Evaluation of virtual pivot and TCP misalignment error by contact force perception for bonnet polishing machine tool

The misalignment errors of the virtual pivot and TCP of the assembled five axis bonnet polishing machine tool should be evaluated for subsequent processing. However, standard R-test equipment is expensive and cannot simultaneously evaluate the TCP misalignment error. This study presents a low-cost method for evaluating misalignment errors utilizing bonnet contact force perception. Experiments and finite element simulation analysis show that the bonnet cap possesses good linear contact stiffness under 0–1-mm tool offset range. The combination of bonnet rotation measurement and low-pass filter is utilized to avoid the consideration of TCP off-axis error. The iterative identification process can well reduce the impact of stiffness model inaccuracy on the identification result. A two-step identification strategy is adopted based on the analysis of Jacobian coefficient of DH error model, and the influence of initial tool offset is further studied. After compensation, the fluctuation amplitude of bonnet contact force decreased from about 4 N to 0.2 N under different tool orientations, and the uncompensated contact force is well predicted. The evaluation results show that the errors of bonnet tool with polishing pad adhered requires more attention than VP misalignment error.