Productivity Engineering by F.W. Taylor

 

Taylor developed productivity science. As an engineering, he  did engineering to implement the discoveries of productivity science. Industrial engineering is primarily engineering to enhance productivity of the engineering processes to give output of goods and services. Industrial engineers need good amount of engineering knowledge and have to use in specific direction to increase productivity of engineering resources used in engineering systems and prevent waste of engineering resources and also output due to defects.

Read the elaborate engineering of the system for circulating cutting fluid (water) by Taylor to get productivity increase of 40% by increase in cutting speed. Industrial engineers have to be educated and trained to do engineering for implementing productivity ideas.

Excerpts for the Art of Metal Cutting by Taylor

SYSTEM FOR CIRCULATING THE COOLING STREAM OF WATER 

593 Cooling the nose of a tool by throwing a heavy stream of water or other fluid directly upon the chip at the point where it is being removed by the tool from the steel forging enables the operator to increase his cutting speed about 40 per cent.‘ The economy realized through this simple expedient is so large that it is a matter of the greatest surprise that experimenters on the art of cutting metals have entirely overlooked this source of gain. So far as the writer is aware, no experiments upon this subject have as yet been published. In spite of the fact that (as a result of our experiments) the whole machine shop of the Midvale Steel Company was especially designed as long ago as 1893 for the use Of a heavy stream of water (super-saturated with soda to prevent rusting) upon each cutting tool, until very recently (1906) practically no other shops in this country have been similarly equipped.

598 (B) With high speed tools a gain of 40 per cent‘ can be made in cutting steel or wrought iron by throwing in the most advantageous manner a heavy stream of water upon the tool.

599 In designing slide rules or tables, etc., for assigning daily tasks to machinists a 33 per cent increase in cutting steel or wrought iron should be allowed for instead of 40 per cent, owing to the fact that workmen are more or less careless in directing the stream of water to the proper spot upon the tool.

600 (C) A heavy stream of water (3 gallons per minute) for a 2-inch by 2-inch tool and a smaller quantity as the tool grows smaller, should be thrown directly upon the chip at the point where it is being removed from the forging by the tool. Water thrown upon any other part of the tool or the forging is much less efficient. 

601 (D) The gain in cutting speed through the use of water on the tool is practically the same for all qualities of steel from the softest to the hardest. 

602 (E) The percentage of gain in cutting speed through the use of water on the tool is practically the same whether thin or thick chips are being removed by the tool. 

603 (F) With modern high speed tools a gain of 16 per cent can  be made by throwing a heavy stream of water on the chip in cutting CAST iron. 

604 (G) To get the proper economy from the use of water in cooling the tool, the machine shop should be especially designed and the machine tools especially set with a view to the proper and convenient use of water. 

605 (H) 1n cutting steel, the better the quality of tool steel, the greater the percentage of gain through the use of a heavy stream of water thrown directly upon the chip at the point where it is being removed from the forging by the tool. The gain for the different types of tools in cutting steel is: a Modern high speed tools 40 per cent; b Old style self-hardening tools 33 per cent; c Carbon tempered tools 25 per cent.

606 This fact, stated in different form is that: The hotter the nose of the tool becomes through the friction of the chip, the greater is the percentage of gain through the use of water on the tool. 

THE PORTION OF THE TOOL ON WHICH THE WATER JET SHOULD BE THROWN

607 A series of experiments has demonstrated that water thrown directly upon the chip at the point where it is being removed from the forging by the tool will give higher cutting speeds than if used in any other way.

611 The most satisfactory results are obtained from a stream of water falling at rather slow velocity, but with large volume, at the proper point upon the tool; since a stream of this sort covers a larger area of the tool and is much freer from splash.

612 This water supply should be delivered through pipes fitted up with universal friction joints, so that the apparatus can be quickly adjusted to deliver the water at any desired point (the pipe being supported by a rigid bracket attached to the saddle of the lathe, preferably on the back side so as to be out of the way). In the case of short lathe beds the water supply can be delivered from overhead through a rubber hose, and in the case of long lathe beds through telescoping pipes attached to the saddle (smooth drawn "brass pipes telescoping inside of ordinary wrought iron pipes, with suitable stuffing boxes being used).

613 About three gallons of water per minute are required for adequately cooling a very large roughing tool, say, 2 inches by 2 inches section; and proportionally smaller quantities as the tool grows smaller.

614 For economy, the same water should be used over and over again, and it should be supersaturated with soda to prevent the machines from rusting. Wrought iron pipes about 1 inches diameter should lead the water from beneath the machine below the floor to the main soda water drains at the side of the shop. These drains are made of pipe from 3 to 5 inches in diameter, with a chain extending through them from one end to the other, the chain being twice as long as the drain through which it extends. In case of sediment forming in this pipe or in case of chips passing by the double sets of screens and double settling pots which should be supplied at each machine, the drain can be quickly cleaned by pulling the chain back- ward and forward through it once or twice.

615 The soda water is returned through this system of underground piping to a large central underground tank, from which it is pumped through a small, positive, continuously running pump, driven by the main line of shafting, into an overhead tank with overflow which keeps the overhead soda water supply mains continually filled and under a uniform head. If the shop is constructed with a concrete floor, a catch basin for the water can be molded in the concrete directly beneath each machine. Otherwise, each machine should be set in a large wrought iron pan or shallow receptacle which catches the Soda water and the chips. In both cases, however, two successive settling pots—independently screened so as to prevent the chips, as far as possible, from getting into the return main—are required beneath each machine. _

616 The ends of the 1} inch wrought iron pipes which lead the water from the machines to a large drain at the side of the shop should be curved up with a sweeping curve so that their outer ends come Close to the top of the floor of the shop. The sediment and chips must be cleaned from these pipes from time to time by means of a long round steel rod up to 1 inch in diameter, which, after removing the plug at the outer end of the drain pipes, is shoved through the pipe. Apparatus of this type has been in successful use for about 23 years with no trouble from clogging.

624 Taylor-White tools similar to those described above, having been carefully standardized, showed a cutting speed when run without water of 60 feet per minute, and when run under a heavy stream of water of 83 feet, thus indicating a gain of 1.39 to 1.0 from the use of water. The average gain then through the use of water on the tool for hard and soft forgings is about 40 per cent.

625 In 1906 after the writing of this paper was well under way it occurred to us that we had accepted as true without verification through accurate experiments the fact that water could not be used in cutting cast iron. This is another of the many instances in which an absolutely erroneous opinion prevails throughout all of our machine shops without any foundation in fact. It is likely, however, that this opinion has become so firmly rooted in the minds of all mechanics and foremen from the fact that a water finish cannot be made on cast iron, while it is in many cases most desirable for steel.

626 To determine the effect of a heavy stream of water in cooling a tool cutting cast iron, experiments (similar to those described above) were made by us during the summer and fall of 1906 on a test piece consisting of exceedingly hard cast iron with cutting tools of three different chemical compositions. 

 Gain through use of water  was 16 per cent.

Productivity Engineering of Belting - F.W. Taylor 1893

Industrial engineering is redesign of engineering products and processes to increase productivity and reduce costs. The call for cost reduction of engineering products and products made through engineering processes was made by the first President of ASME in 1880. Various persons presented papers. Taylor presented his first paper in 1893 and explained how the cost of belt systems used for power transmission can be reduced by keeping records for cost accounting and measurement and taking engineering decisions based on the analysis of data. Many commentators, present in the seminar remarked that the paper was the first systematic effort to collect cost and do engineering decision making based on cost analysis to reduce cost of engineering activity, task or process.

Taylor contributed to development of productivity thought and cost reduction thought in engineering by clearly conceptualizing three important activities of productivity improvement in engineering products and processes.

"Notes on Belting" is the first paper presented by F.W. Taylor on Productivity Engineering. Taylor's commitment to productivity science can be seen in this first paper.  The Paper on Piece Rates presented in 1895 contains both productivity engineering and productivity management aspects.

Taylor wrote in the "Piece Rate" paper that to increase productivity, systematizing that is systematically studying and improving all of the small details in the running of each shop, such as the care of belting, the proper shape for cutting tools, and the dressing, grinding, and issuing tool, oiling machines, issuing orders for work, obtaining accurate labor and material returns, and a host of other minor methods and processes has to be done. Then only on the basis of productivity improvement estimates, piece rates that provide motivation or incentive to operators to participate in the high productivity redesigned process can be given. 

Incentives are not increasing the productivity. Productivity science and  engineering improve productivity. Incentives are part of productivity management, where by operators are recruited and trained to work in high productivity processes.

"Notes on Belting" 

Presented at the New York Meeting (December, 1893) of the American Society of Mechanical Engineers, and forming part of Volume XV. of the Transactions.

You can access it from https://archive.org/stream/transactionsof15amer/transactionsof15amer_djvu.txt

pp. 204-259.

125 paragraphs are there in the paper.

The purpose of the paper was to present conclusions to be used in design and use of belting so as to obtain the greatest economy and the most satisfactory results.

It is important to understand carefully the terms "the greatest economy and the most satisfactory results." Taylor always took care to state multiple objectives involved in decision making. The output of a process has to give the most satisfactory results. The customer satisfaction and stakeholder satisfaction especially that of operators involved in the process were always emphasized by Taylor. The greatest economy is to be obtained by assuring first the satisfaction of results. Effectiveness has to be first designed and then only a redesign can be attempted to find lower cost alternative materials, design alternatives and process alternatives.

Important points are extracted from the paper and are given below. Understanding this paper is a foundation to using engineering knowledge to determine the cost collection criteria, cost analysis and redesign of engineering elements to reduce cost.

In using belting so as to obtain the greatest economy and the most satisfactory results, the following rules should be. observed : 

2. The chief consideration has to be the maximum of work from belting cost. Two most important considerations to realize it are securing the minimum of interruptions to manufacture by increasing the durability of the belt to the maximum. This criterion has not hitherto received due attention in belt system design. The one consideration which should have more weight than all others in making up  rules for the use and care of belting is "how to secure the least possible interruption to manufacture from the breakdowns of belts."

3. It is the writer's judgment that belts should be made heavier and run more slowly than theory and accepted rules would indicate, not only for the sake of reducing the belt bill in the long run, but even more to avoid the frequent interruptions to manufacture. In figuring the total expense of belting, and the manufacturing cost chargeable to this account, I think that most careful observers soon come to the conclusion that by far the largest item in this account is the time lost on the machines while belts are being replaced and repaired. This is certainly the case even where the process of manufacture is such that any one machine can be stopped without affecting the running of its neighbors, but far more so in those establishments where the running of a series of machines is dependent one upon another, and the stoppage of one machine involves delays on others. 

4. While working as foreman of a machine shop, the writer became convinced that the belts, which were laced according to the ordinary rules, were a great source of loss to the company — not so much from the cost of the belting and the labor of lacing as from the incidental delays to the machines, and the diminished output of the shop resulting therefrom. The belting was then shown to be by far the largest source of trouble in the shop.

5. But of equal importance in formulating rules for belting is the knowledge of what tension can be surely maintained through a term of  months, or what elements chiefly affect the durability of belting; yet these considerations appear to have been rather neglected by experimenters. Very little information could be obtained either as to the cost of maintenance of belts, or in regard to the interruptions to manufacture from belting, when used under known and uniform conditions as to tension and general treatment.

8. As a result of experience in the old shop, the tight and loose pulleys on the countershafts were made much larger in diameter and of wider face, so that the belt power from main line to countershafting was made about two and one-half times as great as formerly. All belts were made endless by splicing, glueing, and pegging, instead of lacing or hooking, and double belts were used throughout the shop. 

9. In all cases the countershafts were mounted on independent frames, which could be raised and lowered in tightening the belts by the interposition of wooden packing pieces of varying thickness between the frames and the supporting stringers overhead. For this purpose standard packing pieces, varying by eighths of an inch in thickness, were always kept in the tool room. With this method of tightening it was seldom necessary to resplice a belt, since six to ten inches of stretch could be taken up in the belt, by gradually raising the countershaft, before resplicing became necessary. 

10. Belt clamps were used having spring balances between the two pairs of clamps, so that the exact tension to which the belt was subjected was accurately weighed when the belt was first put on, and each time it was tightened. 

11. Experience soon demonstrated about the length of time that each belt would run without requiring to be tightened, and at approximately regular periods the spring-balance belt clamps were put on to each belt and the tension of same weighed, and the countershaft raised just enough to maintain the belt at its proper tension. For this reason, it was a matter of very rare occurrence that a belt slipped during working hours. And as the belts were generally tightened on Sundays (the shop working night and day), the minimum of delay was caused on the machines from this source.

14. At intervals of about three months for the first two years of the test, and after this time at intervals of about five months, each belt was scraped clean, and greased with the kind of dubbing recommended by the maker of the belt. 

15. An accurate account was kept of the original cost of each belt, and every item of expenditure, both for labor and materials used in the maintenance and care of same ; also the exact stretch of each belt was recorded, and its method of treatment throughout (*both engineering details and cost are recorded). 

18. In considering the results of the experiments made,  it should be borne in mind that the belts called "shifting" are those running from the main line of shafting to tight and loose pulleys on the countershafts, and that these belts were used so as to have about two and one half times as great transmitting power as the ordinary belting rules would demand; while the "cone" belts, extending from the countershaft to the machine, are used according to the ordinary rules for belting. 

63. In considering the above table, the most interesting and important fact noticeable is the superiority of the shifting to the cone belts in every respect except first cost, and this superiority is even much greater than the figures would indicate, since, generally speaking, the cone belts which are still in use are nearly worn out, having reached a point at which it is doubtful whether it pays to repair them, while the shifting belts are, to all appearances, in almost as good condition as when they first went into use, and should last twice as long as they have already. 

I think it would be safe to say that the life of the shifting belts will be three times that of the cone, and already the total cost of the shifting belts per year of service is less than that of the cone.  

66. It is interesting to note that after 8.8 years of life the total cost of maintenance and repairs of the shifting belts  amounts to only 30.4% of the original cost, while with the cone belts the maintenance and repairs through a life of 6.7 years amounts to one and one-half times the first cost.

67. In the writer's judgment, by far the greatest point of advantage of the shifting belts lies in the fact that the interruptions to manufacture go down drastically.  Each shifting belt required tightening or  repair  on an average only 6 times during nine years, while the cone belts averaged 32 interruptions to manufacture in 0.7 years. The shifting belts ran on an average twenty-two months without tightening, while the cone belts ran only two and one-half months. 

73. Summarizing the above, we may state that the total life of belting, cost of maintenance and repairs, and the interruptions to manufacture caused by belts, are dependent upon

 (1) the "total load" to which they are subjected, more than upon any other condition; and that, in our judgment, the other conditions chiefly affecting the durability of belting are: 

(2) Whether the belts are spliced, or fastened with lacing or belt hooks. 

(3) Whether they are properly greased and kept clean and free from machinery oil. 

(4) The speed at which they are run. 

77. Based on the evidence, the most economical total load for belting must lie between 174 lbs. and 357 lbs. per square inch of section of belt.

For several years past the writer has used the following rules with satisfaction, and he believes them to represent the most economical practice: 

80. The average total load on belting should be 200 to 225 lbs. per square inch section of belt. 

81. Six- and seven-ply rubber belts, and all double leather belts except oak tanned and fulled, will transmit economically a pull of 30 lbs. per inch of width to the rim of the pulley. 

82. Oak tanned and fulled double leather belts will transmit economically a pull of 35 lbs. per inch of width. 

83. The most economical speed for belting is 4,000 to 4,500 feet per minute. 

103. If the principle is correct, of using thick belts on account of their lateral stiffness and consequent durability, it becomes of the utmost importance to determine the minimum diameter of pulley which can be used with a given thickness of belt, and still have the belt last well. The writer is quite sure that double leather belts 2 inch thick will last well and give excellent satisfaction on pulleys as small as 12 inches in diameter, as he has had many belts in use for years under these conditions. 

For some time past he has had a triple leather belt 12 inches wide, 0.56 inch thick, running about 4,5(>0 feet per minute, with an idler pulley pressing lightly upon it, and transmitting about 100 H.P. to a pulley 12 inches diameter. This belt has up to date given excellent satisfaction, and has already lasted much longer than the two double leather belts which preceded it. 

The writer feels certain, from his experience, that it is safe and advisable to use — 

A double belt on a pulley 12 inches diameter or larger, 

A triple belt on a pulley 20 inches diameter or larger, 

A quadruple belt on a pulley 30 inches diameter or larger ; 

and it his opinion that it is advisable to use double, triple, and quadruple belts on pulleys respectively as small as 9 inches, 15 inches, and 24 inches diameter.

104. Regarding the question of fastening the two ends of the belt together, I think it safe to say that the life of belting will be doubled by splicing and cementing the belt, instead of lacing, wiring, or using hooks of any kind. When belts are subjected to the most severe usage, the spliced portion should be riveted, iron burrs being preferable to copper.

109. The best location for the idler pulley on high-speed belts is on the slack side of the belt, and about one-quarter way from the driving pulley. In this position it wears the belt far less than if placed close to the driven pulley, as is customary ; and the tendency of the idler to guide the belt off the pulley, in case it is slightly misplaced or the belt stretches unevenly, is far less. The writer is aware that this is contrary to the accepted theories on the subject, and has only arrived at this conclusion after repeated trials.

112. The faces of pulleys should, where practicable, be made about one-quarter wider than the belts which run on them, to allow for possible uneven stretch or running of belt, and a certain amount of chasing. 

120. Belts should be cleaned and greased every five or six months, just enough grease being put on to keep the surface of the belt moist and prevent it from cracking. It was found in the above experiment  that every three mouths was oftener than belts required greasing. 

123. Serious repairs to belting, as well as to all other machinery in a mill, should be prevented as far as possible by systematic and careful inspection at regular intervals, and the writer has found a tickler, having a portfolio for every day in the year, from which reminders to inspect and examine are issued daily, an invaluable aid in caring for the machinery of an establishment. With this method, a belt should rarely slip or give out while in use, and most repairs can be made out of working hours.

124. Much time is saved by having all of the repairs and adjustments to belting made by one or two men. A day laborer can soon be taught to repair belting after working hours, and do it much more thoroughly and systematically than if it is attended to by the high-priced men who run the machines during working hours. 

125. In figuring the probable running expenses of an establishment, it is frequently desirable to know about what the yearly belt bill will average. This issue was discussed in the paper.

Industrial engineers have to observe number of modification done by Taylor and the observation of the benefits of those changes. Industrial engineers have to make engineering changes and they must have the engineering knowledge to make those changes. They have to develop engineering expertise in the operations being done in their plant, at minimum operations and processes under their service and continuously increase knowledge of developments in the field.

Also, Taylor paid attention to maintenance. He even mentioned a planning device to send instructions at the appropriate time to maintenance persons. He also indicated that maintenance can be even taken up by part-time persons who come after the production shift is over and do it. Thus we can the areas of productivity science, productivity engineering and productivity management in the project carried out by Taylor in the area of belt systems in 1893. Taylor carried the idea of life of belt and the interruptions in production that a smaller life of belt will created into machining wherein he developed the equation for optimal tool life.

The core of industrial engineering is improvement of engineering elements for cost reduction or productivity improvement.

DISCUSSION.
Henry B. Towne 

The paper is remarkable as covering the record of an unusually large series of experiments, inaugurated on an intelligent and exceptionally comprehensive plan, and subsequently consistently carried out during the extraordinary period of nine consecutive years, under conditions not of the laboratory but of actual practice. 

Perhaps the most salient fact, and the most important conclusion of Mr. Taylor's argument, relates to the value of increased thickness of belts, and to the larger and more general use of double belts.

Having established this point, the paper then presents rules, and the experience on which they are based, governing the conditions under which thick belting can most efficiently and economically be used. These rules pertain less to the theory than to the practice of belting, and cover the questions of speed, diameter of pulleys, modes of tightening, distance of pulleys apart, kind and frequency of dressing, methods of lacing, and the efficiency of different kinds of belting. Previous investigations have dealt rather with the theory of belting than with the question of economy in its application and use. One of the most valuable features of Mr. Taylor's work consists in determining the conditions of application and use of belting which conduce to the lowest ultimate cost (productivity science). In other words, his investigation carries the subject through the field of mechanics into that of economics, and reduces the equation finally to a commercial form. However much the theoretical questions involved may interest the student and engineer, the commercial facts are those which chiefly interest and concern the mill manager and owner. In this, as in similar cases, the work of investigation is not completed until it has been carried to a point which includes both mechanical and commercial factors. 

Mr. Taylor has wisely refrained from further investigation of it, and has directed his observations and reasoning to the other elements involved, and for the purpose of arriving at rules for obtaining the best practical and commercial results. The records show that the coefficient of friction is itself variable, and depending upon the kind of belt, the condition of its surface, and other variable factors. Hence it follows that, in most cases, it is practically unnecessary to consider the coefficient of friction at all, the other more important factors which determine the proper conditions of use, especially those conducive to the best economy, and which determine the transmitting efficiency, being such always as to preclude any slipping of the belt.

Mr. Taylor properly gives much consideration to the question of economy in time of men and machines by using belting in such way as to secure the maximum freedom from interruptions to manufacture due to this source. In this, as in other details, his rules aim to secure results which are not only the best mechanically, but also commercially. The evidence submitted sustains the assertion that to obtain the highest economy in belting it is necessary to limit both the initial tension and the total load to a point much below that fixed by former rules, and which Mr. Taylor has sought to deduce from his observations. To accomplish this reduction he resorts to the use of double belting, adduces facts tending to show that this can be used on  pulleys as small as twelve inches in diameter, and shows that, for most uses, double belting is not only as available as single, but in many respects better, more desirable, and more economical. 

He sums up his conclusions by the statement that the most important factor in determining the life of belting, and its cost for maintenance and repairs, is the total load to which it is subjected, meaning thereby the pull per square inch of cross section. Other things being equal, therefore, the life and economy of the belt will, up to a certain limit, vary directly with its thickness. The conclusions on this point are that the point of best economy of total load lies between 174 lbs. and 357 lbs. per square inch of cross section. Mr. Taylor states that he has adopted a limit from 200 to 225 lbs. If his other conclusions and rules are adopted, further observation and experiment may well be addressed to a closer determination of this factor, in order to either verify the correctness of Mr. Taylor's conclusions, or else to indicate a new and better value for adoption in practice.