[flow-mo-mcfadden]

Hi all,

I'm working a flow optimization problem in Ansys Workbench 19.2 and 19.3 using Fluent. In setting my case up, I've chosen to use the "Solution Steering" controls in order to gain convergence in a reasonable amount of time with my limited computing power. I've spot checked several of the design points and had good results.

My optimization problem uses Workbench's Design of Experiments and parameterizes two pressure inputs in the model. Now when I run the DoE, many of my design points fail to converge, even ones that I manually spot-checked and had good results when inputting those values directly into fluent GUI. The only think I found peculiar is that when running the design points through the DoE and monitoring the solution, I do not see the text output associated with the "Solution Steering" controls which should follow the initialization.

Any ideas on what I need to do to make this work. I've tried with similar outcomes on both 19.2 and 19.3.

[flow-mo-mcfadden]

Well let me answer my own question. Wrestled with this for two days and then 30 seconds after giving up and writing a post here, I saw another option.

Initialization Method under Solution Properties in Workbench. Use "Program Controlled" and on "Solver Controlled". Now everything works....I think

[flow-mo-mcfadden]

Ok, I'm in more trouble. In my attempt to use the "Program Controlled" option, I did not test the entire Design of Experiments, rather, I just changed one pressure parameter and ran a couple design points, instead of including parameters associated with the geometry.

Can anyone confirm that the Response Surface Optimization uses the exact same solution process for each design point as specified in the solution cell. That is, I should be able to open Fluent, input the exact values from my design point, and get the exact same output values as the Response Surface DoE but it seems this is not the case. I would like to enforce this solution process behavior but I do not know how.

[antoinel]

Did you resolve this issue? I am experiencing the same problem when I solve in Design of Experiments vs solving in FLUENT. The output parameters are not equal in the slightest.

[flow-mo-mcfadden]

I made several attempts to workaround this issue. My issue was fixed in Fluent2020R1. Hopefully you are not stuck with version 19.

[antoinel]

Thanks Flow for the info. I changed the solution method from program controlled to solver controlled as you suggested and it worked. I was able to simulate the design points in stand alone FLUENT and the solutions were equal. I am stuck with release 19.