The current function to compute the crack opening loop in the P_RAJ algorithm is too slow for purpose. This is due to the many small case distinctions the FKM nonlinear guideline is demanding for the calculation. So the only way to gain performance is to rewrite it in cython.