This paper assesses factors that impact the simulation accuracy of a 2D FE transient axis-symmetric model of a Thomson Coil actuator. 3D FEA is used to show that geometric differences between a planar coil and the intrinsic assumptions of a 2D axial model may result in important differences in coil inductance. It is then shown that inductance and resistance compensation of the 2D model can be used to produce an accurate prediction of the TC performance. Detailed parameters of a prototype TC test system are used as inputs in the compensated 2D axial model and excellent agreement is observed between 2D FE simulations and experimental results. It is also shown that armature vibration modes explain the presence of apparent speed oscillations in the experimental results not present in numerical simulations. The compensated 2D axis-symmetric model shows good accuracy compared with experimental results for the investigated scenarios, even when armature flexing is considered.