Papers by Author: Shu Wen Lin

Abstract: The geometric error of machine axis can be equivalent to the differential movement, regarded as a differential operator based on its ideal position. Thus, a new modeling method for multi-axis CNC machines based on differential transform theory is proposed in this paper. Firstly, the workpiece coordinate system is selected to observe the error of tools cutter location and orientation. Then, the kinematics chain of the machine will be consolidated into one, and the closed chain can be changed to open series type. The transform matrix is adopted to convert the differential operator between the different basis, because of the geometric errors have a different basis between measurement and modeling. Eventually, the general geometric error model for multi-axis machines is established after a XYFZ type three-axis machine is studied in detail.

Abstract: In this paper, the jerk-limited feedrate planning is used to adaptively change feedrate with consideration of machine dynamics in the sensitive corners. The curvature is introduced to detect sensitive corner by the look-ahead module to avoid the situation that the curvature is zero. The parameters of every deceleration stages are solved with a series of formula first, and then the feedrate profile is discretized with consideration of the deceleration time, which must be an integer multiple of the interpolation sample time. In addition, the deceleration start point estimation method is described in detail, and a number of deceleration terminal speeds are added into the discrete feedrate profile to drive the machine tool to the right position.

Abstract: This paper proposes an improved adaptive NURBS curve interpolation scheme, which keeps a constant feedrate most of the time and adaptively changes feedrate with consideration of machine dynamics. It reserves the advantages of adaptive planning, which has a shorter machining time and a better machining performance, and removes abrupt feedrate changes at the beginning and end of sensitive corner. Furthermore, this method reuses the previous interpolation values during look-ahead planning rather than discard them.