CFX-Solver
 

Hi.

The solver stopped my simulations after appr. 4 timesteps given the following error message:

An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver could not be started, or exited with return |
| code 255: . No results file has been created.


Is there any thing that contribute to that failure?
1、improper mesh ?
2、some settings which have a close relation with numerical stability when the continuity, momentum, energy and mass transport equations are solved together
Many thanks!

It could as well be a license problem. So check your license server to make sure you have available license. Jimmy

I suppose there is nothing to do with my license.
But other factors are responsible for the failure.

I can say this is DEFINITELY not a solver/mesh failure. It might be a memory/license issue. try increasing the default memory from 1.0 to 2.0

If it's the memory issue,there would be some warnings saying that "there is not enough memory on the computer".
I have used my friends' computer, the message is also accurate the same like this.

try this....

start cfx stand alone. SAVE THE PROJECT!

import your mesh and definition file into this new project.

try running now ....don't use wb

I did get a similar error but it has been long and i am trying to remember what was wrong. Have you checked to see whether your time step has been adequately given? It could be the value there (run time) is not adequate. Are u doing your unsteady simulation from a converged steady state one? Jimmy

Thanks for your answer.
My simulation is unsteady.
Because the rotate speed of the gear is 2000r/min, the time step,that is dt, is (1/400)[s]/30 (1/400 is the time for each gear to pass through,and I divide the time into 30 steps) and total time is (12/400)[s] which means a period time.
So what can I do?

Hi, my suggestion would be to look for a similar tutorial in CFX or contact ANSYS for support. Not sure whether you have a local coordinate system about which your model is rotating? I think that might be creating the problem.

Thanks for your answer.
Would you explain it more specifically? I can't not understand your words fully.

Thanks for your answer!
Would you explain it more specifically? I was confused with your words.

+--------------------------------------------------------------------+
| Warning! |
| |
| After waiting for 60 seconds, 1 solver manager process(es) appear |
| not to have noticed that this run has ended. You may get errors |
| removing some files if they are still open in the solver manager. |
+--------------------------------------------------------------------+



Above is more information about my model>

My reply was assuming you're using workbench. This kind of problem happened to me when i'm transferring files from a local computer to server. There were possibly some files lost.

You know you can use CFX in standalone mode without using WB. It will in Programfiles>ANSYS>Fluid Dynamics>CFX (assuming you are using windows). And then import your mesh file and the ccl file and see if it works

HTH

By looking into the snapshot of the model and notice message, it seems that there is no adequate fluid flow channel to make it through, and the inflow back to inlet which is different from the BC you set, that is why an artificial wall was built.

Thanks. My model was in CFX stand alone.

I have changed the settings ,both the "outlet" and "inlet", but the message in CFX-Solver still the same shown on Fig above.

Forget the memory errors.

Your problem is definately setup wrong or ill-formed. CFX is blocking off the inlet to keep flow from trying to travel back out.

Are your gears turning? Is there flow through even if the gears are not turning? Is your inlet set right (i.e. you didnt specify a cartisian component of the flow to force it out?)

QUOTE]Thanks a lot.If the gear is not turning,the total fluid area is static which is the initial state of fluid area.

what do you mean by the inlet setting?you can check the model before meshed. it's the "coordinate 0",which was set in fluid area.

Could anyone offer some advice on my problem?I'm in great gratitude .

i got this error few days ago! then I used a different computer and it worked, besides problem setup, it seems to me that it might be because of some missing files or memory problems. the other computer is a server and it just worked fine. try re-installing ansys or increase your system resources.
in the solver manager increase the memory allocation from default 1.0 to 2.0

For me it still sounds like wrong BC settings or maybe even domain initialisations. The fact that cfx is running and trying to place a wall on the inlet to prevent flow from flowing OUT does not sound like files missing or insufficient memory but simply physically wring boundary conditions.

shanxuewenjdx maybe you could specify how you defined the BC's (inlet/outlet etc.)??? and the domain initialisations? Just wondering if you maybe have defined a higher pressure on the opening side than on the inlet side wich would cause a "reverse" flow...presuming that your gears are able to rotate and are not static.

Stefan, maybe this thread helps: http://www.cfd-online.com/Forums/cfx...fx-solver.html

Glenn was suggesting that maybe the solver has diverged (or did some division by 0) which caused the solver to crash.

Have you done a sensitivity check on the mesh? How many elements does it have? What is the Courant number for your simulation?

Best regards

Are you running in parallel? I have seen this behaviour occcasionally from parallel runs of the CFX-Solver, in that the real problem is disguised, and associated messages are not output.

If you can run it in serial, and there is a problem with the setup, then the solver should write some messages which are more useful than just 'error code 255'.

CG

many thanks!
I run it in serial. But if it was something with my setup,is it related to my inlet(I set it 0MPa) or outlet(20MPa)? After reading the CFX-help file ,it is recommended that inlet:0MPa,outlet:2.667kg/s according to the real situation.But it still can't be right

Hi,guys. I forgot the mesh statitics in CFX-SOLVER when the solver failed to solve the simulation. So is it for sure that the failure is caused by coarse mesh?

the picture I copied from CFX-solver

And some other pictures

The average density of your fluid is 9e-7. Are you sure this is correct?

I chose water as fluid material.anything wrong?

Well, given that water has a density of about 1.000 kg/m3 at room temperature and it is incompresible, 9e-7 sure means something is wrong.

What are the Inlet conditions?? In CFD-Post plot some variables (temperature, velocities, etc) to check if the problem is on a specific area of your domain.

I suspect that you are using 0 reference pressure and almost zero inlet/outlet pressure, leading to very low density fluid as the pressure is just about absolute zero. The reference pressure needs to be set at a suitable level for the simulation - if this thing is close to atmospheric pressure then use 1 [atm] for the reference pressure.

Follows are my further information in my cfx-pre

And these pictures.

Thanks a lot. For more information, u can see more pictures above.the Inlet:2.37m/s.
As to the density of fluid,is the pre-setting I have finished is wrong or the cfx-solver itself?

So kind you are. u can check further information from the above pictures.

Correct me if I am wrong, but it does not seem that the gears are turning. How can there be a flow if the gears are not turning?

I see your inlet and outlet are both 20MPa, and your reference pressure is 1atm. In that case you should use a reference pressure of 20.1MPa and an inlet and outlet pressure of 0 Pa.

The gear1 and gear2 are turning around their geometric centers respectively. the angular velocity both are 2000r/min,u can check that.

Thanks. there are 3 boundaries: inlet,outlet,inletsmall(you can check the first picture attached under the thread).the inlet:2.37m/s;outlet(opening) 20MPa;inletsmall(close to the outlet;set as opening )20MPa.The inletsmall I have modified ,not as OUTLET type