CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Help needed with laminar couette flow boundary conditions (http://www.cfd-online.com/Forums/openfoam-solving/62779-help-needed-laminar-couette-flow-boundary-conditions.html)

bengt March 19, 2009 09:25

Help needed with laminar couette flow boundary conditions
 
Dear Foamers,

first of all, THANKS A LOT for the great support in the forum and the great job the OpenFoam development team is doing!! Still being a beginner with cfd and OpenFoam, I am currently trying to simulate a microscopic incompressible couette flow for a measured sliding contact geometry, using simpleFoam. The aim is a comparison of the calculated friction torque with the experimental values. The computational domain (170 x 170 Ám) is a portion of a fully flooded tribological sliding contact, as similarly described by P. Brajdic in the OpenFoam thesis repository. The fluid film thickness between the fixed upper wall and the moving lower wall (sliding in x-direction) is in the range of approx. 0,1..10 Ám; the blockMesh consists of approx. 2,8Mcells of hex type and checkMesh tells me that the mesh quality is fine. The main problem now is the correct definition of the boundary conditions, since it's very difficult to find information on this rather special case. After some trials with exploding models :D, I am currently using the following, but I am not sure if they are correct:

"inlet" (fluid film cross section normal to x): p = 0, u = zeroGradient
"outlet"(see above): p = 0, u = zeroGradient
side1 (fluid film cross section parallel to x): p = 0, u = zeroGradient
side2 (as above): p = 0, u = zeroGradient
fixed (upper) wall: p = zeroGradient, u = (0, 0, 0)
moving (lower) wall: p = zeroGradient, u = (Ux, 0, 0)

The patch types in the boundary file are all "patch" except for the walls being set to type "wall".
Generally, after approximately 4,5 thousand steps the results for x are looking fine (at least for me :rolleyes:), but there seems to be some convergence issue concerning Uy, Uz and p. However, the fluid film thickness (or gap height) does not vary at all in the sliding direction x (one fluid film cross section simply extruded in x), so I think that the algorithm only deals with the numerical errors regarding Uy, Uz and p :confused:. The logfile entries are as follows:

Time = 4795

smoothSolver: Solving for Ux, Initial residual = 8.28444717201096e-07, Final residual = 1.17861528964069e-16, No Iterations 4
smoothSolver: Solving for Uy, Initial residual = 0.000420799816002115, Final residual = 1.15677735105817e-14, No Iterations 4
smoothSolver: Solving for Uz, Initial residual = 0.00045689356317463, Final residual = 1.21403757603192e-14, No Iterations 4
GAMG: Solving for p, Initial residual = 0.00246394853928313, Final residual = 1.12631494042653e-05, No Iterations 3
time step continuity errors : sum local = 6.29921027180131e-09, global = -1.88504368307571e-10, cumulative = -8.84159167321985e-06
ExecutionTime = 168453.17 s ClockTime = 168462 s

Time = 4796

smoothSolver: Solving for Ux, Initial residual = 8.27636882199578e-07, Final residual = 1.17828578411332e-16, No Iterations 4
smoothSolver: Solving for Uy, Initial residual = 0.000420804884209455, Final residual = 1.15685143238247e-14, No Iterations 4
smoothSolver: Solving for Uz, Initial residual = 0.000456888391577294, Final residual = 1.21400758956957e-14, No Iterations 4
GAMG: Solving for p, Initial residual = 0.0024639398031611, Final residual = 1.12633772620776e-05, No Iterations 3
time step continuity errors : sum local = 6.29377884031872e-09, global = -1.88393757795753e-10, cumulative = -8.84178006697764e-06
ExecutionTime = 168488.27 s ClockTime = 168497 s

Time = 4797

smoothSolver: Solving for Ux, Initial residual = 8.26829853129783e-07, Final residual = 1.17743398700947e-16, No Iterations 4
smoothSolver: Solving for Uy, Initial residual = 0.000420809949537425, Final residual = 1.15680804424955e-14, No Iterations 4
smoothSolver: Solving for Uz, Initial residual = 0.000456883224658707, Final residual = 1.21399060586643e-14, No Iterations 4
GAMG: Solving for p, Initial residual = 0.00246393105191049, Final residual = 1.12636035684504e-05, No Iterations 3
time step continuity errors : sum local = 6.2883465087325e-09, global = -1.88280746920101e-10, cumulative = -8.84196834772456e-06
ExecutionTime = 168523.53 s ClockTime = 168532 s

The solver tolerances for both U and p are 1e-7 and 0.01. Ux is in the range of 0..20e-3 m/s.

Could anybody please give me some feedback if my assumptions are correct? I'd appreciate your help very much!!

Have a nice day, kind regards
Bengt


All times are GMT -4. The time now is 08:56.