Jackson Mark. Machining with Abrasives

Подождите немного. Документ загружается.

Chapter 5

Surface Integrity of Materials Induced

by Grinding

L.C. Zhang

Abstract This chapter discusses some fundamentals in the surface integrity of

workpiece materials generated by grinding, including surface roughness, micro-

structural changes and residual stresses induced by mechanical loading, thermal

heating and phase transformation. The discussion concludes that different classes of

materials have distinctive material removal mechanisms in grinding and hence have

their individual surface integrity characteristics. The workpiece materials to discuss

include metals, ceramics, semiconductors and composites.

Keywords Grinding  Surface integrity  Grinding wheels  Resid ual stress 

Grinding mechanisms

5.1 Introduction

Grinding is a process that uses abrasives as the cutting edges to remove materials

from a workpiece. Generally speaking, a grinding process involves four groups of

variables that comprehensively affect the surface integrity of a ground component.

These variables are:

(a) The performance variables of a grinding machine such as its loop stiffness,

positioning accuracy, stability and thermal deformation.

L. Zhang (*)

School of Mechanical and Manufacturing Engineering, The University of New South Wales,

Sydney, NSW 2052, Australia

e-mail: Liangchi.Zhang@unsw.edu.au

M.J. Jackson and J.P. Davim (eds.), Machining with Abrasives,

DOI 10.1007/978-1-4419-7302-3_5,

Springer ScienceþBusiness Media, LLC 2011

245

(b) The grinding wheel variables such as the size of abrasives, properties of the

abrasive material, wheel baking temperature, wheel structure, concentration,

porosity, hardness and properties of bond materials.

toughness and phase transformation threshold.

(d) Grinding parameters such as the wheel speed, table speed, depth of cut, coolant

type and its supply method, dressing conditions and properties of dressers.

Although the conventional grinding of metals is mainly a mechanical process,

physical and chemical means have been extensively incorporated into the technology

nowadays to improve the grindability and quality in dealing with difﬁcult-to-

grind materials, such as ceramics and silicon. In a sense, grinding is one of the

most complex material removal processes. In addition to the complexity in processing,

applications of grinding in various areas of high technology, such as optics and

electronics have created further requirements for the quality of ground elements.

Apart from the low surface roughness and high accuracy, which are the traditional

requirements for grinding, the high technological applications of ground components

often need tailored sub-surface properties, such as distribution of residual stresses and

allowed variation of micro-structures. These imply that one must understand and

consider the principal factors that determine the surface integrity of an element sub-

jected to grinding [1].

Investigations on grinding can be grouped roughly into three categories, experi-

mental study, theoretical modelling and the combination of the two, of which each

has its merits and disadvantages. For instance, an adequate theoretical modelling

enables one to understand the major roles of a grinding parameter, while an

experimental study can often be done under limited conditions. Nevertheless, it is

difﬁcult for a theoretical model of grinding to give quantitative predictions because

assumptions and simpliﬁcations are usually inevitable in modelling. On the other

hand, real production conditions are much easier to achieve in experimental studies.

This chapter will discuss the surface integrity generation mechanisms in the

grinding of various workpiece materials – metals, ceramics, silicon and ﬁber-

reinforced composites.

5.2 Residual Stresses and Subsurface Microstructures

5.2.1 Grinding of Metals

For a metal workpiece subjected to grinding, mechanical plastic deformation, ther-

mal plastic deformation and irreversible deformation due to phase transformat ions

are the major sources of residual stress generation [1]. Zhang and his co-workers have

systematically analyzed grinding temperature and phase transformation [2–5]and

have investigated the variation of residual stresses caused by surface mechanical

246 L.C. Zhang

loading [6], thermal plastic deformation [7], coupling of thermal plastic deformation

and phase transformation [8], and by the combination of mechanical and thermal

plastic deformation [9]. A conclusion of their research is that residual stress distribu-

tions in a ground component are sensitive to the combination of grinding-induced

permanent deformation and that the change of the deformation sources can change the

nature of residual stresses (i.e., compressive or tensile). Generally, mechanical plastic

deformation produces compressive surface residual stresses, thermal plastic deforma-

tion generates tensile surface residual stresses, but phase transformation may lead to

either compressive or tensile surface residual stresses depending on the volume

increase or decrease of the transformed material phase. Under certain grinding

conditions, plastic deformation by all the three sources can occur simultaneously in

some metals. In the following, we will discuss surface grinding and the full coupling

effect of all the sources of irreversible deformation on the variation of residual

stresses [10, 11]. The reader who is interested in the mechanisms of residual stress

generation by individual sources can ﬁnd details in [2–9].

The deformation in a workpiece during surface grinding can be considered as a

plane-strain problem, and the heat ﬂu x generated by grindin g can be approximated

by a triangular heat source, moving along the positive direction of x-axis on the

workpiece surface, as shown in Fig. 5.1. The surface mechanical stress due to

grinding can also be simulated by a moving triangular traction with its peak being

coincident with that of the heat ﬂux. In this way, the mechanical cutting of the

grinding wheel in the ﬁnite element analysis can be simulated by equivalent nodal

forces. The direction of the horizontal nodal forces depends on the type of grinding

processes, being positive for an up-grinding and negative for a down-grinding.

Meanwhile, the type of grinding processes alters the height of the heat ﬂux, q

, the

height of the mechanical stress traction, p

, and the location of the peaks of the

surface mechanical traction and heat ﬂux, l

¼ 2z

, respectively, where L

is the

contact length between the grinding wheel and the workpiece, as shown in Fig. 5.1.

In the ﬁgure, T

¼ 25



C is the ambient temperature.

The cooling effect of grinding ﬂuid can be modelled by a variable coefﬁcient of

convection heat transfer, that is,

Fig. 5.1 A theoretical model

of surface grinding with convection

5 Surface Integrity of Materials Induced by Grinding 247

hT; 2x=Lc

 1

whT; 2x=Lc



(5.1)

where q

is the convection heat ﬂux, h is the heat transfer coefﬁcient of coolant, T is

grinding generated temperature rise, w is an effective cooling factor over the

workpiece surface, with w ¼ 1 indicating a uniform convection over the whole

workpiece surface, 0  w < 1 representing a less effective cooling, and w > 1

standing for a super-cooling inside the gri nding zone. To account for the property

change of work materials by phase transformation, a proper constitutive model was

developed by Mahdi and Zhang [8] which was then integrated with the ﬁnite

element method [10, 11]. It was found that when a workpiece experiences

the critical temperature variation in grinding, phase change occurs at a certain

distance away from the grinding zone, as shown in Fig. 5.2, where Taust is the

austenite transition temperature of the work material (EN23 steel in this case). The

phase change starts earlier if the coefﬁcient of convection heat transfer, H, is higher,

regardless of the variation of table speed. Moreover, less cooling in the grinding

zone (0  w < 1) accelerates the initiation of phase transformation if H is high, for

instance, H ¼ 1.

The variation of grinding conditions does not inﬂuence the two major surface

residual stresses, s

and s

, when phase transformation occurs, as shown in

Fig. 5.3. The residual stress s

is nearly constant across the martensite zone

(Fig. 5.3b). The longitudinal residual stress s

, however, changes linearly, with a

limit, within the martensite zone. At the boundary of the martensite zone, a rapid

change of residual stresses occurs due to the sudden change of work material proper-

ties. Compared with s

, s

is more affected by the types of grinding operations, the

ratio of horizontal to vertical forces and the fully coupled thermo-mechanical

Fig. 5.2 Onset of phase change vs. heat source location

248 L.C. Zhang

grinding conditi ons with phase change. The maximum s

and s

are 1.5 and 2.6

times higher, respectively, than the initial yield stress of the work material, and

they are tensile at the ground surface. These characteristics are similar to those

of purely thermal residual stresses without a phase change [7]. However, the

coupling of phase transf ormation with thermal and mechanical deformation

brings about much higher tensile residual stresses at the workpiece surface.

The coupling of mechani cal loading with thermal loading and phase transforma-

tion has a minor inﬂuence on residual stress distributions for a wide range of

magnitude of surface traction, as shown in Fig. 5.4. Moreover, at a higher ratio of

horizontal to vertical traction m (e.g., m ¼ 0.3 in the case of Fig. 5.4b), the increase

of traction slightly decreases the longitudinal surface residual stress s

. The reason

for that is related to the reduction of the longitudinal strain within the grinding

Fig. 5.3 Residual stresses without surface cooling (H ¼ 0, l

¼ 0.25, VL

/4a ¼ 1where a is

thermal diffusivity). (a) s

/Y, and (b) s

/Y, where Y is the yield stress of the workpiece material

5 Surface Integrity of Materials Induced by Grinding 249

zone. Compared with thermal deformation and phase transformation, the effect

of mechanical traction is minor. In the case with martensite phase transformation,

the strains at the austenising grinding temperature play a central role in the

formation of tensile residual stresses.

The effect of grinding conditions on the nature of residual stresses can be under-

stood more deeply by comparing the inﬂuence of the individual causes and with

different combinations. Figure 5.5 shows the role of each mechanism of different

grinding conditions on the longitudinal residual stress s

. Under sole mechanical

conditions (Case 1), a very small, almost negligible residual stress is developed. Sole

thermal grinding conditions without surface hardening (Case 2) lead to a tensile

Fig. 5.4 The effect of mechanical traction on residual stresses (H ¼ 0, l

¼ 0.25, VL

/4a ¼ 1

and q

¼ 60 MW/m

). (a) m ¼ 0.1 and (b) m ¼ 0.3

250 L.C. Zhang

residual stress, which decreases gently with the subsurface depth. When the

phase transformation (surface hardening) is coupled, see Case 3, a surface layer with

a greater tensile residual stress is generated. The variation of s

also becomes sharper

compared with Case 2. By coupling the mechanical grinding conditions

of Case 1 with those of Case 3, a considerable decrease in residual stress occurs, see

Case 4. This means that mechanical grinding conditions may have stronger effects on

residual stresses when combined with thermal conditions only, particularly at a lower

value of VL

/4a.This is similar to the results of thermo-mechanical grinding

conditions reported by [9] when the work material properties are temperature-

independent.

The above study can be brieﬂy summarized below: (1) The surface hardening

associated with a phase change is slightly increased if cooling is more effective;

(2) The inﬂuence of mechanical grinding conditions on residual stress distribution

is minor when phase change takes place; (3) All the components of surface

residual stresses become tensile when phase transformation occurs in the

EN23 steel,which is independent of the cooling conditions and the type of grinding

operations (up- or down-grinding); and (4) The distribution of the residual stress in

the grinding direction is nearly linear in both the martensite and non-martensite

zones, but the location of the maximum residual stress is related to the depth of

martensite and surface mechanical loads. The understanding above, though from

EN23 steel only, has been important to the late development of the grinding-

hardening technology and its optimization [12–15].

Fig. 5.5 Effect of individual grinding conditions. Case 1: P

/Y ¼ 2.0, m ¼ 0.1; Case 2: q

¼ 80

MW/m

,Pe¼ 1, H ¼ 0; Case 3: Case 2 coupled with phase transformation; Case 4: Case 3

coupled with Case 1

5 Surface Integrity of Materials Induced by Grinding 251

5.2.2 Grinding of Ceramics

High surface integrity of hard, brittle monocrystalline materials with mirror-like

and fracture-free surfaces can be achieved by ultra-precision grinding [16, 17].

It has been shown both theoretically and experimentally that the mechanism of

material removal can be purely ductile when sufﬁcient independent slip systems are

activated [18, 19]. However, in the case of grinding polycrystals, certain fractured

surface areas always appear, causing the degradation of their surface integrity

[18–21]. Zarudi and Zhang [22] pointed out that defects such as pores or voids in

a ceramic workpiece, a result of the sintering process of ceramics, contribute

signiﬁcantly to the surface integrity of ground ceramic components. Alumina is

known as a material of low density with high temperature strength and high

resistance to wear and corrosion. These properties make it attractive to industry

for producing electronic and optical components that require high quality of surface

and damage-free subsurface. Thus in this section of the chapter, we will aim to

understand the formation mechanism of the surface and subsurface structures of

alumina in ultra-precision grinding. The conclusions obtained, however, will be

applicable to other ceramics of similar properties.

The materials used were polycrystalline alumina of 99.99% purity with average

grain diameters of 1 and 25 mm (Kyocera, Japan). The grinding experiments were

conducted on a precision surface grinder, Minini Junior CF CNC M286. A diamond

grinding wheel of SD4000L75BPF (diameter 305 mm, peripheral speed 27 m/s)

was used. The depth of cut was ﬁxed at 100 nm but the table speed was varied from

0.02 to 1 m/min. A water based coolant Syntilo 3 (99% water, 1% mineral oil) was

applied. The topog raphy of the ground surfaces was explored by a high resolution

scanning microscope (HRSEM), JSM-6000F. The surface roughness was measured

by an atomic force microscope (AFM) and the subsurface structure of the speci-

mens was studied by a transmission electron microscope (TEM), EM 430. Detailed

procedures can be found in [18, 19, 22].

5.2.2.1 Surface Topography After Grinding

The surface topography of ground alumina is shown in Fig. 5.6. Mirror surfaces

were obtained under all the table speeds used. Plastic grooves were clearly seen on

the ground surface, suggesting the ductile mode of material removal. However,

some pits were also observed. The variation of Rms roughness vs. table speed for

both types of alumina is shown in Fig. 5.7. It is clear that Rms roughness increased

with the increase of table speed. For the 1 mm-grained alumina, the Rms roughness

increased from 30 to 50 nm with the increase of table speed from 0.02 to 1 m/min.

For the 25 mm-grained alumina, the effect was more pronounced and the Rms

roughness varied from 33 to 90 nm. It is worth noting that the Rms roughness was

inﬂuenced by pits on the surface, as shown in Fig. 5.8 (determined by the technique

of image analysis). It is obvious that the surface covered by pits decreases

252 L.C. Zhang

signiﬁcantly with the decrease of table speed, i.e., with the decrease of the uncut

chip thickness. Furthermore, the pits in the 1 mm-grained alumi na (1.5% under the

table speed 0.02 m/min) is less than those in the 25 mm-grained alumina (5% under

the same table speed).

The nature of pits can be explored by looking into their topography. As shown in

Fig. 5.9, the central part of a pit resembles the cross-section of pores that were

formed during the specimen processing (ceramic sintering). The peripheral edge of

the pit has the characteristics of fracture. The fracture around the edge can be

explained by the effect of pore on the grinding process. Brieﬂy speaking, a pore in

the material can act as a stress raiser and promote the microcracking near the

pore edge during the abrasive cutting of the grinding wheel. Thus for a polycrystal-

line alumina with pores, it is impossible to obtain a perfect surface without any edge

mirocracking around a surface pore. The reader can ﬁnd more detailed discussions

about the fracture mechanics analysis in reference [22].

Fig. 5.6 Surface topography after grinding (Table speed 0.02 m/min): (a) the 1 mm-grained

alumina; (b) the 25 m m-grained alumina

0.02

0.1

0.2

0.5

Rms

roughness

(nm)

Table speed (m/min)

Fig. 5.7 Effect of table speed on Rms roughness. S1: the 1 mm-grained alumina, and S2: the

25 mm-grained alumina

5 Surface Integrity of Materials Induced by Grinding 253

5.2.2.2 Subsurface Structure

Figure 5.10 shows that ductile-mode grinding created a layer with an extremely

high density of dislocations, which was immediately beneath the ground surface.

This layer spreads to the depth of 100–150 nm under the surface depending on the

table speed. Following this, a zone with a much lower density of dislocations

appears. Dislocations could be distinguished easily there. The depth of the second

zone is also dependent on table speed and varies from 150 to 500 nm.

Microcracks were observed only in the vicinity of the pit edges, which are

neither radial nor lateral cracks. The microcracking-free behaviour in the ﬁrst

layer can be explained by the nucleation of sufﬁcient number of twin and slip

systems that satisfy the von Mises criterion [23]. The second zone had less than ﬁve

independent slip and twin systems and thus microcracking should be highly

possible there. However, the above experimental observations do not provide any

0.02

0.1

0.2

0.5

Surface covered

by pits (%)

Table speed (m/min)

Fig. 5.8 Effect of table speed on surface covered by pits. S1: the 1 mm-grained alumina, and

S2: the 25 mm-grained alumina

Fig. 5.9 The topography of pits (HRSEM) on ground surfaces

254 L.C. Zhang