Discrepancies in spin tracking simulations with vertical quad offset and radial magnetic field

[vcilento]

I am reaching out to discuss some discrepancies I have encountered in the spin tracking simulations using BMAD. Specifically, I modeled the lattice with the same effect but under different settings.

My goal was to simulate the spin buildup for a lattice with magnetic quadrupoles in the presence of a radial magnetic field located in a quadrupole in the middle of the machine, along with horizontal betatron oscillations.
I introduced the radial magnetic field in the lattice file (attached with the name lattice_radial_field.bmad) as bl_vkick. Additionally, I added two magnetic correctors (VK) before and after the quadrupole to correct the vertical closed orbit distortion in the rest of the ring.
I then compared these results with those obtained by applying a vertical offset (y_offset) of 0.1 mm (attached with the name lattice_quad_offset.bmad).
Theoretically, both effects should produce the same vertical spin buildup (attached particle_tracking.init). However, as you can see from the attached plots, the results differ between the two cases.
I also performed simulations for different values of the action variable, J_x, to see if there was a systematic trend over time, as expected and confirmed by the simulations.

To clarify, the radial field value was computed based on the vertical offset to achieve the same effect, with a 0.1 mm offset corresponding to an integrated field of 8.17e-6 Tm.
input files.zip

[DavidSagan]

@vcilento One possibility is that the quadrupole with the kick has some sensitivity to the number of tracking steps num_steps. What happens if you vary the number of steps?

[vcilento]

@DavidSagan thanks for the input, I tried to vary the num_steps and I have no change in the results.