Validation of the implementation of angle-dependent SIDM based on the effective drag force and momentum diffusion approach (fSIDM scheme). Each panel shows the angular distribution of beam particles within the deflection test set-up at three different times t (solid lines), compared to a Gaussian Molière distribution with width related to the transfer cross section (dashed lines), see Eq. (27). Here we assume beam-target scattering without recoil. The two panels correspond to two values of the transfer cross section, differing by a factor of ten. The distribution is expected to be identical at times rescaled by a factor of one-tenth, which we observe to be the case.