[mad]

turbulent viscosity limited to viscosity ratio of 1.000000e+05 in 2....5 cells;
I model a simple wind tunnel that a cone test as sample in this wind tunnel. the flow is air and model prepare in 3D. up stream and down stream checked and territory is acceptable. wind velocity is about 2-4 m/s. and turbolant model is k-w sst. problem is this message in every iteration after about 50-60 iterate.after this solution didnt converge never and residual goes up or has linear motion. in all solution Under relaxation factor for Turbulent Viscosity is 0.6. Please help in this matter. i have not time...

[kbaker]

This may due to bad mesh quality you need to examine your mesh quality in Gambit see Gambit help for more information about examining quality.

[mad]

hi dear Kbaker;
thanks for reply,
it was checked in Fluent too and report is:
Orthogonal Quality ranges from 0 to 1, where values close to 0 correspond to low quality.
Minimum Orthogonal Quality = 4.66412e-01
Maximum Aspect Ratio = 1.71025e+01
as my previous case, its not bad quality.
thanks again
Please guide me in new

[vasava]

You can plot the turbulent viscosity, find the area of problem and modify mesh in that region. You may also use adaptive mesh.

[RodriguezFatz]

By "bad mesh quality" Khalid also means a too coarse mesh. Do as vasava said and check where the error happens. Maybe your grid isn't fine enough there.

[mad]

thanks all guys
will do all your recommendation and share what happen.
thanks again

[diamondx]

show us your mesh...

[mad]

thanks Ali. I will. let me to modify the risk area mesh. may be some thing change.
thanks again

[kbaker]

In addition to what guys mentioned I add this If you sure your mesh is good try to select smaller time step (fro example 1e-05 sec or 1e-06 sec) and see is this error still reported or disappear? if it disappear try to increase the time step gradually but not jump to higher one selecting of proper time step is very critical and important to let solution converge. And If you solve steady state case try to change solution methods for simple or reduce the under-relaxation factors. Hope this solved what you need.

Originally Posted by mad

thanks Ali. I will. let me to modify the risk area mesh. may be some thing change.
thanks again

[mad]

hi Khalid Baker,
the quality of mesh is good but i'm not sure about size. With one test, in larger cells the problem is growing. and about smaller, the machine capacity is the problem (with size 1 of meshing, it will be 24000000 meshes).
did you mean changing the type of case from steady to transient? and then change the time step (in time dependence box)
realy thank you
thanks

[kbaker]

No I not mean transforming case from steady to unsteady its your decision about the case type reference to what you think about it. I mentioned the last suggestions for you for both steady and unsteady cases and let the choice for you about the modifications in your case.

Originally Posted by mad

hi Khalid Baker,
the quality of mesh is good but i'm not sure about size. With one test, in larger cells the problem is growing. and about smaller, the machine capacity is the problem (with size 1 of meshing, it will be 24000000 meshes).
did you mean changing the type of case from steady to transient? and then change the time step (in time dependence box)
realy thank you
thanks

[vasava]

There is an option in Fluent called 'Mark Poor Elements'. You can look for this topic in fluent help. Although there is no guarantee that the mesh will improve significantly but it will certainly pin-point the problem area.

I am not sure about this but the command 'mesh → repair-improve → report-poor-elements' will some how mark the poor elements. There are some commands in fluent that can improve mesh but I dont remember them now. Look for them. It might do the trick for you.

[mad]

hi again;
thanks Vasava and Baker. your suggestions helped me and i improved meshes.
i decrease mesh size and the problem with viscosity ratio till (850th iteration)already its happened in 50-80th Iteration. but the problem is why solution didnt converge. the continuity does not decrease more than 5e-2 and after that the solution move horizontally without any convergence or divergence. and some times continuty comes up!!
help me please
thanks

[jwillie2000]

In case you try the recommendations above, especially with regards to using a fine mesh and you still get the same problem then you can try to increase the turbulent viscosity under limit and see if that would help you. Jimmy

[mad]

you mean Limits tap in Control menu?
thanks
mad

[jwillie2000]

Yes, but another thing which i hope u have tried is to play with the under relaxation factors? I would usually reduce them to 0.1 for critical flow variables and then increase them gradually till the problem is resolved. That may also help.Jimmy