283. Nothing aids more the proper accomplishment of any object than a correct idea of what is wanted. The following definition is given by Mr. W. B. Adams, of what good superstructure should be:—

“The principal requirements of permanent way are: That it shall be well drained, especially in contiguity to the substructure; that the weight and damaging power of the locomotives and rolling stock should be considered the data for calculation; that the strength, hardness, and tenacity of rails, and the immobility of the substructure should be adapted to the hardest work to which the railway is to be subjected; that the substructure should have an amount of bearing surface proportioned to the load to be borne, and the nature of the rail and ballast; and a sufficiently fair hold in the ground to prevent looseness or lateral motion, from the side lurches of the engines and trains; that the rails should possess so much vertical and lateral stiffness, either in themselves or in their fastenings, as to prevent all deflection; and have sufficient hardness of surface not to laminate or to disintegrate beneath the rolling loads; also, to have sufficient breadth or tread surface to diminish the crushing effect of the wheels.

“They should be as smooth as possible, to prevent concussion, and be laid at the proper angle, and the curves regularly bent, so as to insure the accurate tread of the wheels. The joints should be so made that the rails may practically become continuous bars, yet with freedom to contract and expand without being too loose. And with all this there should be interposed between the rails and the solid ground, some medium sufficiently elastic to absorb the effect of the blows of the wheels, without being crushed or forced down into the ballast, and yet stiff enough to keep the upper surface of the rails in a uniform plane.”

TIMBERWORK.

284. The timber-work supporting the rails consists either of cross ties of wood, hewn flat on top and bottom, of dimensions from 6 × 7 to 7 × 9, and 2½ or 3 feet longer than gauge; or of longitudinal sawed timbers rectangular in section, placed directly beneath the rail, and giving it a bearing throughout the whole length.

Longitudinal bearings seem to possess no advantage over cross ties, but are subject to some decided disadvantages. In case of removal, two rails at least must be taken up to admit of the replacing a timber; while with cross ties any one may be taken out and replaced without even affecting the immediate passing of a train. A continued bearing is no better than a broken one, as the strength of the timber itself offers very little resistance to the weight of a locomotive. Strength is not to be expected in the timber-work; it is only the elastic medium between the rail and the ground serving to maintain the rail in a proper position. The strength is in the rail. The distance at which to place cross ties depends upon the weight of engines traversing the road, the nature of the ballast, and the strength of the rail; somewhere between two and four feet from centre to centre.

The amount of superficial bearing which the timber-work ought to give per lineal foot of rail, is differently estimated by different engineers.

Upon the 4′ 8½″ gauged roads of America, 1¾ square feet per lineal foot of rail has been allowed.

Several of the English roads give the following:—

If ties are made eight inches wide, and eight feet long; we have the following amounts of bearing surface per lineal foot of the rail with different distances of the ties.

Of course the longer the tie is made the greater may be the distance between, provided the rail will bear it. Mr. Peter Barlow, in his report of August, 1835, to the directors of the Liverpool and Manchester Railway, fixes the following dimensions for superstructure.

At the time these dimensions were given, however, much less weight was applied to the rails than at the present day. As the bearing is increased, the rail must become heavier and more expensive; but the number and cost of the ties is lessened. The report above referred to, concludes that five feet bearings, involving heavier rails, would cost no more, after the road bed is consolidated, than shorter ones; but that on embankments and soft subsoils, it would be at first somewhat more expensive.

285. The object of the ballast is, first, to transfer the applied load over a large surface; second, to hold the timber-work in place, horizontally; and third, to carry away the rain water from the superstructure; it also furnishes the means of adjusting the timber-work to the proper position. It should be at least one half way up the depth of the tie, and deep enough below the under surface to prevent the timber being forced down by the passing weight. From various observations it appears that there should be one and a half times the depth of the tie of ballast, beneath the under surface; or the whole depth of ballast should be from two to two and one half times the depth of tie.

For ballast, broken stone, gravel, or other dry, durable, and porous material, is suitable.

A perfectly inelastic road bed is not to be desired. Something is necessary to absorb the shocks given by the wheels, and prevent their reaction against the machinery. To supply this amount of elasticity, and to transmit the weight evenly to the ground, is the duty of the ballast and timber-work.

Of late years there has been applied, in England, cast-iron hemispherical bowls, designed to take the place of both tie and chair. Such answers very well when there is no lack of ballast, and where wooden ties are worth from seventy-five cents to one dollar each.

SECTION OF THE RAIL.

286. A good rail must be able to act as a girder, or supporter, between the ties, as a lateral guide upon curves; and must possess a top surface of sufficient hardness and size to resist the rolling wear of the wheels.

Thus, at sixty dollars per ton, each square inch of section costs $943.20 per mile, or $94,320 per one hundred miles, whence the necessity of rolling the rail to the form which shall give the greatest strength with the least weight.

The sections most in use in America are shown in fig. 136, and 137.

Fig. 136 gives the most direct bearing, is compact, and brings the fibres at top and bottom more directly in opposition with the compressive and extensive strains. The top of the rail being curved to a radius of ten or twelve inches, the load is applied nearly to a single point; whence the whole resistance in fig. 137, depends upon the lateral resistance of the piece a b c d to being pushed down.

An objection is sometimes made to fig. 136, on the ground that it splits off on the line n n: this will not be the case when the head is joined to the web by a proper curve, as in fig. 136. This splitting off happens full as often in fig. 137 as may be seen where it is in use; and it might be supposed to act in that manner; because if the weight is transferred at all from the point of application to the web, it must be in the direction e f.

The rails in present use upon our roads, weigh from fifty to seventy-five lbs. per lineal yard; and are laid upon cross ties placed at a distance of from two and one half to four feet from centre to centre.

OF THE ACTUAL DIMENSIONS OF RAILS.

Digesting carefully the results of the experiments of Barlow, Fairbairn, and Hodgekinson, and the experience of Mr. W. B. Adams, and other English engineers; also the conclusions arrived at by the Berlin Convention of 1850, appointed to determine the best form of section, we come to the following limiting dimensions.

THE HEAD.

Mr. Barlow limits the width of head at two and one half inches as the maximum; the Berlin Convention, at two and one fourth inches; W. B. Adams, at two and one half; and all of the above recommend supporting the edges of the head well from the rib.

THE VERTICAL RIB.

The experiments of the Prussian engineers fix the thickness for a rail four inches high, at one half of an inch, and a rail four and one half inches high, at 0.6 or 6
10 inch. Mr. Barlow makes it six tenths of an inch for a four and one half inch rail, and 0.75, or three fourths inch for a rail four and five eighths inches high, and for four and three fourths inches high, 0.8, eight tenths inch.

THE BOTTOM FLANGE.

The use of this is more for bearing and fastening, than for supporting strength. The Prussian engineers make three and one half inches an ample base for a rail five inches high. The edge for one half or three fourths of an inch, should be nearly horizontal, or parallel with the base, to allow the spike to have a good bearing.

OF THE INCLINATION.

As the tread of the wheel is conical, the top of the rail must be inclined to fit this cone, otherwise the wear will come upon the inner edge of the rail only. This may be done in two ways; by placing the rail base level, and inclining the vertical axis of the cross section of the rail, and making the tread square with that axis; or by making the rail section true, and inclining the base, either by cutting the tie, or by a wedge placed between the rail and the tie.

OF THE HEAD CURVATURE.

If the top surface of the rail were perfectly flat, and the wheel tire does not happen to fit it exactly, (from want of the proper position of the rail, by settling, or other cause,) the wheel will bear entirely upon one edge, and would soon destroy the rail. To remedy this, a slight convexity is given to the top. Mr. Clark (in R. R. Mach.), recommends the top to be curved to a radius of ten or twelve inches.

OF THE VERTICAL DEPTH (HEIGHT) OF THE RAIL.

Mr. Barlow’s general results are as follow:—

In the London edition (1836) of Barlow’s Strength of Materials, page 402, in a report to the London and Birmingham Railway Co., upon the best form and upon the strength of rails; after a carefully conducted set of experiments, and an elaborate theoretical deduction of results, the writer comes to the following five sections of rails possessing the maximum strength, with the least weight.

This table shows the ratio of material which should be placed in the top and bottom.

With the above dimensions, and joining the curve of the head to the rib at two and one fourth inches from the top of the head, we obtain a strong and well-shaped rail, with the least material possible. See fig. 136.

As an example of the application of the above, the table below has been formed, showing four standard forms, which will be found to unite all of the requirements of good rails; the general form being that of fig. 136.

and the following figures show the weights which should be applied to differently spaced sleepers.

The amount of inclination or bevel to be given to the cross section of the rail, depends directly upon the cone of the wheel, and indirectly upon the gauge of the track. (See Chapter XIV. part 2.) The radius of curvature being averaged at 2°, or 2,865 feet,

in the width of the rail, or two and one half inches.

The above dimensions embrace all of the best results of experiment and experience, and at the same time satisfy the conditions demanded by the mechanical and physical nature of the material—iron.

CHAIRS AND JOINTS.

287. The chairs most common at present are made of a wrought iron plate, with two lips, either cut and punched up, or forged up, to hold the lower web of the rail. Such chairs weigh from six to ten pounds each, and are less liable to break than the common form of cast-iron chairs. It is probable that a cast-iron chair may be made, however, with properly shaped lips, and so hollowed out as to be at once strong and light. (See Clarke’s R. R. Machinery, “Permanent Way.”)

Of late the chair of Mr. David L. Davis, of Dedham, Mass., has attracted considerable attention, and bids fair to be the means of obtaining a better rail surface than has heretofore been possible. This gentleman has been for twenty years Road-master of the Boston and Providence Railroad, and has had ample opportunity for considering the subject of track laying in every respect. The rail bears upon a cap of wrought iron, which rests upon a piece of rubber, lying in the chair. The testimony of the leading managers of the New England Railroads bears witness of the excellence of the arrangement.

The practice of notching each end of the rail causes the expansion to be exerted directly against the fastenings, which should not be the case. Some point should be fixed longitudinally, to resist the end shocks from the wheel. This point should be either the centre or one end of the rail. End chairs may hold the rail laterally, and vertically, but not longitudinally.

The weakest part of the track is that, where, to resist the concussions of the wheels it should be strongest, namely, at the joint: here we lose the strength of the rail and depend entirely upon the tie. The flattened ends of rails which have been laid for a few years show the bad effect of the common joint. The complete remedy for this is, so splicing the rail that it is as strong at the joint as elsewhere. The method termed “fishing,” is not much more expensive than the ordinary method of jointing, it is perfectly effectual, and has had the test of long and successful use. It consists in bolting a plate two and one half feet long, two and one half or three inches wide, and from one third to one half inch thick, to the ends of both rails making the joint; one plate being placed on each side. The plates are convexed a little from the rail as in fig. 138, so that being sprung by screwing on the nuts, the latter shall not work loose by the vibration of the rail.

In the above arrangement there is no tie below the joint, but the latter lies midway between two sleepers.

Another method of “fishing” is, to place a piece of ᕼ or ⊤ iron beneath the rail, bolting it firmly to the lower flanges.

In bolting rails together at the ends, the bolt holes must be cut a little larger than the bolt, to allow for the expansion of the iron.

The effect of the joint upon the passing carriage, is the jumping motion; the middle of each rail being a summit, and the end a depression, (the strength at the joint being taken away); and if the joints are not opposite to each other, there is generated a very injurious and dangerous side rocking. Figs. 138, 138 A, and 138 B, show the methods of fishing.

To avoid the wear caused by frequent joints, various forms of compound rails have been proposed; consisting of two or more parts breaking joint. One form has been contrived in which the section is vertically halved; another of three parts, a head placed on top of a double vertical web. Fig. 139 shows what would seem to answer any purpose (if compound rails are at all allowable). The joint is here divided into four parts, so that the strength of the bar at any point is reduced only one fourth. In bolting the parts together the joints should be left open enough (see in advance) to allow for contraction; and the bolt-holes, as before noticed, should be longer than the bolts. (This enlargement, extending only in the direction of contraction, and not in the line of the force.) The upper part of such a rail should be hardened to resist the rolling of the wheels, while the webs must possess the strength to act as a girder.

It is questionable whether, by dividing the rail, particularly when it is done horizontally, we do not prevent the mutual extensile and compressive actions which ought to have place in the top and bottom; for we cannot make the bolts perfectly tight because of expansion.

Some of the compound rails which have been laid in America have given good results, others have not.

Mr. W. B. Adams observes, that a compressed rail to be as strong as a sixty pound whole rail, must weigh ninety lbs. per yard.

Some engineers have proposed such a rail that when one side becomes worn it may be turned over so that the lower may become the upper table. This is quite wrong in principle; as when the lower fibres have been subjected for some time to extension, they are entirely unfitted to oppose compression.

OF THE LIFE OF RAILS.

288. The time which a rail will last, depends upon the form and weight, and on the quality of the iron; and upon the number, weight, and speed of engines and cars passing over it.

Note.—The effect of quality is altogether too little regarded in America. How worthy of attention it is may be seen by the following.

Upon the same road were used two kinds of seventy-two pound rails, each five inches deep, and having a bearing surface of 2.7 inches in width. The one was worn out with a tonnage of 41,000,000 tons, the other of 22,000,000 tons; the difference being entirely in the quality of the iron.

Upon the Philadelphia and Reading Railroad there have been used forty-five pound rails of reheated and refined iron, which have lasted for eighteen years; and that with a very heavy traffic upon them. While upon other American roads, English sixty pound rails have required renewing in one, two, three, and four years.

The durability of rails is practically independent of time, and depends entirely upon the amount of work done. The repairs of iron, depending upon flaws and other physical defects, will be greater at the commencement of operations than afterwards. After the first one or two years the regular depreciation begins. The first Liverpool and Manchester rail weighed thirty-five lbs. per yard, and the locomotive seven and a half tons. As the traffic increased, so did the necessary weight of engines, and a corresponding increase in the strength and weight of rails was also rendered necessary. In 1831, the average weight of engines with tenders was eighteen tons. In 1855, the maximum engine with tender, fuel, and water weighed sixty tons; and in like manner the rails increased from thirty-five to eighty-five lbs. per yard.

Messrs. Stephenson and Locke, in a report to the London and North-western Railroad Company, in 1849, recommend the adoption in future of an eighty-five lb. rail.

Upon the roads of Belgium are used rails of fifty-five and sixty-four lbs. per yard; but it is asserted that an eighty lb. rail would allow of ten times more traffic.

For the average of American roads, when the iron is good, (in quality,) fifty-five, sixty, and at most sixty-five lbs., will probably be found ample for the heaviest traffic: the rail being of the form already given, and supported on ties not more than two and a half feet from centre to centre.

Mr. Belpaire, (of the Belgium engineers,) concludes, from many experiments, that in sixty miles, each engine abrades 2.2 lbs.; each empty car 4½ oz.; and each ton of load 1.4 oz.; the amounts being in direct ratio to the several weights.

Captain Huish, of the London and North-western Railroad, (England,) estimates (Report of April, 1849) that fifty trains per day, or 18,250 trains per annum, for twenty years, would wear out a seventy lb. rail.

The Belgian engineers have concluded that 3,000 trains per annum, for one hundred and twenty years, would wear out a fifty-five lb. rail.

Now 120 × 3,000 = 360,000 Belgian, and 20 × 18,250 = 365,000 English, a very satisfactory coincidence, as the different observers did not know of each other’s proceedings. The difference, 5,000 trains, being accounted for by the use of heavier engines upon the roads of England.

From the above results the following table is formed, showing the life of rails under from two to one hundred trains per day. American roads being less nicely finished, as regards the road-bed, will of course wear out rails faster than the roads of Europe. The table will serve as a base for estimates.

Probably one half of the above numbers of years would show the full life of rails upon American roads.

As those rails which are most used wear out the soonest, they should be made accordingly heavier. Such are those at depot grounds and at sidings.

Note.—From the reports of the Reading (Penn.) Railroad it appears that in 1846 153
209 of the damaged rails were split; and that in 1845 285
295 were split.

As regards the quality of railroad iron, it is generally notoriously bad, and its makers know it as well as those who buy it. Railroad companies are not willing to pay for good iron. Comparisons between American and English iron amount to little. First rate iron can be made in England or in America, and so can that which will last about two years. Time will convince companies that the most expensive iron is the cheapest.

TABLE OF THE WEIGHT PER MILE OF DIFFERENT RAILS.

TRACK-LAYING.

289. As wrought iron expands 0.0000068 of its length per degree (Fahrenheit) of heat, a change of 130° will cause the following expansions:—

and that the track may be kept in the right vertical and horizontal line, rails laid in cold weather must not be placed in contact; but separated by space enough to allow expansion to take place. In hot weather they may be placed close together. Calling 100° the maximum and -30° the minimum, we form the following table for the average lengths of rail, (20 feet).

The proper distance of rails may be fixed by the use of the steel plates shown in figs. 140 and 140 A, which are marked with the temperature, according to their thickness, as in the above table.

To incline the rail base may be used, when the rail is not bevelled, wedges one foot long and six inches wide, spiked with the rail to the tie. When the chairs are of cast-iron, they may be cast to the required slope.

FROGS.

290. When one line of rail crosses another, a contrivance called a frog is used; see figs. 141 and 142.

That the wheel may run smoothly from a to c, fig. 141, the rail b f must be cut at D, and the rail a c must be cut at the same point. Cutting the two gives the form shown in the figure, and further developed in fig. 142.

In order that the flange of the wheel shall not leave the line a c, when at the break D, the guard rail m m is used to confine the opposite wheel. It should be placed at a distance of two inches from, and parallel with, the main rail g g, from opposite six inches below the frog point at s, to six inches above the shoulder at s′. From the ends of the parallel line n n the guard rail should gently curve away at both ends. Thus the wheel will be gradually brought into the right line, kept so until the break in the rail is passed, and finally easily released. To place and maintain the guard rail in the right position, it is well to put both it and the main rail into a double chair, which is spiked to the sleeper.

The form and dimensions of the cast-iron frog depends upon the angle at which the cutting rails cross, and upon the size of the wheel tire.

To draw the frog, proceed as follows:—

Let a c b be the angle. Parallel with and two inches from b c draw d e, e being in a c produced. In the same manner fix the point g. At the width of the rail head (from 2¼ to 2½ inches) draw, parallel to a c, L 8. The point 8 is the limit to the solid steel. At double the rail width, or 4½ inches, draw, also, parallel to a c, 16. 6; 5. 6 is the limit of the flat steel, generally about half an inch in thickness. This is the least amount of steel allowable; it is best to steel the whole tongue, and all of that part of the wings acted upon by the wheels. The geometric point is generally very thin, and is omitted to a distance far enough back to make the point a third or half an inch wide, which is rounded off; e L and d k are made two and a half inches; as also f m and g n; k 10 and m 11 are made six or seven inches, and joined to d and f by a curve, abrupt at first, but afterwards more gentle. The distances, 5 a and 6 b must be such that a 9 is three and one eighth inches, (depending upon the breadth of rail base,) o m″ is from three to four inches. At the other end of the frog e h must be enough to make s t at least an inch, when e h and i g are from three to four inches; i m′ being, as at the other end, three or four inches. The steel plates N N are one half inch in thickness. The surface, N, is two inches above the bottom, M. The lower plate, M, is two inches thick. A B, C D, and E F are six or seven inches wide, and one inch thick. The spike holes 11
16 square, the spike being one half inch. The sharp edges, i g, e h, a c, b c, should be rounded off to fit the wheel at A, fig. 142 A. The surface of the tongue N 9 should be formed to a double incline to fit the wheel cone.

Note.—Fig. 142 A gives the shape and dimensions of the largest tires.

Another method of making a frog is to cut and weld the rails a and b of the track, as in fig. 143. The continuations of these rails are bent as shown in the figure.

The whole angle is placed upon a firm wooden bearing.

There is no weaker part of the track than the frog. To make up the strength at such places a heavy longitudinal timber twelve feet long will answer a good end.

291. The object of the switch is to adjust a single line of rails to two or more pairs, so that any two lines may be made continuous. The form in general use consists of two rails, as at a b, a b, fig. 144, moving upon a and a as centres. Here the tangent point of the turnout curve is at c. The data given for the switch are the length of switch rail and the motion at the toe (c) (which determine the direction of the starting tangent) and the radius of curvature of the turnout curve. The required elements are, the angle of frog at b and the distance from a to the point of the frog.

The following formula and table are by Josiah Hunt, Esq., (at present chief engineer of the Hannibal and St. Joseph Railroad, Mo.). The formula was first published in Appleton’s Mechanics’ Magazine, vol. 1, p. 575.