Milling force identification from acceleration signals using regularization method based on TSVD in peripheral milling

Milling forces are important signals that can be used to monitor the condition of machine operating and milling process. In general, milling forces can be measured by special equipment, for example, Kistler dynameters. However, dynameters are very expensive and require particular geometry of workpiece, which causes that dynameters are limited in research laboratories rather than actual production. In this paper, a force identification method in frequency domain, the regularization method based on the truncated singular value decomposition (TSVD), is employed for milling force reconstruction using acceleration signals in peripheral milling process. The two acceleration sensors are mounted on the spindle box in the feed and cross-feed directions for reconstructing the milling forces. The regularization method based on TSVD is adopted for inverse problem in force identification because of its high identification accuracy, which can possess ill-conditioned matrixes well. Finally, the method is validated by numerical simulation and cutting experiments. In the experiment, the milling forces and acceleration signals from spindle box are acquired synchronously. Then, the frequency response functions between milling tool nose and acceleration sensors in the feed and cross-feed directions are obtained by impact tests. With the acceleration signals and frequency response functions, the milling forces can be reconstructed. Identified results show that the calculated forces and the measured forces are in good agreement on the whole, which verify the effectiveness of the employed method on milling force identification.