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.

Random vibration for a structural shell -contact problem

Please login with a confirmed email address before reporting spam

Dear people

I would like some advice on how to approach a random vibration issue related to the ROM (reduced order model) calculation process. I am following the example with the bracket model for random vibrations as a guiding step and I apply it to my model which has shell type physics. Concisely the steps followed, six new nodes, new values for forces, points set for measurement in Def\Variables, calculate the ROM for pre-stress Eigen (contact type prorblem), Global\ Set interpolation-piecewise, assign parameters to rand vibration node. and finally I update the solution for ROM. When I go to results to get from a global evaluation sweep the PSD of the response at a point (specified at variables) it gives a message that there is no such expression (rvib1.psd). I took it a step further by loading the model in Matlab and running the study there. Where with eval I ask it to provide data related to response for all axes. The matrices return empty. The information I find for what to consider related to the ROM solving process are limited so a direction where to look for or some advice will be appreciated.

Thank you, Theodore


2 Replies Last Post Mar 10, 2020, 10:21 a.m. EDT
Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Mar 9, 2020, 11:34 a.m. EDT

The key here is that rvib1.psd cannot be identified. Check
1. That the tag of the node where you reference the PSD actually is 'rvib1'
2. That you are using the right dataset

-------------------
Henrik Sönnerlind
COMSOL
The key here is that rvib1.psd cannot be identified. Check 1. That the tag of the node where you reference the PSD actually is 'rvib1' 2. That you are using the right dataset

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Mar 10, 2020, 10:21 a.m. EDT
Updated: 4 years ago Mar 10, 2020, 10:22 a.m. EDT

Hej Henrik,

Tack för att svara varje fråga jag har.

Before and while I started implementing the model I checked all messages in the forum I could find. The things you mentioned above seem to be common mistakes. In my case the second puzzled me a little bit since Comsol's example indicates to take as database the first solution from ROM and not the second one. A main difference is that instead of calculating the eigenfrequencies and using that as training study I use my prestressed results since it is a contact problem. A hint I was given by you on what to calculate. I follow in parallel the random vibration example for the bracket but I can't get the psd of the vibrations at a point. I am interested in extracting the DOFs over the hole structure, something that ROM should provide.

I have also a question about ROF (rain on the roof) excitation and if this is covered by just the fact that I use broadband white noise and , maybe, by using the load group node for a series of forces. I am interested in the force correlation data.

Vänlig hälsning, Theodore

Hej Henrik, Tack för att svara varje fråga jag har. Before and while I started implementing the model I checked all messages in the forum I could find. The things you mentioned above seem to be common mistakes. In my case the second puzzled me a little bit since Comsol's example indicates to take as database the first solution from ROM and not the second one. A main difference is that instead of calculating the eigenfrequencies and using that as training study I use my prestressed results since it is a contact problem. A hint I was given by you on what to calculate. I follow in parallel the random vibration example for the bracket but I can't get the psd of the vibrations at a point. I am interested in extracting the DOFs over the hole structure, something that ROM should provide. I have also a question about ROF (rain on the roof) excitation and if this is covered by just the fact that I use broadband white noise and , maybe, by using the load group node for a series of forces. I am interested in the force correlation data. Vänlig hälsning, Theodore

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.