Hi

Try going to "Materials" and check whether all materials copper and water have a "mu".

If mu is viscosity, this might be challenging for a solid object. If this is the case, your simulation space may be set incorrect at "laminar flow" physics. Probably, you want to simulate the flow within the tube, but not the flow of the shell itself.

It seems that yu have defined a material property named "mu" in your physics, but not in your parameters; usually mu is the greek letter for dynamic viscosity. I suppose that this is the case because you have a "fluid flow module", where Navier-Stokes equations have to be set and solved on your computational grid (mesh).

Before starting the stationary solver, check if dynamic voscity in the physical subonde Fluid Properties 1 (below Fluid Flow module) has been defined properly.

To check this, go into your "Parameters 1" or in "Definition", where you should have defined the variable "mu"; if "mu" is not defined into these menu, then define its value (or expression as a function of other parameters).

Let me know if it is clear and if the problem is solved.

Best regards,

AP

Jasper Giesler

I fix the issue of the property of the fluid

but I got another issue which is

( Undefined value found. - Detail: NaN or Inf found when solving linear systems using SOR. )

how in this situation ؟

look at the pics, I got the solution but I used input temp and surface instead of heat flux and the result its seems not right

Alessio Pricci

Thanks my freind but i got anther issue which is

(Undefined value found. - Detail: NaN or Inf found when solving linear system using SOR.)

My mesh is normal because my PC cant handle the fine or finer mesh its take 20 GB of memory and don't solve in the end , I got final message ( out of memory ) so what can is do in these case

Alessio Pricci

I fix the issue of 'mu' fluid property

but for Navier-Stokes equations has to be set in mesh layer , I didn'tknow about it , I apologise I'm new for the software

Dear Sezar,

Navier-Stokes equations are embedded in Fluid Flow module; these equations are discretized by Finite Elements on your mesh.

Your problem seems to be related to the mesh size; let's see this discussion:

https://www.researchgate.net/post/Getting-error-in-COMSOL-like-NaN-of-Inf-found-when-solving-linear-system-using-Vanka-What-may-be-the-possible-solution-for-this

How to deal with that? A finer mesh would be useful, but if the message "out of memory" is returned the problem is that you RAM is not enough. Comsol takes a lot of RAM.

One tip: in you Fluid Properties 1 (under Fluid Flow module) let's change your discretization scheme from P2+P1 to P1+P1; in this way, you solve all variables (velocities and pressure) on first order shape functions.

This is not good when you want an accurate solution, but it can save a lot of degrees of freedom, allowing for more complicated meshes.

The other solution is to buy an hardware with more RAM and/or change software (e.g., Ansys Fluent, which handle very weel thermo- and fluid dynamics problems).

Best regards,

AP

Alessio Pricci

Thanks my friend

but I take another procedure, I made work work plane and portionmt domain into a quarter of the original shape and made finer meshing,

notes ..

i used heat flax instead of input temperature ..

if this didn't work . can i send you the mph file , if that possible?