This work is titled in similar terms as The Art of Machining by F.W. Taylor and examines productivity
Taylor, in this country, and Nicolson, in England, have done classical work in the field of lathe-tool action.
This paper describes an attempt to investigate the fundamental principles underlying the action of milling.
It is shown in the present investigation that metal is removed more efficiently with thick chips than with thin chips. It follows from this that, other conditions being equal, including speed and feed per minute, the cutter with the fewest teeth gives the greatest efficiency. However, it is evident that the efficiencies of two cutters with different numbers of teeth are equal provided the table feeds be adjusted so that the same feed per tooth is effected. This gives a definite working theory on the influence of spacing.
6 In addition to the present investigation, experiments published in substantiation of the advantage of wide spacing agree in confirming this theory. It is definitely established that for a given material, tooth shape and sharpness, thickness of chip is the sole criterion of the efficiency with which metal is removed in milling and that increase of spacing over that required for free cutting is a handicap.' Present-day high-powered cutters have several times the chip space needed. Limitation of machine power has doubtlessly been the chief factor in giving a false bias to the influence of spacing.
7 The work described in this paper started in May, 1920, to find a rational basis for the action of a milling cutter so far as this could be removed from the region of empiricism.
9 The chip taken by a plain slab mill starts infinitely thin and its thickness gradually increases to a maximum just before the finish, from which point it quickly decreases. This decrease is practically instantaneous in an unexaggerated chip. The question arises, how does the force vary throughout the cutting of this chip?
11 The first problem is to find how the tangential force varies in relation to chip thickness. For measuring this force while the chip is being formed under reasonable commercial velocity conditions, a dynamical measuring method is the only solution, as a chip is completely cut in about one tenth of a second.
16 The analysis and computation of smoked-paper records is given fully in Appendix No. 1. An example of a chip study is given in Fig. 3', where it is very clearly shown that the force employed is not proportional to the chip thickness. In other words, material is removed more efficiently as the chip becomes thicker. This is confirmed later by experiments on varying feeds and furnishes the true explanation of the supposed advantages of coarse-tooth cutters.
17 Over twelve hundred chips were cut and the energy computed. These were taken at three different cutting speeds, viz., 17.5, 32 and 44 ft. per min., respectively. The average of the energy required at different speeds was remarkably equal. If the energy used at lowest speed be taken as 100, the results are: Speed in ft. per min . 17.5, 32, 44 Energy . . . 100, 101.8, and 97.4 As there appears no law in this and the variation is slight, presumably due to experimental error, the conclusion was reached that speed does not influence energy required.
Influence of Rake and Feed on Cast Iron. The results of the analysis of sixteen sets of chip formations in cast iron are given in Table 1 and shown graphically in Figs. 4 and 5. Each set consists of three chips formed at the different speeds as stated in Par. 17.
The area under each curve gives a numerical representation of the average production of each tool over a certain region of chip weight. Taking the 0-deg. rake tool as 100 we have: Production of 0-deg. rake tool = 100 Production of 10—deg. rake tool = 133 Production of 20-deg. rake tool = 138.4 Production of 30-deg. rake tool = 142.1
Influence of Rake and Feed on Bronze. The results of the analysis of sixteen sets of chip formations in bronze are given in the following table. As in the last section the consistent superior economy of heavy feeds is observed.
This information is condensed as before and results in: I Production of 0-deg. rake tool 100 Production of 10-deg. rake tool = 113.2 Production of 20-deg. rake tool = 123 Production of 30-deg. rake tool = 118
INFLUENCE OF RAKE AND FEED IN CUTTING MACHINE STEEL
The consistent superiority of heavy feeds is again noted. Again condensing the above information: Production of 0-deg. rake tool = 100 Production of 10-deg. rake tool = 118.7 Production of 20-deg. rake tool = 157.2 Production of 30-deg. rake tool = 172
Influence of Rake and Feed on Carbon Tool Steel. The results of the analysis of sixteen sets of chip formations in tool steel, are given in Table 3, and graphically in Figs. 8 and 9. Again condensing the above information: Production of 0-deg. rake tool = 100, Production of 10-deg. rake tool = 106, Production of 20-deg. rake tool = 112.2, Production of 30-deg. rake tool = 112.0.
Effect of Lubrication. Various experiments were run with lubricant. The results must be taken qualitatively rather than quantitatively. The action of lubrication in milling appears to be different from its action in lathe work. In lathe action, lubrication proper of the cutting tool is probably non-existent because of the unlikelihood of the lubricant finding its way to the point of the tool during cutting. The result in the lubricant being useful in lathe work only in proportion to its heat-carrying capacity and to its freedom from properties injurious to the mechanical condition of the machine.
In milling there is a decided influence entirely apart from heat-conducting properties. If the surface exposed by the last chip was very slightly smeared with an oiled finger (the oil film was scarcely discernible), the energy consumed in taking the next chip was from 10 per cent to 25 per cent lower in nearly all cases.
On plotting batches of results, however, the surprising situation was found of oil proving advantageous with low rakes in bronze and detrimental with high rakes. The existence of some unknown varying factor was suspected and the result would have been rejected and forgotten but for the fact, that the same effect showed up in experiments with machine steel made at a different time.
The authors hesitate to accept the foregoing unreservedly until further investigation sheds some light on its rationality.
Effect of Chatter. When chatter occurs, violent fluctuation of energy results. The energy consumed was usually greater than with smooth cutting. The authors believe that, due to chatter alone increase of energy invariably results and that, where a decrease? This can be brought about by the chatter causing an increase of force normal to the work and this in turn springing the work or cutter.
29 Effect of Clearance. Clearance was found not to affect energy consumed. Chatter, wear of tool and liability of tool to snip are affected by clearance.
30 Method of Determining Machineability. The word “machineability” is suggested to indicate the ease with which a given material can be machined. This of course presupposes standardized conditions of tool and size and shape of chip. If all conditions be standardized except the quality of the material being cut, then the machineability can be expressed in foot-pounds of energy required to remove one cubic inch of the material under investigation.
31 It is well known that machineability is a continual bone of contention in machining departments and that no accepted satisfactory way has yet been devised for its measurement. It is generally admitted that even though chemical analysis, tensile strength, elongation, Brinell hardness number and scleroscope hardness number are known, yet the machineability is an unknown factor. In fact, the War Department gave to the American Research Council as their major topic the development of some definite means of determining machineability.
33 To summarize: The authors had, at this state of development, no definite idea leading to a rational structure which would connect up the results. They regard the work in this section as useful in opening up territory, thereby enabling them to formulate more definite layouts for subsequent work, and believe that broadly the following conclusions are justified:
(a) As a chip thickens, metal is removed more easily per unit volume
(b) As rake angle is increased, metal is removed more easily per unit volume. Also this improvement is continuous but not uniform
(c) The advantage of rake depends on the material and is more pronounced with ductile materials
(d) The effect of lubrication proper (apart from cooling action) is decided, but the information available is inadequate to formulate it.
34 Numerous experiments on power consumption have been made by different investigators, but the method employed has been to measure the electrical input required to drive the machine. This confuses the unknown action of the cutter with the unknown machine loss and has therefore been rejected for the present investigation.
35 Features to be Measured: It was thought desirable to measure both the vertical and the horizontal components of the force exerted by the cutter. This necessitated developing an instrument that would measure both force and location, i.e., in all, three quantities: magnitude, direction and location. Or, which amounts to the same thing, the two components and the location may be measured.
39 Influence of Rake on Machine‘ Steel. The results of fourteen experiments are shown in Table 4 and graphically in Fig. 12. The noticeable consistency in the spacing of the pairs of values given as results of the zero-rake cutter is due to the second three having been taken in a different part of the block from the first three. The results plotted in Fig. 12 show the decided advantage of rake. The advantage of heavy chips is also clearly shown except in the case of the 20-deg. tool.
It is important to note that Carl Barth and Gilbreth made comments on this paper.
Gilbreth said his methods could be used to photograph the cutting action of the milling machine.
Two Interesting Papers in ASME Transactions of 1923, Vol. 45
https://ia804503.us.archive.org/19/items/sim_american-society-of-mechanical-engineers-transactions_1923_45/sim_american-society-of-mechanical-engineers-transactions_1923_45.pdf
Prof N.N. Sawin's Paper - A brief review
Boston and Kraus presented their work regarding milling in two parts in 1932. They evaluated the effects of cutter width, variations in feed and depth of cut, tool dulling and cutting fluids on milling energy, surface finish and tool life. They discussed the surface finish obtained by both the up and down milling processes. It was determined that the shinier surface produced by up milling was probably due to a burnishing action of the cutter at the start of each cut. They observed that although not as shiny, the surface produced by down milling was flatter and freer from ridges, but that the surface often had a dull torn appearance. An investigation of cutter dulling was made and it was determined that the energy required to power the cutter increased rapidly and uniformly as the cutter became dull.
In their second part, Boston and Kraus further discussed the difference between up and down milling and their respective finishes. The adherence of chips to the cutting edge in relation to the surface finish obtained was discussed. They were the first investigators to present work done regarding the presence of a built-up edge.
Boston, 0. W. and Kraus, Charles E., "The Elements of Milling," Transactions of the ASME, Vol. 54, 1932, p. RP-71.
Boston, 0. W. and Kraus, Charles E., "The Elements of Milling, Part 2," Transactions of the ASME, Vol. 56, 1934, pp. 355-371.
