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.

Path-following arc-length solution

Please login with a confirmed email address before reporting spam

Hello,

I need to trace solution of snap-through plate deformation problem using COMSOL Structural Mechanics Module and I want to develop arc-length method for it. Has anybody here experience in doing so in Matlab interface to Comsol?

Thanks
Peter

4 Replies Last Post 6 juil. 2010, 11:47 UTC−4
COMSOL Moderator

Hello Peter Cendula

Your Discussion has gone 30 days without a reply. If you still need help with COMSOL and have an on-subscription license, please visit our Support Center for help.

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 juil. 2010, 08:48 UTC−4
Hello,

Have you managed to solve your problem of path-following ? I would be really interested to get some advice !

Thanks
Hello, Have you managed to solve your problem of path-following ? I would be really interested to get some advice ! Thanks

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 6 juil. 2010, 11:42 UTC−4
Here is one answer from comsol support:


Dear Peter,
True arc length continuation is rather tricky to implement, at least
inside the COMSOL Multiphysics GUI. It may be slightly easier from Matlab.
But if you are looking to model a snap-through behavior where you know that
the displacement of some point is monotonous during the process, it is far
easier to prescribe this displacement for each solution step and solve for
the corresponding applied load.

In the attached model, which simulates snap-through of a quarter cylinder,
the magnitude of the applied load, "f", is a Global Equation variable,
while the corresponding equation specifies that the displacement measured
in the mid point must equal the displacement prescribed by the parametric
solver. Note that this type of models with large displacement often require
manual scaling of the variables (on the Advanced page in Solver Settings)
and converge better if the damped Newton method is turned off (on the
Stationary page, leading to a full Newton method).

Best regards
Nils Malm

==================================================
Nils Malm
COMSOL AB
Tegnérgatan 23
SE-111 40 Stockholm
Here is one answer from comsol support: Dear Peter, True arc length continuation is rather tricky to implement, at least inside the COMSOL Multiphysics GUI. It may be slightly easier from Matlab. But if you are looking to model a snap-through behavior where you know that the displacement of some point is monotonous during the process, it is far easier to prescribe this displacement for each solution step and solve for the corresponding applied load. In the attached model, which simulates snap-through of a quarter cylinder, the magnitude of the applied load, "f", is a Global Equation variable, while the corresponding equation specifies that the displacement measured in the mid point must equal the displacement prescribed by the parametric solver. Note that this type of models with large displacement often require manual scaling of the variables (on the Advanced page in Solver Settings) and converge better if the damped Newton method is turned off (on the Stationary page, leading to a full Newton method). Best regards Nils Malm ================================================== Nils Malm COMSOL AB Tegnérgatan 23 SE-111 40 Stockholm

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 6 juil. 2010, 11:44 UTC−4
And another reply from Support about the access to previous solution in the parametric solver:

Date 03-01-2010:

Dear Peter,

For a transient (or parametric) analysis it is still not possible to
access the previous solution in the current calculation.
However the upcoming version, COMSOL 4.0 will have a new operator that
will make possible for the solver to access the value of a variable at a
previous calculation.

Currently with version 3.5a, it is a rather complicated task to proceed.
Using MATLAB is indeed a solution as MATLAB can use persistent variable
type. The persistent variable are kind of global variable that remain their
value until you reassign them a new value.

You would then need to create a MATLAB function that will store as a
persistent variable the current solution (that can be then used for the
next time step).
The main problem is that COMSOL assemble and solve the problem using
block. This means that the same function can be called several time at the
same time step... and so the persistent variable would be rewritten even
though you are still remaining at the same time step.

This is definitely not an easy task and I would rather recommend you to
wait for 4.0 to be released (end of march/april as expected plan)

Best regards,
______________________________________
Remi Magnard

COMSOL AB
Tegnérg. 23
111 40 Stockholm
Sweden
Tel : +46 8 412 9500
Fax : +46 8 412 9510
www.comsol.com
And another reply from Support about the access to previous solution in the parametric solver: Date 03-01-2010: Dear Peter, For a transient (or parametric) analysis it is still not possible to access the previous solution in the current calculation. However the upcoming version, COMSOL 4.0 will have a new operator that will make possible for the solver to access the value of a variable at a previous calculation. Currently with version 3.5a, it is a rather complicated task to proceed. Using MATLAB is indeed a solution as MATLAB can use persistent variable type. The persistent variable are kind of global variable that remain their value until you reassign them a new value. You would then need to create a MATLAB function that will store as a persistent variable the current solution (that can be then used for the next time step). The main problem is that COMSOL assemble and solve the problem using block. This means that the same function can be called several time at the same time step... and so the persistent variable would be rewritten even though you are still remaining at the same time step. This is definitely not an easy task and I would rather recommend you to wait for 4.0 to be released (end of march/april as expected plan) Best regards, ______________________________________ Remi Magnard COMSOL AB Tegnérg. 23 111 40 Stockholm Sweden Tel : +46 8 412 9500 Fax : +46 8 412 9510 www.comsol.com

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 6 juil. 2010, 11:47 UTC−4
Thanks a lot,

I think i will wait for the 4.0 version ;)

Francois
Thanks a lot, I think i will wait for the 4.0 version ;) Francois

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.