different result between seriel and parallel running

cyw · January 15, 2018, 04:05

Dear Foamers,

I implemented level set method into interFoam. I found an interesting problem. For serial and parallel running, the result is different. Anyone meets this problem? Anyone knows why?

Artur.Ant · August 22, 2018, 16:33

I also found this problem in my simulations. By the way in some cases I observed that turbulent kinetic equation (in rhoSimpleFoam) after the first iterations has residuals of order of e-09 and this is impossible. Solving the same problem in serial way gave good, realistic and pretty stable (steady state) results.
Have you solved this difference in results between serial and parallel solver?

January 15, 2018, 04:05	different result between seriel and parallel running	#1
cyw Member Yuanwei Cao Join Date: Jun 2017 Location: Munich,Germany Posts: 54 Rep Power: 8	Dear Foamers, I implemented level set method into interFoam. I found an interesting problem. For serial and parallel running, the result is different. Anyone meets this problem? Anyone knows why?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
overlapGgi parallel 1.6-ext	linnemann	OpenFOAM Running, Solving & CFD	7	January 24, 2011 11:03
Cyclic patches and parallel postprocessing problems	askjak	OpenFOAM Bugs	18	October 27, 2010 03:35

August 22, 2018, 16:33		#2
Artur.Ant Member Arthur Join Date: Aug 2014 Location: Italy Posts: 47 Rep Power: 11	I also found this problem in my simulations. By the way in some cases I observed that turbulent kinetic equation (in rhoSimpleFoam) after the first iterations has residuals of order of e-09 and this is impossible. Solving the same problem in serial way gave good, realistic and pretty stable (steady state) results. Have you solved this difference in results between serial and parallel solver?