Dear Dr. Amine,
With your help I am able to run the UDF after allocating the gradients using either of the above stated methods. And udf compile nicely, and runs without any segmentation fault error. But, my values are not realistic, some random distance (a very large number like 1.797693143E+306) i am getting.
I was wondering if you could tell me whether your results ( from that small case) are realistic? It would give me a hope to further work on the udf to get realistic values. But, if you are also getting some unrealistic numbers from this macro then I will think of some alternative.
I went through the paper your mentioned where you have talked about node based and cell based smoothing of interface to compute area by Mag(Grad(alpha) method. I will compare the results of gas liquid mass transfer from my geometrical approach with the grad(alpha) formulation combined with node/cell based smoothing.
Waiting for your postive response.