Volume 16 - Machining Part 18 doc

A 10-year research program on tool materials for high-speed machining concluded that the two major wear mechanisms associated with high-speed machining are high-speed chemical dissolutio

Wear Mechanisms. A 10-year research program on tool materials for high-speed machining concluded that the two major wear mechanisms associated with high-speed machining are high-speed chemical dissolution wear and high-speed diffusion-limited wear (Ref 37) In the range of cutting speeds used in high-speed machining, the chemical dissolution of the tool material into the workpiece is the most important contributor to wear In essence, the tool material dissolves into the flowing chip The tool material that is the most resistant to dissolution exhibits the least wear

The second consideration is that of diffusion-limited wear As the cutting speed is increased, the cutting temperature rises

to a level at which seizure of the chip material occurs everywhere on the tool face This layer of adherent material becomes saturated with tool constituents and serves as a diffusion-boundary layer, reducing the rate of transport of tool material into the chip and consequently the wear rate The wear phenomenon becomes increasingly diffusion-limited, and the observed dramatic decrease in wear occurs with increasing speed Because the diffusivity increases exponentially with temperature, a further increase in cutting speed beyond the speed for minimum wear produces a rapid increase in wear rate, as shown in Fig 10

Fig 10 Effect of cutting speed on the wear rate of cubic boron nitride tooling Workpiece: AISI 4340 steel (35

HRC) Source: Ref 37

Cutting Tool Selection. Consideration of the wear mechanisms of tool materials during high-speed machining

suggests three possibilities for tool development:

• Pick a tool material that is so chemically stable with respect to the workpiece that chemical dissolution

of the tool does not occur to a significant extent, even at the melting point of the workpiece

• Promote the transition to the diffusion limited wear regime

• Isolate the tool from the workpiece rather than a modification of the tool material itself If a protective layer could be introduced between the tool and the chip, transport of the tool constituents into the chip could be prevented The use of viscous lubricants in the profile milling of titanium has proved successful (Ref 38)

Tool Systems for Aluminum Alloys. In aluminum alloys, the cutting temperature is limited by the low melting point and high thermal conductivity of aluminum Chemical effects are minimal, and wear is primarily a result of the abrasion

of the tool material by hard second-phase particles Abrasion resistance increases with the hardness of the tool material; high-speed steel tooling and cemented carbide tooling are suitable for machining most structural alloys, and polycrystalline diamond is preferred for the highly abrasive cast aluminum-silicon (10 to 20% Si) alloys (see the article

"Machining of Aluminum and Aluminum Alloys" in this Volume for a discussion of machining high-silicon aluminum

alloys) Existing tool materials are adequate for machining aluminum alloys at any conceivable speed, with spindle speed and horsepower design limitations setting the upper limit on cutting speed

Tool Systems for Steel. The oxides are the only potential tool materials that are not limited by their chemical stability Therefore, the most promising area for tool material development is in improving the toughness and flow strength of the oxides A second possibility involves the development of tool materials with improved hot strength The development of tool materials other than cubic boron nitride (CBN) with high hot strength in the range of 1300 to 1400

°C (2400 to 2600 °F) might allow a transition from dissolution-limited to diffusion-limited wear, with a corresponding increase in tool life Finally, the wear of CBN in the machining of steels of moderate hardness (35 to 50 HRC) should be investigated to determine whether a transition to a low-wear regime (similar to that which has been observed in hard steels) occurs at sufficiently high cutting speeds If so, machining in the range of 1200 m/min (4000 sfm) and above may

be feasible

Tool Systems for Superalloys. The recommendations are similar to those for steel The oxides are quite chemically stable with respect to nickel and cobalt, making the development of tough oxide materials a priority In addition, the transition to diffusion-limited wear is known to occur at high speeds; therefore, any new high hot strength compositions will likely find application in the machining of superalloys A possible example is the class of tool materials based on silicon nitride (Si3N4) and alloys of Si3N4 and aluminum oxide (Al2O3), referred to as SiAlON (see the article "Ceramics"

in this Volume) These tool materials are very effective in machining nickel-base alloys at high speeds In light of the relatively poor chemical stability of these materials, it is suspected that they represent the second example (in addition to CBN) of a tool material that has sufficient hot strength to enable the very high speed wear transition

Tool Systems for Titanium Alloys. Titanium has low thermal conductivity, low specific heat, and a high melting point These properties ensure that cutting temperatures will be high at even moderate cutting speeds In addition, titanium is highly chemically reactive with all known tool materials, causing rapid wear It is quite likely that the most wear resistant materials tungsten carbide and diamond have already been identified Therefore, the most promising area for investigation is the development of effective lubrication techniques to reduce the interaction between the tool and the chip In addition, new tool geometries, such as the ledge tool described below, have also increased the productivity of machining titanium alloys

Alternative Cutting Tool Geometries (Ref 27, 28)

Rapid tool wear remains a problem in the machining of titanium and other difficult-to-machine alloys, even though cutting speeds for titanium have been recently increased three- to fivefold through judicious choice of cutter grades and geometries, fluids, and machining parameters Partial solutions to the tool life problem lie in new tool geometries and the use of rotating cutters

The Ledge Tool. The concept underlying the ledge tool (Fig 11) is very simple: The tool contains a ledge that is allowed to wear away at a controlled rate with only minimal increase in force Thus, tool wear has not been reduced, but tool life has been greatly increased The size of the ledge (overhang) equals the depth of cut desired, and its thickness equals the ultimate flank wear width to be tolerated In turning, the square tool with the ledged side is brought against the workpiece so that clearance is available between the edge of the ledge and the finished surface of the workpiece (that is, the end cutting edge angle, 1°), as shown in Fig 11 Because the depth of cut is the same or less than the width of the ledge, only the ledge portion of the tool does the cutting and wearing The ledge wear back is due to a combination of flank wear and microchipping wear With this tool, cutting speeds for titanium alloys can be increased five times over conventional speeds, with long tool life ( 30 min) and good finish (Fig 12) The sparking that accompanies tool wear at high cutting speeds has been eliminated by surrounding the tool/workpiece interface with inert gas (N2) or by submerging the workpiece in a cutting fluid A combination of a flood lubricant and an inert gas also serves to eliminate sparks and to keep the workpiece cool

Fig 11 Schematic of ledge tool, which is designed to increase productivity during the high-throughput

machining of titanium (a) Ledge tool mounted on a conventional toolholder (b) A turning operation using a ledge tool Source: Ref 27

Fig 12 Variation of ledge wear back with cutting time for different carbide grades when turning Ti-6Al-4V at

180 m/min (600 sfm) with a depth of cut of 0.75 mm (0.030 in.) and a feed of 0.023 mm/rev (0.009 in./rev) Source: Ref 27

Ledge tools have also been evaluated in the face milling of forged Ti-6Al-2Sn-4Zr-2Mo The results have been comparable to those in turning except the rate of ledge wear back was about three times faster in milling; on a microscopic level, the mode of tool wear is the same Figure 13 shows the variation of ledge wear with cutting time and volume of material removed

Fig 13 Variation of ledge wear back with cutting time (a) and with the volume of material removed (b) for

different carbide grades when face milling a Ti-6Al-2Sn-4Zr-2Mo alloy (36 HRC) Tool: 0.75 mm × 1.0 mm (0.030 in × 0.040 in.) Cutting speed: 155 m/min (515 sfm) Chip load per tooth: 0.23 mm (0.009 in.) Axial depth: 0.75 mm (0.03 in.) Source: Ref 28

Another incremental approach for increasing tool life when machining titanium alloys is a new cutting geometry comprised of a high clearance angle (10 to 15°) together with a high negative-rake angle (-10 to -15°) This geometry will also allow the use of a conventional insert on a modified toolholder (Ref 39)

Rotary tool machining is another technique that holds promise for extending tool life In this method, a tool with a circular cutting edge is allowed to rotate about its own axis, either self-propelled by the cutting process or driven externally at the desired speed As cutting proceeds, new portions of the cutting edge are continuously brought into contact with the workpiece at the cutting zone Thus, increased tool life can be anticipated because of lower temperatures and reduced chemical reactions at the moving chip/tool interface, and reduced cutting forces can be anticipated because of increased options for modifying the chip formation

Applications of High-Speed Machining (Ref 26)

High-speed machining is used in the defense and airframe industries to manufacture aircraft engine propulsion components and in the automobile industry When high-speed machining is accompanied by higher feed rates and spindle power, the higher spindle speeds allow higher removal rates Increased productivity, however, necessitates high-speed, high-power, compact spindle designs; low-inertia feed tables; fast feed drives; quick-response numerical control; and a totally integrated machining system

A typical integrated machining system consists of multiple machining cells that will fabricate large machine parts under hierarchical computer control in an effective, cost-efficient manner (Ref 40) Benefits of the integrated machining system include:

• Reduced labor requirements

• Improved throughput with the application of high-speed and high-throughput machining

• Enhanced production flexibility and reduced work-in-process through establishment of a serial production environment

• Improved product quality and enhanced industrial base

Figure 14 shows an isometric view of an integrated machining system Such a system can take up to more than 9000 m2(100,000 ft2) of floor space

Fig 14 Integrated machining center for the high-speed and high-throughput machining of aluminum and

titanium, respectively AGV, automated guide vehicle; AS, automated storage; RS, retrieval system Source: Ref 40

Airframe and Defense. Most airframe manufacturers have implemented high-speed machining Its primary application is in end milling with small-size cutters Aluminum alloys are the common work materials used, so tool wear

is not a limitation, especially with carbide cutters The ideal candidates for high-speed machining are parts whose machining time is a significant fraction of the floor-to-floor time The ideal cuts are long, straight ones that enable the use

of high feed rates and consequently high removal rates Although some parts may fall under this category, there are others that are suitable for high-speed machining but require extensive pocketing and complex contouring Both pocketing and contouring can involve frequent accelerations and decelerations in feed rates

Tool-changing time can be reduced by using as few cutters as possible, eventually using only one, smallest-diameter cutter capable of generating all the radii on the part Modifying the design of the part may enable the use of high-speed machining and may be desirable if the modification can be achieved without compromising the part performance specifications Further, although high-speed steel tools may be satisfactory for some applications, carbide cutters will not only extend tool life but also provide about three times the stiffness, which is essential for machining long, thin webs with slender end mills

The integrated machining system shown in Fig 14 is used to machine both aluminum and titanium For the high-speed machining of aluminum parts, metal removal rates of 3300 cm3/min (200 in.3/min) are possible using machines with up to

55 kW (75 hp) spindle drives and spindle speeds approaching 20,000 rev/min For the high-throughput machining of titanium alloys, metal removal rates of the order of 165 cm3/min (10 in.3/min) using 95 kW (125 hp) spindle drives and up

to 833 rev/min spindle speeds are possible Many high-speed machining applications, however, do not require such machine capabilities For example, a 15 kW (20 hp), 20,000 rev/min spindle is used to machine A7 wing spars made of 7057-T6 aluminum (Ref 26) The feed rate is 15,000 mm/min (600 in./min) on long, external tapered flanges and 7500 mm/min (300 in./min) in pocket areas; the metal removal rate is 1300 cm3/min (80 in.3/min)

Retrofitted machining centers with 20 kW (26 hp), 18,000 rev/min high-speed spindles are used for the high-speed machining of 7075-T6 aluminum parts for the Trident missile (Ref 26) The worktable has a feed capacity of 5000

mm/min (200 in./min) The high-speed system incorporates such safety features as a double lock for the toolholder and sensors that can shut down the machine quickly if tool breakage creates an imbalance

A major commercial airline manufacturer uses six-axis machining centers for the contour milling of aluminum honeycomb for engine nacelles (Ref 26) The system has a 3.3 m (11 ft) diam rotary table capable of a full 360° rotation The spindle is mounted on the gantry and can tilt 240° and rotate 360° A unique feature of this system is the use of a graphite-fiber reinforced plastic ram This ram provides the required stiffness and reduced weight necessary for the high feed rates of low-inertia parts The high-speed machining systems utilize high-speed (24,000 rev/min) 11 kW (15 hp) spindles

Aircraft Engine Propulsion. Nickel-base superalloys and titanium alloys are the work materials most often used in aircraft engine propulsion components These cause rapid tool wear at high speeds, which constitute a major limitation in the high-speed machining of propulsion parts Until recently, the cutting speeds possible with nickel-base superalloys were about 30 m/min (100 sfm) with carbide cutters and about 9 m/min (30 sfm) with high-speed steel tools The development of CBN and ceramics such as hot-pressed Al2O3 plus TiC, SiAlON, and silicon carbide (SiC) whisker-reinforced Al2O3 has made possible an increase in cutting speeds from 30 to 180 m/min (100 to 600 sfm) for machining nickel-base superalloys These ceramic tool materials are much tougher and more consistent in performance than the ceramics introduced in the 1950s SiAlON and SiC whisker-reinforced Al2O3 are recommended for roughing, and hot-pressed Al2O3-TiC for finishing applications Although new, tougher ceramic tool materials, perhaps based on ceramic composites, will undoubtedly be developed in the future, it is unlikely that cutting speeds will exceed 600 m/min (2000 sfm), because tool wear will continue to be a limitation As described earlier in this article, innovative tool designs and geometries, such as the ledge tool, have improved productivity in titanium machining

Automobile Industry. High-speed machining in the automobile industry is performed on gray cast iron and aluminum alloys, especially the high-silicon type Silicon nitride ceramic and polycrystalline diamond are the important tool materials that permit higher speed with longer tool life Gray cast iron can be machined at a speed of about 1500 m/min (5000 sfm) with Si3N4 tools, and aluminum alloys with high silicon content (10 to 20%) can be machined at about 750 m/min (2500 sfm) with polycrystalline diamond tools

In working with these materials, the current high-speed machining systems need to provide more power and stiffness and improved chip-handling means, controls, and safety features High-speed machining with these materials may require spindle power from 150 to 375 kW (200 to 500 hp) Chip removal rates can reach 16,000 cm3/min (1000 in.3/in.), necessitating efficient chip disposal systems Because products are mass produced in this industry, the nonmachining time should be minimal

Implementing High-Speed Machining (Ref 26)

There are several factors to be considered by companies planning to employ high-speed machining systems Whether high-speed machining is economically appropriate to the application, the relative value of various systems, and how well the system can be integrated into existing operations must all be evaluated Other areas to consider are the research and development support required, system reliability and maintenance, overall safety aspects, general acceptance on the manufacturing shop floor, capital and other investment costs, the skills required, corporate goals, and the financial health

of the company The overall goal should be to improve productivity, reduce costs, and produce parts of a given size, shape, finish, and accuracy at competitive cost

Productivity and overall costs depend on cutting time, noncutting time, labor, and overhead High-speed machining can decrease cutting time by increasing cutting speed Noncutting time can be decreased by the automatic loading and unloading of parts, automatic tool changing, in-process inspection, in-process sensing, and adaptive control Labor costs can be decreased because with high-speed machining fewer operators are needed to work on fewer, more efficient machine tool systems Similarly, overhead costs can be reduced by operating fewer, more efficient machine tools on two

or more shifts and during holidays and at night

In order to evaluate the influence of cutting speed on productivity, floor-to-floor time can be regarded as the sum of the cutting time and the noncutting time Figure 15 shows the percentage decrease in floor-to-floor time with an increase in cutting speed for different ratios of cutting time to floor-to-floor time, using conventional, currently used speeds of whatever the value may be for a given material and process as a base When the cutting time is a significant fraction of the floor-to-floor time and when tool wear at high speed is not significant (solid lines, Fig 15), the cutting speed can be increased considerably to effect a significant reduction in floor-to-floor time If, however, this ratio is low (bottom curve,

Fig 15), an increase in cutting speed by even an order of magnitude or more will result in only a marginal decrease in floor-to-floor time In this case, unless the noncutting time could be decreased significantly, high-speed machining would not be advantageous

Fig 15 Variation of percent decrease in floor-to-floor time with cutting speed See text for details Source: Ref

26

When an increase in cutting speed does not contribute to a significant reduction in floor-to-floor time and when tool wear

is significant at high speeds, as in the machining of titanium alloys (dashed line, Fig 15), high-throughput machining can

be adopted to reduce noncutting time and to increase productivity Similarly, where heavier cuts can be made on the part without affecting either the part finish and accuracy requirements or the cutting tool performance, greater depths of cut (as

in high removal rate machining) are recommended instead of higher speed (see the article "High Removal Rate Machining" in this Volume)

Future Needs (Ref 1). High-speed machining can be cost effective only if other aspects of machining, including reduction of noncutting times and labor costs, can also be improved For this reason, automated machining, such as the integrated machining system discussed earlier, is receiving much attention Dynamic in-process inspection is an integral part of automated machining Sensors and diagnostics have been identified as critical to effective in-process inspection Recently, proximity sensors for measuring tool wear, vibration sensors for measuring tool-touch (between cuts), and lasers for measuring workpiece dimensions have been studied The integrated machining system shown in Fig 14 utilizes

a coordinate-measuring machine as its inspection module

Some of the specific needs for the increased use of high-speed machining are:

• Integrated sensing techniques for monitoring machining processes

• Sensors and diagnostics for detecting tool breakage (including incipient tool breakage)

• Linkage of signal analysis to tool wear mechanisms and machining variables

• Self-teaching adaptive control systems

• Faster response in machine tool controls when machining aluminum

• Better cutting tools (composition and geometry) for machining titanium

• Water-base cutting fluids for machining titanium

• More reliability and predictability in cutting tools

• Greater machine tool rigidity for increased productivity

• More notch-resistant cutting tools for superalloys

In addition to the above-mentioned technical needs, significant reductions in noncutting time can be achieved by procedures that to a large extent involve improved management These procedures include setup, part load/unload, queing, chip cleanup, tool change, maintenance, scheduling, and general operator efficiency Until improvements in these procedures are achieved, cutting times will remain only a small fraction of the overall manufacturing sequence, and the full advantages of high-speed machining will not be realized

References

1 D.G Flom, High-Speed Machining, in Innovations in Materials Processing, G Bruggeman and V Weiss,

Ed., Plenum Press, 1985, p 417-439

2 B.F von Turkovich, Influence of Very High Cutting Speed on Chip Formation Mechanics, in Proceedings

of NAMRC-VII, 1979, p 241-247

3 The 12th American Machinist Inventory of Metalworking Equipment 1976-78, Am Mach., Dec 1978, p

133-148

4 R.I King and R.L Vaughn, A Synoptic Review of High-Speed Machining from Salomon to the Present, in

High-Speed Machining, American Society of Mechanical Engineers, 1984, p 1-13

5 R.L Vaughn, Ultra-High Speed Machining, Am Mach., Vol 107 (No 4), 1960, p 111-126

6 R.L Vaughn, "Recent Developments in Ultra-High Speed Machining," Technical paper 255, Vol 60, Book

1, Society of Manufacturing Engineers, 1960

7 R.L Vaughn, "Ultra-High Speed Machining Feasibility Study," Final Report, Contract AF 33 (600) 36232, Production Engineering Department, Lockheed Aircraft Corporation, June 1960

8 R.L Vaughn, L.J Quackenbush, and L.V Colwell, "Shock Waves and Vibration in High-Speed Milling," Technical Paper 62-WA-282, American Society of Mechanical Engineers, Nov 1962

9 R.L Vaughn and L.J Quackenbush, "The High-Speed Milling of Titanium Alloys," Technical Paper MR 66-151, Society of Manufacturing Engineers, April 1966

10 R.I King and J McDonald, Product Design Implications of New High-Speed Milling Techniques, Trans ASME, Nov 1976

11 F.J McGee, "Final Technical Report for Manufacturing Methods for High-Speed Machining of Aluminum," Technical Registry No 6089, Manufacturing Methods and Technology Branch (DRDMI- EAT), U.S Army Missile Research and Development Command, Feb 1978

12 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Semi-annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No SRD-80-018, Feb 1980

13 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No SRD-80-118, Aug 1980

14 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Semi-annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No SRD-81-018, Feb 1981

15 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Semi-annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No SRD-81-062, Aug 1981

16 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Semi-annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No SRD-82-027, Feb 1982

17 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Annual Technical Report, Air Force Control No F33615-79-C-5119, GE Report No SRD-82-070, Aug 1982

18 D.G Flom, Ed., "Advanced Machining Research Program (AMRP)," Semi-annual Technical Report, Air Force Contract No F33615-79-C-5119, GE Report No 83-SRD-012, Feb 1983

19 D.G Flom, R Komanduri, and M Lee, "Review of Past Work in High-Speed Machining," Paper presented

at the TMS-AIME Meeting (Louisville, KY), The Metallurgical Society, Oct 1981

20 R Komanduri and J Hazra, A Metallurgical Investigation of Chip Morphology in Machining an AISI 1045

Steel at Various Speeds up to 10,100 SFPM, in Proceedings of NAMRC-IX, Society of Manufacturing

Engineers, 1981

21 R Komanduri and B.F von Turkovich, New Observations on the Mechanism of Chip Formation When

Machining Titanium Alloys, Wear, Vol 69, 1981, p 179-188

22 R Komanduri, "Titanium A Model Material for Studying the Mechanism of Chip Formation in Speed Machining," Paper presented at the TMS-AIME Meeting (Louisville, KY), The Metallurgical Society, Oct 1981

High-23 R Komanduri, Some Clarifications on the Mechanics of Chip Formation When Machining Titanium

Alloys, Wear, Vol 76, 1982, p 15-34

24 R Komanduri and R.H Brown, The Mechanics of Chip Segmentation in Machining, J Eng Ind (Trans ASME), Vol 103, Feb 1981, p 33-51

25 R Komanduri, T Schroeder, J Hazra, B.F von Turkovich, and D.G Flom, On the Catastrophic Shear

Instability in High-Speed Machining of an AISI 4340 Steel, J Eng Ind (Trans ASME), Vol 104, May

1982, p 121-131

26 R Komanduri, High-Speed Machining, Mech Eng., Dec 1985, p 65-76

27 R Komanduri, D.G Flom, and M Lee, Highlights of the DARPA Advanced Machining Research Program,

J Eng Ind (Trans ASME), Vol 107, Nov 1985, p 325-335

28 D.G Flom, "Advanced Machining Research Program (AMRP) Final Technical Report," Air Force Contract

No F33615-79-C-5119, GE Report No 83-SRD-040, Oct 1983

29 D.R.C Durham, "Physical Metallurgy of Deformation Localization," Topical report prepared for DARPA, UV-81-DD1, Aug 1981

30 B.F von Turkovich, Influence of Very High Cutting Speed on Chip Formation Mechanics, in Proceedings

of NAMRC-VII, 1979, p 291-297

31 D.R.C Durham and B.F von Turkovich, Material Deformation Characteristics at Moderate Strains and

High Strain Rates, from Metal Cutting Data, in Proceedings of NAMRC-X, 1982, p 324-331

32 B.F von Turkovich and D.R.C Durham, Machining of Titanium and Its Alloys, in Proceedings of the Symposium on Advanced Processing Methods for Titanium (Louisville, KY), The Metallurgical Society,

35 T.A Schroeder and J Hazra, High Speed Machining Analysis of Difficult-to-Machine Materials, in

Proceedings of NAMRC-IX, Society of Manufacturing Engineers, 1981

36 J.P Kottenstette and R.F Recht, Ultra-High-Speed Machining Experiments, in Proceedings of NAMRC-X,

1982, p 263-270

37 B.M Kramer, On Tool Materials for High-Speed Machining, in High-Speed Machining, American Society

of Mechanical Engineers, 1984, p 127-140

38 J Jensen, "High-Speed Milling of Titanium," M.S thesis, Massachusetts Institute of Technology, 1983

39 R Komanduri and W Reed, Jr., Evaluation of Carbide Grades and a New Cutting Geometry for Machining

Titanium Alloys, Wear, Vol 92, 1983, p 113-123

40 Integrated Machining System An Overview, LTV Aircrafts Products Group, Military Aircraft Division,

1988

High Removal Rate Machining

Introduction

HIGH REMOVAL RATE (HRR) MACHINING involves the use of extremely rigid, high-power, high-precision machines, such as roll turning lathes, to achieve material removal rates far beyond the capacity of conventional machine tools Material removal rates as high as 6000 cm3/min (370 in.3/min) have been reached using multiple ceramic cutters on high-carbon (0.8% C) heat treated cast steel rolls This article will review the machine requirements, cutting parameters, and applications associated with HRR machining Additional information can be found in Ref 1, 2, 3, 4, and 5

Acknowledgement

The editors would like to thank Jack Binns, Sr., inventor of the "Super-Lathe," for his valuable contributions to this article

Machine Requirements

The key to success in HRR machining is extreme rigidity in machine, workpiece, and cutting tool setups Figure 1 shows

a "Super-Lathe" capable of making cuts as deep as 25 mm (1 in.) at feeds up to 1.3 mm/rev (0.050 in./rev) on hardened steel and chilled cast iron rolls Metal removal rates as high as 4500 kg/h (10,000 lb/h) can be achieved with such machines, which have power ranges from 100 to 450 kW (150 to 600 hp) and can produce up to 400 kN (95,000 lbf) of output torque at the spindle Workpieces of softer materials can also be turned, producing large, segmented chips, as shown in Fig 2

Fig 1 Superlathe used for HRR machining This 300 kW (400 hp) machine, which is computer numerically

controlled, can perform rough cuts on 1290 mm (50.75 in.) diam rolls and finishing cuts on workpieces up to

1200 mm (47.25 in.) in diameter at maximum roll lengths of 7.8 m (25 ft, 7 in.) Courtesy of J Binns, Sr., Binns Machinery Products

Fig 2 A large AISI 1055 steel (180 HB) chip cut on an HRR machining lathe Courtesy of J Binns, Sr., Binns

Machinery Products

The lathe bed, which is the main structural member of the "Super-Lathe," is designed to withstand applied loads with minimal deflection The structure is an elongated steel box fabricated from hot-rolled ASTM A 36 steel that has a high elastic modulus The six exterior plates of the box become the principally stressed members When fabricated correctly, the unit needs no vibrational damping, and when subjected to maximum rated loads, maximum unit stresses of only approximately 3450 kPa (500 psi) will develop in the critically stressed members

The tailstock of the machine features a rectangular quill that can withstand radial loads two or three times as great, for a given amount of deflection, as an equivalent circular quill with an identical bore and wall thickness The tailstock is also fabricated from A 36 structural steel plate

The method of clamping the tailstock to the bed again provides for maximum rigidity There are four clamp bars at the bottom edge of the tailstock; when pulled up with 100 J (75 ft · lbf) of torque, the force required to separate the tailstock from the bed is approximately 5500 kN (625 tonf)

The headstock is also made of heat-treated wrought steel In the power shafting, no shaft between gears is smaller in diameter than the length between the gears The main spindle bearing, the bull gear, and the ring gear (when used) are all the same nominal diameter as the maximum size of the roll the lathe will turn Additional information on superlathe gearing and electrical drive systems can be found in Ref 2 and 3 Workpieces are gripped by an eight-jaw chuck

The carriage, which carries the cutter, must be capable of withstanding the maximum torque force and feed forces with only a fraction of a thousandth of an inch of deflection The weight and mass of the cross slide and saddle are important considerations because inertia can substantially reduce the changes in deflection of the carriage as the feed and cut forces vary at high frequency For example, cutting forces suddenly drop from a maximum to almost zero when each segment of the chip ruptures from the workpiece Antifriction ways (with linear ball and roller bearings) to the cross slide and bed are

a major factor in steadying the carriage

Tooling for HRR machining is also of considerable importance Because the cutter shank is the link between the insert cutting edge and the machine, the same considerations in terms of rigidity must be given it as the bed, headstock, and carriage A number of criteria must be considered in shank design The first is that the larger and more massive the shank, the better it will cut A large monobloc shank is both more rigid and a better heat sink than either a small shank or one comprised of a small interchangeable toolholder placed in a large shank In addition, a properly matched shank socket and insert will reduce insert breakage Because aluminum oxide (Al2O3) and Al2O3 + TiC inserts are low in transverse rupture strength as compared to alternative cutting tool materials (Table 1), it is essential that the internal stresses from the cutting forces and thermal distortion be confined as much as possible to the compressive state

Table 1 Transverse rupture strengths of common cutting tool materials

Transverse rupture strength Tool material

Coated tungsten carbide 1030-2070 150-300

Inserts should be mechanically clamped in place with only enough force to secure them while the cut is starting After the cut has begun, the cut force will hold the inserts When initially clamping the inserts, it is important to have them solidly against the stops in the socket

Evaluating Machine Rigidity. The manner in which chips flow from the work indicates the variation in deflections of the cutting edge with respect to the workpiece and therefore provides a good evaluation of rigidity Adequate rigidity has been achieved when the chip peels off as a smooth, continuous ribbon from the work (Fig 3) After a short period of smooth flow, the ribbon will fall apart into short lengths from its own weight or movement, then drop into the chip chute

If the chips fly off like snowflakes, there is insufficient rigidity, and the insert will soon break

Fig 3 Continuous 115 mm (4 in.) wide ribbon of machined material peeling off an HRR cutter Courtesy of J

Binns, Sr., Binns Machinery Products

Cutting Tools and Cutting Parameters

Although cemented carbide and cubic boron nitride (CBN) tools can be used for HRR machining, Al2O3 + TiC tooling is normally used, particularly for finishing cuts on ferrous alloys (surface finishes of 0.25 to 0.75 m, or 10 to 30 in., can

be produced) Alumina-base inserts with a density of 4.30 g/cm3 and a hardness of 95 HRA are commonly used Carbide tools are used for some roughing cuts Both round and rectangular insert configurations are used, with the latter being the most common (Fig 4) For rectangular inserts, negative-rake angles of 5 to 25° with 5 to 10° clearance are used;

side cutting edge angles (lead angles) range from at least 30° up to 90° for lap cutting at fast speeds Depths of cut to 75

mm (3 in.) are possible Speeds range from 6 to 240 m/min (20 to 800 sfm), with feeds reaching as high as 6.4 mm/rev (0.250 in./rev) at low speeds The effects of cutting parameters on tool life are reviewed in Ref 3 and 4 Detailed information on cemented carbide, CBN, and ceramics can be found in the Section "Cutting Tool Materials" in this Volume

Fig 4 Ceramic insert (bottom) used for HRR machining, along with the resulting chip (top) The insert cutting

edge is that adjacent to the scale Courtesy of J Binns, Sr., Binns Machinery Products

As stated earlier in this article, metal removal rates for ferrous materials can range as high as 6000 cm3/min (370 in.3/min) with multiple inserts A 375 kW (500 hp) lathe can make cuts at more than 105 m/min (350 sfm) at a feed of 2 mm/rev (0.08 in./rev) and depths of cut of 75 mm (3 in.) in as-forged 52100 bearing steel rolls This results in about 7700 kg (17,000 lb) of chips produced per hour, which equals an energy rate of metal removal of 44 cm3/min/kW (2 in.3/min/hp) Turning tests on Inconel 718 and Ti-6Al-4V have produced metal removal rates of more than 800 cm3/min (50 in.3/min) The workpiece finish was between 0.75 and 2.5 m (30 and 100 in.) With aluminum, it has been shown that removal rates of 3300 cm3/min (200 in.3/min) are attainable on each insert of a gang tool (Ref 5)

References

1 Staff Report, Mech Eng., March 1984, p 55-59

2 J Binns, Sr., "Rough Turning and Hogging With Ceramic Cutters," Papers presented at the Creative Manufacturing Seminars, American Society of Tool and Manufacturing Engineers, 1963-1964

3 J Binns, Sr., "Ceramic Cutter Performance on Rough Turning and Hogging Cuts," Technical Paper 633, American Society of Tool and Manufacturing Engineers, 1964

4 J Binns, Sr., Super-Lathe for Roll Turning, Iron Steel Eng., Oct 1961

5 D.G Flom, Ed., "Advanced Machining Research Program," Annual Technical Report, Air Force Contract

No F33615-79-C-5119, General Electric Co Report No SRD-82-070, Aug 1982

to achieve desired accuracies for small and medium-sized production runs The Parsons Corporation subcontracted the development of the control system to the Servomechanism Laboratory of the Massachusetts Institute of Technology (MIT) The challenge was met successfully by MIT, and in 1952 a three-axis Cincinnati Hydrotel milling machine controlled with digital technology was developed This digital technology was termed numerical control (NC) Thus, the primary motivations for the development of NC systems for machine tools were the demand for high accuracy in the manufacture of complicated parts and the desire to shorten production time

Evolution of Numerical Control

The first NC controllers in the 1950s used vacuum tubes and were extremely large Controllers in the early 1960s used transistors in their logic circuits and digital control loops These units were more reliable The third generation of NC controllers used integrated circuit chips and consequently became less expensive, more reliable, and smaller Many of these third-generation controllers are still in operation In all of these systems, the numerical data required for producing a part is maintained on punched tapes and inserted to the controller through a built-in tape reader

An important advance in the philosophy of machine tool numerical control that took place in the early 1970s was the shift toward the use of computers instead of controller units in NC systems This produced both computer numerical control (CNC) and direct numerical control (DNC) Computer numerical control is a self-contained NC system for a single machine tool including a dedicated computer controlled by stored instructions to perform some or all of the basic NC functions With DNC, several machine tools are directly controlled by a central computer Of the two types of computer control, CNC has become much more widely used for manufacturing systems (for example, machine tools, welders, and laser beam cutters) mainly because of its flexibility and the lower investment required The preference of CNC over DNC

is increasing as a result of the availability and declining costs of minicomputers and microcomputers

One of the objectives of CNC systems is to replace as much of the conventional NC hardware as possible with software and to simplify the remaining hardware There are many ways in which functions can be shared between software and hardware in such systems, but all involve some hardware in the controller dedicated to the individual machine (Ref 1) This hardware must contain at least the servoamplifiers, the transducer circuits, and the interface components, as shown in Fig 1

Fig 1 Block diagram of a CNC machine

The software of a CNC system consists of at least three major programs:

of the machine

The CNC controllers of the 1980s are more powerful and more user friendly than earlier units They incorporate shooting features such as on-board diagnostics, which allow self-testing of the controller, and simulation mode, which is used to test part programs without generating axes motions Many controllers offer high-level programming facilities, three-dimensional tool path animation with graphics, tool data base, and preselection of cutting parameters The modern machine controller is a workstation in a Local Area Network (Ref 2) It is capable of communicating with other controllers and of being integrated into a flexible manufacturing system in the future, thus permitting the gradual construction of a full flexible manufacturing system

trouble-Numerical control was introduced and developed in the metalworking industry, and the largest concentration of NC equipment remains in metalworking shops Numerical control has been successfully implemented for turning, milling, drilling, grinding, boring, punching, and electrical discharge machines It is interesting to note that numerical control has made possible the development of machines with basic capabilities that far surpass those of conventional machines For example, sophisticated NC milling machines maintain control over five axes of motion and can literally sculpt complex surfaces (Ref 3) A new breed of NC machine tool is the machining center and the turning center, which incorporate the functions of many machines into a single device A machining center can access multiple tools to perform such operations

as milling, drilling, boring, and tapping (Fig 2) A turning center is a powerful lathe equipped with an automatic tool changer Other types of NC machines include welding machines, drafting machines, tube benders, inspection machines, and wiring machines in the electronics industry

Fig 2 CNC machining center Courtesy of Cincinnati Milicron

Fundamentals of Numerical Control

Numerical control equipment has been defined by the Electronic Industries Association (EIA) as a system in which actions are controlled by the direct insertion of numerical data at some point The system must automatically interpret at least some portion of these data

In a typical NC or CNC system, the numerical data required for producing a part are maintained on a disk or on a tape and called the part program The part program is arranged in the form of blocks of information Each block contains the numerical data required to produce one segment of the workpiece profile The block contains, in coded form, all the information needed for processing a segment of the workpiece: the segment length, its cutting speed, feed rate, and so on Dimensional information (length, width, and radii of circles) and the contour form (linear, circular, or other) are taken

from an engineering drawing Dimensions are given separately for each axis of motion (X, Y, and so on) Cutting speed,

feed rate, and auxiliary functions (such as coolant on and off, spindle direction, clamp, and gear changes) are programmed according to surface finish, tolerance, and machining requirements (Ref 1, 2, 3, 4, 5, and 6)

Compared to a conventional machine tool, the NC system replaces the manual actions of the operator In conventional machining, a part is produced by moving a cutting tool along a workpiece by means of powered slides that are engaged and disengaged by an operator Contour cuttings are performed by an expert operator by sight On the other hand, the operators of NC machine tools need not be skilled machinists They only have to monitor the operations of the machine, operate the tape reader, and load and remove the workpiece Most intellectual operations that were formerly done by the operator are now included in the part program However, because the operator works with a sophisticated and expensive system, intelligence, clear thinking, and good judgment are essential qualifications of a good NC operator

Preparing the part program of an NC machine tool requires a part programmer The part programmer should possess knowledge and experience in mechanical engineering fields Knowledge of tools, cutting fluids, fixture design techniques, machinability data, and process engineering are all of considerable importance The part programmer must be familiar with the function of NC machine tools and machining processes and must decide on the optimum sequence of operations The part program can be written manually, or a computer-assisted language, such as the automatically programmed tool language, can be used

In NC machines, the part dimensions are presented in part programs by integers In CNC machines, the dimensions in part programs are sometimes expressed as numbers with a decimal point, but are always stored in the computer as integers Each unit of these integers corresponds to the position resolution of the axes of motion and is referred to as the basic length unit (BLU) The BLU is also known as the increment size or bit weight, and in practice it corresponds approximately to the accuracy of the NC system

In NC and CNC machine tools, each axis of motion is equipped with a separate driving device, which replaces the handwheel of the conventional machine The driving device may be a dc motor, a hydraulic actuator, or a stepping motor The type selected is determined by the power requirements and the machine

An axis of motion in numerical control means an axis in which the cutting tool moves relative to the workpiece This

movement is achieved by the motion of the machine tool slides The main three axes of motion are referred to as the X-,

Y-, and Z-axes The Z-axis is perpendicular to both X and Y in order to create a right-hand coordinate system For

example, in a vertical drilling machine (as one faces the machine), a +X command moves the worktable from left to right,

a +Y command moves it from front to back, and a +Z command moves the drill up, away from the workpiece The X-, Y-, and Z-axes are always assigned to create a right-hand cartesian coordinate system

Each axis of motion also has a separate control loop The control loops of NC or CNC systems use two types of feedback devices (Fig 1): tachometers to monitor velocity and encoders or other position transducers (for example, resolvers) to measure position The controller compares the actual position with the required one and generates an error The control loop is designed in such a way as to reduce the error, that is, the loop is a negative-feedback type A common requirement

of continuous-path NC and CNC systems is the generation of coordinated movement of the separately driven axes in order to achieve the desired path of the tool This coordination is accomplished by interpolators Numerical control systems contain hardware interpolators, but in CNC systems interpolators are implemented by software

to another during the machining of the workpiece With a conventional machine tool, the work must be stopped at such points because the operator must go to the next step The operator must stop the cutting process frequently and measure the part dimensions to ensure that the material is not overcut It has been proved that the time wasted on measurements is frequently 70 to 80% of the total working time (Ref 1) The rate of production is also decreased because of operator fatigue In NC systems, these problems do not exist Because the accuracy is repeatable with numerical control, inspection time is also reduced

Numerical control produces higher-quality parts and makes possible the accurate manufacture of more complex designs without the usual loss in accuracy encountered in conventional manufacturing Producing a part that must be cut with an accuracy of 0.01 mm (0.0004 in.) or better may take a considerable amount of time using conventional methods In numerical control using single-axis motion, obtaining such accuracies is the state-of-the-art, and they are maintained throughout the entire range of cutting speeds and feed rates

Another intangible advantage of numerical control is the production of complex parts that are not feasible in conventional manufacturing Complex-contour cutting in three dimensions cannot be performed by manual operation Even when it is possible, the operator must manipulate the two handwheels of the table simultaneously while maintaining the required accuracy; thus, it becomes possible only when the part is simple and requires relatively low accuracies It is obvious that

in such work the NC machines save a considerable amount of time

Compared to conventional machining methods, the NC machine tool has the following advantages:

• Complete flexibility; a part program is needed only for producing a new part

• Accuracy is maintained through the full range of speeds and feeds

• The possibility of manufacturing a part of complicated contour

• A shorter production time

• Higher productivity achieved by saving indirect time, such as setting up and adjusting the machine and using one operator to monitor several machining operations, or by using completely automatic operation

in unmanned production

Programming

Most NC and CNC machine tools use off-line programming methods, which can be either manual or computer assisted, such as programming with the aid of the automatically programmed tool (APT) language During off-line programming, the machine remains in operation while a new part program is being written Typically, when a part program is ready, it is stored on a punched tape or a floppy disk The tape or disk is taken to the machine shop and loaded into the machine tool controller, and the part is subsequently produced

Manual part programs are written by programmers First, the programmers must determine the machining parameters and the optimum sequence of operations to be performed Based on this sequence, they calculate the tool path and write a manuscript Each line of the manuscript, which is referred to as a block, contains the required data for transferring the cutting tool from one point to the next, including all machining instructions that should be executed either at the point or along the path between the points

The EIA standard RS-273-A ("Interchangeable Perforated Tape Variable Block Format for Positioning and Straight Cut Numerical Controlled Machines") provides a line format for point-to point (PTP) and straight-cut NC machines A typical line according to this standard is as follows:

N102 G01 X-52000 Y9100 F315 S717 T65432 M03 (EB)

The letter and the number that follows it are referred to as a word For example, X-52000 and M03 are words The first

letter of the word is the word address Word addresses are denoted as follows: N, sequential number, G, preparatory

function; X and Y, dimensional words; F feed rate code; S, spindle speed code; T, tool code; M, miscellaneous function; and EB, end-of-block character The EB character is not printed but only punched or coded, and it is usually the carriage-return code, thus permitting a new line to begin immediately afterward The EB character indicates to the NC controller that the current reading is completed and that the axes of motion must start up When this motion is accomplished, the next block is read

According to EIA standards RS-273-A and RS-274-D ("Interchangeable Variable Block Data Contouring Format for Positioning and Contouring\Positioning for Numerically Controlled Machines"), two block formats are available: the word address format and the tab sequential format In the word address format, each word must be headed by the word address The controller uses the address letter to identify the word that follows it In this type of format, words need not be arranged in any specific order within the block because the letter identifies the corresponding word However, because the standards prescribe a definite sequencing of words, the format recommended above is used for PTP programming

In the tab sequential format, a tab character is inserted between each two words in the block, and the address letter is omitted In this format, words must be arranged in a specific order When a word is not needed in a particular block, it can

be omitted, but the corresponding tab character must be punched

Manual part programming can be applied to PTP systems, but it is too complex for contouring systems Therefore, a great number of computer software systems have been developed to assist in NC part programming The APT system is the most widespread and the most comprehensive one It is available on many computers and is widely used by many manufacturers of NC equipment

The first prototype of the APT software system was developed in 1956 by the Electronic System Laboratory of MIT The program was further developed by the cooperative efforts of 21 industrial companies sponsored by the Aerospace Industries Association with assistance from MIT As a result of these efforts, a system called APT II was produced in

1958, and a more effective system, the APT III, was distributed in 1961 The Illinois Institute of Technology Research Institute was elected to direct the future expansion of the program, and the capabilities of APT are continually being expanded The current APT language has a vocabulary of more than 600 words

The APT program is a series of instructions that specifies the path the tool must follow to produce a part To communicate the tool path to the computer, one must provide the computer with geometric descriptions of the part

surfaces The APT language enables the programmer to accomplish this and then to specify the manner in which the tool should move along these surfaces The geometric expressions describing the part and the motion statements represent about 70% of the average program

A geometric expression defines a geometric shape or form For each geometric form, there are many different methods of definition Definitions of at least 16 different geometric forms are contained in APT: the most useful ones are POINT, LINE, PLANE, CIRCLE, CYLINDER, ELLIPS, HYPERB, CONE, and SPHERE Several examples of definitions are presented below

Most APT statements are divided into two sections, major and minor, which are separated by a slash The major section appears to the left of the slash and is generally one word containing one to six letters The minor section, if required, appears to the right of the slash and contains modifiers to the major portion For example, a point can be defined by:

POINT/X-coordinate, Y-coordinate, Z-coordinate

An example of a corresponding geometric statement is:

PT2 = POINT/3,4

where PT2 is the symbolic designation of a point whose X-coordinate is 3 and whose Y-coordinate is 4

A line can be defined by a point and a tangent circle (Fig 3):

LINE/symbol for a point, , TANTO, symbol for a circle

The modifiers LEFT or RIGHT are applied looking from the point toward the circle Examples are:

L1 = LINE/P1, LEFT, TANTO, CIRI L2 = LINE/P1, RIGHT, TANTO, CIRI

Fig 3 Line definition by a point and a tangent circle

Once the required part has been defined with the geometric expressions, tool movements are specified using the motion statements Each motion statement will move the tool either to a new location or along a surface specified by the statement Examples of motion statements are:

GOTO/HOLE2 GOLFT/L1, PAST, L2

It should be noted that L1, L2, and HOLE2 are identifying names Identifying names are given to geometric expressions and cannot be APT words In addition to the geometric and motion statements, there are other kinds of statements and features One of the most useful statements is the CLPRNT The CLPRNT is an instruction to the APT system to produce

a printed list of all the cutter location coordinates that have been computed The computation results are those of the APT program before postprocessing

The output of the APT program is sent as input to another program called the postprocessor The latter is a program written specifically for each CNC machine tool system, and it includes information about the particular machine (size, accelerations, coolant, and so on) The output of the postprocessor is the NC part program that is loaded onto the machine controller to produce the required part

Numerical Control

Numerically controlled machine tool systems are in some ways related to industrial robot systems In both, the axes of a mechanical device are controlled to guide a tool that performs a manufacturing process task (Fig 4) In both systems, each axis of motion is equipped with a separate control loop and a separate driving device The difference is in the process and the mechanical device The process in numerical control may be drilling, milling, grinding, or welding, and in robotics, painting, welding, assembly, or handling In numerical control, the mechanical handling device is the machine tool, and in robotics, the manipulator Machine tools, however, are more rigid and usually have perpendicular axes of motion, which result in simpler control strategies and better position accuracy than those of robot arms

Fig 4 Schematic of an NC or industrial robot system

A common feature in numerical control and robotics is that the required path of the tool is generated by the combined motion of the individual axes In numerical control, the tool is the cutting tool, such as a milling cutter or a drill; in robotics, the tool is the instrument at the far end of the manipulator, which might be a gripper, a welding gun, or a paint-spraying gun Robot systems, however, are more complex than machine tool CNC systems for the reason discussed below

Machine tools require control of the position of the tool cutting edge in space In many cases, the control of three axes is adequate Robots require the control of both the position of the tool center point and the orientation of the tool; this is achieved by controlling six axes of motion (or degrees of freedom)

Some robot systems use more than six axes, and some CNC machines use more than three axes of motion A typical example is the addition of a rotary table to a three-axis milling machine On the other hand, many robot systems use fewer than six axes, and in such cases the wrist section contains fewer than three degrees of freedom (Ref 7) Similarly, there are CNC machines with only two numerically controlled axes of motion For example, only two numerically controlled axes are required for turning parts on a lathe because the parts are symmetrical

System Structure

Computer numerical control systems can be divided into PTP and continuous-path or contouring systems (Ref 1, 2, 3, 4,

5, and 6) A typical PTP system is encountered in a CNC drilling machine In a drilling operation, the machine table moves until the point to be drilled is exactly under the tool, and then the hole is drilled The table then moves to a new point, and another hole is drilled This process is repeated until all the required holes on the part are drilled The machine table is then brought to the starting point, and the system is ready for the next part

In more general terms, the PTP operation is described as follows The machine tool moves to a numerically defined position, and then the motion is stopped The tool performs the required task with the machine table stationary Upon completion of the task, the machine tool moves to the next point, and the cycle is repeated

In a PTP system, the path and the velocity while traveling from one point to the next are insignificant Therefore, as shown in Fig 5, a basic PTP system would require only position counters for controlling the final position of the machine table in order to bring it to the target point (Ref 8) The coordinate values for each desired position are loaded into the counters with a resolution that depends on the basic length unit of the system During the motion of the table, the encoder

at each axis transmits pulses, each representing the tool travel of 1 BLU along the axis Each axis of motion is equipped with a counter to which the corresponding encoder pulses are transmitted At the beginning of a motion from a point, each axial counter is loaded by the corresponding required axial incremental distance (in BLUs) to the next point During the motion of the table, the contents of each counter are gradually decremented by the pulses arriving from the corresponding encoder When all counters are at zero, the machine table is in its new desired position

Fig 5 Block diagram of a point-to-point system

In continuous-path CNC machines, the tool performs the task while the axes of motion are moving, as in milling machines The task of the controller in milling is to guide the tool along the programmed path In continuous-path systems, all axes of motion may move simultaneously, each at a different velocity These velocities, however, are coordinated under computer control in order to trace the required path or trajectory In a continuous-path operation, the position of the tool at the end of each segment (motion step), together with the ratio of axes velocities, determines the generated contour, and at the same time the resultant velocity also affects the quality of the surface

Continuous-path systems require interpolators for determining the path between given end points (Fig 6) Part programs supply only the end points of the segments along the contour, and the system computer interpolates the path between the points and generates in real time the commands to the individual axes of motion Typically, NC and CNC machines are capable of linear and circular interpolation

Fig 6 Block diagram of a contouring system

The principle of operation for linear interpolators of machine tools is relatively simple (Ref 9) The axes of motion in machine tools are perpendicular to each other, thus creating a cartesian coordinate system A motion along a straight path

with a length L employs the following relationships:

The distances x, y, and z are the components of L in the X, Y, and Z directions, respectively, and V is the required velocity

along the path (referred to as feed rate in machining) Equations 1, 2, and 3 provide the basis of linear interpolator algorithms for machine tools Circular interpolator algorithms are more complicated and can be found in the literature (Ref 10, 11, 12)

Acceleration and deceleration are achieved by varying V in Eq 1, 2, and 3 according to a predetermined formula

Machining feed rates (the tangential velocity of the tool along the contour) in machine tools are small; and therefore, paths that include acceleration and deceleration periods are not often encountered

Adaptive Systems

Computer numerical control machines in production are programmable systems that can repeat a sequence of programmed operations as long as necessary However, these systems are unable to sense and respond to any changes in their working environments For example, assume that a CNC lathe is used to turn a batch of cylindrical workpieces to a

200 mm (8 in.) radius from raw material having radii from 204 to 207 mm (8.04 to 8.16 in.) The machining feed rate in the part program is calculated as the maximum allowable feed rate needed to remove a maximum load of 4 mm (0.16 in.) without cutter breakage, although with the smaller load the feed rate can be increased without risking the cutter This and similar situations can be remedied only if the systems are able to adapt to the changing conditions in their environments; such changes would be sensed with suitable sensors This is basically the motivation for introducing adaptive control systems in machining The objectives of these adaptive controls are to increase productivity in rough machining and to improve part accuracy in fine cutting (Ref 13) The article "Adaptive Control" provides a more thorough discussion of adaptive control for machine tools

References

1 Y Koren, Computer Control of Manufacturing Systems, McGraw-Hill, 1983

2 P Ranky, Computer Integrated Manufacturing, Prentice-Hall, 1986

3 R.S Pressman and J.E Williams, Numerical Control and Computer-Aided Manufacturing, John Wiley &

Sons, 1977

4 N.O Olesten, Numerical Control, Wiley-Interscience, 1970

5 M.P Groover, Automation, Production Systems, and Computer-Aided Manufacturing, Prentice-Hall, 1980

6 J Pusztaai and M Sava, Computer Numerical Control, Reston Publishing, 1983

7 Y Koren, Robotics for Engineers, Mc-Graw-Hill, 1985

8 J.T Beckett and H.W Mergler, Analysis of an Incremental Digital Positioning Servosystem With Digital

Rate Feedback, J Dyn Sys Meas Contr (Trans ASME), Vol 87, March 1965

9 G Ertell, Numerical Control, Wiley-Interscience, 1969

10 B.W Jordan, W.J Lenon, and B.D Holm, An Improved Algorithm for Generation of Nonparametric

Curves, IEEE Trans Comp., Conf C-22, No 12, Dec 1973, p 1052-1060

11 O Masory and Y Koren, Reference-Word Circular Interpolators for CNC Systems, J Eng Ind (Trans ASME), Vol 104, 1982

12 A.N Poo and J.G Bollinger, Dynamic Errors in Type I Contouring Systems, IEEE Trans Ind Appl., Vol

1A-8 (No 4), July 1972, p 477-484

13 G.A Ulsoy, Y Koren, and F Rasmussen, Principal Developments in Adaptive Control of Machine Tools,

J Dyn Sys Meas Contr (Trans ASME), Vol 105 (No 2), June 1983, p 107-112

In CNC systems, the relative position between the tool and the workpiece is controlled; however, the part programmer must specify the cutting speed and feed rates The determination of these cutting parameters requires experience and knowledge regarding workpiece and tool materials, machine characteristics, coolant effects, and so on Their selection directly affects such important economic factors as product dimensional accuracy, surface finish, metal removal rates, tool wear rates, and tool breakage The main focus of adaptive control is the improvement of these production and product quality related factors in the actual machining process This is accomplished using the measurement and control of certain process variables in real time as well as several alternative strategies for the AC algorithm A schematic of a typical AC configuration for a machine tool is shown in Fig 1 It is clear that the adaptive control represents a process control system that operates in addition to the CNC position, or servo, control system In this discussion, the different types of AC systems that have been proposed are considered first and then the economic benefits of AC systems

Fig 1 Schematic of an adaptive control system incorporated into a CNC machine tool system Source: Ref 3

Adaptive control systems use a variety of sensors and control strategies Depending on these factors, as well as the particular process being controlled, the AC systems can conveniently be classified as (Ref 3, 4):

• Adaptive control with optimization (ACO), in which an economic index of performance is used to optimize the cutting process using on-line measurements

• Adaptive control with constraints (ACC), in which the process is controlled using on-line measurements

to maintain a particular process constraint (for example, a desired force or power level)

• Geometric adaptive control (GAC), in which the process is controlled using on-line measurements to maintain desired product geometry (for example, dimensional accuracy or surface roughness)

These three types of AC systems are discussed in some detail in the sections that follow It is worth mentioning here that the ACO system is the most general However, because it is difficult to implement an ACO system, the suboptimal ACC and GAC strategies are often used These can typically provide much of the benefit to be expected from ACO systems and are easier to implement The ACC-type systems are well suited for rough cutting, and the GAC-type systems are typically used in finishing operations

The economic benefits of AC systems can be significant, particularly under varying cutting conditions (Ref 5) However, the main objective is improvement in productivity, for example, by increasing the metal removal rate (MRR) This is illustrated in Fig 2 for a milling operation with a variable depth and/or width of cut With adaptive control, the feed rate can be increased when the depth and/or width of cut is small, and reduced if either becomes larger By contrast, in conventional milling, the smallest feed rate would be selected based on the worst-case conditions for that particular part

Fig 2 Comparison of feed rates in adaptive and conventional (nonadaptive) milling when cut varies (a)

Variable depth (b) Variable width Source: Ref 3

Some typical results demonstrating the economic benefits of adaptive control are given in Fig 3 These compare machining costs in adaptive and conventional (non-adaptive) machining and show that improvements in productivity of 20% to 80% can be expected with the use of AC systems The actual improvement depends on the material being machined and the complexity of the part being produced Improvements or consistency in product quality can also be an objective of AC systems (usually GAC systems) Another benefit is the reduced part programming time because feeds and speeds are adjusted on-line The economic impact of adaptive control also depends on the percentage of total production time allocated to actual cutting, and currently this is often only 5 to 20% As automation reduces the percentage of production time required for part setup and tool changing, the percentage of time the machine spends in actual cutting increases, and the advantages of adaptive control become even more significant

Fig 3 Machining cost comparison for adaptive and conventional (nonadaptive) machining Bar graphs (a) and

(b) compare a Cincinnati Milacron system with nonadaptive methods under the following conditions: (a) the machining of mild steel with a 0.05 mm (0.002 in.) tolerance and a cut 0.75 mm (0.003 in.) deep and 25 mm

(1 in.) wide, and (b) the machining of stainless steel with a 0.25 mm (0.01 in.) tolerance and a cut 2.5 mm (0.1 in.) deep and 25 mm (1 in.) wide Bar graphs (c), (d), and (e) compare a Bendix system with nonadaptive methods when variable-depth cuts (top graph) and a 1 mm (0.05 in.) depth of cut (bottom graph) are used Machining conditions are as follows: (c) machining of 4140 steel with a high-speed steel cutter, (d) machining

of 4140 steel with a carbide cutter, and (e) machining of stainless steel with a high-speed steel cutter Source: Ref 3

Adaptive Control With Optimization

This section describes two systems, a milling and a grinding system, based on an ACO strategy While ACO systems represent the most general class of AC systems, they have not found application in industry The two systems described here indicate the ACO strategy and also help to illustrate the limitations that have prevented their use in industry The first system described is a laboratory system developed in the mid-1960s for milling and used industrially in a few installations The ACO system described for grinding is a more recent development that has found off-line application in industry However, its on-line implementation has so far been limited to the laboratory

ACO System for Milling. The best-known ACO system is probably the system developed by The Bendix Corporation for the U.S Air Force in the early 1960s (Ref 5, 6) The structure of this ACO system for milling is illustrated in Fig 4 This system will not be described in detail here, but requires some mention because of its historical significance and because it illustrates many aspects of ACO systems It uses several measurements of process variables (cutting torque, tool temperature, and machine vibration) and an economic index of performance for on-line adjustment of the cutting feed

and speed The performance index (J) can be expressed as:

(Eq 1)

where MRR = waV is the metal removal rate (in.3/min), w is the milling width (in.), a is the depth-of-cut (in.), V is the milling feed rate (in./min), TWR = K1(MRR) + K2 + K3(dT/dt) is the tool wear rate (in./min), is the tool temperature (°F), T is the cutting torque (lbf · ft), W0 is the allowable width of flank wear (in.), f is the feed (in./rev), N is the spindle speed (rev/min), C1 is the cost of machine and operator per unit time ($/min), C2 is the cost of tool and regrind per tool

change ($/change), and t1 is the tool changing time (min)

Fig 4 Schematic of the Bendix ACO system for milling Source: Ref 3

The constants K1, K2, and K3 depend on tool and workpiece materials, and is an adjustable parameter in the range of 0

1 that determines the type of performance index, J When = 1, the J represents the reciprocal of the cost per unit produced; when = 0, the J represents the production rate; and for intermediate values of , the J represents a weighted

combination of these two objectives, which can be adjusted to represent the profit (that is, maximum production with minimum cost)

The objective of the ACO system, then, is to maximize this economic index of performance, J, subject to constraints on

maximum spindle speed, minimum spindle speed, maximum torque, maximum feed, maximum temperature, and maximum vibration amplitude This constrained optimization problem is solved on-line using gradient methods, with increments in feed of 0.003 mm/rev (0.0001 in./rev) and increments in spindle speed of 10 rev/min Although this ACO system was successfully demonstrated in the laboratory, it was not widely accepted by industry The main problem was

the need for a tool wear sensor (for tool wear rate, or TWR, in Eq 1) that can operate in an industrial environment The

calculation of the tool wear rate from the temperature and torque measurements as used in the Bendix system was not satisfactory for the entire range of feeds and speeds, and reliable temperature measurements were difficult to obtain in a production environment Accurate on-line tool wear estimation remains an important topic of current research, as discussed in the section "Trends in AC Systems" in this article

ACO System for Grinding. A successful ACO system for grinding is based on the measurement of the power at the

grinding wheel, as illustrated in Fig 5 This measured power, P, is compared to the maximum allowable power for burning, Pb, and the power error generated, e, is used to adjust the workpiece spindle speed, nw, and the radial infeed

velocity, vf The adjustment strategy uses a function Gb to calculate the allowable power for burning and the optimal locus

algorithm, based on a model of the grinding process mechanics, to maximize vf and calculate the corresponding value of

nw such that P = Pb Thus, the system is designed to operate at (or near) the power constraint for burning The sampling

period is denoted by T, and is 0.5 s A surface roughness constraint can also be included Experimental evaluation of the

system was performed in the laboratory, and excellent results were obtained, as shown in Fig 6 These results illustrate the successful operation of the system at the specified power level through ten cycles, despite the progression of grinding

wheel wear; this was achieved by increasing the radial infeed velocity, v This optimal locus approach may also be

applicable to the development of ACO systems for metal removal operations other than grinding, and as such is a topic of current research (Ref 8)

Fig 5 Schematic of an ACO system for grinding Source: Ref 7

Fig 6 Grinding of repetitive cycles with power limit constraint wheel, 32A46K8VBE; workpiece, SAE 4340 steel,

240 HB; wheel diameter, 440 mm (17.3 in.); workpiece diameter, 165 mm (6.50 in.); grinding width, 32 mm (1.26 in.); material removed from workpiece radius per cycle, 0.3 mm (0.012 in.); single-point dressing only before cycle 1 with a dressing depth of 10 m (400 in.), and a dressing lead of 5 m (200 in.) Source: Ref

7

Adaptive Control With Constraints

In general, commercial AC systems used in production today for rough milling and turning are of the ACC type This is because ACO systems are rather complex and will require further research before they can be fully implemented The ACC systems can provide much of the benefit of ACO systems and are relatively easy to implement In this section, a general discussion of ACC systems is followed by a description of a particular ACC system for a turning operation

Fundamentals of ACC Systems. Typically, ACC systems are based on the measurement of a single process variable, such as force, torque, or motor current, and try to maintain that variable at some predetermined reference value If this reference value is determined to ensure a relatively high production rate without excessive wear rates or breakage, it

provides good, although suboptimal, performance Most ACC systems attempt to maximize the metal removal rate, MRR,

by maximizing one of the machining variables, for example, operating at a maximum feed rate compatible with maintaining a constant load on the cutter, as was illustrated in Fig 2 for a slab milling operation In this case the average

feed rate with the ACC system is larger than the programmed feed rate, particularly when there are significant variations

in the depth and/or width of cut in the particular milling operation The most commonly used constraints in ACC systems are the cutting force, cutting torque, and the machining power The machining variables commonly manipulated to meet

these constraints are the feed rate (V) and the spindle speed (N) The machining feed is defined by the ratio:

(Eq 2)

where p = 1 in turning and drilling and is the number of teeth on the cutter in milling The main cutting force component

is typically modeled as being proportional to the depth of cut and the feed:

where KS is the specific cutting force coefficient, and the exponent, u, is typically in the range 0.6 < u < 1.0 Both KS and

u depend on workpiece and tool material properties and must be empirically determined by means of machining tests The

cutting torque T is proportional to the force times the workpiece radius in turning and to the force times the tool radius in milling and drilling The machining power P is proportional to the force times the spindle speed

An ACC system for turning is shown in Fig 7 and described in Ref 9 The main cutting force component is measured

and sampled (typically every 0.1 s) The sampled force value F is then compared to a desired, or reference, force value Fr The force error is provided as input to the ACC controller, which in turn produces a feed rate command signal A positive

error increases the feed rate and consequently increases the actual force F (see Eq 3) Some results of laboratory

experiments in which the depth of cut varies in a stepwise fashion are shown in Fig 8 To completely eliminate the force error, an integral control strategy has been used; that is, the feed rate command signal is proportional to the integral of the force error

Fig 7 An ACC system for turning Source: Ref 9

Fig 8 Turning experiments with an ACC system with a medium controller gain and step changes in depth of

cut The plots show the effect of changes in (a) depth of cut on (b) feed and (c) cutting force with increasing time Source: Ref 9

However, the ACC system can still have poor performance problems, and even tool breakage or controller instability (Ref

10, 11) The experimental results in Fig 9 illustrate the instability that occurs because of changes in the cutting process parameters (for example, depth of cut, spindle speed, and even feed rate) The cutting process, as shown in Fig 7, is part

of the control loop, and unless the controller gain is varied to compensate for the process parameter variations, consistent closed-loop performance cannot be obtained Thus, the process parameter variations require the use of techniques from adaptive control theory (Ref 1) In other words, the controller gains must be adapted on-line to the changing process parameters Such variable-gain ACC systems have been developed and are described in the section "Trends in AC Systems" in this article