Online interpolation of 5-axis machining toolpaths with global blending
Online interpolation of 5-axis machining toolpaths with global blending
复制标题
DOI：
10.1016/j.ijmachtools.2022.103862
复制
发表时间：
2022-02
影响因子：
14
通讯作者：
Shingo Tajima;B. Sencer
中科院分区：
文献类型：
--
作者：
Shingo Tajima;B. Sencer
现代高速5轴加工刀具路径由密集离散的刀具中心点（TCP）和刀具方向矢量（TOV）组成。本文提出了一种全局插补技术，可以混合非常密集的离散5轴刀具路径内的用户定义的混合公差和插补他们在实时使用最小的计算费用。使用有限脉冲响应（FIR）低通滤波代替几何样条来混合TCP和TOV，并平滑地对它们进行插值，以生成同步的5轴运动。基于轨迹几何学、刀具运动学和FIR滤波器动力学模型，建立了TCP和TOV的全局融合误差模型。这些全局混合误差通过调节加工进给率来限制。换句话说，调整加工进给速率，即减小和增大，以控制沿5轴加工刀具路径的全局混合精度沿着。提出了一种加窗算法来解决这一新的进给速度调度问题。由于所提出的插补方法是基于数字滤波（卷积），它使在线（实时）插补的密集离散的5轴加工刀具路径的实时，并在同一时间内提供控制的频谱内插轨迹。建议的全球混合策略提供了显着减少加工周期时间相比，传统的本地混合技术，并产生更平滑，频率控制的轨迹与一小部分的计算工作量相比，基于样条的插值。说明性的例子和实验结果提供展示的功能和验证所提出的技术的有效性。
Modern high-speed 5-axis machining toolpaths consist of densely discretized tool center points (TCP) and tool orientation vectors (TOV). This paper presents a global interpolation technique that can blend very densely discretized 5-axis toolpaths within user defined blending tolerance and interpolate them in real-time using minimal computational expense. Instead of geometric splines, Finite Impulse Response (FIR) low-pass filtering is used to blend TCP and TOV and interpolate them smoothly for synchronized 5-axis motion generation. Global blending errors of TCP and TOV are modeled based on path geometry, tool motion kinematics and dynamics of the FIR filter. These global blending errors are confined by modulating the machining feedrate. In other words, machining feedrate is adjusted, i.e. reduced and increased, to control the global blending accuracy along the 5-axis machining toolpath. A windowing scheme is proposed to solve this new feedrate scheduling problem. Since the proposed interpolation method is based on digital filtering (convolution), it enables online (real-time) interpolation of densely discretized 5-axis machining toolpaths in a real-time and at the same time provides control over the frequency spectrum of interpolated trajectories. The proposed global blending strategy delivers significant reduction in machining cycle-time as compared to conventional local blending techniques and generates smoother, frequency-controlled trajectories with a fraction of the computational effort required as compared to spline-based interpolation. Illustrative examples and experimental results are provided to showcase functionality and validate effectiveness of the proposed technique.