Simulating Syst Simulating Systems in Gr ems in Ground V ound - - PowerPoint PPT Presentation

Simulating Syst Simulating Systems in Gr ems in Ground V

• und Vehicle

hicle Design Design

Frederic ederick J. k J. Ross ss Direct Director

• r, Gr

, Ground T

• und Transpor

ansportation tation

Simulat Simulating Syst 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

• f S

Simulation

– Simulation growth impact on virtual Powertrain simulation

Ag Agenda

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

Simulati Simulating Syst ng Systems ems

– 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

• f 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.

Im Impact of Simul pact of Simulation on V ation on Vehicle Design hicle Design

J G F E D C B A H Today: Data Freeze Hardware Phase Assessment Vehicle Functions

• Dig. Ph. Digital Phase

Digital Prototypes Simulation Series Digital Phase Digital Phase Hardware Phase Hardware Phase Series Hardware Hardware Phase Hardware Phase Simulation Previous: Simulation Hardware Phase Hardware Phase

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

Digital prototypes have allowed decrease in hardware phase. That reduces turn-around time, and the expensive build/test

• f 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

Durability (BiW) Crash NVH Ride/Handling

HVAC/ Thermal Comfort Durability Chassis

Aerodynamics Climate Control Heat Protection Manufacturing Powertrain Transmission

Digital Pr Digital Prototype becomes enab ype becomes enabler ler for adv r advance simulation nce simulation

– Simulation for more advance analysis then just component design – Simulation includes multi-physics. – Simulation can involve motion as needed as well. Whatever best helps engineer design their product efficiently. – In the past, these would not have been possible until hardware of the vehicle has been produced.

Generation of a Generation of a Digital Pr Digital Prototype ype

Data F Data Freeze defi eeze defines dig s digital pr l prototype

– As with a real prototype, design teams work together to meet a goal for the design freeze. – Review board checks, to make sure all components are fitted together and data pool is complete.

Data Filter: Filters data for simulation

– Data needed for simulation is filtered from the

• verall 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.

Deflection speed v Damping force F Grade 1 Grade 2

Geometrical Data Functional Data

Vehicle Thermal Ma hicle Thermal Mana nagement R gement Roadmap map

1 2 3 4 5 6 7 1 2 3 4 5 6 7

Front End Air Flow

• Top Tank Temperature

Prediction

• Turn‐Around: 1 Day

1

Local Component Temperature

• 30‐60 Solids
• Local to a component
• 2‐3 Weeks

2

Total Vehicle Simulation

• Using existing sub‐models
• 2 Week Assembly

4

Underbody Temperature

• ~ 100 Solids
• Includes Exhaust System, hangers,

engine mounts, heat shields

• 3‐4 Weeks

3

Power Train Cooling

• Full Engine CHT model
• Induction System
• Exhaust System
• Oil Flow
• 4‐6 Weeks

5

Full Vehicle Thermal Management

• Conduction/Radiation using

Radtherm

• Includes Drive Cycle Simulation
• 5‐6 Weeks Modeling time?

6

Full Vehicle Thermal Management

• Co‐Simulation from STAR‐

CCM+ to STAR‐CCM+

• 4000 Solid Components
• Includes Drive Cycle

Simulation via Ports

• 6‐7 Weeks Modeling Time

7

8

GUM: Grand Unified Model

• Complete vehicle simulation
• 4000+ Solid Components
• Cabin Thermal Comfort
• Vehicle Aerodynamics
• HVAC Simulation
• Electronics Cooling
• Co‐Simulation STAR‐CCM+ to STAR‐CCM+

8

Note: Times are estimated on past projects. Actual times depend on CAD

Complexity Increased ROI

Front End Air Flo

• nt End Air Flow/A

/Aer erodynamics: 1 Da

• dynamics: 1 Day T

y Turn- rn-Around und

Challenge: Data Filtering

Large CAD database needs to be quickly moved from 1000’s of CAD parts to few boundaries needed for CFD.

Solution:

OpenRoad

• 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
• ptimization with Heeds

1

Deflection speed v Damping force F Grade 1 Grade 2

Geometrical Data Functional Data

OpenR OpenRoad ad Data Pr Data Processing

• cessing

Data Filtering

• Typical path comes from:

1. Vismockup with PlmXML & JtOpen 2. ANSA

• Some pre-processing may be

necessary

• Heat exchanger cores: set

upstream/downstream interfaces

• Fan Interfaces
• Close large wholes

Functional Data:

• Vehicle Speed.
• Heat Exchangers
• Porous Media
• Heat Rejection
• Secondary Fluid Stream
• Fan RPM
• Wall Temperature

Directly entered in OpenRoad

1 VisMockup Plmxml JtOpen STAR-CCM+ Surface Repair ANSA Organizatio n/Repair Nastran Export

Functional Data

Vehicle A hicle Aerodynamics Optimization dynamics Optimization Automation and Robustness Enable Optimization

Challenge:

Current design loops for shape

• ptimizations 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

Typical Methods

1. CAD Parameterization 2. Surface morphing

Methodology:

• Vehicle is broken down to sub-system
• Surface repair phase only affects creation
• f data not in CAD.
• Heeds can drive change in CAD to feed

new information to OpenRoad

• Full model is rebuilt with each design step

Alternative: Simulation file can be used for the baseline. Can use Optimate or Heeds to modify part.

VisMockup Plmxml JtOpen STAR-CCM+ Surface Repair ANSA Organizatio n/Repair Nastran Export

1 2

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

• nent Temperature:

erature: Brak Brake Cooling 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

2

Underb derbody Thermal Management

• dy Thermal Management

STAR AR-CCM+ Enables F

• CCM+ Enables Full A

ll Automat mation

• n
• Clients are minimizing turn-

around to go straight from CAD to results.

• Relies on surface wrapping for

providing clean closed surfaces

• Includes heat exchanger models
• May include secondary stream for

top tank temperature

• Preparation work primarily done

in CAD.

• Minimize user interaction to clean

up surfaces from CAD since focus

• n design teams that are not

experts with the CFD tool.

Good example of automatic procedure can be seen with the OpenRoad App.

3

Solid Components

Driv Drive Cy e Cycle Simulation cle Simulation

Challenge

Often, the means of failure of an engine or component occurs from operators drive cycle. As a vehicle travels in stop/go traffic, the heat load builds up and can cause overheating or more severe thermal loading issues. This can be costly to fix once the design has been finalized.

Solution Solution

Use virtual vehicle to simulate different drive cycles critical for performance. Drive cycles are simulated using a solid model to model transient thermal environment, coupled to different steady state vehicle operating conditions which make up the complete drive cycle.

Impact Impact

Thermal loading caused by critical drive cycles can now be simulated on a virtual vehicle. This identifies heating issues prior to having initial

• prototype. This reduces design time, and costs
• f prototype testing.

respo sponse se from ev

• m event

ent Unsteady Uphill Drive

Averaged cell voltage, ΔUmax=80mV

16

Ext Extending Simulation t nding Simulation to Hybrids/Electrical: Hybrids/Electrical: Lithium Ion Ba Lithium Ion Batt tter ery thermal ma y thermal mana nagement gement

• Challenge: Ensure sufficient cooling for set of lithium ion cells electrically loaded

Courtesy of ASCS Validation, Stuttgart, Germany

Temperature plots for two cells of the stack , ΔTmax=2°C Heat transported by coolant

• Solution: Coupled electrochemical-thermal performance for complete battery

system simulated with STAR-CCM+; excellent correlation with test data

• Impact: Able to use 1 simulation tool to accurately predict

battery cooling system performance; avoid overheating

GUM: Grand U GUM: Grand Unified Model nified Model

8 Grand Unif Grand Unified ied Model Model

– Presented at SAE 2012:

• SAE 2012-01-0635

– Used for final vehicle testing – Combines:

• Aerodynamics
• Vehicle Thermal Management
• Passenger Thermal Comfort
• DeIce/Defog

Com Comple lete flo flow cir circui uit of air fr

• f air from e
• m exteri

rior t

• r to
• utle
• utlets in the trun

in the trunk. k.

Powertrain Simulation R rain Simulation Roadmap admap

Component Analysis

• Port Flow
• Coolant Flow
• Intake/Exhaust Manifold flow

1

Transient Behavior

• Couple Simulation to 1D Code
• Look at EGR mixing
• Exhaust manifold temperatures

2

System Analysis

• Coolant Filling
• Crank Case Ventilation
• Oil Circuits
• Turbo Charger

3

Environment Evaluation

• Dynamometer Testing
• Engine in Vehicle
• Drive Cycle Simulation

4 Increased ROI Complexity

Validate Troubleshoot Predict Automate Optimize

Component Analysis System Analysis

19

PowerT rTrain: Simulating Syst rain: Simulating Systems ems

Challenge

Engines that pass the dynamometer still fail when installed into the vehicle. Once vehicle design has been finalized, it can be costly to adjust cooling to the engine. At early design stages, it is important to determine possible thermal issues.

Solution Solution

Use existing geometry of the engine in dynamometer and place engine in vehicle. Includes:

• Cooling Air Flow
• Air Induction System
• Coolant Flow Network
• Oil Flow

Impact Impact

• Reduce prototype of engine/vehicle

construction.

• Reduce time to find out thermal failures.
• Reduce cost
• Reduce time to production.
• Improve information on failure cause.

Powertrain Data Pr rain Data Processing

• cessing

Direct CAD Reader

• Interfaces defined in CAD brought
• ver to STAR-CCM+
• Fluid extraction used to separate

interior air/coolant/oil circuits

• Enables conformal interfaces with

polyhedral cells

Functional Data:

• Cycle average tool to extract

cylinder data from ICE simulation

• 1D tool used to specify boundary

conditions during transient

• Co-simulation used to pass data

between fluid and solid networks

• Enables coupling to steady

state fluid with transient thermal if needed.

• Material properties linked in from
• utside spread sheets?

Optimization

• Baseline Simulation file built
• CAD parameters can be exposed

in STAR-CCM+

Direct CAD Reader

'Associativ 'Associative' geome e' geometry transf transfer t er to STAR-CCM+ with a AR-CCM+ with a single click single click Design changes can be continuously performed

– One click update of design changes from CAD or from STAR-CCM+ – Parameters are selectively chosen that maybe modified in STAR-CCM+

Tree structure and naming are maintained in STAR-CCM+ during each transfer

V8.06: Bi-directional V8.06: Bi-directional Geometr eometry T y Transf ansfer (1/2) er (1/2)

CAD STAR-CCM+

Contr Control o l over ho er how geome w geometry is is transf transferred erred Coincident f Coincident faces de ces detect cted aut ed automatically matically and and par part contacts creat contacts created af ed after r transf transferring geome erring geometry fr from an

• m an assembly

assembly

– Default setting – Part contacts are always detected when transferring the full CAE model

Multi-body Multi-body par parts transf s transferred as erred as com composit

• site par

e parts Option t Option to transf transfer hidden geome er hidden geometry and and design parame design parameters rs

V8.06: Bi-directional V8.06: Bi-directional Geometr eometry Transf ransfer (2/2) r (2/2)

EGR Cooler EGR Cooler: 1 : 1 Da Day Se y Setup tup Metal T Metal Temperature Prediction erature Prediction

Challenge

EGR thermal loads can cause fatigue stress and failure. Boiling can also cause engine failure

Solution Solution

CAD import with volume extraction for coolant and exhaust gas simplify building EGR cooler. Parts base Parts based mesh meshing hel ng help enab p enable qui e quick desi ck design studi gn studies es to opti to optimize fl e flow.

• w.

Header designs for flow uniformity can be linked to Heeds to optimize the design

Impact Impact

Improved efficiency of EGR Cooler. Coolant Prediction Coolant Prediction

• Temperature/Pressure Loss

mperature/Pressure Loss

• Onset

set of

• f Boilin

iling

• Fl

Flow Uni

• w Uniformi

mity ty

Exhaust Prediction Exhaust Prediction

• Temperature/Pressure Loss

mperature/Pressure Loss

• Force

Force o

• n Fl

Flap

• Fl

Flow l

• w leakage

eakage St Structu cture

• Temperature/Thermal St

mperature/Thermal Stress ss

• Heat T

Heat Transfer ansfer

24

Coolant Jack Coolant Jacket Analysis Analysis Automation/Optimization mation/Optimization

Challenge

Coolant flow through the engine is critical for cooling of the head and block. CFD is a standard tool to help optimize the flow path. The challenge is how to optimize the coolant path quickly and efficiently.

Solution Solution

• Direct CAD import, volume extraction

makes it easy to extract coolant passage.

• Custom tool helps automate process from

CAD to PowerPoint report.

• Optimate can be used to automate gasket

sizes to maximize cooling efficiency

• 5 Man Hours
• 300 Designs Evaluated

Impact Impact

• Automation helps reduce dependency/skill
• f the engineer and enables the engineer to

concentrate on the engine design.

• Heat transfer coefficients can be quickly

generated and mapped to FEA model.

Left Bank Left Bank Right Bank Right Bank

Intake Po Port F Flow

• w A

Analysis Automated V Virt rtual P Port rt F Flow T Tool

Challenge

Combustion efficiency depends a lot on the intake air flow, tumble, and swirl to get complete, and fast burn. CFD has proven to be a valuable tool to

• ptimize port flow. Engineer needs quick design

studies to evaluate flow efficiency at different valve lifts.

Solution Solution

• Automated tool has been built and designed.
• Port Flow optimization.
• Follows work from established best practices.
• Pass data to other software/databases without

manual interactions.

Impact Impact

• Reduce errors in simulation.
• Leverage product expertise without needing

software expertise.

• Leverage the expertise of analysis to the

experts.

Return to a focus on Design as opposed to Analysis!

26

Transmissions: Splash lubri ansmissions: Splash lubrication simulation using CFD cation simulation using CFD

Challenge

Provide adequate lubrication and cooling of important components, such as bearings, gear contacts, clutches, synchronizers, etc, while working with minimum losses in the transmission.

Solution Solution

Transmission housing plays major role in guiding flow to critical areas. Simulating the splash lubrication with CFD provides visual showing the motion

• f the oil. The housing can then be

modified to bring oil to critical locations.

Impact Impact

Improved reliability of the transmission as well as fuel consumption by minimizing the losses in the engine.

27

Pist Piston Cooling

• n Cooling

Oil Spra Oil Spray Injection Modelling y Injection Modelling

Challenge

Important to cool piston during

• peration. In manufacturing of engines,

it is hard to determine if oil is directed to desired hot spot. Failure to cool piston becomes reliability issue.

Solution Solution

Simulate oil injection on to piston

• Volume of Fluid is used to

simulate oil.

• Mesh morphing used for piston

motion.

Impact Impact

• Improved cooling of engine.
• Ensures oil reaches critical areas of

the piston.

Syst System Simulation is em Simulation is im impacting design pacting design

– Reduce turn-around time in design – Reduce costs from reducing number of prototypes

Simulation is e Simulation is expanding panding

– Users are looking at replacing more expensive tests with simulation – As capability grows and mature in simulation tools, so does the demand on extending the features.

Design Exploration Design Exploration Gr Growing ing

– With increased automation provided by the STAR-CCM+ suite, optimization expanding to provide engineers with best design, in the shortest design cycle.

Summar Summary

Thank Y Thank You! u!