POLYFLOW - trouble with the rate of deformation tensor (D11+D22 does not vanish)
|
When running a 2D (or 3D) incompressible fluid flow case with POLYFLOW, it can be apparently found that the trace of the rate of deformation tensor (the sum of the diagonal components) does not vanish. Depending on the problem definition, this may be due to either round-off errors or just a simple permutation of the tensor component names. This should be easily checked on a Poiseuille flow, e.g. with the velocity along the x-direction and the gradient along the y-direction. Two scenarios: - If the component D12 is not vanishing (and reports the expected result), the calculation is OK, and one deals only with possible round-off errors. - If the component D12 does not report the expected result, the trouble originates from a permutation of the tensor component names. The solution is: * 2D flows: the names D11, D12, D22 respectively contain D11, D22, D12; * 3D flows: the names D11, D12, D22, D13, D23, D33 respectively contain D11, D22, D33, D12, D23, D13. |
||
|