英语论文原文.pdf

Ultrasonically assisted turning of aviation materials: simulations and experimental study V.I. Babitsky, A.V. Mitrofanov, V.V. Silberschmidt * Wolfson School of Mechanical and Manufacturing Engineering, Loughborough University, Leicestershire LE11 3TU, UK Abstract Ultrasonically assisted turning of modern aviation materials is conducted with ultrasonic vibration (frequency f ? 20 kHz, amplitude a ? 15 lm) superimposed on the cutting tool movement. An autoresonant control system is used to maintain the stable nonlinear resonant mode of vibration throughout the cutting process. Experimental comparison of roughness and roundness for workpieces machined conventionally and with the superimposed ultrasonic vibration, results of high-speed fi lming of the turning process and nanoindentation analyses of the microstructure of the machined material are presented. The suggested fi nite-element model provides numerical comparison between conventional and ultrasonic turning of Inconel 718 in terms of stress/strain state, cutting forces and contact conditions at the workpiece/tool interface. ? 2004 Elsevier B.V. All rights reserved. Keywords: Ultrasonic machining; Turning; Finite element modelling; Microstructure 1. Introduction Turning is a machining process, where a thin surface layer of the treated material is removed from a work- piece by a sharp wedge-shaped cutting tool forming a cylindrical surface. This technology has been used for centuries mainly for cutting various types of metallic materials. However, in the recent years, a range of new alloys and composite materials has been developed for various engineering applications. Many of these new materials become much more diffi cult to cut with the existent, conventional turning (CT) technology. Conventional machining of modern nickel- and tita- nium-based superalloys, used in aerospace applications, causes high tool temperatures and subsequent fast wear of cutting edges even at relatively low cutting speeds. A growing demand for machining these intractable materials requires new advanced turning technologies. Such a technology was introduced in 1960s: high- frequency ultrasonic vibration, superimposed on the conventional movement of the cutting tool (Fig. 1), has proved to be eff ective in machining intractable metal alloys as well as brittle materials, such as ceramics and glass. This technology, called ultrasonically assisted turning (UAT), demonstrates a range of benefi ts in machining hard metal alloys: a decrease in cutting forces of up to several times 14, improvement in surface fi nish by up to 50% compared to CT 5 and noise reduction 6. As for machining of brittle materials, ceramics and glass presently require prolonged and expensive post-processing to obtain the surface quality required for optical components; UAT allows obtaining mirror surface fi nish in machining these materials as well as considerable reduction in tool wear and average cutting forces 7,8. Nevertheless, up to the present day UAT has not been widely introduced into industrial environment. The main reason for it is sensitivity of the UAT process to the load applied to the cutting tip, resulting in the loss of cutting effi ciency when the load changes or a diff erent tip is used. However, this limitation has recently been eliminated with the invention of the autoresonant con- trol system. This system stabilises the turning process with ultrasonic vibration and makes this process highly controllable. The detailed description of this novel control system is given in 9,10. The experimental part of this paper studies UAT with autoresonant control in comparison to conventional turning. Another important issue concerning UAT is the mechanics of this process. There are only a few sources *Corresponding author. Tel.: +44-1509-227504; fax: +44-1509- 227502. E-mail address: v.silberschmidtlboro.ac.uk (V.V. Silberschmidt). 0041-624X/$ - see front matter ? 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ultras.2004.02.001 Ultrasonics 42 (2004) 8186 attempting to describe the processes in the workpiece/ cutting tool interaction zone and their infl uence on the structure of the machined material 3,6,11. These works study mostly the dynamics of the ultrasonic machine unit and not the response of the treated material to this technology, while a clear understanding of mechanical processes in the material during UAT would certainly allow a further development of the UAT technology. The main aim of this paper is to study experimentally and numerically the material mechanics of the UAT process. 2. Experimental studies The experimental setup used to study UAT is shown in Fig. 1. The workpiece is clamped in the chuck of the universal lathe and rotates with a constant speed. High frequency electric impulses, fed to the input of the ultrasonic transducer, excite vibration in piezoceramic rings due to the piezoelectric eff ect. The vibration amplitude is intensifi ed in the concentrator and trans- mitted to the tool holder at the thin end of the con- centrator. Resultant vibration of the cutting tip fi xed in the tool holder reaches 15 lm (i.e. 30 lm peak-to-peak) at a frequency of about 20 kHz. The vibration can be applied either in the direction tangential to the surface of the workpiece (parallel to the X-axis, Fig. 1b), or in the feed direction, i.e. along the axis of the workpiece (Z-axis, Fig. 1b). A self-sustained resonant mode of vibration of this cutting system is implemented via the autoresonant control system, which is described in detail in 9,10. A range of turning tests has been conducted to compare the usage of UAT and CT for machining avi- ation materials. The detailed description of these tests can be found in 5. Among the materials used for the tests is Inconel 718a high-grade heat-resistant Ni-based superalloy widely used in the aerospace industry. This material is very abrasive and causes the tool blunting and high cutting temperatures when machined conven- tionally. The surface quality obtained by turning is one of the crucial factors in metal cutting and is extremely sensitive to any changes in the machining process. The surface fi nish of specimens is compared in terms of average roughnessmeasurementRaandmeasurementof roundness (the peak-to-valley measure), using the Tay- lor HobsonTalysurf 4 surface measurement instrument. The following cutting parameters are used to machine tested specimens: depth of cut d 0:8 mm, feed rate s 0:05 mm/rev, and cutting speed v 17 m/min. The same parameters are used for both UAT and CT, with superimposed ultrasonic vibration in the feed direction applied for UAT. Fig. 2a shows representative axial profi les of the machined surface of the Inconel 718. It is obvious that magnitudes of Raare reduced by nearly 50% for speci- mens machined with UAT. Furthermore, the regularity of the surface profi le is greatly improved, as the surface Fig. 2. Surface quality of Inconel 718 specimens machined with UAT and CT: axial surface profi les (a), roundness profi les (b). Cutting parameters: d 0:8 mm, s 0:05 mm/rev, v 17 m/min. Fig. 1. Experimental setup for ultrasonically assisted turning (a), and a scheme of relative motion of the workpiece and cutting tool in orthogonal UAT with tangential vibration (b). 82V.I. Babitsky et al. / Ultrasonics 42 (2004) 8186 becomes smoother in the axial direction. A considerable improvement is also obtained for roundness of ma- chined workpieces (Fig. 2b): a peak-to-valley value of roundness measures 4.20 lm for CT, whereas it attains only 1.89 lm for UAT. Hence, the roundness is im- proved by 40% when ultrasonic vibration is superim- posed upon the movement of the cutting tool. It is worth noticing that similar results have been obtained by other researchers 7,12 utilising vibration in the tangential direction. Apparently, the reason for these improvements is the change of the nature of the cutting process, which is transformed into the one with multiple-impact high- frequency interaction between the cutting tool and chip due to applied ultrasonic vibration. This leads to changes in material deformation processes and fric- tion forces, and increase in the dynamic stiff ness of the lathe-tool-workpiece system 6,11 due to the vibration frequency levels considerably exceeding its natural fre- quency. In addition to measurements of the surface quality, the microstructure of the machined surface has been investigated. Inconel 718 workpieces are machined un- der the same cutting conditions (v 3:6 m/min, d 0:1 mm, s 0:03 mm/rev) with application of ultrasonic vibration in tangential direction and without it. Then, nanoindentation analyses of the surface layers are per- formed with the NanoTest Platform made by Micro Materials Ltd. According to the results of these tests, the width of the hardened surface layer, which results from the extensive deformation and high temperature pro- cesses during the turning procedures, for the ultrasoni- cally machined specimen is half the size that of the conventionally machined one (40 and 80 lm, respec- tively). Furthermore, the average hardness of this layer for UAT (about 15 GPa) is a half of that for CT and considerably closer to the hardness of the untreated material (about 7 GPa). The hardness of the material nonlinearly increases with a rise in the level of the residual plastic strains. Hence, nanoindentation tests indicate lower residual strains in the surface layer for workpieces machined with UAT and a conclusion can be drawn that the UAT procedure is considerably more delicate to the workpiece material. 3. Numerical analysis of UAT Finite element (FE) simulations are a major tool for modelling of machining processes. It has been used for modelling of turning for some 30 years. The overview of the state of the art in metal cutting simulations can be found in 13,14. However, up to the authors knowledge, no models for UAT have been developed until now. The two-dimensional FE model of both CT and UAT de- scribed in this paper is based on the commercial FE code MSC.Marc 15. An orthogonal turning process, i.e. the cutting process where the tool edge is normal to both cutting and feed directions, with tangential vibration is considered. Fig. 1b shows a scheme of the modelled relative motion of the workpiece and cutting tool; the rotation axis of the cylindrical workpiece is orthogonal to the plane of the fi gure. The workpiece moves with a constant velocity, whereas the tool vibrates harmonically around its equilibrium position with frequency f 20 kHz and amplitude a 15 lm, corresponding to the values used in experimental studies. Other parameters of simulations are: uncut chip thickness t1 0:1 mm (which corresponds to the depth of cut), rake angle of the tool c 10?, cutting speed V 9 m/min. Such parameters of vibration and of the cutting process provide separation of the cutter from the chip within each cycle of ultrasonic vibration. The material constants for aged Inconel 718 are taken from 16. Kinematical boundary conditions for the workpiece are applied to its left, right and bottom sides (Fig. 1b), whereas its top surface is free: VxjAH V ; VxjFG V ; VxjHG V ; VyjHG 0: Thermal boundary conditions include convective heat transfer from the workpiece, chip and tool free surfaces to the environment: ?koT=on hT ? T1, where k is the conductivity, h is a convective heat transfer coeffi - cient, T1 is the ambient temperature. The thermal fl ux passing from the chip to the cutter along the contact length Lc(Fig. 1b) is described as follows: q HTchip? Ttool , where H is a contact heat transfer coeffi cient, Tchip and Ttoolare chip and tool surface temperatures, respectively. The model takes into consideration the following factors, important for metal turning simulations and aff ecting stress and strain generation: (1) contact inter- action and friction at the tool-chip interface; (2) non- linear material behaviour, including strain-rate eff ects, namely the dependence of the materials yield stress on strain rates; (3) thermomechanical coupling, i.e. inter- connection between mechanical and thermal parts of the problem. As follows from FE simulations, the UAT process during one cycle of vibration could be divided into four main stages. During the fi rst stage (Fig. 3a), the cutter approaches the chip; in the second stage, the cutting tool contacts the chip and starts penetrating into the work- piece causing the chip separation. The attainment of the maximum penetration depth is characterized by the highest level of generated stresses in the process zone and marks the end of the second stage (Fig. 3b). The following stage is unloading: the velocity direction of the tool changes and it moves backwards, but remains in contact with the chip even after the moment when the speed of the tool exceeds the cutting speed (due to the elastic spring-back of the chip). During this phase, V.I. Babitsky et al. / Ultrasonics 42 (2004) 818683 the elastic strains in the process zone decrease. The last stage, starting with the full separation of the cutting edge from the chip, is the withdrawal of the cutter from the chip (Fig. 3c). The intermittent character of the chip-cutting tool contact determines the main diff erences in the stress distribution for CT and UAT. The stress state during CT is nearly quasistatic, as shown in numerical simulations (Fig. 3d), with the highest equivalent stresses concen- trated in primary and secondary shear zones, i.e. zones around line BE (Fig. 1b) and next to the rake face EK. Conversely, the stress state in UAT is inherently tran- sient: stresses reach maximum levels, similar to those of CT, during the penetration part of the cycle of ultrasonic vibration (Fig. 3b), whereas the stress magnitude is sig- nifi cantly lower during the other stages of the cycle (Fig. 3a and c), when the cutter moves away from the chip or not in contact with it. Hence, average stresses generated in the material and, consequently, the integral level of interaction forces between the cutting tool and work- piece are considerably smaller for UAT. This explains a reduction in average cutting forces (by several times) reported in many experimental studies 1,3,4. 4. Study of chip formation A chip formation process is one of the most impor- tant characteristics in metal cutting. Hence, it is of particular interest to study the diff erences in the chip formation arising from superimposed ultrasonic vibra- tion. In the experimental part of this study, a high-speed digital camera (Kodak Ektapro HS Motion Analyzer 4540) is used for real-time observations of the chiptool interaction during both UAT and CT of Inconel 718. Filming speed is in the range from 9000 up to 27000 frames per second with the area of the image comprising about 4 mm2. Fig. 4a demonstrates a frame of the UAT fi lming showing an interaction between the cutting tip and workpiece. The diff erences between CT and UAT in chip separation manifest in such specifi c features of the process as the size and shape of the process zone, and the type of the produced chip. The area of the visible process zone for UAT and its size in the radial (vertical, Fig. 4a) direction are considerably smaller than those for CT. Deformation processes for UAT are localized in the direct vicinity of the cutting edge along the surface of the workpiece and are not observed underneath the cutter, in contrast to the CT process. This correlates well with the results of nanoindentation tests indicating smaller hardened layer for the UAT machined surface. Finally, high-speed observations showed that super- imposed ultrasonic vibration makes the process of chip formation more regular, resulting in an incremental, continuous chip formation process. In contrast, the observed CT process produces essentially segmented chip, due to forced irregular vibration of the cutting tool and tearing-like chip separation. A scanning electron microscope (SEM) study of microstructure of chips produced with UAT and CT confi rms this observation, revealing a continuous chip with small serrations for UAT and strongly segmented chip with vivid shear bands for CT. Fig. 3. Distribution of equivalent stresses during UAT at diff erent moments of a single cycle of vibration: cutter approaching the chip (a), cutter in full contact with the chip (b), and cutter moving away from the chip (c) and CT (d). Fig. 4. Chip formation: frames of high-speed fi lming (a) and numerical (FE) simulations (b). 84V.I. Babitsky et al. / Ultrasonics 42 (2004) 8186 A shape of the numerically modelled chip (Fig. 4b) is in a good agreement with that experimentally observed. This plot of the FE-simulation results provides an in- sight into the distribution of equivalent plastic strains in the cutting region. Plastic strains attain maximal values near the cutting edge and along the newly formed sur- face of the specimen. 5. Study of tool and chip temperatures During the turning process, the work of plastic deformation and friction between the cutting tool and workpiece can result in a signifi cant temperature in- crease in both the workpiece and cutter. The tempera- ture increase, in its turn, changes material properties, such as the yield stress, temperature conductivity and specifi c heat, thus infl uencing deformation processes in the workpiece. Evolution of temperatures during turn- ing is, therefore, an important characteristic feature of the comparison between CT and UAT. Infrared fi lming has been employed to experimentally measure the temperature distribution in the area of interaction between the workpiece and cutting tip. Agema 880 Thermovision System was used to perform these measurements. Single frames with infrared pic- tures were taken during CT and UAT turning processes. A quasistatic temperature distribution was measured, since a possible change of temperature during a single cycle of