CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Siemens (
-   -   strange elapsed time (

azmir February 4, 2008 20:33

strange elapsed time
Has anyone else seen alternating elapsed time for iteration? This just occurred recently on v326. The elapsed time changes from 20-30 secs to 300-400 secs alternatively. I wonder if this is due to server or bugs? FYI, very similar model has been run without problem before this.

1 3.45E-04 1.13E-04 1.25E-04 1.11E-03 4.97E-04 3.48E-03 0.00E+00 7.89E+00-1.86E-01 8.62E-02-1.45E+01 1.12E+00 9.54E+00 4.52E+01

2 3.63E-05 1.04E-04 4.25E-05 4.80E-04 9.21E-03 5.80E-03 0.00E+00 1.04E-01-1.90E-03-5.81E-03 0.00E+00 3.12E-04 5.43E-04 1.22E+02 FINISH ITERATION NO. 434 CPU TIME IS 68321.23 ELAPSED TIME IS 68319.31 MEMORY (DYNAMIC) USED IS 574 MBYTES ---- ------------------------------------------------------- ITER. NO 435 -------------------------------------------------------

1 3.43E-04 1.13E-04 1.24E-04 1.55E-03 4.99E-04 3.49E-03 0.00E+00 7.89E+00-1.86E-01 8.59E-02-1.45E+01 1.12E+00 9.58E+00 4.52E+01

2 3.57E-05 1.03E-04 3.97E-05 6.55E-04 9.17E-03 5.77E-03 0.00E+00 1.04E-01-1.90E-03-5.82E-03 0.00E+00 3.11E-04 5.41E-04 1.22E+02 FINISH ITERATION NO. 435 CPU TIME IS 68350.86 ELAPSED TIME IS 68348.94 MEMORY (DYNAMIC) USED IS 574 MBYTES ---- ------------------------------------------------------- ITER. NO 436 -------------------------------------------------------

1 3.44E-04 1.12E-04 1.25E-04 6.31E-04 5.00E-04 3.48E-03 0.00E+00 7.89E+00-1.86E-01 8.58E-02-1.45E+01 1.12E+00 9.61E+00 4.52E+01

2 3.66E-05 1.03E-04 4.83E-05 2.72E-04 9.15E-03 5.74E-03 0.00E+00 1.04E-01-1.89E-03-5.84E-03 0.00E+00 3.10E-04 5.39E-04 1.22E+02 FINISH ITERATION NO. 436 CPU TIME IS 68764.05 ELAPSED TIME IS 68762.09 MEMORY (DYNAMIC) USED IS 574 MBYTES ---- ------------------------------------------------------- ITER. NO 437 -------------------------------------------------------

1 3.41E-04 1.12E-04 1.24E-04 1.10E-03 5.02E-04 3.48E-03 0.00E+00 7.89E+00-1.87E-01 8.56E-02-1.46E+01 1.13E+00 9.65E+00 4.52E+01

2 3.57E-05 1.02E-04 4.31E-05 4.83E-04 9.10E-03 5.71E-03 0.00E+00 1.04E-01-1.89E-03-5.85E-03 0.00E+00 3.08E-04 5.38E-04 1.22E+02 FINISH ITERATION NO. 437 CPU TIME IS 68790.10 ELAPSED TIME IS 68788.15 MEMORY (DYNAMIC) USED IS 574 MBYTES ---- ------------------------------------------------------- ITER. NO 438 -------------------------------------------------------

1 3.38E-04 1.11E-04 1.24E-04 6.03E-04 5.03E-04 3.48E-03 0.00E+00 7.89E+00-1.87E-01 8.55E-02-1.46E+01 1.13E+00 9.69E+00 4.52E+01

2 3.59E-05 1.02E-04 4.18E-05 2.53E-04 9.06E-03 5.68E-03 0.00E+00 1.04E-01-1.88E-03-5.87E-03 0.00E+00 3.07E-04 5.36E-04 1.22E+02

murash-v February 5, 2008 06:19

Re: strange elapsed time
I think is all good. What is the problem ?? For example you may have different piso stages for each timestep....

azmir February 5, 2008 21:58

Re: strange elapsed time
I am not sure, I thought this was not normal because different and alternating elapsed times on each iteration. Do you think it's normal for steady state problem using SIMPLE? I think i've run this model last year but no strange elapsed time like this time (this model is slightly modified geometry).

Pauli February 6, 2008 19:31

Re: strange elapsed time
The times will vary per iteration because each iteration/time step has a different number of sweeps (sub-iterations). If you normalize the times with total number of sweeps the results should become a bit more consistent.

And yes you can get rather large (10X) variation in sweeps between consecutive iterations. I've never found an answer to that mystery.

azmir February 10, 2008 01:24

Re: strange elapsed time
Thanks for that info. So something like does occur on rare cossasions, without solution? I thought it could be a bug or something else... almost contacted hardware (server) supplier about this!

All times are GMT -4. The time now is 10:18.