Added splitting the forces into their contribution. There seems to be a sign mistake in F_tau and an overflow for Fp of the nonisothermal model.