Next must he established the allowable de viations from the model. These are controlled mainly by two factors: first, the proper func tioning of the whole,—thus the largest shaft must always assemble into the smallest bearing; second, how small a variation can be main tained in manufacture. This done, means of testing the accordance of the parts with these standards must be furnished; namely, gauges representing the largest and smallest permis sible size for each important dimension. An inspection organization should be carefully trained to use these, and throw out any piece not conforming to them, to save further labor on it or mischief in assembling, and prevent more of its sort being made. Such a service can retrench much useless expense; in fact, if in any large plant it does not save' far more than its cost in a year, something is very wrong.
But further uniformities than size, which gauges and drawings concern solely, are called for. Physical characteristics must conform also. This is effected by specifications dealing itemwise with strength, hardness and finish, composition of materials, and vital points to be watched most closely in course of making; and prescribing both frequency of tests (enough to establish sure accuracy) and manner of manufacture, also in detail any special treat ment contingently called for, as heat treatment of steel.
Equipment to obtain results simply and in expensively is a business necessity. It must be carefully suited to the special requirements of the product, as a type of appliance very satisfactory in one class of manufacturing may be quite unfit for another.
These steps are indispensable for inter changeability. The model should remain fixed till a new or improved design is devised. The tolerances are subject to some experimental change, but should be put • on a fixed basis at the earliest feasible moment; and never altered to suit inaccuracies in manufacturing methods or other temporary convenience, as the ultimate loss outweighs the saving. The specifications of course change with new methods or materials. All these are so essen tially interwoven that slighting any one makes the rest unsure, and every change should be carefully considered as to affecting the whole system. A well-balanced whole will give far more efficient results than one with some ele ments much in advance of the others.
The Model.— This affords for all dimen sions a direct comparison against a physical standard, the most uniform method available with the usual measuring equipment. The whole object will be to reproduce it in quantity as closely as possible, and such reproduction should be carefully kept in mind in developing it. To aid in this, no operations on it should be performed by hand tools (as a file), because such cuts are hardest and costliest to re produce. All cuts should be machining cuts.
But its development consists of far more than merely producing a piece of work as a standard for other production. Every cut taken should be governed, where possible, by some profile plate, size block, angle plate, form tool or master plate that will be available later for construction of the manufacturing equip ment. This will eliminate much of the neces sity for several independent settings to repro duce the. same sizes or shapes, ensure much greater uniformity throughout, and in the long run prove a real economy. Every step must be considered as to how it affects not only the model itself, but the reproduction of these same sizes, profiles and locations in the work ing tools and equipment.
To make the model, as often done, an in dependent unit merely for comparison of meas urements, is a wasteful economy far over balanced by increased expense of making the necessary tools. The more intricate the model. the greater the economy in liberal use of master plates, templets and all other useful special equipment. It should be very carefully kept, not handled or hit, and used only in cases of question and for verifying tools and gauges as needed. It must not be confused with the ex perimental models; the work is assumed to be past the experimental stage. This ultimate standard is in a class by itself, and must be done by skilled craftsmen of the highest type. At present the manufacturers of firearms are the leaders in this branch.
The Tolerances, or Allowed Variations. —Several factors are involved in establishing these. First, they must never be so great as to injure the operation of the mechanism; yet since expense of manufacturing increases very fast with severity of requirements, economy demands as large a tolerance as consistent with proper functioning. Second, those on one piece must not overlap the sizes fixed for com panion parts. The extreme sizes must be compared with each other; so that, for example, the largest pin will always enter the smallest hole, the smallest never be too loose in the largest hole. Third, it must always be kept in mind that the only dimension on any part controllable within close limits is the distance between where it is supported and the surface being machined by the cutting tool. Thus the method of holding the work determines the correct locations of the dimensions; or, vice versa, those locations determine the holding points for machining. Fourth, the accuracy of the machining facilities available determine the actual possibilities of closeness, and hence should be carefully considered in the original design. The actual extent of tolerances will be a matter of experiment or experience. The same care should he exercised, and the same factors considered, in altering a tolerance as in establishing it.