1. Trang chủ
  2. » Kỹ Thuật - Công Nghệ

Machinability and Surface Integrity Part 4 pps

10 264 0
Tài liệu đã được kiểm tra trùng lặp

Đang tải... (xem toàn văn)

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 10
Dung lượng 341,5 KB

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

Chatter is not unduly affected by the feedrate selected, but feed does have an effect on the predictable severity of vibration during machining, NB As no cutting force exists if the vibr

Trang 1

either entirely eliminating it, or at the very least,

min-imising its affect on the overall machining process

Chatter during machining can result from a range of

multifarious and often linked-factors, they include:

• Depth of cut (DOC) – can be considered as the

prin-cipal cause and, for the prospective control of

chat-ter The DOC delineates the chip width, acting as

the feed-back gain within the closed-loop cutting

process,

NB The machining processes ‘stability limit’ –

be-ing the threshold between stable cuttbe-ing and

chat-ter – can be dechat-termined from trial-and-error by

simply incrementally increasing the DOC until the

commencement of chatter, then‘backing-off’ at this

level The prediction of chatter’s onset can be found

analytically, this value being based upon thorough

knowledge of material stiffness and cutting system

dynamics

• Rotational speed – is probably the simplest

param-eter to modify, thereby altering chatter and its

as-sociated amplitude,

NB The peripheral speed of either the rotating

tool, or workpiece, affects the phase-shift between

overlapping surfaces and its associated vibration

regeneration

• Feed – for milling operations the feed per tooth

de-fines the average uncut chip thickness (t),

influenc-ing the magnitude of the cuttinfluenc-ing process Chatter

is not unduly affected by the feedrate selected, but

feed does have an effect on the predictable severity

of vibration during machining,

NB As no cutting force exists if the vibration

oc-curs in the ‘Y’ direction – resulting in loss of

con-tact between the tool and workpiece – the

maxi-mum amplitude of chatter vibration will be limited

by its feed

 ‘Gain’ , can be practically defined in the following way: ‘The

ratio of the magnitude of the output of a system with respect

to that of the input – the conditions of operation and

measure-ments must be specified’ (Smith, 1993, et al.).

• Cutting stiffness (Ks) – is a material property con-nected to: shear flow stress; hardness, as well as work-hardening characteristics of the workpiece, this factor often being referred to in a metaphorical sense of its material’s machinability characteristics,

NB Materials that might offer poorer comparative

machinability, for example titanium, require con-siderably higher cutting forces leading to a greater

displacement in the ‘Y’ direction and as such, offer

a less stable cutting action

• Width of chip (total) – is equivalent to the product

of the DOC multiplied by the number of cutting edges engaged in the cut Furthermore, the total cut width will influence the stability of the cutting process,

NB At a preset D OC corresponding to that of the

‘stability limit’ , increasing the number of engaged cutting edges, will result in chatter, or vice-versa

• Cutting tool geometry – influences both the

direc-tion and the magnitude of the cutting force, in particular the quantity of the force component in

the modulation direction ‘Y’ So, an increased force occurring in the ‘Y’ direction, causes amplified

dis-placement and vibration at 90° to the surface, creat-ing ideal conditions for chatter Other cuttcreat-ing insert geometrical factors that can influence the cutting stability include the following:

Back rake angle (α) – as it is inclined to a more

positive angle, the length of the commencement

of the shearing zone decreases, this in turn,

re-duces the magnitude of the cutting force (F c) As the back rake inclination becomes larger, then this directs the cutting force in a more tangential manner, thereby reducing the force component

in the ‘Y’ direction – creating improved stability

at higher speeds,

NB An insufficient feedrate in comparison to the

insert edge radius produces a less efficient cutting action, with more tool deflection and reduced ma-chining stability

Clearance angle – reduction (γ) – has the effect

of increasing the frictional contact at the inter-face between the tool and workpiece, possibly having a process damping effect This potential stabilising effect could be the result of energy

Trang 2

dissipation – heat transformation, which could

result in decreased tool life, with the

superflu-ous effect of thermal distortion of the machined

part, or an increase in the workpiece’s

heat-af-fected zone (HAZ),

NB On a newly-fitted cutting insert, if initial wear

occurs, this can sometimes have a stabilising effect

for the onset of chatter

– Nose radius – size, insert shape – diamond

tri-angular, square, round, plan approach angle

– positive, neutral, negative – all influence the

area of the chip shape and its corresponding ‘Y’

direction The orientation of the modulation

direction ‘Y’ toward a dynamically more-rigid

direction angle, allows a decrease in vibrational

response, giving greater overall process stability

– having notably less chattering tendencies

As machining process stability is a direct result of

characteristics of dynamic force displacement between

both the workpiece and the cutting insert, all of the

various factors of a machining system: machine tool;

spindle; tooling; workpiece; workholding – in varying

degrees, can influence chatter To increase process

sta-bility of the machining system, it is necessary to

maxi-mise the dynamics, this being the overall product of its

static stiffness and damping capacity Further,

machin-ing stability can be increased by utilismachin-ing toolmachin-ing with

the greatest possible diameter with the minimum of

tool overhang By way of a caution concerning chatter

frequency, this normally occurs near the most flexible

vibrational mode of the machining system

7.3.3 Stability Lobe Diagrams

In Fig 157c, a ‘Stability lobe diagram’ (SLD) is

de-picted, which relates to the: total cut width that can be

machined, to the tooling’s rotational speed, for a

speci-fied number of cutting inserts For example referring

to the: Degarmo, et al (2003) diagram, suppose the

total width of cut was maintained below a minimum

level, then the process stability would exhibit ‘speed

 If the total cut width was maintained below a minimum level,

in practical terms this would be of limited value for many

ma-chining systems.

independence’ , or an ‘unconditional stability’ Hence,

at relatively slow speeds an increased stability can

be achieved within the process damping region – as

shown The ‘conditional stability’ lobe regions of the

diagram, permit an increased total cut width (i.e the

DOC x number of cutting edges, these being engaged

in the cut) at dynamically preferred speeds, at which

the phase-shift ‘ε’ between overlapping, or consecutive

cutting paths approaches zero In Fig 157c, stability

lobe number ‘N’ refers to the complete vibration cycles

existing between overlapping surfaces Moreover, the higher speeds correspond to lower lobe numbers, pro-viding the utmost potential increase in the total cut width and material removal rate – this being due to the greater lobe height and width If the total cut width exceeds the stability threshold – even assuming that the cutting process is operating at the desired speed, chatter will occur So, the larger the total cut width above the ‘stability limit’ , the more unstable and ag-gressive the chatter vibration becomes

Referring to the diagrammatic representation of the SLD on the graph in Fig 157c, if a chatter

con-dition arises, such as that found at point ‘a’ , the

ro-tational speed is attuned to the initial recommended

speed (i.e when ‘N=’), resulting in stable machining

at point ‘b’ on this diagram The D OC can be incremen-tally increased until the onset of chatter again – as the

threshold stability is crossed at point ‘c’ By utilising a

hand-held ‘speed analyser’ whilst the chatter

contin-ues – under the previously-selected operating

condi-tions, this will result in the ‘analyser’ giving a modified

speed recommendation that corresponds to point ‘d’

Now, if required, the DOC can be progressively

incre- ‘Speed analysers’ , are normally hand-held devices that

pro-duce dynamically-favoured speed recommendations and are

commercially available Such ‘speed analyser’* equipment

when utilised for a cutting process, can show the relative mo-tion between the tooling and the workpiece and recommends the appropriate speed to avoid chatter-effects

*‘Speed analysers’ can be successfully used for many industrial

applications, such as those involving: High-speed; Thin-chip, hardened-die machining; multi-point cutting operations – milling, etc.; Turning and boring operations These ‘speed anal-ysers’ can also be employed for workpiece compositions rang-ing from ductile metals (i.e aluminium and steel grades) and brittle materials (i.e cast irons and brasses, etc.), together with some non-metallics (plastics, etc.) and composite materials (carbon fibre, etc.)

Trang 3

mentally increased to point ‘e’  – this being a

‘safe-limit’ for the optimum machining operation

7.4 Milled Roundness –

Interpolated Diameters

Circular features such as bosses, circular rebates, etc.,

can be CNC milled by utilising a specific

word-ad-dress ‘circular interpolation’  command This CNC

function creates precise and accurate circular control

in two slideways simultaneously, while the milling

cutter mills around the workpiece, as depicted in Fig

158 Here, the milling cutter’s rigidity plays an

impor-tant role in the quality of the final machined feature,

this being based upon the ‘rigidity square rule’  The

deflected milling cutter illustrated in Fig 158-right,

having lack-of-rigidity will produce some unwanted

effects on the final milled part Cutter deflection not

only introduces the potential for chatter vibration,

but if used to mill up to square shoulder, its deflection

distorts the component geometry and introduces

har-monic variation to the circular interpolated feature

So that minimal change takes place in a milled profile,

it is advisable to keep to cutter lengths having short

 Generally-speaking, it is not advisable to attempt to maintain

both the DOC and the total cut width at the stability

thresh-old , because any variation in the: workpiece affecting its

cut-ting stiffness ‘K s’; speed errors; or perhaps small changes in

the overall dynamic characteristics of the machining system,

could result in crossing the stability limit, creating severe

chatter For example, in a milling application, the amplitude

of chatter vibration can be limited by a provisional feed per

tooth reduction , until an established and desired speed has

been achieved offering a stable DOC.

 ‘Circular interpolation’ , is a block of entered information

di-recting the CNC system to cut, either an arc, or a circle, (e.g

G02 – in a clockwise, or G03 anti-clockwise direction).

 ‘Rigidity square rule’ – for milling cutters states: ‘Cutter

rigid-ity decreases by the ‘square’* of the distance from the holder’

(Smith, 1993, et al.).

*For example, if a cutter ‘stood-out’ from its respective

tool-holder by 50 mm to mill a circular feature (Fig.158 – left), then,

if all other machining conditions remained the same and, then

cutter was replaced by one of 100 mm long (Fig 158 – right),

it would now be 4 times less rigid, causing serious tool

deflec-tion.

stand-off distances, conducive with correct and cur-rent operational practices

There are several distinct problems involved in the milling high-quality circular interpolated features and,

a slight digression into basic machine tool induced-er-rors is necessary to clarify the circumstances for the problems exhibited in Fig 159 Most of today’s

ma-chine tools have what is termed ‘orthogonally-orien-tated axes’ and in the case of the popular three-axis

vertical machining centre configurations, if the axes have not been recently calibrated, then considerable

‘error’8 can be introduced into the final milled part

features It has been well-proven that a machine tool

equipped with three orthogonal sideways: ‘X-axis’;

‘Y-axis’ – in the horizontal plane, together with the

‘Z-axis’ – in the vertical plane, can introduce up to 21

kinematic ‘errors’ into the cutting process The kine-matics for any machine tool are quite complex, when it has the ability to provide motion to all its axes simulta-neously, although these errors are often small, they are

 ‘Orthogonally-orientated axes’ , (is briefly mentioned in

Foot-note 2) refers to the fact that each axis is positioned at 90° with respect to each other, often situated on top of another axis For example, on a typical 3-axis vertical machining centre,

the ‘Y-axis’ sits on top of the ‘X-axis’ , but at right-angles to it, conversely, the ‘Z-axis’ is situated at 90° to these axes – hence

the term ‘orthogonal’

NB Non-orthogonal machine tools exist, often having

com-plex ‘kinematics’* between five and six axes Therefore with

these machine tools, in order to machine (i.e mill) a

straight-line all the axes must be in synchronised control to achieve

this linear action.

*Kinematics, comes from the Greek word ‘Kinesis’ , which

means ‘Motion’ It can be defined as: ‘The study of motion

with-out regard for the cause‘ (Lombardi, 2001) In machine tool

terminology, it refers to the translational effects of both lin-ear and angular motions It is principally concerned with the

effects of the ‘degrees of freedom’ for a ‘free-body’ in

three-di-mensional space (also see: Footnote 47, in Chapter 3).

 ‘Error’ is now not considered as an appropriate metrological

term for any form of calibration, the recommended term

to-day, is: ‘uncertainty’*.

*‘Uncertainty’ , has been simply defined as: ‘The doubt that

exists about the result of any measurement’ (Bell/NPL, 1999)

This is why today, uncertainty in measurement is a

combina-tion of many factors, some physical, while others are induced

Hence, another term, along with all of these uncertainty

fac-tors has been coined, which is its ‘Uncertainty budget’ – this

being a simple mathematical calculation, based upon a sum-mary of these uncertainty calculations.

Trang 4

Figure 158 The effect of increased milling cutter length on the resultant circular interpolated profile on the workpiece

.

Trang 5

Figure 159 The generated errors produced when circular interpolating at high feedrates when high-speed machining

.

Trang 6

but significant ‘errors’ , which can be said to be

simplis-tically produced as a result of:

• Linear motions (six) – created by the displacement

of the forward-and-backward motion of the X-, Y-

and Z-axes slideway movements, introducing

par-ticular non-linearities into the slideway

position-ing,

• Rotational motions (three) – yaw, pitch and roll for

each axis All of these partial rotational motions can

be practically-described in the following manner:

• Yaw is the side-to-side ‘crabbing-motion’ along the

slideway,

NB ‘Yaw’ is normally the result of too much

clear-ance (i.e ‘slop’) in the adjacent slideway members

• Pitch introduces a backward-and-forward

rock-ing (pitchrock-ing) action normal to the slideway, as the

moving element traverses along the axis,

NB ‘Pitching’ is probably due to the

‘profile/wavi-ness’ (i.e long-frequency effects) in its respective

slideway

• Roll is the clockwise-and-anticlockwise rotational

motion along the slideway

NB ’Roll’ could be introduced by two ‘adjacent

ways’ situated on each slideway, but not being

coin-cident with respect to each other (i.e laying in the

same respective plane), causing a limited pivoting

action – along the ‘line-of-sight’ of the axis as it

tra-verses along its length

• Squareness (three) – these ‘errors’ occur due to the

fact that each axis may not be at 90° (i.e square) to

one another

These types of 21 ‘kinematic machine-induced

er-rors’ can be appreciably reduced by the application of

calibration through laser-based techniques To a lesser

extent, these ‘errors’ can be minimised via ballbar

ar-tifact-based methods, offering a quick ‘health-check’

by either static, or dynamic assessment techniques

The results of either the laser, or ballbar, can be fed

back into the machine’s CNC controller for dynamic

corrections as cutting takes place, offering a

consid-erable improvement in the machine’s subsequent

ac-curacy and precision The above machine tool

calibra-tion techniques are somewhat beyond the scope of the

present discussion, the same could be said for

‘ther-mally-induced errors’ , however, they can also influence

the machined part surface and the machine tool’s

pro-filing abilities Moreover, ‘error-mapping techniques’

and sophisticated in-process control by an associated

‘dynamic error compensation system’ , have been shown

to extensively reduce the effects of the ‘variety of

er-rors’ that can be present on the machine tool, but once again, these topics are mentioned only for further re-search applications – as necessary

The circular interpolated milled profile shown in Fig 159, shows significant departures from roundness

of the milled workpiece, which is a function of most of the previously discussed kinematically- and thermally-induced machine tool ‘errors’ , together with the possi-bility of some ‘load-induced errors’ This diagrammatic representation (i.e Fig 159), indicates that several

‘errors’ on the milled circular interpolated profile are present At relatively slow simultaneous

feeding-mo-tions of the two axes (‘X-’ and ‘Y-axis’), it will generate

a reasonable facsimile of the required circular feature However, then by somewhat increasing this milled in-terpolation speed, the apparent roundness will appre-ciably degrade, the reasons for this degradation, might

be the result of:

• Servo-spikes – these unwanted effects occur at the

‘axis transition points’  at their respective 90° angu-lar intervals, often termed ‘quadrant-points’ ,

• Back-lash – possibly resulting from any form of

axis reversals, originating from the recirculating ballscrews0, creating a slight ‘off-set’ , or ‘mismatch’

at the axis transition points,

• Servo-errors – when both axes are simultaneously

moving, their respective linear speed should be

 ‘Axis transition points’ , are where the ‘servo-spikes’ occur

They result from a reversal of one of the axes at this angular position and, its associated motor power-surge creating this

‘spike’ Normally, the ‘spike’ is associated afterward by a cor-responding, but very small localised slack here, as axis take-up

begins once more at these ‘quadrant-points’ on the

circular-interpolated feature (i.e see the inset and magnified diagram

in Fig 159).

0 ‘Recirculating ballscrews’ , are not supposed to have any

ap-preciable back-lash present, as they are normally pre-stressed

by applying loads by the application of either: tension-, or

compression-shimming However, as the pitch of any the screw has minute errors present, these are usually ‘mapped-out’ by

the original machine tool builder – using the recognised

In-ternational Standard laser-calibration techniques Although,

once the machine tool has been operating for sometime and either local ballscrew-wear occurs, or perhaps the machine

has had the occasional ‘tool-crash’ , this can introduce and

af-fect both its pitching- and back-lash-errors.

Trang 7

perfectly matched, allowing either a partial arc,

or circular feature to be reproduced If

non-syn-chronised motion occurs, often termed

‘servo-mis-match’  between these two axes, then an elliptical

profile – usually inclined at an 45° angle occurs,

• Squareness – when orthogonal (squareness) is not

maintained between the two interpolating axes,

then the net result will look similar to that of a

milled angular elliptical profile shape, which is

un-affected by the selected circular interpolation

rota-tional direction

Considerably more machine tool-induced factors can

affect a milled circular interpolated profile These

‘er-rors’ can be found, isolated and then reduced by

di-agnostically interrogation by using dynamic artefacts,

such as the ballbar Ballbars and their associated

in-strumentation can not only find the sources of error,

they can prioritise their respective magnitudes – to

show where the main ‘error-sources’ occur, then

in-stigate any feed corrections into the CNC controller

to nullify these ‘machine-induced errors’ As a result

of eliminating such ‘error-sources’ , this enables the

milled circular contouring and overall performance to

be appreciably enhanced

7.5 Machined Surface Texture

Introduction to Surface Texture Parameters

When a designer develops the features for a component

with the requirement to be subsequently machined

utilising a computer-aided design (CAD) system, or

by using a draughting head and its associated

draw-ing board, the designer’s neat lines delineate the

de-sired surface condition, which can be further specified

by the requirement for specific geometric tolerances

In reality, this designed workpiece surface condition

cannot actually exist, as it results from

process-in-duced surface texture modifications Regardless of the

method of manufacture, an engineering surface must

have some form of ‘topography, or texture’ associated

 ‘Servo-mismatch’ , can often be mistaken for a ‘squareness

er-ror’ , but if the contouring interpolation direction is changed,

from G02 (clockwise) to G03 (anti-clockwise) rotation, then

an elliptical profile will ‘mirror-image’ (‘flip‘) to that of the

op-posite profile – which does not occur in ‘squareness errors’.

with it, resulting from a combination of several inter-related factors, such as the:

• Influence of the workpiece material’s microstruc-ture,

• Surface generation method which includes the cut-ting insert’s action, associated actual cutcut-ting data and the effect of cutting fluid – if any,

• Instability may be present during the production machining process, causing induced chatter, result-ing from poor loop-stiffness between the machine-tooling-workpiece system and chosen cutting data,

• Inherent residual stresses within the workpiece

can occur, promoted by internal ‘stress patterns’ 

causing latent deformations in the machined com-ponent

From the restrictions resulting from a component’s manufacture, a designer must select a functional sur-face condition that will suit the operational constraints for either a ‘rough’ , or ‘smooth’ workpiece surface This then raises the question, posed well-over 25 years ago –

which is still a problem today, namely: ‘How smooth is smooth?’ This question is not as superficial as it might

at first seem, because unless we can quantify a surface accurately, we can only hope that it will function cor-rectly in-service In fact, a machined surface texture condition is a complex state, resulting from a combi-nation of three distinct superimposed topographical conditions (i.e as diagrammatically illustrated Fig 160a), these being:

 ‘Stress patterns’ , are to be expected in a machined

compo-nent, where: corners, undercuts, large changes in cross-sec-tions from one adjacent workpiece feature to another, etc., produce localised zones of high stress, having the potential outcome for subsequent component distortion ‘Modelling’

a component’s geometry using techniques such as: finite ele-ment analysis (FEA), or employing photo-elastic stress analy-sis* models or similar simulation techniques, will highlight

these potential regions of stress build-up, allowing a designer

to nullify, or at worst, minimise these potential undesirable stress regions in the component’s design.

*Photo-elastic stress analysis displays a stress-field, normally

a duplicate of the part geometry made from a thin two-dimen-sional (planar) nematic liquid crystal, or more robustly from

a three-dimensional Perspex model, which is then observed through polarised light source This polarised condition, will highlight any high-intensity stress-field concentrations in the part , which allows the ‘polarised model’ to be manipulated

by applying either an un-axial tension, or perhaps a bi-axial bending external stress to this model, showing dynamically its potential stress behaviour during its intended in-service con-dition.

Trang 8

Figure 160 Surface texture comprises of: ‘long-’, ‘medium-’ and ‘short-components’, together with the

‘direction of the dominant pattern’ – superimposed upon each other [Courtesy of Taylor Hobson]

Trang 9

1 Roughness – comprising of surface irregularities

occurring due to the mechanism of the machining

production process and its associated cutting insert

geometry,

2 Waviness – that surface texture element upon

which roughness is superimposed, created by

fac-tors such as the: machine tool, or workpiece

deflec-tions, vibrations and chatter, material strain and

other extraneous effects,

3 Profile – represents the overall shape of the

ma-chined surface – ignoring any roughness and

wavi-ness variations present, being the result of perhaps

the long-frequency machine tool slideway errors

The above surface topography distinctions tend to be

qualitative – not expressible as a number – yet have

considerable practical importance, being an

estab-lished procedure that is functionally sound The

com-bination of roughness and waviness surface texture

components, plus the surface’s associated ‘Lay’  are

shown in Fig 160a The ‘Profile’ is not depicted, as it

is a long-frequency component and at best, only its

partial affect would be present here, on this diagram

The ‘Lay’ of a surface tends to be either: anisotropic,

or isotropic in nature on a machined surface

topog-raphy When attempting to characterise the potential

functional performance of a surface, if an anisotropic

‘lay-condition’ occurs, then its presence becomes of

vital importance If the surface texture instrument’s

stylus direction of the trace’s motion over the assessed

topography is not taken into account, then totally

mis-representative readings result for an anisotropic

sur-face condition occur – as depicted in Fig 160b This

is not the case for an isotropic surface topography, as

relatively uniform set of results will be present,

regard-less of the stylus trace direction across the surface (i.e

 ‘Lay’ , can simply be defined as: The direction of the dominant

pattern’ (Dagnall, 1998).

 ‘Anisotropic, or isotropic surfaces, either condition can be

in-dividually represented on all machined surfaces Anisotropy,

refers to a surface topography having directional properties,

that is a defined ‘Lay’ , being represented by machined

feed-marks (e.g turned, shaped, planed surfaces, etc.) Conversely,

an isotropic surface is devoid of a predominant ‘Lay’

direc-tion, invariably having identical surface topography

charac-teristics in all directions (e.g shot-peening/-blasting and, to a

lesser extent a multi-directional surface-milling, or a

radially-ground surface, etc.).

see Fig 161a – for an indication of the various clas-sifications for ‘Lay’)

Returning once more to Fig 160b, as the stylus trace obliquity changes from trace ‘A’ , inclining to-ward trace ‘E’ , the surface topography when at ‘E’ has now become flat, giving a totally false impression of the true nature of the actual surface condition If this machined workpiece was to be used in a critical and highly-stressed in-service environment, then the user would have a false sense of the component’s potential

fatigue characteristics, potentially resulting in

ei-ther premature failure, or at worst, catastrophic fail-ure conditions In Fig 162, the numerical data (ISO 1302:2001), has been developed to establish and de-fine relative roughness grades for typical production processes However, some caution should be taken when utilising these values for control of the surface condition, because they can misrepresent the actual state of the surface topography, being based solely on a derived numerical value for height What is more, the

‘N-number’ has been used to ascertain the arithmetic

roughness ‘Ra’ value – with more being mentioned on

this and other parameters shortly The actual ‘N-value’

being just one number to cover a spread of potential

‘Ra’ values for that production process

Neverthe-less, this single numerical value has its merit, in that

it ‘globally-defines’ a roughness value (i.e.‘Ra’) and

its accompanying ‘N-roughness grade’ , which can be

used by a designer to specify in particular a desired surface condition, this being correlated to a specific production process The spread of the roughness for a specific production process has been established from experimental data over the years – covering the

maxi-mum expected ‘variance’  – which can be modified

 ‘Fatigue’ , can be defined as: ‘The process of repeated load, or

strain application to a specimen, or component’ (Schaffer, et

al., 1999) Hence, any engineering component subjected to repeated loading over a prescribed time-base, will normally undergo either partial, or complete fatigue.

 ‘Variance’ , is a statistical term this being based upon the

standard deviation, which is normally denoted by the Greek

symbol ‘σ’ Thus, variance can be defined as: ‘The mean of the

squares of the standard deviation’ (Bajpai, et al., 1979)

Thus, σ = √Variance, or more specifically for production

op-erations:

s=��n −  ċ �n

j=(xj− ¯x)

*s = the standard deviation of a sample from a production batch run.

Trang 10

depending upon whether a fine, medium, or coarse

surface texture is obligatory Due to the variability in

any production process being one of a ‘stochastic

out-put’ , such surface texture values do not reflect the

likely in-service performance of the part Neither the

surface topography, nor its associated integrity has

been quantified by assigning to a surface

representa-tive numerical parameters In many instances, ‘surface

engineering’ 8 is utilised to enhance specific

compo-nent in-service condition

It was mentioned above that in many in-service

engineering applications the accompanying surface

texture is closely allied to its functional performance,

predominantly when one, or more surfaces are in

mo-tion with respect to an adjacent surface This close

proximity between two mating surfaces, suggests that

the smoother the surface the better, but this is not

nec-essarily true if the surfaces in question are required to

maintain an efficient lubrication film between them

The apparent roughness of one of these surfaces with

respect to the other, enables it to retain a

‘holding-film’ in its associated topographical ‘valleys’ While

another critical factor that might limit the designer’s

choice of the smoothness of an engineering surface’s

selection, is related to its production cost (i.e see

Fig 161b) Therefore, if the designer requires a very

smooth machined surface, it should be recognised that

its manufacturing time is considerably longer – so its

respective cost will be greater to that of a rough

sur-face, this being exacerbated by a very close

dimen-sional tolerance requirement

 ‘Stochastic processes’ , are defined as: ‘A process which has a

measurable output and operating under a stable set of

condi-tions which causes the output to vary about a central value in a

predictable manner’ (Stout, 1985).

 ‘Surface engineering’ , is applying suitable discrete

technolo-gies to create surface films (e.g 10 to 100 nm thick), or by

ma-nipulating the surface atomic layers (e.g 2 to 10 atomic layers,

approximately 0.5 to 3 nm), to enhance the ‘engineered’

sur-face condition (i.e Source: Vickerman, 2000).

 ‘Surfaces’ , are recognised to have topographical features that

mimic the natural world So a regular/irregular engineering

surface can exhibit both peaks and valleys, not unlike

moun-tainous terrain.

7.5.1 Parameters for Machined

Surface Evaluation

In order that a machined workpiece’s surface texture can be determined using stylus-based (two-dimen-sional) instrumentation, three characteristic lengths are associated with this surface’s profile (i.e see Fig 163a), these are:

1 Sampling length0 – is determined from: the length

in the direction of the X-axis used for identifying

the irregularities that characterise the profile under evaluation Therefore, virtually all surface

de-scriptors (i.e parameters) necessitate evaluation over the sampling length Reliability of the data

is enhanced by taking an average of the sampling lengths as depicted by the evaluation length shown

in Fig 162a Most of today’s stylus-based surface texture instruments undertake this calculation au-tomatically,

2 Sampling length – can be established as: the to-tal length in the X-axis used for the assessment of

the profile under evaluation From Fig 163a, this

length may include several sampling lengths – typi-cally five – being the normal practice in evaluating roughness and waviness profiles The evaluation

length measurement is the sum of the individual

sampling lengths (i.e it is common practice to em-ploy a 0.8 mm sampling length for most surface texture assessments),

3 Traverse length – can be defined as: the total

length of the surface traversed by the stylus in mak-ing a measurement The traverse length will

nor-mally be longer than the evaluation length (i.e see Fig 163a), this is due to the necessity of allowing

‘run-up’ and ‘over-travel’ at each end of the

evalua-tion length These addievalua-tional distances ensure that any mechanical and electrical transients, together filter edge effects are excluded from the measure-ment

0 ‘Sampling length’ , is often termed ‘Meter cut-off’ , or simply the ‘cut-off’ length and its units are millimetres The most

common cut-offs are: 0.25, 0.8, 2.5, 8.0, 25.0 mm The 0.8 mm

sampling length will cover most machining production

pro-cesses In any surface texture evaluation, it is essential that the

cut-off is made known to the Inspector/Metrologist reviewing

this surface topographical data.

Ngày đăng: 05/08/2014, 21:20