DCCLXXX.* NOTE ON THE CARBON-CONTENTS OF PISTON-RODS AFFECTING ENDURANCE AS FATIGUE. THEIR UNDER BY JOSEPH E. JOHNSON, JR., LONGDALE, VA. THE subject of the relative endurance under reversed stresses of materials of different tensile strength and hardness, has recently been attracting a considerable amount of attention from engineers, and the theory, until recently universally held, that for withstanding these repeated or reversed stresses a very soft material was necessary, has had to be revised to the extent of complete abandonment, at least in many cases. It is not with the idea of communicating anything new, but merely as a strictly practical confirmation of the valuable work done by others that the following data are given. About six years ago the company with which the writer is connected bought a compound locomotive of the Baldwin or Vauclain type, no description of which is needed before this Society, except to recall, for the sake of clearness, the fact that the high and low pressure cylinders lie as close together as possible, one vertically above the other, the rods from the two cylinders being fastened to the same crosshead, which is of the four-bar type, and located centrally between the two rods, as shown by the accompanying drawing (Fig. 170). The wings or guiding surfaces are made very long in the direction of the stroke, to overcome the torque set up by the unequal and constantly varying pressures on the high and low pressure pistons respectively. These pressures are made as nearly equal as possible by the steam distribution, but practically there is always considerable difference at some part of the stroke, so that there is a stress tending to tilt the crosshead one way during one stroke and the opposite way during the other. This stress puts a considerable pressure on the diagonally opposite corners of the guiding wings, and, * Presented at the Niagara Falls meeting (June, 1898) of the American Society of Mechanical Engineers, and forming part of Volume XIX. of the Transactions. the reciprocating motion going on while under this pressure, wear takes place on the corner of the wings first, and allows a slight rocking of the crosshead, a complete oscillation occurring at each revolution when running under steam. The piston-rods are fastened to the crosshead with the regular taper fit drawn up to a shoulder by a nut. This connection being rigid, and the opposite end of the rods prevented from vibrating with the crosshead by the fit of the pistons in the cylinder, the rods are bent at the shoulder through a very small arc in each direction vertically, at each revolution. This first locomotive ran for three years and two months, when a duplicate was bought, and the first put in the shop for a general overhauling previous to taking the place of the smaller engines on another part of the road, the new one taking the run of the old one. During the overhauling the piston-rods were renewed, having worn down too small to work well with the metallic packing any longer. The material for the new rods was ordinary "machinery steel," taken from stock on hand. The rods on this engine (No. 4), it should be stated, were straight from shoulder to shoulder, while those of the "duplicate" (No. 5) were reduced in the body, having a collar inch larger than the rod and inch wide next to the shoulder at the crosshead end. After having been in service about fourteen months, one of the low-pressure rods of No. 5 "let go," and smashed the cylinderhead, without, however, doing any very serious damage. Within a few weeks the overhauled engine did the same thing. This was becoming a serious matter, and after some careful consideration the writer ordered some genuine Swedish iron to make rods of. It was beautiful stock, and so soft that it acted almost like lead in the lathe, being very difficult to get a smooth finish on. A set of these was put into one of the engines at once, and ran about four months, when one of them let go in the same way. The rods that broke were all low-pressure ones, due undoubtedly to the fact that in the "emergency," or starting gear, those cylinders get almost full boiler pressure; 180 pounds per square inch. The rods were all broken in the same way, and right in the shoulder, the metal cracked at top and bottom, and the crack gradually widened, as could be seen by the worn appearance of the upper and lower segments of the break, which gradually approached each other until only a narrow horizontal strip of solid metal was left across the middle of the rod when the final rupture occurred. Soon after ordering the Swedish iron, the writer came across one or two articles bearing upon this subject of the endurance of soft and hard steel or iron under fatigue, and describing tests made to elucidate this point, notably those of the Pope Tube Company and the Bethlehem Iron Company, which showed quite clearly that high-carbon steel was infinitely better than low-carbon, and that nickel-steel was better than either for such service; also that very soft material, like Swedish iron, lacked endurance under fatigue. Therefore the breaking of the rod of this material was not a very great surprise, and was met by ordering material for a set of rods of high carbon and one of nickel-steel from the Bethlehem Iron Company. These have now been in considerably over a year, and we hope that they will last long enough to wear out without breaking. The writer had the three rods which had broken, and the one which had worn out, analyzed, to see how they bore out the theory of high-carbon material versus low. It will be seen that these results bear out the theory to a striking extent, there being nothing in No. 1 to cause its far greater endurance except the carbon, and possibly to a slight extent the sulphur, which is also claimed by some to be a hardener. It is very difficult to deduce any quantitative results as to number of reversals of stress producing flexure even approximately, because even given the approximate daily mileage of the engines and the size of the drivers, it is impossible to say what portion of the total running was done under steam, the grades being quite heavy, and the trains running by gravity for nearly half the total distance. If thirty miles per day under steam, twenty-eight days per month, be taken, the diameters of the drivers being thirty-six inches, the revolutions per day would be, say 16,000, and per month say 450,000; this would make for the second and third rods about 6,000,000 double flexures before rupture, and for the Swedish iron rod say 1,800,000. There is no way of giving the amount of flexure; the crosshead probably never tilted more that inch in twenty-four inches to either side of the vertical, but this amount varied as the wear occurred, and was taken up; also it is not possible to tell what portion of the total length of the rod absorbed this flexure, so that it is impossible to give any figures having a scientific value. The theory of the superior endurance of harder materials under fatigue has been explained many times, and by those far more competent to do it than the writer, so that nothing on that subject is said here. As stated at the beginning, this is only intended as a strictly practical confirmation of facts already brought out by the splendid researches of others. DISCUSSION. Mr. Thomas R. Almond.-Hardened steel, if immersed in a liquid such as a solution of cyanide of potassium, may become weakened, and if it remains there long enough it may break. A fracture may be started as soon as immersion takes place, or something similar to a fracture. There is a condition which I do not understand. The liquid seems to get in between the molecules at some portions of the surface, and weakens the material. The first time that I observed this was when cleaning some steel springs which were under tension. I left them in a weak solu tion of cyanide of potassium over night, and to my surprise in the morning they were broken. It set me thinking, and one of the conclusions to which I came was that, where rods are continually under stress, possibly the molecules may become sufficiently separated at the most distressing time of the motion of the rod, and any acid which may be present in the oil or lubricant which is used may possibly have an influence upon the material towards weakening it. I have often thought that the breakages which occur in the crank-pins of locomotives, at the weakest portionthat is, the corner which comes nearest to the wheels-are probably due to acid in the lubricant being absorbed when the stress is greatest. It is easy to understand that the molecular density will be less at that moment, and that there may then occur an absorption which perhaps might not occur before. This continued through a long period of time, and under the same circumstances every time, may very materially assist towards bringing about fractures which are often considered as being very mysterious, more especially when the material used is known to be of high quality. Mr. H. H. Suplee.—I think in considering the breakages which are described in the paper, sufficient emphasis is not given to the peculiar conditions under which the piston-rods in these Vauclain compound engines work. I have had occasion to observe a number of these engines running in and out of Philadelphia, and there is certainly a very severe and very sudden stress on those rods. In the first place, there is nearly always a considerable amount of clearance between the guides, and if they are not so made they soon become so. This permits a tilting of the crosshead which is very marked on starting; and, furthermore, it takes place in a very short space of time, and is very much more in the nature of a blow on the rods than of a bend. When the steam is admitted you can hear the crosshead tilt with a clank like the blow of a hammer, and then there will be a corresponding sharp blow on the return. The bending is not distributed over the whole length of the rod. It occurs in the first few inches of the forward stroke and the same on the back portion of the stroke. It occurs until the engine is well under way, when the pressure in the two cylinders is equalized and the sound disappears. Then I think it is supposed that the pressure in the two cylinders is so nearly equalized that the bending is comparatively slight. The result is that these rods at starting, for a few strokes, are practically being hit a sharp blow at the end of the stroke, and then, after that, these |