7.10 Surface Integrity of Machined Components – Introduction Previously in Section 7.5 concerning machined sur-face texture, the discussion was principally concerned with the resultant
Trang 1the machining operation, but it also provides the user
with validated and relevant data analysis These
posi-tive benefits enable tool designers and users alike, to
design and develop advanced cutting tools and to
un-dertake efficient and optimised machining operations
Beyond the positive advantages of tool optimisation,
simulation can significantly reduce tooling develop-ment costs and lead times to bring a newly-developed product to market The role of machining simulation
is likely to rapidly grow, as more tooling and produc-tion engineers become aquainted with these software packages
Figure 184 The insert’s cutting edge: illustrating the ‘rounding effect’ (exaggerated) or, a manufacturer’s ‘edge preparation’ and
the material flow conditions that arise as a result
.
Trang 27.10 Surface Integrity of
Machined Components –
Introduction
Previously in Section 7.5 concerning machined
sur-face texture, the discussion was principally concerned
with the resultant surface topography where the
topo-graphical information was valid, but disguised the fact
that potential sub-surface material layers might have
compromised and altered the machined component
The concept of the overall functional performance of
a surface and its accompanying sub-surface condition
was recognised by Field and Kahles (1971), where they
used the term ‘Surface Integrity’ to describe its
poten-tial state The overall concept of surface integrity ant
its various generating mechanisms in conjunction with
the production process is known as the ‘unit event’ 8
This unit event has now been reclassified into five
discrete generating mechanisms: chemical,
mechani-cal, mechano-thermal, thermo-mechanical and
ther-mal – the order they are listed reflects their respective
power density per unit area For example, increases in
the power density from the chemical end of the series,
results in an augmented level of thermal energy
enter-ing the surface leadenter-ing to greater thermal damage and
poorer part surface integrity The chemical mechanism
is dominant across all classes of production process to
some degree and that surfaces react with their
imme-diate environment, via absorbates, oxidation, etc., as
illustrated in Fig 185 – more will be said on these
ef- ‘Unit event’ , is a complex interrelated series of reactions with
the potential for distinct zones to be present within the
sur-face vicinity, including a:
Chemically affected layer (CAL) – resulting from chemical
surface changes by the production process, or from
post-production exposure to a local environment,
Mechanically affected layer (MAL) – this may be due to
factors such as material bulk transportation: deposits; laps;
folds and plastic deformation,
Heat affected layer (HAL) – principally concerned with
factors such as: phase transformation; thermal cracking and
retempering,
Stress affected layer (SAL) – is in the main, the result of
residual stresses being a combination of the above (Field
and Kahles, 1971)
–
–
–
–
fects when discussing the machined surface condition
in the following section
7.10.1 Residual Stresses
in Machined Surfaces
A machined surface is the product of either ‘abusive’ ,
or ‘gentle’ machining regimes, these being the direct
result of the cutting process and its chosen machining data Thus, machining being a complex relationship of many interrealated factors, affects the outcome of the production process – see Fig 144 Here, a simplistic schematic diagram attempts to show the complexity
of a machining operation, with the surface integrity grouping indicating for a turning operation the fol-lowing features:
• Surface condition – surface texture and its
associ-ated roundness,
• Micro-structural changes – micro-cracks,
disloca-tions and fissures, etc.,
• Surface displacement – bulk transportation of
ma-terial and residual stresses,
• Surface/sub-surface micro hardness – plastic
de-formation and localised residual stress layers Machined surfaces are even more complex than seem-ingly at first glance, as their performance can be
in-fluenced by either external layers (chemical transfor-mations and plastic defortransfor-mations) and internal layers
(metallurgical transformations and residual stresses)
By way of example, the anisotropic – periodic – longitudinally turned surface illustrated in Fig 185,
is affected by the cutting insert’s tool tip geometry and the regularity of the cusps (i.e peaks and val-leys) – the surface topography being dominated by the pre-selected feedrate A series of other micro-tech-nological features can also occur, these often being superimposed onto the machined surface, typically the result of: tool wear, vibrational influences and to
a lesser extent, machine tool-induced errors In the
circumferential direction the ‘Lay’ is both periodic
and regular, albeit this round generated surface by the turning operation, will probably have some form
of harmonic effects present: departures-from-round-ness characteristics (i.e a combination of harmonic influences present) The exposed sterile surface (Fig 185), is the result of highly localised temperatures and transients, which when turned the machined surface will be instantaneously oxidised and adsorb
contami-360 Chapter 7
Trang 3Figure 185 The cross-section of an anisotropic (i.e periodic) surface, illustrating surface contaminants (oxides and adsorbates),
together with some sub-surface plastic deformation (the residual stress zone) and an unaffected substrate
.
Trang 4nants The outermost adsorbate layer is often termed
the ‘Beilby layer’ 8: ≈1 µm in thickness and consisting
of many complex factors Notably, this ‘layer’ would
more than likely have hydrocarbons present and
wa-ter vapour, that originated in the coolant, or the
at-mospheric environment, respectively Underneath
this metallic surface for work-hardening materials,
there is normally a plastically-strained region that has
usually been metallurgically altered The depth of this
strain-hardened layer will vary somewhat, but it is in
the region of 10 µm, its actual thickness is dependent
upon the amount of plastic deformation induced by
the tool’s passage over the surface and is influenced by
the metallic substrate’s composition The plastic
defor-mation and work-hardening depths8, can penetrate
to fractions of a millimetre this is particular true, if a
‘wiper-insert‘, or roller burnishing tools is employed to
purposely create this localised hardened region to the
component’s surface
Residual Stress Deformations
For any residual stresses acting within a body (i.e
component), they will occur without any external
forces, or moments Internal forces form a system that
is currecntly in a state of equilibrium and if portions
are removed – by machining, the equibrium status is
normally disturbed, resulting in potential component
deformation This effect of machining distortion is
well-known to practising industrial engineers, when,
for example, machining just one side of a thin
compo-nent, this operation will cause a partial release of local
residual stresses causing it to bend and bow If either a
casting, or forging has not been heat-treated for stress
relief and its needs asymmetrical machining (i.e on
one side only), it is likely to deform after unclamping
restraint from its work-holding device on the machine
tool In an attempt to minimise this distortion created
by residual stress release, an experienced machinist will
release the clamping forces after roughing cuts so that
‘Beilby layer’ , on the machined surface is ‘practically
amor-phous’ – this condition being proposed by Sir George Beilby
around the beginning of the 20th century.
As an approximation, the depth of hardness penetration is
ap-proximately 50% to that produced by residual stress
penetra-tion, whereas the observational plastic deformation is about
50% greater than this penetration
the stressed surfaces are equalised, prior to reclamp-ing and takreclamp-ing a finish pass If this unclampreclamp-ing and then re-clamping activity is not possible, components clamped in-situ on the machine tool are occasionally vibrated at their natural frequency, to minimise these induced residual stresses Component deformation is roughly proportional to the removed cross-section of workpiece material Any further finishing is usually concerned with just a light cut to minimise any detri-mental effects resulting from residual stresses by a pre-vious production processing operation, or route The release of internal residual stresses must not be confused with the input of such stresses by machin-ing, as indicated in Fig 186b The machining process generates residual stresses by plastic deformation (Fig 187a), or from localised metallurgical transforma-tions In Fig 186a, the residual stress effects influence
a range of mechanical and physical properties of the workpiece material, such as:
• Deformation – this point has been alluded to above
and can create problem with small workpiece cross-sections,
• Static strength – is affected by the yeild point of the
workpiece material, which in turn, is influenced by the presence of residual stresses,
• Dynamic strength – of the part in-service can often
have its fatigue strength and life affected by the in-fluence of residual stresses present,
• Chemical resistance – if certain metals are
sub-jected to induced residual stresses on exposure to
atmosphere over a period of time, then stress corro-sion may occur,
• Magnetism – residual stresses present, can affect a
component’s magnetic properties, creating distur-bances of the crystalline structure
Taper-Sectioning and Micro-Hardness Assessment
So that an improvement of metallographical inspection
of a sectioned machined surface can be made without unduly affecting any form of surface distortion, ‘taper-sectioning’ has often been utilised A tapered-section (Fig 187b), allows such sub-surface features as: phase transformations; plastic flow zones; localised cracking; bulk transportation and redeposit of material; to be in-vestigated which would otherwise have been missed, if only profilometry (i.e surface topography assessment) had been undertaken
As its name implies, a taper-section overcomes the limitation of perpendicular sectioning By taking an
362 Chapter 7
Trang 5angular planar slice through the components
cross-section, this modified cut angle enhances the substrate
magnification, without unduly distorting exposed
sur-face features – giving greater discretion when
observ-ing, or testing the surface topography In Fig 187b,
an 11° sectional cut improves surface discrimination
by increasing the vertical section magnification by
around five times The taper-section angle (TSA) will
thus be 79°, with the vertical magnification being
ob-tained from the following expression:
TSM = secant (TSA)
Where:
TSM = taper-section magnification, TSA = taper-section angle.
Often, the exposed sub-surface feature of interest that has been plastically deformed, or mechanically altered
is in the main quite small, somewhat less than 0.1 mm
in width If a micro-hardness indentor such as either
Figure 186 The effects of residual stress and deformations of a workpiece by machining [After:
Brinksmeier et al., 1982]
.
Trang 6the Vickers8, or the Knoop8 is utilised (Fig 187c) to
establish hardness readings in the vicinity of this
re-sidual stress zone, then more indentations are possible
using the Knoop, rather than the Vickers indentor,
giv-ing, more discrimination to the ‘foot-printing’
assess-ment A note of caution here when originally
attempt-ing to take the taper-section, is that it is quite possible
to metallurgical alter the sub-surface features, if when
taking the section too much heat is induced when
cut-ting it from the parent component This comment is
also a valid statement for the subsequent grinding and
polishing of the removed taper-section, prior to
metal-lographical/hardness assessment
Surface Condition – Being
Affected by Cutting Speed
Prior to discussing the surface and sub-surface
modi-fications to the machined part – shortly to follow, it
is worth taking a closer look at the series of
photo-micrograph images shown in Fig 188 Here, a group
of identical metallurgical composition ferrous
work-pieces was machined, but at various cutting speeds
It can be demonstrated that the role played in
affect-ing the machined surface condition, is significantly
influenced by the cutting speed, with its
accompany-ing amplification of induced temperature effects as
‘speeds’ are increased Moreover, it can also be said,
‘Vickers indentor’ , has a square-based dymond pyramid with
and indentor included angle of 136° Its indentation is defined
as: ‘The load divided by the surface area of the indentation’ The
Vickers hardness [i.e penetration] number (VPN), may be
determined from the following expression:
VPN = 2Psin(θ/2)/L
Where: P = applied load (kg), L = average length of diagonals
(mm), θ = angle between opposite faces of diamond (136°).
‘Knoop indentor’ , has complex facets to its diamond indentor,
having angle of 130° (Short diagonal) and 172.5° (Long
diago-nal), respectively This facet geometric indentor arrangement
(i.e having a diagonal ratio of 7:1), leaves a significantly
nar-rower and longer surface indentation, to that of the Vickers –
mentioned in Footnote 84 Thus, the Knoop hardness number
(KHN) has been defined by the National Bureau of Standards
(USA), as: ‘The applied load divided by the unrecovered
pro-jected area of the indentation’ The following expression relates
to the Knoop’s surface indentation:
KHN = P/Ap = P/LC
Where: P = applied load (kg), Ap = unrecovered projected
area of indentation (mm), L = length of long diagonal (mm),
C = constant – supplied by indentor manufacturer.
that a material’s properties are dependent on the strain rate, with the type and magnitude of tool wear chang-ing accordchang-ing to the cuttchang-ing speeds, so simplistically speaking:
• Low cutting speeds – wear is normally
character-ised by attrition (i.e mechanical removal of surface layers),
• High cutting speeds – here, attrition gives way to
diffusion type wear and ‘Fick’s laws’ dominate the cutting regime
NB Such ‘broad classifications’ of tool wear
mech-anisms occurring, affects the type of: surface pro-duced; chip formation and strain behaviour
In some interesting trials undertaken by Watson and Murphy (1979) – which highlight the disguised nature
of the underlying factors in surface integrity investi-gations In this practically-based experimental work, they used a cemented carbide insert on an alloy steel (Fig 188) It was found that the feedrate and DOC have
only marginal effects on the sub-surface damage to a machined workpiece, with the cutting speed being the most influential in this situation This fact has been
established in Fig 188, when a range of similar
work-piece specimens was machined with the only variable being the cutting speed, as follows:
machined at a very low cutting speed (2.6 m min–) The chip formation was discontinuous and the sur-face shows an alternating effect of both chip forma-tion and fracture, with some evidence of deposited residual BUE Here, the surface topography is the result of complex interactions by various effects, such as changes in shear angle in the contact area between the tool and chip, plus ‘straining’ causing increases in the chip thickness These phenomena produce a variety of conditions, from strain-to-cracking and visually introduces an irregular and
an alternating surface topography,
range from 11 to 59 m min–, generate a continuous chip formation It is evident from these photomi-crographs (b, c and d), that the surface texture was gradually improving as the cutting speed increased, although even at 59 m min–, there was some indi-cation of debris from re-deposited BUE here (i.e in
‘d’),
speed had been reached (112 m min– – for this
ce-364 Chapter 7
Trang 7Figure 187 The tribological action of machining and its affect on induced residual stresses and the
micro-hardness ‘foot-printing’ technique
.
Trang 8mented carbide insert grade), the surface texture
appears to be in the main, ‘good’ , with only isolated
areas of the topography exhibiting marginal
work-piece side-flow effects,
was increased to 212 m min–, then in these trials,
greater cutting insert wear-rate occurred and was
attributed to appreciable carbide edge breakdown,
although the surface topography indicated that an
excellent surface texture was present
The machined surfaces produced at the lower range of cutting speeds indicated in Figs 188 a to d, shows evi-dence of some re-deposited BUE material to greater-or-lesser extent: having broken away from original
‘BUE mass’ , then being re-deposited over several adjacent machined feed cusps (i.e see Fig 28a, fully-appreciate this effect) To obtain a better and deeper understanding of these machined surface and sub-surface effects at the extreme conditions of either very low, or high cutting speeds: Figs 188 a and f,
respec-Figure 188 Some
photomi-crographs of component surfaces machined at different cutting speeds – otherwise with identical cutting data – illustrating the surface, but not sub-surface steel’s condition [Source: Watson & Murphy, 1979]
.
366 Chapter 7
Trang 9tively, the following comments can be made When
longitudinal taper-sections were taken through these
specimens’ cross-sections, the ground, polished and
etched surfaces reveal their true substrate damage In
the case of Fig 188a, BUE was presents on the
sur-face, moreover, there was a cutting/fracture sequence
indicated with confirmation of work-hardening
hav-ing ‘layered scales’ of with cracks and crevices beneath
them Conversely, the test specimen machined at high
cutting speed (Fig 188f), there is some verification of
a ‘white-layer’ formation – which is a complex
metal-lurgical phenomena found in certain ‘abused’ ferrous
workpiece situations – more will be said on this
condi-tion shortly In fact, the ‘good’ machined surface
to-pography disguises the fact that an underlying
‘white-layer’ condition was present, having a local recorded
hardness of 860 HVPN By way of comparison, if this
same alloy steel composition had received a
‘conven-tional’ hardness heat-treatment process: heated and
water-quenched from 1200°C, then the bulk hardness
would only be approximately 700 HVPN – see Appendix
12 for Hardness Comparison Tables
From these examples of cutting speed investigative
results and the previously mentioned discussion, it is
evident that the ‘optimum’ machined surface texture
is obtained when the cutting speed is closely aligned
to that of the tooling manufacturer’s
recommenda-tions, so here in this case it is ≈112 m min–, with a
correspondingly ‘good’ surface
topography/integ-rity If the cutting speeds had been employed at the
‘higher’ cutting data (i.e 212 m min–), then one could
have been fooled into accepting this apparently
‘im-proved’ surface topography Nevertheless, underlying
this machined surface would be an unstable
sub-sur-face condition, which if used in a stressed and critical
in-service environment, it might potentially fail, by a
reduced fatigue-life – this is why the topic of surface
integrity is so important in today’s climate of potential
industrial litigation, when component failure occurs!
Surface Cracks and White-Layers
If any cracks are present at the free surface which
ex-tends into the material’s substrate, they are potential
sites for premature component failure – for highly
stressed in-service components It has been reported
in the findings of industrial enquiries into the UK
railway industry of late, that despite these railroad
tracks being precision machined and then
occasion-ally inspected by non-destructive (NDT)8 techniques – according to the maintenance schedule, instances have occurred when these rails and particular on high-speed banked corners – have delaminated This catastrophic rail delamination has caused several pas-senger trains to lose contact with the rails and crash, resulting in significant loss of life Hence, the method
of machining – ‘abusive’ – can contribute poor surface
integrity and to the susceptibility of these machined surfaces to prematurely fail In the case of milling op-erations, it has been recognised for a number of years
that up-cut milling – alternatively termed ‘conventional
milling’ (Fig 190a), can introduce a surface tensile re-sidual stress into the surface layers of a milled work-piece If this machined component is then subjected
to both an arduous and potentially fatigue-inducing environment, then the cyclical nature of continuous stressing followed by its immediate stress release, can initiate surface crack sites causing them to open-up, which could result in premature part failure Con-versely, an identical machined component that has
been ‘down-cut’ – otherwise termed ‘climb-milling’
(Fig 190b), will induce surface compressive residual stresses This surface layer with its residual stress com-pression, has invariably been shown to remain closed and thus, avoiding crack propagation and growth, when machined under identical cutting data and en-vironmental circumstances Moreover, for many years,
it has been recommended that for CNC milling appli-cations ‘climb-milling’ not only generates this favour-able machined surface compressive stress effect, but is
a more efficient cutting process and as a result, draws less spindle power In Appendix 13a and b, two useful
‘nomographs, are given to determine either the cutting data (Appendix 13a) this is related to the workpiece’s diameter and, a diagram (Appendix 13b) to obtain the spindle power from the anticipated chip area, respec-tively
In a machined surface, both craters and pits do not pose too great a fatigue problem, as they cannot
achieve the ‘critical radius’ (i.e see Footnote 67)
neces-sary to instigate a site for crack initiation at a
poten- ‘Non-destructive testing’ (NDT), is a range of ‘non-invasive’
sub-surface inspection testing techniques, typically: Eddy-current testing, Ultrasonics tests, X-ray investigation, etc., that
can, in many cases be automated for the detection of otherwise
hidden flaws in the component(s).