Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
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.
Structural optimisation
Posted 7 set 2013, 13:00 GMT-4 Interfacing, Studies & Solvers, Structural Mechanics Version 4.3b 1 Reply
Please login with a confirmed email address before reporting spam
Hi,
I am using a livelink from Solidworks to Comsol in order to import 3D geometry and quickly define some geometry parameters.
I've made a structural analysis with applying torque load to my model and now I would like to use the optimisation tools to lower the total mass and remain the desired stifness (this probably means the control variable should be maximum stress). The model is actually a stator that is loaded with torque end should be optimisated to weight, therefore the inner diameter and the stick depth are the variables, needed for optimisation. My goal is to do a sensitivity analysis using Comsol optimization modul, however I need to go step by step and would like to make first an optimization using the study node and then an additional one using optimization modul (via Add physics) and then finally make the sens. analysis.
As noted already I defined the parameters via solidworks and am not sure if this is the proper approach since these two paramters should be variable and they are defined in the glbal definition under Pi. Would it be wiser to set these two variable parameters manually?
Using optimisation (under the study node) I managed to define the boundary parameters and the objective function equations for the model's mass. I also defined I want a minimization type and made a sum of both objective functions.
When I compute the analysis i get and error in multiphysics compilation. I believe this could be solved using other mesh types, but my problem can also be seen in solidworks, where the model is updated and does not look as it was expected and there is also an error visible in comsol livelink (Updating parameters failed.). The lower and upper boundaries have been exceeeded as it seems and 3/4 of the model are missing.
If anyone has an idea how to solve this please provide me with some guidance. I am pretty much new with Comosl and have searched the commutnity, guides, tutorials, conference proceedings but haven't really found some similair cases.
I also tried to attach the model, but it seems there are some size regulations, as I wasn't able to upload them...
Kind regards,
Matej
I am using a livelink from Solidworks to Comsol in order to import 3D geometry and quickly define some geometry parameters.
I've made a structural analysis with applying torque load to my model and now I would like to use the optimisation tools to lower the total mass and remain the desired stifness (this probably means the control variable should be maximum stress). The model is actually a stator that is loaded with torque end should be optimisated to weight, therefore the inner diameter and the stick depth are the variables, needed for optimisation. My goal is to do a sensitivity analysis using Comsol optimization modul, however I need to go step by step and would like to make first an optimization using the study node and then an additional one using optimization modul (via Add physics) and then finally make the sens. analysis.
As noted already I defined the parameters via solidworks and am not sure if this is the proper approach since these two paramters should be variable and they are defined in the glbal definition under Pi. Would it be wiser to set these two variable parameters manually?
Using optimisation (under the study node) I managed to define the boundary parameters and the objective function equations for the model's mass. I also defined I want a minimization type and made a sum of both objective functions.
When I compute the analysis i get and error in multiphysics compilation. I believe this could be solved using other mesh types, but my problem can also be seen in solidworks, where the model is updated and does not look as it was expected and there is also an error visible in comsol livelink (Updating parameters failed.). The lower and upper boundaries have been exceeeded as it seems and 3/4 of the model are missing.
If anyone has an idea how to solve this please provide me with some guidance. I am pretty much new with Comosl and have searched the commutnity, guides, tutorials, conference proceedings but haven't really found some similair cases.
I also tried to attach the model, but it seems there are some size regulations, as I wasn't able to upload them...
Kind regards,
Matej
Attachments:
1 Reply Last Post 12 set 2013, 13:22 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
I was able to determine the cause of the fault. I managed to find an older post from Ivar, expalining that it is wise to make a parameter sweep before doing the actual optimisation in order to see with which boundary conditions the system will converge...
In addition I have made a simple optimisation, but was not yet able to add a proper constraint to my analysis.
The aim of my analysis is to reduce the model mass, while having the displacement inside my desired values. The first part is working fine, but as I am not able to input a constraint the analysis finishes with the minimum model mass and this reduction in material is not a function of stress or dispalcement, which is my final goal.
Attached is a picture of my design tree.
I am sure there is prety simple solution to this, just I am not able to determine it. Where and what to input to get a constraint in deformation.
Please provide me with some information...
--
Kind regards,
Matej
I was able to determine the cause of the fault. I managed to find an older post from Ivar, expalining that it is wise to make a parameter sweep before doing the actual optimisation in order to see with which boundary conditions the system will converge...
In addition I have made a simple optimisation, but was not yet able to add a proper constraint to my analysis.
The aim of my analysis is to reduce the model mass, while having the displacement inside my desired values. The first part is working fine, but as I am not able to input a constraint the analysis finishes with the minimum model mass and this reduction in material is not a function of stress or dispalcement, which is my final goal.
Attached is a picture of my design tree.
I am sure there is prety simple solution to this, just I am not able to determine it. Where and what to input to get a constraint in deformation.
Please provide me with some information...
--
Kind regards,
Matej
Attachments: