[Piloten]

I'm trying to simulate a transient free surface flow in a closed fishway. I'm now getting an error message 38. The problem is that the simulation is going great in the beginning but all of a sudden the program just freezes. Then when i try to abort the simulation the error code 38 appears.

to simplify you can image this closed fish way as a box with an inlet flow and an outlet flow. The inlet and outlet has been set to boundaries as openings because it is supposed to be gravitational flow. Therefore the turbulence model has been set to SSG.

I have tried to lower the time step from 0.1 in incrementally to 0.001 with no better results. I have also changed the coefficient loop iterations from 10 to 4.


chart_help.png

output_help.zip

[Gert-Jan]

- For reasons of security, I won't open downloaded zip-files. Better add plain text.
- Start with SST as turbulence model. If converged, you can continue your calculation with SSG.
- Did you include the free surface as a free slip wall, or do you really have a free surface with air above? If so, did you initialize the pressures according to what is written in the Help?

[Piloten]

Here is the output help zip. I included the latest information that the output file gave me.

================================================== ====================
| Timestepping Information |
----------------------------------------------------------------------
| Timestep | RMS Courant Number | Max Courant Number |
+----------------------+----------------------+----------------------+
| 1.0000E-03 | 2.72 | 25.71 |
----------------------------------------------------------------------

================================================== ====================
TIME STEP = 33 SIMULATION TIME = 3.3000E-02 CPU SECONDS = 1.686E+04
----------------------------------------------------------------------
COEFFICIENT LOOP ITERATION = 1 CPU SECONDS = 1.686E+04
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom-Bulk | 1.13 | 1.7E-03 | 6.3E-02 | 7.6E-04 OK|
| V-Mom-Bulk | 1.14 | 1.6E-03 | 6.8E-02 | 8.3E-04 OK|
| W-Mom-Bulk | 1.14 | 1.5E-03 | 6.2E-02 | 6.0E-04 OK|
| P-Vol | 0.94 | 1.7E-04 | 1.8E-02 | 8.7 3.1E-02 OK|
+----------------------+------+---------+---------+------------------+
| Mass-Water | 1.00 | 8.1E-03 | 3.3E-01 | 10.5 2.3E-05 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE-Bulk | 0.90 | 3.2E-03 | 1.5E-01 | 10.5 6.6E-05 OK|
| O-TurbFreq-Bulk | 0.86 | 1.9E-03 | 1.6E-01 | 7.4 9.6E-04 OK|
+----------------------+------+---------+---------+------------------+
----------------------------------------------------------------------
COEFFICIENT LOOP ITERATION = 2 CPU SECONDS = 1.696E+04
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom-Bulk | 1.12 | 1.9E-03 | 1.7E-01 | 1.9E-03 OK|
| V-Mom-Bulk | 0.92 | 1.5E-03 | 6.7E-02 | 2.1E-03 OK|
| W-Mom-Bulk | 1.01 | 1.5E-03 | 1.5E-01 | 1.3E-03 OK|
| P-Vol | 0.81 | 1.3E-04 | 3.4E-02 | 8.7 5.3E-02 OK|
+----------------------+------+---------+---------+------------------+
| Mass-Water | 0.29 | 2.4E-03 | 1.7E-01 | 10.6 1.4E-05 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE-Bulk | 0.67 | 2.2E-03 | 9.6E-02 | 10.5 4.5E-05 OK|
| O-TurbFreq-Bulk | 1.23 | 2.4E-03 | 1.5E-01 | 7.4 6.7E-04 OK|
+----------------------+------+---------+---------+------------------+
----------------------------------------------------------------------
COEFFICIENT LOOP ITERATION = 3 CPU SECONDS = 1.706E+04
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom-Bulk | 0.48 | 9.1E-04 | 9.0E-02 | 2.6E-03 OK|
| V-Mom-Bulk | 0.49 | 7.2E-04 | 5.9E-02 | 2.9E-03 OK|
| W-Mom-Bulk | 0.51 | 7.5E-04 | 9.5E-02 | 2.2E-03 OK|
| P-Vol | 0.56 | 7.6E-05 | 1.3E-02 | 8.7 2.0E-02 OK|
+----------------------+------+---------+---------+------------------+
| Mass-Water | 0.39 | 9.3E-04 | 7.8E-02 | 10.5 1.4E-05 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE-Bulk | 0.40 | 8.6E-04 | 6.6E-02 | 10.5 3.8E-05 OK|
| O-TurbFreq-Bulk | 0.60 | 1.4E-03 | 1.7E-01 | 7.4 6.2E-04 OK|
+----------------------+------+---------+---------+------------------+
----------------------------------------------------------------------
COEFFICIENT LOOP ITERATION = 4 CPU SECONDS = 1.716E+04
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom-Bulk | 0.55 | 5.0E-04 | 1.2E-01 | 2.1E-03 OK|
| V-Mom-Bulk | 0.51 | 3.7E-04 | 5.3E-02 | 2.2E-03 OK|
| W-Mom-Bulk | 0.49 | 3.7E-04 | 9.6E-02 | 1.7E-03 OK|
| P-Vol | 0.47 | 3.6E-05 | 8.0E-03 | 8.7 1.8E-02 OK|
+----------------------+------+---------+---------+------------------+
| Mass-Water | 0.56 | 5.2E-04 | 4.0E-02 | 10.6 1.4E-05 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE-Bulk | 0.48 | 4.1E-04 | 5.3E-02 | 10.5 5.0E-05 OK|
| O-TurbFreq-Bulk | 0.36 | 5.2E-04 | 6.3E-02 | 7.4 9.7E-04 OK|
+----------------------+------+---------+---------+------------------+
----------------------------------------------------------------------
COEFFICIENT LOOP ITERATION = 5 CPU SECONDS = 1.726E+04
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom-Bulk | 0.59 | 2.9E-04 | 6.7E-02 | 1.8E-03 OK|
| V-Mom-Bulk | 0.59 | 2.2E-04 | 3.0E-02 | 1.9E-03 OK|
| W-Mom-Bulk | 0.53 | 2.0E-04 | 2.6E-02 | 1.5E-03 OK|
| P-Vol | 0.56 | 2.0E-05 | 3.8E-03 | 8.7 1.7E-02 OK|
+----------------------+------+---------+---------+------------------+
| Mass-Water | 0.65 | 3.4E-04 | 2.1E-02 | 10.6 1.6E-05 OK|
+----------------------+------+---------+---------+------------------+

+--------------------------------------------------------------------+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 38. |
+--------------------------------------------------------------------+

End of solution stage.

+--------------------------------------------------------------------+
| The following transient and backup files written by the ANSYS CFX |
| solver have been saved in the directory |
| K:/TVM/F7013T/Fiskvandring/Andretest1_files/fleraboxesrr_pending/- |
| dp0_CFX_1_Solution_1-1/2_001: |
| |
| 0_full.trn |
+--------------------------------------------------------------------+


+--------------------------------------------------------------------+
| The results from this run of the ANSYS CFX Solver have been |
| written to |
| K:/TVM/F7013T/Fiskvandring/Andretest1_files/fleraboxesrr_pending/- |
| dp0_CFX_1_Solution_1-1/2_001.res |
+--------------------------------------------------------------------+


+--------------------------------------------------------------------+
| For CFX runs launched from Workbench, the final locations of |
| directories and files generated may differ from those shown. |
+--------------------------------------------------------------------+


This run of the ANSYS CFX Solver has finished.

[Piloten]

Originally Posted by Gert-Jan

- For reasons of security, I won't open downloaded zip-files. Better add plain text.
- Start with SST as turbulence model. If converged, you can continue your calculation with SSG.
- Did you include the free surface as a free slip wall, or do you really have a free surface with air above? If so, did you initialize the pressures according to what is written in the Help?

I'll try to simulate and get convergence with sst now. But i have not done any y+ studies so I'm a bit worried with that turbulence model working. However the convergence part of the simulation was not reached with the ssg model. The problem is that cfx just freeezes around 30 iterations.

The initialization is set to air = 1 and water = 0. It is supposed to be a gravitational driven water flow from the inlet.

[Gert-Jan]

- I have no idea what error 38 means........

- Are you calculating the filling of the tank? You are a brave person.

- How does the results look like at iteration 30?

[Piloten]

Yes the idea is to simulate how much and in what way the filling of the tank looks like depending on if the inlet is below or above water levels and also if the outlet is below or above water levels.
So basically the filling with different hydrostatic pressures.

I get the error code 38 and then I can't view any results in post.

[Gert-Jan]

Your calculation stops at coefficient-loop/iteration 5 of timestep 33. How does your result look like at the end of timestep 30?

[Gert-Jan]

You perform a run with a timestep of 1e-3[s] and have finalized 0.033[s] so far.
How long do you need to calculate before your fishtank is filled. Do you really have a viable question for CFD? To me it looks completely unrealistic.

[Piloten]

Originally Posted by Gert-Jan

Your calculation stops at coefficient-loop/iteration 5 of timestep 33. How does your result look like at the end of timestep 30?

I didn't know you could load those iterations as results in post I will have to try that and see what it looks like.

[QUOTE]

You perform a run with a timestep of 1e-3[s] and have finalized 0.033[s] so far.
How long do you need to calculate before your fishtank is filled. Do you really have a viable question for CFD? To me it looks completely unrealistic.

Well the expectation for the tank is that it will be filled different ways depending on the hydrostatic pressure difference between inlet and outlet. So what I want to display using CFD is the different degrees of filling in the tank.

[Gert-Jan]

In CFX-Pre, you should go to Output Control where you can save transient results for every timestep you want. In these files, you can include the variables that are relevant, like volume fraction, velocity and pressure.

[Piloten]

Originally Posted by Gert-Jan

In CFX-Pre, you should go to Output Control where you can save transient results for every timestep you want. In these files, you can include the variables that are relevant, like volume fraction, velocity and pressure.

Now that you mention it, I did save all of the transient results. I'll check it out.
I also got a recommendation from a colleague to run the simulation on CFX as stand alone instead of through workbench which seems to have solved the issue regarding the program freezing.

[Opaque]

May I ask what version of Ansys CFX you are using?

with recent versions, there should be a "trace" file left in the running directory and reported in the output file.

Such file must be provided to Ansys CFX Support, and they can figure out where the CFX solver stopped, and perhaps give a hint on what to do next.