GRIDING DEFECT PREDICTION DEVICE

To provide a grinding defect prediction device which can predict that a surface roughness of a work-piece deteriorates and shape collapse thereof occurs during grinding, and can improve grinding yield.SOLUTION: This grinding defect prediction device comprises a processing defect prediction part 58 which predicts processing defects during grinding on the basis of a change in a first signal strength SP1 in a first frequency band B1 relating to crashing of abrasive grains 38 of a grinding wheel 14, or a change in a signal strength ratio SR(=SP1/SP2) of the first signal strength SP1 to a second signal strength SP2 in a second frequency band B2 relating to friction vibration or elastic vibration that occurs by contact between the abrasive grains 38 of the grinding wheel 14 and a work-piece 44. Therefore, the prediction device can predict processing defects for grinding of the work-piece 44 prior to deterioration of a surface roughness of the work-piece 44 during grinding. Thus, it is reduced that the work-piece 44 during grinding becomes a defective product, and therefore, grinding yield can be improved.SELECTED DRAWING: Figure 1 【課題】研削加工中に被削材の表面粗さが悪化したり、形状崩れが発生したりすることが予測でき、研削歩留りを向上させることができる研削加工不具合予知装置を提供する。【解決手段】研削ホイール１４の砥粒３８の破砕に関連する第１周波数帯Ｂ１の第１信号強度ＳＰ１の変化、または、第１信号強度ＳＰ１と研削ホイール１４の砥粒３８と被削材４４との接触により発生する摩擦振動或いは弾性振動に関連する第２周波数帯Ｂ２の第２信号強度ＳＰ２との信号強度比ＳＲ（＝ＳＰ１／ＳＰ２）の変化に基づいて、研削加工中の加工不具合を予測する加工不具合予知部５８を備えているので、研削加工中の被削材４４の表面粗さの悪化に先立って、被削材４４の研削加工中の加工不具合が予測できる。これにより、研削加工中の被削材４４が不良品となることが低減されるので、研削歩留りが向上する。【選択図】図１