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.

Frequency Response from Modes

Please login with a confirmed email address before reporting spam

Hello everybody,

I am struggling to perform a simple frequency response analysis using modal reduction for a 3-D solid structure. Usually the first step is the solution of eigenfrequencies, the second one a frequency response using the determined modes. How can I do this is Comsol? It's pretty easy in NASTRAN, but what are the parameters, the solution sequence or initial parameters in Comsol?

3 Replies Last Post 7 juil. 2017, 02:44 UTC−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 18 août 2010, 11:13 UTC−4
Hi

first of all 3.5 or 4 ?

In 3.5 you can do an eigensolver then you can export the desired modes to matlab either as matrices or Silmulink I/O, but its different from Nastran, check the doc its poor (no-fwe examples) but in 3.5a you have a short chapter there.

By the way COMSOL does not normalise eigenmodes as Nastran/Ansys/COSMOS ... so the solution provided by COMSOL must be exported to matlab to be renomalise to get the mass particiaption factors out (see other discussion and script code on the Forum), such to sort out how many modes to select. Furthermore the mode export feature does only allow to select a contiguous number of modes.

In 4 I have not tested all this yet, not sure if its already fully in there.

These mode reduction issues and eigenmode mass factors are the (main) things I lack in COMSOL Structural, as the exporting to matlab only works for simple models, any "normal" engineering model is so heavy that the solution matrices saturates the RAm during the transfer to matlab

--
Good luck
Ivar
Hi first of all 3.5 or 4 ? In 3.5 you can do an eigensolver then you can export the desired modes to matlab either as matrices or Silmulink I/O, but its different from Nastran, check the doc its poor (no-fwe examples) but in 3.5a you have a short chapter there. By the way COMSOL does not normalise eigenmodes as Nastran/Ansys/COSMOS ... so the solution provided by COMSOL must be exported to matlab to be renomalise to get the mass particiaption factors out (see other discussion and script code on the Forum), such to sort out how many modes to select. Furthermore the mode export feature does only allow to select a contiguous number of modes. In 4 I have not tested all this yet, not sure if its already fully in there. These mode reduction issues and eigenmode mass factors are the (main) things I lack in COMSOL Structural, as the exporting to matlab only works for simple models, any "normal" engineering model is so heavy that the solution matrices saturates the RAm during the transfer to matlab -- Good luck Ivar

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: 7 years ago 6 juil. 2017, 09:52 UTC−4
Hi all

Lets update this one :) :

New in 5.3 all "mass participation factors" are now implemented, even the rotational ones but in "beta demo" mode, and including added mass components and so.

And a solver State Space Modal Reductions nodes exist (already for several releases) to allow to export the (A,B,C,D) matrices, still I do not know of any Blogs or Application Library illustrating this for a simple example, but there are so many models out floating there, so there might be one !

Finally, the few missing items in Solid Physics, I needed, compared to the older FEM codes, are all here. These items are mandatory to allow me to show my clients the results as they are used too see them, this is essential too, even if the previous COMSOL results were just as correct.

--
Good luck
Ivar
Hi all Lets update this one :) : New in 5.3 all "mass participation factors" are now implemented, even the rotational ones but in "beta demo" mode, and including added mass components and so. And a solver State Space Modal Reductions nodes exist (already for several releases) to allow to export the (A,B,C,D) matrices, still I do not know of any Blogs or Application Library illustrating this for a simple example, but there are so many models out floating there, so there might be one ! Finally, the few missing items in Solid Physics, I needed, compared to the older FEM codes, are all here. These items are mandatory to allow me to show my clients the results as they are used too see them, this is essential too, even if the previous COMSOL results were just as correct. -- Good luck Ivar

Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 7 years ago 7 juil. 2017, 02:44 UTC−4
Updated: 7 years ago 7 juil. 2017, 03:34 UTC−4
Hi Andreas,

Even though Ivar's comments contains a lot of details insight, I interpret the question in more simple way:

"How to perform a frequency domain analysis by mode superposition"

This is quite simple:

1. You add a study of the type Frequency-Domain Modal. It will contain one Eigenfrequency study step and one Frequency-Domain Modal study step.
2. You add your loads, and make sure that they are defined as "Harmonic Perturbation" (Right-click on the load in the Model Builder tree).
3. In the settings for the Eigenfrequency study step, you select which frequencies to solve for.
4. In the settings for the Frequency-Domain Modal study step, you select the range of the frequency sweep.
5. Run.

The tutorial in www.comsol.com/model/various-analyses-of-an-elbow-bracket-4131 shows how to set up most types of structural analyses.

Regards,
Henrik
Hi Andreas, Even though Ivar's comments contains a lot of details insight, I interpret the question in more simple way: "How to perform a frequency domain analysis by mode superposition" This is quite simple: 1. You add a study of the type Frequency-Domain Modal. It will contain one Eigenfrequency study step and one Frequency-Domain Modal study step. 2. You add your loads, and make sure that they are defined as "Harmonic Perturbation" (Right-click on the load in the Model Builder tree). 3. In the settings for the Eigenfrequency study step, you select which frequencies to solve for. 4. In the settings for the Frequency-Domain Modal study step, you select the range of the frequency sweep. 5. Run. The tutorial in https://www.comsol.com/model/various-analyses-of-an-elbow-bracket-4131 shows how to set up most types of structural analyses. Regards, Henrik

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.