Prediction of cumulative surface location error at the contact zone of in-process workpiece and milling tool

•During continuous radial immersion, the conservative stability lobe diagram (SLD) was established efficiently and the convergent radial depth of cut was determined, which makes it convenient to select stable machining parameters.•Based on the impact tests of workpiece, the simulation in ansys was modified and offset so that the mode shapes of workpiece which were closer to the clamping condition was obtained.•The cumulative surface location error (CSLE) model considering the dynamic characteristics of milling tool and in-process workpiece was developed during continuous radial immersion.•With the contour of machined workpiece acquired, the surface location error and the corresponding angles at different cutter locations were measured and analyzed. The surface location error (SLE) is used to evaluate the machining error caused by forced vibration at the contact zone during the stable machining. When the workpiece was machined continuously, the convergent SLE would change with the dynamic response of the workpiece as well. In this paper, we first established the conservative stability lobe diagram (SLD) caused by regenerative effect, and excluded the instability induced by cumulative effect with continuous radial immersion. An assumption was made that multiple immersion-by-immersion was equivalent to one immersion with the same material removal, and then the in-process workpiece (IPW) dynamics was obtained by means of structural dynamic modification. Finally, we used the improved third-order full discretization method to obtain the predicted cumulative surface location error (CSLE) with considering the dynamics of milling tool and workpiece simultaneously. Based on the stable machining parameters, the cutting tests and measurement were performed. The experimental results showed that the proposed model could effectively predict the distribution of CSLE. Under the same spindle speed, the CSLE at different cutter locations were not the same and the difference was obvious, which indicated that the dynamics from the workpiece played an important role in the formation of CSLE during the continuous machining process and should be included in the predicted model. [Display omitted]