Iterative Learning Method for Drilling Depth Optimization in Peck Deep-Hole Drilling

Abstract

Due to the enclosed chip evacuation space in deep hole drilling process, chips are accumulated in drill flutes as drilling depth increases, resulting in the increase of drilling torque and lead to drill breakage. Peck drilling is a widely used method to periodically alleviate the drilling torque caused by chip evacuation; the drilling depth in each step directly determines both drill life and machining efficiency. The existing drilling depth optimization methods face problems including low accuracy of the prediction model, the hysteresis of signal diagnosis, and onerous experiments. To overcome these problems, a novel drilling depth optimization method for peck drilling based on the iterative learning optimization is proposed. First, the chip evacuation torque coefficients (CETCs) are introduced into the chip evacuation torque model to simplify the model for learning. Then, the effect of chip removal process in peck drilling on drilling depth is analyzed. The extended depth coefficient by chip removal (EDCbCR) is introduced to develop the relationship between the extended depth in each drilling step and drilling depth. On the foundation of the modeling above, an iterative learning method for drilling depth optimization in peck drilling is developed, in which a modified Newton's method is proposed to maximize machining efficiency and avoid drill breakage. In experiments with different cutting parameters, the effectiveness of the proposed method is validated by comparing the optimized and measured results. The results show that the presented learning method is able to obtain the maximum drilling depth accurately with the error less than 10%.