Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

lumped versus segregated step

Please login with a confirmed email address before reporting spam

Hi Dear Supports,

I have a sphere in contact with a film. I am using axial symetry and my sphere is pushing against the film in presence of a boundary load. I am using weak constrains to calculate reaction froces on the sphere. Now sth happens in solver:
the mode1.u and mod1.u_lm are in segregated step as it shows the default solver. However, the contact pressure, mod1.Solid.Tn_p1 is in lumped step it fails to converge but when I put it in segregated group 2, by defining a new one, it not only works but the speed of simulation is 5 times faster. Please let me know what happens?

Apart from that, when should I use lower limit step?

--
Marjan

2 Replies Last Post 2012年6月15日 GMT-4 12:42
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 2012年6月14日 GMT-4 15:00
Hi

First of all:
here onthe User Forum, we are all USERS, NOT SUPPORT (to reach the official "support" check COMSOLs support mail address But we try to help each other at best, so welcome to the club ...

I believe your description can come from a few issues: the model has been updated and changed so many times that COMSOL is not manageing to reset the sover sequence fully, and does not propose an optimal sequence, or just that for your model the sequence you tried is much more efficient, while for another model il might not work out at all ...
Perhaps your model would even solve quicker in a one step direct mode ;)

Sometimes, it's worth to reconstruct the model from scartch, after many try&fail steps or regenerate fully either the present solver sequence or another new one.
Important note: if you delete the solver sequence then all your linked postprocessing nodes are also DELETED which might represent a few hours work, so take care, save and think twice;)

--
Good luck
Ivar
Hi First of all: here onthe User Forum, we are all USERS, NOT SUPPORT (to reach the official "support" check COMSOLs support mail address But we try to help each other at best, so welcome to the club ... I believe your description can come from a few issues: the model has been updated and changed so many times that COMSOL is not manageing to reset the sover sequence fully, and does not propose an optimal sequence, or just that for your model the sequence you tried is much more efficient, while for another model il might not work out at all ... Perhaps your model would even solve quicker in a one step direct mode ;) Sometimes, it's worth to reconstruct the model from scartch, after many try&fail steps or regenerate fully either the present solver sequence or another new one. Important note: if you delete the solver sequence then all your linked postprocessing nodes are also DELETED which might represent a few hours work, so take care, save and think twice;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年6月15日 GMT-4 12:42
Hi Dear Ivar,

First of all thanks for mentioning how everything works here. Also thanks for welcoming me but actually it is for a while that I am in club ;).

Yes, I got the point. The model started from comsol 4 and is developing until now (it is almost 1 year and half). So it is really a complex model with 2 different physics and everytime I save a new one and clear solutions and reset history and run for new conditions :). Building it up from zero is a nightmare.

I decided to make a new study to check sth and I figured out this issue plus some more. The current study node was modified by comsol support and is indeed perfect. I just would like to learn more about it.

All the best,
Marjan

Hi Dear Ivar, First of all thanks for mentioning how everything works here. Also thanks for welcoming me but actually it is for a while that I am in club ;). Yes, I got the point. The model started from comsol 4 and is developing until now (it is almost 1 year and half). So it is really a complex model with 2 different physics and everytime I save a new one and clear solutions and reset history and run for new conditions :). Building it up from zero is a nightmare. I decided to make a new study to check sth and I figured out this issue plus some more. The current study node was modified by comsol support and is indeed perfect. I just would like to learn more about it. All the best, Marjan

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.