In recent years， ice crystal icing has emerged as a prominent and challenging research area in aircraft and engine anti-icing. This paper presents a computational model for accurate prediction of the ice crystal icing process. The proposed model encompasses a trajectory model that accounts for irregular ice crystal shapes， an adhesion model， an erosion model， and a thermodynamic model for icing. The multi-step method is employed to simulate the ice accumulation process， involving the updating of the airflow field， ice crystal particle trajectories， and thermodynamic calculations at each time step. Additionally， the dynamic mesh technology is utilized to update the geometric boundaries of the ice formation. The research findings demonstrate that the application of the multi-step method enables stable simulation of ice formation. The stability of the ice shapes is attributed to the combined effects of reduced adhesion on the surface of ice crystals and increased erosion on pre-existing ice， ultimately leading to a dynamic equilibrium between the adhesion and erosion processes. It should be noted that the proposed model and computational approach in this paper are specifically suited for low melting rate conditions， which can guide the precise simulation of three-dimensional structural ice crystal icing in future investigations. While high melting rate scenarios would necessitate consideration of the downstream flow of ice in the future.