Abstract:To obtain the effect of change in design elements of curves on vehicle’ track and speed while driving in curve cutting, two indexes were established to measure cutting effectiveness, i.e. speed increment and track radii increment at mid curve. Then, process of a car driving in curve cutting and lane keeping pattern on test curves was simulated by a “roadway - driver - vehicle” virtual driving system. And the relationships among curve radii R, deflection angle ΔA, roadway width WR, curve length Lh and cutting effectiveness were analyzed. The results show that cutting effectiveness reaches maximum at ΔA= 20deg, and it would decline as ΔA increase. For curves with ΔA less than 10deg, track radii in curve areas is larger than the curve radii even from lane following pattern. And only when ΔA exceeds 20deg, track radii can be in accordance with curve radii. There exists a critical angle ΔA0C, widening the pavement can result in an increase in cutting effectiveness only when ΔA <ΔA0C. Cutting effectiveness would decline as curve radii increase, and the larger ΔA, the faster in decline. For curves with 30 m radii, no matter how a ΔA and WR it has, an obvious speed increment can be gained by cutting. Cutting effectiveness would decrease as Lh increase, when Lh exceeds a critical length, the effectiveness being zero.