Machining of High Strength Steels With Emphasis on Surface Integrity by air force machinability data center_1 pot

In 1938, Ernst in his book on the ‘Physics of Metal Cutting’ , defined machinability in the follow-ing manner: ‘As a complex physical property of a metal involving: • True machinability,

7

Machinability and Surface Integrity

‘It is common sense to take a method and try it

If it fails, admit it frankly and try another

But above all, try something ’

FRANKLIN DELANO ROOSEVELT

(1882 – 1945) [32nd President: United States of America]

7.1 Machinability

Introduction – an Historical Perspective

Today, greater emphasis is being placed on a

compo-nent’s ‘machinability’ , but this term is an ambiguous

one, having a variety of different meanings, depending

upon the production engineer’s requirements In fact,

the machinability expression does not have an

author-itative definition, despite the fact that it has been used

for decades In 1938, Ernst in his book on the ‘Physics

of Metal Cutting’ , defined machinability in the

follow-ing manner:

‘As a complex physical property of a metal involving:

• True machinability, a function of the tensile strength,

• Finishability, or ease of obtaining a good finish,

• Abrasiveness, or the abrasion undergone by the tool

during cutting.’

By 1950, Boulger had summarised these criteria more

succinctly in his statement: ‘From any standpoint, the

material with the best machinability is the one

permit-ting the fastest removal of chips with satisfactory tool

life and surface finish.’ This ‘Boulger definition’ leaves

some unanswered questions concerning

chip-form-ing factors, cuttchip-form-ing forces and, has little regard for

either the physical and mechanical properties of the

material, nor potential sub-surface damage caused by

the cutting edge By 1989, Smith made the point that

in fact machinability, had to address these properties

and the word ‘metal’ should be substituted by the

ex-pression ‘material’ , in a combined general-purpose

definition, as follows: The totality of all the properties

of a work material which affect the cutting process and,

the relative ease of producing satisfactory products by

chip-forming methods.’ Even these definitions still lack

sufficient precision to be of much practical use and by

1999, Gorzkowski, et al., in their powder metallurgy

paper concerning ‘secondary machining’, entitled:

 ‘Secondary machining’ , is a term used to cover any additional

post-machining operations (e.g drilling, turning and milling,

etc.), that has to be undertaken on powder metallurgy (i.e

sintered’) compacts, after compaction and sintering

Nor-mally, these post-sintering production processes, are only

car-ried out to ensure, say: a good turned registered diameter, a

precision cross-drilled hole, precise and accurate screwthread,

an undercut, or similar* – as this is a last resort, as it

adds-value to the overall component’s cost.

‘Machinability’ , stated that: ‘Machinability is a difficult

property to quantify.’ Why is this so? It is probably is

a combination of many inter-related factors, such as: chemical composition of the workpiece, its micro-structure, heat-treatment, purity, together with many more effects which influence the overall machining operation In Fig 144, this diagram attempts to high-light some of the important factors that affect a com-ponent’s machined state – its ‘machinability’ Although even here, an important factor such as power con-sumption is missing, showing that this is by no means

an exhaustive flow-chart of the complex mechanisms that exist when a material is subjected to machining This is probably why it is virtually impossible to state that one, or another material after machining, was ei-ther a ‘good’ , or bad’ one to machine By utilising some

‘impartial and objective testing program’ , it may be

pos-sible to ‘rank’ prospective or current materials, or pro-duction tools – in some way, perhaps by way of a

‘De-sign of Experiments’ (DoE), in combination with ‘Value Analysis’ (VA) approach to the production problem

This strategic technique to the problems of ‘machin-ability comparisons’ of differing factors will shortly be mentioned in more detail, after a brief resumé on just some of the machinability testing techniques favoured today

7.1.1 Design of Machinability

Tests and Experimental Testing Programmes

Over the years, a range of machinability testpieces have been developed – more on this shortly – that are used to assess specific cutting conditions found when machining the actual production part The assessment

of a material’s machinability can be undertaken by two

groups of tests, these are machining and

non-machin-ing testnon-machin-ing programmes The former machinability

group, can be further sub-divided into either ‘ranking’ and ‘absolute’ tests and, it should be mentioned that the latter non-machining tests fall into the ranking category Often, ‘ranking’ tests are termed ‘short

*Powders when they fill the dies and are compacted, cannot

reproduce component features at 90° to the major pressing

di-rection – hence, the powders cannot readily move sideways –

as such, features, like: screwthreads, transversal features (i.e undercuts, etc.), must be machined afterward, hence, the term

‘secondary machining’.

Figure 144 The major factors that influence a machined component’s condition

.

tests’ , conversely ‘absolute’ tests are known as ‘long

tests’ By their very nature, the ‘short tests’ merely

in-dicate the relative machinabilities of two, or more

dif-ferent combinations of tool and workpiece Whereas,

the ‘long tests’ can produce a more complete depiction

of the anticipated conditions for various combinations

of tool and workpiece, but as their name suggests,

they are more time-consuming and costly to develop

and perform Some of these test regimes are briefly

reviewed below, but more information can be obtained

from the listed references at the end of this chapter

‘Ranking’ Machining Tests

A series of these ‘ranking’ tests for fast assessment of

actual production conditions has been devised over

the years and some will be mentioned below, but this is

by no means an exhaustive account of all such testing

programmes, they merely indicate the relatively

well-tried-and-tested techniques, such as:

• ‘Rapid facing test’ – this consists of a turning

op-eration, requiring facing-off a workpiece, preferably

having a large diameter, using an HSS tool  The

machinability is assessed by the distance the tool

will travel radially-outward, from the bar’s centre,

prior to its catastrophic tool failure This

‘end-point’ as it is known, is compared with a similar

trial, where the distance for tool failure by using a

reference material was previously determined,

NB Although the ‘Rapid facing test’ quickly assesses

one particular test criterion that a machinability

rating can be based upon, it suffers from a number

of limitations Firstly, the material’s diameter may

be smaller than that which one would ideally prefer

to use for the test Secondly, if the workpiece

mate-rial’s structure is not homogeneous, then this test

only indicates properties over the diameter-range

 ‘HSS tool material’ is utilised, because under these extreme

machining conditions, it will rapidly promote catastrophic

tool failure as the forces steadily increase together with

esca-lating tool interface temperature, as the tool’s edge is fed

radi-ally-outward during the subsequent facing operation.

 ‘Reference materials’ , are normally those workpiece materials

that are considered to be ‘easy-to-machine’ , as their name

sug-gests they, at the very least, give a ‘base-line’ , or datum, for

some form of machinability comparison.

 ‘Homogeneity of material’ , refers to a uniformity of its

micro-structure and having isotropic properties.

used This latter problem of lack of homogeneity of the workpiece material, can be somewhat lessened

by boring-out the material at the workpiece’s cen-tre, prior to commencing the test

• ‘Constant-pressure test’ – this is quite a popular

testing technique and can be undertaken by a va-riety of methods of machining assessment For example, in turning, machinability is measured by utilising predetermined geometry in association

with a constant feed force The technique has been

used to some effect on the machining of free-cut-ting steels This test is essentially a measure of the friction between the chip and tool, which is re-lated to the specific cutting temperatures generated whilst machining, together with its effects on the tool’s wear-rate,

NB Normally a turning centre has a constant feed

force, in order to obtain relevant data An engine-/ centre-lathe can also be employed to acquire iden-tical data, but a tool-force dynamometer is used

to measure this feed force, then plotting a graph

of this feed force with its associated frictional ef-fects, but this requires more effort and takes longer Similar constant pressure tests can be employed for drilling processes

• ‘Degraded tool test’ – consists of workpiece

ma-chining with a softened (i.e degraded) cutting tool

The test’s ‘end-point’ is determined either: when a specified amount of tool flank/crater wear has been reached, or at catastrophic tool failure,

NB If machinability testing is carried out on softer

and more easy-to-machine materials – typically on various alloys of brass, then just a small variation in softening the tool steel prior to cutting, has a dras-tic effect on the results obtained, but for harder-to-machine materials this effect is significantly less-ened

• ‘Accelerated cutting-tool wear test’ – as an

alterna-tive to deliberately softening the tool (i.e Degrad-ing tool test), in order to speed-up the

machinabil-ity process the cutting speeds are increased If the

cutting speeds are significantly increased, the tool

will not behave according to the predictable tool life

equation – due to the artificially-elevated cutting

temperature generated

NB It is not prudent practice to extrapolate

tool-life data beyond that actually obtained during

test-ing in order to obtain quantitative information

about other ranges and conditions, with differing

operations and parameters As a result, this test is

usually classified as a ‘ranking test’

‘Ranking’ – Non-Machining Tests

Whenever there seems to be a need to experiment with

material cutting using perhaps one of the techniques

just mentioned, it is important to establish whether

any savings gained will be recouped in the actual

pro-duction operation If a company is unsure of the likely

cost benefits of such testing, then a strong case can be

made not to test the material at all! Fortunately,

non-machining tests exist that can be utilised in these

doubt-ful situations, rather than ‘working blindly’ – with no

relevant cutting data, to base the applied cutting

con-ditions upon Several of these ‘ranking’ non-machining

tests can be employed, such as:

• Chemical composition test – a variety of tests have

been developed by which workpiece materials are

‘ranked’ according to their primary constituents It

is obvious that the results from such tests are only

relevant when materials of similar type, having

identical processing conditions/thermal history,

are to be machined

 Taylor’s tool life equation(s), has been utilised for many years,

to determine the ‘end-point’ of a cutting insert’s useful life,

under steady-state cutting conditions The basis of the general

formula: VcTα = C, has been modified and expanded to obtain

an equation for the ‘economical cutting-edge life’ for a

speci-fied feed, as follows:

Te = (1/α – 1)(C´t/C´m + tc)

Where: Te = economical tool life, α = slope of the VT-curve (i.e

measured from a plotted graph), C´t = cutting-tool cost per

cutting edge (i.e see ‘Machining costs’ – later in the chapter),

C´m = machine charge per minute (i.e normally established by

the machine shop management), tc = tool-changing time for

the cutting operation – this will vary according to whether the

tooling is of the conventional, or quick- change type.

 ‘Thermal history’ , refers to the heat treatment thermal cycle

that the component in question was processed, describing the

time at temperature, with any modifications to the

tempera-ture-induced regime on the heat-treated part.

NB Given the above limitations, these tests have

proved to be quite valid and successful for screen-ing a workpiece material prior to actual machinscreen-ing Typical examples of this test type, rank materials using a V0 scale – giving cutting speeds in m min–

and the machinability index of 100 (i.e utilised by the ‘Volvo test’ – not shown) A close correlation be-tween the chemical composition test and ‘absolute tests’ has been obtained with accuracies claimed

to within 8% For example, the relationship be-tween chemical composition and cutting speed is: Cutting speed (V0) = 161.5 – 141.4 × %C – 42 – 4 ×

%Si – 39.2 × %Mn – 179.4 × %P + 121.4 × %S

• Microstructure tests – are principally concerned

with the type of microstructure present in say, a steel workpiece, specifically: inclusion type, shape and dispersion The test method gives a good in-dication of the likely machinability, but requires highly-specialised laboratory equipment for such a metallographical investigation although materials can only be ranked, as either: good, bad, or indiffer-ent

NB Early work here, primarily investigated

low-to-medium carbon steel microstructures, notably considering the spacing between pearlite laminae achieved by heat treatment The pearlite-to-ferrite proportions clearly influenced the materials hard-ness value (e.g Brinell) When a cutting speed was selected (e.g V80), a machinability rating could be obtained for either life at: a constant speed (min-utes), or relative speed for a constant tool life (m min–) It has been observed that when >50%

pearlite was present, combined with a relatively high bulk hardness, then good machining characteristics

occurred In recent years, commercially-available

steels have trace elements added to aid machinabil-ity, the so-called free-machining steels Typically, sulphur and manganese additions, create manga-nese sulphide, with their shape, size and distribu-tion within the steel’s matrix, playing a major role

in aiding machinability factors

 ‘Bulk hardness’ , is a term that is used to state the overall

hard-ness of the test specimen, not its micro-hardhard-ness – which only

establishes localised hardness levels.

• Physical properties test – requires specialist

equip-ment in order to perform this test The physical

properties of the workpiece material are utilised in

order to determine its machinability ranking

NB Researchers, have produced a general

machin-ability equation using a dimensional analysis

tech-nique and, by utilising conventional test methods

to establish and measure its: thermal conductivity,

harness (Brinell), percentage reduction in area,

together with the test sample’s length This

‘Physi-cal properties test’ , gives close agreement with the

V0 cutting speed for a range of ferrous alloys,

al-though when brittle materials are assessed, the lack

of a yield-point8 and the much smaller reductions

in area – after tensile testing – may cause potential

ranking problems

‘Absolute’ Machining Tests

As their name implies, the ‘absolute tests’ are utilised

in order to obtain a comprehensive data-gathering

machining-based activity, on particular types of

work-piece and cutting tool combinations Many of these

‘absolute testing’ techniques have been devised, with

several of them listed below, including the:

• Taper-turning test – being undertaken by turning a

tapered workpiece As a result of turning along the

taper, the cutting speed will proportionally increase

with increasing taper diameter – this also being in

proportion to the cutting time By originally

estab-lishing the cutting speed, the changing-rate of the

 ‘Yield-point’ , refers to the strain* at which deformation

be-comes permanent, when the material is subjected to some

form of mechanical-working The yield-point strain for

fer-rous and many ductile materials is well-defined, illustrating a

‘sharp’ transition from elastic-to-plastic deformation – where

a permanent ‘set’ occurs However, this is not the case for

many brittle materials, here when say, a tensile test is

con-ducted, an artificial ‘proof-stress’ value is used to intersect the

stress/strain curve plotted, to establish its safe-working level

of operation – see the relevant References for more in-depth

details.

*‘Strain’ , is a measure of the change in the size, or shape

of a body – referring to its original size, or shape For

ex-ample, linear strain is the change per unit length of a linear

dimension – after some form of mechanical working For a

tensile test specimen that has been subjected to a tensile test, it

refers to its linear dimensional change from its original gauge

length.

cutting speed in conjunction with the amount of tool flank wear – for two separate tests – allows the

values of the constants (i.e.‘α’ and ‘C’) in Taylor’s

equation for tool wear – see Footnote 5 – to be de-rived and, the tool life established for a range of fu-ture cutting tests As the DOC must be consistently maintained throughout the test, either a CNC pro-gram must be written – using one of the standard

‘canned-routines’ available, or a taper-turning at-tachment is necessary on an engine-/centre-lathe,

NB Some major advantages accrue from this

com-prehensive testing technique, not least of which is that results are valid for a range of pre-selected cut-ting speeds and, the test is of relatively short du-ration, but closely agree with many thorough and longer test methods Although, the results obtained may not be representative of actual cutting condi-tions, owing to the fact that the cutting tool, ma-chines at differing diameters throughout the taper turning test

• Variable-rate machining test – achieves similar

results to the previously described ‘Taper-turn-ing test’ In this case, the increase in cutt‘Taper-turn-ing speed

is obtained by turning a parallel testpiece axially, whilst simultaneously increasing the cutting speed

as the tool traverses longitudinally along the work-piece Once again, the constants are derived for the

‘Taylor equation’ after a minimum of two tests have been completed,

NB The main advantages of this method over the

‘Taper-turning test’ , are that a standard testpiece can be used and the results probably reflect truer actual turning conditions – in that consistent diam-eters are being turned, although this argument is somewhat debased, if the turning of complex free-from component geometry is demanded for the production part

• Step-turning test – was developed to overcome

some of the problems associated with the two pre-viously described testing techniques In the ‘Step-turning test’ method, a range of discrete diameters and speeds are utilised to determine the ‘Taylor’s constants’ This test, shows close agreement with re-sults obtained from the two previously-mentioned

‘absolute test’ methods,

• HSS tool wear-rate test – this test assesses

machin-ability by measurement of the tool’s flank wear,

pro-duced when machining free-cutting steels, with the

major parameters being the elemental additions to

the metallurgical composition of these steel grades

NB These tests are undertaken in a similar manner

to the: ISO 3685:1977 Standard, for a long ‘absolute

test’ , but it was withdrawn in mid-1984

All of the above ‘absolute testing’ programmes, relate

to turning operations, principally due to the fact that

the tool is engaged in the workpiece test sample for a

reasonably lengthy period of time This tool/workpiece

engagement, allows for ‘steady-state’ conditions to be

developed, having the additional benefit of producing

relatively consistent ‘Taylor constants’ From a more

practical viewpoint, the author has developed some

other testpieces, which have proved somewhat useful

in actual industrial machining applications, where a

more representative machinability situation was

de-manded Just some of these testpieces, along with a

discussion of their relative merits, will now proceed

Practical Testpieces – for CNC Applications

The premise behind the development of the testpiece

depicted in Fig 145, was to attempt to ‘mirror’ the

ac-tual production operations and to a lesser extent, the

physical geometry of a particular component part

Here, the component geometry was devised to be

ma-chined on either a machining centre, or a turning

cen-tre with the facility of driven tooling and at the very

least, having an indexing workholding spindle/chuck

With this testpiece, the part is preferably a thick-walled

tube that can be bored out, OD turned, circular

inter-polated (i.e milled), drilled and tapped – as the

drill-ing size, is also an M6x1 tappdrill-ing size This allows the

component’s geometric features to be inspected

‘On-machine’ – using metrological inspection routines in

association with touch-trigger probes and,

‘Off-ma-chine’ employing a CNC Co-ordinate Measuring

Ma-chine (CMM) These identical parts were from a series

of exhaustive tests undertaken on both ferrous metals

and aerospace-grade aluminium stock Of particular

note, was that when a milled circular interpolated

fea-ture – the boss, was assessed on the machining centre,

it gave more accurate readings than its equivalent

in-spection routine on the CMM This perceived

differ-ence in accuracy and precision, was the result of part

changes caused by both relaxation of the clamping

forces – upon release – and the greater temperature

differential between these workpieces when inspected

on the CMM However of note, was the fact that in general for the inspection of part features, the CMM showed a four times improvement in repeatability, to that of the touch-trigger probing undertaken on the machine tool, as the following Table 9 indicates: The above type of practical ‘testing regimes’ are

gen-erally termed: ‘Production Performance Tests’ (PPT)

Typically, these PPT’s can be utilised to determine the maximum production rate – in parts per hour Al-though it must be said, that with shifts normally con-sisting of between 6 to 8 hours duration of potential

‘in-cut time’ , this to a certain extent, limit’s the achiev-able machined surface finish requirement, particularly

if a ‘Sister tooling strategy’ is not operated One of the main problems connected with PPT’s, is that invari-ably free-cutting metals are usually selected for long-term testing, meaning that any wear-related data takes awhile to accrue Despite this slight reservation, actual cutting data can be employed, which represents almost optimum machining conditions, leading the way to

Table 9 A comparison of the machined component

tes-tpiece accuracies by either: ‘On-’ , or ‘Off-machine’ inspection procedures

- equipped with Renishaw touch-trigger probes:

Machining Centre

Scope Full range of: X-, Y- and Z-axes Direction of test Uni-directional

Positional Accuracy ±13 µm X-axis ±8 µm

Y-axis ±5 µm Z-axis ±6 µm

* Machine tools here, are part of a fully-industrial Flexible Manufac-turing Cell (FMC), comprising of Cincinnati Milacron equipment: 200/15 Turning Centre, 5VC Vertical Machining Centre, T3 776 Ro-bot- equipped with twin back-to-back grippers – for component loading/unloading, LK Micro4-CMM, DeVlieg Tool Presetting Ma-chine, Component workstation, Cell Controller, all equipped with Sandvik Coromant quick-change tooling (Block Tools and Varilock Tooling), plus DNC-link to a CAD/CAM workstation – being desi-gned and developed by Cincinnati Milacron and the Author, when acting as an Industrial Engineering Professor at the Southampton Solent University.

.

Figure 145 General machinability test piece for CNC machine tools.

NB Holes marked ‘A, B and C’ are machined at different cutting speeds, as are the turned, bored and milled dimensions.

.

‘full’ production operational machining, meaning that

with some degree of confidence, manufacturing

dic-tates and objectives will be met

In Fig 146, a commercial (PPT) testpiece has been

developed showing typical machining data employed,

based upon the secondary machining operations

de-manded by many companies on Powder Metallurgy

(P/M) components – where light finishing cuts, or

ac-curate and precise screwthreads are demanded Here,

the cutting insert can turn three different diameters

– usually in some form of arithmetic progression, so

that feedrate longitudinally can be metrologically

as-sessed Moreover, the insert’s passage over the surface

can be metallographically-inspected and a

micro-hardness ‘footprint’ across a tapered section can be

undertaken, to see if any surface/sub-surface

modifica-tions have occurred More will be said on this subject

later in the chapter, when discussing the effects of

‘ma-chined surface integrity’ This design of using a

thick-walled tube (Fig 146), that can be produced from

ei-ther wrought stock, or P/M compact processing – the

latter, giving a controlled ‘density’0 across and along

the part, makes it particularly ‘ideal’ for any secondary

machining machinability trials Boring operations can

also be conducted on such a testpiece geometry,

al-lowing roundness parameters and its associated

‘har-monic profile’ to be metrologically assessed, in

conjuc-tion with any ‘eccentricity’ with respect to the OD and

 ‘Arithmetic progressions’ , are normally utilised for many

ap-plied machining (PPT) trials as they give a ‘base-line’ for the

research work and increase at a controlled amount For

ex-ample, a feedrate, could begin and increase as follows: 0.1, 0.4,

0.7, 1.0, 1.3, … mm rev– – with the ‘common difference’ being

3 As a mathematical expression, this simple arithmetic

pro-gression, can be written as follows:

a, a+d, a+2d, a+3d, a+4d, a+5d, … where the ‘common

differ-ence’ is ‘d’ , giving the:

n´th term as: a+(n–1)d.

0 ‘P/M Density’ , refers to either the uncompacted, or

free-par-ticulates and is termed its ‘Apparent density’ (AD) This term

AD, is used to refer to the loose material particulates prior to

PM compacting, to describe the density of a powder mass

ex-pressed in grammes per cubic centimetre of a standard volume

of powder This AD differs from that of its ‘compacted density’

– which will vary depending upon the consolidation (i.e

com-pacting) technique utilised For example, double-compaction

– pressing the powder in the dieset from both ends, or

us-ing ‘floatus-ing diesets’ – to simulate double compaction, in this

latter case, pressing from one end only, will produce a more

uniform bulk density throughout the ‘green compact’ as it is

known – prior to its subsequent sintering process

ID – these machined surfaces both being produced in

a ‘one-hit machining’ operation – then inspected by a suitable roundness testing machine

The main advantage of using industrial-based (PPT) testpieces similar to that shown in Fig 146, is

that ‘canned-cycles’, can be used to produce the un-dercuts, turning passes, or screwcutting operations

on each part Moreover, optional ‘programmed-stops’

can be written, allowing the research-worker/operator,

to have the facility to stop machining at a convenient point as desired, at the press of a button – giving a measure of control to the automated CNC machining processes If a series of testpieces are to be machined, it

is important that all of the parts machining sequences are known and that they are laid-out in a consequtive logical fashion This allows one to measure the dete-rioration with machining time for the sequence of tes-tpieces produced To this end, not only should some unique and logical part numbering system be used, but in the case of P/M testpieces, the top and bottom for each compact should be established As when each one was initially compacted, its local density have var-ied and, for consistency for all machining undertaken with each test piece, it needs to be held in the same orientation

Often it is possible to amalgamate two previous ranking machining test regimes into one, this is the

case with ‘Accelerated Wear Test’ (AWT) illustrated

in Fig 147, this test being a combination of both the:

‘Rapid Facing’ and ‘Degraded Tool’ tests – previously

described In the case of the AWT technique, this

hy-brid test’s aim is to assess the relative machinability of either wrought, or secondary machined P/M compacts

 ‘Canned-cycles’ , this is a preset sequence of events that is

ex-ecuted by issuing a single command, which may remain active throughout the program, or in this case will not, for a par-ticular ‘canned-cycle’ * For example, once the preset values/ dimensions together with the required tool offsets have been

established, then a preparatory function entitled a ‘G-code’

can be used, such as a G81 code, which would initiate a sim-ple drilling cycle, in association with the following G84 code which would then specify a tapping cycle on this drilled hole,

or alternatively, a G32 code commences a threading cycle and

so on – which considerably reduces both the complesaty and overall length of a CNC program

*G-codes fall into two categories, they are either ‘modal’ , or

‘non-modal’ A ‘modal’ G-code, remains ‘active’ for all

subse-quent programmed blocks, unless replaced by another ‘modal’ G-code Conversely, a ‘non-modal’ G-code will only affect the

programmed block in which it appears.