Principles of interchangeable manufacturing; a treatise on the basic principles involved in successful interchangeable manufacturing practice, covering design, tolerances, drawings, manufacturing equipment, gaging and inspection

Front matter

i'

fS'K^-' -I :ti^ 'it

CORNELL

Preface

While many articles dealing with various phases of interchangeable manufacturing have appeared from time to time in the technical press, no complete and comprehensive treatise dealing with this subject as a whole has heretofore been available to those interested in interchangeable manufacturing in the machine building and metal working fields.

The development of interchangeable manufacturing is closely interwoven with many distinctly American manufacturing methods and processes. Every large American industry has contributed its share to the progress made in interchangeable manufacturing. Different plants working along independent lines have often achieved the same results by widely different methods. The author has attempted to define and emphasize the underlying basic principles, using specific methods only when necessary to illustrate the appUcation of these principles in actual manufacturing processes. He has gathered the information upon which this treatise is based from many manufacturing plants, both large and small, in this country and in Canada. He has seen every method discussed in successful operation, some in one plant, some in another — but not all in any one.

For more than ten years the author has been in constant touch with many of the detailed manufacturing problems that arise in the production of interchangeable mechanisms in large quantities. During the World War his work took him for four years into many manufacturing plants in connection with ordnance work, first for private corporations and later for the Ordnance Department. When engaged in this work it became apparent to him that the absence of common methods of interpretation of drawings, tolerances, and specifications, the lack of uniform gaging methods and misunderstanding of many of

VI PRKFACE

the factors of interchangeable manufacturing, presented an urgent need for a complete treatise on this subject.

In arranging the material available on the subject of interchangeable manufacturing, the author has first taken up the general principles involved in the industrial apphcation of this method of production, and has then devoted a separate chapter to the definition of the terms used, so that there will be no misunderstanding as to the meaning of the terms used later in the book. The influence of interchangeable manufacturing processes on machine design and the purposes of models are then dealt with, followed by a complete and minute discussion on the dimensioning of drawings intended for use in interchangeable manufacturing. This is followed by a discussion of the principal elements that govern mechanical production, the equipment required for interchangeable manufacturing (including machines, jigs and fixtures); the gaging equipment necessary; and the principles of inspection and testing. Special chapters are also devoted to the manufacture for selective assembly, and methods used in small quantity production on an interchangeable basis. An entire chapter deals with the service factor in interchangeable manufacturing, because in the final analysis no manufactured machine or device is ever purchased for itself alone, but is acquired for the service which it is supposed to render.

The Pratt & Whitney Co., Hartford, Conn., with whose cooperation this treatise is written, submits it to the public as a part of the company's contribution to the art of interchangeable manufacturing with the hope that it will assist manufacturers and mechanics to employ effectively the principles of interchangeable manufacturing and to reap the benefits that a rational application of these principles make possible. The author also wishes to acknowledge at this time the assistance that has been given him by many other manufacturing plants that he has visited. To name them all would mean a long list of prominent plants manufacturing machine tools, automobiles, tractors, ordnance, tj^ewriters, watches, phonographs, instruments, etc.

Eaele Buckingham

Chapter I PRINCIPLES OF INTERCHANGEABLE MANUFACTURING

Pages Economy — Extent of Interchangeability — Clearances — Tolerances — Component Drawings — Specifications — Gages for Checking Resiilts — Manufacturing Equipment — Production Problems — Inspection of Product 1-17

Chapter II TERMS USED IN INTERCHANGEABLE MANUFACTURING

Interchangeability — Selective Assembly — Function — Limit — Tolerance — Basic and Model Size — Maximum and Minimum Metal Size — Maximum and Minimum Clearance — • Interference — Operating, Functional, and Clearance Surfaces — Elementary and Composite Surfaces — Compound Tolerances — Register Points — Unit Assembly — Component and Operation Drawings 18-28

Chapter III MACHINE DESIGN IN INTERCHANGEABLE MANUFACTURING

Classes of Design — Simplifying Design — Choice of Materials — Clearances and Tolerances — Application of Interchangeable Principle — Advantages of Unit Assembly — Designing for Assembling and Service 29-39

Chapter IV Purpose Of Models

Manufacturing Model to Test Functioning — Experimental Model — Testing Tolerances — Model for Standard of Precision 40-45

vii

Vlll CONTENTS

Chapter V PRINCIPLES IN MAKING COMPONENT DRAWINGS

Pages

Functional Drawings — Manufacturing Drawings — ■ Laws of Dimensioning — Inspection Gage Requirements — Composite Surfaces — Compound Tolerances — Force Fits — Profile Surfaces — Dimensioning Holes — Location of Holes — Concentricity and Alignment — Gears 46-76

Chapter VI PRACTICE IN MAKING COMPONENT DRAWINGS

Maintaining Functional Requirements of a Mechanism — Basic Dimensioning — Maintaining a Common Locating Point — Length Dimensions from Common Locating Point — Drawing of Separate Parts — Compound Tolerances 77^104

Chapter VII ECONOMICAL PRODUCTION

Specifications — Functions and Requirements of Product — Manufacturing Data — Factory Cost — Direct Labor Cost — Machine Hour Rate — Product Overhead — Clerical and Accounting Work — Specific and General Information 105-120

Chapter VIII EQUIPMENT FOR INTERCHANGEABLE MANUFACTURING

Selection of Machine Tools — Designing Jigs and Fixtures — Cutting Tools — Locating Points — Chip Clearances — Checking and Testing Jigs and Fixtures — Maintaining Tolerances — Special Equipment for Machining Automobile Transmission Cases — Drilling Holes Simultaneously — Pneumatic Clamping Devices — Multiple-Tool Facing Bar — MilUng and Drilling Fixtures 121-176

Chapter IX GAGES IN INTERCHANGEABLE MANUFACTURING Classification According to Use — Accuracy — Working and Inspection Gages — Interchangeability between Parts made in

CONTENTS IX

Pages Different Shops — Snap, Ring, and Plug Gages — Contour or Profile Gages — Receiving Gages — Flush-pin Gages — Sliding Bar Gages — Depth and Length Gages — Hole Gages — Gaging Threads — Tolerances on Threaded Parts — Wing and Indicator Gages — Functional Gages — Gaging Gears — Master and Reference Gages 177-216

Chapter I Principles Of Interchangeable Manufacturing

Interchangeable manufacturing consists of machining the component parts of a given mechanism in the manufacturing departments within such limits that they may be assembled in the assembling department without fitting or further machining. Component parts may also be replaced or transferred from one mechanism to another without detriment to the functioning and without machining. The advantages of such a method of manufacture are self-evident, and need not be dwelt upon further. It is obvious that with proper equipment and control, the component parts of a mechanism can thus be manufactured in large quantities at a low direct labor cost.

Economy of Interchangeable Mantifacturing. In all private industrial enterprises ultimate economy is the controlling factor of any method of procedure. This does not necessarily mean that the methods adopted always are actually the most economical. Methods which will promote this economy are, however, the ideals toward which manufacturers are constantly striving. Now, a careful analysis will show that interchangeability does not always result in ultimate economy. In such cases the attempt to maintain it is a fault, not a virtue.

To make this point clear, consider the matter first from the standpoint of production alone. The equipment and preparation necessary to produce interchangeable parts are expensive. In making only a small number of special mechanisms, it would be gross extravagance to maintain any high degree of interchangeability. Viewed simply as a question of production, the problem of interchangeable parts is solved by establishing a balance between manufacturing and assembling costs, whether the quantity of production be great or small, whether the mechanism involved be a standard or a special product.

Ultimate economy, however, must include the factor of service. Suppose automobiles, typewriters, sewing machines, or sporting rifles are sold. Parts will wear out or be broken by accident. The maintenance of service stations, where extra parts are quickly available, tends to keep customers satisfied. Service stations will be least expensive if the product is truly interchangeable and the agent can replace a part with the aid of a screwdriver or wrench — or, still better, if the customer can replace it himself. Since the advent of the automobile, people have been much more interested in things mechanical than before, and have taken pride in making their own repairs. The more nearly interchangeable mechanisms are made, the more this desirable trait is fostered and the less will service stations cost. Ultimate economy, then, requires that service costs be balanced against total productive costs.

Degree of Interchangeability Desirable. It should not be assumed from this that entire interchangeabiHty or none at all must be had. In almost dvery mechanism certain parts are begun as separate units in order to simplify the manufacture, but are later permanently assembled into a single unit and machined to completion as such. In many such cases, the expense of attaining interchangeability would be too great to justify the attempt, because of the many mechanical difi&culties to be overcome. It would be more economical, in case of breakage, to discard and replace the entire assembled unit. In other cases, the functional requirements may be so severe that a system of selective assembly will prove to be the proper course, although this entails carrying a double or triple number of spare parts in service stations, or involves some fitting when replacing unserviceable parts.

In general, however, interchangeability is a desirable goal, and is readily attained in the majority of cases if the proper attention is given to the basic principles governing it, including the design of the mechanism and the process of manufacture; yet it is limited in several directions by the inadequacy of many present manufacturing conditions. With improved facilities, it may be that in future years a much greater degree of interchangeability will be possible than at present.

The following paragraphs, which are based on manufacturing conditions as they now exist, trace the progress of a commodity through all stages of its manufacture, from its inception as a mechanical project to the final testing that determines its successful completion. An attempt has been made to single out for special comment those factors which make possible, or promote, the interchangeable manufacture of its parts.

Design as it Affects Success. The development of any new mechanism starts with a mental conception of some function to be performed. This conception then takes detailed form, first mentally, then on paper, and finally in metal. The experimental model — if such be constructed — is usually made by the cut-and-try method. Little attention is paid in the beginning to future manufacturing requirements. The main object is to construct a mechanism that will function properly regardless of the exact design. When this end is reached, what may be called the inventive or functional design has demonstrated its success.

Before manufacturing is begun, however, a manufacturing design must be perfected which will modify the inventive design so as to allow its economical production on a large scale. Several manufacturers recognize this twofold nature of designing, and maintain a separate department for each type. Indispensable as is the original invention, it is the manufacturing design which largely determines the success or failure of a given project. This manufacturing designing necessarily continues throughout the whole course of production because of the almost infinite number of petty detailed questions involved, only a few of which can be foreseen and answered in advance. One of the important functions of an engineering department is to keep in close touch with the progress of the work in the shops, deduce general principles therefrom, and apply these principles not only to the work in hand, but also to all new work that may be developed.

The Manufacturing Model. Assume that the functional requirements of the mechanism are established and that the manufacturing design has been adopted. The first concern is to test this design as far as possible. The most certain method of accompHshing this is to develop a physical model. Such a model must not be confused with the experimental model, as its purpose is quite different. The experimental model shows that the mechanism will perform certain functions. The manufacturing or physical model, if properly developed, proves that the mechanism, as modified and developed to facilitate manufacture, still retains the functional advantages of the experimental model. The manufacturing model is naturally an expensive piece of equipment, but if a large output of a new commodity is under consideration, it is money well invested. In the case of a small total output, a "pilot" mechanism is often built for this purpose, which is not set aside for future reference but incorporated in the product itself.

There are many other services which a manufacturing model is capable of rendering. It may serve as a physical standard of dimensions for the future product. In this case, it must be made with much greater care than if it were to be used merely to test the functioning of the manufacturing design. Such a model will be of great value as a reference at all times during production. In itself, it comprises an effective functional gage to test any completed part. It should be used but rarely, however, for that purpose. In addition, the component parts of the model are of great assistance in checking the manufacturing equipment in the early stages of the work.

Clearances. Clearances are vital factors in interchangeable manufacturing. Fits can be secured without interchangeability, but the latter cannot be maintained without proper clearances. It is self-evident that a certain space must be left between operating parts. The minimum clearances should be as small as the assembling of the parts and their proper operation under service conditions will allow. The maximum clearances should be as great as the functioning of the mechanism permits. The variation between a maximum and a minimum clearance determines the manufacturing tolerance. It is clear, then, that determining at the outset the permissible clearances establishes also the extent of the tolerances which control the final inspection.

Clearances should be one of the principal considerations in developing the manufacturing design. This design should aim to allow the greatest possible amount of clearance between companion parts. The more the design lends itself to this end, the greater the economy of manufacture and the greater the degree of interchangeability obtainable. In determining which parts of a mechanism can be made interchangeable, this matter of permissible clearances plays the largest part. A mechanism which is so designed that it cannot permit fairly liberal clearances is not a suitable one to be manufactured on a strictly interchangeable basis with the standard equipment now available. Every operating part of a mechanism must be located within reasonably close clearances in each plane. After such requirements of location are met, all other surfaces should have Hberal clearances, unless the factor of strength is the controlling one.

Manufacturing Tolerances. The general tendency in the past has been to establish manufacturing tolerances by trying to hold the product as closely as possible to a fixed size. The natural result of this policy is that the tolerances established on paper are often exceeded; yet the actual working variations remain unrecorded, because it is argued that under certain conditions the original requirements might be met and, therefore, the tolerances noted are the proper ones, even though they are not maintained. Every effort to make the recorded tolerances represent the actual working tolerances is opposed on the ground that such a procedure would lower the shop standards. As a matter of fact, it is hard to understand how anything could lower the standards of the shop more than the absolute disregard of the rules it is supposed to be obeying.

There is a further argument for the acceptance of liberal tolerances. Too often in manufacturing concerns, and especially in the case of interchangeable manufacturing, one finds details being made ends in themselves rather than means to a larger end. In producing a component part, the main object should not be to demonstrate how closely a fixed size can be approached; the aim should be to construct, as economically as possible, a mechanism that will satisfactorily perform certain functions. The knowledge of how accurately a machining operation can be performed is indeed invaluable in making the manufacturing design; but when that design has once been completed, interest should shift to the proper functioning of the completed mechanism. Finally, it may be said that in most cases the tolerances originally fixed are increased during the process of manufacturing without detriment to the mechanism. It is rarely that a tolerance has to be reduced.

The proper minimum clearances can be determined quite readily and definitely for most cases in the early stages of the work — ■ the manufacturing model is of great value in this respect — but the maximum clearances become estabhshed only after extended experience with the particular mechanism. In many cases the extreme maximum is never found, because long before that point is reached, the tolerances have become so liberal that there is no need, from the standpoint of economical production, to increase them further.

Component Drawings. Component drawings have two main functions to perform. The first is to give such information about the design and the tolerances that the manufacture of the product can begin. This does not seem like a very difficult task, but the notation of the tolerances on component drawings has created new problems of interpretation that have not, as yet, been fully solved. At the present time, the language of drawings is not altogether clear and exact.

The first tendency in introducing tolerances on drawings seems to have been to attempt to express a permissible variation on every dimension given. The results obtained in the shop depend, then, upon the particular combination of dimensions used. Different organizations using different combinations could obtain radically different results; and of the possible number of different combinations there is no end.

The existence of a tolerance on a drawing is an acknowledgment that variations are inevitable in the physical dimensions of the product. Any dimension given on such a drawing without a tolerance should not be construed to denote an absolute size without error, but rather to indicate either that the permissible variation for that point or surface is controlled by tolerances given on other co-related dimensions, or that the dimension is so relatively unimportant that no attempt had been made to determine its permissible variation.

In making component drawings, the effort should be made to so give the dimensions and necessary tolerances that it would be possible to lay out one, and only one, representation of the "maximum metal" condition and one, and only one, of the "minimum metal" condition. If such lay-outs were superimposed, the difference between them would represent the permissible variation on every surface. Any condition of the product which fell within the zone thus established could be considered as meeting the requirements of the drawing. If one will make a few such lay-outs, it will soon be clear to him that there are always a number of dimensions that should be given without tolerances if drawings are to be kept consistent and intelligible.

Information on Component Drawings. It must be realized at the start that it is impossible in every case to give on one component drawing all the dimensions that are needed to construct the patterns, tools, gages, and other manufacturing equipment, without introducing many inconsistencies. Certain dimensions could be correct if one set of holding points and one series of operations were to be used, but would be incorrect under different conditions. If the component drawings are made so that they represent the proper completed conditions — call them inspection gage requirements if you wiU — the end in view is attained. Any figures that the shop desires to use are correct if they insure this result.

It is impossible to amplify this point without entering into a prolonged discussion of the effect of using different holding or registering points in the manufacturing processes. Yet it may be of interest to know that several manufacturing plants solve this problem by adding operation drawings, which give only the specific dimensions required at a particular operation. Some of the dimensions are duplicates of those on the component drawing, while others are computed to serve their restricted purpose. This proves an effective means of recording additional information required in the manufacturing departments, which cannot be put on component drawings without danger of misuse.

After production is well under way, the component drawings have served their first purpose. In the meantime, the actual manufacturing operations have made available a store of new information regarding the proper conditions to be maintained. It should be the second function of the component drawings to record as much of this information as possible. Conflicting information or misinformation should be eliminated at the same time; in short, the drawings should be revised to agree with actual conditions and requirements. It has been a great fault in the past to neglect this second function almost entirely. It is a difficult task to make the component drawings represent from the first conditions ,that must be maintained. In time, the shop will discover many of them, often after bitter experience, even though they have been omitted from the component drawings. Frequently, however, it happens that this information does not make its way back to the office, but is retained by the shop men among themselves. Often this is the fault of the office, which is prone to consider such information as criticism, so that the shop, after a few rebuffs, makes no further attempt to pass it along. It is most essential, however, that such information be recorded in permanent form, not only because of its value to the work in hand, but also because of its helpful application to new work in the future.

Dimensioning of Component Drawings. The problem of the proper dimensioning of component drawings is strictly a mathematical one. There are a few basic principles in regard to it as fixed and simple as Newton's three laws of motion, but even more difficult at times to apply correctly. When either of the two following principles is violated, trouble will inevitably follow;

1. In interchangeable manufacturing, there is but one dimension (or group of dimensions) in the same straight line that can be controlled within fixed tolerances. That is the distance between the cutting surface of the tool and the locating or registering surface of the part being machined. Hence, it is incorrect to locate any point or surface with tolerances from more than one point in the same straight line.

2. Dimensions should be given between those points which it is essential to hold in a specific relation to each other. The majority of dimensions, however, are relatively unimportant in this respect. It is good practice to establish locating points in each plane, and to give, as far as possible, all such dimensions from these common locating points.

There are also a few other general principles which it is good practice to follow. Although violations of them are not errors in themselves, they lead to many unnecessary errors. In all of this work it must be realized that it is impossible to create anything that is altogether fool-proof; the best that can be expected is to make conditions such that little or no excuse remains for making a mistake. The three following principles are of this order:

1. The basic dimensions given on component drawings for interchangeable parts should be the maximum metal sizes, except for force fits and other unusual conditions. The direct comparison of the basic sizes should check the "danger zone" or the minimum clearance conditions in most cases. It is evident that these sizes are the most important ones, as they control the interchangeability. They should be the first determined and, once established, they should remain fixed if the mechanism functions properly and the design is unchanged. The direction of the tolerances, then, would be such as would increase this clearance. For force fits, such as taper keys, etc., the basic dimensions should be those which determine the minimum interference (which is the "danger zone" in this case) and the direction of the tolerances for this class of work should be such as would increase this interference.

2. Dimensions should not be duplicated between the same lO INTERCHANGEABLE MANUFACTURING

points. The duplication of dimensions causes much needless trouble, due to changes being made in one place and not in the others. It causes less trouble to search a drawing to find dimensions than it does to have them duplicated and, though more readily found, inconsistent.

3. As far as possible, the dimensions on companion parts should be given from the same relative locations. This procedure assists in detecting interferences and other improper conditions.

If careful thought is given to these component drawings, much time and effort will be saved later in the shop. If they are neglected, all the future work will suffer. A large percentage of the mistakes made in the manufacturing departments may be traced back to improper component drawings.

Specifications for Interchangeable Manufacturing. The information that can be included on component drawings, except in the case of a very simple or familiar mechanism, is seldom sufficient in itself to enable the manufacturer to proceed intelligently with a new product. It is desirable that he know the particular purpose for which the mechanism is to be made. The better he is informed on this subject, the greater service he can render in promoting its economical manufacture and future development. Specifications are supposed to supplement the drawings by giving all the needed additional information which has no place on the drawings. I say "supposed" because it is only in rare cases that the specifications commonly encountered give all the desirable information. They usually deal with only the most exacting requirements and make no mention of the others, thus establishing a severe precedent for the solution of all questions in regard to the requirements, important and unimportant. They seldom indicate the essential object in view, namely, the economical production of mechanisms which will function satisfactorily.

Specifications should state the end to be accomplished, and should give all possible information to assist in the attaining of that end. Any unusual conditions should be explained in detail. All exacting requirements should be specified with the reasons

INTERCHANGEABLE MANUFACTURING II

for the same, including requirements of functioning and of materials to be used. But they should not stop here. The less exacting conditions shovdd be noted also. If a certain material is specified, and the chief consideration is economy, it should be so stated, with the substitution allowed. The material that might be most economical under one set of conditions might be otherwise under different circumstances. Parts which are detailed on the drawings but for which commercial articles can be substituted should be so designated. The specifications should hst those parts which must be made interchangeable and those which need not be. A description of the tests for materials, for physical dimensions, and for functioning should be included. In fact, any information that will assist in the manufacture of the product should be given. Some of it will specify the results to be obtained; more of it should be information to assist the manufacturer, not hard and fast rules which he must follow regardless of consequences.

Such specifications would undoubtedly be far from complete at first. Provision should be made to keep them abreast of the actual progress of the work. The shop should use them as a place to record as much of the experience gained as possible. If certain methods have been found unsatisfactory, here is an ideal place to record the fact, and perhaps save a duplication of the mistake in the future. If other methods have proved satisfactory, they, too, should by all means be recorded. In fact, specifications of this kind, although they would in time become voluminous, would be a history of a mechanism and furnish valuable data to assist in developing new mechanisms.

Written specifications are held in low esteem by the majority of manufacturers. They do without written specifications for their own products, and when obliged to meet them for contract work, find them an additional annoyance instead of a help. This is due, in large measure, to the fact that this subject, as also the matter of tolerances, has been regarded as an end in itself instead of as a means to a larger end. Manufacturers do have specifications, although they are seldom called by that name and are seldom written or grouped together for ready reference. Some of them may be found in the cost and production records, some in the shop correspondence, but most of them are carried in the memories of the foremen and older employes who maintain the traditions of the shop.

Gages for Checking Results. Thus far those elements which form the groundwork for actual manufacturing operations have been discussed. The manufacturing design has been developed; it has been tested with the manufacturing model; the first guess as to the proper manufacturing tolerances has been made; all suitable and available information has been recorded on the component drawings; and the specifications to supplement these drawings have been partially developed by recording there all further information available that will assist in accomplishing the main purpose, namely, to produce satisfactory mechanisms as economically as possible. The means of carrying on the work of actual production and the facilities that should be provided for checking the results, must now be considered.

There are two important reasons for inspecting the product during manufacture: First, spoiled parts must be eUminated as soon as possible to save the expenditure of useless effort on unserviceable pieces. Second, the completed components must be checked before assembly to eliminate the unserviceable parts and thus insure the proper functioning of the mechanism. For these purposes, gages are extensively employed.

A gage should be provided whenever its use is more economical than the use of standard measuring instruments. For example, if the total production of a certain mechanism amounts to about a dozen units, it is extravagance to provide special gages. On the other hand, if this production amounts to several thousand units, a complete set of gages is both desirable and necessary. The extent to which gages are necessary, therefore, depends in great measure upon the amount of the total production. Furthermore, gages should be provided to check only those conditions which it is essential to maintain. The nature and extent of the gages required depend upon the manufacturing conditions. In many cases, a check on one or two points is sufficient to detect any unsatisfactory results. Under varying manufacturing conditions different faults must be guarded against. Gages are a preventive and not a cure. The point to be emphasized is that they should be provided whenever their addition will result in the production of more or better components with a total expenditure of the same or less effort.