The Machine Work Study was done by Taylor over a period of 26 years on metal cutting over the period of 1880-1906 and the results of the study as productivity science of metal cutting were presented in the 1906 conference of ASME. Taylor himself was the president of ASME and he gave the presentation as Presidential Address.
Results of 50,000 experiments
The following is a record of some of our more important steps: _
33 (A) In 1881, the discovery that a round-nosed tool could be run under given conditions at a much higher cutting speed and there- fore turn out much more work than the old-fashioned diamond- pointed tool.
34 (B) In 1881, the demonstration that, broadly speaking, the use of coarse feeds accompanied by their necessarily slow cutting speeds would do more work than fine feeds with their accompanying high speeds.
35 (C) In 1883, the discovery that a heavy stream of water poured directly upon the chip at the point where it is being removed from the steel forging by the tool,would permit an increase in cutting speed, and, therefore, in the amount of work done of from 30 to 40 per cent. In 1884, a new machine shop was built for the Midvale Steel Works, in the construction of which this discovery played a most important part; each machine being set in a wrought iron pan in which was collected the water (supersaturated with carbonate of soda to prevent rusting), which was thrown in a heavy stream upon the tool for the purpose of cooling it. The water from each of these pans was carried through suitable drain pipes beneath the floor to a central well from which it was pumped to an overhead tank from which a system of supply pipes led to each machine. Up to that time, so far as the writer knows, the use of water for cooling tools was confined to small cans or tanks from which only a minute stream was allowed to trickle upon the tool and the work, more for the purpose of obtaining a water finish on the work than with the object of cooling the tool; and, in fact, these small streams of water are utterly inadequate for the latter purpose. So far as the writer knows, in spite of the fact that the shops of the Mid- vale Steel Works until recently have been open to the public since 1884 no other shop in this country was similarly fitted up until that of the Bethlehem Steel Company in 1899, with the one exception of a small steel works which was an offshoot in personnel from the Midvale Steel Company.
36 (D) In 1883, the completion of a set of experiments with round nosed tools; first, with varying thicknesses of feed when the depth of the cut was maintained constant; and, second, with varying depths of cut while the feed remained constant, to determine the efiect of each of these elements on the cutting speed.
37 (E) In 1883, the demonstration of the fact that the longer a toolis called upon to work continuously under pressure of the shaving, the slower must be the cutting speed, and the exact determination of the effect of the duration of the cut upon the cutting speed.
38 (F) In 1883, the development of formula: which gave mathematical expression to the two broad laws above referred to. Fortunately these formulae were of the type capable of logarithmic expression and therefore suited to the gradual mathematical development extend- ing through a long period of years, which resulted in making our slide rules, and solved the whole problem in 1901.
39 (G) In 1883, the experimental determination of the pressure upon the tool required on steel tires to remove cuts of varying depths and thickness of shaving.
40 (H) In 1883, the starting of a set of experiments on belting described in a paper published in Transactions, Vol. 15 (1894).
41 (J) In 1883, the measurement of the power required to feed a round-nosed tool with varying depths of cut and thickness of shaving when cutting a steel tire. This experiment showed that EVERY MACHINE TOOL required as much pressure to feed it as to drive the cut. This was one of the most important discoveries made by us, and as a result all steel cutting machines purchased since that time by the Midvale Steel Company have been supplied with feeding power equal to their driving power and very greatly in excess of that used on standard machine tools.
42 (K) In 1884, the design of an automatic grinder for grinding tools in lots and the construction of a tool room for storing and issuing tools ready ground to the men.
43 (L) From 1885 to 1889, the making of aseries of practical tables for a number of machines in the shops of the Midvale Steel Company, by the aid of which it was possible to give definite tasks each day to the machinists who were running machines, and which resulted in a great increase in their output.
44 (M) In 1886, the demonstration that the thickness of the chip or layer of metal removed by the tool has a much greater effect upon the cutting speed than any other element, and the practical use of this knowledge in making and putting into everyday use in our shops a series of broad-nosed cutting tools which enabled us to run with a coarse feed at as high a speed as had been before attained with r‘ound- nosed tools when using a fine feed, thus substituting, for a considerable portion of the work, COARSE FEEDS AND 1-non srnnns for our old maxim of coansn FEEDS AND snow srnnns.
45 (N) In 1894 and 1895, the discovery that a greater proportional gain could be made in cutting soft metals through the use of tools made from self-hardening steels than in cutting hard metals,the gain made by the use of self-hardening tools over tempered tools in cutting soft cast iron being almost 90 per cent, whereas the gain in cutting hard steels or hard cast iron was only about 45 per cent. Up to this time, the use of Mushet and other self-hardening tools had been almost exclusively confined to cutting hard metals, a few tools made of Mushet steel being kept on hand in every shop for special use on hard cast- ings or forgings which could not be cut by the tempered tools. This experiment resulted in substituting self-hardening tools for tempered tools for all “ roughing work” throughout the machine shop.
46 (P) In 1894 and 1895, the discovery that in cutting wrought iron or steel a heavy stream of water thrown upon the shaving at the nose of the tool produced a gain in the cutting speed of SELF-HARDEN- mo TOOLS of about 33 per cent. Up to this time the makers of self- hardening steel had warned users never to use water on the tools.
47 (Q) From 1898 to 1900, the discovery and development of the Taylor-White process of _treating tools; namely, the discovery that tools made from chromium—tungsten steels when heated to the melting point would do from two to four times as much work as other tools. This is the discovery of modern high-speed tools.
48 (R) In 1899- 1902, the development of our slide rules, which are so simple that they enable an ordinary workman to make practical and rapid everyday use in the shop of all the laws and formulae deduced from our experiments.
49 (S) In 1906, the discovery that a heavy stream of water poured directly upon the chip at the point where it is being removed from CAST IRON by the tool would permit an increase in cutting speed, and therefore, in the amount of work done, of 16 per cent.
(T) In 1906, the discovery that by adding a small quantity of vanadium to tool steel to be used for making modern high speed chromium-tungsten tools heated to near the melting point, the red hardness and endurance of tools, as well as their cutting speeds, are materially improved.
51 We regard as of by far the greatest value our mathematical work on experimental data which has resulted in the development of the slide rules; i. e., the mathematical expression of the exact effect upon the cutting Speed of such elements as the shape of the cutting edge of the tool, the thickness of the shaving, the depth of the cut, the quality of the metal being cut and the duration of the cut, etc. This work enables us to fix a daily task with a definite time allowance for each workman who is running a machine tool, and to pay the men a bonus for rapid work.
52 The gain from these slide rules is far greater than that of all the other improvements combined, because it accomplishes the original Object, for which in 1880 the experiments were started; i. e., that of superseding “ rule of thumb” by scientific control.
53 By far the most difficult and illusive portion of this work has been the mathematical side: first, finding simple formula: which expressed with approximate accuracy the effect of each of the numer ous variables upon the cutting speed; and, second, finding a rapid method of using these formulae in the solution of the daily machine shop problems.
63 In the second portion of this paper will be given in detail a statement of the appliances, methods and principles which we believe to be necessary to use in order to obtain reliable results. For the pur- pose of a. more general discussion of the subject, however, it seems important to anticipate this portion of the paper by describing in detail the standard which we have finally adopted as a true criterion for determining the effect of each of the variables upon the cutting speed.
64 The efect of each variable upon the problem is best deter- mined by finding the exact rate of cutting speed (say, in feet per minute) which shall cause the tool to be completely ruined after having been run for 20 minutes under uniform conditions.
65 For example, if we wish to investigate the effect which a change in the thickness of the feed has upon the cutting speed,~it is necessary to make a number of tools which are in all respects uniform, as to the exact shape of their cutting edge, their clearance and lip angles, their chemical composition and their heat treatment. These tools must then be run one after another, each for a period of 20 minutes, throughout which time the cutting speed is maintained exactly uniform. Each tool should be run at a little faster cutting speed than its predecessor, until that cutting speed has been found which will cause the tool to be completely ruined at the end of 20 minutes (with an allowance of a minute or two each side of the 20-minute mark). In this way that cutting speed is found which corresponds to the particular thickness of shaving which is under investigation.
66 A change is then made in the thickness of the shaving, and another set of 20-minute runs is made, with a series of similar uniform tools, until the cutting speed corresponding to the new thickness of feed has been determined; and by continuing in this way all of the cutting speeds are found which correspond to the various changes of feed. In the meantime, every precaution must be taken to maintain uniform all the other elements or variables which affect the cutting speed, such as the depth of the cut and the quality of the metal being cut; and the rate of the cutting speed must be frequently tested during each 20-minute run to be sure that it is uniform.
67 The cutting speeds corresponding to varying feeds are then plotted as points upon a curve, and a mathematical expression is found which represents the law of the effect of feed upon cutting speed. We believe that this standard or method of procedure constitutes the very foundation of successful investigation in this art; and it is from this standpoint that we propose to criticise both our own experiments and those made by other investigators. For further discussion of our standard method of making experiments see Par. 137.
68 It was only after about 14 years’ work that we found that the best measure for the value of a tool lay in the exact cutting speed at which it was completely ruined at the end of 20 minutes. In the meantime, we had made one set of experiments after another as we successively found the errors due to our earlier standards, and realized and remedied the defects in our apparatus and methods; and we have now arrived at the interesting though rather humiliating con- clusion that with our present knowledge of methods and apparatus, it would be entirely practicable to obtain through four or five years of experimenting all of the information which we have spent 26 years in getting.
69 The following are some of the more important errors made by us:
70 We wasted much time by testing tools for a shorter cutting period than 20 minutes, and then having found that tools which were apparently uniform in all respects gave most erratic results (particularly in cutting steel) when run for a shorter period than 20 minutes; we erred in the other direction by running o.ur tools for periods of 30 or 40 minutes each, and in this way used up in each single experi- ment so much of the forging that it was impossible to make enough experiments in cutting metal of uniform quality to get conclusive results. We finally settled on a run of 20 minutes as being the best all-round criterion, and have seen no reason for modifying this conclusion up to date.
71 We next thought a proper criterion for judging the effect of a given element upon the cutting speed lay in determining the particular cutting speed which would just cause a tool to be slightly discolored below the cutting edge at the end of the 20 minutes. After wasting six months in experimenting with this as our standard, we found that it was not a true measure; and then adopted as a criterion a certain definite dulling or rubbing away of the cutting edge. Later it was found, however, that each thickness of feed had corresponding to it a certain degree of dullness or injury to the cutting edge at which point regrinding was necessary (the thicker the shaving the duller the tool should be before grinding); and a third series of experiments was made with this as a standard. While experimenting on light forgings a standard dullness of tool was used which was just sufficient to push the forging and tool apart and so slightly alter the diameter of the work.‘ All of these criterions were discarded, however, when in 1894 we finally bit upon the true standard, above described, of completely ruining the tool in 20 minutes.
72 As will be pointed out later in the paper, this standard demands both a very large and expensive machine to experiment with, and also large, heavy masses of metal to work upon, which is unfortunate; but we believe without apparatus and methods of this kind it is out of the question to accurately determine the laws which are sought. See paragraphs 210-263.
73 Experiments upon the art of cutting metals (at least those experiments which have been recorded) have been mainly undertaken by scientific men, mostly by professors. It is but natural that the scientific man should lean toward experiments which require the use of apparatus and that type of scientific observation which is beyond the scope of the ordinary mechanic, or even of engineers unless they have been especially trained in this kind of observation. It is perhaps for this reason more than any other that in this art several of those elements which are of the greatest importance have received no atten- tion from experimenters, while far less fruitful although more complicated elements, have been the subject of extended experiments.
74 As an illustration of this fact we would call attention to two of the most simple of all of the elements which have been left entirely untouched by all experimenters, namely: a the effect of cooling the tool through pouring a heavy stream of water upon it, which results in a gain of 40 per cent in cutting speed; b the effect of the contour or outline of the cutting edge of the tool upon the cutting speed, which when properly designed results in an equally large percentage of gain.
-------------------
The problem before us may be again briefly stated to consist cf a careful study of the effect which each of the twelve following variable elements has upon the selection of the cutting speed and feed and therefore on the cutting time.
a. The quality of the metal which is to be cut, i. e., its hardness or other qualities which affect the cutting speed;
b. The diameter of the work;
c The depth of the cut, or one-half of the amount by which the forging or casting is being reduced in diameter in turning;
d. The thickness of the shaving, or the thickness of the spiral strip or band of metal which is to be removed by the tool, measured while the metal retains its original density ; not the thickness of the actual shaving, the - metal of which has become partly disintegrated; e The elasticity of the work and of the tool;
f. The shape or contour of the cutting edge of the tool, together with its clearance and lip angles;
g. The chemical composition of the steel from which the tool is made, and the heat treatment of the tool ;
h. Whether a heavy stream of water, or other cooling medium, is used on the tool;
j. The duration of the cut, i. e., the time which a tool must last under pressure of the shaving without being reground; '
k. The pressure of the chip or shaving upon the tool;
l. The changes of speed and feed possible in the lathe; m The pulling and feeding power of the lathe at its various speeds.
137 Before entering upon the details of our experiments, it seems necessary to again particularly call attention to the fact that “standard cutting-speed” is the true criterion by which to measure the
https://babel.hathitrust.org/ cgi/ssd? id=mdp.39076000032131
About Carl Barth
https://www.naha.stolaf.edu/pubs/nas/volume13/vol13_7.htm
72 As will be pointed out later in the paper, this standard demands both a very large and expensive machine to experiment with, and also large, heavy masses of metal to work upon, which is unfortunate; but we believe without apparatus and methods of this kind it is out of the question to accurately determine the laws which are sought. See paragraphs 210-263.
73 Experiments upon the art of cutting metals (at least those experiments which have been recorded) have been mainly undertaken by scientific men, mostly by professors. It is but natural that the scientific man should lean toward experiments which require the use of apparatus and that type of scientific observation which is beyond the scope of the ordinary mechanic, or even of engineers unless they have been especially trained in this kind of observation. It is perhaps for this reason more than any other that in this art several of those elements which are of the greatest importance have received no atten- tion from experimenters, while far less fruitful although more complicated elements, have been the subject of extended experiments.
74 As an illustration of this fact we would call attention to two of the most simple of all of the elements which have been left entirely untouched by all experimenters, namely: a the effect of cooling the tool through pouring a heavy stream of water upon it, which results in a gain of 40 per cent in cutting speed; b the effect of the contour or outline of the cutting edge of the tool upon the cutting speed, which when properly designed results in an equally large percentage of gain.
-------------------
The problem before us may be again briefly stated to consist cf a careful study of the effect which each of the twelve following variable elements has upon the selection of the cutting speed and feed and therefore on the cutting time.
a. The quality of the metal which is to be cut, i. e., its hardness or other qualities which affect the cutting speed;
b. The diameter of the work;
c The depth of the cut, or one-half of the amount by which the forging or casting is being reduced in diameter in turning;
d. The thickness of the shaving, or the thickness of the spiral strip or band of metal which is to be removed by the tool, measured while the metal retains its original density ; not the thickness of the actual shaving, the - metal of which has become partly disintegrated; e The elasticity of the work and of the tool;
f. The shape or contour of the cutting edge of the tool, together with its clearance and lip angles;
g. The chemical composition of the steel from which the tool is made, and the heat treatment of the tool ;
h. Whether a heavy stream of water, or other cooling medium, is used on the tool;
j. The duration of the cut, i. e., the time which a tool must last under pressure of the shaving without being reground; '
k. The pressure of the chip or shaving upon the tool;
l. The changes of speed and feed possible in the lathe; m The pulling and feeding power of the lathe at its various speeds.
The ultimate object of all experiments in this field is to learn how to remove the metal from our forgings and castings in the quickest time, and that therefore the art of cutting metals may be briefly defined as the knowledge of how, with the limitations caused by some and the opportunities offered by others of the above twelve variable elements, in each case to remove the metal with the highest appropriate cutting speed.
137 Before entering upon the details of our experiments, it seems necessary to again particularly call attention to the fact that “standard cutting-speed” is the true criterion by which to measure the
To give another illustration of our practical use of this standard. If, for example, we wish to determine which make of tool steel is the best, we should proceed to make from each of the two kinds to be tested a set of from four to eight tools. Each tool should be forged from tool steel, say, 5- inch x 1§ inch and about 18 inches long, to exactly the same shape, and after giving the tools made from each type of steel the heat treatment appropriate to its chemical composition, they should all be ground with exactly the same shaped cutting edge and the same clearance and lip angles. One of the sets of eight tools should then be run, one tool after another, each for a period of 20 minutes, and each at a little faster cutting speed than its predecessor, until that cutting speed has been found which will cause the tool to be completely ruined‘ at the end of 20 minutes, with an allowance of a minute or two each side of the 20-minute mark.
Every precaution must be taken throughout these tests to maintain uniform all of the other elements or variables which affect the cutting speed, such as the depth of the cut and the quality of the metal being cut. The rate of the cutting speed must be frequently tested during each 20-minute run to be sure that it is uniform throughout.
Throughout this paper, “the speed at which tools” give out in 20 minutes, as described above, will be, for the sake of brevity, referred to as the “standard speed.” ~ 141 After having found the -“standard speed” of the first type of tools, and having verified it by ruining several more of the eight tools at the same speed, we should next determine in a similar manner the exact speed at which the other make of tools will be ruined in 20 minutes; and if, for instance, one of these sets of tools exactly ruins at a cutting speed of 55 feet, while the other make ruins at 50 feet per minute, these “standard speeds," 55 to 50, constitute by far the most important criterion from which to judge the relative economic value of the two steels for a machine shop.
https://babel.hathitrust.org/ cgi/ssd? id=mdp.39076000032131
About Carl Barth
https://www.naha.stolaf.edu/pubs/nas/volume13/vol13_7.htm
Ud. 3.4.2026
Pub. 28.5.2020
