You should diagnose your forces and factors affecting them. As an example, in orthogonal cutting, the thrust force is highly dependent of frictional stress and pressure on the tool meanwhile for cutting force the flow stress plays an important role.

Since you have 50% error, you should look for significant factors that can change your results significantly. My suggestion is that first make sure you did everything right (J-C parameters, damage parameters, contact modeling, etc), then try to adjust your simulation by your solution method. (explicit/implicit, kinematic/penalty, ALE/mesh refinement, etc). 

Thanks Ali, indeed I tried changing mesh, friction as well as workpiece clamping, only hanging the workpiece clamping did improve the thrust force results slightly but not significantly, my initial outcome is that the JC parameters are the main factor causing such low thrust force, however for Al2024 there are only several JC damage parameters which we can use for FE modelling. 

Can you upload your picture of simulation? It would be better if you make pic with the close view of the tool-wp interaction. 

I still don't know your interaction properties. Are you willing to model the process with the most complete(fully thermo-mechanical properties) solution or you would like to simplify it? 

I am using General contact and coulomb friction as penalty for contact between the drill and the workpiece, I am not using temp-disp in the model, so the model does not account for temperature effects which could also be a reason for underestimated thrust force since the model ignores the effect of temperatures propagated at the drilling zone.

hi there, How u got trust force and torque in Abaqus. How to measure them?

Thank you