§ 6. Expansion and Contraction of Wrought-iron Bars.

In the use of cast iron for railways, the bars or rails are necessarily short; and, therefore, the effect of the variation of temperature does not produce any practical evil, each length of rail being supported by a block. But in the use of malleable iron for railways, the rails are made of considerable length, of now generally fifteen feet, and, consequently, the amount of expansion and contraction upon that length of iron is considerable; and the rail being supported by several intermediate bearings, between the ends, it becomes a subject of importance to ascertain the effect, in practice, of the expansion and contraction of the rails, from the variation of temperature in summer and winter.

The amount of expansion and contraction of wroughtiron bars, by a variation of temperature, has been ascertained by several experimentalists. Mr. Smeaton found the expansion equal to Toth part of its length for each degree of temperature. Professor Daniel more recently, and with more accurate apparatus, found the expansion 5th part. Taking, therefore, the variation of temperature, between summer and winter, to be equal to 76°, a bar of malleable iron will expand and contract, between the two extremes of temperature, the 2000th part of its length. Taking railway bars, at fifteen feet in length, the amount of expansion and contraction will, therefore, be equal to the 11th part of an inch, for a range of temperature of 76°.

Amongst railway engineers, it has become a subject of discussion, whether the peculiar application of bars of iron as rails upon railways, and laid down upon

blocks of stone, imbedded below the surface, did, or did not, follow the precise law of expansion and contraction of common bars of iron; and this was brought more prominently forward by the knowledge, that, when railway bars were laid upon blocks of stone, there did not appear the same twisting or bending of the rails, in summer, as with bars laid upon timber.

To determine this question, which is of great importance in railway practice, I had a double length of rails laid down upon the Killingworth railway, 152 feet in length, and composed of fifteen railway bars of nearly fifteen feet each long. The joinings of each bar with the next, were so constructed, that they could not separate, being formed with half-lap joints, and bolted together by a bolt, completely filling the hole in each half-lap. By this plan, all the bars were combined, and, therefore, the expansion and contraction were the same as for one bar 152 feet in length. The rails were supported, at each three-feet length, upon blocks of stone of three cubic feet each. The mode of measuring the amount of expansion and contraction was as follows:— On each side of the rails, at each end of the length, the subject of the experiment, wooden piles were driven into the ground, sufficiently deep to be quite firm and immoveable, the heads of which were level with the under side of the rails. Pieces of iron were fastened to the head of these piles, on each side of the rail, upon which a line was cut, transversely, from one side of the rail to the other. A similar line was cut upon the rail, corresponding with the mark on the iron of the pile; and the temperature taken, when the lines on the rail corresponded with the lines upon the iron of the pile. These, as before stated, were done at both ends of each length of rails so bolted together. When, therefore, the

rails contracted, the lines upon the rails would be short of those upon the posts, and when expanded, they would extend beyond. Observations were made, therefore, at certain intervals, during the whole of the winter of 1836, and summer of 1837, and the extent of extension and contraction accurately measured. It is necessary to observe, that the joint at the end of this length of road, was not firmly united with the other part of the railway, but sufficiently loose to allow the rails to stretch

or contract.

This experiment, therefore, gives the expansion and contraction of a length of railway or bar of iron in use equal to 152 feet, and the result of the experiment made during the whole year was, that the range of temperature, when the observations were made, was 50°, viz. from 31° to 81°, and the extent of contraction and expansion equal to 685 of an inch; which gives the amount, equal to about the 15th part of an inch, for a fifteen-feet rail, between summer and winter, or from the freezing point to 81°.

We see, therefore, that the expansion of railway bars in use, does not vary from that shewn by common bars of iron, subjected to experiment by a variation of temperature, produced by artificial means; and that the amount of expansion or contraction of a railway bar, fifteen feet in length, is about the 750th part of an inch for each degree of temperature.

§7.-Experiments on the Strength of Cast-iron Rails.

We shall now give an account of some experiments on the strength of cast-iron rails, made at Walker Foundry, near Newcastle-upon-Tyne, the property of Messrs. Losh, Wilson, and Bell.

The rails were all cast from the same pattern, the difference in weight being accidental. Sections

98 2 14

100 3 14

similar to Figs. 1 and 3, Plate II., two inches and a quarter broad on the upper side, and tapering away to one inch and a half in the middle, and again swelling out at the bottom into the square c b, each side of which is seven eighths of an inch. Extreme depth, in the centre, six inches, gradually decreasing towards the ends, or points of support, in a parabolic form, to four inches. In the experiments, the rails were fastened in the usual manner to the chairs, which were fixed upon beams of wood; distance between the points of support, three feet, and nine and a half inches.

In comparing the strength of the different rails with each other, we find a great variation, not only between the different kinds of metal, but also in rails cast from the same metal. The only constant and regular law appears to be, that the weight or specific gravity of rails, formed of a mixture of different kinds of metal, is uniformly greater than of one description of metal separately, and, also, that such a mixture makes the rails invariably stronger. This is a very useful discovery, and enables the founder, by mixing different metals in the proper proportions, to form a rail much stronger, with the same weight of metal, than otherwise could be done by casting them of any particular kind of metal alone. The depth of the middle section of the cast-iron rail renders them very rigid, and the deflection is comparatively trifling before fracture. The recorded weights are those which produced fracture; in loading the rails, the weight should, of course, be much less than that which breaks the rail. Inequalities of the road, or occasional obstacles occurring upon the surfaces of the rails, will sometimes produce jerks, or shocks to the wheels of the carriages, and the reaction will transfer those to the rails, and cause blows, which, from the