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.

Initial Stress doesn't work????

Please login with a confirmed email address before reporting spam

Dear all,

A 3D simulation of beam's mechanical modes. Beam size is 30um*400nm*240nm. The beam is fixed on both ends. At first no initial stress is applied. In theory that the internal tensile stress will increase the frequancies of the first a few modes. But when I add the inital stress to be 1e9. The frequency doeson't change at all.

Can anyone help me about this? Thanks.

Best,
Jiangjun

5 Replies Last Post 2010年10月18日 GMT-4 09:21
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 2010年10月13日 GMT-4 02:01
Hi

eigenfrequency analysis is a linear analysis, typically linearised around the initial value of by default "0" (also for the stress) so be sure you really apply the stress corrrectly, i.e. by a first stationary analysis and then save the results to restart from there

--
Good luck
Ivar
Hi eigenfrequency analysis is a linear analysis, typically linearised around the initial value of by default "0" (also for the stress) so be sure you really apply the stress corrrectly, i.e. by a first stationary analysis and then save the results to restart from there -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年10月17日 GMT-4 23:24
hi, Ivar,

Thanks for your kind reply. The structure can be considered as a string with length L. If the string is tightly fixed (with internal tensile stress) at both ends. Then its mechanical frequency can be high.

According to your reply, I fixed one end of the beam, and applied a static force on the other end. Then the internal stress and the displacement can be obtained and used as the initial value for the eigenvalue problem. here, I don't know how to make the "initial value" work... I added a second application mode for the eigen value analysis. How should I set the boundary and inital conditions for it? I tried a lot but it seems that these conditions take no effects.

Thanks.

Best,
Jiangjun
hi, Ivar, Thanks for your kind reply. The structure can be considered as a string with length L. If the string is tightly fixed (with internal tensile stress) at both ends. Then its mechanical frequency can be high. According to your reply, I fixed one end of the beam, and applied a static force on the other end. Then the internal stress and the displacement can be obtained and used as the initial value for the eigenvalue problem. here, I don't know how to make the "initial value" work... I added a second application mode for the eigen value analysis. How should I set the boundary and inital conditions for it? I tried a lot but it seems that these conditions take no effects. Thanks. Best, Jiangjun

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年10月18日 GMT-4 00:14
Hi, Ivar,

This time I tried to the thermal expansion as the first step to get the inital stress. It works well:
The beam is fixed on both ends. The temperature drops from 800K to 300K. Which lead to tensile stress in the beam. (It is very clear from the figures that the share stress is 0 and the tensile stress is several GPa, almost homogeneous in the beam.)

But still problems happens when I tried to use these initial info to the eigen value mode which is followed after the thermal expansion step:
the initial stresses are set to be sx_smsld, sy_smsld etc. and in the solver manager, I select to solve the for the smsld2 with the current solution as the initial value.
The eigen values are still the same as those without tensile stress.

BTW, I am using the comsol3.5a. Any futher suggestions? Thanks in advance.

Best,
Jiangjun
Hi, Ivar, This time I tried to the thermal expansion as the first step to get the inital stress. It works well: The beam is fixed on both ends. The temperature drops from 800K to 300K. Which lead to tensile stress in the beam. (It is very clear from the figures that the share stress is 0 and the tensile stress is several GPa, almost homogeneous in the beam.) But still problems happens when I tried to use these initial info to the eigen value mode which is followed after the thermal expansion step: the initial stresses are set to be sx_smsld, sy_smsld etc. and in the solver manager, I select to solve the for the smsld2 with the current solution as the initial value. The eigen values are still the same as those without tensile stress. BTW, I am using the comsol3.5a. Any futher suggestions? Thanks in advance. Best, Jiangjun

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 2010年10月18日 GMT-4 02:19
Hi

in V3.5 you must go through matlab to get the initial stress loaded into the eigenfrequency model, in V4 its OK (without matlab). there is an example with a string under variable tension (violin cord) somewhere

--
Good luck
Ivar
Hi in V3.5 you must go through matlab to get the initial stress loaded into the eigenfrequency model, in V4 its OK (without matlab). there is an example with a string under variable tension (violin cord) somewhere -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年10月18日 GMT-4 09:21
Hi, Ivar,

Thanks for your kind reply. It is very helpful to me.

I am now trying to do what you have said. Another thing is if I turn on the "large deformation" in "application mode properties". Seems that the eigenfrequence can change with the intitial stress. But is this correct?

Thanks.

Best,
jiangjun
Hi, Ivar, Thanks for your kind reply. It is very helpful to me. I am now trying to do what you have said. Another thing is if I turn on the "large deformation" in "application mode properties". Seems that the eigenfrequence can change with the intitial stress. But is this correct? Thanks. Best, jiangjun

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.