Cutting temperatures in hard turning chromium hardfacings with PCBN tooling

Cutting temperatures in hard turning chromium hardfacings with PCBN tooling

Journal of Materials Processing Technology 147 (2004) 38–44 Cutting temperatures in hard turning chromium hardfacings with PCBN tooling X.J. Ren a,∗ ...

395KB Sizes 2 Downloads 62 Views

Journal of Materials Processing Technology 147 (2004) 38–44

Cutting temperatures in hard turning chromium hardfacings with PCBN tooling X.J. Ren a,∗ , Q.X. Yang b , R.D. James c , L. Wang d a

School of Engineering, Liverpool John Moores University, Liverpool L3 3AF, UK b School of Materials Science and Engineering, Yanshan University, PR China c School of Engineering, University of Hull, Hull, UK d Engineering Department, University of Oxford, Oxford, UK Received 24 April 2003; accepted 24 October 2003

Abstract Hard turning of high chromium hardfacing materials is a new application field for PCBN tools. An investigation of the cutting temperature is essential in understanding the machining mechanism of the material, thus to improve the process efficiency. In this work, cutting temperatures of two typical hardfacing materials were studied using a mixed experimental and numerical approach— remote thermocouple technique and finite element (FE) simulation. The effect of microstructure and machining parameters on the cutting temperatures were comparatively investigated using titanium alloy as a reference material. The average cutting temperatures of the hardfacings were found to be ranged from 600 to 700 ◦ C and increased with higher cutting speed and feedrate. A hardfacing with larger carbide grains showed lower cutting temperatures and exhibited lower increase rate with cutting speed and feedrate. © 2003 Elsevier B.V. All rights reserved. Keywords: Cutting temperature; Hardfacing; Remote thermocouple; Finite element analysis; PCBN; Hard turning

1. Introduction High chromium hardfacing materials are widely used in industry due to its excellent wear resistance [1]. The wear resistance of these materials was mainly achieved by a high hardness and high carbide contents and this makes machining of these hardfacings extremely difficult [2,3]. The development of ultra-hard CBN material has opened up the possibility to machine these materials by turning or milling instead of grinding, thus improving the productivity and reducing the cost [4]. The chip formation process of hardfacing and the tool wear characteristics have been previously investigated [5,6]. This paper further reports on measurement and prediction of machining temperatures of hardfacings at different cutting regimes, which is another key factor of the machinability of materials, and the effect of microstructure on the cutting temperatures. The temperature generated during a machining operation is influenced, mainly, by the properties of the workpiece and the tool material, the machining parameters used and ∗ Corresponding author. E-mail addresses: [email protected], [email protected] (X.J. Ren).

0924-0136/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2003.10.013

other condition, e.g. coolant. Factors, such as tool wear, surface quality and chip formation, etc. are all affected by the cutting temperature. In many cases, the temperature was a limiting factor of the cutting tool efficiency [7–9]. When cutting some very hard materials using ceramic tools, e.g. case hardened alloy steel and super alloys, sufficient cutting heat is needed to soften the workpiece material [10–12]. In this process, an ideal machining temperature would generate local ductility of the workpiece material without causing significant deterioration of the tool strength [13]. In this context, the thermal phenomenon is of significant relevance in the machining mechanics of difficult-to-cut materials. Hard turning of high chromium hardfacings is a similar operation and an investigation of the cutting temperature would provide beneficial information for the understanding of the cutting mechanism, thus improving the process efficiency. Several experimental techniques have been used to measure cutting temperatures, including tool–workpiece thermocouple techniques, radiation, metallographic techniques and embedded thermocouple [7,9,14,15]. Each technique has its own advantage over others in certain applications depending on the tool materials, workpiece materials and other machining conditions. For applications with very hard tool materials, embedded thermocouple techniques were widely

X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44


used due to its lower cost and easy to operate [13,14]. One method of this technique involves positioning the thermocouple at the back of the cutting insert, i.e. at the interface of the insert and its support [16,17]. This method is much easier to operate and avoids the difficulty of producing holes in the tool materials, which is extremely difficult when using ultra-hard tools. The temperature of the real cutting edge can be deduced based on theoretical temperature fields [16–19], or finite element simulation [15,20]. Another advantage of this method is that the required thermocouples can be built into the tool holder, making the method attractive for routine measurements and process monitoring [21]. In this work, cutting temperatures when machining two typical high chromium hardfacings were investigated by means of a thermocouple located between the tool (PCBN) and its supporting shim. A finite element (FE) model was used to predict the chip–tool interface temperature and its distribution in the machining system. The cutting temperatures of hardfacing with different microstructures under various cutting conditions were determined. The effect of workpiece microstructure and cutting parameters on the cutting temperatures and their influence on the operating wear mechanism of cutting tools are discussed.

2. Experimental procedure A hardfacing layer, nominally 6 mm thick, had been deposited on a mild steel bar using a flux-cored arc welding (FCAW) machine. Hardfacings of two typical microstructures were used as the workpiece materials in the current work (Fig. 1). The solidification begins with the formation of primary (Cr, Fe)7 C3 carbides, the residual liquid decomposing eventually by a ternary eutectic reaction into a mixture of austenite and more (Cr, Fe)7 C3 [22]. Hardfacing A is characterised by a relatively fine structure, the needle-like bright image and the tangled structure is the eutectic of fine carbides (1 ␮m in diameter) in an austenite matrix. The much larger grains are the primary carbides (8–15 ␮m in diameter). In hardfacing B (Fig. 2), a large portion of primary carbides has formed with the remaining part transformed into mixture of austenite and finer carbides. The primary carbides are much bigger than those in hardfacing A, up to 20–35 ␮m in cross-section. The hardness of hardfacings A and B are 57 and 59 HRC, respectively. The titanium alloy used in the test is an as-rolled and annealed TA48 titanium alloy. It has a microstructure consisting of elongated alpha phase in a fine dark-etching beta matrix and a hardness of 41 HRC. The microstructure and high hardness made these materials very difficult to machine and only PCBN tools with high volume fraction could be used [4]. The inserts used in this work are PCBN tools with high cBN content (over 90 vol.%) and a ceramic binder (AlN). The microstructure of the PCBN material, after leaching out the binder phase, is shown in Fig. 2. It can be seen that CBN particles, with

Fig. 1. Microstructure of the hardfacings. (a) Hardfacing with fine structure (hardfacing A) and (b) hardfacing with coarse structure (hardfacing B).

an average grain size of 5–8 ␮m, have formed a PCBN network during synthesis. This accounts for the high hardness and the ability of the tool material to withstand the high mechanical loading developed during the cutting process. The hardness of the PCBN material measured with a load of 2.0 kg at 20 and 600 ◦ C are HK 36.5 and 15.5 GPa, respectively. This made the conventional method, i.e. drilling holes close to the cutting tip, very difficult. Similarly, implanted workpiece thermocouple was also not applicable due to the fact that the carbides in the workpiece are hard

Fig. 2. Microstructure of the PCBN tool material after the binding phase has been leached out using boiling hydrochloride solution.


X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44

Fig. 3. Schematic to show the set-up of the temperature measurement system.

and brittle. A remote thermocouple method with thermocouple(s) at the interface between the tool insert and supporting shim, coupled with FE modelling, would provide the best solution. The machining temperature of titanium alloy, which has been extensively studied [7,23,24], was also measured under similar cutting conditions as a basis for comparison. The turning tests were carried out, without a coolant, on a Churchill ‘Computurn’290 CNC lathe. The cutting tools used were solid circular PCBN inserts (RNMN0700300). A thin thermocouple (Ø0.5 mm) was mounted in a shallow groove on the silver steel shim, so it could detect the average temperature developed at the interface of the insert and the shim (Fig. 3). The K-type thermocouple used was a base metal system using nickel alloys, with a positive arm of nickel/chromium and a negative arm of nickel/aluminium. The junction was insulated by magnesium oxide (MgO) with a sheath of 310 stainless steel. Before each test, the system was calibrated in the laboratory. The reliability of the technique had been checked in the preliminary tests by repeating the same cutting condition and thermocouple condition several times using titanium alloy as the workpiece material, the results were consistent and satisfactory.

3. The finite element modelling The purpose of the FE modelling are in two folds, one is to investigate the temperature distribution within the machining system, especially temperature distribution in the tool insert and the tool–shim interface where the thermocouple is located. This will provide useful guidance for the experimental work. The other is to reversely calculate the

tool–chip temperature from the measured tool–shim temperature using a similar approach as [16,17,19]. A two-dimensional heat transfer simulation model was used as shown in Fig. 4. The set-up of the simulation model includes the tool insert, the supporting shim, the top clamp, the tool holder and two clamping blocks on the tool turret. The remaining parts of the turret and the lathe were not included because thermocouple measurements demonstrated that within the shank, there are very low temperature variations with respect to room temperature [25]. The model was meshed with 1030 elements. The element type used was Plane 77, which is a 2D quadrilateral element with eight nodes (ANSYS). For the tool holder and the clamping blocks of the turret, the heat exchange will not be as significant as in the parts near the insert and therefore a coarser mesh was applied to reduce processing time and increase efficiency. A much finer mesh is used at the insert tip as it approached the cutting zone to increase the accuracy (Fig. 5). It is known that the maximum tool face temperature tend to be close to the tool tip when machining difficult-to-cut materials, hence it was assumed that the heat generated from both the deformation process in the chip and the friction between the chip and the tool supplies a uniform heat force at the cutting tip [13]. The contact length data were determined from experimental measurements [5,23]. In the initial stage, i.e. at time zero, the whole system was assumed to be at a constant ambient temperature 293 K (20 ◦ C). In early stages (within 1 min) the heat generated over the flank face of the tool is very limited [25], so the temperature rise of the system was mainly due to the heat generated over the tool–chip contact on the rake face. This input was simulated in the model by applying ther-

X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44


Fig. 4. The FE model: the set-up of the simulation model and the meshing method.

mal loading (temperature) onto the contact region (Fig. 5). Three materials properties were specified for the model: density (ρ), thermal conductivity (k) and specific heat (Cp) . Density is assumed to be constant, while the other two properties are temperature-dependent, especially for the tool material. The value used for the thermal conductivity (k) of PCBN was varied from k(20 ◦ C) = 100 W/(m K) to k(900 ◦ C) = 130 W/(m K); the value of specific heat ranged from 1000 J/(kg K) (at 20 ◦ C) up to 1966 J/(kg K) (at 900 ◦ C) [3]. The density used for PCBN was 3120 kg/m3 . For the tool steel, constant thermal parameters were used (k = 47 W/(m K), Cρ = 486 J/(kg K) and ρ = 7800 kg/m3 ) since they would not vary significantly within the temperature range encountered in the tests [25,26].

Fig. 5. The FE model: meshing of the cutting tip and the temperature applied on the contact length.

4. Results and discussions 4.1. Experimental results Fig. 6 shows the increase of the measured temperature at the tool–shim interface with the cutting time (S = 65 m/min, F = 0.25 mm per revolution). Each data point on the plot represents the mean of three repeated tests. When the cutting started, the tool–shim interface temperature increased rapidly and after 15–20 s, it exhibited a gradual increase. The temperature when machining the three workpiece materials showed the same trend of increase and a similar period before reaching a stable stage. This result is in good agreement with other works [19]. The measured temperature in machining the titanium alloy was significantly higher than the hardfacings. The measured temperatures when machin-

Fig. 6. Temperature increases at the tool–shim interface with cutting time.


X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44

increasing feedrates while the temperature for the titanium alloy showed a more significant increase than that of the hardfacings. A higher rate of increase was also observed when machining hardfacing A than hardfacing B. SEM observations of the used cutting tools at the point of stopping of the cutting showed very little flank wear for all the three workpiece materials. This suggested that the heat generated over the flank face of the tool is probably very limited in this early stage (within 1 min). The three materials were machined with the same tool and cutting conditions; therefore, the tool–chip temperatures (θt ) could be analysed in terms of the specific cutting energy (u) and the kρC value of the workpiece material (Eq. (1)):   Vt 1/2 θt ∼ u (1) kρC Fig. 7. Effect of cutting speed on tool–shim interface temperature.

ing hardfacing A (fine structure) was lower than that of titanium alloy but is obviously higher than that of hardfacing B, which suggested that the microstructure of hardfacing material has significant effect over the cutting temperature. Average measured temperatures between 30 and 60 s were used to represent the temperature reached under different cutting parameters. Five speeds have been used (35, 45, 65, 85 and 125 m/min) to assess the effect of the cutting speed on the tool–shim temperature and the results are shown in Fig. 7. At the highest speed (125 m/min), machining of hardfacing B was extremely unstable due to very high cutting forces and no stable result was obtained. As shown in the curves, the measured temperatures increased with the cutting speeds in all cases while the curve for the titanium alloy was more linear than the hardfacings. The temperature when machining hardfacing A showed a much more significant increase than that of hardfacing B. Fig. 8 shows the temperatures measured at different feedrates. All the measured temperatures increased with

where u is the specific cutting energy, Vt is the cutting speed, k is the thermal conductivity of the workpiece, ρC is the volume specific heat of the workpiece material. The specific cutting energy (u) principally depends on the cutting force, which is associated with the strength of the material. The titanium alloy is softer than the other two materials and its specific cutting energy (u) has been reported to be relatively lower than iron-based alloys [7]. Cutting forces through the indicator on the machining centre also showed that the cutting forces in machining hardfacings were much higher than in machining the titanium alloy. Therefore, it could not be the reason for the higher temperature generated in machining the titanium alloy. The kρC of titanium is much lower than the hardfacings due to its lower thermal conductivity and density. Estimation based on published results of other similar iron-based materials suggested that the kρC value of the hardfacing materials are above 0.03, which is much higher than titanium alloy (about 0.01) [7,9]. This is probably the basic reason for the higher temperature in machining the titanium alloy. The different temperatures for hardfacings A and B showed that the microstructure had significant effect on the cutting temperature. This is possibly due to the fact that hardfacing A has more ductile matrix, which can absorb more energy through plastic deformation and resulted in more cutting heat. In addition, because the thermal conductivity of carbides is higher than austenite [7], hardfacing B probably has a higher kρC value due to the large amount of primary carbides, therefore resulted in lower cutting temperatures. Subsequently, higher temperatures soften the workpiece material more and reduce tool wear rate; this is in agreement with the tool wear tests [3]. 4.2. FE modelling results

Fig. 8. Effect of feedrate on tool–shim interface temperature.

4.2.1. Temperature distribution in the machining systems In the simulation process, the temperature of the tool insert and its holding system increases with cutting time. For an input average temperature on the contact length, the temperature distribution of the system at any time could be determined. Fig. 9 shows a typical temperature distribution

X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44

Fig. 9. Temperature distribution in the tool insert and the supporting system.

for the system (at time = 30 s) with an applied tool–chip temperature of 1173 K. The temperature contours are represented by different grey scales, which allow the temperature regions to be clearly distinguished. As shown in Fig. 9, extensive heat exchange occurred in the tool insert and the supporting shim; in addition, the temperature of the top clamp in contact with the tool insert has significant changed. The tool holder and the turret clamp were basically still in their original temperature, this shows that the original assumption used in this work, i.e. model this sub-system instead of the whole machine as a heat sink, was valid. Fig. 10 shows a close up view of the temperature distribution of the tool insert. The high temperature was con-

Fig. 10. Temperature distribution in the PCBN insert and the tool–shim interface (applied chip–tool temperature = 1073 K).


Fig. 11. Predicted tool–chip temperatures at different cutting conditions.

centrated in a small portion of the tip and the temperature dropped quickly from the tool surface into the bulk of the tool insert. The temperature distribution was further analysed by plotting the data against the distance away from the cutting end (Fig. 10). The temperature drops along the tool–shim interface away from the end immediately beneath the cutting zone. At the thermocouple location (0.5 mm from the edge) used in this work, the temperature gradient is not very significant; this could reduce the error due to uncertainty of the thermocouple positions. 4.2.2. Predicted chip–tool temperatures As shown in Fig. 10, with a temperature input to the insert tip, there will be unique value at the thermocouple temperature at the tool–shim interface. Using this relation, the average tool–chip temperature could be predicted based on the temperature measured at the tool–shim interface and the chip–tool contact length. Fig. 11 shows the average predicted chip:tool temperature of the two hardfacings and the titanium alloy. The predicted average chip:tool temperature of titanium alloy (650–750 ◦ C) agrees with some reported experimental results, at similar cutting conditions [23,24]. The simulated chip–tool temperature when machining hardfacing A is ranged between 650 and 700 ◦ C, while the temperature for hardfacing B is significantly lower (600–650 ◦ C). An important aspect of machining hard materials is the generation of sufficiently high temperatures to soften the workpiece material whilst the tool material retains its strength [10]. Compared with other engineering materials [9], the cutting temperature generated in cutting hardfacings is in a modest range. The cutting temperature will not cause the deterioration of the mechanical strength of the PCBN tools, which could retain its hardness up to a temperature of 800 ◦ C [27]. In addition, the temperature will not cause significant thermal/chemical wear process [25,28–30]. However, the temperature are also not high enough to soften the carbides within the hardfacing itself [4]. Hence, the effect of hard turning cannot be fully achieved because the hard-


X.J. Ren et al. / Journal of Materials Processing Technology 147 (2004) 38–44

ness of the carbides remains sufficiently high to adversely affect the performance of cutting tools. This is probably one of the main factors limiting the machinability of the hardfacing materials.



5. Conclusions [13]

The cutting temperatures of titanium alloy and chromium hardfacings have been evaluated using thermocouples at tool–shim interface and the average tool–chip temperatures were predicted using FE method. The temperatures for hardfacing of a fine structure were predicted to be 650–700 ◦ C, while the temperatures for hardfacing with coarser carbides were around 600 ◦ C. The cutting temperatures were found to increase with cutting speeds as well as with feedrate, the increase rate was affected by the workpiece microstructure. The temperature for hardfacing of finer structure increased more rapidly than that of coarse carbides with higher speed and feedrate.






Acknowledgements [19]

The authors would like to express their gratitude to Mr. R.W. Swain, Mr. G. Robinson and Mr. D. Wright of the School of Engineering at the University of Hull for their technical support in the cutting tests and temperature measurements. References



[22] [23]

[1] Wear Resistant Surface in Engineering: A Guide to their Production, Properties and Selection, International Research and Development Co. Ltd., Department of Trade and Industry, London, 1985. [2] I.M. Hutchings, A.N.J. Stevenson, Wear of hardfacing white cast irons by solid particle erosion, Wear 186–187 (1995) 150–158. [3] X.J. Ren, Ph.D. Thesis, University of Hull, UK, 2000. [4] E.J. Brookes, R.D. James, X.J. Ren, Machining of welded hardfacing materials, in: Proceedings of the Fourth International Conference on Behavior of Materials in Machining: Opportunities and Prospects for Improved Operations, Stratford-Upon-Avon, UK, 1998, pp. 189–198. [5] X.J. Ren, R.D. James, E.J. Brookes, Wear characteristics of CBN tools in machining hardfacing materials, J. Superhard Mater. 1 (1999) 36–45. [6] X.J. Ren, R.D. James, E.J. Brookes, L. Wang, Machining of high chromium hardfacing materials, J. Mater. Process. Technol. 115 (2001) 423–429. [7] M.C. Shaw, Metal Cutting Principles, Clarendon Press, Oxford, 1984. [8] E.M. Trent, Metal Cutting, 3rd ed., Butterworths, London, 1991. [9] D.A. Stephenson, J.S. Agapiou, Metal Cutting Theory and Practice, Marcel Dekker, New York, 1996. [10] P.J. Heath, Structure, properties and applications of polycrystalline cubic boron nitride, in: Proceedings of the 14th North American


[25] [26]


[28] [29]


Manufacturing Research Conference, Minneapolis, MN, USA, 1986, pp. 66–80. A.M. Abrao, D.K. Aspinwall, M.L.H. Wise, A review of polycrystalline cubic boron nitride cutting tool developments and application, in: Proceedings of the 13th International MATADOR Conference, Manchester, UK, 1993, pp. 169–180. W. Konig, A. Berktold, J. Liermann, N. Winands, Top quality components not only by grinding, Ind. Diamond Rev. 3 (1994) 127–132. T.I. El-Wardany, E. Mohammed, M.A. Elbestawi, Cutting temperature of ceramic tools in high speed machining of difficult-to-cut materials, Int. J. Mach. Tools Manuf. 36 (5) (1996) 611–634. A.M. Abrao, D.K. Aspinwall, Temperature evaluation of cutting tools during machining of hardened bearing steel using polycrystalline cubic boron nitride and ceramic cutting tools, Mater. Sci. Technol. 13 (1997) 445–451. M.B.D. Silva, J. Wallbank, Cutting temperature: prediction and measurement methods—a review, J. Mater. Process. Technol. 88 (1999) 195–202. M.P. Lipman, B.E. Nevis, G.E. Kane, A remote sensor method for determining average tool–chip interface temperatures in metal cutting, ASME J. Eng. Ind. 89 (1967) 333–338. J.G. Chow, P.K. Wright, On-line estimation of tool–chip interface temperature for a turning operation, ASME J. Eng. Ind. 110 (1988) 56–64. A.R. Shaik, S. Raman, F. Civan, P.H. Cohen, A new forward temperature estimator for remote thermocouple sensing in machining, Int. J. Mech. Sci. 37 (5) (1995) 511–526. W.C. Chen, C.C. Tsao, P.W. Liang, Determination of temperature distributions on the rake face of cutting tools using a remote method, Int. Comm. Heat Mass Transfer 24 (2) (1997) 161–170. J. Mackerle, Finite-element analysis and simulation of machining—a bibliography (1976–1996), J. Mater. Process. Technol. 86 (1999) 17–44. G.E. D’Errico, An adaptive system for turning process control based on tool temperature feedback, J. Mater. Process. Technol. 78 (1998) 43–47. S. Atamert, H.K.D.H. Bhadeshia, Microstructure and stability of Fe–Cr–C hardfacing alloys, Mater. Sci. Eng. A130 (1990) 101–111. F. Nabhani, The performance of ultra-hard cutting tool materials in machining aerospace alloy TA48, Ph.D. Thesis, University of Hull, UK, 1991. Z.A. Zoya, R. Krishnamurthy, The performance of PCBN tools in machining of titanium alloys, J. Mater. Process. Technol. 100 (2000) 80–86. M.A. Younis, Mechanical and thermal stress in clamped, brazed, and bonded carbide tools, ASME J. Eng. Ind. 114 (1992) 377–385. L.S. Casto, L.E. Valvo, M. Piacentini, V.F. Ruisi, E. Lucchini, S. Maschio, Cutting temperatures evaluation in ceramic tools: experimental tests, numerical analysis and SEM observations, CIRP Ann. 43 (1994) 73–76. T.K. Harris, E.J. Brookes, C.J. Taylor, The flow stress of PCBN cutting tool materials at high temperatures, Int. J. Refract. Met. Hard Mater. 19 (2001) 267–273. C.A. Brookes, W.A. Lambert, Ultra Hard Material Application Technology, De Beers, Hornbeam Press Ltd., 1982. C.A. Brookes, R.D. James, F. Nabhani, A.R. Parry, Structure, composition, and integrity of workpiece: tool interfaces in metal cutting and grinding, in: Proceedings of the IMechE Seminar on Tribology in Metal Cutting and Grinding, 1992, pp. 41–48. W. Konig, A. Neises, Wear mechanisms of ultrahard, non-metallic cutting materials, Wear 162–164 (1993) 12–21.