Modern Machine-Shop Practice Part 97

Fig. 1417, for example, represents the ordinary form, the points being rounded; hence, when the legs are closed the point of contact between the inside and outside calipers will be at A, while when they are opened out to their fullest the points of contact will be at B. This may, however, be remedied to a great extent by bevelling off the ends from the outside as shown in Fig. 1416.

[Ill.u.s.tration: Fig. 1418.]

The end faces of outside calipers should be curved in their widths, as in Fig. 1418, so that contact shall occur at the middle, and it will then be known just where to apply the points of the inside calipers when testing them with the outside ones.

Inside and outside calipers are capable of adjustment for very fine measurements; indeed, from some tests made by the Pratt and Whitney Company among their workmen it was found that the average good workman could take a measurement with them to within the twenty-five thousandth part of an inch. But the workman of the general machine shop who has no experience in measuring by thousandths has no idea of the accuracy with which he sets two calipers in his ordinary practice. The great difference that the one-thousandth of an inch makes in the fit of two pieces may be shown as in Fig. 1419, which represents a collar gauge of 5/8 inch in diameter, and a plug 1/1000 inch less in diameter, and it was found that with the plug inserted 1/8 inch in the collar it could be moved from A to B, a distance of about 5/16 inch, which an ordinary workman would at once recognise as a very loose fit.

If the joints of outside calipers are well made the calipers may upon small work be closed upon the work as in Fig. 1420, and the adjustment may be made without requiring to tap or lightly knock the caliper legs against the work as is usually done to set them. But to test the adjustment very finely the work should be held up to the light, as in Fig. 1421, the lower leg of the calipers rested against the little finger so as to steady it and prevent it from moving while the top leg is moved over the work, and at the same time moving it sideways to find when it is held directly across the work. For testing the inside and outside calipers together they should for small diameters be held as in Fig. 1422, the middle finger serving to steady one inside and one outside leg, while one leg only of either calipers is grasped in the fingers.

[Ill.u.s.tration: Fig. 1419.]

[Ill.u.s.tration: Fig. 1420.]

[Ill.u.s.tration: Fig. 1421.]

[Ill.u.s.tration: Fig. 1422.]

For larger dimensions, as six or eight inches, it is better, however, to hold the calipers as in Fig. 1423, the forefinger of the left hand serving to rest one leg of each pair on the contact being thus tested between the legs that are nearest to the operator.

The adjustment of caliper legs should be such that contact between the caliper points and the work is scarcely, if at all, perceptible. If with the closest of observation contact is plainly perceptible, the outside calipers will be set smaller than the work, while in the case of inside calipers, they would be set larger; and for this reason it follows that if a bore is to be measured to have a plug fitted to it, the inside calipers should have barely perceptible contact with the work bore, and the outside calipers should have the same degree of contact, or, if anything, a very minute degree of increased contact. On the other hand, if a bore is to be fitted to a cylindrical rod the outside calipers should be set to have the slightest possible contact with the rod, and the inside ones set to have as nearly as possible the same degree of contact with the outside ones, or, if anything, slightly less contact.

For if in any case the calipers have forcible contact with the work the caliper legs will spring open and will therefore be improperly set.

Calipers should be set both to the gauge and to the work in the same relative position. Let it be required, for example, to set a pair of inside calipers to a bore, and a pair of outside calipers to the inside ones, and to then apply the latter to the work. If the legs of the inside calipers stand vertical to the bore for setting they should stand vertical while the outside calipers are set to them, and if the outside calipers are held horizontally while set to the inside ones they should be applied horizontally to the work, so as to eliminate any error due to the caliper legs deflecting from their own weight.

[Ill.u.s.tration: Fig. 1423.]

To adjust calipers so finely that a piece of work may be turned by caliper measurement to just fit a hole; a working or a driving fit without trying the pieces together, is a refinement of measurement requiring considerable experience and skill, because, as will be readily understood from the remarks made when referring to gauge measurements, there are certain minute allowances to be made in the set of the calipers to obtain the desired degree of fit.

In using inside calipers upon flat surfaces it will be found that they can be adjusted finer by trusting to the ear than the eye. Suppose, for example, we are measuring between the jaws of a pillow-block. We hold one point of the calipers stationary, as before, and adjust the other point, so that, by moving it very rapidly, we can just detect a sc.r.a.ping sound, giving evidence of contact between the calipers and the work. If, then, we move the calipers slowly, we shall be unable, with the closest scrutiny, to detect any contact between the two.

Calipers possess one great advantage over more rigid and solid gauges, in that the calipers may be forced over the work when the degree of force necessary to pa.s.s them on indicates how much the work is too large, and therefore how much it requires reducing. Thus, suppose a cylindrical piece of work requires to be turned to fit a hole, and the inside calipers are set to the bore of the latter, then the outside calipers may be set to the inside ones and applied to the work, and when the work is reduced to within, say, 1/100 inch the calipers will spring open if pressed firmly to the work, and disclose to the workman that the work is reduced to nearly the required size. So accustomed do workmen become in estimating from this pressure of contact how nearly the work is reduced to the required diameter, that they are enabled to estimate, by forcing the calipers over the work, the depth of the cut required to be taken off the work, with great exact.i.tude, whereas with solid gauges, or even caliper gauges of solid proportions, this cannot be done, because they will not spring open.

The amount to which a pair of calipers will spring open without altering their set depends upon the shape: thus, with a given joint they will do so to a greater extent in proportion as the legs are slight, whereas with a given strength of leg they will do so more as the diameter of the joint is large and the fit of the joint is a tight one. But if the joint is so weak as to move too easily, or the legs are so weak as to spring too easily, the calipers will be apt in one case to s.h.i.+ft when applied to the work, and in the other to spring so easily that it will be difficult to tell by contact when the points just touch the work and yet are not sprung by the degree of contact. For these reasons the points of calipers should be made larger in diameter than they are usually made: thus, for a pair of calipers of the shape shown in Fig. 1410, the joint should be about 1-1/4 inches diameter to every 6 inches of length of leg. The joint should be sufficiently tight that the legs can just be moved when the two legs are taken in one hand and compressed under heavy hand pressure.

[Ill.u.s.tration: Fig. 1424.]

For measuring the distance of a slot or keyway from a surface, the form of calipers shown in Fig. 1424 is employed; the straight leg has its surface a true plane, and is held flat against the surface B of the slot or keyway, and the outside or curved leg is set to meet the distance of the work surface measuring the distance C. These are termed keyway calipers.

There are in general machine work four kinds of fit, as follow: The working or sliding fit; the driving fit; the hydraulic press fit; and the shrinkage fit. In the first of these a proper fit is obtained when the surfaces are in full contact, and the enveloped piece will move without undue friction or lost motion when the surfaces are oiled. In the second, third, and fourth, the enveloped piece is made larger than the enveloping piece, so that when the two pieces are put together they will be firmly locked.

It is obvious that in a working or sliding fit the enveloped piece must be smaller than that enveloping it, or one piece could not pa.s.s within the other. But the amount of difference, although too small to be of practical importance in pieces of an inch or two in diameter and but few inches in length, is appreciable in large work, as, say, of two or more feet in diameter. A journal, for example, of 1/10 inch diameter, running in a bearing having a bore of 1/1000 inch larger diameter, and being two diameters in length, would be instantly recognised as a bad fit; but a journal 6 inches in diameter and two diameters or 12 inches long would be a fair fit in a bearing having a bore of 6-1/1000 inches. In the one case the play would be equal to one one-hundredth of the shaft's diameter, while in the other case the play would equal but one six-thousandth part of the shaft's diameter. In small work the limit of wear is so small, and the length of the pieces so short, that the 1/1000 of an inch a.s.sumes an importance that does not exist in larger work.

Thus, in watch work, an error of 1/1000 inch in diameter may render the piece useless; in sewing machine work it may be the limit to which the tools are allowed to wear; while in a steams.h.i.+p or locomotive engine it may be of no practical importance whatever.

A journal 1/10 inch in diameter would require to run, under ordinary conditions, several years to become 1/1000 inch loose in its bearing.

Some of this looseness, and probably nearly one half of it, will occur from wear of the bearing bore; hence, if a new shaft of the original standard diameter be supplied the looseness will be reduced by one-half.

But a 6-inch journal and bearing would probably wear nearly 1/1000 inch loose in wearing down to a bearing which may take but a week or two, and for these reasons among others, standard gauges and measuring tools are less applicable to large than to small work.

The great majority of fits made under the standard gauge system consist of cylindrical pieces fitting into holes or bores. Suppose then that we have a plug and a collar gauge each of an inch diameter, and a reamer to fit the collar gauge, and we commence to ream holes and to turn plugs to fit the collar gauge, then as our work proceeds we shall find that as the reamer wears, the holes it makes will get smaller, and that as the collar gauge wears, its bore gets larger, and it is obvious that the work will not go together. The wear of the gauge obviously proceeds slowly, but the wear of the reamer begins from the very first hole that it reams, although it may perform considerable duty before its wear sensibly affects the size of the hole. Theoretically, however, its size decreases from the moment it commences to perform cutting duty until it has worn out, and the point at which the wearing-out process may have proceeded to its greatest permissible limit is determined by its reduction of size rather than by the loss of its sharpness or cutting capacity. Obviously then either the reamer must be so made that its size may be constantly adjusted to take up the wear, as in the adjustable reamer, or else if solid reamers are used there must be a certain limit fixed upon as the utmost permissible amount of wear, and the reamer must be made above the standard size to an amount equal to the amount of this limit, so that when the reamer has worn down it will still bore a hole large enough to admit the plug gauge. To maintain the standard there should be in this case two sets of gauges, one representing the correct standard and the other the size to which the reamer is to be made when new or restored to its proper size.

The limit allowed for reamer wear varies in practice from 1/1000 to 1/10000 of an inch, according to the requirements of the work. As regards the wear of the standard gauges used by the workmen they are obviously subject to appreciable wear, and must be returned at intervals to the tool room to be corrected from gauges used for no other purpose.

To test if a hole is within the determined limit of size a limit gauge may be used. Suppose, for example, that the limit is 1/1000 of an inch, then a plug gauge may be made that is 1/1000 of an inch taper, and if the large end of this plug will enter the hole, the latter is too large, while if the small end will not enter, the hole is too small.

When only a single set of plug and collar gauges are at hand the plug or the collar gauge may be kept to maintain the standard, the other being used to work to, both for inside and outside work. Suppose, for example, that a plug and collar gauge are used for a certain piece of work and that both are new, then the reamer may be made from either of them, because their sizes agree, but after they have become worn either one or the other must be accepted as the standard of size to make the reamer to. If it be the collar gauge, then the plug gauge is virtually discarded as a standard, except in that if the plug gauge be not used at all it may be kept as a standard of the size to which the collar gauge must be restored when it has worn sufficiently to render restoration to size necessary. If this system be adopted the size of the reamer will be constantly varying to suit the wear of the collar gauge, and the difficulty is encountered that the standard lathe arbors or mandrels will not fit the holes produced, and it follows that if standard mandrels are to be used the reamers must when worn be restored to a standard size irrespective of the wear of the gauges, and that the standard mandrels must be made to have as much taper in their lengths as the limit of wear that is allowed to the reamers. Suppose, for example, that it is determined to permit the reamer to wear the 1/2000 of an inch before restoring it to size, then in an inch mandrel the smallest end may be made an inch in diameter and the largest 1-1/2000 inch in diameter, so that however much the reamer may be worn within the limit allowed for wear the hole it produces will fit at some part in the length of the standard mandrel. But as the reamer wears smaller its size must be made as much above its designated standard size as the limit allowed for wear; hence, when new or when restored to size, the reamer would measure 1-1/2000 inches, and the hole it produced would fit the large end of the mandrel. But as the reamer wore, the hole would be reamed smaller and would not pa.s.s so far along the mandrel, until finally the limit of reamer wear being reached the work would fit the small end of the mandrel. The small end of the mandrel is thus the standard of its size, and the wear of the collar gauge is in the same direction as that of the reamer. Thus, so long as the collar gauge has not worn more than the 1/2000 of an inch it will, if placed upon the mandrel, fit it at some part of its length.

Now suppose that the plug gauge be accepted as the standard to which the reamer is to be made, and that to allow for reamer wear the reamer is made, say, 1/2000 inch larger than the plug gauge, the work being made to the collar gauge. Then with a new reamer and new or unworn gauges the hole will be reamed above the standard size to the 1/2000 inch allowed for reamer wear. As the reamer wears, the hole it produces will become smaller, and as the collar gauge wears, the work turned to it will be larger, and the effect will be that, to whatever extent the collar gauge wears, it will reduce the permissible amount of reamer wear, so that when the collar gauge had worn the 1/2000 inch the work would not go together unless the reamer was entirely new or unworn.

In a driving fit one piece is driven within the other by means of hammer blows, and it follows that one piece must be of larger diameter than the other, the amount of the difference depending largely upon the diameter and length of the work.

It is obvious, however, that the difference may be so great that with sufficiently forcible blows the enveloping piece may be burst open. When a number of pieces are to be made a driving fit, the two pieces may be made to fit correctly by trial and correction, and from these pieces gauges may be made so that subsequent pieces may be made correct by these gauges, thus avoiding the necessity to try them together.

In fitting the first two pieces by fit and trial, or rather by trial and correction, the workman is guided as to the correctness of the fit by the sound of the hammer blows, the rebound of the hammer, and the distance the piece moves at each blow. Thus the less the movement the more solid the blow sounds, and the greater the rebound of the hammer the tighter the fit, and from these elements the experienced workman is enabled to know how tightly the pieces may be driven together without danger of bursting the outer one.

What the actual difference in diameter between two pieces may require to be to make a driving fit is governed, as already said, to a great extent by the dimensions of the pieces, and also by the nature of the material and the amount of area in contact. Suppose, for example, that the plug is 6 inches long, and the amount of pressure required to force it within the collar will increase with the distance to which it is enveloped by the collar. Or suppose one plug to be 3 inches and another to be 6 inches in circ.u.mference, and each to have entered its collar to the depth of an inch, while the two inside or enveloped pieces are larger than the outside pieces by the same amount, the outside pieces being of equal strength in proportion to their plugs, so that all other elements are equal, and then it is self-evident that the largest plug will require twice as much power as the small one will to force it in another inch into the collar, because the area of contact is twice as great. It is usual, therefore, under definite conditions to find by experiment what allowance to make to obtain a driving or a forcing fit. Thus, Mr.

Coleman Sellers, at a meeting of the Car Builders a.s.sociation, referring to the proper amount of difference to be allowed between the diameters of car axles and wheel bores in order to obtain a proper forcing or hydraulic fit, said, "Several years ago some experiments were made to determine the difference which should be made between the size of the hole and that of the axle. The conclusion reached was that if the axle of standard size was turned 0.007 inch larger than the wheel was bored it would require a pressure of about 30 tons to press the axle into the wheel." The wheel seat on the axle here referred to was 4-7/8 inches in diameter and 7 inches long. It is to be remarked, however, that the wheel bore being of cast iron and the axle of wrought iron the friction between the surfaces was not the same as it would be were the two composed of the same metal. This brings us to a consideration of what difference in the forcing fit there will be in the case of different metals, the allowance for forcing being the same and the work being of the same dimensions.

Suppose, for example, that a wrought-iron plug of an inch in diameter is so fitted to a bore that when inserted therein to a distance of, say, 2 inches, it requires a pressure of 3 lbs. to cause it to enter farther, then how much pressure would it take if the bore was of cast iron, of yellow bra.s.s, or of steel, instead of wrought iron. This brings us to another consideration, inasmuch as the elasticity and the strength of the enveloping piece has great influence in determining how much to allow for a driving, forcing, or a shrinkage fit.

Obviously the allowance can be more if the enveloping piece be of wrought iron, copper, or bra.s.s, than for cast iron or steel, because of the greater elasticity of the former. Leaving the elasticity out of the question, it would appear a natural a.s.sumption that the pieces, being of the same dimensions, the amount of force necessary to force one piece within the other would increase in proportion as the equivalents of friction of the different metals increased.

This has an important bearing in practice, because the fit of pieces not made to standard gauge diameter is governed to a great extent by the pressure or power required to move the pieces. Thus, let a steel crosshead pin be required to be as tight a fit into the crosshead as is compatible with its extraction by hand, and its diameter in proportion to that of the bore into which it fits will not be the same if that bore be of wrought iron, as it would be were the bore of steel, because the coefficient of friction for cast steel on cast iron is not the same as that for steel on wrought iron. In other words, the lower the coefficient of friction on the two surfaces the less the power required to force one into the other, the gauge diameters being equal. In this connection it may be remarked that the amount of area in contact is of primary importance, because in ordinary practice the surfaces of work left as finished by the steel cutting tools are not sufficiently true and smooth to give a bearing over the full area of the surfaces.

This occurs for the following reasons. First, work to be bored must be held (by bolts, plates, chuck-jaws, or similar appliances) with sufficient force to withstand the pressure of the cut taken by the cutting tool, and this pressure exerts more or less influence to spring or deflect the work from its normal shape, so that a hole bored true while clamped will not be so true when released from the pressure of the holding clamps.

To obviate this as far as possible, expert workmen screw up the holding devices as tight as may be necessary for the heavy roughing cuts, and then slack them off before taking the finis.h.i.+ng cuts.

Secondly, under ordinary conditions of workshop practice, the steel cutting tools do not leave a surface that is a true plane in the direction of the length of the work, but leave a spiral projection of more or less prominence and of greater or less height, according to the width of that part of the cutting edge which lies parallel to the line of motion of the tool feed, taken in proportion to the rate of feed per revolution of the work.

Let the distance, Fig. 1424A, A to B lie in the plane of motion of the tool feed, and measure, say, 1/4 inch, the tool moving, say, 5/16 inch along the cut per lathe revolution. Suppose the edge from B to D to lie at a minute angle to the line of tool traverse, and the depth of the cut to be such that the part from B to C performs a slight cutting or sc.r.a.ping duty, then the part from B to C will leave a slight ridge on the work plainly discernible to the naked eye in what are termed the tool marks.

The obvious means of correcting this is to have the part A B of greater width than the tool will feed along the cut, during one revolution of the work (or the cutter, as the case may be); but there are practicable obstacles to this, especially when applied to wrought iron, steel, or bra.s.s, because the broader the cutting edge of a tool the more liable it is to spring, as well as to jar or chatter, leaving a surface showing minute depressions lying parallel to the line of tool feed.

If the cutting tool be made parallel and cylindrical on its edges, and clearance be given on the front end of its diameter only, so as to cut along a certain distance only of its cylindrical edge, the rest being a close fit to the bore of the work, the part having no cutting edge, that is, the part without clearance, will be apt to cause friction by rubbing the bore of the work as the tool edge wears, and the friction will cause heat, which will increase as the cut proceeds, causing the hole to expand as the cut proceeds, and to be taper when cooled to an equal degree all over. This may be partly obviated by giving the tool a slow rate of cutting speed, and a quick rate of feed, which will greatly reduce the friction and consequently the heating of the tool and the work. On cast iron it is possible to have a much broader cutting edge to the tool, without inducing the chattering referred to, than is the case with wrought iron, steel, or bra.s.s, especially when the finis.h.i.+ng cut is a very light one. If the finis.h.i.+ng cut be too deep, the surface of the work, if of cast iron, will be pitted with numerous minute holes, which occur because the metal breaks out from the strain placed on it (and due to the cut) just before it meets the cutting edge of the tool.

Especially is this the case if the tool be dull or be ground at an insufficiently acute angle.

When the work shows the tool marks very plainly, or if of cast iron shows the pitting referred to (instead of having a smooth and somewhat glossy appearance), there will be less of its surface in contact with the surface to which it fits, and the fit will soon become destroyed, because the wearing surface or the gripping surface, as the case may be, will the sooner become impaired, causing looseness of the fit. In the one case the abrasion which should be distributed over the whole area of the fitting parts is at first confined to the projections having contact, which, therefore, soon wear away. In the other case the projecting area in contact compresses, causing looseness of the fit.

Hydraulic press or forcing fits.--For securing pieces together by forcing one within the other by means of an hydraulic press, the plug piece is made a certain amount larger than the bore it is to enter, this amount being termed the allowance for forcing. What this allowance should be under any given conditions for a given metal, will depend upon the truth and smoothness of the surfaces, and on this account no universal rule obtains in general practice. From some experiments made by William Sellers & Co., it was determined that if a wheel seat (on an axle) measuring 4-7/8 inches in diameter and 7 inches long was turned 7/1000 of an inch larger than the wheel bore, it would require a pressure of about thirty tons to force the wheel home on the axle.

At the Susquehanna shops of the Erie railroad the measurements are determined by judgment, the operatives using ordinary calipers. If an axle 3-1/2 diameter and 6 inches long requires less than 25 tons it is rejected, and if more than 35 tons it is corrected by reducing the axle.

In order to insure a proper fit of pieces to be a driven or forced fit it is sometimes the practice to make them taper, and there is a difference of opinion among practical mechanics as to whether taper or parallel fits are the best. Upon this point it may be remarked that it is much easier to measure the parts when they are parallel than when they are taper, and it is easier to make them parallel than taper.

On the elevated railroads in New York city, the wheel bores being 4-1/8 inches in diameter and 5 inches long, the measurements are taken by ordinary calipers, the workmen judging how much to allow, and the rule is to reject wheels requiring less than about 26 tons, or more than about 35 tons, to force them on. These wheels form excellent examples, because of the excessive duty to which they are subjected by reason of the frequency of their stoppage under the pressure of the vacuum brake.

The practice with these wheels is to bore them parallel, finis.h.i.+ng with a feed of 1/4 inch per lathe revolution, and to turn the axle seats taper just discernible by calipers.

This may, at first sight, seem strange, but examination makes it reasonable and plain. Let a wheel having a parallel bore be forced upon a parallel axle, and then forced off again, and the bore of the wheel will be found taper to an appreciable amount, but increasing in proportion as the surface of the hole varied from a dead smoothness; in other words, varying with the depth of the tool marks in the bore and the smoothness of the cut.