Development of autonomous NC machine tool to realize NC programless machining operation for advanced manufacturing systems

研究了为实现先进制造系统的数控无编程加工，开发自主数控机床所需的功能。取得了以下研究成果.在预加工过程中预测切削力和加工误差的加工仿真虚拟加工仿真器（VMSim）已开发用于此目的。VMSim可以提前预测整个加工操作的切削力和加工误差。通过使用该功能，可以在自主数控机床的前处理加工中发现隐藏在刀具路径和切削条件规划中的问题.基于仿形铣削原理的真实的实时刀具轨迹生成技术研究开发了一个虚拟仿形铣削系统，该系统能够真实的实时生成刀具轨迹。虚拟仿形铣削系统检测虚拟仿形器和虚拟主模型之间的碰撞，并生成刀具路径以跟随雕刻表面，用于在过程中加工期间进行粗切削。实验表明，虚拟仿形铣削系统可以在普通数控机床上实现无程序数控加工.自适应控制和反馈算法避免加工故障研究了一种用于检测加工状态的监控系统。在该系统中，主轴和进给电机的转矩可以估计数控控制信号的基础上控制理论。在不增加传感器的情况下，可以检测出切削力矩或切削力，但需要进一步研究消除电机摩擦力矩的方法，并研究了避免加工故障的反馈算法。通过使用上述虚拟仿形铣削系统，除了进给速度之外，还可以改变轴向和径向切削深度，以避免加工故障。利用VMSim仿真软件验证了算法的有效性。

Required functions to develop an autonomous NC machine tool, which realizes NC programless machining operation for advanced manufacturing system, have been studied. Following research results were obtained.1. Machining simulation to predict cutting force and machining error in pre-process machiningA Virtual Machining Simulator (VMSim) has been developed for this purpose. The VMSim makes possible to predict cutting force and machining error for whole machining operation in advance. By using this function, the problem hidden in planning of tool paths and cutting conditions can be detected in pre-process machining for an autonomous NC machine tool.2. Real time tool path generation to optimize cutting conditions during in-process machiningA virtual copy milling system, which generates tool paths in real time based on copy milling principle, has been developed. The virtual copy milling system detects collision between a virtual tracer and a virtual master model, and generates tool paths to follow a sculpture surface for rough cutting during in-process machining. An NC programless machining operation can be achieved experimentally by the virtual copy milling system using conventional CNC machine tool.3. Adaptive control and feed back algorithm to avoid machining troubleA monitoring system to detect machining condition has been studied. In the monitoring system, rotational torque of both spindle and feed motors can be estimated from NC control signals based on the control theory. It is possible to detect cutting torque or cutting force without additional sensors, but further investigation is required to eliminate frictional torque of the motors.Furthermore, a feed back algorithm to avoid machining troubles has been studied. By using the virtual copy milling system mentioned above, even the axial and radial depths of cut beside the feed speed can be changed to avoid machining troubles. The effectiveness of the algorithm is verified using the VMSim mentioned above.