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 The information provided may be out of date.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.

parametric sweeps: geometry and frequency sweep

Please login with a confirmed email address before reporting spam

Hi

I'm using comsol 4.1. I have a problem about parametric sweeps

I have a model, eg. just a cube, light is excited on one of the facet of the cube, to simulate the light going through the cube, I want to sweep the frequency of the light. in addition, I also need to change the size of the cube.

for the frequency sweep, we can just specify the frequency range in GUI, then run the study, we can get solutions for different frequency. But for sweep of the size of cube, is there anyone knows how to realize this in the GUI (maybe it is impossible to do so in GUI).

For comsol v3.5a, I used a script containing a loop which change the size parameter of the cube.
But for the V4.1, I still don't know how to write a script.



8 Replies Last Post 2011/05/09 3:47 GMT-4
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 2011/02/10 5:40 GMT-5
H

you have two types of paramtetric sweeps: the "continuation" internal parameter sweep you define in the solver node i.e. stationary, or frequency sweep. And the full Parametric sweep. Th latter loops aroud the geoemtry too, so this one allows you to also update your geoemtry and mesh if required.

note that you might nest the sweeps to, apart that it makes looooong runs ;)

--
Good luck
Ivar
H you have two types of paramtetric sweeps: the "continuation" internal parameter sweep you define in the solver node i.e. stationary, or frequency sweep. And the full Parametric sweep. Th latter loops aroud the geoemtry too, so this one allows you to also update your geoemtry and mesh if required. note that you might nest the sweeps to, apart that it makes looooong runs ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2011/02/14 4:23 GMT-5
Thank you

I know how to do it now
Thank you I know how to do it now

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2011/05/06 13:30 GMT-4

H

you have two types of paramtetric sweeps: the "continuation" internal parameter sweep you define in the solver node i.e. stationary, or frequency sweep. And the full Parametric sweep. Th latter loops aroud the geoemtry too, so this one allows you to also update your geoemtry and mesh if required.

note that you might nest the sweeps to, apart that it makes looooong runs ;)

--
Good luck
Ivar


How to do the Full Parametric Sweep to change the geometry exactly? And how to do 3 parametric sweeps at once for the geometry exactly? This is urgent...please advise
[QUOTE] H you have two types of paramtetric sweeps: the "continuation" internal parameter sweep you define in the solver node i.e. stationary, or frequency sweep. And the full Parametric sweep. Th latter loops aroud the geoemtry too, so this one allows you to also update your geoemtry and mesh if required. note that you might nest the sweeps to, apart that it makes looooong runs ;) -- Good luck Ivar [/QUOTE] How to do the Full Parametric Sweep to change the geometry exactly? And how to do 3 parametric sweeps at once for the geometry exactly? This is urgent...please advise

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 2011/05/06 14:47 GMT-4
Hi

I haven tried more than 2 levels in V4, not sure it can handle 3 levels, you get a lot of data like that !

Perhaps if you get the outer one to save the results as a series of files with 2 levels, it couldwork (for the postprocessing/results)

--
Good luck
Ivar
Hi I haven tried more than 2 levels in V4, not sure it can handle 3 levels, you get a lot of data like that ! Perhaps if you get the outer one to save the results as a series of files with 2 levels, it couldwork (for the postprocessing/results) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2011/05/07 15:22 GMT-4

Hi

I haven tried more than 2 levels in V4, not sure it can handle 3 levels, you get a lot of data like that !

Perhaps if you get the outer one to save the results as a series of files with 2 levels, it couldwork (for the postprocessing/results)

--
Good luck
Ivar


Hello,

I keep getting this error

Parameter list must be strictly monotone.

Although my parameters are both increasing and have the same number of intervals
[QUOTE] Hi I haven tried more than 2 levels in V4, not sure it can handle 3 levels, you get a lot of data like that ! Perhaps if you get the outer one to save the results as a series of files with 2 levels, it couldwork (for the postprocessing/results) -- Good luck Ivar [/QUOTE] Hello, I keep getting this error Parameter list must be strictly monotone. Although my parameters are both increasing and have the same number of intervals

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2011/05/07 15:24 GMT-4
Besides, does a parametric sweep solve each value of the second term for all values of the first term and so on?
Besides, does a parametric sweep solve each value of the second term for all values of the first term and so on?

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 2011/05/08 4:56 GMT-4
Hi

for multi parameter definitions, you must provide all parameter values in a 2xn or 3xn or mxn... list and the first of each pair (triple ...) must be monotonously increasing. Check the doc, and the FORUM there are several threads already on the subject

--
Good luck
Ivar
Hi for multi parameter definitions, you must provide all parameter values in a 2xn or 3xn or mxn... list and the first of each pair (triple ...) must be monotonously increasing. Check the doc, and the FORUM there are several threads already on the subject -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2011/05/09 3:47 GMT-4

Hi

for multi parameter definitions, you must provide all parameter values in a 2xn or 3xn or mxn... list and the first of each pair (triple ...) must be monotonously increasing. Check the doc, and the FORUM there are several threads already on the subject

--
Good luck
Ivar


Hi,Ivar
My problem is the same as the above. The model contain 3 physical field, ElectroicMagnet, Heat transfer, Solid mechanic. The solver include 3 Segregated solvers, each for one field variable. When parametric sweep is conducted, the first two given values can get through successfully, and the RAM consumes about 1.8G of the total 3.25G. And then, it will stop and warn "LU out of memory", what confused me is that the RAM taken by the solver is no more than 2.5G, and there is still free memory,how can this happen? And why the previous given values can get through? I suspect there is something related to the very poor memory management of COMSOL.
[QUOTE] Hi for multi parameter definitions, you must provide all parameter values in a 2xn or 3xn or mxn... list and the first of each pair (triple ...) must be monotonously increasing. Check the doc, and the FORUM there are several threads already on the subject -- Good luck Ivar [/QUOTE] Hi,Ivar My problem is the same as the above. The model contain 3 physical field, ElectroicMagnet, Heat transfer, Solid mechanic. The solver include 3 Segregated solvers, each for one field variable. When parametric sweep is conducted, the first two given values can get through successfully, and the RAM consumes about 1.8G of the total 3.25G. And then, it will stop and warn "LU out of memory", what confused me is that the RAM taken by the solver is no more than 2.5G, and there is still free memory,how can this happen? And why the previous given values can get through? I suspect there is something related to the very poor memory management of COMSOL.

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.