Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Geometric optimization v4.1

Please login with a confirmed email address before reporting spam

Hello,

I would like to find the optimum geometry of my structure in v4.1, but I have some problems.

I have - as a simple example - a beam in a 3D model. The thickness and the length of the beam are global parameters. I used the "Piezoelectric Devices interface" to clamp the two side of the beam and to apply a pressure on the upper boundary. I would like to find the optimum thickness (of a defined interval) to get a maximal deflection of the beam. I can do it with a parametric sweep, but I would like to use the optimization module as the model will get more complicated.

I set the point objective as the deflection and the global control variable as the thickness. COMSOL returnes the initial value of control variable as result. The geometry did not change. Do I have to define the parameter for thickness in a different way?

Has anyone examples for geometry optimization?

Thank you very much!



2 Replies Last Post 02.05.2011, 12:13 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.04.2011, 15:17 GMT-4
Hi

I haven't checked if the 3.5 models are all ported yet to v4. But there were some shape optimisation examples in 3.5

Now for the geometry, I believe the "DG" "deformed Geometry" is better suited than a full parametric sweep over a parameter changing the geometry dimensions (even if that should work too).

If you want to see the change on your geometry you must loop through the geometry + remesh, and that implies an (external or full) Parametric Sweep node before your solver node.

Therefore I believe the "DG" physics is easier, it makes a "mesh" change of your geometry, so you might use easier the "optimisation" module with a (solver internal) "Continuation (parametric) sweep"

Unfortunately, I do not have any examples at hand, you will have to search the forum, and the conf papers

--
Good luck
Ivar
Hi I haven't checked if the 3.5 models are all ported yet to v4. But there were some shape optimisation examples in 3.5 Now for the geometry, I believe the "DG" "deformed Geometry" is better suited than a full parametric sweep over a parameter changing the geometry dimensions (even if that should work too). If you want to see the change on your geometry you must loop through the geometry + remesh, and that implies an (external or full) Parametric Sweep node before your solver node. Therefore I believe the "DG" physics is easier, it makes a "mesh" change of your geometry, so you might use easier the "optimisation" module with a (solver internal) "Continuation (parametric) sweep" Unfortunately, I do not have any examples at hand, you will have to search the forum, and the conf papers -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 02.05.2011, 12:13 GMT-4
Thank you very much for your fast reply!

I have tried the deformed geometry in the optimization module now.
The geometry changed and I get a value for the objective function which is different from the initial value. So I guess it's the optimized result.
But where can I find the corresponding value for the control variable? I only get the initial value.




Thank you very much for your fast reply! I have tried the deformed geometry in the optimization module now. The geometry changed and I get a value for the objective function which is different from the initial value. So I guess it's the optimized result. But where can I find the corresponding value for the control variable? I only get the initial value.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.