
[Sponsors] 
June 26, 2014, 11:33 
delta T decreases suddenly in pimplefoam

#1 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
hi
I have simulated tiny pipe with newtoniam incompressible flow which is laminar by pimplefoam but the time steps are getting smaller and smaller during the solution process(it start from 0.003 and after 10 time steps gets to 10e23 and more)and finally it is divrged .courant was set as 1 and nothing has changed by increasing the amounts of nOuterCorrectors ,nCorrectors and nNonOrthogonalCorrector.what should I do ?any suggestion? Code:
Create time Create polyMesh for time = 0 Time = 0 Mesh stats points: 9078 faces: 90732 internal faces: 83908 cells: 43660 faces per cell: 4 boundary patches: 3 point zones: 0 face zones: 1 cell zones: 1 Overall number of cells of each type: hexahedra: 0 prisms: 0 wedges: 0 pyramids: 0 tet wedges: 0 tetrahedra: 43660 polyhedra: 0 Checking topology... Boundary definition OK. Cell to face addressing OK. Point usage OK. Upper triangular ordering OK. Face vertices OK. Number of regions: 1 (OK). Checking patch topology for multiply connected surfaces... Patch Faces Points Surface topology body 6496 3278 ok (nonclosed singly connected) outlet 164 98 ok (nonclosed singly connected) inlet 164 98 ok (nonclosed singly connected) Checking geometry... Overall domain bounding box (0.0215399 0.0115756 0.002365) (0.0198627 0.00851412 0.002365) Mesh (nonempty, nonwedge) directions (1 1 1) Mesh (nonempty) directions (1 1 1) Boundary openness (3.2476e18 2.81028e17 1.76588e17) OK. Max cell openness = 1.42371e16 OK. Max aspect ratio = 6.62732 OK. Minimum face area = 4.49238e08. Maximum face area = 2.90028e07. Face area magnitudes OK. Min volume = 3.78117e12. Max volume = 4.92374e11. Total volume = 8.03586e07. Cell volumes OK. Mesh nonorthogonality Max: 56.6451 average: 19.2493 Nonorthogonality check OK. Face pyramids OK. Max skewness = 0.702453 OK. Coupled point location match (average 0) OK. Mesh OK. End 

June 27, 2014, 06:56 

#2 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
there is no idea?


June 27, 2014, 07:10 

#4 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
thanks for your answer
these are fv solution and scheme Code:
solvers { p { solver PCG; preconditioner DIC; tolerance 1e06; relTol 1e05; } pFinal { solver PCG; preconditioner DIC; tolerance 1e08; relTol 0; } "(Ukepsilon)" { solver PBiCG; preconditioner DILU; tolerance 1e05; relTol 0.1; } "(Ukepsilon)Final" { $U; tolerance 1e05; relTol 0; } } PIMPLE { nOuterCorrectors 4; nCorrectors 2; nNonOrthogonalCorrectors 3; pRefCell 0; pRefValue 0; residualControl { U { tolerance 1e2; relTol 0; } p { tolerance 1e2; relTol 0; } } } relaxationFactors { fields { } equations { "p.*" 0.3; "U.*" 0.7; "k.*" 0.4; "epsilon.*" 0.5; } } Code:
ddtSchemes { default Euler; } gradSchemes { default Gauss linear; grad(p) Gauss linear; grad(U) Gauss linear; } divSchemes { default none; div(phi,U) Gauss limitedLinearV 1; div(phi,k) Gauss limitedLinear 1; div(phi,epsilon) Gauss limitedLinear 1; div(phi,R) Gauss limitedLinear 1; div(R) Gauss linear; div(phi,nuTilda) Gauss limitedLinear 1; div((nuEff*dev(T(grad(U))))) Gauss linear; } laplacianSchemes { default none; laplacian(nuEff,U) Gauss linear corrected; laplacian((1A(U)),p) Gauss linear corrected; laplacian(DkEff,k) Gauss linear corrected; laplacian(DepsilonEff,epsilon) Gauss linear corrected; laplacian(DREff,R) Gauss linear corrected; laplacian(DnuTildaEff,nuTilda) Gauss linear corrected; } interpolationSchemes { default linear; interpolate(U) linear; } snGradSchemes { default corrected; } fluxRequired { default no; p ; } 

June 27, 2014, 07:23 

#5 
Senior Member
Tushar Chourushi
Join Date: Jul 2009
Location: IITIndore, India
Posts: 319
Blog Entries: 1
Rep Power: 10 
You're welcome!
In fvscheme file try these instead of mentioned ones. Best is to try with 1st order and check your results. divSchemes { default none; div(phi,U) Gauss upwind; div(phi,k) Gauss upwind; div(phi,epsilon) Gauss upwind; div(phi,R) Gauss upwind; div(R) Gauss linear; div(phi,nuTilda) Gauss upwind; div((nuEff*dev(T(grad(U))))) Gauss linear; } Let us know about your results. post your reply here  Best Regards! 

June 27, 2014, 07:44 

#6 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
that works and everyting is ok
thank you but may I ask you to explain how did you find that I mean what the big differnces are there between these two div schemes? thanks 

June 27, 2014, 07:58 

#7 
Senior Member
Tushar Chourushi
Join Date: Jul 2009
Location: IITIndore, India
Posts: 319
Blog Entries: 1
Rep Power: 10 
Good!
Actually, there are various ways to avoid such situations. One such problem could be due to the above situation. You can find the reference for these in any of the CFD books. These "div schemes" differ in order of accuracy and many other facts. Quick Hint: Upwind scheme is unconditionally stable of all. 

June 27, 2014, 15:47 

#8 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
thanks
As you know yhis is a first order scheme, thus too diffusive and I will not get a good solution,am I right?and what is the solution 

June 28, 2014, 00:47 

#9 
Senior Member
Tushar Chourushi
Join Date: Jul 2009
Location: IITIndore, India
Posts: 319
Blog Entries: 1
Rep Power: 10 
Please, excuse for late reply. I was away from the desk.
Yes you're correct. With increased mesh resolution you can obtain very close solutions. I don't know the details of your problem case. It's always wise to compare the CFD results with the available resources (experimental or analytical). Anyways, with experience and literature survey you can gain grip over many of these CFD techniques. Best of Luck for your future endeavour.  Best Regards! 

June 28, 2014, 01:14 

#10 
Member
Ali reza
Join Date: Mar 2014
Posts: 95
Rep Power: 4 
thanks for protect and excuse me for lots of question (I am so curious ) but I promise that this is the last one
I think one way to decrease inaccuracy in using upwind linear in starting the algorithm with upwind and then changing it to more accurate schemes but I dont know when is the right time to do that,I mean is there any clue in the terminal that can aware us? thanks alot 

June 28, 2014, 01:34 

#11 
Senior Member
Tushar Chourushi
Join Date: Jul 2009
Location: IITIndore, India
Posts: 319
Blog Entries: 1
Rep Power: 10 
Actually, you can get comparable solution with upwind scheme too depending upon your problem case. Sometime, results obtained are comparable with the available resources. You can find many literature in which 1st order scheme give very near comparable solutions.
This is because CFD solution doesn't depend on these factors only. There are many other factors which are equally important for CFD solutions. It's always best to monitor residuals. Generally, convergence check tell us when to stop our iterations. I hope I tried to answer your queries.  Best Regards! 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
pimpleFoam: turbulence>correct(); is not executed when using residualControl  hfs  OpenFOAM Running, Solving & CFD  3  October 29, 2013 09:35 
Delta wing cfd  H_R  FLUENT  16  September 19, 2013 04:41 
Smoothing of IDDES delta  Fuchs  OpenFOAM Running, Solving & CFD  0  July 15, 2013 11:53 
Differences simpleFoam vs. pimpleFoam / RASModel.H vs turbulenceModel.H  uli  OpenFOAM Programming & Development  7  January 26, 2013 16:01 
Delta Wing Structured Grid  Riaan  FLUENT  3  December 31, 2004 13:03 