1 of 19

Optimization of periodic fins using streamwise periodic flow solver with iso-thermal boundaries in SU2

Nitish Anand, Carlo de Servi

Energy Technology Unit

Thermal Energy Systems Group

​

​

This research was funded by Vlaanderen Agentschap Innoveren and Ondernemen (VLAIO), Belgium, through the project IAMHEX (grant number HBC.2021.0801).

vito.be

©VITO – Not for distribution

2 of 19

Motivation & Background

1

©VITO – Not for distribution

3 of 19

Why a periodic flow solver?

Reduced-order Modeling

vito.be

©VITO – Not for distribution

4 of 19

Boundary Condition: Iso-thermal Wall

 

Suitable temperature scaling

Energy Equation

 

 

 

 

 

Temperature periodicity

 

 

vito.be

©VITO – Not for distribution

5 of 19

Streamwise Periodic Energy Equation for Isothermal Walls

Implementation, Verification & Optimization

2

©VITO – Not for distribution

6 of 19

Implementation in SU2

feature_isothermalSWP

su2code/SU2

Draft #1799

Flow Source Terms: Tobias Kattmann

Extended: Additional Energy equation source terms

2D/3D generalization

vito.be

©VITO – Not for distribution

7 of 19

Iso-thermal Solver: Temperature profile

Laminar

Turbulent

vito.be

©VITO – Not for distribution

8 of 19

Verification of temperature profiles

vito.be

©VITO – Not for distribution

9 of 19

Gradient Validation

Laminar

Turbulent

Not for external distribution

FD: Finite Difference, ADJ: Adjoint

vito.be

©VITO – Not for distribution

10 of 19

CAD-Based Shape Optimization

Laminar

Turbulent

+5%

(feasible design)

+6%

(infeasible design)

 

 

 

 

Shortcomings !

vito.be

©VITO – Not for distribution

11 of 19

FFD-Based Shape Optimization

Laminar

Turbulent

 

 

 

vito.be

©VITO – Not for distribution

12 of 19

Topology Optimization of HEX Manifold

3

©VITO – Not for distribution

13 of 19

Upgraded Legacy Code

feature_porous_media_topopt

su2code/SU2

Not yet

Original Implementation: Dr. Thomas D. Economon

Extended: Objective Function

Multi-zone communication

Python optimization framework

vito.be

©VITO – Not for distribution

14 of 19

Objective Functions: Power Dissipation

Simulation Courtesy: Sanjay Vermani, Researcher, VITO.

Iso-surface porosity 0.5

vito.be

©VITO – Not for distribution

15 of 19

Objective Functions: Maldistribution Function

Simulation Courtesy: Sanjay Vermani, Researcher, VITO.

Iso-surface porosity 0.5

vito.be

©VITO – Not for distribution

16 of 19

Objective Functions: ED+MDF

Simulation Courtesy: Sanjay Vermani, Researcher, VITO.

Combined Objective: Inlet manifold only

Iso-surface porosity 0.5

vito.be

©VITO – Not for distribution

17 of 19

Conclusions & �Way Forward

4

©VITO – Not for distribution

18 of 19

Conclusions and Perspective

  • Verification of 2D/3D iso-thermal SWP 🡪 Completed

​

  • Tuning the optimization problem to suit the test case

​

  • Setting up high-fidelity flow solver for verification study in turbulent regime 🡪 Long term goal

​

  • Increase robustness of fluid topology optimization problem 🡪 ongoing

​

LES - Teaser

vito.be

©VITO – Not for distribution

19 of 19

Thank you!

Dr.ir. Nitish Anand

nitish.anand@vito.be

Dr.ir. Carlo de Servi

carlo.deservi@vito.be

This research was funded by Vlaanderen Agentschap Innoveren and Ondernemen (VLAIO), Belgium, through the project IAMHEX (grant number HBC.2021.0801).

vito.be

©VITO – Not for distribution