Modern Machine-Shop Practice Part 98

Let the length of the wheel bore be 7 inches long, and the amount allowed for forcing be .004 inch, and one end of the wheel bore will have been forced (by the time it is home on the axle) over the length of 7 inches of the axle-seat, whose diameter was .004 larger than the bore: a condensation, abrasion, or smoothing of the metal must have ensued.

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

Now the other end of the same bore, when it takes its bearing on the shaft, is just iron, and iron without having suffered any condensation.

If the tool marks be deep, those on one end will be smoothed down while those at the other remain practically intact. Clearly then, for a parallel hole, a shaft having as much taper as the wheel bore will get in being forced over the shaft best meets the requirements; or, for a parallel shaft or seat, and a taper hole (the taper being proportioned as before), the small end of the taper hole should be first entered on the shaft, and then when home both the axle and the wheel-bore will be parallel.

It may be remarked that the wheel seat on the axle will also be affected, which is quite true, but the axle is usually of the hardest metal and has the smoothest surface, hence it suffers but little; not an amount of any practical importance.

In an experiment upon this point made in the presence of the author by Mr. Howard Fry and the master mechanic of the Renovo shops of the Philadelphia and Erie railroad, an axle seat finished by a Whitney "doctor," and parallel in diameter, was forced into a wheel having a parallel bore, and removed immediately. On again measuring the axle, the wheel-seat was found to be 1/1000 taper in its length.

The wheel-bore was found to be but slightly affected in its diameter, which is explained because it being very smooth, while the turning marks in the axle were plainly visible, the abrasion fell mainly upon the latter.

When the enveloping piece or bore is not solid or continuous, but is open on one side, the degree of the fit may be judged from the amount that it opens under the pressure of the plug piece.

[Ill.u.s.tration: Fig. 1425.]

Thus the axle bra.s.ses of American locomotives are often made circular at the back, as shown in Fig. 1425, and are forced in endways by hydraulic pressure. The degree of tightness of the bra.s.s within the box may, of course, be determined by the amount of pressure it requires to force it in, but another method is to mark a centre punch dot as at J, and before the bra.s.s is put in mark from this dot as a centre an arc of a circle as L L. When the bra.s.s is home in the box a second arc K is marked, the distance between L and K showing how much the bra.s.s has sprung the box open widening at H. In an axle box whose bore is about 4 inches to 5 inches in diameter, and 6 inches long, 1/32 inch is the allowance usually made.

Shrinkage fits are employed when a hole or bore requires to be very firmly and permanently fastened to a cylindrical piece as a shaft. The bore is turned of smaller diameter than its shaft, and the amount of difference is termed the allowance for shrinkage. The enveloping piece is heated so as to expand its bore; the shaft is then inserted and the cooling of the bore causes it to close or contract upon the shaft with an amount of force varying of course with the amount allowed for contraction. If this allowance is excessive, sufficient strain will be generated to burst the enveloping piece asunder, while if the allowance for shrinking is insufficient the enveloping piece may become loose.

The amount of allowance for shrinkage varies with the diameter thickness, and kind of the material; but more may be allowed for wrought iron, bra.s.s, and copper, than for cast iron or steel.

Again, the smoothness and truth of the surfaces is an important element, because the measurement of a bore will naturally be taken at the tops of the tool marks, and these will compress under the shrinkage strain, hence less allowance for contraction is required in proportion as the bore is smoother.

In ordinary workshop practice, therefore, no special rule for the amount of allowance for shrinkage obtains, the amount for a desultory piece of work generally being left to the judgment of the workman, while in cases where such work is often performed on particular pieces, the amount of allowance is governed by experience, increasing it if the pieces are found in time to become loose, and decreasing it if it is found impossible to get the parts together without making the enveloping piece too hot, or if it is found to be liable to split from the strain.

The strength of the enveloping piece is again an element to be considered in determining the amount to be allowed for shrinkage. It is obvious, for example, that a ring of 8 inches thick, and having a bore of, say, 6 inches diameter, would be less liable to crack from the strain due to an allowance of 1/50 inch for contraction, than would a ring of equal bore and one inch thick having the same allowance. The strength or resistance to compression of the piece enveloped in proportion to that enveloping it, is yet another consideration.

The tires for railway wheels are usually contracted on, and Herr Krupp states the allowance for contraction to be for steel tires 1/100 inch for every foot of diameter; in American practice, however, a greater amount is often employed. Thus upon the Erie railroad a 5 foot tire is given 1/16 inch contraction. The allowance for wrought iron or bra.s.s should be slightly more than it is for steel or cast iron, on account of the greater elasticity of those metals.

Examples of the practice at the Renovo shops of the Pennsylvania road are as follows:

Cla.s.s E, diameter of wheel centre, 44 inches; bore of steel tire, 43-15/16 inches.

Cla.s.s D, diameter of wheel, 50 inches; bore of tire, 49-9/16 inches.

It is found that the shrinkage of the tire springs or distorts the wheel centre, hence the tires are always shrunk on before the crank-pin holes are bored.

Much of the work formerly shrunk on is now forced on by an hydraulic press. But in many cases the work cannot be taken to an hydraulic press, and shrinkage becomes the best means. Thus, a new crank pin may be required to be shrunk in while the crank is on the engine shaft, the method of procedure being as follows: In heating the crank, it is necessary to heat it as equally as possible all round the bore, and not to heat it above a _very dark_ red. In heating it some dirt will necessarily get into the hole, and this is best cleaned out with a piece of emery paper, wrapped round a half-round file, carefully blowing out the hole after using the emery paper. Waste or rag, whether oiled or not, is not proper to clean the hole with, as the fibres may burn and lodge in the hole; indeed, nothing is so good as emery paper.

It is desirable to heat the crank as little as will serve the purpose, and it is usual to heat it enough to allow the pin to push home by hand.

It is better, however, to overheat the crank than to underheat it, providing that the heat in no case exceeds a barely perceptible red heat. If, however, the crank once grips the pin before it is home, in a few seconds the pin will be held so fast that no sledge hammer will move it. It is well, therefore, to have a man stationed on each side of the crank, each with a sledge hammer, and to push the crank pin in with a slam, giving the man in front orders to strike it as quickly as possible at a given signal; but if the pin does not move home so rapidly at each blow as to make it appear certain that it will go home, the man at the rear, who should have a ten-pound sledge, should be signalled to drive out the crank pin as quickly as he possibly can for every second is of consequence. All this should be done so quickly that the pin has not had time to get heated to say 100 at the part within the crank.

So soon as the pin is home, a large piece of wetted cotton waste should be wrapped round its journal, and a stream of water kept running on it, to keep the crank pin cold. At the other end water should be poured on the pin end in a fine stream, but in neither case should the water run on the crank more than can be avoided. Of course, if the crank is off the shaft, the pin may be turned downward, and let project into water.

The reasons for cooling the pin and not the crank are as follows: If the crank be of cast iron, sudden cooling it would be liable to cause it to split or crack. If the crank pin is allowed to cool of itself, the pin will get as hot as the crank itself, and in so doing will expand, placing a strain on the crank that will to some extent stretch it.

Indeed, when the pin has become equally hot with the crank it is as tight a fit as it will ever be, because after that point both pieces will cool together, and shrink or contract together, and hence the fit will be a looser or less tight one to the amount that the pin expanded in heating up to an equal temperature with the crank.

The correct process of shrinking is to keep the plug piece as cold as possible, while the outside is cooled as rapidly as can be without danger of cracking or splitting.

The ends of crank pins are often riveted after being shrunk in, in which case it is best to recess the end, which makes the riveting easier, and causes the water poured upon its face to be thrown outward, thus keeping it from running down the crank face and causing the crank to crack or split.

It sometimes becomes necessary and difficult to take out a piece that has been shrunk in, and in this event, as also in the case of a piece that has become locked before getting fully home in the shrinking process, there is no alternative but to reheat the enveloping piece while keeping the enveloped piece as cold as can be by an application of water.

The whole aim in this case is to heat the enveloping piece as quickly as possible, so that there shall be but little time for its heat to be transmitted to the piece enveloped. To accomplish this end melted metal, as cast iron, is probably the most efficient agent; indeed it has been found to answer when all other means failed.

[Ill.u.s.tration: Fig. 1426.]

[Ill.u.s.tration: Fig. 1427.]

The fine measurements necessary for shrinkage purposes render it necessary, where pieces of the same form and kind are shrunk on, to provide the workmen with standard gauges with which the work may be correctly gauged. These often consist of simple rods or pieces of iron wire of the required length. Figs. 1426 and 1427, however, represent an adjustable shrinkage gauge designed by H. S. Brown, of Hartford, Connecticut. Fig. 1427 is a sectional, and Fig. 1426 a plan side view of the gauge. A is a frame, containing at its lower end a fixed measuring piece B, and provided at its upper end with a threaded and taper split hub to receive externally the taper-threaded screw cap C, and threaded internally to receive a tube E, which is plugged at the bottom by the fixed plug F. The adjustable measuring leg G is threaded with the tube E, so as to be adjustable for various diameters of boxes, but it is locked when adjusted by the jamb-nut H. The operation is as follows: The cap-nut C and jamb-nut H are loosened and screwed back, allowing stem G and tube E to be adjusted to the exact size of the shaft for which a shrinkage fit is to be bored, as, say, in an engine crank. In setting the gauge to the diameter of the shaft, the cap end C and jamb-nut H are screwed home, so as to obtain a correct measurement while all parts are locked secure. The cap-nut C draws the split hub upon the tube E, and the jamb-nut H locks up G to E, so that the shaft measurement is taken with all lost motion, play and spring of the mechanism taken into account, so that they shall not vitiate the measurement. This being done, C is loosened so that E can be rotated, and raised up (by rotating) to admit the shrinkage gauge-piece J, whose thickness equals the amount to be allowed for the size of borer to be shrunk on the shaft. J being inserted, E is rotated back so as to bind J between the end of E and the foot piece B, when C is screwed down, clamping E again.

Thus the measuring diameter of the gauge is increased to an amount due to the thickness of the gauge-piece J. At the right of Fig. 1426 an edge and side elevation of J is shown, the 12/1000 indicating its thickness, which is the amount allowed for shrinkage, and the 6-inch indicating that this gauge-piece is to be used for bores of 6 inches in diameter.

The dotted circle K K L L represents a bore to which the gauge is shown applied.

The system of shrinking employed at the Royal Gun Factory at Woolwich, England, is thus described by Colonel Maitland, superintendent of that factory:--

"The inside diameter of the outer tube, when cold, must be rather smaller than the outside diameter of the inner tube: this difference in the diameter is called the 'shrinkage.' While the outer coil is cooling and contracting it compresses the inner one: the amount by which the diameter of the inner coil is decreased is termed the 'compression.'

Again, the outer coil itself is stretched on account of the resistance of the inner one, and its diameter is increased; this increase in the diameter of an outer coil is called 'extension.' The shrinkage is equal to compression plus the extension, and the amount must be regulated by the known extension and compression under certain stresses and given circ.u.mstances. The compression varies inversely as the density and rigidity of the interior ma.s.s; the first layer of coils will therefore undergo more compression than the secondhand the second more than the third, and so on.

"Shrinking is employed not only as an easy and efficient mode of binding the successive coils of a built-up gun firmly together, but also for regulating as far as possible the tension of the several layers, so that each and all may contribute fairly to the strength of the gun.

"The operation of shrinking is very simple; the outer coil is expanded by heat until it is sufficiently large to fit easily over the inner coil or tube (if a large ma.s.s, such as the jacket of a Fraser gun, by means of a wood fire, for which the tube itself forms a flue; if a small ma.s.s, such as a coil, in a reverberatory furnace at a low temperature, or by means of gas). It is then raised up by a travelling crane overhead and dropped over the part on to which it is to be shrunk, which is placed vertically in a pit ready to receive it.

"The heat required in shrinking is not very great. Wrought iron, on being heated from 62 Fahr. (the ordinary temperature) to 212, expands linearly about 1-1000th part of its length; that is to say, if a ring of iron 1000 inches in circ.u.mference were put into a vat of boiling water, it would increase to 1001 inches, and according to Dulong and Pet.i.t the coefficient of expansion, which is constant up to 212, increases more and more from that point upward, so that if the iron ring were raised 150 higher still (_i.e._ to 362) its circ.u.mference would be more than 1002 inches. No coil is ever shrunk on with so great a shrinkage as the 2-1000th part of its circ.u.mference or diameter, for it would be strained beyond its elastic limit. Allowing, therefore, a good working margin, it is only necessary to raise a coil to about 500 Fahr.,[22] though in point of fact coils are often raised to a higher degree of temperature than this in some parts, on account of the mode of heating employed.

Were a coil plunged in molten lead or boiling oil (600 Fahr.) it would be uniformly and sufficiently expanded for all the practical purposes of shrinking, but as shrinkings do not take place in large numbers or at regular times, the improvised fire or ordinary furnace is the more economical mode, and answers the purpose very well.

[22] The temperature may be judged by color; at 500 F. iron has a blackish appearance; at 575 it is blue; at 775 red in the dark; at 1,500 cherry red, and so on, getting lighter in color, until it becomes white, or fit for welding, at about 3,000.

"Heating a coil beyond the required amount is of no consequence, provided it is not raised to such a degree of temperature that scales would form; and in all cases the interior must be swept clean of ashes, &c., when it is withdrawn from the fire. With respect to the modes of cooling during the process of shrinking, care must be taken to prevent a long coil or tube cooling simultaneously at both ends, for this would cause the middle portion to be drawn out to an undue state of longitudinal tension. In some cases, therefore, water is projected on one side of a coil so as to cool it first. In the case of a long tube of different thickness, like the tube of a R. M. L. gun, water is not only used at the thick end, but a ring of gas or a heated iron cylinder is applied at the thin or muzzle end, and when the thick end cools the gas or cylinder is withdrawn from the muzzle, and the ring of water raised upward slowly to cool the remainder of the tube gradually.

"As a rule, the water is supplied whenever there is a shoulder, so that that portion may be cooled first and a close joint secured there; and water is invariably allowed to circulate through the interior of the ma.s.s to prevent its expanding and obstructing or delaying the operation; for example, when a tube is to be shrunk on a steel barrel, the latter is placed upright on its breech end, and when the tube is dropped down on it, a continual flow of cold water is kept up in the barrel by means of a pipe and syphon at the muzzle. The same effect is produced by a water jet underneath, when it is necessary to place the steel tube muzzle downward for the reception of a breech coil. As to the absolute amount of shrinkage given when building up our guns, let us take the 12-1/2-inch muzzle-loading gun of 38 tons as an example.

SHRINKAGES OF COILS OF 12.5 INCH R. M. L. GUNS.

-------------+---------------------------+--------------------------- | Shrinkages. | +-------------+-------------+ | | In terms of | Coils. | In Inches. | diameter. | Remarks.

+------+------+------+------+ | Rear.|Front.| Rear.|Front.| -------------+------+------+------+------+--------------------------- | | | D | D | Breech-piece| .022 | .026 | --- | --- | Shrunk on A tube.

| | | 857 | 807 | | | | | | | | | D | D | B coil | .055 | .01 | --- | --- | " "

| | | 561 | 190 | | | | | | | | | D | | B tube | .035 | nil. | --- | nil. | " "

| | | 668 | | | | | | | | | | D | D | C coil | .03 | .06 | ---- | --- | Shrunk on to breech piece | | | 1134 | 729 | and rear end of 1 B coil."

The objections to fitting work by contraction where accuracy is required in the work are, that if the enveloping piece is of cast iron its form is apt to change from being heated. Furthermore, if the enveloping piece, which is always the piece to be heated, is of unequal thickness all round the bore, the thin parts are apt to become heated the most, and to therefore give way to the strain induced by contraction when cooling, which, while not, perhaps, impairing the fit, may vitiate the alignment of parts attached to it. Thus, a crank pin may be thrown out of true by the alteration of form induced first by unequal heating of the metal round the crank eye, enveloping the shaft; and secondly, because of the weakest side of the eye giving way, to some extent, to the pressure of the contracting strain. To counteract this, the strongest part of the enveloping piece should be heated the most, or if the enveloping piece be of equal strength all round its bore, it should be heated equally all round. To effect this object heated liquids, as boiling water, or heated fluids, as melted lead, may advantageously be employed.

In some practice, locomotive wheel tires are heated for shrinking in boiling water. The allowance for shrinkage is from .075 millimetre to every metre in diameter, which is .02952 inch to every 39.37079 inches of diameter.

The employment of hot water, however, necessitates that the tires be bored very smoothly and truly, and that the wheel rim be similarly true and smooth, otherwise the amount of expansion thus obtained will be insufficient to maintain a permanent fit under the duty to which a wheel tire is submitted.

Shrinking is often employed to strengthen a weak place or part, or one that has cracked. The required size is, in this case, a cylindrical surface that is not a true cylinder, obtained by a rolling wheel rotated by friction over the surface to be enveloped by the band. Or if the surface is of a nature not to admit of this, a strip of lead or piece of lead wire may be lapped round it to get the necessary measurements.