Find your FIT A Comparison of Strategies for Simulating Vehicle Heat - - PowerPoint PPT Presentation

▶
find your fit
SMART_READER_LITE
LIVE PREVIEW

Find your FIT A Comparison of Strategies for Simulating Vehicle Heat - - PowerPoint PPT Presentation

Feb 16, 2023 234 likes •663 views

Find your FIT A Comparison of Strategies for Simulating Vehicle Heat Protection Test Cycles in 3D The Thermal Management Process 2 The Thermal Management Process The ideal process minimizes the cost of each step Source Model Meshing

slide-1
SLIDE 1

Find your FIT

A Comparison of Strategies for Simulating Vehicle Heat Protection Test Cycles in 3D

slide-2
SLIDE 2

2

The Thermal Management Process

slide-3
SLIDE 3

3

The Thermal Management Process

Meshing Source Inputs Model Construction Calculation Model Revision Post Processing Communicate Results

The ideal process minimizes the cost of each step

slide-4
SLIDE 4

What is the most effective thermal management process?

  • Needs to support high volume production work
  • Easily adapt to specialized jobs
  • Minimize resource requirements

4

Meshing Source Inputs Model Construction Calculation Model Revision Post Processing Communicate Results

slide-5
SLIDE 5

Find your FIT

5

Calculation

Speed (m/s)

Method 1 Method 2 Method 3

FIT

slide-6
SLIDE 6

Simulation methods Approach to evaluation Results Conclusions

6

slide-7
SLIDE 7

Different Strategies

7

  • CHT - Conjugate Heat Transfer
  • Step-wise
  • Psuedo Transient
  • 1D Surrogate
  • 2D surrogate
slide-8
SLIDE 8

8

Conjugate Heat Transfer

Methods Approach Results Conclusions

  • Solves as one solution
  • Very detailed
slide-9
SLIDE 9

9

CFD Coupling

CFD TAITherm

Surface temps (Twall) Convection coefficients or fluid velocities & fluid temperatures (h and Tfluid)

slide-10
SLIDE 10

10

Test Cycle Duration(s) Transient Solid Domain - TAITherm

Tw HTC & Tf Tw

Coupling Interval

Step Wise

Tw HTC & Tf Tw Repeated Process

Methods Approach Results Conclusions

CFD CFD

slide-11
SLIDE 11

Psuedo-Transient

11

CFD

Steady-state repr. t=0

TAITherm

Steady-state at t=0

CFD

Steady-state repr. t=4

1 2 3 4 5 6 7 8 9 Time

CFD

Steady-state repr. t=9

TAITherm Initial Thermal Model (estimated convection) TAITherm

Transient from t=0 to t=4

1

2 3 4 5

Methods Approach Results Conclusions

slide-12
SLIDE 12

CFD

Steady-state repr. t=0

TAITherm

Steady-state at t=0

CFD

Steady-state repr. t=4

1 2 3 4 5 6 7 8 9 Time

CFD

Steady-state repr. t=9

TAITherm Initial Thermal Model (estimated convection) TAITherm

Transient from t=0 to t=4

1

2 3 4 5

CFD

Steady-state repr. t=4

TAITherm

Transient from t=0 to t=4

Heat Transfer Coefficient Time

Run Thermal model with initial CFD HTC and Tf Tw CFD Data from time = 0 CFD Data from time = 4, prior to transient coupling CFD Data from time = 4, after 1st transient coupling loop

Methods Approach Results Conclusions

slide-13
SLIDE 13

Pseudo-Transient

13

CFD

Steady-state repr. t=0

TAITherm

Steady-state at t=0

CFD

Steady-state repr. t=4

1 2 3 4 5 6 7 8 9 Time

CFD

Steady-state repr. t=9

TAITherm Initial Thermal Model (estimated convection) TAITherm

Transient from t=0 to t=4

1

2 3 4 5

CFD

Steady-state repr. t=4

TAITherm

Transient from t=0 to t=4

Heat Transfer Coefficient Time

Run Thermal model with updated CFD HTC and Tf Tw

slide-14
SLIDE 14

Pseudo-Transient

14

CFD

Steady-state repr. t=0

TAITherm

Steady-state at t=0

CFD

Steady-state repr. t=4

1 2 3 4 5 6 7 8 9 Time

CFD

Steady-state repr. t=9

TAITherm Initial Thermal Model (estimated convection) TAITherm

Transient from t=0 to t=4

1

2 3 4 5

CFD

Steady-state repr. t=4

TAITherm

Transient from t=0 to t=4

Heat Transfer Coefficient Time

Run Thermal model with updated CFD

slide-15
SLIDE 15

Psuedo-Transient

15

CFD

Steady-state repr. t=0

TAITherm

Steady-state at t=0

CFD

Steady-state repr. t=4

1 2 3 4 5 6 7 8 9 Time

CFD

Steady-state repr. t=9

TAITherm Initial Thermal Model (estimated convection) TAITherm

Transient from t=0 to t=4

1

2 3 4 5

TAITherm

Transient from t=4 to t=9

6

Methods Approach Results Conclusions

slide-16
SLIDE 16

16

Surrogate Modeling Process

Sample range of vehicle operating conditions Compute a steady state CHT solution at each operating condition Fit an equation to the convective boundary conditions Run transient thermal model using surrogate model to approximate convective boundary conditions 1D 2D

Uniform Sampling of Vehicle Speed OLHC of Vehicle Speed and Inlet Temperature Coupled CHT solutions Linear Interpolation Gaussian Anisotropic Kriging Coupled CHT solutions Leveraged Existing Software Features Custom Developed Coupling Harness

Methods Approach Results Conclusions

slide-17
SLIDE 17

17

Traditional Conjugate Heat Transfer Simulation Stepwise Transient Surrogate Models Pros Cons

  • High Accuracy
  • Easiest process
  • Large computational

costs

  • Inflexible resource

allocation

  • Steady fluids

assumption

  • Reduced runtimes
  • Flexible resource

allocation

  • Reduced runtimes
  • Models can be reused
  • Flexible resource allocation
  • Flexible post analysis
  • ptions
  • Many samples required
  • Complex process
  • Steady sample point

assumption Psuedo Transient

  • Reduced runtimes
  • Flexible resource

allocation

Methods Approach Results Conclusions

  • Complex process
  • Steady state fluid

assumptions

slide-18
SLIDE 18

CoTherm

Process automation software from ThermoAnalytics

slide-19
SLIDE 19

19

The Thermal Management Process

Meshing Source Inputs Model Construction Calculation Model Revision Post Processing Communicate Results

  • CoTherm
slide-20
SLIDE 20

Drive Cycle Extension – 1D Surrogate

  • Inputs:
  • Thermal/CFD models
  • Drive cycle data

20

  • Output:
  • Transient thermal

model

  • Determines coupling

points based on Drive Cycle Profile

  • Runs steady thermal-CFD

cases

  • Imports CFD results into

transient thermal model

  • Runs transient thermal

model

  • CoTherm
slide-21
SLIDE 21

Psuedo Transient Method

  • Inputs:
  • Base Thermal/CFD

models

  • Boundary conditions
  • Coupling interval

21

  • Output:
  • Merged thermal

model with all CFD points

  • Automatically sets

up SS CFD models

  • Couples Thermal and

CFD models

  • Merges thermal

models

  • CoTherm
slide-22
SLIDE 22

22

  • Selected highly simplified engine

bay geometry

  • 34,602 surface elements
  • 275,748 volume elements

Methods Approach Results Conclusions

slide-23
SLIDE 23

23

50 100 150 200 250 300 350 400 200 400 600 800 1000 1200 1400 200 400 600 800 1000 1200 1400 1600 1800 2000

Tf(K) HTC (w/m2K) Time (s)

HTC Tf

Crank Case Manifolds Head Block

Methods Approach Results Conclusions

slide-24
SLIDE 24

24

Inlet Pressure Outlet Pressure Outlet

295 300 305 310 315 320 325 330 335 2 4 6 8 10 12 14 16 500 1000 1500 2000

Inlet Temperature (K) Inlet Speed (m/s) Time (s)

Inlet Speed Inlet Temperature

Methods Approach Results Conclusions

slide-25
SLIDE 25

25 5 10 15 20 25 30 35 40 500 1000 Speed (m/s) Time (s)

Cycle 1

5 10 15 20 25 30 35 40 1000 2000 3000 4000 Speed (m/s) Time (s)

Cycle 3

5 10 15 20 25 30 35 40 500 1000 1500 2000 Speed (m/s) Time (s)

Cycle 2

Duration (s) Max Speed (m/s)

  • Avg. Speed (m/s)

Volatility 1350 1800 3600 38.8 18.0 2.98 36.5 12.9 0.27 31.4 18.6 0.06

Methods Approach Results Conclusions

slide-26
SLIDE 26

26

Methods Approach Results Conclusions

Cycle 1 Cycle 2 Cycle 3

slide-27
SLIDE 27

27

Stepwise Transient Prediction - Temperature

Cycle 1 Cycle 2 Cycle 3

6 node average

5 10 15 20 25 30 35 40 45 300 350 400 450 500 550 200 400 600 800 1000 1200 1400 Vehicle Speed (m/s) Temperature (K) Time (s) CHT Stepwise - 30s Vehicle Speed 5 10 15 20 25 30 35 40 300 320 340 360 380 400 420 500 1000 1500 2000 Vehicle Speed (m/s) Temperature (K) Time (s)

CHT Stepwise - 30s Vehicle Speed

Methods Approach Results Conclusions

300 320 340 360 380 400 420 440 500 1000 1500 2000 2500 3000 3500

Temperature (K) Time (s)

slide-28
SLIDE 28

28

5 10 15 20 25 30 35 40 45 300 350 400 450 500 550 500 1000 Vehicle Speed (m/s) Temperature (K) Time (s) CHT 1D Surrogate 2D Surrogate Vehicle Speed

5 10 15 20 25 30 35 40 300 320 340 360 380 400 420 440 500 1000 1500 2000

Vehicle Speed (m/s) Temperature (K) Time (s)

CHT 1D Surrogate 2D Surrogate Vehicle Speed

Cycle 1 Cycle 2 Cycle 3

5 10 15 20 25 30 35 300 320 340 360 380 400 420 440 500 1000 1500 2000 2500 3000 3500

Vehicle Speed (m/s) Temperature (K) Time (s)

CHT 1D Surrogate 2D Surrogate Vehicle Speed

Surrogate Model Transient Prediction – Temperature

6 node average

Methods Approach Results Conclusions

slide-29
SLIDE 29

29

Cycle 1 Cycle 2 Cycle 3

5 10 15 20 25 30 35 40 45 300 350 400 450 500 550 200 400 600 800 1000 1200 1400

Vehicle Speed (m/s) Temperature (K) Time (s)

CHT Psuedo-Transient 30s Vehicle Speed 5 10 15 20 25 30 35 40 300 320 340 360 380 400 420 500 1000 1500 2000

Vehicle Speed (m/s) Temperature (K) Time (s)

CHT Psuedo-Transient 30s Vehicle Speed

Psuedo Transient Prediction – Temperature

6 node average

Methods Approach Results Conclusions

300 320 340 360 380 400 420 440 1000 2000 3000

Temperature (K) Time (s)

slide-30
SLIDE 30

30

0.0 1.0 2.0 3.0 4.0 5.0 6.0 1D Surrogate 2D Surrogate Stepwise - 30s Psuedo-Transient 30s

RMSE (K)

Manifold

Cycle 1 Cycle 2 Cycle 3 0.0 1.0 2.0 3.0 4.0 5.0 6.0 1D Surrogate 2D Surrogate Stepwise - 30s Psuedo-Transient 30s

RMSE (K)

Top Wall

Cycle 1 Cycle 2 Cycle 3

55 node average 6 node average

0.5 1 1.5 2 2.5 3 3.5 4 1D Surrogate 2D Surrogate Stepwise - 30s Psuedo-Transient 30s

RMSE(K)

Total Accuracy

Methods Approach Results Conclusions

slide-31
SLIDE 31

31

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18

1 D S u r r

  • g

a t e 2 D S u r r

  • g

a t e S t e p w i s e 3 s C H T S

  • l

v e T i m e P s u e d

  • T

r a n s i e n t 3 s

CPU-hr/sec. duration

Cycle Compute Cost

Cycle 1 Cycle 2 Cycle 3

Methods Approach Results Conclusions

0.002 0.111 0.004 0.281 0.006 0.000 0.050 0.100 0.150 0.200 0.250 0.300

1 D S u r r

  • g

a t e 2 D S u r r

  • g

a t e S t e p w i s e 3 s C H T S

  • l

v e T i m e P s u e d

  • T

r a n s i e n t 3 s

CPU-hr/sec. duration

Total Compute Cost

slide-32
SLIDE 32

32

Methods Approach Results Conclusions

0.5 1 1.5 2 2.5 3 3.5 4 4.5 0.0001 0.001 0.01 0.1 1

RMSE (K)

CPU-hr/sec 1D Surrogate Stepwise 30s 2D Surrogate Psuedo-Transient 30s CHT Solve Time

slide-33
SLIDE 33

33

Methods Approach Results Conclusions

0.5 1 1.5 2 2.5 3 3.5 4 0.0001 0.001 0.01 0.1 1

RMSE (K) CPU-hr/sec

1D Surrogate Stepwise 30s 2D Surrogate Psuedo-Transient 30s CHT Solve Time

Cycle 1

slide-34
SLIDE 34

34

Cycle 3

6 node average

Methods Approach Results Conclusions

Psuedo Transient Prediction – Temperature

300 320 340 360 380 400 420 500 1000 1500 2000 2500 3000 3500

Temperature (K) Time (s)

CHT Psuedo-Transient 30s Psuedo-Transient 60s Psuedo-Transient 120s Psuedo-Transient Manual

slide-35
SLIDE 35

35

Methods Approach Results Conclusions

2.40 2.45 2.50 2.55 2.60 Psuedo-Transient 30s Psuedo-Transient 60s Psuedo-Transient 120s Psuedo-Transient Manual

RMSE (K)

Manifold

Cycle 3

6 node average

2.70 2.72 2.74 2.76 2.78 2.80 2.82 2.84 2.86 2.88 2.90 Psuedo-Transient 30s Psuedo-Transient 60s Psuedo-Transient 120s Psuedo-Transient Manual

RMSE (K)

Top Wall

Cycle 3 2.6 2.62 2.64 2.66 2.68 2.7 Psuedo-Transient 30s Psuedo-Transient 60s Psuedo-Transient 120s Psuedo-Transient Manual

RMSE(K)

Total Accuracy

55 node average

slide-36
SLIDE 36

36

Methods Approach Results Conclusions

2.645 2.65 2.655 2.66 2.665 2.67 2.675 2.68 20 40 60 80 100 120 140

RMSE (K) Number of Coupling Points

Cycle 2

Psuedo-Transient 30s Psuedo-Transient 60s Psuedo-Transient 120s Psuedo-Transient Manual

slide-37
SLIDE 37

37

Methods Approach Results Conclusions

5000 10000 15000 20000 25000 1D Surrogate Stepwise 30s Psuedo-Transient 30 sec Psuedo-Transient 60 sec Psuedo-Transient 120 sec Psuedo-Transient Manual Selection

Time (s)

Total Solve Time Sample Point Computation Time Thermal Simulation Time CFD Run Time

5 10 15 20 25 30 35 40 1000 2000 3000 4000 Speed (m/s) Time (s)

Cycle 3

slide-38
SLIDE 38
  • Significant cost differences between the methods
  • Step-wise and Psuedo-Transient coupling offers a good balance of accuracy and run time
  • Finding the number of coupling points that balance accuracy and computational costs is important
  • Surrogate models offer significant savings, but sacrifice accuracy
  • Further Research
  • Model sizes
  • Time Stepping
  • Other coupling methods
  • Sampling method for surrogate models
  • Surrogate model interpolation methods

38

Methods Approach Results Conclusions

slide-39
SLIDE 39

Find your FIT

39 Speed (m/s)

Method 1 Method 2 Method 3

FIT

techsupport@thermoanalytics.com

slide-40
SLIDE 40

Questions?

Thank you for your attention

slide-41
SLIDE 41

Technical Support

  • https://support.thermoanalytics.com
  • Submit & Check Status of Requests
  • techsupport@thermoanalytics.com
  • Secure Large File Uploads
  • Software Downloads
  • Technical Library
  • Webinar Videos
  • FAQs
  • Papers & Presentations
  • Spreadsheet Tools
  • Training Videos
  • Feature Requests

41

slide-42
SLIDE 42

1.

Disch, M., Widdecke, N., Wiedemann, J., Reister, H. et al., "Numerical Simulation of the Transient Heat-Up

  • f a Passenger Vehicle during a Trailer Towing Uphill Drive," SAE Technical Paper 2013-01-0873, 2013

2.

Kaushik, S., "Thermal Management of a Vehicle's Underhood and Underbody Using Appropriate Math- Based Analytical Tools and Methodologies," SAE Technical Paper 2007-01-1395, 2007

3.

Pryor, J., Pierce, M., Fremond, E., and Michou, Y., "Development of Transient Simulation Methodologies for Underhood Hot Spot Analysis of a Truck," SAE Technical Paper 2011-01-0651, 2011

4.

Haehndel, K., Pere, A., Frank, T., Christel, F. et al., "A Numerical Investigation of Dampening Dynamic Profiles for the Application in Transient Vehicle Thermal Management Simulations," SAE Technical Paper 2014-01-0642, 2014

42

Thanks and References