Tool Path Generation Functionality and Ultrasonic Assisted Machining of Ceramic Components using Multi-axis Machine Tools

Size: px

Start display at page:

Download "Tool Path Generation Functionality and Ultrasonic Assisted Machining of Ceramic Components using Multi-axis Machine Tools"

Phebe Lester
5 years ago
Views:

Tool Path Generation Functionality and Ultrasonic Assisted Machining of Ceramic Components using Multi-axis Machine Tools B. Lauwers, D. Plakhotnik, M. Vanparys, W. Liu Dept.

industries (automotive, medical, general mechanical engineering,..).

1 Tool Path Generation Functionality and Ultrasonic Assisted Machining of Ceramic Components using Multi-axis Machine Tools B. Lauwers, D. Plakhotnik, M. Vanparys, W. Liu Dept. of Mechanical Engineering, Katholieke Universiteit Leuven, Leuven, Belgium Abstract Machining of complex shaped components in various materials, including ceramics, is an important challenge for many industries (automotive, medical, general mechanical engineering,..). Multi-axis machine tools (fiveaxis milling/grinding, multi-axis turn/mill centers) offer the necessary flexibility to perform these machining operations. This paper describes on-going and planned research activities at the K.U.Leuven on machining processes using multi-axis machine tools. First, the paper describes some recent developments in tool path generation functionality for multi-axis machining operations. Second, ultrasonic assisted machining of various ceramic materials (Al 2 O 3, ZrO 2, SiC,..) on a multi-axis machining platform is discussed. Finally, this paper also describes planned research activities to be performed on a recently installed multi-axis machine tool, the Mori Seiki NL2000Y/500. Keywords: Multi-axis machine tool, Tool Path Generation, Ultrasonic Assisted Machining 1 INTRODUCTION Multi-axis machine tools are increasingly used in modern manufacturing industries. The flexibility offered by more than 3 axes allows efficient machining of complex shaped parts within one set-up. In addition, more efficient machining (higher material removal, better surface quality) is obtainable by multi-axis milling. The efficient NC-programming of these machine tools is known to be very hard and time consuming. The way a part is machined is not unique and the efficiency of several strategies is often difficult to estimate without computer support. New methodologies and concepts for tool path planning and generation for multi-axis milling of complex shaped surfaces have been worked out by the K.U.Leuven research group [1,2]. These developments, dealing with part geometry analysis, part segmentation, part set-up, machining strategy evaluation and multi-axis tool path generation, only take the part geometry (and properties) into account. The multi-axis machine tool itself has however an important influence on the machining performance. Due to differences in speed and the acceleration profiles of linear and rotary axis, the way a part (e.g. complex shaped surface) is machined can strongly influence the machining time. Also more simple parts with a number of features on different faces, can often be machined in different ways by multi-axis machine tools (e.g. multi-axis turn/mill centre). The sequence in which the features are machined can result in different real machining times. For example, the tool movement from one feature to another feature position can result in differences of the total machining time. Industries are increasingly demanding for the machining of complex shaped ceramic components (e.g. ZrO 2, Al 2 O 3, B 4 C, SiC,..). These ceramic components find their applications where high hardness and good resistance against heat, chemicals, erosion and wear are required. Today the most common manufacturing processes used to machine these ceramic materials (in their final state) are grinding and Electrical Discharge Machining (EDM). Compared to grinding, NC-controlled Wire EDM, die sinking EDM as well as milling EDM allow to machine complex shapes within one set-up [3]. EDM is however only feasible under the condition that the ceramic has a maximum electrical resistivity of about 1 Ωm. To obtain a certain electrical conductivity, second phases are added such as TiN, WC or TiCN, which result in so called ceramic composites. Ultrasonic assisted machining is another process which is gaining industrial interest for the machining of ceramic components. Since 2006, the K.U.Leuven research group (Lauwers) also focuses on multi-axis ultrasonic assisted milling/grinding of ceramic components. The ultrasonic vibration is generated within dedicated HSK-based tool holders and implemented on a multi-axis machining platform. The multi-axis capabilities allow machining of complex shaped components in various ceramic materials. There is however a strong need from industry to machine ceramic parts with an overall axi-symmetric shape. Figure 1 shows parts made by EDM, but which could be machined on the Mori Seiki NL2000Y/500, a machining center which has recently been installed through MTTRF at K.U.Leuven. Powder spray nozzle Mould insert Figure 1: Examples of industrial ceramic components (B 4 C) with an overall axi-symmetric shape. This paper describes on-going research activities at the K.U.Leuven on multi-axis tool path generation functionality and machining of ceramic components. To perform this research, multi-axis machine tools (DMG Sauer Ultrasonic 70/5, MAHO 600C, Mori Seiki NL2000Y/500) are used. This paper also briefly describes the planned research activities,

2 especially these where the recently installed Mori Seiki machining center will be used. 2 DEVELOPMENT OF ADVANCED TOOL PATH GENERATION FUNCTIONALITY 2.1 Concept for efficient NC-programming As described in last years MTTRF paper [2], a new concept (Figure 2) for tool path planning and tool path generation has been worked out by Lauwers within the on-going EU research project NEXT (Next Generation Production Systems, ). The scheme in Figure 2 is clearly divided into 2 parts (dashed horizontal line). Within the off-line part (upper), an initial (draft) tool path is generated. Off-line means that the generation of the tool path is done prior to machining (but the software itself could also run on the controller). The initial tool path is tagged with extra information which allows adaptation of the tool path to a certain extend during machining. Within this research, an XML-based tool path description is proposed, because this allows to add tags and attributes to the tool path in a flexible way. planning and evaluation will not be purely based on the part geometry, but also on machine information (e.g. axes dynamics), stored within the machine model. During machining, the controller is able to adapt the tool path (= tuning of the tool path), which results in a more optimized tool path trajectory. Adaptations can mean geometrical corrections of the tool path, interpolation corrections, speed and feed corrections, change of PLCparameters, etc.. Within this research, tool path planning and generation is based on real machine tool movements, including the time required to perform these moves. Therefore, information about the machine tool (work space, speed and acceleration profiles,..) is required. This information is stored in the machine model. 2.2 Machine model (machine time estimation) The machine model can have various levels of detail. Currently, the implemented machine tool model is a software module (algorithm) estimating machine times based on given speed and acceleration profiles for the different axes. Within the current software, an unlimited jerk is allowed (acceleration is zero or maximum). In addition, corner smoothing is implemented, so the machine can deviate from the corner within in a pre-specified tolerance zone. The time estimation module takes a sequence of tool movements as input. Based on the pre-specified corner tolerance, the maximum allowable feeds in the corners (movement in X,Y,Z) are calculated (Figure 3, green arrows). The red lines represent the maximum feed (based on maximum axes feeds) for each tool segment. Based on these maximum allowable speeds, acceleration and deacceleration profiles are defined. Last, the acceleration and speed profiles are fitted to the speed profiles of the rotary axes and a total machining time is calculated. Figure 3: Calculation of speed and acceleration profliles for the linear axes. Figure 2: Proposed model for advanced tool path planning and generation. The off-line part deals with tool path planning, operations evaluation and advanced tool path generation. Compared to previous developments described in section 1 [1], tool path 2.3 Tool path planning and strategy evaluation Strategy evaluation (which tool path pattern is the best), requires a fast evaluation of the performance of a given operation. An operation can be evaluated by different means. In previous research [1,2], the performance of a given strategy was evaluated based on the machining strip width and total tool path length. Only the geometrical properties of the part were taken into account. A larger

machining strip width results in less tracks and hence a shorter tool path length.

To perform the evaluation, the part surface, represented as a STL-file (triangulated surface) is scanned along a specified tool path pattern. Of course, different tool path patterns can be evaluated.

The minimal machining strip width in the band is taken as worst case and defines the position of the next scanning band.

Figure 4 shows a strategy evaluation for a complex shaped surface. The trajectory length and machining time is given as a function of the scanning angle α.

Time estimations have been calculated for a multi-axis milling machine (rotary axis in the table) with the following conditions: maximum feed: 5000mm/min; acceleration for the linear axes: 0.

3 machining strip width results in less tracks and hence a shorter tool path length. Within this research, the efficiency of a tool path strategy is defined by the estimated machining time (taking into account the kinematics and dynamics behaviour). To perform the evaluation, the part surface, represented as a STL-file (triangulated surface) is scanned along a specified tool path pattern. Of course, different tool path patterns can be evaluated. Along the tool path pattern, the machining strip width is calculated for each consecutive STL vertex lying in the scanning band. The minimal machining strip width in the band is taken as worst case and defines the position of the next scanning band. At each vertex, the tool posture (X,Y,Z,I,J,K) is converted to machine axes values. A sequence of machine motions (typical for one scan) is then sent to the machine time estimation module. Figure 4 shows a strategy evaluation for a complex shaped surface. The trajectory length and machining time is given as a function of the scanning angle α. An end-mill tool (Ø: 10 mm) is used and the tool inclination is calculated for a maximal strip width. Time estimations have been calculated for a multi-axis milling machine (rotary axis in the table) with the following conditions: maximum feed: 5000mm/min; acceleration for the linear axes: 0.1g; maximum angular feed for the A-axis: 39 rev/min; maximum angular feed for the B-axis: 17.4 rev/min; acceleration for the rotary axes: 500rev/min 2. From Figure 4, it can be concluded that deviations of trajectory length are less pronounced than of machining time. This means that including the machine tool behaviour is an important issue in tool path planning and generation. A next example is the machining of a curved region of the complex shaped part presented in Figure 5. A 3-axis milling strategy is compared to a 5-axis milling strategy. Interesting is the difference between the machining time estimation and the time normally calculated by the postprocessor (trajectory length/feed). Tool path generation and simulation for a complex shaped part Strategy I 3-axis machining Tool: ball nose, Ø 4 mm Tolerance: 0.001mm Feed: 1500 mm/min (2000 mm/min for engage/retract) Strategy II 5-axis machining Tool: flat end mill, Ø 4 mm Tolerance: 0.001mm Feed: 600 mm/min (1500 mm/min for engage/retract) Results Trajectory length: 5812 mm Machining time: 6.69 min Postprocessor time: 3.84 min (trajectory length/feed): Results Trajectory length: 1302mm Machining time: 1.96 min Postprocessor time: 1.45 min (trajectory length/feed): 300% 250% 200% 150% 100% 50% 0% Machining time Trajectory length Scanning angle (α) Figure 4: Estimation of machining time and trajectory length for ZIG-milling strategy as a function of the scanning angle. Figure 5: Comparison between two multi-axis milling strategies based on trajectory length and estimated machining time. 2.4 Planned research Planned research includes further development of tool path planning and generation functionality for multi-axis machining operations. In short-term, the machine tool model will be extended with more controller related features. As an example, the Siemens VNCK kernel will be integrated within the system and compared to the currently developed software. Further testing of this concept by integrating other controller and/or machine tool emulators (e.g. for the available Mori Seiki machining center) for tool path generation is certainly an interesting topic to be investigated. In the longer run, new concepts for tool path generation, including machine information, will be developed.

3 ULTRASONIC ASSISTED MACHINING OF CERAMIC COMPONENTS As described above, ultrasonic assisted machining gains more and more interest within industry for the fabrication of ceramic components.

The mechanical ultrasonic vibration is generated by a piezo-actuator integrated within the actor system of the tool holder (Figure 6a).

In order to evaluate the effect of the ultrasonic vibration, samples have been machined with and without vibration.

There is no large difference in surface roughness between samples machined with or without ultrasonic vibration. For the ZrO 2 material, the surface roughness is 0.

Ultrasonic on Ultrasonic off (a) (b) Figure 6: (a) Tool holder system (piezo actuator integrated within HSK tool holder), (b) Example of diamond embedded tools. 3.

of proper machining strategies. Most investigations [4,5] state that material removal is based on a combination of plastic deformation and brittle removal.

4 3 ULTRASONIC ASSISTED MACHINING OF CERAMIC COMPONENTS As described above, ultrasonic assisted machining gains more and more interest within industry for the fabrication of ceramic components. Recent and on-going research activities are focused on (multi-axis) ultrasonic assisted grinding (UAG). The mechanical ultrasonic vibration is generated by a piezo-actuator integrated within the actor system of the tool holder (Figure 6a). Figure 7 shows surface topographies of two investigated materials, Al 2 O 3 and ZrO 2. In order to evaluate the effect of the ultrasonic vibration, samples have been machined with and without vibration. Similar to literature, plastic deformation as well as brittle removal can be observed. For the samples made without vibration, more grooves, typical for a grinding operation can be observed. There is no large difference in surface roughness between samples machined with or without ultrasonic vibration. For the ZrO 2 material, the surface roughness is 0.34μm compared to 0,39μm when no ultrasonic vibration is applied. In case of Al 2 O 3, the surface roughness is 0.37 μm (ultrasonic ON ) compared to 0.43 μm in case no ultrasonic vibration is applied. Ultrasonic on Ultrasonic off (a) (b) Figure 6: (a) Tool holder system (piezo actuator integrated within HSK tool holder), (b) Example of diamond embedded tools. 3.1 Identification of material removal mechanisms in UAG of ceramic materials Knowledge of the occurring material removal mechanisms and the resulting surface texture is important for the development of proper machining strategies. Most investigations [4,5] state that material removal is based on a combination of plastic deformation and brittle removal. Literature about the influence of machining parameters on the type of material removal mechanism is however limited. Within this research, the occurring material removal mechanism of various ceramic materials (Al 2 O 3, ZrO 2, SiC), representing a broad range of mechanical properties, have been investigated. Table 1 gives the material properties for the different materials. In order to investigate the material removal mechanisms, slots have been machined as such that the tool moves in a horizontal plane (feed direction perpendicular to the ultrasonic vibration direction). Table 1: Mechanical properties of investigated materials. Density (g/cm 3 ) E-modulus (GPa) Fracture Toughness (MPa.m 1/2 ) Hardness (kg/mm 2 ) Al 2 O Bending Strenth (MPa) ZrO ** SiC * (*) 4 points bending strength (**) 3 points bending strength Al 2 O 3 Depth slot: 0,02mm, tool: 2mm, 4500 rpm, feed: 50mm/min, frequency: 17400Hz, cut-dept: 5μm ZrO 2 Depth slot: 0,02mm, tool: 5mm, 4500 rpm, feed: 500mm/min, frequency: 18404Hz, cut-dept: 5μm Figure 7: Surface topographies of machined samples, with and without ultrasonic vibration (Al 2 O 3, ZrO 2 ). In case of planar grinding of SiC (tool: 60 mm), a strong increase in surface quality could be obtained when applying ultrasonic vibration. Figure 8 shows two sample parts, one machined with ultrasonic vibration (left) and one without ultrasonic vibration (right). In both cases, the other parameters were set to: feed rate: 200 mm/min, spindle rotation: 3000 rpm, cutting depth: mm, tool diameter: 60 mm, tool wall thickness: 3 mm. The measured surface roughness for the part with ultrasonic vibration ON was 0.03μm Ra, while a value of 0.12μm Ra has been measured without ultrasonic vibration. The surface roughness has been measured using a Taylor Hobson System (pin radius: 2μm, cut-off length: 0.25mm).

Figure 8: Machining of SiC (left picture: ultrasonic ON ; right picture: ultrasonic OFF ).

Force measurements however only gave a small difference in the Z-force (along the tool axis) of 3 4 N between ultrasonic ON and OFF (average force: 80N).

having done the same type of experiments. The reason is simply that the tool moves perpendicular to the vibration direction.

Slot1 Slot 2 Slot 3 A more detailed investigation of the material removal mechanism of Al 2 O 3 showed that samples machined with similar parameters do not always

To investigate this, five slots (length 10 mm, depth 0.3 mm for the first slot, 0.

frequency: 16589 Hz, cutting depth: 0,01 mm, spindle rotation speed: 4500 rpm).

From the pictures, it is clear that some samples show more brittle removal (e.g. slot 1) than other slots (e.g. slot 3).

But it can be logically said that sharp grains which largely stick out of the tool surface, generate grooves in the workpiece surface and hence plastic deformation.

The tool wear mechanisms are a combination of the mechanism described by [4,5,6].

Figure 10 presents measured surface roughness values as function of the machined volume. Each experiment was started with a re-dressed tool.

It is clear from this figure that there is quite some variation in the surface roughness, which is most probably influenced by the status of the tool (grain

5 Figure 8: Machining of SiC (left picture: ultrasonic ON ; right picture: ultrasonic OFF ). According to literature, the advantage of ultrasonic assisted machining is the strong reduction in cutting force. Force measurements however only gave a small difference in the Z-force (along the tool axis) of 3 4 N between ultrasonic ON and OFF (average force: 80N). The same results were obtained by other research institutes (in the frame of the IWT/ EU CORNET project Ultrasonic assisted grinding of brittle hard materials UAG) having done the same type of experiments. The reason is simply that the tool moves perpendicular to the vibration direction. Larger force reductions can be obtained in case of drilling operations where the feed direction is similar to the vibration direction. Slot1 Slot 2 Slot 3 A more detailed investigation of the material removal mechanism of Al 2 O 3 showed that samples machined with similar parameters do not always have a similar surface texture. The reason for the difference in surface texture could be related to the evolution of the tool wear. To investigate this, five slots (length 10 mm, depth 0.3 mm for the first slot, 0.1mm for the other slots) were machined with the same tool (without redressing) and the same parameters (tool diameter: 3 mm, feed rate: 250 mm/min, ultrasonic frequency: Hz, cutting depth: 0,01 mm, spindle rotation speed: 4500 rpm). Figure 9 shows the surface topographies of all 5 samples, together with the status of the tool. From the pictures, it is clear that some samples show more brittle removal (e.g. slot 1) than other slots (e.g. slot 3). It is however difficult to draw clear conclusions. But it can be logically said that sharp grains which largely stick out of the tool surface, generate grooves in the workpiece surface and hence plastic deformation. A tool surface containing flattened grains results in general in brittle removal. The status of the tool surface clearly evolves during machining. The tool wear mechanisms are a combination of the mechanism described by [4,5,6]. It seems that the grains first flatten, after which the flattened grains are splintered or dislodged. Figure 10 presents measured surface roughness values as function of the machined volume. Each experiment was started with a re-dressed tool. Experiment 2 is related to the slots shown in Figure 9 (roughness value for slot 3 is not available). It is clear from this figure that there is quite some variation in the surface roughness, which is most probably influenced by the status of the tool (grain flattening, grain splintering, grain dislodgment). Roughness Ra [µm] 1 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 Experiment 1 Experiment 2 Experiment Machined volume [mm³] Figure 10: Evolution of the surface roughness. Slot 4 Slot 5 Figure 9: Surface topographies of five Al 2 O 3 samples machined with the same tool. 3.2 Development of machining strategies On-going research on machining strategies in ultrasonic assisted grinding include the machining of simple features such as holes, protrusions, as well as more complex shapes (e.g. spherical shape). Initial experiments have been performed to machine more complex shapes. As an example, the mould insert part (Figure 1, right) has been machined by UAG in Al 2 O 3 (Figure 11). The part has been machined in 3 steps: a roughing step and two finishing steps. Roughing is performed by a layer by layer strategy (cavity machining), using a tube shaped tool, 5 mm. The finishing operations are performed by a downwards spiral strategy using a ball nose type tool, 10 mm.

mechanism, surface roughness). The ultrasonic assisted turning will be performed on the at K.U.Leuven available Mori Seiki machining center.

Figure 12 shows the influence of the cutting depth on the edge quality of a protrusion (workpiece material SiC) machined by ultrasonic assisted grinding.

12mm Process parameters Tool: 5mm Feed rate: 50mm/min Spindle speed: 5000 rpm Frequency: 18404 Hz Cutting depth: 5 µm Strategy II Cutting depth: 2 µm 0.

6 mechanism, surface roughness). The ultrasonic assisted turning will be performed on the at K.U.Leuven available Mori Seiki machining center. CAD geometry Final shape Figure 11: UAG of concave spherical shape. An important issue in the machining of prismatic parts is the avoidance of edge chipping. Figure 12 shows the influence of the cutting depth on the edge quality of a protrusion (workpiece material SiC) machined by ultrasonic assisted grinding. A larger cutting depth yields larger forces and hence more edge chipping. Strategy I Cutting depth: 5 µm 0.12mm Process parameters Tool: 5mm Feed rate: 50mm/min Spindle speed: 5000 rpm Frequency: Hz Cutting depth: 5 µm Strategy II Cutting depth: 2 µm 0.04mm Process parameters Tool: 5mm Feed rate: 50mm/min Spindle speed: 5000 rpm Frequency: Hz Cutting depth: 2 µm Figure 12: Strategy development for the machining of protrusions made of SiC. 3.3 Ultrasonic assisted turning planned research Planned research at K.U.Leuven also aims ultrasonic assisted turning of ceramic components. This objective is supported by two main reasons. First, there is quite some industrial demand for axi-symmetric ceramic parts (see for example mould insert, Figure 1). Second, the research on ultrasonic assisted turning, where tools are used with a defined cutting edge, will hopefully support the knowledge generation in ultrasonic assisted machining. It might be clear from the description of experiments in ultrasonic assisted grinding, that there are quite some influencing parameters, which makes it often difficult to draw clear conclusions (e.g. influence of tool wear on material removal 4 CONCLUSIONS This paper described on-going developments at K.U.Leuven on advanced tool path generation for multi-axis machine tools and research on ultrasonic assisted machining of ceramic materials/components. Main issue in tool path generation is the integration of machine information (kinematics and dynamics). So far, research on ultrasonic assisted machining has been mainly focused on ultrasonic assisted grinding. In order to have a better understanding, near future research aims the ultrasonic assisted turning of ceramic materials. ACKNOWLEDGEMENT The authors want to thank the Machine Tool Technologies Research Foundation (MTTRF) for the support in making a CNC Vertical Machining Center (Mori Seiki NL200Y/500) available to the Katholieke Universiteit Leuven. In addition, the research has been made possible by the financial support of FWO (Research Foundation Flanders, project Multi-axis rotating ultrasonic machining of ceramic components ), IWT (Institute for the Promotion of Innovation by Science and Technology in Flanders, IWT/ EU CORNET project Ultrasonic assisted grinding of brittle hard materials UAG) and the European Union (project FP6 Next Generation Production Systems ). REFERENCES [1] Lauwers, B., P.P. Lefebvre, 2006, Part Analysis Algorithms for Efficient 5-Axis Milling Strategy Planning of Sculptured Surfaces, 2 nd International Conference on High Performance Cutting, Vancouver, CD-Rom. [2] Lauwers, B., Lefebvre, P., Kruth, J.P., Van Brussel, H., Development of Strategies and Systems for the Machining of Complex Shaped Components and Ceramic Materials, Proceedings of MTTRF 2007 meeting, July 10-12, 2007, Nagano, Japan, pp [3] Lauwers, B., Kruth, J.P., Brans, K., Development of Technology and Strategies for the Machining of Ceramic Components by Sinking and Milling EDM, Annals of the CIRP, Vol. 56/1, 2007, pp [4] Daus, N.-A., Ultraschallunterstützes Quer-Seiten- Schleifen, PhD thesis; Technische Universität Berlin [5] Uhlmann, E., Hübert, C, Advances in Ultrasonic Assisted Grinding of Ceramic Materials, Advances in Science and Technology, Vol. 45, pp , [6] Zeng, W.M., Li, Z.C., Pei, Z.J., Treadwell, C., Experimental observation of tool wear in rotary ultrasonic machining of advanced ceramics, International Journal of Machine Tools and Manufacture, Vol. 45, Nr , pp , 2005.

A study of accuracy of finished test piece on multi-tasking machine tool

A study of accuracy of finished test piece on multi-tasking machine tool M. Saito 1, Y. Ihara 1, K. Shimojima 2 1 Osaka Institute of Technology, Japan 2 Okinawa National College of Technology, Japan yukitoshi.ihara@oit.ac.jp