Reprinted from American Machinist Vol. 20, No. 50, 1897.
A most disastrous boiler explosion occurred on the afternoon of November 22d at the Graves Elevator Works, Rochester, New York, accompanied by circumstances which make it peculiarly worthy of more than passing notice. The boiler plant was a thoroughly first-class one in every respect. There were two boilers, only one of which was in use at the time of the explosion. It was not old, was well made and well taken care of. There is absolutely no evidence of carelessness or of improper conditions, so far as they could have been known at the time. The usual pressure of 90 pounds was on the boiler at the time of the explosion. The effects were such as should be expected. The boiler was practically blown to pieces; walls were knocked down, and bricks were thrown, some of them to a great distance. The only two persons in the boiler room were instantly killed, and buried under the debris. Three others were seriously injured, and several received slighter wounds. One of the men killed was the regular fireman of the boiler, and the other was Patrick Shields, an inspector of the Hartford Steam Boiler Inspection and Insurance Company. Mr. Shields had been in the boiler room not more than five minutes, and was found seated in a chair, covered with bricks and other material, and shockingly mutilated.
The following extract from a letter from J. M. Allen, president of the Hartford S. B. I. & I. Company gives some additional information, and also reveals the probable cause of the explosion:
'The boiler which exploded in the works of the Graves Elevator Company, of Rochester, New York, by which one of our inspectors lost his life, was a boiler that was only four years old. It was 66 inches diameter, 16 feet long, constructed of ****-inch steel plates; dome, 36 x 39 inches. There were 98 3-inch tubes; was well built and well set, and the boiler room was regarded as one of the best equipped in that vicinity. When the boiler was constructed, the plates were all tested by coupons being sheared from their edges, and the tests were very thorough and very satisfactory.
'Now, as to the cause of the explosion, our chief inspector has made a careful examination, and has satisfied himself that it resulted from one of those incipient cracks that are hidden in the joints by the over-lap of the sheet. We have found several similar cases. The cracks are so located that they cannot be discovered by an inspection, as from an internal inspection the inner lap does not show any fracture whatever, and the outer lap covers the fracture which has not worked through the plate. As to the cause of this kind of fracture, we have pretty well satisfied ourselves that it comes from an effort to bring the plates together at the edges after they have been rolled; that is, the outer lap sometimes 'cocks up' a little above the lower plate, and is brought into place by hammering. We are led to arrive at this conclusion from the fact that, in visiting boiler works, we have seen this kind of practice, and have protested against it, as it is very liable to start an incipient fracture which, under the conditions of use and pressure, is liable to develop into a dangerous weakness. There seems to be no other well-founded reason for such a weakness being developed. The holes are punched in the plate and considerable metal is taken out, and then the effort to bring the lap of the plate together would be liable to bring a pretty severe stress at the points where the plate is forcibly bent.
'I do not wish to be hypercritical in regard to the construction of boilers by the better class of boiler makers, and could not say that this was the practice in the works where this boiler was made; for the boilers are regarded as first class in material and workmanship. At the same time, there seems to be a point which it is well to consider, and which we have written about, and concerning which we have sent out printed information among all the boiler makers of the land.
'Mr. Shields, the inspector who lost his life by this accident, made inspections of the boilers in a certain portion of the western part of the State of New York. He had been in Buffalo, and arrived in Rochester that day, and was busy looking over the various boilers insured, on what we call our visits of inspection; that is, external inspections to see that gages, safety valves and all attachments were in proper working order. The internal inspection had been made some time previous. He had just entered the boiler room: had not been there five minutes; was chatting with the boiler tender, when the explosion occurred without a moment's warning, so far as we can learn, without even a hiss of escaping steam. The examination of the fragments showed a fracture on one side of the boiler, which was constructed in three sections of plates, and the joint which ruptured was up out of the fire some distance, and was covered by the brickwork of the setting. I do not think any inspection, save cutting the boiler to pieces, could have detected this defect. There had been no indications of leakage, so far as we are able to ascertain no indication whatever of any weakness; but the crack was there.'
The present writer knew Mr.Shields for a number of years as a shopmate, and he has been met occasionally since he was in the service of the Hartford Company. He was an excellent boiler maker, thoroughly informed as to the best boiler practice, and was a faithful, thorough and reliable inspector. That he should have lost his life under the circumstances narrated, is one of the remarkable facts of life which continually challenge fiction.
In this connection, it seems to be no less than our duty to reproduce from the Locomotive a portion of the article which it has twice printed upon this danger which lurks unseen in boiler construction. The article referred to was entitled 'Cracked Plates.' The part especially bearing upon the cause of the explosion referred to above is as follows:
'The accompanying cut shows a crack due to a different cause, and it ought to carry with it a useful lesson. It represents a piece of plate that was cut from a boiler in active service, and which was believed to be in good condition. The boiler from which it was taken was 48 inches in diameter, with tubes 15 feet long, and the plates were of steel, of an inch thick. The piece of plate shown in the cut formed the edge of one of the sheets, where two sections of the shell were united by a longitudinal, double-riveted lap-joint. It was taken from the upper part of the boiler, and was not exposed to the fire. It contained one well-marked crack extending completely through the plate, besides many other shorter ones running into one another in all sorts of ways, some of them extending through the plate, and others not quite through it. All these cracks were entirely covered by the inside lap of the joint, so that they could not be seen from the interior of the boiler: and on the outside, the boiler was covered at this point by a thick layer of non-conducting asbestos covering. We mention these points in order that the reader may understand how easy it would be to overlook this defect. Yet it would not be putting the case too strongly to say that although the boiler appeared to be in good condition, it was actually on the verge of explosion. For a considerable distance along the joint the strength of the plate was entirely destroyed, and at other places it was held together by the merest skin of metal, as was afterwards shown by breaking the plate across along the line of the cracks. The fractured area was almost entirely black, though bright spots were noticeable at intervals of two or three inches or so.
'The cause of this defect will be sufficiently obvious to those who are familiar with the processes of boiler making. In rolling plates into the cylindrical form, preparatory to riveting them up into shells, it is customary to bend one end of the plate to what is judged to be the proper radius, by the use of the sledgehammer. The plate is then run through the rolls and rolled into shape, the end that was previously bent being introduced first. When the plate has been rolled all but the last 5 or 6 inches, the last end slips off the first roll, and the rolls can no longer 'grip' the sheet. The result is, that the last end of the sheet is not bent to the proper radius, but remains straight, or nearly so.
'In order to bring the outer lap to the proper curvature, it is customary for one man to hold a sledge against the projecting edge of the lap, while another workman strikes the shell on the inside. In this way the lap is bent down into place, and after the shell has been brought to conform with the 'sweep' or templet, in every part, it is ready for riveting.
'Now it will be seen that the treatment required for bringing the laps together in this manner is rather violent; and it follows that nothing but the best of materials will stand it without being greatly distressed and permanently weakened. Under the sledging operation the material is likely to be strained beyond its elastic limit, unless it possesses great ductility. The greatest strain on it comes on the outer lap, at or near the line where it touches the inner one, or along one of the lines of rivet holes, where the plate is weakened by the loss of material. We have no doubt that the cracks shown in the cut were started in this way, and that they afterwards crept into the plate gradually as the boiler yielded slightly under varying pressures, until they reached the highly dangerous state described above.
'If the sledging were done while the sheet was hot, it would not be so objectionable; but the great majority of boiler makers will not attempt to heat the plate before sledging the lap down, because when the sheets are hot they are apt to buckle out of shape and give great trouble. If the sheets are to be sledged cold, the proper way to do it is to bend each end to the proper radius before beginning the operation of rolling. A convenient way to do this is to lay the ends of the sheet over the upper roll, and bring it down to the proper radius very gradually.
'In the early days of steel boilers, before the manufacture of that material was understood as well as now, plates were much more apt to be injured by sledging than they are at present. Steel having a high tensile strength is almost certain to be deficient in ductility; and for this reason it is customary, in the specifications sent out from this office, to make the maximum allowable strength of plate 65,000 pounds to the square inch, when such plate is to be exposed to the fire. We also specify that the steel used shall show an elongation of 25 percent, in a length of 8 inches, that it shall show a reduction of area of not less than 56 percent, and that its elastic limit shall be at least 50 percent of its ultimate strength. The plate should also be capable of being bent double and hammered, when either hot or cold, without showing cracks; and it is also desirable that it should stand this same test after being heated and quenched in water. Steel that possesses these qualities makes excellent boilers, and it will stand a great deal of abuse in the boiler shop without developing defects in after service.'