Increasing Part Geometric Accuracy in High Speed Machining using Cascade Iterative Learning Control

Abstract

High speed machining provides high productivity and low machining cycle times. Post machining, there can exist differences between desired and measured part geometry due to tool deflection induced from higher feedrates. Reducing the feedrate leads to an increase in machining time. Using predicted drive responses on a virtual CNC with an integrated surface location error model, this research is the first time Iterative Learning Control (ILC) has been applied to reduce part geometry errors from tool deflection. Validation machining trials demonstrated that the ILC scheme improved machining performance whilst maintaining machining times when compared to a baseline part program.