|
[Sponsors] |
October 4, 2020, 03:40 |
Cloud case differs from local case
|
#1 |
Member
Bob Tipton
Join Date: Apr 2020
Posts: 34
Rep Power: 6 |
After weeks of experimenting I was able to reproduce a reference solid hemisphere's Cd to 0.020%.
I did this on my laptop. I then tested the same case on my laptop using mpi and 4 cores. The results are identical to all decimal places. I then moved the mpi case to a cloud service (KaleidoSim). The cloud service is has the same version number of openFoam that I do. The cloud service is off by 1.025% and doesn't converge. I checked the forces output file and the cloud service differs on step #1. I did this test preparatory to running a lot of trade studies so that I could have reasonable assurance my work was correct and reproducable. What settings, options etc. could cause such a discrepency? Yes, I did an exact zip and upload of my local case sans outputs. Both are running rev 1912. I know it's out of date, but it the configuration that works best for my needs. Local forces file - # Force # CofR : (0.0000000e+00 0.0000000e+00 0.0000000e+00) # # Time (total_x total_y total_z) (pressure_x pressure_y pressure_z) (viscous_x viscous_y viscous_z) 1 (1.1153609e+01 2.1528108e+00 2.1528031e+00) (1.1131152e+01 2.1531862e+00 2.1531784e+00) (2.2456886e-02 -3.7546594e-04 -3.7535068e-04) 2 (1.2810712e+01 3.2416530e+00 3.2415862e+00) (1.3145809e+01 3.2409055e+00 3.2408373e+00) (-3.3509685e-01 7.4750946e-04 7.4895280e-04) 3 (4.8591313e+00 3.3006178e+00 3.3005076e+00) (5.2341243e+00 3.2997321e+00 3.2995069e+00) (-3.7499292e-01 8.8572079e-04 1.0007073e-03) 4 (-3.7884967e+01 2.7486277e+00 2.7511909e+00) (-3.7596978e+01 2.7647277e+00 2.7650021e+00) (-2.8798872e-01 -1.6099998e-02 -1.3811225e-02) 5 (-7.5107130e+01 1.9971352e+00 1.9984947e+00) (-7.4980961e+01 2.0141339e+00 2.0137086e+00) (-1.2616914e-01 -1.6998650e-02 -1.5213851e-02) 6 (-7.3778802e+01 1.2963405e+00 1.2966028e+00) (-7.3807799e+01 1.3120553e+00 1.3117404e+00) (2.8996651e-02 -1.5714811e-02 -1.5137587e-02) Cloud file: # Force # CofR : (0.0000000e+00 0.0000000e+00 0.0000000e+00) # # Time (total_x total_y total_z) (pressure_x pressure_y pressure_z) (viscous_x viscous_y viscous_z) 1 (1.1153825e+01 2.2716479e+00 2.2716248e+00) (1.1131368e+01 2.2720232e+00 2.2720001e+00) (2.2456466e-02 -3.7531164e-04 -3.7526665e-04) 2 (1.2806504e+01 3.2872884e+00 3.2873017e+00) (1.3146464e+01 3.2864989e+00 3.2865089e+00) (-3.3996042e-01 7.8950895e-04 7.9272825e-04) 3 (4.7604402e+00 3.1481441e+00 3.1478024e+00) (5.1414958e+00 3.1472298e+00 3.1467241e+00) (-3.8105569e-01 9.1430933e-04 1.0782581e-03) 4 (-4.0348198e+01 2.2899967e+00 2.2917158e+00) (-4.0055823e+01 2.3057917e+00 2.3059401e+00) (-2.9237504e-01 -1.5794964e-02 -1.4224236e-02) 5 (-7.7588518e+01 1.2172655e+00 1.2171502e+00) (-7.7462248e+01 1.2330427e+00 1.2324310e+00) (-1.2627011e-01 -1.5777169e-02 -1.5280799e-02) 6 (-7.5514934e+01 2.7729261e-01 2.7652271e-01) (-7.5546921e+01 2.9218414e-01 2.9157532e-01) (3.1987429e-02 -1.4891532e-02 -1.5052609e-02) |
|
Tags |
mpi cloud |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[solidMechanics] Support thread for "Solid Mechanics Solvers added to OpenFOAM Extend" | bigphil | OpenFOAM CC Toolkits for Fluid-Structure Interaction | 686 | December 22, 2022 10:10 |
AMI interDyMFoam for mixer nu problem | danny123 | OpenFOAM Programming & Development | 8 | September 6, 2013 03:34 |
dynamic Mesh is faster than MRF???? | sharonyue | OpenFOAM Running, Solving & CFD | 14 | August 26, 2013 08:47 |
[blockMesh] BlockMeshmergePatchPairs | hjasak | OpenFOAM Meshing & Mesh Conversion | 11 | August 15, 2008 08:36 |
Free surface boudary conditions with SOLA-VOF | Fan | Main CFD Forum | 10 | September 9, 2006 13:24 |