Figure 4.14b, for turning two differ-ent Ti-alloys with a carbide tool, is an example of where a break point falls within the prac-tical cutting speed range.. Figure 4.14c, for face mill
Trang 1a ductile cast iron, shows ceramic tools having a larger n-value than a carbide tool – for the white alumina, n≈ 1.2 This very high value indicates a mechanical wear mechanism reducing in intensity with increasing cutting speed Figure 4.14(b), for turning two differ-ent Ti-alloys with a carbide tool, is an example of where a break point falls within the prac-tical cutting speed range Figure 4.14(c), for face milling a grey cast iron, shows a condition in which tool life decreases with reducing cutting speed
Taylor’s equation – influence of feed and depth of cut
Tool life is influenced by feed and depth of cut, as well as by cutting speed Additional life equations are
f T n
2= C2; dT n
and these may be combined with equation (4.2) (replacing n there by n1) to give
V 1/n
1f 1/n
2d 1/n
When tool life is limited by thermal damage mechanisms, n1< n2< n3: i.e cutting speed has a larger influence on life than does feed than does depth of cut, reflecting the influences of these variables on cutting temperature If, however, tool life is determined by
chipping and fracture failures, n2and n3can become smaller
4.2.3 Tool life fluctuations
It is almost impossible to keep cutting conditions exactly constant in practical machining Even if it were possible, it would be found that tool life and failure are phenomena based
on probability Fluctuations cannot be avoided in these However, the range of fluctuations
is influenced by the damage mechanism It is easy to imagine larger fluctuations when chipping, or fracture rather than abrasion is the main mechanism
Figure 4.15 shows the cumulative probability of flank wear development after 1 min of
Tool life 133
Fig 4.15 Distributions of flank wear after turning free-cutting steel B1112 and difficult-to-cut sintered steel and
Inconel 718
0.05
5 10 20 40 60 80
99 85
Flank wear VB (mm)
P10 Sintered steel
P10 B1112
TiC-Al2O3 ceramic Inconel 718
VBmax
B112 - P10, V = 200m/min, d = 0.5mm, f = 0.1mm/rev Sintered steel - P10, V = 200m/min, d = 0.5mm, f = 0.1mm/rev
Inconel 718 - Al2O3-TiC ceramic, V = 200m/min, d = 0.5mm, f = 0.19mm/rev
Trang 2turning a resulphurized free machining steel and a sintered steel with a P10 carbide tool (plotted on a Weibull chart) Abrasion was the main cause of tool wear with the free machining steel, while edge chipping was the mechanism with the sintered steel The different slopes of the Weibull plots are clear The figure also shows the distribution for turning Inconel 718 with an Al2O3/TiC ceramic tool As well as the greater amount of wear, the similarity of slope between this and the sintered steel observations is striking Figure 4.16 is an example of tool wear and wear distribution influenced by the machine tool It gives the results of face milling a quenched die steel with an Al2O3/TiC ceramic tool, in the same conditions apart from the machine tool used Tool wear was by edge chip-ping or fracture Machine B obviously provides better resistance against this type of damage This is due to a better stiffness, maybe a better dynamic stiffness
4.3 Summary
This chapter complements Chapter 3 on tool bulk properties, by focusing on the mecha-nisms of cutting edge damage and their characteristic developments with time Cutting edges experience much higher normal and shear stresses than almost any other type of bearing surface and, at high cutting speeds, high temperatures are also generated It is not surprising that tool lives are measured in minutes rather than in hours, and certainly not in days
Abrasion occurs with all tools if the work material has hard enough phases, and self-abrasion follows from other mechanical causes of damage Mechanical damages, of
99 90 70 50 30
10
3
1
Machine tool A
Machine tool B
Numbers of impact until tool fracture Cutting speed : 220 m min-1, Depth of cut : 0.1 mm, Feed rate: 0.1 mm/tooth, Cutter Dia : 80 mm
Fig 4.16 Distributions of tool life limited by fracture when milling a quenched die steel (HRC60) with an Al2O3/TiC ceramic tool, on two different milling machines
Trang 3increasing size – from attrition, to chipping, to fracture – increase the more brittle is the tool material and they are relatively insensitive to temperature changes Thermal damages follow diffusion and chemical reactions They are very sensitive to temperature and are particularly variable from one tool and work combination to another Adhesive wear depends on both mechanical and thermal factors, and passes through a maximum rate as temperature increases
For all these reasons of complexity and further influences of mode of cutting, and of the machine tools themselves, on tool life, it has not been attempted to provide comprehensive guidance on tool damage rates Rather, the goal has been to emphasize what phenomena can occur, and what their effects look like, so mechanisms limiting life in different circum-stances may be recognized and sensible directions for improved performance may then be investigated
References
Cook, N H (1973) Tool wear and tool life Trans ASME J Eng Ind 95B, 931–938.
Dawihl, W (1941) Die Vorgange beim Verschleiss von Hartmetallegierungen Stahl und Essen 61,
210–213.
Gregory, B (1965) Surface interaction of cemented carbide tool material and Armco iron Brit J.
Appl Phys 16, 689–695.
Kitagawa, T., Maekawa, K., Shirakashi, T and Usui, E (1988) Analytical prediction of flank wear
of carbide tools in turning plain carbon steels (Part 1) Bull Jap Soc Prec Eng 22(4), 263–269.
Naerheim, Y and Trent, E M (1977) Diffusion wear of cemented carbide tools when cutting steel
at high speeds Metals Technology 4, 548–556.
Narutaki, N and Yamane, Y (1976) Wear mechanism of carbide tool based on the reaction between
tool and work material (Part 1 – reaction test) Bull Jap Soc Prec Eng 10(3), 95–100.
Narutaki, N and Yamane, Y (1993) High-speed machining of Inconel 718 with ceramic tools.
Annals CIRP 42(1), 103–106.
Takeyama, H and Murata, R (1963) Basic investigation of tool wear Trans ASME J Eng Ind 85,
33–38.
Trent, E M (1952) Some factors affecting wear on cemented carbide tools Proc I Mech E Lond.
166, 64–74.
Trigger, K J and Chao, B T (1956) The mechanism of crater wear of cemented carbide tools Trans
ASME 78,1119–1126.
Uehara, K (1976) On the generating mechanism of wear particles on carbide cutting tools J Japan
Soc Prec Eng 42(6), 445–452.
Usui, E., Shirakashi, T and Kitagawa, T (1978) Analytical prediction of three dimensional cutting
process Pt 3 Trans ASME J Eng Ind 100, 236–243.
Yamane, Y and Narutaki, N (1983) The effect of atmosphere on tool failure in face milling (1st
report) J Jap Soc Prec Eng 49(8), 521–527.
References 135
Trang 4Experimental methods
Previous chapters have presented optical and electron microscope pictures of chip sections and worn tools, and the results of cutting force and temperature measurements In addition
to cutting force measurements, acoustic emission is also used to study the health of a cutting process This chapter explains a number of these experimental methods
5.1 Microscopic examination methods
5.1.1 The quick-stop technique
Direct observations as well as theoretical analyses are needed to clarify chip formation mechanisms Ideally, such observations would be during cutting, to follow dynamic
Fig 5.1 The principle of a quick-stop device for use in turning
Trang 5variations of chip flow Although video cameras have been used to gain an external overview of dynamic chip motions, and it is possible to look through transparent tools (for example made from diamond or sapphire) directly at the chip contact, it is difficult, in general, to resolve much because of the small scale of the deforming region and usually the high cutting velocities Experimentalists are prepared to lose dynamic information to gain microscopic detail, by freezing the motion, for later study The quick-stop technique
is a popular method for achieving this The machining process is stopped quickly, by moving the tool and work material apart at a speed greater – preferably much greater – than the cutting speed The chip is left attached to the work (sometimes with a fragment of the cutting edge attached as well) The photographs in Figure 2.4 are polished and etched sections of quick-stopped chips
Figure 5.1 is a schematic view of a quick-stop device for use with a stationary tool and
a moving workpiece, such as in turning, while Figure 5.2 shows a device that could be used for a stationary work and moving tool, as in milling In Figure 5.1, the tool is supported at
a pivot point and by a shear pin A mass M is made to strike the tool holder with a speed
VM If the impact force is much greater than needed to break the shear pin, the mass will then cause the tool holder to swing quickly away from the chip The tool holder’s velocity
VTdoes not instantaneously reach the cutting velocity V that is necessary for cutting to stop, because of its inertia However, to minimize the retraction time, M and VMshould be made large and the inertia of the holder should be made small
In practice, VMis frequently made large by firing the mass M from a gun (although for
low cutting speed turning tests, hitting the tool holder with a hammer can be sufficient)
A device that uses a humane killer gun (normally used for stunning animals prior to
slaughter) with its captive bolt as the mass M was reported to achieve a tool displacement
of 2.5 mm in 1.2 × 10–4s (Williams et al., 1970) If this is assumed to have occurred at
approximately constant acceleration, and it is supposed that, for a successful quick-stop,
VTmust reach V in a cut distance less than f/10, then this device can be used successfully,
provided
Microscopic examination methods 137
Fig 5.2 The principle of a quick-stop device for use in milling
Trang 6V[m/min] ≤ 354f [mm] (5.1)
For a feed of 0.13 mm, the largest allowable cutting speed is 128 m/min, while for f =
0.5 mm, the largest speed is 250 m/min These speeds are larger than those represented in Figure 2.4, but are not large compared with what can be of interest in modern high speed machining The acceleration required of the tool increases as the square of the cutting veloc-ity, so successful quick-stops become rapidly more difficult as the cutting speed increases
A similar discussion could be developed in terms of the device of Figure 5.2 However,
in milling, it is more difficult to guide the work material away from the cutting edges, and the work and its holder have higher inertia than the tool and its holder in turning The quick-stop must be synchronized with the intermittent cutting action There must be a very special reason to pursue a quick-stop in milling, to make the difficulties worthwhile Quick-stops can show different results, depending on the adhesion between the chip and the tool (Figure 5.3) If there is low adhesion, a clean separation between the two will occur, as shown in Figure 5.3(a) Coated tools usually show this behaviour If there is high adhesion relative to the strength of the chip or tool, any of the results of Figure 5.3(b) to (d) can occur If it is particularly desired to preserve the chip/tool interface, a result like Figure 5.3(d) can be engineered by artificially weakening the tool with a notch or crack on its rake face
Fig 5.3 Modes of quick-stop separation
Trang 75.1.2 Other chip form and wear observations
Careful observation of tools and chips after machining can often reveal useful information, without the need for quick-stops For example, the built-up edge (BUE) formed in machin-ing is usually unstable It is carried away on the back surface of chips, so observation of the chips (Figure 5.4) can give information as to whether BUE is formed or not It is obvi-ous that information about wear is obtained by looking at the cutting tools at any time after cutting
Chapter 4 has shown examples of SEM and EPMA used to study wear and contact conditions in great detail The magnifications of these techniques are not always necessary
In many cases, a low magnification optical microscope,× 10 or × 20, is enough Such a
microscope on an X–Y measurement stage is commonly used in laboratories or machine
shops to record wear images and their sizes Wet photography and printing paper used to
be used for archiving information for many years Now, a high quality CCD camera and a personal computer with a large memory can do the job
5.2 Forces in machining
5.2.1 Resultant forces
Forces in machining can be measured in two main ways: directly or indirectly Direct measurements involve mounting a tool (in turning) or the tool or workpiece (in milling) on
a dynamometer, which responds to the forces by creating electrical signals in proportion
to them These measurements are used when the forces need to be known accurately both
in magnitude and direction, for example if thrust, feed and the main cutting forces in turn-ing are required (Figure 5.5), or the torque and thrust force in drillturn-ing are needed Indirect measurements involve deductions from the machine tool behaviour For exam-ple, the power used by the main spindle motor increases with the main cutting force or torque; and that used by the feed motions can be related to the feed force Particularly with
Forces in machining 139
Fig 5.4 The back surface of chips formed from 0.15% C steel by P20 carbide tools: (a) with built-up edge, v = 40 m min –1 , d = 2.0 mm, f = 0.08 mm rev –1 ; (b) without built-up edge, v = 100 m min –1 , d = 2.0 mm, f = 0.12 mm rev –1
Trang 8NC machines, which are fitted with high sensitivity and response main and feed drive motors, indirect methods can be used to determine the active forces Indirect methods are less accurate than direct methods, but can be sufficient for monitoring purposes The main consideration here will be direct methods
Tool dynamometers – general points
A tool dynamometer should have high sensitivity, high rigidity, high frequency response, high linearity and low drift Sensitivity is expressed as electrical output per unit force input Useful dynamometers must be able to discriminate at least 1% of full scale output Rigidity depends strongly on the dynamometer’s construction The force sensing trans-ducer is usually the least rigid element of a dynamometer’s structure: different types of element are considered in the following subsections
Frequency response depends on a dynamometer’s natural frequency and damping char-acteristics In line with elementary dynamics, these may be described in terms of the response of a viscously damped elastic system subjected to a harmonic forcing system:
Figure 5.6 shows how the amplitude ratio (the response relative to the response in static conditions) of such a system varies with frequency ratio (the frequency relative to the system’s undamped natural frequency of k/m) and damping factor c/cc, where ccis the critical damping coefficient The figure shows that for a linear response between amplitude and force (and hence a linear response between a dynamometer’s output and force), a damping factor slightly less than 1, around 0.7, is desirable and then a dynamometer could
be used at frequency ratios up to 0.2 to 0.3
Fig 5.5 Forces acting on cutting tool in turning
Trang 9Linearity and drift are usually more influenced by the electrical elements (including signal amplification) than by the mechanical elements of a dynamometer Systems with linearity better than 0.5% of full scale output are required Drift, which describes the stabil-ity of output (both from the dynamometer transducer and amplification system) over time, can be a problem with cutting force dynamometers because of the sensitivity of electrical elements to temperature changes and the tendency of machining to heat its surroundings
Strain gauge dynamometers
A common type of dynamometer uses strain gauges to sense elastic strains caused by cutting forces Figure 5.7 shows a basic elastic beam type dynamometer with gauges bonded to its surface It also shows an example of a wire-type gauge and a Wheatstone bridge and
ampli-fier system usually used to measure strain changes in the gauges The main cutting force FC
will cause the beam to bend, so that the gauge on the top surface will be placed in tension, that
on the bottom surface will be placed in compression, and those gauges on the side surfaces (at the neutral axis) will experience no strain Likewise, a feed force will strain the side-face gauges but not those at the top or bottom The arrangement shown in Figure 5.7 is not sensi-tive to force along the axis of the beam as this causes equal strain changes in all gauges The fractional resistance change of a strain gauge (DR/R) is related to its fractional
length change or direct strain (DL/L) by its gauge factor Ks:
For wire strain gauges, Ks is typically from 1.75 to 3.5 Strains down to 10–6 may be detected with a bridge circuit The upper limit of strain is around 2 × 10–3, determined by the elastic limit of the beam
A disadvantage of the simple cantilever dynamometer is that the gauges’ strains depend
Forces in machining 141
Fig 5.6 The frequency response of a damped forced vibration system
Trang 10basically on the moment applied to the section at which they are positioned They there-fore depend on the gauges’ distance from where the load is applied, as well as on the size
of the load Better designs, less sensitive to where the load is applied, are the octagonal ring and parallel beam designs shown in Figure 5.8 Supporting the load on well-separated thin sections results in the sum of the strains in the gauges being unchanged when the point
of application of the load is changed, even though the strains are redistributed between the sections It is possible to connect the strain gauges in a bridge circuit so that the output is not sensitive to where the force is applied
The choice of parallel beams or octagonal rings is a matter of manufacturing choice For both, it is important, as a matter of convenience, to minimize cross-sensitivity between the different orthogonal components of electrical output and mechanical input For the parallel beam design, this is achieved by manufacturing the two sets of beams perpendic-ular to each other For the octagonal ring design, it is important to choose a particperpendic-ular shape
of octagon When a circular ring (Figure 5.9) is loaded radially there is zero strain at the positions B and B′, ± 39.9˚ from the point of application of the radial load; likewise when the ring is loaded tangentially, there is zero strain at A and A′, ± 90˚ from the load Gauges placed at A and A′ will respond only to radial loads; and at B and B′ only to tangential
loads The strains will depend on the loads and the ring dimensions (radius R, thickness t and width b) and Young’s modulus E as
Fig 5.7 A strain gauged cantilever dynamometer with its bridge circuit