My question is about implementing an integral constrain to transform my frame of reference to one determined by the solution such that the solution appears stationary.
I'm using a phase-field model for a circle which moves along the x axis. The centroid of the circle is given by \int{ x[0]phasedx} / \int{phase*dx}. In FEniCS this is calculated by
assemble(x[0]*phase*dx(mesh))/assemble(phase*dx(mesh))
I would like to add a convective term to my phase field equation such that the centroid is fixed (the speed at which it moves is not known). Can anyone point me in the right direction please?
Thanks,
Mike
