Hello, This is regarding the defination of "test functions" in a vector values case. In the documentation of "Handling vector valued problems", it is mentioned that the solution vector is defined as: U=(u,p) transposed, where u is vector
(Question-1): Does this mean that the DOFs are arranged (at least conceptually) as: U = (u1,v1,u2,v2 ... u4,v4,p1...p4) transposed? (for a 4 noded 2D element) It is also mentioned that the test functions 'V' is chosen as vector valued: V=(v,p) transposed, where v is a vector (Question-2): Does this mean that the shape factors are arranged as: V = [ N1 , N1 , N2 , N2,...,N4,N4, Np1 ,...Np4.] transposed ? (N1.. is the shape factor terms for velocity and Np is the shape factor terms for pressure) This seems to make sense when we see that the return value type of fe_values[velocities].divergence (i, q) is a double,as div.U will be scalar.... BUT if V is simply V=[N1,N1.Np1.....] then V.U = N1*u1 + N1*v1 + Np1*p1 + N2*u2 + N2*v2 + Np2*p2.....!!!! which does not make sense...please explain... Abhishek
_______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
