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 The information provided may be out of date.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.

Warning messege

Please login with a confirmed email address before reporting spam

Hi dear all,

I made a simple (3D) model (a tube in which air is flowing) and at the first stages of my modelling I just made my geometry , materials and air flow, so I computed this simple model and it seems everything was defined to the software correctly so that withour any error I could get the results, but there was a Warning:

warning !
Ill-conditioned preconditioner. Increase factor in error estimate to 73.88653904508402

Do you know what is this for?

With best regards,
Semirams


4 Replies Last Post 26 apr 2011, 16:15 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 apr 2011, 10:49 GMT-4
In the initial conditions, give the model as starting point the velocity that you used at your inlet.
That might remove your warning message,

Luis
In the initial conditions, give the model as starting point the velocity that you used at your inlet. That might remove your warning message, Luis

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 26 apr 2011, 15:06 GMT-4
Hi

My advice is also going in that direction, I have noticed, that helping the solver with some "good" initial condition guesses, improves a lot the time to results. In fact for simple fluid flow cases, I mostly apply a parabolic velocity profile, as initial and inlet conditions, and for the initial pressure drop I use Poiseuille law to have a pressure gradient along the flow line.

Then too check the Reynolds cell number to be sure you are in laminar flow conditions, when you apply it, this might give an indication about the mesh size ratio.

Finally use "boundary mesh layers along the non-slip walls, this is not yet always implemented in the "default physics settings of COMSOL mesher

--
Good luck
Ivar
Hi My advice is also going in that direction, I have noticed, that helping the solver with some "good" initial condition guesses, improves a lot the time to results. In fact for simple fluid flow cases, I mostly apply a parabolic velocity profile, as initial and inlet conditions, and for the initial pressure drop I use Poiseuille law to have a pressure gradient along the flow line. Then too check the Reynolds cell number to be sure you are in laminar flow conditions, when you apply it, this might give an indication about the mesh size ratio. Finally use "boundary mesh layers along the non-slip walls, this is not yet always implemented in the "default physics settings of COMSOL mesher -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 apr 2011, 15:26 GMT-4
Hi Ivar,
Could you please guide me how to calculate pressure drop for the flow between two parallel plates (square duct). I think Poiseuille law is for the circular pipe flow..
thanks
Mayur
Hi Ivar, Could you please guide me how to calculate pressure drop for the flow between two parallel plates (square duct). I think Poiseuille law is for the circular pipe flow.. thanks Mayur

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 26 apr 2011, 16:15 GMT-4
Hi

you are right for Poiseuille, I overlooked your 2D parallel plate comment, went back to my pipe issue ;)

My main comment was if you have difficulties to get the solver to converge, any good estimates is better than a flat "0". You need some pressure build-up to drive a flow, then you must use your common physics sense, and perhaps iterate, or repeat the solver sequence with something close to the first solution found (used as initial conditions for second attempt) just to check that the solution remains the same

--
Good luck
Ivar
Hi you are right for Poiseuille, I overlooked your 2D parallel plate comment, went back to my pipe issue ;) My main comment was if you have difficulties to get the solver to converge, any good estimates is better than a flat "0". You need some pressure build-up to drive a flow, then you must use your common physics sense, and perhaps iterate, or repeat the solver sequence with something close to the first solution found (used as initial conditions for second attempt) just to check that the solution remains the same -- Good luck Ivar

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.