Simulating Syst Simulating Systems in Gr ems in Ground V ound Vehicle hicle Design Design Frederic ederick J. k J. Ross ss Direct Director or, Gr , Ground T ound Transpor ansportation tation
Agenda Ag Simulating Syst Simulat ng Systems – How does this apply to ground transportation? – How CAD/PLMXML are key links to simulating systems. – How simulation has matured over time to enable simulation of systems. Ve Vehicle R Roadmap S Simulation – Simulation growth with the vehicle chassis teams have grown over time from simple front end air flow, to complex drive conditions Powe Powertrain Ro Roadmap o of S Simulation – Simulation growth impact on virtual Powertrain simulation
Simulating Syst Simulati ng Systems ems What is meant What is meant b by sim simulati lating sys systems ems? Operating systems under various real-world – tasks through simulation Example of different systems – • Vehicle Chassis • Powertrain • HVAC • Exhaust • Transmission Interdisciplinary field of engineering – • 1D System Tool Analysis • Structural Studies • Multi-body: Ride/Handling • Fluid Dynamics – Vehicle Chassis Systems – Powertrain Systems Complex interaction among multiple variables/physics within – a system • Water/Air interaction • Fluid/solid interaction • Flow/Thermal/Stress simulations Deals with work processes, optimization, risk management – Wh Why y are w are we seeing increase seeing increase of user of users s simulating simulating syst systems? ems? Parts can be optimized early on in the concept phase where – previously expensive prototypes needed to be built.
Impact of Simul Im pact of Simulation on V ation on Vehicle Design hicle Design Digital prototypes have allowed decrease in hardware phase. That reduces turn-around time, and the expensive build/test of the hardware prototype. Efficiency: Virtual Data Freeze allows for digital evaluation to be made J H G F E D C B A Previous: Hardware Hardware Phase Hardware Phase Hardware Phase Hardware Phase Series Simulation Simulation Today: Digital Phase Digital Phase Hardware Phase Hardware Phase Dig. Ph. Digital Phase Digital Prototypes Series Hardware Phase Simulation Data Freeze Assessment Vehicle Functions Digital prototypes have allowed decrease in hardware phase. That reduces turn-around time, and the expensive build/test of the hardware prototype. *Exert from 2005 presentation from Walter Bauer on Virtual Product Development process at Daimler AG.
Simulation using the Digital Pr Simulation using the Digital Prototype ype Aerodynamics Heat Protection Digital Pr Digital Prototype becomes enab ype becomes enabler ler for adv r advance simulation nce simulation HVAC/ NVH – Simulation for more advance Thermal Comfort analysis then just component design – Simulation includes multi-physics. Manufacturing Climate Control Transmission Powertrain – Simulation can involve motion as needed as well. Whatever best helps engineer design their product efficiently. Durability Ride/Handling Crash Durability Chassis (BiW) – In the past, these would not have been possible until hardware of the vehicle has been produced.
Generation of a Digital Pr Generation of a Digital Prototype ype Data F Data Freeze defi eeze defines dig s digital pr l prototype As with a real prototype, design teams work – Damping force F together to meet a goal for the design freeze. 3000 2500 Grade 1 Review board checks, to make sure all – 2000 1500 Grade 2 1000 components are fitted together and data pool is 500 0 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 complete. Deflection speed v Data Filter: Filters data for simulation Geometrical Data Functional Data Data needed for simulation is filtered from the – overall data pool, and provided for the virtual simulation. • Key component for data transfer Example of data filters: – • Red Cedars Heeds • Custom tool designed to pull data together. – OpenRoad • CAD plugin can help provide data filter PLM (product lifecycle management) tools – enable communication between different tools. Analysis Response Feeds back into the data pool for design – improvement.
Vehicle Thermal Ma hicle Thermal Mana nagement R gement Roadmap map GUM: Grand Unified Model • Complete vehicle simulation • 4000+ Solid Components 8 • Cabin Thermal Comfort • Vehicle Aerodynamics • HVAC Simulation • Electronics Cooling Full Vehicle Thermal • Co ‐ Simulation STAR ‐ CCM+ to STAR ‐ CCM+ Management • Co ‐ Simulation from STAR ‐ CCM+ to STAR ‐ CCM+ 7 • 4000 Solid Components Increased ROI • Includes Drive Cycle Simulation via Ports • 6 ‐ 7 Weeks Modeling Time Full Vehicle Thermal Management • Conduction/Radiation using 6 Radtherm Includes Drive Cycle Simulation • • 5 ‐ 6 Weeks Modeling time? Underbody Temperature Power Train Cooling • ~ 100 Solids • Full Engine CHT model 3 • Includes Exhaust System, hangers, • Induction System 5 engine mounts, heat shields • Exhaust System • 3 ‐ 4 Weeks • Oil Flow • 4 ‐ 6 Weeks Front End Air Flow • Top Tank Temperature 1 Total Vehicle Simulation Prediction 4 • Using existing sub ‐ models • Turn ‐ Around: 1 Day • 2 Week Assembly Local Component Temperature • 30 ‐ 60 Solids 2 • Local to a component 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 • 2 ‐ 3 Weeks Complexity Note: Times are estimated on past projects. Actual times depend on CAD
Front End Air Flo ont End Air Flow/A /Aer erodynamics: 1 Da odynamics: 1 Day T y Turn- rn-Around und 1 Challenge: Data Filtering Large CAD database needs to be quickly moved from 1000’s of CAD Damping force F parts to few boundaries needed for 3000 2500 Grade 1 CFD. 2000 1500 Grade 2 1000 500 0 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 Solution: Deflection speed v OpenRoad Geometrical Data Functional Data • Provides part filtering with link to boundary setup for the simulation. • Forms template for the full simulation process including dual stream heat exchangers. Impact: • Enables users to quickly predict drag and/or front end air flow. • Enabler for more complex studies such as component temperature prediction, soiling, aero-acoustics • Runs fully in batch: good for optimization with Heeds
OpenR OpenRoad ad Data Pr Data Processing ocessing 1 Data Filtering • Typical path comes from: 1. Vismockup with PlmXML & JtOpen VisMockup ANSA 2. ANSA • Some pre-processing may be Plmxml Organizatio necessary JtOpen n/Repair • Heat exchanger cores: set upstream/downstream STAR-CCM+ interfaces Nastran Surface • Fan Interfaces Export • Close large wholes Repair Functional Data: • Vehicle Speed. • Heat Exchangers • Porous Media • Heat Rejection • Secondary Fluid Stream • Fan RPM • Wall Temperature Directly entered in OpenRoad Functional Data
Vehicle A hicle Aerodynamics Optimization dynamics Optimization Automation and Robustness Enable Optimization Challenge: Current design loops for shape optimizations allow significant improvement in drag. Turn-around needs to be quick. Solution: Shape optimization using CFD provides quick alternatively to determine shape sensitivity enabling drag reduction during vehicle design cycle. Sherpa enables user to find optimum design with fewer evaluations. Impact: Provides fast, robust turn-around for design optimization
Vehicle Optimization hicle Optimization 1 2 Typical Methods 1. CAD Parameterization 2. Surface morphing Methodology: • Vehicle is broken down to sub-system VisMockup ANSA • Surface repair phase only affects creation Plmxml Organizatio of data not in CAD. • Heeds can drive change in CAD to feed JtOpen n/Repair new information to OpenRoad • Full model is rebuilt with each design step STAR-CCM+ Nastran Surface Alternative: Export Repair Simulation file can be used for the baseline. Can use Optimate or Heeds to modify part.
Vehicle A hicle Aerodynamics Optimization dynamics Optimization Adjoint Solver Challenge: Running design optimization can require evaluation of a lot of different design variables which drive high computational resources for early design studies Solution: Adjoint solver allows many design variables to be evaluated on a single solution. By hooking adjoint with Heeds, a range of design parameters can quickly be achieved. Impact: Reduces computation cost for key drag reduction locations
Local Com Local Component T onent Temperature: erature: 2 Brake Cooling Brak e Cooling Modeling Brak Modeling Brake Coolin e Cooling • Coupling between separate Star-CCM+ analyses • Moving Reference Frame (MRF) to capture the influence of rotating components on fluid motion • Mixed Free and Forced Convection Flows with Radiation • Detailed component temperatures and flow visualization • Ability to analyze a variety of braking scenarios of varying durations
Recommend
More recommend
