Surface Integrity Cutting Fluids Machining and Monitoring Strategies_6 pdf

9.6.4 Artefact Stereometry: for Dynamic Machine Tool Comparative Assessments Introduction The use of machinable artefacts for the assessment of machine tools such as machining centres,

(Fig 239b) So, by the simple action of turning a hand

wheel at its end, the tools could be simultaneously

opened and closed – for the required trepanned

diam-eter This simultaneous tooling action was achieved,

by the singular rotating action of both the φ20mm by

4 mm pitch left- and right-hand (i.e M 20 × 4)

square-threaded leadscrews (Fig 239c)

One of the major advantages of an UHSM

trepan-ning operation over its equivalent turtrepan-ning counterpart,

is that the cutting forces are virtually ‘cancelled-out’ , in

a similar fashion to a conventional ‘balanced turning’

operation (Figs 41 and 238 – top right) Here in this

instance, one tool is set and positioned slightly ahead

of the other, thereby not only reducing the overall DOC,

but allowing the ‘trailing tool edge’ to effectively act as

a ‘finishing tool’ This tooling positioning strategy

pro-duced an improved trepanned surface texture, while it

significantly reduced the harmonic

departures-from-roundness, as metrologically assessed later on the

roundness testing machine Moreover, by effectively

‘halving’ the DOC, this allowed for an improvement in

the chip-streaming behaviour to be attained

In a later modification to the trepanning fixture (i.e

not shown), a large micrometer drum with its

inte-grated vernier scale was fitted in place of the knurled

adjustable hand-wheel (i.e see Fig 239a), allowing for

some considerable discretion over the linear tooling’s

diametral adjustment With such a large trepanning

fixture – having the opposing tooling widely-spaced, it

is vital that these tools are centralised directly beneath

the machine’s spindle Otherwise, there is a possibility

of both sine and cosine errors being present, creating

‘Abbé-type errors’ , when adjusting and setting these

tools for their diametral in-feed

UHSM – Trepanning Operation

This preliminary work on UHSM by trepanning, has

shown that with a suitably robust tooling fixturing and

allowing a large (indirect) range of tooling diameter

adjustment – via the twin leadscrews, then not only is

the process feasible, but it offers considerably improved

machining performance and an inherent improvement

in trepanned surface and roundness characteristics,

over vertical turning processes Possibly in a later

modification to a heavily-revised tooling adjustment

system, it might be possible to employ twin coaxial

ballscrews, with CNC servo-control, allowing

auto-matic control for machining tapers and profiling to the

workpiece – by utilising the supplementary rotary axis

control in the machine’s CNC controller Moreover,

one limitation to this UHSM trepanning technique is

the length of longitudinal cut that can be taken, prior

to the Z-axis motion causing the rotating part to foul

on the central portion of the trepanning fixture This problem can be mitigated against, by increasing the relative stand-off height of the twin-tooling from the top of the fixture by mounting each toolholder in an extended tool block, so allowing greater Z-axis feeding

to be undertaken Moreover by rearranging the tools

in relation to the workpiece, it would be possible to

‘turn’ shallow, depth internal trepanned features UHSM by trepanning offers significant advantages over ‘conventional’ vertical turning, in that, in this cur-rent work, if was found that the trepanned workpiece surface and roundness were significantly improved from the previously discussed UHSM by vertical turn-ing, described in Section 9.6.2

9.6.4 Artefact Stereometry:

for Dynamic Machine Tool Comparative Assessments

Introduction

The use of machinable artefacts for the assessment of machine tools such as machining centres, has been utilised for some of years (i.e typically: NAS Stan-dard 979: 1969; ISO StanStan-dard: 10791-7: 1997; Knapp, 1997), being developed just for this purpose Both the NAS and ISO Standard testpieces incorporated nota-ble prismatic and rotational characteristics, manu-factured to specific geometric and dimensional toler-ances, such as: at the top, an φ110 mm circular feature;

6 mm below this round shape, an 110 mm diagonal feature is cut; a central φ30 mm though-running hole

is produced; with a series of counter-bored holes at four equi-spaced quadrants are generated these be-ing situated 6 mm below the diagonal shape Taken in cross-section, the geometry of the machinable arte-facts resembles a stepped component, having an over-all height of 50 mm In fact, this type of artefact has long been employed by industry to establish the over-all machining performance capabilities of a particu-lar machine tool under test However, although this prismatic and rotational featured machinable artefact achieves some measure of conformance and indicates the likely operational performance of the machine tool, it does tend to have several significant limita-tions, such as the:

• Overall dimensional size of the artefact is quite small – when compared to that of the volumetric envelope of typical industrial machining centres,

Figure 240 Artefact stereometry, illustrating its integrated volume geometries, for a:

1 (right) conic frustum,

2 (right) cylinder,

3 rectangular volume of machine tool’s axes.

[Source: Smith, Sims, Hope & Gull, 2001]

.

• Circular feature cannot be directly compared to

that of diagnostic instrumentation – such as the

Ballbar, as the diameter of this rotational feature

differs from that of the standard Ballbar sizes,

• Weight of the artefact does not realistically

com-pare to any workspaces normally placed on the

ma-chine tool in its ‘loaded-state’ , meaning that ‘true’

machine tool loading-conditions are not directly

comparable

With these machinable testpiece limitations in mind,

it was thought worthwhile developing a new

calibra-tion strategy for such machine tools, but here, under

more realistic ‘loaded conditions’ , also this new

arte-fact being more directly comparable to diagnostic

instrumentation (i.e such as the Ballbar), but having

considerably larger volumetric size and weight, with

the capacity for reuse of the expensively-produced

precision part of the machinable artefact’s assembly

Stereometric Artefact – Conceptual Design

Stereometry has been a concept that has often been

over-looked, but it deals with the volumetric content of

a range of geometric shapes However, if this

‘volumet-ric concept’ is carefully integrated into a single artefact,

it could be employed for calibration work on machine

tools such as machining centres (i.e see Fig 240) Here,

the cylinder was represented by three machinable

aero-space aluminium disks (grade: 2017F – produced from

6 mm sheet, to nominally slightly >φ300 mm) each one

being set 100 mm apart in height (i.e disks: 1, 2 and 3)

and after machining, the disks were exactly φ300 mm

(i.e see Fig 241) The conic frustum included angle

was 22.5°, this being the result of producing 4

equi-spaced holes in each disk Starting on the bottom (disk

1), then stopping the machine and fitting the middle

disk (disk 2) and drilling the 4 holes and likewise

up-ward to the top disk (disk 3), while simultaneously

producing a 3-dimensional Isosceles triangle (Fig

241) Each disk had these individual holes being set

at an angular relationship of 90° equi-spaced apart, so,

when they are taken as a ‘volume’ , a conic frustum is

produced (Fig 242b) These geometric and

volumet- ‘Isosceles triangle’ , has two sides with two angles being equal,

but in this case, with the geometry of a right-angled triangle

NB These side lengths and associated angles can be varied,

so long as they both (i.e lengths, or angles) remain of identical

proportions.

ric relationships were intrinsically set and datumed to

a centrally-machined slot in the base of the precision

mandrel This fact, meant that the exact angular and

volumetric relationships remained in-situ, when the stereometric artefact was then taken off the machine tool for subsequent analyses

Stereometric Artefact – Machining Trials

Prior to the stereometric artefact having its machin-able disks milled, the initial test machine tool (i.e in the initial trials on a Cincinnati Milacron Sabre 500 equipped with a Fanuc OM CNC controller) was fully diagnostically calibrated by: Laser interferometry; long-term dynamic thermal monitioring of its duty-cycles in both a loaded and unloaded condition; to-gether with Ballbar assessment Prior to discussing the actual machining of the disks, it is worth taking a few moments to consider the precision mandrel that accurately and precisely locates each disk in the de-sired orientation, with respect to each other and the machine tool’s axes This mandrel body was produced from a eutectic steel (0.83% carbon), which after through-hardening to 54 HRC, was precision

cylindri- ‘Eutectic steel’ or ‘Silver-steel’ as it is generally known, due to

its almost ‘shiny appearance’ when compared to other grades

of plain carbon steels In brief, this 0.83% carbon content steel

is so-called a eutectic* steel as it relates to the eutectic com-position derived from the iron-carbon thermal equilibrium

diagram Producing an 100% pearlitic structure (i.e hence its

‘metallographic-brilliance’ , or its ‘irridescence’) when viewed under a microscope, exhibiting fine alternate layers of: FeC and Fe To harden eutectic steel, its temperature is raised slightly above the ‘arrest point’ (i.e arrest point here, equals 723°C, so hardening could be undertaken at ≈765°C) into

‘γ-solid solution’ (i.e austenitic region), then rapidly quenched

and agitated in water to prevent carbon atomic diffusion (i.e

undertaken at greater than the ‘critical cooling velocity’), with

the carbon atoms now being effectively ‘fixed’ – though not

intrinsically part – of the atomic lattice structure This carbon

entrapment, creates intense local strains that block dislocation movement Hence, the resulting structure is both hard and

ex-tremely strong, but also very brittle Microscopically, the

hard-ened structure appears as an array of random needles, being completely different from the original pearlitic structure This needle-like structure formed by trapped carbon atoms in an

iron crystal lattice is termed, ‘martensite’ Thus, the degree of

hardness – after quenching, being proportional to its lattice

strain After hardening, the mandrel needed to be tempered

Tempering is a controlled heat-treatment process to allow some of the trapped carbon to escape from the interstitial spaces between the iron atoms distorted lattice structure,

where they eventually form particles of cementite

cally-ground on the three register diameters, with the

top and bottom faces being surface ground Previous to

this heat-treatment and the grinding processes,

dow-elling datums (i.e φ6 mm) were drilled and reamed,

then 3 equi-spaced tapped clamping holes were

pro-duced for each disk, along with a ground tenon groove

in the base – all these features being orientated to the

geometry of the machines axes (Fig 241)

Several unique features are introduced within the

machinable portions of the disks, such as:

• These aerospace-grade aluminium disks were

milled to φ300 mm diameter, which directly

cor-responded to the radial path of the Ballbar (i.e see

Fig 242a) – used previously for diagnostic machine

tool assessment, ensuring that some degree of

cor-relation occurred between them,

• The three Z-plane disk heights of: 70, 170 and

270 mm (i.e modified from the original design Fig

241), coincided with both the X-Y plane table

po-sition and vertical heights utilised for the Ballbar

plots, creating a reasonably large cylindrical

volu-metric envelope (Fig 242b) Moreover, the

stereo-metric artefact was both designed and orientated to

coincide with the start and finish positions of the

Ballbar’s polar traces,

• The 4 circular interpolated holes (φ10 mm) on each

disk (i.e see Fig 241), were geometrically

posi-tioned to form a three-dimensional Isosceles

tri-angle at the three Z-axis heights for each quadrant

of these disks – with the 1st an 3rd holes relating to

the axes transition points in the X-Y planes Thus,

each of the interpolated milled holes in the face of

separate disk’s, produced the geometric

stereom-etry of a conic frustum, having an included angle of

22.5° – when the angular orientation of the middle

disk is ‘software-realigned’ to produce a

straight-line relationship (i.e see Fig 242b),

NB The temperature at which tempering is undertaken is

critical, thus between 200–300°C, atomic diffusion rates are

slow with only a small amount of carbon being released,

thereby the component retains most of the hardness So if

higher ‘soaking-temperatures’ are employed (i.e between

300–500°C), then this creates greater carbon diffusion

form-ing cementite, with a correspondform-ing drop in the component’s

bulk hardness.

* A eutectic structure is a two-phase microstructure resulting

from the solidification of a liquid having the eutectic

compo-sition: the phases exist as fine lamellae that alternate with one

another (Sources: Thelning, 1981, Alexander et al., 1985;

Cal-lister, Jr et al., 2003)

• Overall weight of the mandrel and three disk as-sembly was 38 kg, consequently, this could be considered as a realistic ‘loaded condition’ for the machine tool to operate under, from a practical sense

In order to minimise the milling forces on the ma-chinable disks, HSM was employed using a

spindle-mounted ‘Speed-increaser’ (Fig 243a) equipped with

a φ6 mm slot drill The HSM speed-increaser was oper-ated under the following conditions: 18,000 rev min–1;

at a circular interpolation feed of 750 mm min–1; with the disks having 1 mm of excess stock for each ma-chinable disk – to be milled by circular interpolation

In Fig 243a, the last machinable disk has been located and clamped and the whole mandrel-and-disk assem-bly was nearing completion, having previously had its φ10 mm quadrant-positioned holes for each disk ma-chined by small circular interpolated motions by the slot drill (i.e see the sectional details of the φ10 mm hole geometry in each disk’s quadrant co-ordinates, as illustrated in Fig 241)

Stereometric Artefact – HSM Results

After HSM by milled interpolation on the vertical machining centre, the complete artefact with its ma-chinable disks in-situ, was carefully removed from the machine tool, then automatically-inspected for its quadrant hole positions and disk diameters, on an Eastman bridge-type Co-ordinate Measuring Machine (CMM) This CMM having previously been thermally

error-mapped, then checked with a ‘Machine

Check-ing Gauge’  (MCG) – prior to artefact inspection The CMM utilised a specially-made and calibrated

 ‘Speed-increasers’ , are a means of multiplying the rotational

speed of the machine’s spindle, by utilising a fixed relationship geared head Here, this actual speed-increaser had a 3:1

gear-ing ratio, equatgear-ing to a top speed of 18,000 rev min–1, when it

is operating at the top speed for this particular machine tool (i.e 6,000 rev min–1)

NB Normally, these HSM milling/drilling geared heads are

limited to a certain proportion of running time per hour at

its top speed, as they could over-heat and thereby damage the

bearing/gearing mechanism.

 ‘Machine Checking Gauge’ (MCG), is utilised to check a

CMM’s repeatability and accuracy and to detect for any po-tential ‘lobing-type errors’ from the ‘triggering-positioning’ mechanism of the touch-trigger probe, these being invariably used on such machines

Figure 241 Artefact stereometry was designed for the volumetric and positional uncertainites on machining

centres, by: HSM interpolation of machinable disks [Source: Smith, Sims, Hope & Gull, 2001]

.

Figure 242 HSM (milling) of three machinable disks in-situ on a stereometric artefact, on a vertical machining centre [Source:

Smith, Sims, Hope & Gull, 2001]

.

cranked-probe – with its calibration obtained from

the ‘reference measurement sphere’ being located on

the CMM’s table, utilised to inspect the φ10 mm hole

geometry and there respective co-ordinate positions

This probe arrangement was swapped for a

conven-tional ‘touch-trigger probe assembly’ to measure the

machinable disk diameters – while holding the same

cartesian co-ordinate relationships as when it was

originally UHM Later – without ‘breaking-down’ ,

while still maintaining the same angular orientation,

this stereometric artefact assembly was inspected on a

roundness testing machine (Taylor Hobson: ‘Talyrond

265’) for individual disk parameters of roundness and

cylindricity assessment – for the ‘three-disk

relation-ship’ The results of all of these ‘averaged’ roundness

measurements and Ballbar polar plots are graphically

depicted as histograms in Fig 243c

When a comparison is made of these results from

three individual and completely differing

inspec-tion procedures, namely: Ballbar; CMM; and

Taly-rond, they show some degree of measurement

con-sistency individually, but less so when each disk data

is grouped For example, in the case of the Ballbar, it

indicated a 1 µm variation (i.e range) from the

top-to-bottom disks, while having a mean value of 17.5 µm

The Talyrond polar plots (i.e ‘Least Squares Reference

Circle’ : departures from roundness) also produced

consistent roundness results, ranging from <4 µm,

having a mean value of ≈14 µm Conversely, the

larg-est variability occurred with the CMM, producing a

 ‘Cylindricity’ , can be defined as: ‘The minimum radial separation

of 2 cylinders, coaxial with the fitted reference axis, which totally

enclose the measured data’ (Source: Taylor Hobson, 2003)

NB A ‘working definition for cylindricity’ , might be: ‘If a

perfectly flat plate is inclined at a shallow angle and a parallel

cylindrical component is rolled down this plate If it is a truly

round cylinder then as the component rolls, there should be

no discernible radial/longitudinal motion apparent’ (Sources:

Dagnall, 1996; Smith, 2002)

 ‘Least Squares Reference Circle’ (LSC1), can be defined as: ‘A

line, or figure fitted to any data such that the sum of the squares

of the departure of the data from that line, or figure is a

mini-mum’ This is also the line that divides the profile into equal

minimum areas

NB This LSC1 is the most commonly used ‘Reference Circle’

The ‘out-of-roundness’ , or ‘departures-from-roundness’ as it

is now known, is then expressed in terms of the maximum

departure of the profile from the LSC1 (i.e the highest peak

to lowest valley – on the ‘polar plot’) (Source: Taylor Hobson,

2003)

range of 15 µm, with a mean value of ≈23 µm Prior to discussing why the CMM results significantly varied from those obtained by both the Ballbar and Talyrond,

it is worth visually looking at a comparison between the general profiled shapes of typical ‘polar plots’ pro-duced by both these techniques In Fig 242a, a rep-resentative ‘polar plot’ from a Ballbar is shown and, likewise in Fig 242c, one from a Talyrond is depicted Their respective profiled shape geometry in terms of harmonics, is remarkably alike, illustrating the same generally similar lobed-shape combined with its iden-tical angular orientation

Returning to the CMM results, only a few data points are utilised to obtain a measured diameter, while with the Ballbar and Talyrond alike, they liter-ally take thousands of data points to obtain the polar plotted profiles If, when the CMM touches each of the machinable disk’s profile with the ‘touch-trigger probe’ , this co-ordinate’s data could have been ob-tained at the extremes of the elliptical shape, namely,

at its major and minor diameters, this may account for such a variation in both the range and discrepancies, when compared to the data obtained by the Ballbar and Talyrond

The four φ10 mm holes in each disk that were pro-duced by HSM utilising circular interpolation at their respective quadrant positions (Figs 241 and 242b), are given in the form of tabulated data in Table 16 – in terms of their positional accuracy and radial change, from their theoretical centres From this φ10 mm hole

data, then the radial change for each disk, from the

top, middle and bottom disks, was: 46 µm: 45 µm: and

42 µm: respectively, giving a positional uncertainty across these disks of 4 µm

Conversely, if the difference is considered for the

three stacked disks with respect to their angular rela-tionships to each other, at: 0°; 90°; 180°; 270°; then their

angular positional changes are: 46 µm; 26 µm; 32 µm; and 49 µm; giving a positional uncertainty of 23 µm

This positional uncertainty is still relatively small con-sidering that in this case, each hole’s position is on a

different Z-axis plane – spanning 200mm in height

Although if one considers the ‘Grand mean’ for both cases then they have a positional uncertainty of just

1 µm, which for a machine tool that at this time was around three years old is quite exceptional – having by now, undertaken considerable industrial machinabil-ity trials for the automotive and aerospace industries, but admittedly, this vertical machining centre had pre-viously been both Laser- and Ballbar-diagnostically corrected – showing the ‘true’ relevance of calibration

to resolve and reduce any ‘errors’!

This initial calibration work using the

stereomet-ric artefact, has shown that its overall positional

un-certainty when utilised for HSM accuracy and

preci-sion assessments in combination with machine tool

diagnostics, then compared to other more-exacting

inspection techniques and measurements, has been

successful Hence, a large artefact having replaceable

machining components (disks), has proved to be a valid

means to verify the actual ‘machine tool’s health’ – in

a real sense, as it is undertaken in a loaded state, while

partaking in high-speed cutting trials Once, the

part-program has been written, then the whole machining

activity can be completed in a relatively short

time-frame Although admittedly, the additional

metro-logical inspection and data analyses takes sometime

to complete, but these inspection functions can be

un-dertaken while the machine tool is still in operation –

producing quality machined components

9.7 HSM: Rotating

Dynamometry

High-speed rotating dynamometers are being utilised

across diverse fields of industry, where: hard-part

ma-chining such as die-sinking are necessary; free-form

sculptured parts are required; expensive and delicate

workpieces that need closely-monitored cutting

con-ditions; together with applied and fundamental ma-chinability research programmes HSM rotating cut-ting force dynamometry, allows one to insert the unit into the machine tool’s spindle, where it measures the forces acting on the tool This measurement data are directly amplified within the dynamometer, then radio transmitted to a specially-configured receiver (Fig 244) There are many technical advantages why

a rotating dynamometer is preferred to that of its spa-tially-stationary counterpart (i.e platform-type: see Fig 237), such as:

edge – meaning that the cutting force near to the

cutting process is measured during engagement with

the workpiece, not a reaction force,

maintain-ing a uniform level of natural frequency – while in

contrast, a stationary version (e.g a platfrom-type), has the workpiece’s mass (i.e normally placed onto the dynamometer) which continuously diminishes

as the machining operation takes place,

re-maining constant throughout the machining op-eration – unlike that of a stationary platform

dyna-mometer,

occupy/ori-entate themselves into any position, during ma-chining – whereas the stationary platforms have

certain volumetric space-requirements, having just one orientation to the applied cutting forces

Table 16 The φ10 mm hole positional deviations for the truncated frustum – based upon the three-dimensional Isosceles

tri-angle in the disks at four quadrants (from the theoretical), in terms of their radial change

Position of holes: Top disk Middle disk Bottom disk Range Mean Grand Mean

972

NB Values in: µm

[Source: Smith, Sims, Hope and Gull, 2001]

.

Figure 243 Technique of manufacture of the aerospace aluminium disks and the ‘averaged’ tabulated inspection data

[Source: Smith, Sims, Hope & Gull, 2001]

.

The HSM rotating cutting force dynamometer depicted

in Fig 244, is of compact dimensions having an overall

protrusion length of 106 mm, having a standard collet

chuck adaptor With its short overall length, this HSM

dynamometer has the sensor screwed onto the sensor

bearer, giving a high moment of bending resistance,

which is reflected in the small amount of ‘cross-talk’

values of radial forces on the FZ-force component (i.e

FRadial → FZ)

The rotating HSM dynamometer was designed to

have a ‘ceiling-speed’ in accordance with ISO

Stan-dard: 15641: 1998, which specifies a test speed of

40,000 rev min–1 Although the manufacturer originally

tested the antenna casing up to speeds of 56,000 rev

min–1, corresponding to a centrifugal acceleration of:

aZcentrifugal = 130,000 g

Any HSM tooling assembly, such as a rotating dy-namometer must be balanced with particular care, in this case, there is a pre-balance quality of 6.3 for the antenna casing and sensor, with the final balance be-ing 2.5 g-mm, which ensures that the cuttbe-ing force measurements are not adversely affected

The overall compact construction minimising ac-celeration masses (i.e weighing just 5.3 kg) for this HSM rotating dynamometer, with its low-level of natural frequency, having both a very low response threshold coupled to high resolution precision, pro-vides ‘sound’ cutting force and torque information making it an ideal ‘tool’ for either qualitative, or quan-titative machinability research work Hence, the signal quality achieved by this instrument, makes it possible

to utilise this HSM rotating cutting force

dynamom-Figure 244 A high-speed rotating

cutting force dynamometer, employed during drilling and multi-axis milling operations [Courtesy of Kistler Instru-mente AG]

.