中英文翻译-机床床主轴轴承的监控-英文版.PDF

Int J Adv Manuf Technol (2006) 28: 993-1005 DOI 10.1007/s00170-004-2449-0 S. Saravanan ?9 G.S. Yadava ?9 P.V. Rao Condition monitoring studies on spindle bearing of a lathe Received: 10 June 2004 / Accepted: 18 October 2004 / Published online: 16 November 2005 ?9 Springer-Verlag London Limited 2005 Abstract In modern industry, machinery must become increas- ingly flexible and automatic. In order to increase productivity, enhance quality and reduce cost, machine tools have to work free of any failure. When a failure occurs in a machine tool, it is necessary to identify the causes as early as possible. Machine tool condition monitoring is very important to achieve this goal. Condition monitoring is generally used on the critical subsystem of any machine tool. This paper endeavors to focus on the con- dition monitoring aspects on the machine tool element. In the present study, a critical subsystem has been identified based on the failure data analysis. Condition monitoring techniques like vibration monitoring, acoustic emission, Shock Pulse Method (SPM) and surface roughness have been successfully used for fault identification. Keywords Acoustic emission ?9 Bearing - Lathe ?9 Shock pulse method ?9 Surface roughness ?9 Vibration Abbreviation SPM - Shock Pulse Method 1 Introduction Machine tool technology, the backbone of the worlds manufac- turing industry, is developing faster than ever. Despite these rapid advances, developments in the strategy and techniques of ma- chine tool maintenance have fallen behind in the general pattern of progress. This is of serious concern, since it can be argued that present maintenance techniques will be unable to cope with new demands. In the modern automated manufacturing process, ma- S. Saravanan ?9 G.S. Yadava () Industrial Tribology, Machine Dynamics and grease mixtures with 90 Ixm size silica particle with 5%, 10% and 15% concen- tration by weight are identified as B 1, B2 and B3, respectively. The details of grease mixtures are given in Table 2. Table 2. Details of grease mixture used for experimentation Grease Concentration of Particle size of mixture contamination contamination identification (%wt) (lxm) Remarks HG A 1 5 53 A 2 10 53 A 3 15 53 B 1 5 90 B 2 10 90 B 3 15 90 Sample taken from bearing under healthy condition Delibeartely prepared Delibeartely prepared Delibeartely prepared Delibeartely prepared Delibeartely prepared Delibeartely prepared 3.3 Experimental procedure Accelerometers were mounted at locations 1 and 2 to receive the spindle bearing vibration. The overall vibration levels and vibration spectrums were recorded in FFF Analyser (ONO SOKKI) for all the cutting conditions, as shown in Table 1. The acoustic emission hydrophone SEH receives the acoustic waves through a jet of cooling lubricant which is connected directly to cutting zone, as shown in Fig. 4. The measuring jet was made as short as possible since this has better charac- teristics with respect to ground noise and damping. The flow quantity was also kept as low as possible to minimize the impact noise. The shock pulse measurements were carried out on test bear- ings to ensure that the bearing fault due to contaminated lu- brication condition existed. Surface roughness measurements of a turned workpiece was carried out as an offline monitoring method. Talysurf-6 (Taylor Hobson) was used for this purpose. To minimise the effect of cutting tool wear and breakage on acoustic emission and surface roughness, care was taken to en- sure cutting tool sharpness. For each set of experiments a fresh cutting tool insert was used; i.e., the tool was not allowed to wear more than a standard specification 28. 3.4 Signal acquisition for healthy condition Turning operations were performed with the combinations of three independent cutting parameters, namely cutting speed, feed rate and depth of cut. Three different cutting speeds (i.e. 78, 98 and 125 m/rain) were used. Feed rate values of 0.304 mm/rev, 0.384 mm/rev, 0.480 mm/rev, and 0.528 mm/rev were kept for the entire experiment. Depth of cut varied from 0.2 mm to 0.8 mm. While varying one parameter, the value of all other parameters were kept constant. In the first stage, machining op- erations were performed on a defect-free, “healthy“ lathe. The overall vibration level (rms velocity) and frequency spectrum were recorded at both the headstock-beating locations. The over- all vibration levels for a healthy condition lathe was monitored at locations 1 and 2, as shown in Figs. 5-8 respectively. The fre- quency spectrums of the defect-free lathe are shown in Figs. 9 and 10 for monitoring locations 1 and 2, respectively. The AE values for the healthy condition are shown in Fig. 11. The AE smoothed signal for healthy conditions is given in Fig. 12. Sur- face roughness measurements (in txm) on the turned workpieces 997 are shown in Figs. 13 and 14 for a defect-free lathe. The grease sample was taken from the taper roller bearing after complet- ing the experimentation. Flame atomic absorption spectrometry (FAAS) analysis was carried out at Department of Chemistry on the used healthy grease (HG) sample; this is shown in Fig. 15. 3.5 Signal acquisition for faulty bearing conditions The next stage of experimental investigations considers defective taper roller bearings. In our study, a new set of taper roller bear- ings were used for each step. In the first step, the test bearings were filled with a particular contaminated grease mixture. The amount of grease added can be obtained from 29: G - 0.005D B, where G = grease quantity, g D = bearing outside diameter, mm B = beating width, mm. The taper roller bearing was lubricated with deliberately con- taminated grease mixture A1. The lathe operated for 8 hours Fig. 5. Overall vibration levels for 1.8 healthy and faulty bearing condi- tions at location 1 for a speed of 125 m/min, a feed of 0.528 mm/rev, and a contamination particle size of 53 ltm at various %wt 1.6 O v 1.4 1.2 0.8 0.6 0.4 0.2 OIQIQIQDNeDmNDQimD OtOli i I i O i g O 6 ?9 g J Feed 0 = 0.584 mm/rev o = 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev ,O Healthy , J J ?9 5 % silica- AI /“ 10% silica-A2 ,“ * S ?9 / mD 15% silica -A3 . .“ , , # , ?9176 O“ ?9 / ,l , , . /“ I . .-?9149149 j“ t n“ , l* “ S o.*?9 o“ J o?9 I“ .12“ “ . J“ ?9 s j ?9 . ?9 “149149176176176176 “ l“ , ./11 . . o .“ - ?9 , - . o.-“ e, -“:“ “ or“ . o“ ,- ?9 - “r?9149 .“ ?9 - _,x .,. :,- ,. - -x o- - “ .,c“ .; - “ Lr. ?9 - .- ?9 .“ .“ “ “ “ - K. “ . s t :.: - , ?9 No Cut 0.2 0.4 0.6 0.8 Depth of Cut (mm) 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 IA 1.2 1 0.8 0.6 0.4 998 Fig. 6. Overall vibration levels for healthy and faulty bearing condi- tions at location 1 for a speed of 125 m/rain, a feed of 0.528 mm/rev, and a contamination particle size of size 90 Ixm at various %wt 2.75 2.5 2.25 2 75 a5 1.25 re 1 0.75 0.5 0.25 0 Feed O - 0.584 mm/rev c = 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev Healthy “ . 5 % silica- AI . -“ .; .“ 10% silica-A2 “ “4“ “ .“ 411 . - B“ ,- ,r .- “ r -.-“ .- :. - “ .- . . :_-.-“ -_ . :.,- , 0,5 without cutting. For every 30 minutes of run, a 15 minute in- terval was provided to avoid overheating the motor and bearing elements. It was regularly observed from the shock pulse meas- urement (SPM) that the test bearing had entered the faulty zone. A turning operation was performed and parameters were meas- ured (as for the healthy condition). The overall vibration levels for faulty bearing conditions at locations 1 and 2 are shown in Figs. 5-8, respectively. The frequency spectrums of faulty bearing conditions A1 and B1 at locations 1 and 2 are shown in Figs. 16-19, respectively. The AE values (in decibels) for various feed and depth of cut conditions at a constant spindle speed (125 m/min) tinder faulty bearing conditions are shown in Figs. 11 and 20. The AE smoothed signal for faulty bearing con- dition A1 is given in Fig. 12. Surface roughness measurements Fig, 8. Overall vibration levels for healthy and faulty bearing conditions at location 2 at a speed of 125m/rain, a feed of 0.528 mm/rev, and a contamination par- tic|e size 90 Ixm at arious %wt I.g 1.6 1,4 I 1,2 1 0 .-g 0.8 0.6 0.4 0.2 0 Feed 0 = 0.584 mm/rev - 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mrrdrev Healthy -“ “ -. . 5 % silica -AI / ,El“ . - ,o silica-. .,.-. .-. - -“ ,- -.- “ - -“ -“.- -n ?9 15%siLica-A3 “ A,(*“ /%“ .“ ./ Ce“ .“ .0 .,- “ “ .“.- “ ., 0,02 0.01 0 0 57.5 425.75 - - ?9 , , ?9 T 50 100 150 200 250 300 350 400 450 500 Frequency (Hz) 28.75 Fig. 10. Frequency spectrum for healthy condition at location 2 at a speed of 125m/min, a feed of 0.528 mm/rev, and a depth of cut of 0.8 mm 0.35 o.3 o.25 0.1 0.05 I O 0 56.25 50 1 O0 150 200 260 300 350 400 450 500 Frequency (Hz) 1000 Fig. 11. Acoustic emission levels of healthy and faulty bearing conditions at a speed of 125m/min, a feed of 0.528mm/rev, and a contamination particle size of 53 I-tm at vari- ous %wt 41 39 37 A 111 ,“ 35 .o ._m E 33 tu ._o o 29 27 25 Feed 0 = 0.584 mm/rev o = 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev Healthy 5 % silica- AI .NNmm 10% silica- A2 15% silica- A3 . . . . - - . . . . “- -=o . e-“S . . -S. = :-“-“ -a O I B 4t-; “. - “ . _ .-,-,r - - -S-.:.-. - o “ .“ #- _ - - .-.-r - -Q . .- # .4r.“.“ s.,( (3“ .- “ . .If- ?9 ?9 -O“ “ “ : .- “- -. e “ :. . .- : . . a“ -“. .- - - - r , r r , , 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Depth of Cut (mm) (in Itm) were carried out on the turned workpiece for defective bearing condition A1, and are shown in Figs. 13 and 14. The experiments were repeated for other grease mixtures for all the cutting conditions shown in Table 1. FAAS analysis was also car- ried out for contaminated grease mixtures, and the proportions of iron content is shown in Fig. 15. Fig. 12. Acoustic emission smoothed signals for healthy and faulty bearing conditions 4 Results and discussion 4.1 Vibration monitoring 4.1.1 Overall vibration The graph of overall vibration levels at locations 1 and 2 for defect-free lathe conditions (Figs. 5 and 6) shows continuous lines. This indicates that vibration levels increase with an in- crease in the depth of cut, at a constant feed. With an increase in feed (keeping depth of cut constant), the overall vibration level shows an increasing trend. The values of overall vibration level are higher at location 2 than at location 1. This is because at lo- cation 2 the headstock spindle is supported on a bush bearing, whereas a pair of taper roller bearings are used at location 1. A study of the overall vibration levels at location 1 for various faulty bearing conditions of 53 Ixm silica contamination (A1, A2 and A3 grease mixtures) of the bearings in Figs. 5 and 6 shows that overall vibration levels are higher than those recorded in the healthy condition. The rate of increase with the depth of cut is also more than that in measured under defect-free conditions. With an increase in feed (keeping depth of cut constant), the in- crease in overall vibration level is more in the case of healthy conditions. At location 2, the overall vibration level values show Fig. 13. Surface roughness of healthy and faulty bearing conditions at a speed of 125 m/rain, a feed of 0.528 mm/rev, and a contamination particle size of 53 txm at various %wt 5.5 4.5 4 35 lao r 2.5 1.5 Feed 0 = 0.584 mm/rev = 0,480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev - Healthy . 5 % silica-A! - _ - -O 10% silica-A2 41, - - _ _rt 15% silica-A3 - -“ - ,J3“ - - - -. _- - - .e,- - -.4:1 .r . .- . . . . . - - - . - - - : ,.- : .-.- .- .,d W- o .,_.-.,- X., “,r. - ,.4X“ /- .,-_. - 176 . . 176 “ i i i 0.2 0.4 0.6 0.8 Depth of Cut(mrn) 1001 Fig. 14. Surface roughness of healthy and faulty bearing conditions at a speed of 125 m/min, a feed of 0.528 mm/rev, and a contamination particle size of 90 ttm size at various %wt 5.5 4.5 - 3.5 2.5 1.5 Feed 0 = 0.584 mm/rev = 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev - Healthy . 5 % siliea-Al 4 . “ -2SS 15% siliea-A3 - - o “. - - /D“ “ “ .,-O ;S :2 “ ?9 -. - J, - - - “- -“I- ,. ,-“ ,4“ 1,dr- 0.2 0.4 0.6 0.8 Depth of Cut(mm) a similar trend, but the increase in the vibration level is much lower than that at location 1. This is because the taper roller bear- ings in location 1 were lubricated with deliberately contaminated grease mixtures. Figures 7 and 8 show the overall vibration lev- els at locations 1 and 2 for faulty bearing conditions of 90 Ixm sized silica contamination (B 1, B2 and B3 grease mixtures) of the lathe. The increase in vibration level is much greater as com- pared to the 53 Ixm sized silica contamination for all cutting t002 Fig. 15. Amount of iron content (ppm) on different grease mixtures according to FAAS study Fig. 16. Frequency spectrum for defective bearing condition at location 1 at a speed of 125 m/min. a feed of 0.528 mm/rev, a depth of cut of 0.8 mm, and a 53-txm silica contamination, 5 wt. % Fig. 17. Frequency spectrum for defective beating condition at location 2 at a speed of 125 m/min, a feed of 0.528 mm/rev, a depth of cut of 0.8 mm, and a 53-1Lm silica contamination, 5 wt. % Fig, 18. Frequency spectrum for defective bearing condition at location 1 at a speed of 125 m/min, a feed of 0.528mm/rev, a depth of cut of 0.8mm, and a 90-1m silica contamination, 5 wt.% 0.4 - 0.35 0,3 0.25 0.2 0.15 .o o.1 tn 0.05 28.75 7.5 k 1003 218.75 56.75 270 50 1 O0 150 200 250 300 350 400 450 500 Frequency (Hz) Fig. 19. Frequency spectrum for defective bearing condition at location 2 at a speed of 125 m/min, a feed of 0.528mm/rev, a depth of cut of 0.8mm, and a 90-1tm silica contamination, 5wt.% E C .o .0 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 28.75 218.75 0 50 100 150 200 Frequency (Hz) 250 300 350 400 450 500 Fig. 20. Acoustic emission levels of various faulty bearing condi- 40 tions at a speed of 125 m/min, a feed of 0.528mm/rev, and 5 wt. % at various contamination sizes 38 36 34 r,1 32 30 28 Feed 0 - 0.584 mm/rev = 0.480 mm/rev A = 0.384 mm/rev x = 0.304 mm/rev 53 t/m silica-A1 , 4), -“ “ “ 90 lm silica -B 1 . “ “ / A i i i i i 0.2 0.4 0.6 0.8 1 Depth of Cut (ram) 1004 conditions. The surface waviness created due to the bigger par- ticle may be responsible for higher vibrations 30. 4.1.2 Frequency spectrum Figures 9 and 10 show the frequency spectrums of the defect-free lathe at a spindle speed of 125 m/rain, a feed of 0.528 mm/rev, and a depth of cut of 0.8 mm at locations 1 and 2, respectively. Significant peaks can be observed at frequencies of 28.75 Hz, 57.5 Hz and 451.25 Hz for location 1. Location 2 (Fig. 10) has significant peaks at 28.75 Hz and 56.25 Hz frequencies. IX com- ponent (i.e. 28.75 Hz) is predominant at both locations. This indicates that there may be certain amount of residual imbalance in the headstock spindle. Figures 16 and 17 show the frequency spectrums of faulty bearing condition (A1) at a spindle speed of 125 m/min, a feed of 0.528 mm/rev, and a depth of cut of 0.8 mm for locations 1 and 2, respectively. The bearing defect frequencies are given as follows 31: Defect on ball/roller, Defect on inner race, Defect on outer race, t21 N Fb = - d 1- cos Fi= 1+ cos/ 1-n d cos/3) N n Fo=(1 b 1- ( d )N Defect on cage, Fc = 1 - - cos/3 120 In the present case, the values of Fb, Fi, Fo and Fc are found to be 5.36, 268.23, 211.76 and 11.76 Hz, respectively. The corres- ponding bearing defect frequencies for the inner race (265 Hz), the roller (7.5 Hz), the cage ( 13.75 Hz) and some higher frequen- cies can be observed in Figs. 16 and 17. The frequency spectrums of faulty bearing condition B 1 at locations 1 and 2 are shown in Figs. 18 and 19, respectively. The corresponding peaks of roller (7.5 Hz), outer (218.75 Hz) and inner (218 Hz) defect frequen- cies were observed with higher amplitudes than was the case in faulty bearing condition A1. The increase in amplitude of bear- ing defect frequencies were observed from Figs. 16-19 as the size of the silica particle increases. Thus, it can be concluded that there is certain residual im- balance in the spindle because the 1X component (i.e. 27.5 Hz) is predominant in all the cases (healthy and faulty bearing con- ditions). Significant peaks are observed at the bearing defect frequencies. Also, an increase in amplitude was observed as par- ticle size of the contamination increased. 4.2 Acoustic emission monitoring Figure 11 shows that AE values for a defect-free lathe increase continuously with an increase in the depth of cut, at a constant feed. With an increase in feed for the same depth of cut, the AE level shows an increasing trend. The AE smoothed signals for healthy and faulty bearing conditions are shown in Fig. 12. The AE levels for defective bearing A1, as shown in Figs. 11 and 20, demonstrates that AE levels are higher here as compared to healthy conditions. The rate of increase with the depth of cut is also more than the case of healthy conditions. With an increase in feed (with the same depth of cut), the increase in AE level is more than that under healthy conditions. The study of AE lev- els for defective bearing condition A1 (Fig. 11) shows a higher AE level compared to a defect-free lathe. But at higher feed and higher depth of cut, the increase in AE level is much more as compared to healthy bearing conditions, and is shown in Fig, 11. Figure 20 depicts the relationship between the particle size of the contaminants and AE level. With larger particle sizes there is a considerable increase in the AE level, irrespective of depth of cut. But in earlier research 19, AE pulse count measurements on a deep groove ball bearing indicated that the AE pulse count decreased with an