Robotic Conformance Grinding Modeling, Control and Optimization

In precision manufacturing, there are strict accuracy requirements regarding the final dimensions of workpieces. Conformance grinding, which refers to a grinding process that can remove a specific amount of material from a specific region of a workpiece, is typically employed as the final machining process to adjust dimensions to fit into their accuracy tolerances. For free-form workpieces, conformance grinding requires high expertise and skill, and currently can only be performed by manual operations, which leads to low efficiency and inconsistent workpiece quality.;In this research, we propose a comprehensive system for pre-process and in-process control of a robot for the conformance grinding of workpieces. The system adopts a framework that can be readily implemented into commercial industrial robot systems. In this framework, the robotic grinding path, the robot velocity and contact force along the path are planned in the pre-process phase. In the in-process phase, the robotic system is controlled by a position-force controller to track the planned grinding path, robot velocity, and contact force to accomplish the grinding.;An offline programming and pre-process calibration methodology is employed to generate the robotic grinding path. A system error compensation method is proposed to enhance the path accuracy. Dynamic models of the robot, the grinder and the grinding process are built. A target tracking controller is designed for the control of the robot position and the contact force during grinding. A material removal model, consisting of a superposition sub-model and a shape-dependent sub-model, are proposed and experimentally verified to precisely model the material removal rate of the grinding process. The planning of grinding parameters for the conformance grinding process is formulated as an optimization problem. Parameters are solved by minimizing the difference between expected material removal and actual material removal. The effectiveness of this planning strategy is demonstrated on a comprehensive robotic grinding simulation platform.