Development and evaluation of an on-machine optical measurement device

https://doi.org/10.1016/j.ijmachtools.2006.11.007Get rights and content

Abstract

Demand for fabricating micro-features such as fine holes, micro-cavity for injection moulding, and micro-pin using both conventional (turning, milling, etc.) and non-conventional edge detection method (EDM), wire cut EDM, etc.) processes is increasing significantly. To successfully achieve micro-machining, development of a miniature machine tool, process technology, and on-machine measurement is essential. However, in such tool-based micro-machining processes, proper tool shape monitoring, precision processing, and dimensional control require significant attention. Since these are tool-based machining processes, tool shape monitoring and control are also important technologies to be established.

In this study, an on-machine measuring device was developed based on non-contact optical method to inspect dimensions of the fabricated tools (e.g. electrodes for EDM) as well as the wear of tools used for the respective processes. The developed inspection system uses a laser light source and a photo-electronic device. To minimize errors due to the change of tool measurement position and the Fresnel diffraction of laser light, an edge detection algorithm using a linear discrimination function is proposed in this study. Furthermore, an intensity measuring method was added for specimen with a smaller diameter. The experimental results show that the developed on-machine optical inspection system has the accuracy and stability to effectively monitor the fine dimensions of tools and their wear.

Introduction

In order to maintain accuracy of a component which is fabricated by tool-based machining, it is necessary that the dimension of tool must be under control. In-process estimation of tool wear in the machining process is also important for scheduling tool changing time and for adaptive process control and optimization. Sensing methods of in-process estimation of tool wear can be divided into direct and indirect measurements. In case of indirect measurement, relationships are established between tool wear and other parameters which are related to tool wear and are easier to measure. Some parameters include cutting forces, roughness of machined surfaces, temperature and thermoelectric effect, acoustic emission and vibrations, etc. [1], [2], [3], [4], [5]. On the other hand, direct measurement involves measuring the wear and evaluating the volumetric loss from the tool due to tool wear [6], [7]. In high accuracy micro-machining at the micron level, the indirect measuring signals become very small, and the uncertainty of sensed signal increases owing to the effects of size.

Micro-machining quite often requires a combination of both conventional and non-traditional machining for achieving dimensional accuracy. For example, fabricating an electrode by turning, and subsequently the same electrode is used for edge detection method (EDM), achieves the best dimensional accuracy. By using on-machine tool fabrication such as wire electro discharge grinding (WEDG), very fine electrodes can be prepared without any possible clamping error [8]. Nevertheless, to ensure an exact dimension of the electrode, inspection should be carried out on-machine without taking off the electrode from the machine to eliminate re-clamping error of the electrode. Since tool wear is quite significant in EDM, on-machine tool inspection is also required to compensate for the electrode wear.

In this paper, an optical tool wear measuring system using the laser light source and a photo-electronic device was developed and tested. By scanning the shape of tool and comparing it with a fresh tool, the amount and shape of tool wear could be measured. To measure tool wear, two edge detection algorithms were tested and compared. From the experimental results, an edge detection algorithm using linear discrimination function was found to be suitable for the measurement of the electrode without any stand-off dependent error.

Section snippets

Design of the optical measurement system

A variety of direct measurements have long been used to measure the dimensional change of tool for machining. One popular method of dimensional measurement uses a video camera to capture an image of the tool. Image processing techniques are then used to determine the location/shape of the object. In this study, a quadrant photo diode array was used to detect tool shape in order to develop an affordable on-machine measurement system at a lower system cost, and Fig. 1(a) shows the configuration

Experimental set-up

In this study, experiments were conducted on a miniature desktop machine tool, which was designed for multi-purpose micro-machining capability such as micro-milling, micro-drilling, micro-turning, and micro-EDM/ECM. A hybrid machining technique can be used to fabricate a micro-mechanical structure [8]. Fig. 4 shows the on-machine measurement device attached on the machine table. Tool electrodes are not rotating during measurement.

For measuring experiments, cylindrical specimens were machined

Change of sensor signal and normalizing

Fig. 5 shows a typical pattern of the change of sensor signals recorded as the cylindrical specimen passes across the laser light. The diameter of specimen is 3.939 mm and the measuring stand-off from the photo-detector to the specimen is 40 mm. As discussed above, the signals show fluctuations when the electrode partially blocks optical detector. Signals are gradually decreasing when the whole sensor elements are blocked by the electrode. This is different from Fig. 3, where the signals show

Conclusions

In this paper, an optical on-machine measurement system has been developed based on edge detection methods and intensity measurement. These methods have been tested to measure the diameter of specimens. From the study, the followings conclusions have been drawn:

  • (1)

    Two kinds of edge detection methods are applied to measure the diameter of specimens. One is maximum differential method, and the other is linear discrimination function.

  • (2)

    Maximum differential method shows slightly larger estimation than

Acknowledgments

The authors would like to express their heartfelt thanks to the Mitsutoyo Association for Science and Technology Japan for their generous financial support towards the development of this on-machine measurement device.

References (12)

There are more references available in the full text version of this article.

Cited by (0)

View full text