As a cost-effective hole production technique, friction drilling is widely used in industrial and automotive manufacturing. Compared with the traditional bolted connection, it enables the fastening of thin metal sheets and thin-walled tubular profiles. Friction drilling results in higher thread length and joint strength, thus better fulfilling the demand for lightweight structures. However, in the numerical simulation of friction drilling, the traditional finite element method encounters difficulties caused by the extreme deformation and complex failure of the material. A large number of elements are usually deleted due to the failure criterion, which significantly reduces the solution accuracy. The development of meshless methods over the past 20 years has alleviated this problem. Especially the Smoothed Particle Galerkin (SPG) method proposed in recent years and incorporating a bond-based failure mechanism has been shown to be advantageous in material separation simulations. It does not require element removal and can continuously evolve each particle's information such as strain and stress after the material failure. Therefore, the SPG method was used in this research for the simulation of frictional drilling of HX220 sheet metal. First the particle distance and the friction coefficient were varied to investigate the applicability of the SPG method to the friction drilling process. Predicted and experimental results were compared and found to be in high agreement. Furthermore, the influence of input parameters, such as sheet thickness, feed rate and rotational speed, on axial force as well as torque of the tool and the surface temperature of the workpiece during friction drilling was investigated numerically.