wallShearStress VS wallGradU
I am using simpleFoam solver with k-Epsilon model to simulate 2D backward-facing step, and comparing the numerical results with the experimental data from NASA in 1990. (Here are the link to the exp. data : http://www.grc.nasa.gov/WWW/wind/val.../bstepdata.txt)
I tried to compare the value of Cf defined as 2tao/rho*u*u, with tao = mu*(du/dy). I used 2 utilities wallShearStress and wallGradU to calculate Cf, here are the phenomenons I found:
For wallShearStress, if one wants to compare to the experiemental data, then a minus sign must be put in front. After adding the sign, the numerical results can be compared to the experimental data.
For wallGradU, the sign is fine, however, the value is only half of the value from wallShearStress, which means the data can not be comparable.
I check the definition of wallShearStress, and found that inside twoSymm(T), there is a factor 2 in the construction of the Tensor(only the upper triangular of the Tensor is counted), it may be the reason for the 2 times difference.
Now I am a little bit torn in between these two utilities: wallShearStress is comparable but with a wrong sign; the sign of wallGradU is OK but incomparable, for a simple Benchmark test as backward-facing step, k-Epsilon model seems to work well.
Here comes my question: when I (or someone else) wants to calculate Cf, which utility can be used?
Btw, the OF Version is 2.1.1
Many thanks in advance!
|All times are GMT -4. The time now is 19:48.|