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.
Time dependant analysis in solid mechanics
Posted 14 mar 2011, 13:27 GMT-4 Structural Mechanics 3 Replies
Please login with a confirmed email address before reporting spam
Hi everyone,
I am a new user of comsol, currently exploring the software capabilities with a trial license.
I am analysing a 300 mm x 300 mm x 3 mm aluminum plate with an impulse load at the neutral axis. ( the impulse load is 1 micro second in duration, triangular and has a peak value of 1000 N).
the aim is to get the stress waveform on the surface of the plate ....
i am using time dependant study, and i seem to be having some problems getting the correct waveforms.
Could someone provide any suggestions,
Thank you in advance
I am a new user of comsol, currently exploring the software capabilities with a trial license.
I am analysing a 300 mm x 300 mm x 3 mm aluminum plate with an impulse load at the neutral axis. ( the impulse load is 1 micro second in duration, triangular and has a peak value of 1000 N).
the aim is to get the stress waveform on the surface of the plate ....
i am using time dependant study, and i seem to be having some problems getting the correct waveforms.
Could someone provide any suggestions,
Thank you in advance
3 Replies Last Post 15 mar 2011, 11:16 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
check the forum with a search, there are several discussions about similar subjects.
I suppose its related to the rapid pulse: first be sure you define a smooth pulse (at least derivable once, of not twice, i.e. use a Gaussian or the pulse step)
Then be sure you give the solver enough time steps during the impulse to allow it to follow , and set the solver to intermediate stepping an not automatic (automatic is OK for decay or asymptotic solutions, if you have periodical or short pulses, you should rather sue strict (your time steps) or intermediate (at least one inbetween yours)
Another point: if you apply a pulse load, set it over an boundary (area in 3D, length>0 in 2D) and not only on a point, that is a singularity for the FEM and would give you rather false results
--
Good luck
Ivar
check the forum with a search, there are several discussions about similar subjects.
I suppose its related to the rapid pulse: first be sure you define a smooth pulse (at least derivable once, of not twice, i.e. use a Gaussian or the pulse step)
Then be sure you give the solver enough time steps during the impulse to allow it to follow , and set the solver to intermediate stepping an not automatic (automatic is OK for decay or asymptotic solutions, if you have periodical or short pulses, you should rather sue strict (your time steps) or intermediate (at least one inbetween yours)
Another point: if you apply a pulse load, set it over an boundary (area in 3D, length>0 in 2D) and not only on a point, that is a singularity for the FEM and would give you rather false results
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
thank you Ivar,
while I have done everything you have said about the time steps, I didn't give much thought about the shape of the function or applying the load over an area....
I can't say I don't get any results,
The result I get deviates a lot from the experimental waveforms and the results from other software simulations....
I am trying to decide which FEM software is more suited for the research
it is going to involve time varying loads and extracting the stress waveforms at certain locations of the specimen.....
I will re-run the simulation with the suggested changes......
thanks again
while I have done everything you have said about the time steps, I didn't give much thought about the shape of the function or applying the load over an area....
I can't say I don't get any results,
The result I get deviates a lot from the experimental waveforms and the results from other software simulations....
I am trying to decide which FEM software is more suited for the research
it is going to involve time varying loads and extracting the stress waveforms at certain locations of the specimen.....
I will re-run the simulation with the suggested changes......
thanks again
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
COMSOL mostly calculates correctly, so either you might have missed something in the model build up and the BC definitions, or it's the other tools that are wrong, and oversimplifies things ;) Who knows ?
Your local COMSOL rep should also be interested in showing that COMSOL is OK, have you tried to ask him ?
--
Good luck
Ivar
COMSOL mostly calculates correctly, so either you might have missed something in the model build up and the BC definitions, or it's the other tools that are wrong, and oversimplifies things ;) Who knows ?
Your local COMSOL rep should also be interested in showing that COMSOL is OK, have you tried to ask him ?
--
Good luck
Ivar