Evaluation of DC Servo Machine Tool Feed Drives as Force Sensors

Abstract

Unmanned machine tools as part of an automated factory require reliable, inexpensive sensors to provide machine and process information to the controller. The electric current in the DC motor of a CNC machine tool can be inexpensively measured and used to calculate the tool/workpiece cutting force and the forces associated with drive system components. In order to characterize the bandwidth, sensitivity and accuracy of current monitoring on the feed system of a CNC lathe, a dynamic lumped parameter model of this sensor system is developed. The model is used to identify the system components that have a dominant effect on the behavior of the sensor. Tests were conducted in order to determine the model parameters, verify the model, and determine the signal-to-noise (S/N) ratio of the sensor. The bandwidth of this sensor is predicted to be 80 Hz. Tests show that the S/N ratio is low but can be improved by a trade-off with the system bandwidth. The bandwidth is limited by the characteristics of the SCR amplifier. In addition, the sensitivity and accuracy of calculating the feed force component of the cutting force from the total current used by the feed motor is limited by the pitch of the ball screw and friction coefficient variations in the slide. Feed system design changes, to improve the S/N ratio of the feed system as a tool and machine force sensor, are discussed.

Methods and Research for Multi-Component Cutting Force Sensing Devices and Approaches in Machining

“…Stein et al [42] proposed a dynamic lumped parameter model to characterize the current of a DC motor of a CNC machine tool, the electric current was used to calculate the cutting force and the forces associated with drive system components. Altintas [43,44] described a model of a vertical milling machine feed drive control system to access the cutting force and predict the tool failures in milling, and the feasibility of the cutting force prediction approach based on armature current measurement was illustrated.…”

Dynamic Characterization and Interaction Control of the CBM-Motus Robot for Upper-Limb Rehabilitation

“…Previous examples of robot control based on current monitoring or force sensorless approaches can be found in the literature. They resort to artificial neural networks [42] and neuro-fuzzy approaches [43], [44] for current characterization in a structured environment, or else to observers for force/torque estimation [45], [46], and are typically applied to the industrial context, such as for indirect cutting force measurement [42], [43]. Only recently, an extension to force control of a mechanical finger for prosthetics has been published [47].…”

Driver current analysis for sensorless tool breakage monitoring of CNC milling machines

“…Among several techniques to detect the cutting force during a milling process [1][2][3], driver current monitoring is the cheapest method because it follows a sensorless approach by measuring the control current in the servo driver directly and then extract the cutting force component [4,5]. However, this signal contains undesired components in the form of high-frequency noise, current control commutation and ball screw effects.…”