Grinding of cermets with cup-wheels

Transcription

1 Grinding of cermets with cup-wheels Jeffrey Badger 1,4,a*, Radovan Drazumeric 2,4,b and Peter Krajnik 3,4,c 1 The Grinding Doc, 65 W. 85 th St. #2A New York, NY, USA 2 Univerisity of Ljubljana, Faculty of Mechanical Engineering, Askerceva 6, 1000 Ljubljana, Slovenia 3 Chalmers University of Technology, Department of Materials and Manufacturing Technology, Hörsalsvägen 7B, SE Gothenburg, Sweden 4 The International Grinding Institute, 127 W. 83rd St., #424, New York, NY 10024, USA a jeffrey.badger@grindinginstitute.com, b radovan.drazumeric@fs.uni-lj.si, c krajnik@chalmers.se Keywords: Cermets, loading, cup wheel, dulling Abstract. Cup-wheels are frequently used to grind cermets, a difficult-to-grind material. An investigation was made into the transient geometry of the cup-wheel rim, grit dulling, wheel loading, and wheel self-sharpening with chip thickness. Tests were performed on a saw-tip grinding machine and specific energies, G-ratios and rim geometries were measured. Results showed that, like grinding of tungsten-carbide, loading is prevalent. However, unlike grinding of tungsten-carbide, grit dulling is also prevalent and wheel conditioning is of limited use. Much better results, particularly with respect to surface finish, can be obtained if the wheel is trued to a predetermined geometry. In addition, grinding parameters must be chosen to induce wheel self-sharpening. Practical recommendations are given. 1. Introduction Cup-wheels are frequently used in grinding, particularly when grinding sawblade tips in tungstencarbide and cermets. While a few articles have been written showing experimental results when using cup-wheels [Kumar, 1995; Toyoma et. al. 1993; Mezger, 1986], very little has been written about the geometry of cup-wheel rim at the wheel/workpiece interface, or the important issue of wheel geometry and how this changes as the wheel wears. In addition, grinding of difficult-to-grind materials such as cermets poses specific challenges, such as rapid wheel loading, high specific energies and low G-ratios. Several studies have shown that small, friable grits work better than large, tough grits, and some indications are that more aggressive grinding conditions are helpful [Kumar, 1995; Tyomoa et. al. 1993]. However, nothing has been written about whether these high specific energies are due to loading alone, or if diamond-grit dulling also is a factor, nor has any research been done about the effect of aggressiveness [Badger, 2008] on wheel wear, specific energies and grit self-sharpening. This article explores both of these two subjects, cup-wheel geometry and cermets. It begins with a discussion on the fundamental parameters of material-removal rate and transient chip-thickness in cup-wheels. Then experimental results are given for grinding of cermets with cup-wheels in a sawblade application. The effect of different aggressiveness values are shown with respect to their effect on specific energies, wheel wear, wheel-self-sharpening and dressing intervals. Finally, a novel method is given for truing cup-wheels to achieve shorter cycle times, lower grinding temperatures, less wheel wear, less frequent dressing intervals and better surface finishes. 2. Cup-wheel geometry and taper development Cup-wheel grinding can be divided into two types: a) plunge grinding, and b) traverse grinding. In plunge grinding, the workpiece is plunged either radially into the wheel on the outer-diameter face or axially on the bottom face. In addition, the workpiece may be oscillated back and forth. This is a form of face-grinding and the primary input is the feedrate. In traverse grinding, a fixed depth of cut is taken and the workpiece is traversed across the bottom face of the wheel. Here the primary inputs are the depth of cut and the feedrate. In addition, the in-feed will be either on one side of the wheel or on

2 both sides. Plunge grinding is more common when grinding inserts; traverse grinding is more common when grinding sawblade tips. The focus here is on traverse grinding. Figure 1. Sawblade grinders in Sheffield, England, approximately Nowadays, sawblades are typically made in tungstencarbide and ground with cup-wheels. [Saw grinders, Trades of Sheffield, etching, Ill. London News, January 6, 1866.] During traverse grinding, the diamond wheel is typically trued with a silicon-carbide or aluminumoxide truing wheel. If the wheel is resin-bonded or metal-bonded, it is then conditioned with an aluminum-oxide dressing stick to clear away bond material. Typically this trying action is performed perpendicular to the axis of rotation. This is shown in (a) of Figure 2. As a result, when grinding commences, with a depth of cut of ae and a feedrate of vw, all of the grinding action occurs on the leading edge of the wheel and the contact length is simply the depth of cut, ae. During this time, the aggressiveness is enormously high and the wheel soon wears away to develop a taper, as shown in (b) of Figure 2. The grinding action then occurs on this taper and the aggressiveness decreases drastically. This taper eventually encroaches on the trailing edge of the wheel. At this point, the surface finish becomes poor and the wheel is sent for trueing, where the cycle begins again. If the in-feed is set to occur on both sides of the wheel, then two tapers develop, meeting in the middle of the wheel. This is shown in (d) in Figure 2. A taper, along with flat, can also be trued into the wheel, as show in (c) in Figure 2. The fundamental relationships can be expressed as follows. The material-removal rate, Qw, depends on the depth of cut, ae, the feedrate, vw, and the width of cut, bw, as: Qw= vw ae bw (1) A straight taper and a flat dressed into the wheel in shown in details in (e) in Figure 2. Here the rim width, wrim, is divided into the flat width, wflat, and the taper width, wflat, such that the taper angle, taper, is defined as: tan( taper) = htaper / wtaper (2) wrim = wtaper + wflat (3) For a freshly trued wheel without a taper, the situation is simply face grinding [Malkin and Guo, 2008] and the aggressiveness is: aggr = 10 6 v w v s (4)

3 After the taper breaks in, the grinding action shifts from the front face to the taper. For a wheel trued with a taper with small values of taper, the contact width is simply the taper width, such that wtaper ~ ltaper. Here the aggressiveness on the taper is then: aggr = 10 6 v w sin(α v taper ) (5) s which applies to both a tapered wheel (0º< taper<90º) and a straight-trued wheel ( taper=90º), in which case it reduces to equation (4). If the depth of cut is greater than the taper height, such that ae>htaper, then grinding will occur both on the front face ( taper=90º) and on the taper, with two different aggressiveness values. However, most wheels with a taper trued into them are ground such that ae=htaper. The flat portion of the wheel improves the surface finish, as here the aggressivensss is theoretically zero. The number of times a point on the workpiece experiences one revolution of the grinding wheel can be expressed as the overlap ratio of the flat, Uflat, from the angular velocity of the wheel (or, in practical terminology, wheel RPM/60) by: U flat = w flat v w (6) ω s where a higher value of Uflat would give a better surface finish. When the taper encroaches on the end of the wheel, Uflat approaches zero and the surface finish deteriorates drastically. If we take a typical set of parameters (ae=0.050 mm; vw=3000 mm/min.; vs=20 m/s wtaper=3 mm), for a freshly trued wheel we get a value of aggr=2500. After a 3 mm wide taper has broken in (or been trued in) we get a value of aggr=42. This results in a greater degree of rubbing, less cutting, and poorer self-sharpening. This is one of the reasons why production companies have difficulties with cup-wheel grinding. A typically scenario is as follows: (1) The operator trues the wheel straight; (2) He grinds, but the aggressiveness is so large that he slows down the feedrate; (3) a short taper quickly develops; (4) a longer taper develops, the aggressiveness decreases and the workpiece begins to burn; (5) He slows down the feedrate in an attempt to reduce the burn; the aggressiveness decreases and the amount of rubbing increases, possibly exacerbating the burn; (6) the taper continues to develop to the point where there is no longer a flat on the wheel (and Uflat 0); at which point the operator is forced to true the wheel and begin the cycle again. Rim Geometry: (a) after straight trueing (e) CONFIDENT Slide 6 detailed view of (c) IAL v s after straight truing & break-in via grinding, in-feed both sides (d) (b) (c) after straight truing & break-in via grinding l taper taper h taper a e a e a e after taper trueing w flat w rim w taper v w Figure 2. Taper geometry in traverse cup-wheel grinding.

4 3. Experimental Tests were done on a Vollmer sawblade grinder. Power was measured with The Grindometer, a device that measures current and voltage in all three phases and rectifies the signal to obtain true power in kilowatts. This was then converted to specific energy by e=p/q. Dimensional measurements were made on every tenth tooth and corrections to the infeed were made to reflect this. Wheel wear was measured by monitoring the infeed corrections and the G-ratio was calculated, with G-ratio defined as the volume of material ground divided by the volume of wheel worn away. The workpiece material was cermet. The diamond wheel was trued with a silicon-carbide wheel on a separate truing station and then conditioned with a 400-mesh aluminum-oxide dressing stick. Grinding parameters were as follows: vw=4 mm/s; ae=0.1 mm; Q w=0.4 mm 2 /s; ds=125 mm; dg=54 m; vs=18 m/s; wrim=3 mm; wflat=1 mm; wtaper= 2 mm; taper=2.86º, aggr=11.1; sawblade tips width w=1.9 mm; length l=2 mm; Ns=60 teeth/blade; N=35 blades, total volume ground per mm rim width per mm wheel circumference V w=0.68 mm 3 /mm/mm. Tests were run under three sets of parameters, given in Table 1. Table 1. Test parameters vw vs Q'w aggr Nt Parameter Set 1 4 mm/s 18 m/s 0.4 mm2/s teeth Parameter Set 2 1 mm/s 13 m/s 0.1 mm2/s teeth Parameter Set 3 4 mm/s 13 m/s 0.4 mm2/s teeth 4. Results Results of specific energy (via power measurements) are shown in Figure 3. Here we can see that the power begins at a value of around e~100 J/mm 3 and gradually increases, approaching a nearsteady-state value of e~650 J/mm 3. Although much higher than carbide, these values Kumar [Kumar, 1995] and Toyoma [1993] for similar grit sizes. Then, at around the 25 th sawblade the power drops suddenly before increasing again. This appears to be the collapse phenomenon [Badger, 2009] due to high normal forces from excessive loading or excessive grit dulling. In production-grinding operations, at the first sign of trouble operators will often decrease feedrates. To explore this, the feedrate was decreased by 75%, to vw=1 mm/s. The specific energy began to increase rapidly, reaching a value of e~1900 J/mm 3. This is enormously high, even for cermets, and still had not reached a steady-state value. The feedrate was then increased back to vw=4 mm/s and the wheel speed was decreased from vs=18 mm/s to vs=13 mm/s. Immediately the specific energy dropped back down to e~600 J/mm 3 and then continued to decrease to what appeared to be a steady-state value of e~500 J/mm 3. Wheel wear was also measured via the change in part dimension. The results are given in Figure 4. Here we can see very steady wear for the first parameter set, giving a value of G-ratio=10. Then, when the feedrate was decreased, the wheel wear rate increased, to G-ratio=7. This is counter to what is typically seen, with larger aggressiveness values giving lower G-ratios. However, at these high specific energies, resin-softening is likely a large contributing factor to wheel-wear rates, as found when grinding of tungsten-carbide [Badger, 2015] in spite of the smaller forces on individual grits with the smaller aggressiveness. Then, when the feedrate was increased again and the wheel speed dropped, the wheel-wear rate decreased again, to G-ratio=14.

5 wheel wear wheel (mm wear depth) (mm) Specific Energy (J/mm 3 ) 3 ) 2000 number of sawblades ground Parameter Set 2 feedrate = 1 mm/s wheel speed = 18 m/s Parameter Set 1 wheel speed = 18 m/s Parameter Set 3 wheel = 13 m/s break-in Volume volume of of material cermet removed per per wheel wheel area, area, V '' V w (mm w (mm 3 /mm 3 /mm 2 ) 2 ) Figure 3. Specific energy at three different parameter sets number of sawblades ground G-ratio = 10 (Aggr=11.1) Parameter Set 1 wheel = 13 m/s Parameter Set 3 Parameter Set 2 G-ratio = 7 (Aggr=2.8) feedrate = 1 mm/s 43 G-ratio = 14 (Aggr=15.4) break-in volume Volume of of cermet material removed removed per per wheel wheel area, area, V '' w (mm V w (mm 3 /mm 32 /mm ) 2 ) Figure 4. Wheel wear at three different parameter sets.

6 specific energy 5. Discussion To gain a better understanding of the relationship between aggressiveness, grinding power, wheel wear and self-sharpening, specific energy was plotted vs. the Aggressiveness. This is shown in Figure 5. Typically a higher Aggressiveness (and, in turn, a larger chip-thickness) leads to larger forces on individual grits and, therefore, larger wheel wear and smaller G-ratios. However, this is not always the case, particularly in difficult-to-grind materials when using resin bonds [Badger, 2015]. Because the specific energy was transient due to grit dulling and/or loading the range of values are plotted, with an arrow indicating whether they increased or decreased throughout the test. The results show that specific energies were higher at lower Aggressiveness values. More importantly, when the transient condition is considered, the differences in specific energy are drastic, with very high values of e~1900 J/mm 3. In addition, it can be seen that self-sharpening of the wheel was very poor at the low aggressiveness values. This has important implications in terms of the operator s choice of speeds and feeds. Typically, when an operator experiences a problem for example, burn or chatter his first reaction is to decrease the feedrate. This may initially help the situation (Figure 3, sawblade 27, 28, 29). However, the poor self-sharpening means that specific energies will rise drastically, and eventually exacerbate the problem (figure 3, sawblade 33, 34, 35). A similar situation can be seen with the G-ratio. Typically a higher Aggressiveness leads to larger forces on individual grits and, therefore, larger wheel wear and smaller G-ratios [Malkin, 2008]. However, this is not always the case, particularly in difficult-to-grind materials when using resin bonds [Badger, 2015]. Figure 6 shows higher G-ratios at higher Aggressiveness values. This indicates that softening of the resin due to higher temperatures J/mm J/mm J/mm 3 Parameter Set 2 feedrate = 1 mm/s wheel = 18 m/s 1400 J/mm J/mm J/mm J/mm J/mm 3 Parameter Set 3 wheel = 13 m/s 400 J/mm J/mm 3 Parameter Set 1 wheel = 18 m/s 0 J/mm aggressiveness number, Aggr Figure 5. Specific energy vs. Aggressiveness.

7 G-ratio Parameter Set 3 12 wheel = 13 m/s 10 8 Parameter Set 2 Parameter Set 1 wheel = 18 m/s 6 feedrate = 1 mm/s wheel = 18 m/s aggressiveness Aggressiveness, number, Aggr Aggr Figure 6. G-ratio vs. Aggressiveness. Grinding of tungsten-carbide is accompanied by loading of the swarf within the pores of the wheel [Badger, 2015; Metzger and Torrance, 1990], which leads to rubbing against the workpiece and higher grinding temperatures. However, no evidence of diamond-grit dulling has been seen when grinding tungsten-carbide [Malkin and Zelwer, 1980; Badger, 2015] even though it has been shown to occur when grinding harder ceramic materials [Luo et. al., 1997]. Loading has been seen when grinding cermets [Metzger and Torrance, 1990]. However, it has not been established whether the high specific energies typically associated with cermets grinding [Kumar, 1995; Toyoma et. al., 1993] are due solely to loading, or whether diamond-grit dulling is also occurring. To investigate this, electron-microscope photos were taken of the grinding wheel a) after truing and sticking; b) after grinding; and then c) after stick-conditioning. Photographs of the wheel after grinding (b) showed severe loading, such that the diamond grits could barely be seen above the surface. Theoretically, proper stick-conditioning removes loaded material but leaves the diamond grits in place. Therefore, photographs of the wheel after sticking (c) should reveal whether diamond dulling is occurring. Figure 7 shows the wheel after grinding and sticking. Here we can see very dull diamond grits. These grits also exhibit striations, showing the direction of wear. In addition, several of the grits do not show dulling. It is assumed that these were either grits that were below the surface and had yet to make contact with the workpiece, or had fractured. Considering that grit dulling is occurring, and that other researchers have found that friable grits perform much better than tough grits, the results here indicate that aggressive grinding conditions are much preferred, most likely because they result in a steady fracturing of diamond grits before they become excessively dull. In the tests here, true steady-state conditions were never reached. However, extreme dulling and loading accumulation were seen at Aggr = 2.8, and it appears that a steady-state condition could possibly be reached somewhere between Aggr=12 to Aggr=15. This is very important if reasonable periods between wheel trueings are to be achieved. This is particularly important considering that many operators decrease feedrates when they run into trouble. It is important that they realize the negative effect this has. One option would be to decrease wheel speed proportionately to the decrease in feedrate. In other words, in the case above, a decrease in feedrate from vw=4 mm/s to vw=1 mm/s should be accompanied by a decrease in wheel speed from vs=18 mm/s to vs=4.5 mm/s. This may seem extreme, but would ensure the same aggressiveness, provided the spindle motor is able to provide ample torque at these low speeds. Another option is to use wheels with a smaller rim width.

8 mm mm Figure 7. SEM photos of wheel after grinding and stick-conditioning. 6. Conclusions A new type of wheel geometry in cup-wheel grinding has been introduced, where a taper and a flat are trued into the wheel. This allows both a high feedrate and a good surface finish. The correct choice of grinding parameters is enormously important when grinding cermets. If the aggressiveness is too small, wheel self-sharpening is poor and huge specific energies are reached along with low G- ratios, presumably due to temperature-induced resin-softening. If chip-thickness is high, specific energies are lower, wheel-self-sharpening is better, wheel wear is lower, and more-steady-state conditions are achieved. Both loading and diamond-grit dulling occurs when grinding of cermets. 7. Acknowledgements The authors would like to thank Jürgen Hauger, Patrick Kopf, Daniel Schöllhorn and Holger Reisch at Vollmer in Biberach, Germany; Peter Allen at Vollmer USA; and David Spelbrink of LANDS Superabrasives in New York City. References Badger, J. (2008) Aggressiveness in grinding, CIRP 3rd International Conference on High Performance Cutting, Dublin, Ireland, pp Badger, J. (2009,2) Factors affecting wheel collapse in grinding, CIRP Annals Manufacturing Technology, 58/1, Badger, J. (2015) Grinding of sub-micron-grade carbide: Contact and wear mechanisms, loading, conditioning, scrubbing and resin-bond degradation, CIRP Annals. Kumar, C, 1995, Grinding of Cermet Tool Materials, Proc. of the SuperGrind Conference, Luo, S., Liao, Y., Chou, C., Chen, J., 1997, Analysis of the Wear of a Resin-bonded Grinding Wheel in the Grinding of Tungsten Carbide, Jrnl. of Materials processing Technology, 69: Malkin, S., Zelwer, O. (1980) Grinding of WC-Co Cemented Carbides, Transaction of the ASME Journal of Engineering for Industry, 102: Malkin, S., Guo, C.; 2008, Grinding Technology: Theory and Applications of Machining with Abrasives, Second Edition; Metzger, J.L., Superabrasive Grinding, Publisher: Butterworth-Heinemann, 1986 Metzger, J., Torrance, A., 1990, Dry vs. Wet Grinding of Carbide, Industrial Diamond Review, 6: Toyama, Isao; Inasaki, Ichiro; and Shiratori, Hidehisa High Efficiency Grinding of Cermet, Japan Society for Mechanical Engineers, 59/558, , No ,