International Journal of Mechanical And Production Engineering, ISSN: 2320-2092, Volume- 5, Issue-2, Feb.-2017 http://iraj.in DETECTION OF TOOL WEAR IN DRILLING PROCESS USING VIBRATION SIGNAL ANALYSIS
1M. AMARNATH, 2P. RAMAKRISHNA, 3H. CHALLADURAI
Tribology and Machine Dynamics Laboratory, Department of Mechanical Engineering, Indian Institute of Information Technology Design and Manufacturing Jabalpur, Jabalpur 482001, India
Abstract- Drilling is a machining process where a multi-point tool is used to produce desired holes by removing unwanted material. In this process, the contact between the cutting tool and the work piece generates forces which in turn create torques on the spindle and drive motors. Excessive forces and torques can cause tool failure, spindle stall, undesired structural deflections etc. The vibration levels, cutting forces, torques and power directly affect the other process parameters viz. machinating accuracy, tool wear and chip morphology.Therefore, these parameters are often monitored and regulated due to which productivity is maximized. This paper addresses the results of experimental investigations carried out to monitor drill wear and its corresponding effects on parameters such as vibrations levels, hole accuracy, chip morphology etc. during drilling of AISI 1040 steel. Results highlight the importance of vibration signal analysis techniques to assess tool wear severity.
I. INTRODUCTION (PVD) method. Results highlighted the advantages of TiN coating in enhancement of hardness, surface Metal cutting processes are widely used in finish and compressive strength on the surface of the manufacturing industry which includes operations specimen. Wardany et al. (1995) carried out such as turning, milling, drilling and grinding. The experiments to investigate the wear and failure of the trend towards automation in machining has been drill using vibration signature analysis. Vibration driven by the need to maintain product quality and signal features extracted from time domain and increase in production rates. Machining operations frequency spectra were used to detect the simulated are shape transformation processes where metal is faults on the drills. removed from a stock of material (i.e. workpiece) to produce a desired part (Tlusty, 1986). The performance of machining is measured in terms of Further, frequency domain analysis of vibration cutting forces, tool wear, power consumption and signals acquired in transverse and thrust directions of surface finish of the finished product. In order to drilling operation was considered as a drill breakage ensure product quality, production time, safety and monitoring index. This paper presents the results of overall economy it is desired to implement condition experimental investigations carried out to assess tool monitoring techniques in machining process. wear in drilling, statistical parameter and frequency Objectives of the monitoring of machining process spectrum analyses of vibration signals acquired from are related to the performance of machine tool, the work piece provide promising results to detect detection of tool wear, dimensional tolerance, surface wear propagation on cutting edges of drill during roughness, energy usage, chip formation etc. machining. (Devillez et al., 2004, Zhang et al., 2001, Dhar et al., 2006). II. EXPERIMENTAL SETUP
Dhar et al. (2006) carried out experimental Fig.1 shows the schematic representation of investigations to assess the drill wear. Authors have experimental setup. Drilling operation was carried out observed that the high temperature generated in the on conventional drilling machine of spindle power drilling process causes a large amount tool wear, 2.2 kW which has with automatic feeding mechanism which affects the accuracy of hole, chip shape and facility. About 300 through holes were machined on color of the chip. To minimize these drawbacks, a mild steel work piece of thickness 10 mm by using high pressure coolant was applied to tool – work high speed steel (HSS) drill. Lateral vibration signals piece interface. Hocheng (2001) carried out data of tool-work piece interface was acquired using experiments to improve mechanical properties of accelerometer, vibration signals were processed and AISI D2 tool steel, initially the specimen was stored using Dytran data acquisition system and machined using electrical discharge machining. personal computer. Further, these vibration signals Further, Titanium nitride (TiN) coating was applied were post processed to obtain statistical parameters on the specimen using physical vapour deposition and frequency spectra.
Detection of Tool Wear in Drilling Process Using Vibration Signal Analysis
136 International Journal of Mechanical And Production Engineering, ISSN: 2320-2092, Volume- 5, Issue-2, Feb.-2017 http://iraj.in
Fig.1 Experimental setup
III. RESULTS AND DISCUSSIONS
High speed steel drill of 6 mm diameter and 93 mm 1.14 ) overall length was used to drill through holes in a 2 mild steel plate of 10 mm thickness. The speed and feed rates are 140 rpm and 0.02 mm/ rev respectively, 0.76 about 300 through holes had been drilled during experiments. Fig. 2(a) - (b) show vibration spectra in 0.38 the frequency range of 0.5 – 3.5 kHz, a gradual s / (m Amplitude increase in trend in the amplitudes indicate wear severity on the drill. 0.00 0 1000 2000 3000 4000 5000 Frequency (Hz) From this observation, it is evident that the pattern of frequency vibration spectra is sensitive to degree of wear on the drill rather than operating conditions of the drilling process. Further, the most commonly used 1.14 ) statistical parameters viz. root mean square (RMS), 2 kurtosis and crest factor were considered to monitor 0.76 tool wear propagation.
0.38
Fig. 3-5 show the statistical parameters of vibration s / (m Amplitude signals acquired from tool-work piece interface, which indicate wear propagation on the drill with 0.00 respect to the number of holes drilled. Root mean 0 1000 2000 3000 4000 5000 square (RMS) value indicates all the energy of the Frequency (Hz) signal including all the noise. Fig. 2 Vibration frequency spectra (a) Fresh drill (b) After 300 holes
Detection of Tool Wear in Drilling Process Using Vibration Signal Analysis
137 International Journal of Mechanical And Production Engineering, ISSN: 2320-2092, Volume- 5, Issue-2, Feb.-2017 http://iraj.in Fig. 3 depicts the variation of RMS values of IV. SUMMARY AND CONCLUSIONS vibration signals, a gradual increase in trend is observed with the increase in number of holes. Crest In the present work, experiments were carried out to factor is used to detect the severity of the wear on the assess the wear on the high speed steel twist drill used drill. Fig. 4 shows crest factor values versus number in machining mild steel work piece. Diagnosis of of holes drilled in work piece, increase in trend in wear propagation on the drill was made using lateral the crest factor values indicates the severe wear of the vibration data acquired from the work piece-tool drill. The kurtosis is the fourth normalized moment interface. The following conclusions were drawn of a vibration signal which provides measure of from the experimental results. peakedness of the signal. 1. Frequency spectrum analysis revealed the diagnostic information of wear propagation on the drill, a gradual increase in characteristic frequency amplitudes highlighted the severity of wear on the cutting edges. 2. From the frequency spectrum analysis of HSS drill, it was observed that dominant frequency amplitudes are sensitive to the degree of wear on the drill. 3. Statistical parameters of vibration signals such as RMS and crest factor values showed over all increasing trend as a function of drilled holes, which highlighted wear propagation on drill. Fig. 3 RMS values versus drilled holes 4. The kurtosis values fail to provide tool wear diagnostic information due to spurious effect of noise present in the vibration signals.
REFERENCES
1. Tlusty, J., Smith S., and Winfough, B., (1996), “Techniques for use of long slender end mills in high speed machining”, Annals of the CIRP 45(1),393-396. 2. Devillez, Lesko, S. and Mozer W., (2002) “Cutting tool crater wear measurement with white light interferometry”, Wear, 256, 56 – 65. 3. Zhang, M., Liu, Y.B. and Zhou, H., (2001) “Wear mechanism maps of uncoated HSS tools drilling die-cast aluninium Fig. 4 Crest factor versus drilled holes alloy”, Tribology International, 34, 721-731. The variation of kurtosis values with increase in 4. Dhar, N.R., Islam, M.W.S. and Mithu, M.A.H., (2006) “The influence of minimum quality of lubrication (MQL) on number of holes is shown in the Fig. 5, these values cutting temperature, chip and dimensional accuracy in show uneven trend which is due the higher turning, AISI-1040 steel”. Journal of Material Processing susceptibility to spurious effect of noise and high Technology, 171, pp. 93-99. frequency vibration signals. Hence, in some cases the 5. Dhar, N.R., Rashid, M.H. and Siddiqui, A.T., (2006) “Effect of High Pressure Coolant on Chip, Roundness Deviation and adverse effect of noise on the kurtosis values is more Tool Wear in Drilling Aisi-4340 Steel”, ARPN Journal of than the benefit gained from the higher sensitivity of Engineering and Applied Sciences, 1(3), 52–59. kurtosis to detect incipient faults [8]. 6. Guu, Y.H. and Hocheng, H., (2001) “Improvement of fatigue life of electrical discharge machined AISI D2 tool steel by TIN coating”, Materials Science and Engineering A318, 155- 162. 7. Wardany, T.I.E.L., Gao, D. and Elbestawi, M.A., (1995) “Tool condition monitoring in drilling using vibration signature analysis”, International Journal of Machine Tools and Manufacture, vol.36, 687-711. 8. Oack, H. and Laprao, K.A., (2005), “HMM based fault detection and diagnosis scheme for roller element bearing”, Journal of Vibration and Acoustics 127(4), 299–306 .
Fig. 5 Kurtosis values versus drilled holes
Detection of Tool Wear in Drilling Process Using Vibration Signal Analysis
138