The vehicle-terrain interaction model plays a critical role in off-road vehicle driving visual simulations. This paper present a new terrain deformation model for simulating vehicle-terrain interaction. The model simplifies the Bekker’s classic terramechanics model, and calculates terrain deformation according to the vehicle load, velocity, tire size, and soil concentration in real time. Simulation results show that the simplified terramechanics model is credible and real-time. This method can simulate vehicles on common types of soft terrains such as clay, sand, and be used in visual simulation. Introduction Off-road vehicle (including wheeled and tracked) has a good passing ability and been widely used in the modern military, agriculture, construction and other fields. But the driver training for the vehicle mostly uses real vehicles, that is low efficiency, high power consumption, pollution of the environment, long training period, so more and more vehicle driving simulators are developed and utilized. But most vehicle driving simulators is developed by visual simulation software such as OpenGVS, Vega Prime, Unity3D. There are still many problems, such as lack of real-time dynamic interaction between wheel and ground, lack of real-time terrain deformation, lack of immersion and realism. And most vehicle simulators only make models for road and other man-made road surface, and not consider the condition that vehicles drive on off-road. In addition, as the development of vehicle-terramechanics and multi-body dynamics, domestic and foreign scholars have done a lot of research on vehicle dynamics. By establishing completed and detailed vehicle model in DADS, ADAMS / ATV, RecurDyn and other multi-body simulation software, the researchers can simulate the dynamic behaviors in the real driving process. But a few more degrees of freedom, system complexity, dynamic response slow, the calculation speed not achieving the real-time requirements of visual simulation, so that models are not conducive to the development of visual simulation program. Although these are quite good research for virtual reality simulation technology and vehicle-terramechanics in domestic and foreign, the development work effectively combining of both in a real-time interactive simulation platform still needs to be further deepened. To solve the above problems, by analyzing the interactive force between driven wheels and soft terrain, on the basis of Bekker’s pressure model, Wong’s pressure theory and Janosi’s shear theory the vehicle-terramechanics semi-empirical model has been simplified and the simplified model keep authenticity and real-time. So that it can be applied to virtual driving off-road vehicles and other visual simulation fields. The Vehicle-terramechanics Semi-empirical Model The typical representative of vehicle-terramechanics semi-empirical method is Bekker’s study. This method base the classical soil mechanics (such as the doctrine of Coulomb, Terzaghi, Rankin and Plante) and present a series of semi-empirical soil-vehicle system engineering model (formula) through a large number of laboratory simulation experiments[2].Semi-empirical model characterize the physical properties of terrain. The relationship between track and terrain can be represented by a semi-empirical and provides a theoretical basis for vehicle’s performance prediction. This method has been the most applications in vehicle-terramechanics research. The model can calculate the wheel International Conference on Materials Engineering and Information Technology Applications (MEITA 2015) © 2015. The authors Published by Atlantis Press 983 sinkage, the horizontal force, torque and so on, after knowing the parameters of terrain and wheel vertical load. The interaction force analysis between driven wheels and soft terrain is shown in Fig. 1. The interaction force performance as continuous normal stress σ and shear stress τ . According to Bekker’s pressure model[2] and Wong’s pressure theory[6][7], the formula of normal stress as follows: 1 1 k ( ) ( k ) (r(cos cos )) c b φ σ θ θ θ = + ⋅ − (1)