An automated flank wear measurement of microdrills using machine vision
Introduction
Measuring the wear of the microdrill (diameter of 0.3–0.02 mm) is important in a printed circuit board (PCB) production. While a drill begins to wear, the cutting forces increases and the temperature of the drill raises, which speeds up the physical and chemical reactions associated with drill wear, and causes rapid deterioration of the drill quality [1]. In PCB manufacturing, a worn-out microdrill damages the quality of the surface finish and the dimensions of the drilled hole. Tool wear not only reduces the part geometry accuracy directly but also increases the cutting forces drastically. The diagram of the microdrill and its characteristics are shown in Fig. 1. The cutting plane (also called first facet or lip relief plane) is consisted of four edges, of which the cutting lip and chisel edge are two important cutting edges on the first facet. The centering point is the first contact point of a microdrill with the material surface, and cutting lips are the major cutting edges for material removal. Chisel edges are comprised of two intersecting planes that define two primary cutting edges of the microdrill. They remove the material by extrusion and cut at a highly negative rake angle. Moreover, for smaller diameters, various drilling parameters, such as web thickness, point angle, spindle speed, and feed, need to be further analyzed. The effect of retraction rate for the influence on cutting life is also concerned.
Many researchers have used various sensors, acoustic emission (AE) [5], dynamometers [6], vibration [7], ultrasonic vibrations [8], and motor spindle speed and power and power consumption [14] to monitor drill wear. Lin and Ting [3] established the relationship between force signals and flank wear and other cutting parameters when drilling a copper alloy. Ertunc and Loparo [4] developed a decision fusion center algorithm, which combines the outputs of the individual methods to make a global decision for the wear status of the drill. Li et al. [10] presented a hybrid learning method to map the relationship between the features of cutting vibration and the tool wear condition.
Several researchers have examined the usage of machine vision for the measurement of tool wear. For example, Nickel et al. [11] presented the machining performance of HSS twist drills that were plasma-nitrided before applying PVD TiN coating, and evaluated and compared it with that of commercially TiN-coated drills. The optical methods were essentially off-line and involved the examination of geometry using a specially adapted toolmaker's microscope for the digital processing of the tool tip image. Pedersen [12] reported that the determination of the wear land is usually based on the high intensity of the reflected light from the wear land. Jeon and Kim [13] employed a suitable thresholding technology to convert the captured grey image into binary image. Subsequently, the amount of flank wear was captured using the binary, where the wear land is white and its background is black. Hazra et al. [2] used three silhouette images of flank surface to measure the geometry of drill points of drills, and fitted the coordinates to a mathematical model of the drill point geometry, which was optimized by genetic algorithm, to estimate five geometry parameters.
Based on the above review, previous researches using various sensors mainly emphasized on the variations of monitors on the tool immediately during the machining process. Other researchers using different algorithms based on vision system measured the degree of tool wear, but they focused only on the measurement of the large-size of drill wear (d > 0.5 mm). A few micrometers of the wear error could affect the quality of the microdrills. Therefore, the primary objective of this research presents a machine vision system to develop an innovation method for measuring flank wear of microdrills and to meet the demand for analyzing the relationship between drilling parameters and tool life.
The remainder of this thesis is organized as follows: the experimental set-up for measuring tool wear is established in Section 2. Section 3 describes measuring wear of microdrill approach using two-dimensional edge diction lines. In Section 4, the measured results are presented to confirm the effectiveness of the proposed method. Finally, Section 5 summarizes the conclusion of this study.
Section snippets
Experimental set-up
Machine vision is the direct method used for tool wear measurement. The experimental of the present study is comprised of two major procedures: (1) several PCB hole-drilling tests with given diameter of 0.2 mm, and the images of flank wear are captured and (2) a machine vision system based on tool microscope to measure the flank wear. The system is detailed as follows.
Measuring method
Machine vision is the direct method used for tool wear measurement. The extent of flank wear can be measured as the distance between the top of the cutting edge and the bottom of the area where flank wear occurs. A simple but effective method of measuring flank wear is using edge detection to identify the width change of the cutting plane. The image can be divided into two parts: object (cutting plane including noise) and background. A rising edge is characterized by an increase in the
Measurement results and discussion
Table 1 shows the measurement results of using the proposed edge detection to compute the height of cutting plane between a pair of edge points in drilling tests from 1000 to 11,000 hits. Three parameters of the wear (flank wear area, average wear height VBave, and maximum wear height VBmax) were computed to evaluate the tool life and measure the degree of wear. The maximum wear height usually occurs at the corners of the cutting lips because these are the locations of the highest cutting
Conclusions
This study has proposed a new machine vision-aided system for measuring the flank wear in a PCB production, and proved the feasibility of measuring frank wear with the hole-drilling tests on a 10-layered PCB. This method has not only measured the flank wear area, the average flank wear height VBavg and maximum wear height VBmax but also the height of cutting plane changed for each position. The wear curve of the experiments of PCB hole-drilling trials is very similar to the Taylor curve. The
Reference (14)
Inspection of reground drill point geometry using three silhouette images
J. Mater. Process. Technol.
(2002)- et al.
Tool wear monitoring in drilling using force signals
Wear
(1995) - et al.
A decision fusion algorithm for tool wear condition monitoring in drilling
Int. J. Mach. Tools Manuf.
(2001) - et al.
The application of acoustic emission for precision drilling process monitoring
Int. J. Mach. Tools Manuf.
(1999) - et al.
Drill wear monitoring using cutting force signals
Mechatronics
(2004) - et al.
Tool condition monitoring in drilling using vibration signature analysis
Int. J. Mach. Tools Manuf.
(1996) - et al.
Tool break detection by monitoring ultrasonic vibrations
Ann. CIRP
(1988)