Int J Adv Manuf Technol (2009) 42:883–891 DOI 10.1007/s00170-008-1646-7 ORIGINAL ARTICLE Ultrasonic assisted dry grinding of 42CrMo4 Taghi Tawakoli & Bahman Azarhoushang & Mohammad Rabiey Received: 8 April 2008 / Accepted: 27 June 2008 / Published online: 18 July 2008 # Springer-Verlag London Limited 2008 Abstract In recent years, many cooling lubricants are classified as health hazards, while their end-of-life treatment poses numerous ecological threats. Dry machining is a solution for greener production and also is a favorable process from an economical point of view. However, compared to other machining processes, conventional grinding has a low material removal rate and involves high specific energy. A major part of the specific energy in grinding is changed to heat that makes harmful effect on surface quality. Therefore, in conventional dry grinding, as there are no cutting fluids to transfer the heat from the contact zone, the temperature of workpiece surface and grinding wheel surface will be increased resulted to thermal damage and poor surface integrity, increasing of wheel wear and inefficient grinding compared to conventional grinding. To make a step forward to pure dry grinding and to eliminate the negative environmental impact of the cutting fluids, a new technique called ultrasonic assisted dry grinding has been used. The advantages of ultrasonic assisted grinding were proved mostly for the brittle material. Our investigations show the improvement on the surface roughness, considerable reduction of the normal grinding force, and thermal damage in case of using ultrasonic assisted dry grinding compared to conventional dry grinding for a soft material, 42CrMo4. A decrease of up to 60% of normal grinding forces has been achieved. Keywords Dry grinding . Ultrasonic machining . Ultrasonic assisted dry grinding . Cooling lubricants . Grinding forces . Surface roughness T. Tawakoli : B. Azarhoushang (*) : M. Rabiey Institute of grinding and precision technology (KSF), Furtwangen University, Villingen-Schwenningen, Germany e-mail: B.Azarhoushang@hs-furtwangen.de 1 Introduction The main functions of cutting fluids in metal removal processes are cooling and lubricating the contact zone, swarf disposal, minimizing the wear of the cutting tool, and cleaning and minimizing corrosion. On the other hand, the disadvantages of using cutting fluids include, from an environmental point of view, health hazards, problems involving the disposal of thousands of tones of a year of used emulsion and oil–water mixtures—as well as a vast amount of grinding slurry, filter fleece, and filter process materials. From an economical point of view, coolant and coolant management costs are almost 20% of the total manufacturing costs. As a result, there is a growing interest in dry machining or processes that require least quantities of lubricant. To decrease the negative environmental impact of the cutting fluids and reducing manufacturing costs, new machining techniques such as dry machining [1–4] are used. Grinding is one of the most difficult processes with regard to eliminating cutting fluids. During grinding, many of the super abrasive grits that are in contact with the workpiece do not perform real cutting, but instead generate heat by rubbing and plowing the workpiece surface in the contact zone. The high heat generation associated with a high negative rake angle and with a great contact length in grinding processes can greatly increase the temperature in the contact zone. Without sufficient cooling and lubrication, this can cause thermal damage on the workpiece surface [5–6]. Due to these technical problems, cutting fluid is almost necessary in most grinding applications, and the methods of minimum grinding fluid or dry grinding have not yet been fully successful in industrial applications [7, 8]. Generally, in conventional dry grinding (CDG), as there are no cutting fluids to transfer the heat from the contact zone, the temperature of the workpiece surface and grinding 884 wheel surface will be increased resulting to thermal damage and poor surface integrity, increasing wheel wear and inefficient grinding compared to conventional grinding. A recent and promising technique to overcome these technological constraints is known as ultrasonic assisted dry grinding (UADG). The principle of this technique is to superimpose high frequency (16–40kHz) and low vibration amplitude (2–15μm) in the feed or crossfeed direction to the tool or the workpiece. This cutting process is different from ultrasonic machining. In ultrasonic machining, metal removal is effected with the help of abrasive grains suspended in a slurry, which are made to strike repeatedly upon the workpiece surface by a tool oscillating ultrasonically [9, 10]. Ultrasonic machining has been only applicable to brittle materials. On the other hand, UADG is a hybrid process of CDG and ultrasonic oscillation. It is applicable to both ductile and brittle materials. By using ultrasonic assisted machining, significant improvements in thrust force, burr size, material removal rate, tool wear, heat generation, noise reduction, and surface finish have been reported. Zhang et al. [11] have both theoretically and experimentally concluded that there exists an optimal vibration condition such that the thrust force and torque are minimized. Jin and Murakawa [12] found that the chipping of the cutting tool can effectively be prevented by applying ultrasonic vibration and tool life can be prolonged accordingly. Azarhoushang and Akbari [13] have achieved significant improvements in the circularity, cylindricity, surface roughness, and hole oversize by applying ultrasonic vibration to the drilling tool without using any cutting fluids. Prabhakar et al. [14] have experimentally demonstrated that the material removal rate obtained from ultrasonic assisted grinding (UAG) is nearly six to ten times higher than that from a conventional grinding process under similar conditions. Mult et al. [15] and Uhlmann [16] Int J Adv Manuf Technol (2009) 42:883–891 found that, for ceramic materials, UAG can be applied as an efficient production technology and the ultrasonic assisted creep feed grinding provides enormously reduced normal forces at slightly increased wheel wear and surface roughness. Tawakoli et al. [17] demonstrated that, in ultrasonic assisted dressing of cubic boron nitride (CBN) grinding wheels, considerable reduction in grinding forces and dresser wear is achievable. In this investigation, a UADG system has been designed, fabricated, and tested. Improvements in the Rz (parameter of surface roughness) of the ground surfaces, reduction of the normal grinding force, and thermal damages on the ground surface due to superimposing of ultrasonic vibration in the dry grinding of 42CrMo4 have been achieved. Globally, the temperature fields measured in chips as well as maximal temperature are higher in the case of the 42CrMo4 steel compared to mild steels and medium carbon steels resulting from the increase of the friction coefficient (for V = 30m/s, μ = 0.2 [18]) and the higher yield stress. Sutter and Ranc [19] have experimentally demonstrated that the temperature fields measured for cutting speed around 20m/s present maximums of 870°C for 42CrMo4 and 630°C for C15 located near the tool–chip interface. Therefore, the 42CrMo4 steel was chosen as the test workpiece in this experimental investigation to emphasize the effect of superimposing ultrasonic vibrations in dry grinding of soft steel. The effect of vibration amplitude, feed speed, and depth of cut on surface roughness and the normal grinding force have been investigated. 2 Design and fabrication of UADG system To study UADG, an actuated workpiece holder has been designed and built. Figure 1a illustrates schematically the (a) Fig. 1 a Scheme of the experimental setup. b Experimental setup for UADG (b) Int J Adv Manuf Technol (2009) 42:883–891 experimental setup. The workpiece holder consists of a piezoelectric transducer, a booster, a horn, and a special fixture. The ultrasonic power supply converts 50Hz electrical supply to high-frequency (21kHz) electrical impulses. These high frequency electrical impulses are fed to a piezoelectric transducer and transformed into mechanical vibrations of ultrasonic frequency (21kHz) due to the piezoelectric effect. The vibration amplitude is then amplified by the booster and the horn and transmitted to the workpiece attached to the horn. The resultant vibration of the workpiece fixed in the tool holder reaches 10μm (i.e., 20μm peak to peak) at a frequency of about 21kHz. Vibration is applied to the workpiece in the crossfeed direction of the grinding wheel. The amplitude of the ultrasonic vibration can be adjusted by changing the setting on the power supply. The experimental setup used to study UADG is shown in Fig. 1b. In the design of the UADG acoustic head, it is considered that the whole structure must possess enough stiffness to withstand the dynamic loads during the grinding operation. The acoustic head parts should have high fatigue resistance and low acoustic losses (meaning that they should not absorb too much energy from the vibrations). Each part of the acoustic head is made of aluminum 7075T6 with high strength, high fatigue resistance, and very good acoustic properties to provide enough stiffness and low acoustic losses. The fixture that clamps the acoustic head is made of steel. 3 Experiments The experimental equipment consists of the following: & & Machine tool: Elb Micro-Cut AC8 CNC universal surface grinding machine Ultrasonic vibration generator: Mastersonic MMM generator-MSG.1200.IX, power of 12,000W, frequency ranges of 17.000 to 46.728kHz Table 1 Major machining parameters 885 & & & Eddy current displacement measurement system: Micro epsilon eddyNCDT 3300, to measure the amplitude of vibration. Measuring ranges 0–0.5mm, linearity 0.2%, resolution 0.005%, measuring rate 100kHz Surface roughness tester: Hommel-Werke, model T-8000 Dynamometer: Kistler piezoelectric dynamometer model 9255B The settings of the main machining parameters for the present study are summarized in Table 1. The tests were carried out for both UADG and CDG with the same instrument. However, during the CDG, the ultrasonic generator was switched off. Every workpiece was divided into three different sections, and UADG experiments were applied on the center section (Fig. 2). 4 Experimental results and discussion Most of the CDGs were unsuccessful due to the thermal damage on the ground workpiece surface. As there were no cutting fluids to transfer the high heat from the contact zone, this result had been expected. Figure 2 shows photographs of the ground surfaces. It is apparent that, in both samples, the middle section (ultrasonically assisted ground surface) has experienced much less thermal damage compared to other sections (conventional ground surfaces). The effect of vibration amplitude, feed speed, and depth of cut on surface roughness and normal grinding force were studied. To achieve reliable data, each test was repeated three times. In all the figures, lines were formed by calculating the least-squares fit through the data points for a second-order polynomial equation. Figure 3 shows that the relationship between vibration amplitude and normal grinding force. The amplitude zero in this figure represents results of CDG. The experimental results show significant improvement for UADG compared to CDG in different vibration amplitudes. Apparently, the reason for these improvements is the change of the nature of the cutting Parameters Grinding wheel Workpiece Grinding conditions Grinding process Dressing conditions Dressing tool Direction of ultrasonic vibration Ultrasonic vibration conditions Vitrified bond CBN grinding wheel, B126 C125; ∅400×16 mm 42CrMo4, 85 HRB, (60×50×30 mm) Feed speed vft =1,000–2,000 mm/min; cutting speed vc =60 m/s; depth of cut ae =0.010–0.030 mm; no coolant (dry grinding) Dry surface grinding Dressing ratio qd =0.8, wheel speed vcd =60 m/s, depth of dressing aed =5 μm, overlapping ratio Ud =4, total depth of dressing aed-total =10 μm Diamond disc dresser radius Rsp =0.2 mm Cross feed direction (perpendicular to feed) Frequency f=21 KHz, amplitude A=10µm 886 Int J Adv Manuf Technol (2009) 42:883–891 Fig. 2 (I) vft =1,000 mm/min, vc =60 m/s, (a) ae =15 μm (CDG), (b) ae =15 μm, A= 10 μm (UADG), (c) ae =10 μm (CDG); II vft =2,000 mm/min vc =60 m/s, (a) ae =25 μm (CDG), (b) ae =30 μm, A= 10 μm (UADG), (c) ae =30 μm (CDG) (I) process, which is transformed into a process with a multiple-impact interaction between the abrasive grits and the formed chip. Figures 4, 5, 6 and 7 compared the normal grinding force and surface roughness produced by UADG with CDG under different depth of cuts. Experiments were carried out at vc = 60m/s, f = 21kHz, A = 10μm. Based on the results from previous stages, it is believed that UADG performs enhanced under these conditions. These conditions are not essentially the optimal ones. For depths of cuts more than 10μm in CDG, thermal damages of the ground surfaces were observed. This phenomenon is shown with a fire symbol in Figs. 4 and 5. It should be noted that the scatter in the measured surface roughness and grinding forces obtained through UADG is much less compared to CDG. It means that using UADG increases the repeatability of the process. The maximum oscillating velocities (up to 80m/min) and accelerations (up to 174,100m/s2) are generated at the Fig. 3 Normal grinding force vs. vibration amplitude (ae =20µm, f=21 kHz) (II) amplitude of 10μm and a frequency value of 21kHz. The larger the vibration amplitude, the greater the material removal rate per active grain and the higher the kinetic energy with which the grits strike the work surface. Due to the high frequency interaction of active grains on the workpiece, the cutting process in UADG becomes discontinuous, and ultrasonic impact action occurs, thus causing the material to begin to rollover more easily as well as more micro cracking propagation in the cutting zone, which both make an effective interaction between grits and workpiece surface. Therefore, the grinding forces and frictional effects are decreased so that less plastic deformation occurs. It has already been proven by some researchers [20] that deformation processes for ultrasonic assisted machining are restricted in the vicinity of the cutting edge along the surface of the workpiece and are not observed underneath the cutter, in contrast to the conventional machining process. Plastic deformation of the machined surface in Int J Adv Manuf Technol (2009) 42:883–891 887 Fig. 4 Grinding normal force vs. depth of cut, vft =1,000 mm/min case of using ultrasonic oscillation is less than that in conventional machining. In addition, the coefficient of friction in grinding decreases with an increase in sliding speed between the grit and the material. As the sliding speed in UADG due to ultrasonic vibration is higher than sliding speed in CDG, the coefficient of friction reduces. This suggests that, in UADG, a fewer number of strong bonds between the grit and the material are formed. Authors assume that, by oscillation of the workpiece in Fig. 5 Grinding normal force vs. depth of cut, vft =2,000 mm/min crossfeed direction, the rubbing and plowing regimes that cause the major part of plastic deformation are reduced so that the grinding specific energy is also reduced and the thermal damage on the ground surface is significantly decreased. Ultrasonic vibrations cause a reduction in friction because they apply an additional stress to assist in breaking the instantaneous welds and they reduce the time that any two asperities on opposite surfaces may remain in momen- 888 Int J Adv Manuf Technol (2009) 42:883–891 Fig. 6 Ra and Rz vs. depth of cut, vft =1,000 mm/min (UADG: A=10µm, f=21 kHz) tary contact and, hence, keep them from forming a stronger weld. In addition, as the sliding speed in UADG due to ultrasonic vibration is higher than sliding speed in CDG, the coefficient of friction reduces. This suggests that, in UADG, a fewer number of strong bonds between the grit and the material are formed and the surface roughness is improved. Figure 8 shows a surface roughness profile that compares an ultrasonically ground surface profile with a conventionally ground surface profile. 4.1 Kinematics of UAG To investigate the theoretical aspects of ultrasonic grinding, it is necessary to study the chip formation process by UADG and compare it with CDG. Fig. 7 Ra and Rz vs. depth of cut, vft =2,000 mm/min (UADG: A=10µm, f=21 kHz) The mean uncut chip area in the grinding process can be calculated as [21]: Acu tot ¼ N mom X Acu;i  Acu
N mom ð1Þ i¼1 where N mom is the momentary number of kinematics cutting grains, Acu, i is the uncut chip area of the active grain i, Acu-tot is the total and Acu is the mean uncut chip area. On the other hand, the total volume of material removal in contact zone can be written as: Vw  Lg
N mom X i¼1 Acu;i ð2Þ Int J Adv Manuf Technol (2009) 42:883–891 889 Fig. 8 Surface roughness profile (vc =60 m/s, vft = 1,000 mm/s, ae =30 μm) where is Lg the kinematics contact length. Using Eq. 1, we have: Vw  Lg
Acu
N mom ð3Þ So far, for the case of UADG and CDG, it can be written sinusoidal path due to the ultrasonic vibration of the workpiece. The length of path due to the vibration depends on the time which a grain is in the contact length. This time can be written as: as: Vw UADG Vw CDG  Lg  Lg UADG CDG
Acu
Acu UADG
N mom CDG :N mom CDG UADG ð4Þ t¼ lg CDG vc  vw ð9Þ ð5Þ as the total material removal is the same for both UADG and CDG, so: Vw UADG ¼ Vw CDG : ð6Þ Looking at the Fig. 9, it can be seen that Acu in each cross-section of the contact length depends on the location of the grain on the contact length as well as the depth of penetration. Theoretically, the ultrasonic vibration has a very low influence on Acu when the direction of the ultrasonic wave is perpendicular to the direction of grain movement on the contact length as in the case of this study. With this analogy, we have: Acu UADG  Acu CDG Using Eqs. 2–7:
 1 Lg UADG N mom UADG ¼ : N mom CDG Lg CDG ð7Þ ð8Þ From Fig. 9, it can be seen that the path that a grain moved on it in the contact zone is the superposition of a rather straight line of kinematics contact length and the Fig. 9 Grinding model for UAG. a Schematic of the motion of a grain on the workpiece surface in UAG. b Path generation of grains by UAG 890 Int J Adv Manuf Technol (2009) 42:883–891 And the path due to ultrasonic vibration Lvib will be: Lvib ¼ 2
p
A
lg
f vc  vw The lower specific energy in Eq. 21 causes the lower heat generation and lower surface temperature. ð10Þ where f is the frequency of the ultrasonic vibration and lg is the geometrical contact length, vc is the grinding speed, vw is the workpiece speed, and A is the amplitude of the ultrasonic vibration. The plus sign in the denominator is related to up-grinding and the minus sign is related to down-grinding. Therefore, in general form: 5 Conclusion Experimental studies of UADG and CDG demonstrate considerable advantages of the former technology for dry grinding 42CrMo4. & Lg UADG lg  Lg CDG þ 2
p
A

f vc  vw ð11Þ Based on Eq. 11, it can be seen that: Lg UADG > Lg ð12Þ CDG Now, assume that: 2
p
A
lg
f ¼ l
Lg vc  vw CDG ð13Þ where l is a constant, then using Eqs. 7, 12, and 13: N mom UADG ¼ ð 1 þ lÞ 1 : N mom CDG ð14Þ By the way, the tangential grinding forces have a direct relation with total mean uncut chip area [21]: Ft / Acu ð15Þ tot and the total tangential force is the summation of forces of all active grain in contact zone, so: Ft / Acu
N mom : ð16Þ & Comparative experiments of the grinding forces demonstrated up to 60% reduction in normal grinding force for the workpieces machined with superimposed ultrasonic vibration. Most of CDGs were unsuccessful due to the thermal damage on the ground workpiece surface. The reason for this phenomenon was due to the absence of cutting fluids in the process and, consequently, the generation of high heat in the contact zone. These improvements are subjected to the change of the nature of the cutting process in UADG, which is transformed into a process with a multiple-impact interaction between the tool and the formed chip resulting in interrupted cutting and reducing the grinding forces, frictional effect, and plastic deformation zone. It was also found that using UADG leads to significant improvements on the Rz parameter. Future studies will include the use of ultrasonic oscillation in the feed direction and the comparison of the corresponding process parameters. Acknowledgments The authors would like to acknowledge the help of Mr. Nima Jandaghi and Mr. Rolf Rinderknecht in conducting experiments. Using eqs. 7 and 14–16: Ft UADG ¼ ð 1 þ lÞ Ft CDG 1 ð17Þ On the other hand, the normal forces and tangential forces have the following relation: Fn ¼ m
Ft ð18Þ where μ is a constant. Therefore, using Eqs. 17 and 18: Fn UADG ¼ ð 1 þ lÞ Fn CDG 1 ð19Þ As the specific energy: ec ¼ Ft
vc b ð20Þ where b is the wheel width, then: eUADG ¼ ð 1 þ lÞ eCDG 1 ð21Þ References 1. Bell DD et al (1999) Modeling of the environmental effect of cutting fluid. Tribol Trans 42(1):168–173 doi:10.1080/10402009908982204 2. Heiler R (2005) Dry machining of internal screw threads. Werkstatt Betrieb 138(10):54–56 3. Tawakoli T, Rabiey M (2007) Dry grinding by special conditioning. Int J Adv Manuf Technol 33(3–4):419–424 (June) 4. Malshe AP, Arumugam PU, Batzer SA (2006) Dry machining of aluminumsilicon alloy using polished CVD diamond-coated cutting tools inserts. Surf Coat Tech 200(11):3399–3403 doi:10.1016/j.surfcoat.2005.08.127 5. Voll M Modelle zur thermischen Optimierung von Trockenschleifprozessen, Dissertation an der Technischen Universität Chemnitz (eingereicht 06.03.2000) 6. Heinzel C Methoden zur Untersuchung und Optimierung der Kühlschmierung beim Schleifen, Dissertation Universität Bremen, 2003 7. T. Tawakoli and M. Rabiey, Trockenschleifen, Grenzen und Möglichkeiten, 6. Seminar ,,Moderne Schleiftechnologie und Feinstbearbeitung“in Stuttgart, Hrsg. T. Tawakoli, 17.05.2006 Int J Adv Manuf Technol (2009) 42:883–891 8. Tawakoli T, Rabiey M An Innovative Concept for Dry grinding with Resin and Vitrified Bond CBN Wheel, International Grinding Conference ISSAT 2007/ISME 9. Goldman R (1962) Ultrasonic technology. Reinhold, London 10. Jana JK, Satyanarayana A (1973) Production of fine diameter holes on ultrasonic drilling marching. J Inst Eng Electron (India), Part MC: Mechanical Engineering Division 54(1):36–40 11. Zhang LB, Wang LJ, Liu XY, Zhao HW, Wang X, Luo HY (2001) Mechanical model for predicting thrust and torque in vibration drilling fibre-reinforced composite materials. Int J Mach Tools Manuf 41:641–657 doi:10.1016/S0890-6955(00)00105-X 12. Jin M, Murakawa M (2001) Development of a practical ultrasonic vibration cutting tool system. J Mater Process Technol 113:342–347 13. Azarhoushang B, Akbari J (2007) Ultrasonic-assisted drilling of Inconel 738-LC. Int J Mach Tools Manuf 47(7–8):1027–1033 June 14. Prabhakar PD, Ferreira PM, Haselkorn M (1992) An experimental investigation of material removal rate in rotary ultrasonic machining. Trans North Am Manuf Res Inst SME 20:211–218 891 15. Mult HC, Spur G, Holl SE (1996) Ultrasonic assisted grinding of ceramics. J Mater Process Technol 62:287–293 16. Uhlmann E (1998) Surface formation in creep feed grinding of advanced ceramics with and without ultrasonic assistance. Annals of CIRP, Vol.47/1/1998 17. Tawakoli T, Westkaemper E, Rasifard A (2007) Ultrasonic assisted dressing of vitrified CBN grinding wheel. 40th CIRP International Seminar on manufacturing Systems, Liverpool, UK 18. Sutter G, Molinari A (2005) Analysis of the cutting force components and friction in high speed machining. J Manuf Sci Eng 127:245–250 19. Sutter G, Ranc N (2007) Temperature fields in a chip during highspeed orthogonal cutting—an experimental investigation. Int J Mach Tools Manuf 47:1507–1517 20. Babitsky VI, Mitrofanov AV, Silberschmidt VV (2004) Ultrasonically assisted turning of aviation materials: simulations and experimental study. Ultrasonics 42:81–86 21. Zeppenfeld C (2005) Schnellhunschleifen von-Titanaluminiden. Dissertation RWTH, Aachen