I'm working on modeling frame dragging effects in a binary pulsar system (specifically PSR J0737-3039) for my PhD research and I've hit a wall with the calculations. I need to account for both the Lense-Thirring effect from each pulsar's spin and the orbital frame dragging from their mutual rotation. My current approach uses the weak-field approximation of the Kerr metric, but I'm getting inconsistent results when trying to combine the individual frame dragging contributions. The timing residuals I'm calculating seem to be off by an order of magnitude compared to observational data.
So going to my specific question, What's the proper way to sum the frame dragging contributions from both the spin and orbital angular momenta in a system where both objects are rapidly rotating neutron stars? I've tried using superposition but I suspect this breaks down due to the nonlinearity of general relativity. Has anyone tackled a similar problem or can point me to relevant literature? I'm particularly interested in approaches that work in the strong-field regime since these pulsars are only about 900,000 km apart.