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
31 dic 2009, 09:22 GMT-5
Hi
You do not have a written out "while" nor an "if", but you have almost the same via a bolean expression, see a couple of older discussion:
search for "conditional" on the forum
Good luck
Ivar
Hi
You do not have a written out "while" nor an "if", but you have almost the same via a bolean expression, see a couple of older discussion:
search for "conditional" on the forum
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
2 gen 2010, 09:16 GMT-5
Thanks Ivar,
I found the conditional search interesting and it may be useful if I am to implement discontinuous functions. However I'm not sure if it will help me with this situation where I simply want to terminate the solving process if the stress level in a component (anywhere on the subbdomain) becomes large than a set value (to eliminate unnecessary solving time). Perhaps what I am trying to do is not actually possible.
I thought that I could use a stop condition (and naiively) expected to be able to use a function like:
abs(sxy_smpn)<STRESS_MAX
However this gives me an error since it's not a scalar value with no variation in space....
Best regards,
Kevin
Thanks Ivar,
I found the conditional search interesting and it may be useful if I am to implement discontinuous functions. However I'm not sure if it will help me with this situation where I simply want to terminate the solving process if the stress level in a component (anywhere on the subbdomain) becomes large than a set value (to eliminate unnecessary solving time). Perhaps what I am trying to do is not actually possible.
I thought that I could use a stop condition (and naiively) expected to be able to use a function like:
abs(sxy_smpn)
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
2 gen 2010, 13:05 GMT-5
Hi
to get a value belonging to a subdomain, you have only the Bounday integration coupling variables, but these do an integration, not really what you want.
Your demand is rather similar to the tread:
"Maximum Temperature plot" from october 09 try a "search"
the reply was wait to V4
good luck
Ivar
Hi
to get a value belonging to a subdomain, you have only the Bounday integration coupling variables, but these do an integration, not really what you want.
Your demand is rather similar to the tread:
"Maximum Temperature plot" from october 09 try a "search"
the reply was wait to V4
good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
4 gen 2010, 06:47 GMT-5
Hi Ivar,
thanks for that update....when is V4 due for release? In the meantime I will try the work-around suggested.
An additional question.... with Comsol (in the case of a contact modelling problem) how can I make one of the bodies totally rigid (eliminating the need for meshing and reducing subsequent processing time)?
Thanks and regards,
Kevin
Hi Ivar,
thanks for that update....when is V4 due for release? In the meantime I will try the work-around suggested.
An additional question.... with Comsol (in the case of a contact modelling problem) how can I make one of the bodies totally rigid (eliminating the need for meshing and reducing subsequent processing time)?
Thanks and regards,
Kevin
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
4 gen 2010, 08:24 GMT-5
Hi
Well for V4, I'm in the same situation as you I'm waiting and expecting, but I prefer astable issue rather than an unstable one, specially with my work to be done the coming weeks (by the way I'm NOT COMSOL, just a fan of multiphysics simulations ;).
Now youre question are touching one item on my several item long lists of improved load cases to be implemented. Unfortunately, I do not have a simple reply,
As you state, taking a given geometry with a density (hence mass and inertia) and transforming it into a simple Rigid Body load case for the I/F surfaces would be very handy, other FEM software have this feature, COMSOL not yet. Using the geometry as a fixed contacct I/F woul also be of interest.
If you do not mesh the geometry you cannot use it in the postprocessing, that is how COMSOL works. You select volumes, faces, edges or points, for your BC's but this is just a shortcut to select all meshed nodes belonging to the geometrical item.
But one could think of meshing a rigid part "simply" with little details and no dependent variables to avoid having it using to much RAM, then perhaps one could use a integration boundary variable or a contact pair. Finally "contact" is detected by a metrology of distance between nodes as they inerlace, you could try to replace one of the node references by a geometrical value.
It's several months since I was last playing with contacts now, so I have forgotten the variable names, would have to re-read the doc, so try to start there.
Good luck
Ivar
Hi
Well for V4, I'm in the same situation as you I'm waiting and expecting, but I prefer astable issue rather than an unstable one, specially with my work to be done the coming weeks (by the way I'm NOT COMSOL, just a fan of multiphysics simulations ;).
Now youre question are touching one item on my several item long lists of improved load cases to be implemented. Unfortunately, I do not have a simple reply,
As you state, taking a given geometry with a density (hence mass and inertia) and transforming it into a simple Rigid Body load case for the I/F surfaces would be very handy, other FEM software have this feature, COMSOL not yet. Using the geometry as a fixed contacct I/F woul also be of interest.
If you do not mesh the geometry you cannot use it in the postprocessing, that is how COMSOL works. You select volumes, faces, edges or points, for your BC's but this is just a shortcut to select all meshed nodes belonging to the geometrical item.
But one could think of meshing a rigid part "simply" with little details and no dependent variables to avoid having it using to much RAM, then perhaps one could use a integration boundary variable or a contact pair. Finally "contact" is detected by a metrology of distance between nodes as they inerlace, you could try to replace one of the node references by a geometrical value.
It's several months since I was last playing with contacts now, so I have forgotten the variable names, would have to re-read the doc, so try to start there.
Good luck
Ivar