proper temperature levels Solution: Validated Conjugate Heat Transfer (CHT) simulations with STAR-CCM+ Impact: • Rapid, reliable A-to-B comparisons • Significantly improved cooling efficiency After Upgrade Transition Model Settings (needed for increased firing temperatures) • Reduced costs; fewer experimental tests “STAR -CCM+, with its high level of automation, meshing capabilities and high solution accuracy, is the best commercial CAE tool to perform fast and accurate simulations of conjugate heat transfer.” – René Braun, B&B-AGEMA 17
Engi ginee eerin ing g Sim imula lati tion on Maturi rity ty Mode del Ultimate Goal: Discover Better Designs Faster = Feasible = Infeasible Optimize Explore Objective 2 Predict Objective 1 Troubleshoot Validate 18
Conjug jugate e Hea eat Transfer er (CHT) Case S e Study dy Presented at the 2014 STAR-Global conference in Vienna The role of CHT analysis in the design process for cooled gas turbine components – Design process of the Kawasaki L30A – Upgrade of an E-class gas turbine – Novel film cooling technologies 19
Des esig ign of the L3 e L30A 0A Kawasaki L30A is the highest efficiency industrial 30 MW GT Full conjugate heat transfer analysis of the first stage vane 20
Des esig ign of the L3 e L30A 0A Boundary Conditions: • For primary gas path: stagnation inlet & pressure outlet specified • For sealing inlets: mass flow inlet specified For cooling inlets (hub and shroud): • stagnation inlet specified For cooling holes: mass flow is calculated (i.e., not specified) 21
Des esig ign of the L3 e L30A 0A All internal geometric detail retained Modeled metal inserts to capture impingement cooling effect 22
Des esig ign of the L3 e L30A 0A 23
Des esig ign of the L3 e L30A 0A Polyhedral mesh with prism layers 13.8M cells Conformal fluid-solid interface valuable for CHT 24
Des esig ign of the L3 e L30A 0A 25
Des esig ign of the L3 e L30A 0A 26
Des esig ign of the L3 e L30A 0A Full conjugate heat transfer (CHT) analysis of the first-row turbine vane Analysis included all geometric detail including vane internals and inserts Very good agreement of results (simulation VS. experiment) Provided a detailed understanding of thermal profile and potential issues 28
Engi ginee eerin ing g Sim imula lati tion on Maturi rity ty Mode del Ultimate Goal: Discover Better Designs Faster = Feasible = Infeasible Optimize Explore Objective 2 Predict Objective 1 Troubleshoot Validate 29
Recommend
More recommend
