With significant technological progress in optics fields, such as VR and 3D sensing technologies, the demand for optical components with complex shapes (e.g., freeform surfaces) has expanded significantly in recent years. Ultra-precision diamond turning is one of the common manufacturing methods used for freeform optics. There are two tool-motion mechanisms for freeform diamond turning: slow tool servo (STS), which moves the entire tool rest, and fast tool servo (FTS), which only drives the diamond tool. In particular, high-frequency and long-stroke FTS units driven by a separate control system from the machine tool controller have recently been developed as a new trend in FTS-based diamond turning. However, this type of independent FTS control system causes a time delay between the machine tool instruction and the FTS tool motion. Even a millisecond-scale time delay can cause a considerable angular misalignment owing to the high rotation rate of the spindle, resulting in a micron-scale form error on the workpiece surface. In this paper, a novel method for precisely measuring the time delay from the machined workpiece surface and compensating for the misalignment based on the measured time delay was proposed. To measure the time delay, spherical dimples were machined on the workpiece surface by using STS and FTS. Subsequently, the angular misalignment and time delay were obtained by comparing the measured surface profiles at specific positions of the spherical dimples. Thus, the time delay can be measured with microsecond-scale accuracy. Based on the measured time delay, the shifted tool position along the spindle axis by the time delay was compensated to the ideal tool position by considering the cutting velocity. Consequently, the angular misalignment was completely eliminated, and submicron-level form accuracy was achieved. This calibration method of angular misalignment for an independent FTS unit is expected to significantly improve the accuracy of ultra-precision machining of freeform optics.
Read full abstract