Table of ContentsPreface Development in the Dressing of Super Abrasive Grinding Wheels High Speed Grinding of Advanced Ceramics: A Review Experimental Investigations on Material Removal
Trang 2Progress in Abrasive and Grinding Technology
Trang 3and Grinding Technology
Special topic volume with invited papers only
Edited by
Xipeng Xu
TRANS TECH PUBLICATIONS LTD Switzerland • UK • USA
Trang 4Copyright 2009 Trans Tech Publications Ltd, Switzerland
All rights reserved No part of the contents of this publication may be reproduced or
transmitted in any form or by any means without the written permission of the publisher
Trans Tech Publications Ltd
Full text available online at http://www.scientific.net
Distributed worldwide by and in the Americas by
Trans Tech Publications Ltd Trans Tech Publications Inc
Trang 5Grinding and abrasive processing of materials are the machining processes that use bonded or loose abrasives to remove workpiece materials Due to the well-known advantages of grinding and abrasive processes, advances in abrasive and grinding technology are of importance to enhance both productivity and part quality In order to introduce the progresses in this field, the vice president of Trans Tech Publications, Thomas Wohlbier, invited me to edit this special volume last year I have invited 21 contributions from different countries and regions in an attempt to gather together the achievements of different researchers into a single publication
The 21 invited papers, review or research, are from Australia, China, Germany, Japan, Singapore, Taiwan (China), UK, and USA The abrasive processes addressed in the volume involve not only grinding and polishing, but also wire sawing and abrasive waterjet machining The topics include either fundamental aspects or novel techniques It is therefore the hope of the editor that this volume will be valuable to production and research engineers, research students and academics in the area
At the completion of this volume, I am grateful to all the contributors for the enthusiasm with which they wrote about their topics Thanks are also given to Mr Guoqin Huang at HuaQiao University for his secretarial and editing work; and Trans Tech Publications for publishing the volume
Xipeng Xu Ph.D
Professor in Manufacturing Engineering
HuaQiao University Quanzhou, Fujian 362021, China Tel.: +86-595-22693598; fax: +86-595-22692667
E-mail address: xpxu@hqu.edu.cn
Trang 6Table of Contents
Preface
Development in the Dressing of Super Abrasive Grinding Wheels
High Speed Grinding of Advanced Ceramics: A Review
Experimental Investigations on Material Removal Rate and Surface Roughness in Lapping
of Substrate Wafers: A Literature Review
A Focused Review on Enhancing the Abrasive Waterjet Cutting Performance by Using
Controlled Nozzle Oscillation
A Review of Electrolytic In-Process Dressing (ELID) Grinding
On the Coherent Length of Fluid Nozzles in Grinding
Surface Characteristics of Efficient-Ground Alumina and Zirconia Ceramics for Dental
Applications
Optimization of Cutting-Edge Truncation in Ductile-Mode Grinding of Optical Glass
On the Polishing Techniques of Diamond and Diamond Composites
Super Polishing Behaviour Investigation of Stainless Steel Optical Lens Moulding Inserts
Corrective Abrasive Polishing Processes for Freeform Surface
Applications of Contact Length Models in Grinding Processes
Polishing Performance of Electro-Rheological Fluid of Polymerized Liquid Crystal
Contained Abrasive Grit
Study on Tribo-Fabrication in Polishing by Nano Diamond Colloid
Efficient Super-Smooth Finishing Characteristics of SiC Materials through the Use of
Fine-Grinding
Polishing of Ultra Smooth Surface with Nanoparticle Colloid Jet
An Experimental Study on High Speed Grinding of Granite with a Segmented Diamond
Wheel
Thinning Silicon Wafer with Polycrystalline Diamond Tools
Mechanisms of Al/SiC Composite Machining with Diamond Whiskers
Effect of Slurry and Nozzle on Hole Machining of Glass by Micro Abrasive Suspension Jets
Experimental Investigation of Temperatures in Diamond Wire Sawing Granite
Trang 7Development in the Dressing of Super Abrasive Grinding Wheels
1
Leibniz Universität Hannover, Institute of Production Engineering and Machine Tools,
An der Universität 2, D-30823, Germany a
denkena@ifw.uni-hannover.de; bleon@ifw.uni-hannover.de; cwang@ifw.uni-hannover.de;
dhahmann@ifw.uni-hannover.de
Keywords: Electro contact discharge dressing, Profile dressing, Microprofiles, Super abrasives
Abstract Harder workpiece materials and increased efficiency requirements for grinding processes
make the use of super abrasive grinding wheels indispensable This paper presents newly developed processes for the dressing of super abrasive grinding wheels The different bond systems of grinding wheels require distinct dressing process In this paper, dressing processes for metal and vitrified bonded grinding wheels are investigated It introduces the method of electro contact discharge dressing for the conditioning of metal-bonded, fine-grained multilayer grinding wheels A description of the essential correlation between dressing parameters and the material removal rate of the bond material is presented The considered parameters are the dressing voltage, the limitation of the dressing current and the feed as well as the infeed of the electrode For the grinding of functional microgroove structures, multiroof profiles with microscopic tip geometries are dressed onto the grinding wheel For this, a profile roller in combination with a special shifting strategy is applied on finegrained vitrified bonded grinding wheels
Introduction
High performance components with high hardness and wear resistance are applied with increasing frequency in order to enhance the efficiency of technical systems Furthermore, miniaturized products and microstructured functional surfaces entail new challenges for machining processes Grinding processes with super abrasive CBN and diamond grinding wheels can be used for the economical machining of such components and microgeometries Depending on the bond system of the grinding wheel, different dressing processes should be used [1, 2] To assure small form and dimensional tolerances over an adequate number of workpieces, the grinding wheels have to be regularly redressed In the following, electro contact discharge dressing for metal bonded grinding wheels and a novel dressing strategy using special shift kinematics for vitrified bonded grinding wheels are described The focus of these dressing processes is the profiling of grinding wheels
Electro Contact Discharge Dressing
The effects of continuous wear on process stability as well as on shape and dimension accuracies of
a component are more significant for fine-grained grinding tools used for micro-machining than they are for “conventional” precision grinding In order to counterbalance those influences, wear-resistant grinding tools and procedures for the regeneration of the tool profile are necessary Due to their high wear-resistance and the resulting profile retention, multilayered, metallically bonded diamond grinding wheels are more suitable for micromachining than vitrified or resin bonded tools The main problem is the dressing of those metallically bonded tools Electro contact discharge dressing is a promising method to cope with this challenge It has so far only been used for sharpening, but not for the dressing of tools [3~5]
In the following, the effects of the process variables on the contact erosive removal of the bond material are described It is determined under which conditions a continuous removal of the bond material and thus a durable dressing effect can be achieved Emphasis is put on the significant
Development in the Dressing of Super Abrasive Grinding Wheels
1
Leibniz Universität Hannover, Institute of Production Engineering and Machine Tools,
An der Universität 2, D-30823, Germany a
denkena@ifw.uni-hannover.de; bleon@ifw.uni-hannover.de; cwang@ifw.uni-hannover.de;
dhahmann@ifw.uni-hannover.de
Keywords: Electro contact discharge dressing, Profile dressing, Microprofiles, Super abrasives
Abstract Harder workpiece materials and increased efficiency requirements for grinding processes
make the use of super abrasive grinding wheels indispensable This paper presents newly developed processes for the dressing of super abrasive grinding wheels The different bond systems of grinding wheels require distinct dressing process In this paper, dressing processes for metal and vitrified bonded grinding wheels are investigated It introduces the method of electro contact discharge dressing for the conditioning of metal-bonded, fine-grained multilayer grinding wheels A description of the essential correlation between dressing parameters and the material removal rate of the bond material is presented The considered parameters are the dressing voltage, the limitation of the dressing current and the feed as well as the infeed of the electrode For the grinding of functional microgroove structures, multiroof profiles with microscopic tip geometries are dressed onto the grinding wheel For this, a profile roller in combination with a special shifting strategy is applied on finegrained vitrified bonded grinding wheels
Introduction
High performance components with high hardness and wear resistance are applied with increasing frequency in order to enhance the efficiency of technical systems Furthermore, miniaturized products and microstructured functional surfaces entail new challenges for machining processes Grinding processes with super abrasive CBN and diamond grinding wheels can be used for the economical machining of such components and microgeometries Depending on the bond system of the grinding wheel, different dressing processes should be used [1, 2] To assure small form and dimensional tolerances over an adequate number of workpieces, the grinding wheels have to be regularly redressed In the following, electro contact discharge dressing for metal bonded grinding wheels and a novel dressing strategy using special shift kinematics for vitrified bonded grinding wheels are described The focus of these dressing processes is the profiling of grinding wheels
Electro Contact Discharge Dressing
The effects of continuous wear on process stability as well as on shape and dimension accuracies of
a component are more significant for fine-grained grinding tools used for micro-machining than they are for “conventional” precision grinding In order to counterbalance those influences, wear-resistant grinding tools and procedures for the regeneration of the tool profile are necessary Due to their high wear-resistance and the resulting profile retention, multilayered, metallically bonded diamond grinding wheels are more suitable for micromachining than vitrified or resin bonded tools The main problem is the dressing of those metallically bonded tools Electro contact discharge dressing is a promising method to cope with this challenge It has so far only been used for sharpening, but not for the dressing of tools [3~5]
In the following, the effects of the process variables on the contact erosive removal of the bond material are described It is determined under which conditions a continuous removal of the bond material and thus a durable dressing effect can be achieved Emphasis is put on the significant
Trang 8volume over the dressing time Qd These parameters all vary depending on the strategy chosen for
As start-up phase for electro contact discharge dressing, the split stroke travel with idle stroke is chosen Thus the effects of the variables can be determined (Fig.1) The aim of the dressing strategy
is to attain an even distribution of graphite particles over the thickness of the grinding wheel
1 The electrode is aligned radially next to the dressing wheel (1)
and the grinding wheel overlap axially (2)
influence on the grinding layer during the electrode withdrawal and to guarantee an even electrode profile (3)
This is to provide a further smoothing of the profile
different partial strokes on the process activity The highest process activity occurs when the diagonal feed of the electrode is carried out and when the electrode and the grinding wheel overlap
In the following axial progress of the electrode, the activity slowly decreases and comes to a standstill when there is no more contact between the two interacting parts The following idle stroke leads to low process activity
Fig 1 Start-up phase of electro contact discharge dressing with idle stroke
The effects of the variables on the process are described by the specific material removal rate
removed from the dressing wheel and the machined volume of the electrode
The experiments were carried out in distilled water, which has proved to be a suitable medium in
the maximal current in a short circuit which can be set at the power supply unit They can be
to attain a dressing effect, the voltage has to exceed a critical value which causes a maximal grain protrusion and a continuous removal of bond material Under given boundary conditions, there is no
As start-up phase for electro contact discharge dressing, the split stroke travel with idle stroke is chosen Thus the effects of the variables can be determined (Fig.1) The aim of the dressing strategy
is to attain an even distribution of graphite particles over the thickness of the grinding wheel
1 The electrode is aligned radially next to the dressing wheel (1)
and the grinding wheel overlap axially (2)
influence on the grinding layer during the electrode withdrawal and to guarantee an even electrode profile (3)
This is to provide a further smoothing of the profile
different partial strokes on the process activity The highest process activity occurs when the diagonal feed of the electrode is carried out and when the electrode and the grinding wheel overlap
In the following axial progress of the electrode, the activity slowly decreases and comes to a standstill when there is no more contact between the two interacting parts The following idle stroke leads to low process activity
Fig 1 Start-up phase of electro contact discharge dressing with idle stroke
The effects of the variables on the process are described by the specific material removal rate
removed from the dressing wheel and the machined volume of the electrode
The experiments were carried out in distilled water, which has proved to be a suitable medium in
the maximal current in a short circuit which can be set at the power supply unit They can be
to attain a dressing effect, the voltage has to exceed a critical value which causes a maximal grain protrusion and a continuous removal of bond material Under given boundary conditions, there is no
Trang 9reached When the off-load voltage increases further, there is no further rise in Q’ds, which means that the run of the curve has approached a critical value In the following, a possible explanation for
the electrode show a distinct distribution At low voltages, only few particles are large enough to enable discharges When the voltage increases, the number of suitable particles and thus the probability of discharge increase An analogy investigation, carried out under the same electric and geometric conditions as in the real process, showed that at about 35 V, discharges even occur without any graphite particles implied This shows that the maximal probability of discharge is reached The limiting factor is that there can only be one discharge at a time
the electro contact discharge process (Fig 2, right) The electrode current in the spark gap occurs at
the sum of the local single currents which cause the removal of the bond material Low current
no more rise in the volume flow rate at the grinding wheel This can be explained by the energy released at each discharge under the assumption of a constant discharge duration The energy released at a discharge and thus the temperature in the metal bond increase with a rise in the current
At a certain energy level, the metal bond starts to melt locally The maximal volume of bond which can be molten is limited by the specific boiling temperature and the specific thermal conductivity The temperature of the molten material cannot exceed the boiling temperature Thermal conductivity limits the volume of material which reaches the melting temperature due to heat dispersion, assuming a constant discharge duration The duration will be determined from the experimental data
Fig 2 Specific material removal rate during electro contact discharge dressing
development of the quality factor in both diagrams can be explained by the constant specific material removal rate at the electrode, which is itself due to constant infeed and feed throughout the investigation Thus the quality factor corresponds to the specific material removal rate
that the run of the curve has approached a critical value In the following, a possible explanation for
the electrode show a distinct distribution At low voltages, only few particles are large enough to enable discharges When the voltage increases, the number of suitable particles and thus the probability of discharge increase An analogy investigation, carried out under the same electric and geometric conditions as in the real process, showed that at about 35 V, discharges even occur without any graphite particles implied This shows that the maximal probability of discharge is reached The limiting factor is that there can only be one discharge at a time
the electro contact discharge process (Fig 2, right) The electrode current in the spark gap occurs at
the sum of the local single currents which cause the removal of the bond material Low current
no more rise in the volume flow rate at the grinding wheel This can be explained by the energy released at each discharge under the assumption of a constant discharge duration The energy released at a discharge and thus the temperature in the metal bond increase with a rise in the current
At a certain energy level, the metal bond starts to melt locally The maximal volume of bond which can be molten is limited by the specific boiling temperature and the specific thermal conductivity The temperature of the molten material cannot exceed the boiling temperature Thermal conductivity limits the volume of material which reaches the melting temperature due to heat dispersion, assuming a constant discharge duration The duration will be determined from the experimental data
Fig 2 Specific material removal rate during electro contact discharge dressing
development of the quality factor in both diagrams can be explained by the constant specific material removal rate at the electrode, which is itself due to constant infeed and feed throughout the investigation Thus the quality factor corresponds to the specific material removal rate
Trang 10Fig 3 Quality factor in electro contact discharge dressing
Fig 4 Specific material removal rate in electro contact discharge dressing
Besides by the electric variables of electro contact discharge dressing, the process is also
identical for all single strokes (see Fig 1)
size of the graphite particles that are cut off from the electrode The mean particle size rises both when the feed or the infeed increase This is due to the increase in the equivalent mean chip thickness The probability of discharge increases with the particle size until the maximal probability
of discharge is reached The limiting factor is that there can only be one discharge at a time
Fig 3 Quality factor in electro contact discharge dressing
Fig 4 Specific material removal rate in electro contact discharge dressing
Besides by the electric variables of electro contact discharge dressing, the process is also
identical for all single strokes (see Fig 1)
size of the graphite particles that are cut off from the electrode The mean particle size rises both when the feed or the infeed increase This is due to the increase in the equivalent mean chip thickness The probability of discharge increases with the particle size until the maximal probability
of discharge is reached The limiting factor is that there can only be one discharge at a time
Trang 11Investigations concerning the graphite particle size for varied feed and infeed values will have to be carried out to proof this assumption
Fig 5 Quality factor in electro contact discharge dressing
the quality factor This can be explained by the fact that the specific material removal rate reaches its limit at higher feed and infeed levels of the electrode while the machined volume of the electrode
influence both the volume flow rate and the quality factor of the dressing process Independent of the examined variables, an increase in the volume flow rate is only possible until a certain critical value is reached In order to obtain the maximal quality factor, a compromise has to be found in the
Both straight as well as v-shaped dressing wheel profiles [6] can be produced with this new process for electro contact discharge dressing
Profile Dressing with Special Shift Kinematics for the Generation of Microprofiles
In recent years, the manufacturing of microstructured functional component surfaces has become
the focus of many research works As a typical example, longitudinal microgroove structures, known as riblets, have been extensively investigated during the last decade and proven to reduce skin friction and wall shear stresses in turbulent flow up to 10% compared with smooth surfaces [7, 8] For most technical applications of riblets, microgroove structures with a width of less than
100 µm and a depth of the half of the width are required on large-area surfaces In comparison to other machining processes, grinding offers high potential for large-area microstructuring The main reason for this is the fact that several groove structures can be produced by one run over the surface with a multiprofiled grinding wheel Grinding wheel profiles with microscopic profile peak geometries have to be generated by a dressing process to produce microgrooves using profile grinding For the current investigations, vitrified bonded wheels are selected due to their good dressability and profile holding properties compared to other bonding systems [1] In the following,
a novel profile dressing method using special profile shift kinematics is introduced
For the dressing of multiprofiled vitrified grinding wheels, there are two main dressing methods (Fig 6) The first method is form dressing using a diamond form roller (Fig 6, left) The contour of the wheel is generated by NC-programs and dressed by the dressing tool along the axial direction The whole wheel profile is generated layer by layer Due to the axial dressing path over all of the
Investigations concerning the graphite particle size for varied feed and infeed values will have to be carried out to proof this assumption
Fig 5 Quality factor in electro contact discharge dressing
the quality factor This can be explained by the fact that the specific material removal rate reaches its limit at higher feed and infeed levels of the electrode while the machined volume of the electrode
influence both the volume flow rate and the quality factor of the dressing process Independent of the examined variables, an increase in the volume flow rate is only possible until a certain critical value is reached In order to obtain the maximal quality factor, a compromise has to be found in the
Both straight as well as v-shaped dressing wheel profiles [6] can be produced with this new process for electro contact discharge dressing
Profile Dressing with Special Shift Kinematics for the Generation of Microprofiles
In recent years, the manufacturing of microstructured functional component surfaces has become
the focus of many research works As a typical example, longitudinal microgroove structures, known as riblets, have been extensively investigated during the last decade and proven to reduce skin friction and wall shear stresses in turbulent flow up to 10% compared with smooth surfaces [7, 8] For most technical applications of riblets, microgroove structures with a width of less than
100 µm and a depth of the half of the width are required on large-area surfaces In comparison to other machining processes, grinding offers high potential for large-area microstructuring The main reason for this is the fact that several groove structures can be produced by one run over the surface with a multiprofiled grinding wheel Grinding wheel profiles with microscopic profile peak geometries have to be generated by a dressing process to produce microgrooves using profile grinding For the current investigations, vitrified bonded wheels are selected due to their good dressability and profile holding properties compared to other bonding systems [1] In the following,
a novel profile dressing method using special profile shift kinematics is introduced
For the dressing of multiprofiled vitrified grinding wheels, there are two main dressing methods (Fig 6) The first method is form dressing using a diamond form roller (Fig 6, left) The contour of the wheel is generated by NC-programs and dressed by the dressing tool along the axial direction The whole wheel profile is generated layer by layer Due to the axial dressing path over all of the
Trang 12profiles to be dressed, dressing using a form roller is a highly time consuming process Furthermore, the actual geometry of the dressing tool is required for the dressing tool correction in the NC-program Due to the ongoing wear of the dressing tool, this actual geometry is very difficult to determine Besides that, the stability of the machine axis control system can be negatively influenced by the temperature effects due to the long dressing time The second method is dressing using a diamond profile roller The negative profile of the grinding wheel is mapped on the dressing roller The whole wheel profile is generated within a plunge movement of the dressing tool toward the grinding wheel Compared to the form dressing process, profile dressing offers a higher dressing efficiency and process stability On the other hand, the dressing force could be higher due to the longer tool contact width, which is to be considered to achieve a stable dressing process
Fig 6 Dressing kinematics for form dressing and profile dressing
In the current riblet grinding studies, multiroof profiles with a microtip geometry on the grinding
microprofiles on grinding wheels with a profile tip radius smaller than 50 µm directly by profile dressing This is due to the limited minimal profile geometry on the diamond profile dressing roller, which can be produced [9] Hence a novel dressing strategy using profile rollers is introduced in the following (Fig 7) With the new dressing strategy, roof profiles with an ideal sharp profile tip can
be produced, if the breakout behavior of the grinding layer is not being considered In the first plunge movement, one flank of the profiles is dressed The second plunge movement is carried out with an axial offset of the dressing roller, whereby the other flank of the profiles is dressed Due to the special process kinematics, all roof profiles on the grinding wheel can be produced within two plunge movements Furthermore, generally the tip areas of the dressing roller profile undertake a higher load than the flank areas Using the new dressing kinematics, the flank areas of the dressing roller profile are deciding for the generation of the tips of the roof profiles and enable a higher wear resistance
profiles to be dressed, dressing using a form roller is a highly time consuming process Furthermore, the actual geometry of the dressing tool is required for the dressing tool correction in the NC-program Due to the ongoing wear of the dressing tool, this actual geometry is very difficult to determine Besides that, the stability of the machine axis control system can be negatively influenced by the temperature effects due to the long dressing time The second method is dressing using a diamond profile roller The negative profile of the grinding wheel is mapped on the dressing roller The whole wheel profile is generated within a plunge movement of the dressing tool toward the grinding wheel Compared to the form dressing process, profile dressing offers a higher dressing efficiency and process stability On the other hand, the dressing force could be higher due to the longer tool contact width, which is to be considered to achieve a stable dressing process
Fig 6 Dressing kinematics for form dressing and profile dressing
In the current riblet grinding studies, multiroof profiles with a microtip geometry on the grinding
microprofiles on grinding wheels with a profile tip radius smaller than 50 µm directly by profile dressing This is due to the limited minimal profile geometry on the diamond profile dressing roller, which can be produced [9] Hence a novel dressing strategy using profile rollers is introduced in the following (Fig 7) With the new dressing strategy, roof profiles with an ideal sharp profile tip can
be produced, if the breakout behavior of the grinding layer is not being considered In the first plunge movement, one flank of the profiles is dressed The second plunge movement is carried out with an axial offset of the dressing roller, whereby the other flank of the profiles is dressed Due to the special process kinematics, all roof profiles on the grinding wheel can be produced within two plunge movements Furthermore, generally the tip areas of the dressing roller profile undertake a higher load than the flank areas Using the new dressing kinematics, the flank areas of the dressing roller profile are deciding for the generation of the tips of the roof profiles and enable a higher wear resistance
Trang 13Fig 7 Dressing strategy for the generation of microprofiles using a diamond profile roller The dressing experiments have been carried out on a high precision surface grinding machine of the type Blohm, profimat 407 with an integrated profile dressing system According to the results of the last studies on form dressing [10], a fine grained vitrified bonded SiC-wheel with an average grain size of 17 µm and an outer diameter of 300 mm was selected at the beginning The influences
of the profile dressing parameters on the grinding wheel topography have already been described in the literature [11, 12] The conclusions show that the actual surface roughness of the grinding wheel
the current application of grinding microgrooves, microscopic wheel profiles are required The main issue of the dressing experiments is to investigate the minimal achievable wheel profile geometry Due to the mechanical load on the vitrified wheel layer during the dressing process, breakouts occur
at the profile tip, where the structure strength is not high enough to withstand the dressing force
dressing results (Fig 6)
varied The new dressing strategy with the special profile shift kinematics has been applied The dressing results show a significant influence of the dressing parameters on the profile height
increase, which causes larger breakouts of the profile tip On the other hand, the profile accuracy improves with a decreasing dressing speed ratio from the down dressing to the up dressing mode
Among the different variations in the matrix, the best result (∆h = 20 µm) has been achieved at
µm/rev has not been investigated
Fig 7 Dressing strategy for the generation of microprofiles using a diamond profile roller The dressing experiments have been carried out on a high precision surface grinding machine of the type Blohm, profimat 407 with an integrated profile dressing system According to the results of the last studies on form dressing [10], a fine grained vitrified bonded SiC-wheel with an average grain size of 17 µm and an outer diameter of 300 mm was selected at the beginning The influences
of the profile dressing parameters on the grinding wheel topography have already been described in the literature [11, 12] The conclusions show that the actual surface roughness of the grinding wheel
the current application of grinding microgrooves, microscopic wheel profiles are required The main issue of the dressing experiments is to investigate the minimal achievable wheel profile geometry Due to the mechanical load on the vitrified wheel layer during the dressing process, breakouts occur
at the profile tip, where the structure strength is not high enough to withstand the dressing force
dressing results (Fig 6)
varied The new dressing strategy with the special profile shift kinematics has been applied The dressing results show a significant influence of the dressing parameters on the profile height
increase, which causes larger breakouts of the profile tip On the other hand, the profile accuracy improves with a decreasing dressing speed ratio from the down dressing to the up dressing mode
Among the different variations in the matrix, the best result (∆h = 20 µm) has been achieved at
µm/rev has not been investigated
Trang 14Fig 8 Dressing results at parameter variations using a diamond form roller
Fig 9 Influences of the profile angle on the profile dressing process Based on the results at the SiC400 wheel, a vitrified bonded CBN wheel with a grain size of
16 µm (MB16) has been applied for the following dressing experiments In comparison with conventional wheels, CBN wheels offer a large potential regarding tool wear resistance However, the bonding system of CBN wheels is generally much harder than that of conventional grinding wheels The pictures in the right in Fig 9 show SEM-pictures of the topography of the SiC400 H (on top) and those of the CBN wheel (below) The SiC grains and the pores are distributed uniformly troughout the bond The bonding bridges are very short shaped At the CBN wheel, the grains build many clusters which are fully surrounded by the bond material During the first dressing experiment at the CBN wheel with a target profile angle of 45°, sidewise profile breakouts could be observed at large areas, which lead to a blunt profile tip geometry (∆h = 45 µm)
When the profile angle is increased from 45° to 90°, the profile holding performance improves and the achieved tip geometry dimension decreases The ∆h is about 7 µm at a profile angle of 90°
Fig 8 Dressing results at parameter variations using a diamond form roller
Fig 9 Influences of the profile angle on the profile dressing process Based on the results at the SiC400 wheel, a vitrified bonded CBN wheel with a grain size of
16 µm (MB16) has been applied for the following dressing experiments In comparison with conventional wheels, CBN wheels offer a large potential regarding tool wear resistance However, the bonding system of CBN wheels is generally much harder than that of conventional grinding wheels The pictures in the right in Fig 9 show SEM-pictures of the topography of the SiC400 H (on top) and those of the CBN wheel (below) The SiC grains and the pores are distributed uniformly troughout the bond The bonding bridges are very short shaped At the CBN wheel, the grains build many clusters which are fully surrounded by the bond material During the first dressing experiment at the CBN wheel with a target profile angle of 45°, sidewise profile breakouts could be observed at large areas, which lead to a blunt profile tip geometry (∆h = 45 µm)
When the profile angle is increased from 45° to 90°, the profile holding performance improves and the achieved tip geometry dimension decreases The ∆h is about 7 µm at a profile angle of 90°
Trang 15and the required profile tip geometry for riblet-grinding in the current study can be achieved This behavior can be explained by the higher structure stability and bonding force, which are due to the sidewise bonding support at the profile tip To analyze the structure stability at the profile tip, FEM-modeling has been carried out at a varied profile angle from 45°, 60° to 90° The wheel profile was modeled as a homogenous body A constant dressing load is applied on the wheel tip and the results show that at the smaller angle of 45° the maximal von Mises stress in the grinding layer is about four times higher than at a 90° profile angle The same trend can also be proved at the SiC wheel during the dressing experiments However the skip is not as significant as with the CBN wheel with a more closed bonding structure To compare the real geometry for different profile angles, the profile tip height at a fixed profile width should be used for the evaluation Furthermore, the resulting profile stability at grinding should also be taken into account
Summary
In this Paper two different processes for the dressing of super abrasive grinding wheels have been presented For metal bonded grinding wheels electro contact discharge dressing was applied For vitrified bonded grinding wheels a novel dressing strategy using special shift kinematics was introduced
The application of electro contact discharge dressing allows generating the topography and geometry of fine grained grinding wheels in one process The topography is generated at lower
further increase in the dressing parameters does not affect a further increase in the specific material
depends on the movement path of the electrode
For the grinding of functional microgroove structures like riblets a novel dressing strategy using
a profile roller in combination with shift kinematics was applied to vitrified bonded SiC and CBN grinding wheels By this method, the limit of the smallest dressable wheel profile tip geometry at profile dressing could be reduced significantly To achieve smaller profile breakouts and higher profile accuracy, a small dressing infeed and the up dressing mode should be chosen at dressing microprofiles Furthermore, the angle of the roof profile has a high impact on the profile stability at dressing especially for super abrasive CBN wheels with a hard bonding system
References
[1] F Klocke, W König: Fertigungsverfahren 2: Schleifen, Honen, Läppen Springer Verlag Berlin (2005), ISBN 978-3-540-23496-8
[2] H.K Tönshoff, B Denkena: Spanen Springer Verlag Berlin (2003), ISBN 978-3-540- 00588-9
[3] Y Falkenberg: Elektroerosives Schärfen von Bornitridschleifscheiben Dr.-Ing Dissertation,
Universität Hannover, Germany (1997)
[4] T Friemuth: Schleifen Hartstoffverstärkter Keramischer Werkzeuge Dr.-Ing Dissertation,
Universität Hannover, Germany (1999)
[5] J Xie, J Tamaki: In-process Evaluation of Grit Protrusion Feature for Fine Diamond Grinding wheel by Means of Electro-Contact Discharge Dressing Journal of Materials Processing Technology, Vol 180 (2006), pp 83-90
and the required profile tip geometry for riblet-grinding in the current study can be achieved This behavior can be explained by the higher structure stability and bonding force, which are due to the sidewise bonding support at the profile tip To analyze the structure stability at the profile tip, FEM-modeling has been carried out at a varied profile angle from 45°, 60° to 90° The wheel profile was modeled as a homogenous body A constant dressing load is applied on the wheel tip and the results show that at the smaller angle of 45° the maximal von Mises stress in the grinding layer is about four times higher than at a 90° profile angle The same trend can also be proved at the SiC wheel during the dressing experiments However the skip is not as significant as with the CBN wheel with a more closed bonding structure To compare the real geometry for different profile angles, the profile tip height at a fixed profile width should be used for the evaluation Furthermore, the resulting profile stability at grinding should also be taken into account
Summary
In this Paper two different processes for the dressing of super abrasive grinding wheels have been presented For metal bonded grinding wheels electro contact discharge dressing was applied For vitrified bonded grinding wheels a novel dressing strategy using special shift kinematics was introduced
The application of electro contact discharge dressing allows generating the topography and geometry of fine grained grinding wheels in one process The topography is generated at lower
further increase in the dressing parameters does not affect a further increase in the specific material
depends on the movement path of the electrode
For the grinding of functional microgroove structures like riblets a novel dressing strategy using
a profile roller in combination with shift kinematics was applied to vitrified bonded SiC and CBN grinding wheels By this method, the limit of the smallest dressable wheel profile tip geometry at profile dressing could be reduced significantly To achieve smaller profile breakouts and higher profile accuracy, a small dressing infeed and the up dressing mode should be chosen at dressing microprofiles Furthermore, the angle of the roof profile has a high impact on the profile stability at dressing especially for super abrasive CBN wheels with a hard bonding system
References
[1] F Klocke, W König: Fertigungsverfahren 2: Schleifen, Honen, Läppen Springer Verlag Berlin (2005), ISBN 978-3-540-23496-8
[2] H.K Tönshoff, B Denkena: Spanen Springer Verlag Berlin (2003), ISBN 978-3-540- 00588-9
[3] Y Falkenberg: Elektroerosives Schärfen von Bornitridschleifscheiben Dr.-Ing Dissertation,
Universität Hannover, Germany (1997)
[4] T Friemuth: Schleifen Hartstoffverstärkter Keramischer Werkzeuge Dr.-Ing Dissertation,
Universität Hannover, Germany (1999)
[5] J Xie, J Tamaki: In-process Evaluation of Grit Protrusion Feature for Fine Diamond Grinding wheel by Means of Electro-Contact Discharge Dressing Journal of Materials Processing Technology, Vol 180 (2006), pp 83-90
Trang 16[6] B Denkena, M Reichstein, D Hahmann: Electro Contact Discharge Dressing for Micro-
Grinding Proceedings of the 6th Euspen International Conference (2006), Baden, Austria, pp
92-95
[7] M.J Walsh: Riblets in Viscous Drag Reduction in Boundary Layers Edited by Bushnell, D.M.:
Progress in Astronautics and Aeronautics, AIAA, Washington DC/USA (1990)
[8] W Hage: Zur Widerstandsverminderung von Dreidimensionalen Riblet-Strukturen und
Anderen Oberflächen Dr.-Ing Dissertation, Technical University Berlin, Germany (2005)
[9] Product information, Dr Kaiser Diamantwerkzeuge GmbH, Germany (2008)
[10] B Denkena, M Reichstein, B Wang: Manufacturing of Micro-Functional Structures by
Grinding Annals of the German Academic Society for Production Engineering (WGP), Vol
XIII/1, pp 31-34 (2006)
[11] R Schmitt: Abrichten von Schleifscheiben Mit Diamantbestückten Rollen Dr.- Ing
Dissertation Technical University Braunschweig, Germany (1968)
[12] E Minke: Handbuch zur Abrichttechnik Riegger Diamantwerkzeuge GmbH (1999)
[13] B Linke: Wirkmechanismen Beim Abrichten Keramisch Gebundener Schleifscheiben Dr.-Ing
Dissertation, RWTH Aachen, Germany (2007)
[6] B Denkena, M Reichstein, D Hahmann: Electro Contact Discharge Dressing for Micro-
Grinding Proceedings of the 6th Euspen International Conference (2006), Baden, Austria, pp
92-95
[7] M.J Walsh: Riblets in Viscous Drag Reduction in Boundary Layers Edited by Bushnell, D.M.:
Progress in Astronautics and Aeronautics, AIAA, Washington DC/USA (1990)
[8] W Hage: Zur Widerstandsverminderung von Dreidimensionalen Riblet-Strukturen und
Anderen Oberflächen Dr.-Ing Dissertation, Technical University Berlin, Germany (2005)
[9] Product information, Dr Kaiser Diamantwerkzeuge GmbH, Germany (2008)
[10] B Denkena, M Reichstein, B Wang: Manufacturing of Micro-Functional Structures by
Grinding Annals of the German Academic Society for Production Engineering (WGP), Vol
XIII/1, pp 31-34 (2006)
[11] R Schmitt: Abrichten von Schleifscheiben Mit Diamantbestückten Rollen Dr.- Ing
Dissertation Technical University Braunschweig, Germany (1968)
[12] E Minke: Handbuch zur Abrichttechnik Riegger Diamantwerkzeuge GmbH (1999)
[13] B Linke: Wirkmechanismen Beim Abrichten Keramisch Gebundener Schleifscheiben Dr.-Ing
Dissertation, RWTH Aachen, Germany (2007)
Trang 17High Speed Grinding of Advanced Ceramics: A Review
School of Engineering, The University of Queensland, Brisbane, QLD4072, Australia
ahan.huang@uq.edu.au
Keywords: High speed grinding, Force, Temperature, Ceramics, Subsurface, Coolant supply
Abstract In this paper, the characteristics of high speed grinding of advanced ceramics, including
alumina, alumina-titania, zirconia, silicon nitride and silicon carbide, were reviewed The associated material removal mechanisms were discussed Pragmatic technologies for the high speed grinding of advanced ceramics were also presented
Introduction
Advanced ceramics have become increasingly more important structural materials in modern manufacturing industries due to their excellent properties, such as high hardness at both ambient and elevated temperatures, low thermal expansion, good wear resistance and chemical inertness [1, 2] Grinding with diamond abrasives is the most commonly used machining process for the fabrication of structural components made of the ceramics [1-4] However, the high cost associated with machining of ceramics components has been a major factor that has hindered their application Consequently, in the past several decades considerable research efforts [5-15] have been directed towards the development of efficient grinding processes for the advanced ceramics
High speed grinding was first developed as a finishing machining process over 40 years ago [16] The grinding process was characterized by the elevated wheel velocity of above 60 m/s, which significantly reduces the maximum chip thickness in material removal, compared to the conventional grinding that normally refers to the grinding process with a wheel velocity of below
40 m/s Apparently, the smaller chip thickness resulted from the elevation of wheel velocity led to
a reduction in grinding force This was in favor of either achieving a higher removal rate or an improved component quality when the high speed grinding technology was applied into the machining of advanced ceramics [17-24] When using the high speed grinding as a finishing operation of a ceramic component, an increase in wheel velocity would enhance the tendency towards ductile material removal in the process [25] This could result in an improved surface quality in comparison to the grinding at a conventional velocity On the other hand, the application of high speed grinding enabled the achievement of a higher material removal rate via increasing either depth of cut or feed rate, potentially without deteriorating the integrity of ground surfaces [26, 27] The above mentioned approaches have been well adopted for the development of'
high speed grinding technologies for advanced ceramics For example, Kovach et al [28] clearly
demonstrated that the application of high speed grinding into the machining of advanced ceramics resulted in an improved surface finish Their results also suggested that a transition from a brittle fracture mode to a low damage ‘ductile’ grinding mode could be achieved by increasing the wheel velocity Similar results were achieved in the high speed grinding of silicon nitride [25] It was also demonstrated by Klocke et al [17] that the optimization of peripheral grinding processes
at high speeds for silicon-infiltrated silicon carbide and alumina was achieved by using an increased material removal rate without an increase in wheel wear Yin and Huang [24] showed that in the grinding of silicon nitride with a vitrified diamond wheel the increase in wheel speed from a conventional velocity of 30 m/s to a high velocity of 160 m/s led to a 5 times higher removal rate and 7 times longer dressing interval without lowering ceramic strength Huang et al [29, 30] demonstrated that the combination of high speed and large depth of cut resulted in a high
High Speed Grinding of Advanced Ceramics: A Review
School of Engineering, The University of Queensland, Brisbane, QLD4072, Australia
ahan.huang@uq.edu.au
Keywords: High speed grinding, Force, Temperature, Ceramics, Subsurface, Coolant supply
Abstract In this paper, the characteristics of high speed grinding of advanced ceramics, including
alumina, alumina-titania, zirconia, silicon nitride and silicon carbide, were reviewed The associated material removal mechanisms were discussed Pragmatic technologies for the high speed grinding of advanced ceramics were also presented
Introduction
Advanced ceramics have become increasingly more important structural materials in modern manufacturing industries due to their excellent properties, such as high hardness at both ambient and elevated temperatures, low thermal expansion, good wear resistance and chemical inertness [1, 2] Grinding with diamond abrasives is the most commonly used machining process for the fabrication of structural components made of the ceramics [1-4] However, the high cost associated with machining of ceramics components has been a major factor that has hindered their application Consequently, in the past several decades considerable research efforts [5-15] have been directed towards the development of efficient grinding processes for the advanced ceramics
High speed grinding was first developed as a finishing machining process over 40 years ago [16] The grinding process was characterized by the elevated wheel velocity of above 60 m/s, which significantly reduces the maximum chip thickness in material removal, compared to the conventional grinding that normally refers to the grinding process with a wheel velocity of below
40 m/s Apparently, the smaller chip thickness resulted from the elevation of wheel velocity led to
a reduction in grinding force This was in favor of either achieving a higher removal rate or an improved component quality when the high speed grinding technology was applied into the machining of advanced ceramics [17-24] When using the high speed grinding as a finishing operation of a ceramic component, an increase in wheel velocity would enhance the tendency towards ductile material removal in the process [25] This could result in an improved surface quality in comparison to the grinding at a conventional velocity On the other hand, the application of high speed grinding enabled the achievement of a higher material removal rate via increasing either depth of cut or feed rate, potentially without deteriorating the integrity of ground surfaces [26, 27] The above mentioned approaches have been well adopted for the development of'
high speed grinding technologies for advanced ceramics For example, Kovach et al [28] clearly
demonstrated that the application of high speed grinding into the machining of advanced ceramics resulted in an improved surface finish Their results also suggested that a transition from a brittle fracture mode to a low damage ‘ductile’ grinding mode could be achieved by increasing the wheel velocity Similar results were achieved in the high speed grinding of silicon nitride [25] It was also demonstrated by Klocke et al [17] that the optimization of peripheral grinding processes
at high speeds for silicon-infiltrated silicon carbide and alumina was achieved by using an increased material removal rate without an increase in wheel wear Yin and Huang [24] showed that in the grinding of silicon nitride with a vitrified diamond wheel the increase in wheel speed from a conventional velocity of 30 m/s to a high velocity of 160 m/s led to a 5 times higher removal rate and 7 times longer dressing interval without lowering ceramic strength Huang et al [29, 30] demonstrated that the combination of high speed and large depth of cut resulted in a high
Trang 18subsurface damage was found to be insignificantly affected by the increase in depth of cut [29, 30] Xie et al [31] employed a depth of cut of 6 mm in the high speed grinding of zirconia,
Apparently, significant progresses have been made in the development of high speed grinding technologies for advanced ceramics
This paper reviews our developments of the high speed grinding technologies for a wide range
of advanced ceramics The high speed grinding characteristics are compared with those obtained
at the conventional grinding The removal mechanisms associated with high speed grinding processes, either in high efficiency removal mode or ductile removal mode, are discussed Pragmatic high speed grinding technologies are also summarized
High Speed Grinding Characteristics
Grinding Force In the high speed grinding, the measured grinding forces were substantially
influenced by the impact of coolant [32-34] The coolant induced force could be as significant as the grinding induced force [32] Therefore, it is necessary to remove the coolant induced force from the measured force In this work, for each grinding condition used two grinding cycles were undertaken without varying any grinding set-up In the first cycle, the grinding wheel cut the workpiece, so the measured force includes both grinding and coolant induced forces In the second cycle, the wheel only touched the workpiece surface, but without cutting Thus, the measured force was only from the coolant contribution The subtraction of the force measured without cutting from that obtained with cutting gave the grinding induced force [30] Moreover, the forces output from the dynamometer were in horizontal and vertical directions The normal and tangential grinding forces were then calculated by including the effect of the depth of cut using the following equations
perpendicular to the workpiece and the normal force vector on the grinding zone, written as
Table 1 shows the forces obtained from the grinding of five advanced ceramics, which were normalized using the wheel width [26] It is seen that the normal force for all the ceramics
Table 1 Effect of grinding velocity on the normal and tangential grinding forces*
*(1) The force data in the table were normalized by the wheel width and has a unit of N/mm
(2) The wheel depth of cut and the feed rate used were kept constant at 80 µm and 500 mm/min., respectively
subsurface damage was found to be insignificantly affected by the increase in depth of cut [29, 30] Xie et al [31] employed a depth of cut of 6 mm in the high speed grinding of zirconia,
Apparently, significant progresses have been made in the development of high speed grinding technologies for advanced ceramics
This paper reviews our developments of the high speed grinding technologies for a wide range
of advanced ceramics The high speed grinding characteristics are compared with those obtained
at the conventional grinding The removal mechanisms associated with high speed grinding processes, either in high efficiency removal mode or ductile removal mode, are discussed Pragmatic high speed grinding technologies are also summarized
High Speed Grinding Characteristics
Grinding Force In the high speed grinding, the measured grinding forces were substantially
influenced by the impact of coolant [32-34] The coolant induced force could be as significant as the grinding induced force [32] Therefore, it is necessary to remove the coolant induced force from the measured force In this work, for each grinding condition used two grinding cycles were undertaken without varying any grinding set-up In the first cycle, the grinding wheel cut the workpiece, so the measured force includes both grinding and coolant induced forces In the second cycle, the wheel only touched the workpiece surface, but without cutting Thus, the measured force was only from the coolant contribution The subtraction of the force measured without cutting from that obtained with cutting gave the grinding induced force [30] Moreover, the forces output from the dynamometer were in horizontal and vertical directions The normal and tangential grinding forces were then calculated by including the effect of the depth of cut using the following equations
perpendicular to the workpiece and the normal force vector on the grinding zone, written as
Table 1 shows the forces obtained from the grinding of five advanced ceramics, which were normalized using the wheel width [26] It is seen that the normal force for all the ceramics
Table 1 Effect of grinding velocity on the normal and tangential grinding forces*
*(1) The force data in the table were normalized by the wheel width and has a unit of N/mm
(2) The wheel depth of cut and the feed rate used were kept constant at 80 µm and 500 mm/min., respectively
Trang 19Table 2 Microstructures and nominal properties of the ceramics studied [26]
Elastic modulus [GPa]
Hardness [GPa]
Fracture toughness [MPa.m1/2]
G
Alumina-Titania < 20% of TiO2, grain size 2
Zirconia Yttria partially stabilized,
Silicon nitride Reaction bonded, grain size
substantially decreases with the increasing velocity However, the magnitudes of the forces are different for different materials The normal force for zirconia is the greatest, and the force for silicon carbide is the smallest, for the same grinding velocity used It is also noted that the decreasing rate with an increase in wheel velocity is more rapid for zirconia and silicon nitride, particularly at lower velocities The tangential forces for silicon nitride and zirconia slightly decrease with the wheel velocity, which have the higher toughness values than other ceramics studied in this investigation The velocity apparently has smaller effect on the materials having higher brittleness, such as alumina, alumina-titania and silicon carbide The material properties
of the ceramics are also summarized in Table 2
Grinding Temperature High speed grinding of ceramics normally requires high grinding
power It is important to understand if this would lead to high grinding zone temperature The grinding zone temperature can be measured using a grindable thermocouple technique [35-39] In our measurement [40], the end surfaces of two ceramic specimens were first polished
A groove, which has a width of 0.6 mm and a depth of 0.15 mm, was fabricated using laser machining on one of the polished surfaces The surface and the groove were further polished to remove the residual damages generated by the grooving Nickel chrome and nickel silicon foils, also named as K-type thermocouple, were placed in the groove, separated by a mica sheet with a thickness of 0.02 mm, as shown in Fig 1 The two ceramic pieces were then firmly glued together The junction of the thermocouple was formed during grinding, which enabled the detection of the grinding temperature signals The temperature measured using this technique was
in fact the temperature distribution of the workpiece along the wheel-workpiece contact arc The maximum value was taken from the temperature curves after filtering out the high frequency noise and temperature spikes, which was also named as the temperature in the contact zone
Fig 1 (a) Thermocouple assembled into two ceramic workpiece, and (b) a detailed view of the thermocouple which consists of nickel silicon foil, 1, mica, 2, and nickel chrome foil, 3 [40] Fig 2 shows representative grinding temperature curves measured using the grindable thermocouples [40] Fig 2(a) shows the typical temperature curve The temperature along the contact arc has small fluctuation, superposed by high-frequency noises and temperature pikes Fig 2(b) shows the curve in Fig 2(a) after filtering out the high frequency noises and temperature
Hardness [GPa]
Fracture toughness [MPa.m1/2]
G
Alumina-Titania < 20% of TiO2, grain size 2
Zirconia Yttria partially stabilized,
Silicon nitride Reaction bonded, grain size
substantially decreases with the increasing velocity However, the magnitudes of the forces are different for different materials The normal force for zirconia is the greatest, and the force for silicon carbide is the smallest, for the same grinding velocity used It is also noted that the decreasing rate with an increase in wheel velocity is more rapid for zirconia and silicon nitride, particularly at lower velocities The tangential forces for silicon nitride and zirconia slightly decrease with the wheel velocity, which have the higher toughness values than other ceramics studied in this investigation The velocity apparently has smaller effect on the materials having higher brittleness, such as alumina, alumina-titania and silicon carbide The material properties
of the ceramics are also summarized in Table 2
Grinding Temperature High speed grinding of ceramics normally requires high grinding
power It is important to understand if this would lead to high grinding zone temperature The grinding zone temperature can be measured using a grindable thermocouple technique [35-39] In our measurement [40], the end surfaces of two ceramic specimens were first polished
A groove, which has a width of 0.6 mm and a depth of 0.15 mm, was fabricated using laser machining on one of the polished surfaces The surface and the groove were further polished to remove the residual damages generated by the grooving Nickel chrome and nickel silicon foils, also named as K-type thermocouple, were placed in the groove, separated by a mica sheet with a thickness of 0.02 mm, as shown in Fig 1 The two ceramic pieces were then firmly glued together The junction of the thermocouple was formed during grinding, which enabled the detection of the grinding temperature signals The temperature measured using this technique was
in fact the temperature distribution of the workpiece along the wheel-workpiece contact arc The maximum value was taken from the temperature curves after filtering out the high frequency noise and temperature spikes, which was also named as the temperature in the contact zone
Fig 1 (a) Thermocouple assembled into two ceramic workpiece, and (b) a detailed view of the thermocouple which consists of nickel silicon foil, 1, mica, 2, and nickel chrome foil, 3 [40] Fig 2 shows representative grinding temperature curves measured using the grindable thermocouples [40] Fig 2(a) shows the typical temperature curve The temperature along the contact arc has small fluctuation, superposed by high-frequency noises and temperature pikes Fig 2(b) shows the curve in Fig 2(a) after filtering out the high frequency noises and temperature
3
Trang 20spikes The curve appears to be consistent with the low-frequency temperature component in Fig 2(a), indicating that the filtering didn't distort the signal of the average temperature Similar results were reported in [35] Fig 2(c) is the temperature curve obtained at the highest wheel velocity of
(from Point A to Point C) After passing Point C, the grinding temperature rapidly increased to
to the insufficient coolant supply caused by the employment of the extremely high wheel velocity [40] The enhanced resistance of the air barrier formed around the wheel periphery at higher wheel velocity to the coolant supply was mainly responsible for the rapid temperature rise The grinding temperatures were measured in the high speed grinding of zirconia and silicon nitride The
undertaken at the wheel velocity of 160 m/s The grinding temperatures measured using the grindable thermocouples are summarized in Table 3
Table 3 Grinding temperatures measured at different wheel velocities*
*The workpiece feed rate and grinding depth of cut were fixed at 2400 mm/min and 1 mm, respectively
Ground Surface Characteristics Table 4 shows the effect of wheel speed on the surface
roughness [26] For the ceramics with relatively high brittleness, such as alumina, alumina-titania and silicon carbide, the increased wheel speed has slightly improved the surface roughness For those ceramics with relatively high toughness, such as zirconia and silicon nitride, the roughness exhibited to be slightly increased with the increasing wheel velocity The relatively great effect from the machine vibration caused by the high wheel speed might cause surface deterioration In the high speed grinding process, the increased wheel velocity has resulted in a smaller undeformed chip thickness, but a greater vibration magnitude For those ceramics with relatively higher toughness, the effect of vibration on the surface roughness appeared to be more dominant than the effect from the reduction in undeformed chip thickness Nevertheless, for all the ceramic materials, the surface roughness wasn’t significantly influenced by the wheel velocity under the grinding condition used Fig 3 shows the surface topographies of two representative ceramics obtained from the lowest and highest wheel velocities
spikes The curve appears to be consistent with the low-frequency temperature component in Fig 2(a), indicating that the filtering didn't distort the signal of the average temperature Similar results were reported in [35] Fig 2(c) is the temperature curve obtained at the highest wheel velocity of
(from Point A to Point C) After passing Point C, the grinding temperature rapidly increased to
to the insufficient coolant supply caused by the employment of the extremely high wheel velocity [40] The enhanced resistance of the air barrier formed around the wheel periphery at higher wheel velocity to the coolant supply was mainly responsible for the rapid temperature rise The grinding temperatures were measured in the high speed grinding of zirconia and silicon nitride The
undertaken at the wheel velocity of 160 m/s The grinding temperatures measured using the grindable thermocouples are summarized in Table 3
Table 3 Grinding temperatures measured at different wheel velocities*
*The workpiece feed rate and grinding depth of cut were fixed at 2400 mm/min and 1 mm, respectively
Ground Surface Characteristics Table 4 shows the effect of wheel speed on the surface
roughness [26] For the ceramics with relatively high brittleness, such as alumina, alumina-titania and silicon carbide, the increased wheel speed has slightly improved the surface roughness For those ceramics with relatively high toughness, such as zirconia and silicon nitride, the roughness exhibited to be slightly increased with the increasing wheel velocity The relatively great effect from the machine vibration caused by the high wheel speed might cause surface deterioration In the high speed grinding process, the increased wheel velocity has resulted in a smaller undeformed chip thickness, but a greater vibration magnitude For those ceramics with relatively higher toughness, the effect of vibration on the surface roughness appeared to be more dominant than the effect from the reduction in undeformed chip thickness Nevertheless, for all the ceramic materials, the surface roughness wasn’t significantly influenced by the wheel velocity under the grinding condition used Fig 3 shows the surface topographies of two representative ceramics obtained from the lowest and highest wheel velocities
Trang 21Table 4 Surface roughness of the ground ceramics (unit is in µm)*
*The workpiece feed rate and grinding depth of cut were fixed at 500 mm/min and 0.5 mm, respectively
Fig 3 SEM micrographs of ground surfaces of zirconia (a, b) and silicon carbide (c, d) [26] Graph (a) and (c) are the ground surfaces obtained at the wheel velocity of 40 m/s and
Graph (b) and (d) are those at obtained at the wheel velocity of 160 m/s
Subsurface Damage A bonded interface sectioning technique [41] was adopted to examine the
grinding induced subsurface damage in our investigations In this method, two specimens were first polished and then bonded together using a cyanoacrylate-based adhesive Clamping pressure was applied during bonding to ensure that a thin adhesive layer joint was achieved which would minimize edge chipping during grinding The grinding was completed using various wheel velocities ranged from 40 to 160 m/s, and different depths of cut, varied from 0.1 to 2 mm, while the table feed rate was maintained unchanged at 500 mm/min The grinding direction was perpendicular to the bonded interface After grinding the bonded specimens were subsequently separated by heating on a hot plate to soften the adhesive The separated specimens were cleaned with acetone in an ultrasonic bath and then gold coated for SEM examination
In the high speed grinding of advanced ceramics, including alumina, alumina-titania and partially stabilized zirconia, an increase in depth of cut did not deepen the subsurface damage layer for the alumina and alumina-titania, but resulted in a slightly deeper damage layer in the zirconia [30] As shown in Table 5, the thickness of the damage layer was below 20 µm for all ground specimens Subsurface characteristics of the ceramics showed that the material removal mechanisms associated with the high speed grinding were influenced by their microstructures and mechanical properties [30] For grinding the alumina and alumina-titania, relatively brittle materials, grain dislodgement was the dominant material removal mode, as shown in Fig 4(a) For grinding zirconia the removal mechanisms included both brittle fracture (Fig 4(b)) and ductile cutting (Fig 3(b))
*The workpiece feed rate and grinding depth of cut were fixed at 500 mm/min and 0.5 mm, respectively
Fig 3 SEM micrographs of ground surfaces of zirconia (a, b) and silicon carbide (c, d) [26] Graph (a) and (c) are the ground surfaces obtained at the wheel velocity of 40 m/s and
Graph (b) and (d) are those at obtained at the wheel velocity of 160 m/s
Subsurface Damage A bonded interface sectioning technique [41] was adopted to examine the
grinding induced subsurface damage in our investigations In this method, two specimens were first polished and then bonded together using a cyanoacrylate-based adhesive Clamping pressure was applied during bonding to ensure that a thin adhesive layer joint was achieved which would minimize edge chipping during grinding The grinding was completed using various wheel velocities ranged from 40 to 160 m/s, and different depths of cut, varied from 0.1 to 2 mm, while the table feed rate was maintained unchanged at 500 mm/min The grinding direction was perpendicular to the bonded interface After grinding the bonded specimens were subsequently separated by heating on a hot plate to soften the adhesive The separated specimens were cleaned with acetone in an ultrasonic bath and then gold coated for SEM examination
In the high speed grinding of advanced ceramics, including alumina, alumina-titania and partially stabilized zirconia, an increase in depth of cut did not deepen the subsurface damage layer for the alumina and alumina-titania, but resulted in a slightly deeper damage layer in the zirconia [30] As shown in Table 5, the thickness of the damage layer was below 20 µm for all ground specimens Subsurface characteristics of the ceramics showed that the material removal mechanisms associated with the high speed grinding were influenced by their microstructures and mechanical properties [30] For grinding the alumina and alumina-titania, relatively brittle materials, grain dislodgement was the dominant material removal mode, as shown in Fig 4(a) For grinding zirconia the removal mechanisms included both brittle fracture (Fig 4(b)) and ductile cutting (Fig 3(b))
20 µµµm
Trang 22Table 5 Depths of subsurface damage layers of the ground ceramics (unit in µm)*
*The grinding depth of cut and workpiece feed rate were 160 m/s and 500 mm/min., respectively
Fig 4 Subsurface damage layers of (a) alumina and (b) zirconia [30] White arrows indicate microcracks along grain-boundaries in (a) and lateral cracks in (b) The black arrow in (a) refers to a grain dislodgement
Grindability of Ceramics Previous studies [e.g 10, 20, 30, 42] have shown that the normal
grinding force is a useful indicator of the grinding performance for brittle materials In the grinding
of ceramics, if the removal is in the brittle regime, usually via grain dislodgement for polycrystalline ceramic materials, the normal force may be represented by the following empirical formula [10],
1/ 2 2 / 5
n
T H E (= G) [26] Fig 5 shows that the
relationship between the normal grinding forces and the ceramic material properties in conventional
more ductility during a grinding process, or the material is more grindable
Pragmatic Technologies for High speed Grinding
The advantages of high speed grinding can only be utilized in an effective and premeditated manner if the machine concepts are adapted to the requirements of this high performance grinding technology [16] When operating at relatively high rotational speed, the grinding wheel/spindle/motor system must run extremely accurately and with minimum vibrations For this reason, a high level of rigidity is required for the entire machine system In addition, precise balancing of the grinding wheel is essential In that case, a suitable stabilizing system in terms of balancing capacity and balancing quality is required Also, when working with high grinding speeds, the coolant supply has to be taken into account Sufficient coolant for the cooling and the lubrication of the grinding process is required In this section, the pragmatic technologies for high speed grinding are presented
*The grinding depth of cut and workpiece feed rate were 160 m/s and 500 mm/min., respectively
Fig 4 Subsurface damage layers of (a) alumina and (b) zirconia [30] White arrows indicate microcracks along grain-boundaries in (a) and lateral cracks in (b) The black arrow in (a) refers to a grain dislodgement
Grindability of Ceramics Previous studies [e.g 10, 20, 30, 42] have shown that the normal
grinding force is a useful indicator of the grinding performance for brittle materials In the grinding
of ceramics, if the removal is in the brittle regime, usually via grain dislodgement for polycrystalline ceramic materials, the normal force may be represented by the following empirical formula [10],
1/ 2 2 / 5
n
T H E (= G) [26] Fig 5 shows that the
relationship between the normal grinding forces and the ceramic material properties in conventional
more ductility during a grinding process, or the material is more grindable
Pragmatic Technologies for High speed Grinding
The advantages of high speed grinding can only be utilized in an effective and premeditated manner if the machine concepts are adapted to the requirements of this high performance grinding technology [16] When operating at relatively high rotational speed, the grinding wheel/spindle/motor system must run extremely accurately and with minimum vibrations For this reason, a high level of rigidity is required for the entire machine system In addition, precise balancing of the grinding wheel is essential In that case, a suitable stabilizing system in terms of balancing capacity and balancing quality is required Also, when working with high grinding speeds, the coolant supply has to be taken into account Sufficient coolant for the cooling and the lubrication of the grinding process is required In this section, the pragmatic technologies for high speed grinding are presented
Trang 230 2 4 6 8
Machine Tool Requirements As mentioned earlier, a high level of rigidity is required for the
grinding wheel/spindle/motor system The spindle idle power of a high speed machine tool usually has to be sufficiently large The machine tools used in our investigations included the Okamoto Precision Surface Grinder (Okamoto Model ACC-63DXNC) and a super-high speed surface grinder developed by the National Engineering Research Center for High Efficiency Grinding in China The former is equipped with a built-in type motor of 15 kW, capable of operating at 20,000 rpm for wheels of 200 mm in diameter or a maximum wheel speed of 210 m/s The latter has a spindle idle power of 40 kW, which enables to run a wheel of 350 mm at the highest velocity of
314 m/s This grinding machine is also equipped with an on-machine dynamic balancer (SBS4500, Schmitt Industries)
Wheel Preparation The wheels for high speed grinding of the advanced ceramics used
diamond abrasives The wheel rims were made of metal alloys The used bonding methods included electroplated, vitrified, metal and resin bonds All types of the wheels needed to be trued and dressed, with exception for the electroplated wheels Electroplated wheels only required 'cleaning' or 'touching up' with an abrasive stick While the primary aim in truing was to produce the required macroscopic wheel shape, the truing process also influenced the microscopic wheel topography Likewise, the dressing process might decrease wheel run-out, thereby actually truing as well as conditioning the wheel The wheels were trued to achieve certain roundness before going for dressing The truing technique used was the motor driven truing using conventional abrasives and steel rollers For rough grinding, the grinding wheel was trued to reach a roundness error of below 10 µm [43] For fine grinding, the roundness error was reduced to below 2 µm [26] After truing, the wheel was dressed using an alumina stick to expose the abrasive grits Different truing rollers had different truing/dressing intensities This helped to achieve different abrasive topographies for specific grinding processes Fig 6 shows the comparison of wheel topographies prepared using two different types of truing rollers [43] The grains with micro-cutting edges dressed by the SiC roller (in Fig 6a) were more frequently observed than that dressed by the steel roller For the wheel dressed using the steel roller, larger cutting edges, as shown in Fig 6b, were often observed
Wheel Balancing It is important to maintain an excellent dynamic balancing of the grinding
wheel used for high speed grinding The vibration amplitude in Fig 7 [25] was measured at the top
of the wheel spindle of the Okamoto grinder The wheel balanced at 40 m/s had relatively low vibration amplitudes at the balanced point and at the wheel speed of 80 m/s However, the vibration amplitude was tripled at 120 m/s and was almost ten times higher at 160 m/s The increased vibration level at higher speeds limited the improvement of surface quality [25] The effect of wheel imbalance on the surface roughness clearly becomes more influential at higher speeds To attenuate the vibration at higher wheel speeds, the wheel had to be balanced at the respective
0 2 4 6 8
Machine Tool Requirements As mentioned earlier, a high level of rigidity is required for the
grinding wheel/spindle/motor system The spindle idle power of a high speed machine tool usually has to be sufficiently large The machine tools used in our investigations included the Okamoto Precision Surface Grinder (Okamoto Model ACC-63DXNC) and a super-high speed surface grinder developed by the National Engineering Research Center for High Efficiency Grinding in China The former is equipped with a built-in type motor of 15 kW, capable of operating at 20,000 rpm for wheels of 200 mm in diameter or a maximum wheel speed of 210 m/s The latter has a spindle idle power of 40 kW, which enables to run a wheel of 350 mm at the highest velocity of
314 m/s This grinding machine is also equipped with an on-machine dynamic balancer (SBS4500, Schmitt Industries)
Wheel Preparation The wheels for high speed grinding of the advanced ceramics used
diamond abrasives The wheel rims were made of metal alloys The used bonding methods included electroplated, vitrified, metal and resin bonds All types of the wheels needed to be trued and dressed, with exception for the electroplated wheels Electroplated wheels only required 'cleaning' or 'touching up' with an abrasive stick While the primary aim in truing was to produce the required macroscopic wheel shape, the truing process also influenced the microscopic wheel topography Likewise, the dressing process might decrease wheel run-out, thereby actually truing as well as conditioning the wheel The wheels were trued to achieve certain roundness before going for dressing The truing technique used was the motor driven truing using conventional abrasives and steel rollers For rough grinding, the grinding wheel was trued to reach a roundness error of below 10 µm [43] For fine grinding, the roundness error was reduced to below 2 µm [26] After truing, the wheel was dressed using an alumina stick to expose the abrasive grits Different truing rollers had different truing/dressing intensities This helped to achieve different abrasive topographies for specific grinding processes Fig 6 shows the comparison of wheel topographies prepared using two different types of truing rollers [43] The grains with micro-cutting edges dressed by the SiC roller (in Fig 6a) were more frequently observed than that dressed by the steel roller For the wheel dressed using the steel roller, larger cutting edges, as shown in Fig 6b, were often observed
Wheel Balancing It is important to maintain an excellent dynamic balancing of the grinding
wheel used for high speed grinding The vibration amplitude in Fig 7 [25] was measured at the top
of the wheel spindle of the Okamoto grinder The wheel balanced at 40 m/s had relatively low vibration amplitudes at the balanced point and at the wheel speed of 80 m/s However, the vibration amplitude was tripled at 120 m/s and was almost ten times higher at 160 m/s The increased vibration level at higher speeds limited the improvement of surface quality [25] The effect of wheel imbalance on the surface roughness clearly becomes more influential at higher speeds To attenuate the vibration at higher wheel speeds, the wheel had to be balanced at the respective
Trang 24grinding speeds Using this approach, the surface roughness was decreased with the increasing wheel speed, indicating that the application of high speed did improve the surface quality [25]
Fig 6 SEM micrographs of a diamond abrasive wheel of mesh size of 120/140 trued by (a) SiC roller (roller/wheel ratio=0.3, truing depth of cut=20 µm, and roller feed rate=200 mm/min.) and (b) Steel roller
(roller/wheel ratio=0.3, truing depth of cut=2 µm, and roller feed rate=100 mm/min.) [43]
0 0.2 0.4 0.6 0.8 1
Fig 7 Spindle vibration obtained at various wheel velocities [25]
Coolant Supply Water-based coolant was supplied using a specially designed shoe nozzle [44]
seal 3 and the nozzle body 4 The working principles of the nozzle are described as follows (i) The momentum of coolant from the upper orifice counteracts with the air surrounding the wheel, which blocks the air circulating in the grinding zone and allows the coolant to enter the enclosed cavity Additionally, the coolant from the upper orifice can clean out the wheel as well (ii) Coolant from the lower orifice is mainly sprayed into the grinding zone, functioning like the conventional coolant supply (iii) There is a gap between the wheel and the lower orifice, which acts as the third orifice, so coolant in the enclosed cavity can flow out from this orifice, whose direction can be altered to the grinding zone by adjustable blocker 6 There are two advantages for the use of the shoe-type nozzle [45] First, it brought the coolant closer to the grinding zone Second, the flow speed at the outlet of' the nozzle was 4 times increased, compared to a normal flat-type nozzle The higher speed flow more easily penetrated into the contact zone, reducing the thermal damage Therefore, it is not necessary to increase the coolant pump pressure However, it is believed that there is an optimal flow rate or speed for the high speed grinding system, because a very high flow speed may cause turbulent flow, which allows more air mixed into the moving coolant [17] The ability of the coolant to dissipate heat from the grinding zone is thus compromised by the poor thermal conductivity of the air-coolant mixture [32] In
Fig 7 Spindle vibration obtained at various wheel velocities [25]
Coolant Supply Water-based coolant was supplied using a specially designed shoe nozzle [44]
seal 3 and the nozzle body 4 The working principles of the nozzle are described as follows (i) The momentum of coolant from the upper orifice counteracts with the air surrounding the wheel, which blocks the air circulating in the grinding zone and allows the coolant to enter the enclosed cavity Additionally, the coolant from the upper orifice can clean out the wheel as well (ii) Coolant from the lower orifice is mainly sprayed into the grinding zone, functioning like the conventional coolant supply (iii) There is a gap between the wheel and the lower orifice, which acts as the third orifice, so coolant in the enclosed cavity can flow out from this orifice, whose direction can be altered to the grinding zone by adjustable blocker 6 There are two advantages for the use of the shoe-type nozzle [45] First, it brought the coolant closer to the grinding zone Second, the flow speed at the outlet of' the nozzle was 4 times increased, compared to a normal flat-type nozzle The higher speed flow more easily penetrated into the contact zone, reducing the thermal damage Therefore, it is not necessary to increase the coolant pump pressure However, it is believed that there is an optimal flow rate or speed for the high speed grinding system, because a very high flow speed may cause turbulent flow, which allows more air mixed into the moving coolant [17] The ability of the coolant to dissipate heat from the grinding zone is thus compromised by the poor thermal conductivity of the air-coolant mixture [32] In
Trang 25addition, a higher speed flow would rebound more quickly, thus obstructing the further supply of the coolant
When using high speed coolant, a coolant prop maybe installed in front of workpiece to guide the coolant into the grinding zone at the beginning stage of grinding [26] Without the coolant prop, the coolant was blocked by the front end of the work-piece and rebounded back, the front surface of the workpiece could be thermally damaged
Fig 8 Schematic illustration of the closed Y-type nozzle system, which consists of grinding wheel, 1, steel
cover, 2, rubber seal, 3, nozzle body, 4, hinge, 5, and adjustable blocker, 6 [44]
Mounting of Ceramic Workpieces It is worthy to point out that the mounting of the
ceramic workpieces could significantly influence the grinding quality in high speed grinding When mechanically holding a workpiece, it is very useful to put a thin copper or aluminum foil between the contact surfaces of the workpiece and the mechanical vice So the vibration induced by grinding could be attenuated by the soft metal layer, not forming a direct impact on the brittle workpiece This could avoid the catastrophic failure of workpieces, especially when grinding at super high speed where grinding-induced vibration was unavoidable
Concluding Remarks
This work demonstrated that the application of high speed grinding into the machining of advanced ceramics could either improve the ground surface quality or increase the machining efficiency This was because the elevation of wheel velocity has reduced the undeformed chip thickness in the material removal process, thus resulted in a decreased normal grinding force
considerably high, provided that the coolant supply was sufficiently effective Nevertheless, to realize the potential of the high speed grinding technology, great care had to be taken on the grinding-induced vibration The reduction of the grinding-induced vibration required a highly rigid machine, appropriate preparation of grinding wheels and accurate dynamic balancing of the wheels used Coolant supply was another key factor in the high speed grinding of advanced ceramics It was extremely important that the coolant could effectively enter the grinding zone when grinding advanced ceramics at relatively high speeds
Acknowledgements
The author is grateful to the experimental assistance and valuable discussion from L Yin, K Ramesh, L Zhou, Y Liu, P.L Teo, J Goh, Z.T Shang, G.Z Xie, H.Q Mi and X.M Sheng The author also wishes to acknowledge the financial support from the Australia Research Council under Discovery Project Program, Chinese Education Ministry under Key-Project Scheme and Singapore Institute of Manufacturing Technology under In-house Project Scheme
addition, a higher speed flow would rebound more quickly, thus obstructing the further supply of the coolant
When using high speed coolant, a coolant prop maybe installed in front of workpiece to guide the coolant into the grinding zone at the beginning stage of grinding [26] Without the coolant prop, the coolant was blocked by the front end of the work-piece and rebounded back, the front surface of the workpiece could be thermally damaged
Fig 8 Schematic illustration of the closed Y-type nozzle system, which consists of grinding wheel, 1, steel
cover, 2, rubber seal, 3, nozzle body, 4, hinge, 5, and adjustable blocker, 6 [44]
Mounting of Ceramic Workpieces It is worthy to point out that the mounting of the
ceramic workpieces could significantly influence the grinding quality in high speed grinding When mechanically holding a workpiece, it is very useful to put a thin copper or aluminum foil between the contact surfaces of the workpiece and the mechanical vice So the vibration induced by grinding could be attenuated by the soft metal layer, not forming a direct impact on the brittle workpiece This could avoid the catastrophic failure of workpieces, especially when grinding at super high speed where grinding-induced vibration was unavoidable
Concluding Remarks
This work demonstrated that the application of high speed grinding into the machining of advanced ceramics could either improve the ground surface quality or increase the machining efficiency This was because the elevation of wheel velocity has reduced the undeformed chip thickness in the material removal process, thus resulted in a decreased normal grinding force
considerably high, provided that the coolant supply was sufficiently effective Nevertheless, to realize the potential of the high speed grinding technology, great care had to be taken on the grinding-induced vibration The reduction of the grinding-induced vibration required a highly rigid machine, appropriate preparation of grinding wheels and accurate dynamic balancing of the wheels used Coolant supply was another key factor in the high speed grinding of advanced ceramics It was extremely important that the coolant could effectively enter the grinding zone when grinding advanced ceramics at relatively high speeds
Acknowledgements
The author is grateful to the experimental assistance and valuable discussion from L Yin, K Ramesh, L Zhou, Y Liu, P.L Teo, J Goh, Z.T Shang, G.Z Xie, H.Q Mi and X.M Sheng The author also wishes to acknowledge the financial support from the Australia Research Council under Discovery Project Program, Chinese Education Ministry under Key-Project Scheme and Singapore Institute of Manufacturing Technology under In-house Project Scheme
Trang 26References
Marcel Dekker, 1999
Noyes Publications/William Andrew Publishing LLC, New York, 2000
[3] I Inasaki: Grinding of Hard and Brittle Materials, Annals of CIRP, Vol 36 (1987),
[6] H.H.K Xu, and S Jahanmir: Microstructure and Material Removal in Scratching of Alumina, Journal of Materials Science, Vol.30 (1995), pp.2335-2247
[7] T.W Hwang and S Malkin: Grinding Mechanisms and Energy Balance for Ceramics, Transaction of ASME: Journal of Manufacturing Science and Engineering, Vol.121 (1999), pp.623-631
[8] S Kohli, C Guo and S Malkin: Energy Partition to the Workpiece for Grinding with Aluminum Oxide and CBN Abrasive Wheels, Transactions of the ASME: Journal of Engineering for Industry, Vol.117 (1995), pp.160-168
[9] T.G Bifano, T.A Dow and R.O Scattergood: Ductile-regime: a New Technology for Machining Brittle Materials, Transaction of ASME: Journal of Engineering for Industry, Vol.113 (1991), pp.184-189
[10] K.L B1aedel, I.S Taylor and C.J Evans: Ductile-regime Grinding of Brittle Materials, in Machining of Ceramics and Composites, ed S Jahanmir, M Ramulu and P Koshy, New York: Marcel Dekker, 1999, pp.139-176
[11] A.G Evans and D.B Marshall: Wear Mechanisms in Ceramics, Fundamental of Friction and Wear of Materials, ed D.A Rigney, Metals Park, Ohio, American Society for Metals,
Diamond-[15] S Jahanmir, H.H.K XU and L.K Ives: Mechanisms of Materials Removal in Abrasive Machining, in Machining of Ceramics and Composites, eds S Jahanmir, M Ramulu, P Koshy, New York: Marcel Dekker, 1999, pp 11 -84
Webster and D Stuff: High Speed Grinding - Fundamentals and State of the Art in Europe, Japan and the USA, Annals of CIRP, Vol.46 (1997), pp.715-724
[17] F Klocke E Verlemann and C Schippers: High-Speed Grinding of Ceramics, in Machining
of Ceramics and Composites, ed S Jahanmir, M Ranlulu, P Koshy, New York: Marcel Dekker, 1999, pp 119-138
Characteristics in High Speed Grinding, PhD Thesis University of Connecticut, 2000
References
Marcel Dekker, 1999
Noyes Publications/William Andrew Publishing LLC, New York, 2000
[3] I Inasaki: Grinding of Hard and Brittle Materials, Annals of CIRP, Vol 36 (1987),
[6] H.H.K Xu, and S Jahanmir: Microstructure and Material Removal in Scratching of Alumina, Journal of Materials Science, Vol.30 (1995), pp.2335-2247
[7] T.W Hwang and S Malkin: Grinding Mechanisms and Energy Balance for Ceramics, Transaction of ASME: Journal of Manufacturing Science and Engineering, Vol.121 (1999), pp.623-631
[8] S Kohli, C Guo and S Malkin: Energy Partition to the Workpiece for Grinding with Aluminum Oxide and CBN Abrasive Wheels, Transactions of the ASME: Journal of Engineering for Industry, Vol.117 (1995), pp.160-168
[9] T.G Bifano, T.A Dow and R.O Scattergood: Ductile-regime: a New Technology for Machining Brittle Materials, Transaction of ASME: Journal of Engineering for Industry, Vol.113 (1991), pp.184-189
[10] K.L B1aedel, I.S Taylor and C.J Evans: Ductile-regime Grinding of Brittle Materials, in Machining of Ceramics and Composites, ed S Jahanmir, M Ramulu and P Koshy, New York: Marcel Dekker, 1999, pp.139-176
[11] A.G Evans and D.B Marshall: Wear Mechanisms in Ceramics, Fundamental of Friction and Wear of Materials, ed D.A Rigney, Metals Park, Ohio, American Society for Metals,
Diamond-[15] S Jahanmir, H.H.K XU and L.K Ives: Mechanisms of Materials Removal in Abrasive Machining, in Machining of Ceramics and Composites, eds S Jahanmir, M Ramulu, P Koshy, New York: Marcel Dekker, 1999, pp 11 -84
Webster and D Stuff: High Speed Grinding - Fundamentals and State of the Art in Europe, Japan and the USA, Annals of CIRP, Vol.46 (1997), pp.715-724
[17] F Klocke E Verlemann and C Schippers: High-Speed Grinding of Ceramics, in Machining
of Ceramics and Composites, ed S Jahanmir, M Ranlulu, P Koshy, New York: Marcel Dekker, 1999, pp 119-138
Characteristics in High Speed Grinding, PhD Thesis University of Connecticut, 2000
Trang 27[19] Inasaki High Efficiency Grinding of Advanced ceramics, Annals of CIRP, Vol.35 (1986), pp.211-214
Electroplated Diamond Wheels, Annals of CIRP, Vol.49 (2000), pp.245-248
[21] T.W Hwang, C.J Evans and S Malkin: High Speed Grinding of Silicon Nitride with Electroplated Diamond Wheels, Part 2: Wheel Topography and Grinding Mechanisms, Transaction of ASME: Journal of Manufacturing Science and Engineering, Vol.122 (2000), pp.42-50
[22] K Ramesh, S.H Yeo, S Gowri and L Zhou: Experimental Evaluation of Super High Speed Grinding of Advanced Ceramics, Journal of Advanced Manufacturing Technology, Vol.17 (2001), pp.87-92
Vitrified Diamond Wheel, International Journal of Japan Society for Precision Engineering, Vol.28 (1994), pp.344-345
[24] L Yin and H Huang: Ceramic Response to High Speed Grinding, Machining Science and Technology, Vol 8 (2004), pp.21-37
[25] H Huang, L Yin and L Zhou: High Speed Grinding of Silicon Nitride with Resin Bond Diamond Wheels, Journal of Materials Processing Technology, Vol.141 (2003), pp.329-336 [26] H Huang and L Yin: Grinding Characteristics of Engineering Ceramics in High Speed Regime, International Journal of Abrasive Technology, Vol.1 (2007), pp.78-93
[27] L Yin, H Huang, K Ramesh and T Huang: High Speed Versus Conventional Grinding in High Removal Rate Machining of Alumina and Alumina-Titania, International Journal of Machine Tools and Manufacture, Vol.45 (2005), pp.897-907
[28] J.A Kovach, S Srinivasan, P.J Blau B Bandyopadhyay, S Malkin and K Ziegler: A
Engineers, Cincinnati, Ohio, 1993, MR93-352
[29] H Huang, Machining Characteristics and Surface Integrity of Yttria Stabilized Tetragonal Zirconia in High Speed Deep Grinding, Materials Science and Engineering A, Vol.345 (2003), pp.155-163
[30] H Huang and Y.C Liu Experimental Investigations of Machining Characteristics and Removal Mechanisms of Advanced Ceramics in High Speed Deep Grinding, International Journal of Machine Tools & Manufacture, Vol.43 (2003), pp.811-823
[31] G.Z Xie, H.W Huang, H Huang, X.M Sheng, H.Q Mi and W Xiong: Experimental Investigations of Advanced Ceramics in High Efficiency Deep Grinding, Chinese Journal of Mechanical Engineering, Vol.43 (2007), pp.176-184
[32] F Klocke and A Baus: Coolant Induced Forces in CBN High Speed Grinding with Shoe Nozzles, Annals of CIRP, Vol.49 (2000), pp.241-244
[33] H Huang, S Kanno, X.D Liu and Z.M Gong: Highly Integrated and Automated High Speed Grinding System for Printer Heads Constructed by Combination Materials, International Journal of Advanced Manufacturing Technology, Vol.25 (2005), pp.1-9 [34] K Ramesh, H Huang and L Yin: Analytical and Experimental Investigation of Coolant Velocity in High Speed Grinding, International Journal of Machine Tools and Manufacture, Vol.44 (2004), pp.1069-1076
[35] A.D Batako, W.B Rowe and M.N Morgan: Temperature Measurement in High Efficiency Deep Grinding, International Journal of Machine Tools & Manufacture, Vol 45 (2005), pp.1231-1245
[19] Inasaki High Efficiency Grinding of Advanced ceramics, Annals of CIRP, Vol.35 (1986), pp.211-214
Electroplated Diamond Wheels, Annals of CIRP, Vol.49 (2000), pp.245-248
[21] T.W Hwang, C.J Evans and S Malkin: High Speed Grinding of Silicon Nitride with Electroplated Diamond Wheels, Part 2: Wheel Topography and Grinding Mechanisms, Transaction of ASME: Journal of Manufacturing Science and Engineering, Vol.122 (2000), pp.42-50
[22] K Ramesh, S.H Yeo, S Gowri and L Zhou: Experimental Evaluation of Super High Speed Grinding of Advanced Ceramics, Journal of Advanced Manufacturing Technology, Vol.17 (2001), pp.87-92
Vitrified Diamond Wheel, International Journal of Japan Society for Precision Engineering, Vol.28 (1994), pp.344-345
[24] L Yin and H Huang: Ceramic Response to High Speed Grinding, Machining Science and Technology, Vol 8 (2004), pp.21-37
[25] H Huang, L Yin and L Zhou: High Speed Grinding of Silicon Nitride with Resin Bond Diamond Wheels, Journal of Materials Processing Technology, Vol.141 (2003), pp.329-336 [26] H Huang and L Yin: Grinding Characteristics of Engineering Ceramics in High Speed Regime, International Journal of Abrasive Technology, Vol.1 (2007), pp.78-93
[27] L Yin, H Huang, K Ramesh and T Huang: High Speed Versus Conventional Grinding in High Removal Rate Machining of Alumina and Alumina-Titania, International Journal of Machine Tools and Manufacture, Vol.45 (2005), pp.897-907
[28] J.A Kovach, S Srinivasan, P.J Blau B Bandyopadhyay, S Malkin and K Ziegler: A
Engineers, Cincinnati, Ohio, 1993, MR93-352
[29] H Huang, Machining Characteristics and Surface Integrity of Yttria Stabilized Tetragonal Zirconia in High Speed Deep Grinding, Materials Science and Engineering A, Vol.345 (2003), pp.155-163
[30] H Huang and Y.C Liu Experimental Investigations of Machining Characteristics and Removal Mechanisms of Advanced Ceramics in High Speed Deep Grinding, International Journal of Machine Tools & Manufacture, Vol.43 (2003), pp.811-823
[31] G.Z Xie, H.W Huang, H Huang, X.M Sheng, H.Q Mi and W Xiong: Experimental Investigations of Advanced Ceramics in High Efficiency Deep Grinding, Chinese Journal of Mechanical Engineering, Vol.43 (2007), pp.176-184
[32] F Klocke and A Baus: Coolant Induced Forces in CBN High Speed Grinding with Shoe Nozzles, Annals of CIRP, Vol.49 (2000), pp.241-244
[33] H Huang, S Kanno, X.D Liu and Z.M Gong: Highly Integrated and Automated High Speed Grinding System for Printer Heads Constructed by Combination Materials, International Journal of Advanced Manufacturing Technology, Vol.25 (2005), pp.1-9 [34] K Ramesh, H Huang and L Yin: Analytical and Experimental Investigation of Coolant Velocity in High Speed Grinding, International Journal of Machine Tools and Manufacture, Vol.44 (2004), pp.1069-1076
[35] A.D Batako, W.B Rowe and M.N Morgan: Temperature Measurement in High Efficiency Deep Grinding, International Journal of Machine Tools & Manufacture, Vol 45 (2005), pp.1231-1245
Trang 28[36] W.B Rowe: Thermal Analysis of High Efficiency Deep Grinding, International Journal of Machine Tools & Manufacture, Vol 41 (2001), pp.1-19
[37] W.B Rowe and T Jin: Temperatures in High Efficiency Deep Grinding, Annals of the CIRP, Vol.50 (2001), pp.205-208
[38] T Jin, W.B Rowe and D McCormack: Temperatures in Deep Grinding of Finite Workpieces, International Journal of Machine Tools & Manufacture Vol 42 (2002), pp.53 -59
[39] T Jin and G.Q Cai, Analytical Thermal Model of Oblique Moving Heat Source for Deep Grinding and Cutting, Transaction of ASME, Journal of Manufacturing Science and Engineering, Vol.123 (2001), pp.185-190
[40] G.Z Xie and H Huang: An Experimental Investigation of Temperature in High Speed Deep Grinding of Partially Stabilized Zirconia, International Journal of Machine Tool and Manufacture, (2008) doi:10.1016/j.ijmachtools.2008.06.002
[41] H.H.K Xu and S Jahanmir, Simple technique for observing subsurface damage in machining
of ceramics, Journal of American Ceramic Society, 77 (1994), 1388-1390
[42] A.G Evans and D.B Marshall: Wear Mechanisms in Ceramics, in Fundamental of Friction and Wear of Materials, ed D.A Rigney, Metals Park, Ohio: American Society for Metals,
1981, pp.439–452
[43] H Huang: Effect of Truing/Dressing Intensity on Truing/Dressing Efficiency and Grinding Performance of Vitrified Diamond Wheels, Journal of Materials Processing Technology, Vol.117 (2001), pp.9-14
[44] H Huang, Z.T Shang, H.Q Mi, X.M Sheng, S.Q Wang, Y Wu and G.Z Xie: The Closed type Nozzle for Ultrahigh Speed Grinding China Patent, CN200520052869.7
Y-[45] Z.T Shang, H Huang, Q Tang and S.H Yin: Coolant Effect on Grinding Performance in High Speed Deep Grinding of 40Cr steel, Journal of Metal Finishing, Vol.106 (2008), pp.16-
[40] G.Z Xie and H Huang: An Experimental Investigation of Temperature in High Speed Deep Grinding of Partially Stabilized Zirconia, International Journal of Machine Tool and Manufacture, (2008) doi:10.1016/j.ijmachtools.2008.06.002
[41] H.H.K Xu and S Jahanmir, Simple technique for observing subsurface damage in machining
of ceramics, Journal of American Ceramic Society, 77 (1994), 1388-1390
[42] A.G Evans and D.B Marshall: Wear Mechanisms in Ceramics, in Fundamental of Friction and Wear of Materials, ed D.A Rigney, Metals Park, Ohio: American Society for Metals,
1981, pp.439–452
[43] H Huang: Effect of Truing/Dressing Intensity on Truing/Dressing Efficiency and Grinding Performance of Vitrified Diamond Wheels, Journal of Materials Processing Technology, Vol.117 (2001), pp.9-14
[44] H Huang, Z.T Shang, H.Q Mi, X.M Sheng, S.Q Wang, Y Wu and G.Z Xie: The Closed type Nozzle for Ultrahigh Speed Grinding China Patent, CN200520052869.7
Y-[45] Z.T Shang, H Huang, Q Tang and S.H Yin: Coolant Effect on Grinding Performance in High Speed Deep Grinding of 40Cr steel, Journal of Metal Finishing, Vol.106 (2008), pp.16-
21
Trang 29Experimental Investigations on Material Removal Rate and Surface Roughness in Lapping of Substrate Wafers: A Literature Review
1Department of Industrial and Manufacturing Systems Engineering, Kansas State University,
Manhattan, KS 66506, USA
aweilong@ksu.edu; bpengfei@ksu.edu; czpei@ksu.edu
Keywords: Lapping, Material removal rate, Sapphire, Silicon, Substrate wafer, Surface roughness
Abstract Lapping is an important material-removal process for manufacturing of substrate wafers
Objectives of lapping include removing subsurface damage in sliced wafers, thinning wafers to target thickness, and achieving a high degree of parallelism and flatness of wafer surfaces This paper reviews the literature on lapping of substrate wafers It presents reported experimental results
on effects of input parameters (lapping pressure, plate rotation speed, abrasive grain size, slurry concentration, and slurry flow rate) on material removal rate and surface roughness
Introduction
Substrate wafers can be made of different materials such as gallium arsenide, germanium, lithium niobate, sapphire, silicon, and silicon carbide [1-10] They are used to fabricate various semiconductor and optical devices [1,3]
In order to manufacture high quality substrate wafers, a series of processes are needed, including crystal growth, shaping, slicing, edge grinding, lapping or grinding, etching, polishing, and cleaning [1,11-16] Fig 1 shows a typical process flow for substrate wafer manufacturing with a brief description for each process in Table 1
Fig 1 Manufacturing processes for substrate wafers (after [1,2-7])
Lapping is an important process in manufacturing of substrate wafers Objectives of lapping include removing subsurface damage in sliced wafers, thinning wafers to target thickness, and achieving a high degree of parallelism and flatness of wafer surfaces [2] Both single-side and double-side lapping processes (as illustrated in Fig 2) have been used to lap substrate wafers [1,2]
1Department of Industrial and Manufacturing Systems Engineering, Kansas State University,
Manhattan, KS 66506, USA
aweilong@ksu.edu; bpengfei@ksu.edu; czpei@ksu.edu
Keywords: Lapping, Material removal rate, Sapphire, Silicon, Substrate wafer, Surface roughness
Abstract Lapping is an important material-removal process for manufacturing of substrate wafers
Objectives of lapping include removing subsurface damage in sliced wafers, thinning wafers to target thickness, and achieving a high degree of parallelism and flatness of wafer surfaces This paper reviews the literature on lapping of substrate wafers It presents reported experimental results
on effects of input parameters (lapping pressure, plate rotation speed, abrasive grain size, slurry concentration, and slurry flow rate) on material removal rate and surface roughness
Introduction
Substrate wafers can be made of different materials such as gallium arsenide, germanium, lithium niobate, sapphire, silicon, and silicon carbide [1-10] They are used to fabricate various semiconductor and optical devices [1,3]
In order to manufacture high quality substrate wafers, a series of processes are needed, including crystal growth, shaping, slicing, edge grinding, lapping or grinding, etching, polishing, and cleaning [1,11-16] Fig 1 shows a typical process flow for substrate wafer manufacturing with a brief description for each process in Table 1
Fig 1 Manufacturing processes for substrate wafers (after [1,2-7])
Lapping is an important process in manufacturing of substrate wafers Objectives of lapping include removing subsurface damage in sliced wafers, thinning wafers to target thickness, and achieving a high degree of parallelism and flatness of wafer surfaces [2] Both single-side and double-side lapping processes (as illustrated in Fig 2) have been used to lap substrate wafers [1,2]
Trang 30Table 1 Description of substrate wafer manufacturing processes (after [1,2-7])
Crystal growth Produce crystal ingots with required purity and crystal structure
Shaping To remove both ends of an ingot, grind the ingot to a required diameter, and
grind flats or a notch on the ingot Slicing Slice an ingot into individual wafers with inner-diameter saw or wire saw Edge grinding Obtain smooth edge surface to prevent wafers from defects like edge chipping
and crevices Lapping Remove damage left by slicing and to achieve a high degree of parallelism
and flatness on wafer surfaces Etching Remove the damage and contamination from wafer surfaces
Polishing Achieve smooth and flat wafer surfaces
Cleaning Remove contamination from wafer surfaces
Inspection Inspect wafer quality
Packaging Package wafers for shipment
(a) Single side lapping (b) Double side lapping
Fig 2 Illustrations of lapping processes (after [1,2])
In double-side lapping (DSL), loose abrasive particles are suspended in a colloidal slurry to abrade material from wafer surfaces [3] Wafers are held in geared carriers which are driven in the planetary motion After a batch of wafers is manually loaded into the holes of the carriers, the upper plate will be forced down by a certain pressure (or weight) The two plates start to rotate either in the same direction or opposite directions [3] During double-side lapping, both sides of the wafers are lapped simultaneously The colloidal slurry is continuously filled into the lapping machine, and
a thin film of slurry is usually present between wafers and the two plates [2,4] The slurry performs the material removal through the abrasive grits as they slide or roll between wafer surfaces and the two plates Important lapping parameters include lapping pressure, plate rotation speed, plate material, abrasive material and grain size, slurry concentration, slurry flow rate, and carrier design [3,5]
Many experimental investigations on lapping of substrate wafers have been reported However, there exist no comprehensive review papers that cover all the experimental investigations reported
up to date on lapping of substrate wafers Such review papers are desirable to not only researchers but also industrial practitioners The objective of this paper is to provide a comprehensive review covering experimental investigations on material removal rate and surface roughness in lapping of substrate wafers
This paper is organized into four sections After this introduction, Section 2 and 3 present reported experimental investigations on material removal rate and surface roughness, respectively Section 4 contains concluding remarks
Table 1 Description of substrate wafer manufacturing processes (after [1,2-7])
Crystal growth Produce crystal ingots with required purity and crystal structure
Shaping To remove both ends of an ingot, grind the ingot to a required diameter, and
grind flats or a notch on the ingot Slicing Slice an ingot into individual wafers with inner-diameter saw or wire saw Edge grinding Obtain smooth edge surface to prevent wafers from defects like edge chipping
and crevices Lapping Remove damage left by slicing and to achieve a high degree of parallelism
and flatness on wafer surfaces Etching Remove the damage and contamination from wafer surfaces
Polishing Achieve smooth and flat wafer surfaces
Cleaning Remove contamination from wafer surfaces
Inspection Inspect wafer quality
Packaging Package wafers for shipment
(a) Single side lapping (b) Double side lapping
Fig 2 Illustrations of lapping processes (after [1,2])
In double-side lapping (DSL), loose abrasive particles are suspended in a colloidal slurry to abrade material from wafer surfaces [3] Wafers are held in geared carriers which are driven in the planetary motion After a batch of wafers is manually loaded into the holes of the carriers, the upper plate will be forced down by a certain pressure (or weight) The two plates start to rotate either in the same direction or opposite directions [3] During double-side lapping, both sides of the wafers are lapped simultaneously The colloidal slurry is continuously filled into the lapping machine, and
a thin film of slurry is usually present between wafers and the two plates [2,4] The slurry performs the material removal through the abrasive grits as they slide or roll between wafer surfaces and the two plates Important lapping parameters include lapping pressure, plate rotation speed, plate material, abrasive material and grain size, slurry concentration, slurry flow rate, and carrier design [3,5]
Many experimental investigations on lapping of substrate wafers have been reported However, there exist no comprehensive review papers that cover all the experimental investigations reported
up to date on lapping of substrate wafers Such review papers are desirable to not only researchers but also industrial practitioners The objective of this paper is to provide a comprehensive review covering experimental investigations on material removal rate and surface roughness in lapping of substrate wafers
This paper is organized into four sections After this introduction, Section 2 and 3 present reported experimental investigations on material removal rate and surface roughness, respectively Section 4 contains concluding remarks
Trang 31Experimental Investigations on Material Removal Rate
To calculate material removal rates in lapping of substrate wafers, some researchers used the following equation:
Table 2 Lapping conditions used by Marinescu et al [2]
A A
B
0.0 0.6 1.2 1.8 2.4
A A
B
0.0 0.6 1.2 1.8 2.4
Lapping time = 25 min
Fig 3 MRR in double side lapping of silicon wafers (after [2])
Li et al [6] studied on the effects of plate rotation speed and lapping pressure on material remove rate (MRR) in lapping of sapphire wafers Their experimental conditions are presented in Table 3 The relationship between MRR and plate rotation speed is shown in Fig 4 With the increase in plate rotation speed, MRR increased remarkably The relationship between MRR and lapping pressure (weight) can be seen in Fig 5 With the increase in lapping pressure, MRR increased linearly However, if the lapping pressure was too high, the lapping abrasive grains were likely to
be crashed, MRR would be reduced Therefore, the lapping pressure (weight) should be increased within a moderate range
Experimental Investigations on Material Removal Rate
To calculate material removal rates in lapping of substrate wafers, some researchers used the following equation:
Table 2 Lapping conditions used by Marinescu et al [2]
A A
B
0.0 0.6 1.2 1.8 2.4
A A
B
0.0 0.6 1.2 1.8 2.4
Lapping time = 25 min
Fig 3 MRR in double side lapping of silicon wafers (after [2])
Li et al [6] studied on the effects of plate rotation speed and lapping pressure on material remove rate (MRR) in lapping of sapphire wafers Their experimental conditions are presented in Table 3 The relationship between MRR and plate rotation speed is shown in Fig 4 With the increase in plate rotation speed, MRR increased remarkably The relationship between MRR and lapping pressure (weight) can be seen in Fig 5 With the increase in lapping pressure, MRR increased linearly However, if the lapping pressure was too high, the lapping abrasive grains were likely to
be crashed, MRR would be reduced Therefore, the lapping pressure (weight) should be increased within a moderate range
Trang 32Table 3 Experiment conditions used by Li et al [6]
(Lapping pressure (weight) = 2.3 kg;
Abrasive grain size = # 240)
(Plate rotation speed = 45 rpm;
Abrasive grain size = # 240 )
Fig 4 Relationship between MRR and plate
rotation speed in lapping of sapphire wafers
(after [6])
Fig 5 Relationship between MRR and lapping pressure in lapping of sapphire wafers
(after [6])
They also studied the effects of abrasive grain size on MRR The MRR was 0.346 μm/min when
# 600 boron carbide slurry was used, while the MRR was 3.4 μm/min when # 240 boron carbide slurry was used under the same lapping condition, where the plate rotation speed was 45 rpm and the lapping pressure (weight) was 3 kg The MRR had increased by an order of magnitude
Dudly [4] studied the effect of lapping pressure and plate rotation speed on MRR in lapping of silicon wafers Their experimental conditions are presented in Table 4 Figs 6 and 7 show the relationship between MRR and lapping pressure for different abrasive grain size and different slurry flow rate, respectively It can be seen that MRR increased as lapping pressure, abrasive grain size, and slurry flow rate increased
Table 4 Experiment conditions used by Dudly [4]
Wang et al [7] studied the effects of lapping parameters (abrasive grain size, lapping plate hardness, lapping pressure, plate rotation speed, and slurry concentration) on MRR in lapping of BK7 glass Their experimental conditions are presented in Table 5 Fig 8 shows MRR for different abrasive grain sizes It can be seen that MRR increased sharply with an increase in abrasive grain size However, the slope of the increasing curve became smaller when the abrasive grain size exceeded a certain value, i.e., 20 μm The MRR for different lapping plate hardness values are
Table 3 Experiment conditions used by Li et al [6]
(Lapping pressure (weight) = 2.3 kg;
Abrasive grain size = # 240)
(Plate rotation speed = 45 rpm;
Abrasive grain size = # 240 )
Fig 4 Relationship between MRR and plate
rotation speed in lapping of sapphire wafers
(after [6])
Fig 5 Relationship between MRR and lapping pressure in lapping of sapphire wafers
(after [6])
They also studied the effects of abrasive grain size on MRR The MRR was 0.346 μm/min when
# 600 boron carbide slurry was used, while the MRR was 3.4 μm/min when # 240 boron carbide slurry was used under the same lapping condition, where the plate rotation speed was 45 rpm and the lapping pressure (weight) was 3 kg The MRR had increased by an order of magnitude
Dudly [4] studied the effect of lapping pressure and plate rotation speed on MRR in lapping of silicon wafers Their experimental conditions are presented in Table 4 Figs 6 and 7 show the relationship between MRR and lapping pressure for different abrasive grain size and different slurry flow rate, respectively It can be seen that MRR increased as lapping pressure, abrasive grain size, and slurry flow rate increased
Table 4 Experiment conditions used by Dudly [4]
Wang et al [7] studied the effects of lapping parameters (abrasive grain size, lapping plate hardness, lapping pressure, plate rotation speed, and slurry concentration) on MRR in lapping of BK7 glass Their experimental conditions are presented in Table 5 Fig 8 shows MRR for different abrasive grain sizes It can be seen that MRR increased sharply with an increase in abrasive grain size However, the slope of the increasing curve became smaller when the abrasive grain size exceeded a certain value, i.e., 20 μm The MRR for different lapping plate hardness values are
Trang 33shown in Fig 9 MRR increased with an increase in plate hardness.The effects of lapping pressure and plate rotation speed on MRR are shown in Figs 10 and 11, respectively The MRR increased dramatically with an increase in the lapping pressure and plate rotation speed The effects of slurry concentration on MRR are given in Fig 12 It can be seen that MRR was in direct proportion to slurry concentration [7]
(Slurry flow rate = 20 ml/min) (Abrasive grain size = 20 μm)
Fig 6 Relationship between MRR and lapping
pressure for different abrasive grain size in
lapping of silicon wafers (after [4])
Fig 7 Relationship between MRR and lapping pressure for different slurry flow rate in lapping
of silicon wafers (after [4])
Table 5 Experiment conditions used by Wang et al [7]
Abrasive grain size = 20 μm;
Lapping pressure = 16.2 kPa;
Plate rotation speed = 50 rpm;
5 8 11 14
shown in Fig 9 MRR increased with an increase in plate hardness.The effects of lapping pressure and plate rotation speed on MRR are shown in Figs 10 and 11, respectively The MRR increased dramatically with an increase in the lapping pressure and plate rotation speed The effects of slurry concentration on MRR are given in Fig 12 It can be seen that MRR was in direct proportion to slurry concentration [7]
(Slurry flow rate = 20 ml/min) (Abrasive grain size = 20 μm)
Fig 6 Relationship between MRR and lapping
pressure for different abrasive grain size in
lapping of silicon wafers (after [4])
Fig 7 Relationship between MRR and lapping pressure for different slurry flow rate in lapping
of silicon wafers (after [4])
Table 5 Experiment conditions used by Wang et al [7]
Abrasive grain size = 20 μm;
Lapping pressure = 16.2 kPa;
Plate rotation speed = 50 rpm;
5 8 11 14
Trang 34Abrasive grain size = 20 μm;
Plate rotation speed = 50 rpm;
Lapping pressure = 16.2 kPa;
Aluminum plate
Fig 12 Effects of slurry concentration on MRR in lapping of BK7 (after [7])
Othman et al [8] investigated the effects of plate rotation speed and lapping pressure on MRR in lapping of GaAs wafers Higher plate rotation speed and lapping pressure could lead to higher MRR Lapping pressure had much more significant effects than plate rotation speed did [8] However, details of their experimental conditions were not provided in their paper
Experimental Investigations on Surface Roughness
Marinescu et al [2] studied the surface roughness in double side lapping of silicon Details of their experimental conditions and parameter settings are presented in Table 2 Fig 13 shows the effects
of plate rotation speed and lapping pressure on surface roughness It can be seen that, for lapping pressure = 121 and 484 psi, surface roughness increased as plate rotation speed increased For lapping pressure = 242 psi, surface roughness was the highest when plate rotation speed was 75 rpm Fig 14 shows the relationship between surface roughness and lapping time It can be seen that, as time increased, surface roughness improved, regardless of what lapping pressure was At low lapping pressure (121 psi), surface roughness did not improve noticeably as lapping time increased
At high pressures, surface roughness improved rapidly as lapping time increased Surface roughness improved more rapidly at initial stages for equal intervals of time
Abrasive grain size = 20 μm;
Plate rotation speed = 50 rpm;
Lapping pressure = 16.2 kPa;
Aluminum plate
Fig 12 Effects of slurry concentration on MRR in lapping of BK7 (after [7])
Othman et al [8] investigated the effects of plate rotation speed and lapping pressure on MRR in lapping of GaAs wafers Higher plate rotation speed and lapping pressure could lead to higher MRR Lapping pressure had much more significant effects than plate rotation speed did [8] However, details of their experimental conditions were not provided in their paper
Experimental Investigations on Surface Roughness
Marinescu et al [2] studied the surface roughness in double side lapping of silicon Details of their experimental conditions and parameter settings are presented in Table 2 Fig 13 shows the effects
of plate rotation speed and lapping pressure on surface roughness It can be seen that, for lapping pressure = 121 and 484 psi, surface roughness increased as plate rotation speed increased For lapping pressure = 242 psi, surface roughness was the highest when plate rotation speed was 75 rpm Fig 14 shows the relationship between surface roughness and lapping time It can be seen that, as time increased, surface roughness improved, regardless of what lapping pressure was At low lapping pressure (121 psi), surface roughness did not improve noticeably as lapping time increased
At high pressures, surface roughness improved rapidly as lapping time increased Surface roughness improved more rapidly at initial stages for equal intervals of time
Trang 35A A
300 600 900 1200 1500
Lapping time = 5 min Plate rotation speed = 50 rpm
Fig 13 Effects of lapping pressure and plate
rotation speed on surface roughness in double
side lapping of silicon wafers (after [2])
Fig 14 Relationship between surface roughness and lapping time in double side lapping of silicon wafers (after [2])
Li et al [6] studied the effects of plate rotation speed and lapping pressure on surface roughness
in lapping of sapphire wafers The experimental conditions are listed in Table 3 Fig 15 shows the relationship between surface roughness and plate rotation speed It can be seen that surface roughness decreased with the increase of plate rotation speed It was found that, because of their large grain size, # 240 boron carbide did not distribute very uniformly in the slurry When plate rotation speed was low, non-uniformly distributed abrasive grains could induce deep scratches; as a result, surface roughness was very high With an increase in the plate rotation speed, abrasive grains could distribute more uniformly on the plate, resulting in improved surface roughness
Fig 16 shows the relationship between surface roughness and lapping pressure (weight) It can
be seen that surface roughness decreased with the increase of lapping pressure Because # 240 boron carbide had a larger size and a certain distribution, when lapping pressure was low, sapphire substrates mainly contacted with large abrasive grains As a result, the pressure on those grains was very high, and hence scratches were relatively deep Moreover, large grains had a certain distance between each other, which also made surface roughness worse When lapping pressure (weight) was higher, small grains could contact with sapphire substrates and leveled off deep scratches induced by large grains Moreover, the distance between grains became smaller; as a result, the surface roughness was improved
Lapping pressure (weight) = 2.3 kg;
Abrasive grain size = # 240)
Abrasive grain size = # 240;
Plate rotation speed = 45 rpm
Fig 15 Relationship between plate rotation speed
and surface roughness in lapping of sapphire
300 600 900 1200 1500
Lapping time = 5 min Plate rotation speed = 50 rpm
Fig 13 Effects of lapping pressure and plate
rotation speed on surface roughness in double
side lapping of silicon wafers (after [2])
Fig 14 Relationship between surface roughness and lapping time in double side lapping of silicon wafers (after [2])
Li et al [6] studied the effects of plate rotation speed and lapping pressure on surface roughness
in lapping of sapphire wafers The experimental conditions are listed in Table 3 Fig 15 shows the relationship between surface roughness and plate rotation speed It can be seen that surface roughness decreased with the increase of plate rotation speed It was found that, because of their large grain size, # 240 boron carbide did not distribute very uniformly in the slurry When plate rotation speed was low, non-uniformly distributed abrasive grains could induce deep scratches; as a result, surface roughness was very high With an increase in the plate rotation speed, abrasive grains could distribute more uniformly on the plate, resulting in improved surface roughness
Fig 16 shows the relationship between surface roughness and lapping pressure (weight) It can
be seen that surface roughness decreased with the increase of lapping pressure Because # 240 boron carbide had a larger size and a certain distribution, when lapping pressure was low, sapphire substrates mainly contacted with large abrasive grains As a result, the pressure on those grains was very high, and hence scratches were relatively deep Moreover, large grains had a certain distance between each other, which also made surface roughness worse When lapping pressure (weight) was higher, small grains could contact with sapphire substrates and leveled off deep scratches induced by large grains Moreover, the distance between grains became smaller; as a result, the surface roughness was improved
Lapping pressure (weight) = 2.3 kg;
Abrasive grain size = # 240)
Abrasive grain size = # 240;
Plate rotation speed = 45 rpm
Fig 15 Relationship between plate rotation speed
and surface roughness in lapping of sapphire
wafers (after [6])
Fig 16 Relationship between lapping pressure (weight) and surface roughness in lapping of
sapphire (after [6])
Trang 36Surface roughness was 60.45 nm with # 600 boron carbide slurry, and 416.21 nm with # 240 boron carbide slurry when other conditions were kept the same (plate rotation speed was 45 rpm and lapping pressure (weight) was 3 kg)
Prochnow and Edwards [9] reported lapping of sapphire wafers Their experimental conditions are shown in Table 6 They used a cast iron lap with # 400 B4C abrasives in water to remove ~50
μm of material, and on a copper Kemet lap with ~3 μm diamond abrasives in water to remove
~30-40 μm of material Each of these two steps was completed within ~15 minutes Then ~1 μm diamond abrasives in water were used on the copper Kemet lap to finish the wafers This step required ~30 minutes The surface roughness (RMS) of the finished wafer was 50-70 Å
Table 6 Experiment conditions used by Prochnow [9]
Wang [10] invented a double-side lapping machine and claimed a sequence of lapping steps for sapphire wafers 25 pieces of sapphire wafers (with 50 mm in diameter and ~0.5 mm in thickness) could be loaded into five carriers (with each having five through holes) The sapphire wafers also had slight rotation during lapping due to unbalanced friction on both sides Lapping pressure could
be adjusted by the air-pressure system mounted atop the upper lapping plate The lower lapping plate was driven to rotate by a driving system The upper lapping plate would automatically rotate
in the opposite direction to the lower lapping plate due to the lapping friction Wang recommended three types of B4C abrasives for lapping sapphire wafers They are W20 (10-20 μm), W14 (7-14 μm), and W7 (3.5-7 μm), respectively He claimed that B4C abrasives performed better than SiC due to their higher hardness and would result in lower machining cost than diamond due to their cheaper powder preparation cost After two steps of lapping, surface roughness could reach 0.3 nm
Concluding Remarks
In lapping of substrate wafers, a change in abrasive grain size could cause an “order of magnitude” change in material removal rate (MRR) and surface roughness Other factors such as plate rotation speed and lapping pressure could affect MRR and surface roughness within the same order of magnitude These results would have practical guidance to manufacturing of substrate wafers For example, in order to reduce lapping time, the slurry with a larger abrasive grain size should be used However, when the substrate thickness approaches the required value, smaller abrasive grains should be used so as to keep surface roughness within an allowable limit
Acknowledgements
This study was supported by the National Science Foundation through the CAREER Award
CMMI-0348290
References
[1] M Quirk and J Serda: Semiconductor Manufacturing Technology, Chap 4 (Pearson Education
International, Columbus, Ohio, 2001), pp 67-90
[2] I.D Marinescu, A Shoutak and C.E Spanu: Abrasives Magazine, Vol Dec-Jan (2002), pp 5-9 [3] M Naselaris: Proceedings of the SPIE, Vol TD03 (2005), pp 118-120
[4] J.A Dudley: Microelectronic Manufacturing and Testing, Vol 9 (1986) No.4, pp 1-6
[5] U Heisel and J Avroutine: CIRP Annals, Vol 50 (2001) No.1, pp 229-232
Surface roughness was 60.45 nm with # 600 boron carbide slurry, and 416.21 nm with # 240 boron carbide slurry when other conditions were kept the same (plate rotation speed was 45 rpm and lapping pressure (weight) was 3 kg)
Prochnow and Edwards [9] reported lapping of sapphire wafers Their experimental conditions are shown in Table 6 They used a cast iron lap with # 400 B4C abrasives in water to remove ~50
μm of material, and on a copper Kemet lap with ~3 μm diamond abrasives in water to remove
~30-40 μm of material Each of these two steps was completed within ~15 minutes Then ~1 μm diamond abrasives in water were used on the copper Kemet lap to finish the wafers This step required ~30 minutes The surface roughness (RMS) of the finished wafer was 50-70 Å
Table 6 Experiment conditions used by Prochnow [9]
Wang [10] invented a double-side lapping machine and claimed a sequence of lapping steps for sapphire wafers 25 pieces of sapphire wafers (with 50 mm in diameter and ~0.5 mm in thickness) could be loaded into five carriers (with each having five through holes) The sapphire wafers also had slight rotation during lapping due to unbalanced friction on both sides Lapping pressure could
be adjusted by the air-pressure system mounted atop the upper lapping plate The lower lapping plate was driven to rotate by a driving system The upper lapping plate would automatically rotate
in the opposite direction to the lower lapping plate due to the lapping friction Wang recommended three types of B4C abrasives for lapping sapphire wafers They are W20 (10-20 μm), W14 (7-14 μm), and W7 (3.5-7 μm), respectively He claimed that B4C abrasives performed better than SiC due to their higher hardness and would result in lower machining cost than diamond due to their cheaper powder preparation cost After two steps of lapping, surface roughness could reach 0.3 nm
Concluding Remarks
In lapping of substrate wafers, a change in abrasive grain size could cause an “order of magnitude” change in material removal rate (MRR) and surface roughness Other factors such as plate rotation speed and lapping pressure could affect MRR and surface roughness within the same order of magnitude These results would have practical guidance to manufacturing of substrate wafers For example, in order to reduce lapping time, the slurry with a larger abrasive grain size should be used However, when the substrate thickness approaches the required value, smaller abrasive grains should be used so as to keep surface roughness within an allowable limit
Acknowledgements
This study was supported by the National Science Foundation through the CAREER Award
CMMI-0348290
References
[1] M Quirk and J Serda: Semiconductor Manufacturing Technology, Chap 4 (Pearson Education
International, Columbus, Ohio, 2001), pp 67-90
[2] I.D Marinescu, A Shoutak and C.E Spanu: Abrasives Magazine, Vol Dec-Jan (2002), pp 5-9 [3] M Naselaris: Proceedings of the SPIE, Vol TD03 (2005), pp 118-120
[4] J.A Dudley: Microelectronic Manufacturing and Testing, Vol 9 (1986) No.4, pp 1-6
[5] U Heisel and J Avroutine: CIRP Annals, Vol 50 (2001) No.1, pp 229-232
Trang 37[6] B Li, X Guo and Y Liu: Semiconductor Technology, Vol 30 (2005) No.9, pp 57-60
[7] Z Wang, Y.L Wu and Y.F Dai: Applied Optics, Vol 47 (2008) No.10, pp 1417-1426
[8] M.K Othman, A Dolah, N.A Omar and M.R Yahya: IEEE International Conference on Semiconductor Electronics (2006), pp 583-585
[9] E Prochnow and D F Edwards: Applied Optics, Vol 25 (1986), pp 2639-2640
[10] K.Q Wang: Lapping/Polishing Machine for Optical Parts and its Application in Lapping/
Polishing Sapphire Wafers, CN Patent 1,546,283, 2003
[11] M.S Bawa, E.F Petro and H.M Grimes: Semiconductor International, Vol 18 (1995), pp 115 -118
[12] T Fukami, H Masumura, K Suzuki and H Kudo: Method of Manufacturing Semiconductor
Mirror Wafers, European Patent Application EP0782179A2, 1997
[13] Z.J Pei, S.R Billingsley and S Miura: International Journal of Machine Tools and Manufacture, Vol 39 (1999), pp 1103–1116
[14] G.J Pietsch and M Kerstan: Simultaneous Double-disk Grinding Machining Process for Flat,
Low-damage and Material-saving Silicon Wafer Substrate Manufacturing, Proceeding of the
2nd Euspen International Conference, Turin, Italy, (2001), pp 644–648
[15] S Wolf and R.N Tauber: Silicon Processing for the VLSI Era, Process Technology, Vol 1 (Lattice Press, Sunset Beach, CA 2000)
[16] R Vandamme, Y Xin and Z.J Pei: Method of Processing Semiconductor Wafers, US Patent
6,114,245, 2000
[6] B Li, X Guo and Y Liu: Semiconductor Technology, Vol 30 (2005) No.9, pp 57-60
[7] Z Wang, Y.L Wu and Y.F Dai: Applied Optics, Vol 47 (2008) No.10, pp 1417-1426
[8] M.K Othman, A Dolah, N.A Omar and M.R Yahya: IEEE International Conference on Semiconductor Electronics (2006), pp 583-585
[9] E Prochnow and D F Edwards: Applied Optics, Vol 25 (1986), pp 2639-2640
[10] K.Q Wang: Lapping/Polishing Machine for Optical Parts and its Application in Lapping/
Polishing Sapphire Wafers, CN Patent 1,546,283, 2003
[11] M.S Bawa, E.F Petro and H.M Grimes: Semiconductor International, Vol 18 (1995), pp 115 -118
[12] T Fukami, H Masumura, K Suzuki and H Kudo: Method of Manufacturing Semiconductor
Mirror Wafers, European Patent Application EP0782179A2, 1997
[13] Z.J Pei, S.R Billingsley and S Miura: International Journal of Machine Tools and Manufacture, Vol 39 (1999), pp 1103–1116
[14] G.J Pietsch and M Kerstan: Simultaneous Double-disk Grinding Machining Process for Flat,
Low-damage and Material-saving Silicon Wafer Substrate Manufacturing, Proceeding of the
2nd Euspen International Conference, Turin, Italy, (2001), pp 644–648
[15] S Wolf and R.N Tauber: Silicon Processing for the VLSI Era, Process Technology, Vol 1 (Lattice Press, Sunset Beach, CA 2000)
[16] R Vandamme, Y Xin and Z.J Pei: Method of Processing Semiconductor Wafers, US Patent
6,114,245, 2000
Trang 38A Focused Review on Enhancing the Abrasive Waterjet Cutting
Performance by Using Controlled Nozzle Oscillation
1
School of Mechanical and Manufacturing Engineering, The University of New South Wales,
Sydney, NSW 2052, Australia a
jun.wang@unsw.edu.au
Keywords: Abrasive waterjet, Machining, Machining performance, Nozzle oscillation
Abstract Increasing the performance of the abrasive waterjet (AWJ) cutting technology for
engineering materials is the ultimate aim of research in this field This paper presents a review on the studies using a controlled nozzle oscillation technique to increase the cutting performance of the AWJ cutting technology and the associated mechanisms primarily based on the work in the author’s laboratory Primary attention is paid to the discussions of the depth of cut, the effect and selection of process parameters and the advantages by using this technique in both single- and multi-pass cutting modes
Introduction
Abrasive waterjet (AWJ) machining has been found to have some distinct advantages over the other machining technologies such as no thermal effect, high machining versatility, high flexibility and small cutting forces [1,2] It is increasingly used by industry to process various materials, particularly difficult-to-machine materials such as ceramics [3-12] and composites [6,13-19] In the last decades, significant research has been carried out to explore the mechanisms of the AWJ machining process [2,20,21] It has been found [3,22] that three cutting zones exist in the processing of ductile and brittle materials under an AWJ, i.e a cutting zone at shallow angles of attack, a cutting zone at large angles
of attack, and a jet upward deflection zone The attack angle is defined as the angle between the jet flowing direction and the target surface The study on layered materials such as polymer matrix composites [16-18] has revealed similar phenomenon in terms of the cutting zones Based on the proposals by Bitter [23] and Finnie [24] for particle erosion of materials, Hashish [22] claimed that the cutting mechanisms in the first two zones could be considered as cutting wear and deformation wear, respectively, while in the third zone the cutting process is considered as being controlled by erosive wear at large particle attack angles [3] Furthermore, it has been found that the surfaces produced by an AWJ consist of an upper smooth zone where the surface is characterized by surface roughness and a lower rough zone where the surface has wavy striations, as shown in Fig 1 In the jet upward deflection zone (for non-through cuts only), a large pocket is formed Research is still being undertaken to gain a deeper understanding of the mechanism of striation formation in order to reduce
or eliminate its formation The geometry of the kerf generated by an AWJ is characterized by a wider entry at the top than the exit at the bottom so that a taper is produced There may be a round corner at the top kerf edges because of water bombardment, and burrs at the exit kerf edges for through cuts of ductile materials as a result of the material plastic deformation, as shown in Fig 1
A large amount of research effort has been directed towards understanding and improving the AWJ cutting performance, such as the kerf quality (kerf taper, surface roughness etc.), material removal rate and depth of cut This includes the study of the jet dynamic characteristics [2, 25-29], and the analysis of the machined surfaces and kerf geometrical features to optimise the cutting process [2,5,16,22, 30-32] In addition, predictive models for material removal rate and the depth of cut have been developed using the erosive theories [22,33], an energy conservation approach [16,34,35], fracture mechanics [36,37], dimensional analysis [9,11,12,19,29] and accumulating the micro-cutting
A Focused Review on Enhancing the Abrasive Waterjet Cutting
Performance by Using Controlled Nozzle Oscillation
1
School of Mechanical and Manufacturing Engineering, The University of New South Wales,
Sydney, NSW 2052, Australia a
jun.wang@unsw.edu.au
Keywords: Abrasive waterjet, Machining, Machining performance, Nozzle oscillation
Abstract Increasing the performance of the abrasive waterjet (AWJ) cutting technology for
engineering materials is the ultimate aim of research in this field This paper presents a review on the studies using a controlled nozzle oscillation technique to increase the cutting performance of the AWJ cutting technology and the associated mechanisms primarily based on the work in the author’s laboratory Primary attention is paid to the discussions of the depth of cut, the effect and selection of process parameters and the advantages by using this technique in both single- and multi-pass cutting modes
Introduction
Abrasive waterjet (AWJ) machining has been found to have some distinct advantages over the other machining technologies such as no thermal effect, high machining versatility, high flexibility and small cutting forces [1,2] It is increasingly used by industry to process various materials, particularly difficult-to-machine materials such as ceramics [3-12] and composites [6,13-19] In the last decades, significant research has been carried out to explore the mechanisms of the AWJ machining process [2,20,21] It has been found [3,22] that three cutting zones exist in the processing of ductile and brittle materials under an AWJ, i.e a cutting zone at shallow angles of attack, a cutting zone at large angles
of attack, and a jet upward deflection zone The attack angle is defined as the angle between the jet flowing direction and the target surface The study on layered materials such as polymer matrix composites [16-18] has revealed similar phenomenon in terms of the cutting zones Based on the proposals by Bitter [23] and Finnie [24] for particle erosion of materials, Hashish [22] claimed that the cutting mechanisms in the first two zones could be considered as cutting wear and deformation wear, respectively, while in the third zone the cutting process is considered as being controlled by erosive wear at large particle attack angles [3] Furthermore, it has been found that the surfaces produced by an AWJ consist of an upper smooth zone where the surface is characterized by surface roughness and a lower rough zone where the surface has wavy striations, as shown in Fig 1 In the jet upward deflection zone (for non-through cuts only), a large pocket is formed Research is still being undertaken to gain a deeper understanding of the mechanism of striation formation in order to reduce
or eliminate its formation The geometry of the kerf generated by an AWJ is characterized by a wider entry at the top than the exit at the bottom so that a taper is produced There may be a round corner at the top kerf edges because of water bombardment, and burrs at the exit kerf edges for through cuts of ductile materials as a result of the material plastic deformation, as shown in Fig 1
A large amount of research effort has been directed towards understanding and improving the AWJ cutting performance, such as the kerf quality (kerf taper, surface roughness etc.), material removal rate and depth of cut This includes the study of the jet dynamic characteristics [2, 25-29], and the analysis of the machined surfaces and kerf geometrical features to optimise the cutting process [2,5,16,22, 30-32] In addition, predictive models for material removal rate and the depth of cut have been developed using the erosive theories [22,33], an energy conservation approach [16,34,35], fracture mechanics [36,37], dimensional analysis [9,11,12,19,29] and accumulating the micro-cutting
Key Engineering Materials Vol 404 (2009) pp 33-44
© (2009) Trans Tech Publications, Switzerland
doi:10.4028/www.scientific.net/KEM.404.33
Trang 39processes of individual abrasive particles [38] It has been found that in order to increase these cutting performance measures, low jet traverse speeds are normally selected at high water pressures Such combinations of the process parameters are not preferred in practice from an economic point of view
As a result, various attempts have been made to increase the cutting performance of AWJ, including the use of water pressure at as high as 690 MPa [39] and very long and thick nozzles for large abrasives [40] The author’s laboratory has developed a number of novel cutting techniques to increase the AWJ cutting performance, such as angling the jet forward in the cutting plane [16,41], controlled nozzle oscillation [9,12,42,43], multipass cutting operations [7,8,43,44], and ultrasonic vibration assisted cutting [45] which can significantly increase the cutting performance without additional costs to the process
This paper focuses on the analysis of using the nozzle oscillation cutting technique to enhance the cutting performance, mostly based on the work in the author’s laboratory while referecnes will be made to other work in the world The studies of the combined use of the nozzle oscillation cutting with the multipass cutting technique will also be reviewed and analysed This review will be limited
to the straight-slit cutting mode and comments on other cutting modes including the scope for future research using this cutting technique will be given in conclusions
Fig 1 Schematic of AWJ produced kerf profile and surface
Nozzle oscillation
Jet traverse motion
Workpiece
Nozzle Kerf
Fig 2 Schematic of controlled nozzle oscillation
The Nozzle Oscillation AWJ Cutting Technique
As stated earlier, the surfaces generated by an AWJ consist of an upper smooth zone and a lower striation zone With an increase in jet traverse speed, the surface roughness increases, so does the striation Chao and Geskin [46] reported that cutting head vibration affected the formation and pattern
of surface striations This phenomenon was further explained as a result of the effect of internal and external factors by Chen et al [47] It is believed that if a jet is used in such a way that the jet can scan the surface being cut, the surface roughness and striation can be reduced Nozzle oscillation was introduced to perform such a scanning action [48] This cutting technique was then successfully used
to improve the kerf quality and depth of cut in processing various materials [3,9,12,42,43] With this
processes of individual abrasive particles [38] It has been found that in order to increase these cutting performance measures, low jet traverse speeds are normally selected at high water pressures Such combinations of the process parameters are not preferred in practice from an economic point of view
As a result, various attempts have been made to increase the cutting performance of AWJ, including the use of water pressure at as high as 690 MPa [39] and very long and thick nozzles for large abrasives [40] The author’s laboratory has developed a number of novel cutting techniques to increase the AWJ cutting performance, such as angling the jet forward in the cutting plane [16,41], controlled nozzle oscillation [9,12,42,43], multipass cutting operations [7,8,43,44], and ultrasonic vibration assisted cutting [45] which can significantly increase the cutting performance without additional costs to the process
This paper focuses on the analysis of using the nozzle oscillation cutting technique to enhance the cutting performance, mostly based on the work in the author’s laboratory while referecnes will be made to other work in the world The studies of the combined use of the nozzle oscillation cutting with the multipass cutting technique will also be reviewed and analysed This review will be limited
to the straight-slit cutting mode and comments on other cutting modes including the scope for future research using this cutting technique will be given in conclusions
Fig 1 Schematic of AWJ produced kerf profile and surface
Nozzle oscillation
Jet traverse motion
Workpiece
Nozzle Kerf
Fig 2 Schematic of controlled nozzle oscillation
The Nozzle Oscillation AWJ Cutting Technique
As stated earlier, the surfaces generated by an AWJ consist of an upper smooth zone and a lower striation zone With an increase in jet traverse speed, the surface roughness increases, so does the striation Chao and Geskin [46] reported that cutting head vibration affected the formation and pattern
of surface striations This phenomenon was further explained as a result of the effect of internal and external factors by Chen et al [47] It is believed that if a jet is used in such a way that the jet can scan the surface being cut, the surface roughness and striation can be reduced Nozzle oscillation was introduced to perform such a scanning action [48] This cutting technique was then successfully used
to improve the kerf quality and depth of cut in processing various materials [3,9,12,42,43] With this
Trang 40cutting technique, a pendulum-like nozzle forward and backward motion in the cutting plane at predetermined frequency and angular amplitude is superimposed to the nozzle traverse motion, as shown in Fig 2 The neutral nozzle position is set at the normal to the work surface and the nozzle performs a repeated motion between its neutral position and the maximum set angular amplitude in the nozzle traverse direction under a constant angular velocity
Earlier studies normally used relatively large oscillation angles of up to 30o [3] To perform oscillation movements at such large angles, small oscillation frequencies had to be used Siores et al [3] found that an oscillation angle in the range of 15-20o, gave the maximum average smooth depth of cut of the two kerf walls when cutting alumina ceramics at the nozzle traverse speed of 0.17-0.5mm/s and water pressure of 345MPa In general, more than 30% increase in the smooth depth of cut can be achieved by using nozzle oscillation They also fund that the oscillation frequency which gave the maximum average smooth depth of cut was, in value, about six times the nozzle traverse speed, e.g for a 0.5mm/s traverse speed, the optimum oscillation frequency is about 3Hz The authors further reported that when the oscillation angle was increased to above 20o and up to 30o, the smooth depth of cut on one side of the kerf walls was increased significantly by up to 105% as compared to the traditional cutting technique (no oscillation and 90o jet impact angle), while the smooth depth on the other side was worse off significantly The use of nozzle oscillation to increase the smooth depth of cut and surface finish has been confirmed when cutting mild steels and aluminium alloys [3,42] and fiber-reinforced composites [49]
Clearly, the nozzle oscillation technique can also increase the total depth of jet penetration; however, in earlier studies using relatively large oscillation angles, this advantage was not significant
so that no quantitative data were reported [3]
Fig 3 AWJ and workpiece interface trace profile at traverse speed of 0.5 mm/s and water pressure
of 276 MPa: (a) traditional AWJ cutting method; (b) with nozzle oscillation at
oscillation angle of 2o and oscillation frequency of 2Hz [42]
Studies also include the mechanisms under which nozzle oscillation cutting improves some cutting performance measures It has been reported [3,42,50] that nozzle oscillation cutting creates a scanning cutting action by the particles which not only reduces the particle interference, but also clears the target surface for more effective cutting by subsequent particles It is also believed [12] that the scanning action has changed the particle attack angles on the target surface which changes the material erosion mechanisms resulting in an increased cutting performance A visualization study has been carried out to explore the underlying mechanism in which nozzle oscillation can increase the depth of cut [42] This study was carried out on a transparent plexiglass The kerf formation process and jet (or particles)-work interference were recorded by a high speed video camera Some typical particle-work interface traces are shown in Fig 3 It has been found that the successive traces of the particles on the cut surface with nozzle oscillation is steeper than those without oscillation, which results in more particle energy in the cutting direction for deeper cuts
cutting technique, a pendulum-like nozzle forward and backward motion in the cutting plane at predetermined frequency and angular amplitude is superimposed to the nozzle traverse motion, as shown in Fig 2 The neutral nozzle position is set at the normal to the work surface and the nozzle performs a repeated motion between its neutral position and the maximum set angular amplitude in the nozzle traverse direction under a constant angular velocity
Earlier studies normally used relatively large oscillation angles of up to 30o [3] To perform oscillation movements at such large angles, small oscillation frequencies had to be used Siores et al [3] found that an oscillation angle in the range of 15-20o, gave the maximum average smooth depth of cut of the two kerf walls when cutting alumina ceramics at the nozzle traverse speed of 0.17-0.5mm/s and water pressure of 345MPa In general, more than 30% increase in the smooth depth of cut can be achieved by using nozzle oscillation They also fund that the oscillation frequency which gave the maximum average smooth depth of cut was, in value, about six times the nozzle traverse speed, e.g for a 0.5mm/s traverse speed, the optimum oscillation frequency is about 3Hz The authors further reported that when the oscillation angle was increased to above 20o and up to 30o, the smooth depth of cut on one side of the kerf walls was increased significantly by up to 105% as compared to the traditional cutting technique (no oscillation and 90o jet impact angle), while the smooth depth on the other side was worse off significantly The use of nozzle oscillation to increase the smooth depth of cut and surface finish has been confirmed when cutting mild steels and aluminium alloys [3,42] and fiber-reinforced composites [49]
Clearly, the nozzle oscillation technique can also increase the total depth of jet penetration; however, in earlier studies using relatively large oscillation angles, this advantage was not significant
so that no quantitative data were reported [3]
Fig 3 AWJ and workpiece interface trace profile at traverse speed of 0.5 mm/s and water pressure
of 276 MPa: (a) traditional AWJ cutting method; (b) with nozzle oscillation at
oscillation angle of 2o and oscillation frequency of 2Hz [42]
Studies also include the mechanisms under which nozzle oscillation cutting improves some cutting performance measures It has been reported [3,42,50] that nozzle oscillation cutting creates a scanning cutting action by the particles which not only reduces the particle interference, but also clears the target surface for more effective cutting by subsequent particles It is also believed [12] that the scanning action has changed the particle attack angles on the target surface which changes the material erosion mechanisms resulting in an increased cutting performance A visualization study has been carried out to explore the underlying mechanism in which nozzle oscillation can increase the depth of cut [42] This study was carried out on a transparent plexiglass The kerf formation process and jet (or particles)-work interference were recorded by a high speed video camera Some typical particle-work interface traces are shown in Fig 3 It has been found that the successive traces of the particles on the cut surface with nozzle oscillation is steeper than those without oscillation, which results in more particle energy in the cutting direction for deeper cuts