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:
9 years ago
11 mag 2016, 01:00 GMT-4
Hi
Well I would rather use a stationary study to get absolute stress values, an eigenfrequency analysis is an "infinite energy analysis" and the resulting displacements, hence strain, and from there stresses are depending on the normalization you choose.e
Eigenfrequencies are ideal to analyze relative shapes, and the natural "eigenmodes" of a structure, but not to be used for any absolute value analysis, for that a stationary or time dependent load should be considered. Or an frequency sweep, but then with some realistic damping to have a system with limited global energy
--
Good luck
Ivar
Hi
Well I would rather use a stationary study to get absolute stress values, an eigenfrequency analysis is an "infinite energy analysis" and the resulting displacements, hence strain, and from there stresses are depending on the normalization you choose.e
Eigenfrequencies are ideal to analyze relative shapes, and the natural "eigenmodes" of a structure, but not to be used for any absolute value analysis, for that a stationary or time dependent load should be considered. Or an frequency sweep, but then with some realistic damping to have a system with limited global energy
--
Good luck
Ivar
Henrik Sönnerlind
COMSOL Employee
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
11 mag 2016, 01:51 GMT-4
Hi,
I completely agree with Ivar in that modal stresses have a very limited usability.
But if we turn to the question about how to evaluate stresses in local directions, you should create a coordinate system aligned with the boundary, and then select it in the Linear Elastic node. The local stress results will then be oriented according to this system.
Regards,
Henrik
Hi,
I completely agree with Ivar in that modal stresses have a very limited usability.
But if we turn to the question about how to evaluate stresses in local directions, you should create a coordinate system aligned with the boundary, and then select it in the Linear Elastic node. The local stress results will then be oriented according to this system.
Regards,
Henrik
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
11 mag 2016, 12:56 GMT-4
Thank you for the replies.
I am not really worried by the absolute values, since I would also require the stored energy (J) and put the stresses together in an analytical expression to quantify my required parameter. I guess in the eigenfreqeuncy analysis, the normalization would be same for the stresses and stored energy. My aim is to get the stresses (normal & shear) at the clamped end, and the energy stored in the structure at its resonant mode. These are the two parameters I need to calculate. Energy stored is easy to evaluate from COMSOL. However, the normal and shear stresses , I am finding a bit difficult to interpret from the results.
@Henrik, are suggesting I should create a coordinate system aligned to that edge, and then resolve the problem using that coordinate system? One question though, for example if I evaluate the stresses at the edge (i.e. line average) sx,sy and sxy, are these by default based on the global coordinate system?
This may be a simple thing to realize, but I don't know why I am finding it difficult :(
Thank you for the replies.
I am not really worried by the absolute values, since I would also require the stored energy (J) and put the stresses together in an analytical expression to quantify my required parameter. I guess in the eigenfreqeuncy analysis, the normalization would be same for the stresses and stored energy. My aim is to get the stresses (normal & shear) at the clamped end, and the energy stored in the structure at its resonant mode. These are the two parameters I need to calculate. Energy stored is easy to evaluate from COMSOL. However, the normal and shear stresses , I am finding a bit difficult to interpret from the results.
@Henrik, are suggesting I should create a coordinate system aligned to that edge, and then resolve the problem using that coordinate system? One question though, for example if I evaluate the stresses at the edge (i.e. line average) sx,sy and sxy, are these by default based on the global coordinate system?
This may be a simple thing to realize, but I don't know why I am finding it difficult :(
Henrik Sönnerlind
COMSOL Employee
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
12 mag 2016, 01:15 GMT-4
Hi,
You do no need to solve again, it is sufficient to do an 'Update solution' as long as the new coordinate system does not change the solution (i.e. as long as the material is isotropic).
Stress component solid.sx is in the global direction, while solid.Sl11 is in the local direction. See also this thread:
www.comsol.com/community/forums/general/thread/34639
Regards,
Henrik
Hi,
You do no need to solve again, it is sufficient to do an 'Update solution' as long as the new coordinate system does not change the solution (i.e. as long as the material is isotropic).
Stress component solid.sx is in the global direction, while solid.Sl11 is in the local direction. See also this thread:
https://www.comsol.com/community/forums/general/thread/34639
Regards,
Henrik