Hello,
I am using a weak contribution to couple two physics. I know that the weak contribution adds a term to the weak formulation in the form of an integration over a line, so that we have 
. This would weakly set the variable H to be equal h, since the integral above would be set to zero. However, when I calculate the integral in postprocessing, I do not obtain zero. If I take the integral over the whole boundary, I get something close to zero (of the order of 1e-5), but if I take the integral in one mesh element edge, I do not get zero. I thought that the weak formulation was applied by element so that the weak contribution would be satisfied by element, but it doesn't seem that way here. Am I missing something?
Thanks,
Alex