1912-014: 7 March 1912

[Page 3][1]

INTERSTATE COMMERCE COMMISSION

REPORT OF THE CHIEF INSPECTOR OF SAFETY APPLIANCES COVERING HIS INVESTIGATION OF AN ACCIDENT WHICH OCCURRED ON THE WABASH RAILROAD NEAR WEST LEBANON, IND., MARCH 7, 1912, ACCOMPANIED BY REPORT OF THE ENGINEER-PHYSICIST OF THE BUREAU OF STANDARDS COVERING HIS INVESTIGATION OF THE BROKEN RAIL CAUSING THIS ACCIDENT.

August 19, 1912.

To the Commission

[¶1]  On March 7,[2] 1912, there was a derailment of a passenger train on the Wabash Railroad near West Lebanon, Ind. This derailment caused the death of 2 passengers and the injury of 76 passengers and 11 employees. After investigation as to the nature and causes of this accident and of the circumstances connected therewith, I beg leave to submit the following report.

[¶2]  Train No. 1, known as the Continental Limited, runs from Buffalo, N. Y. to St. Louis, Mo. At the time of the derailment it consisted of engine No. 619, one express car, one mail car, one combination baggage and express car, one coach, two Pullman sleepers and one dining car.

[¶3]  With Conductor Grant and Engineman Duke in charge this train left Peru, Ind., a division terminal, at 2:19 p.m., 1 hour and 15 minutes late, and left West Lebanon, Ind., the last station which it passed prior to the accident, at 4:38 p.m., 1 hour and 29 minutes late. The derailment occurred at 4:41 p.m., at a point about 2 miles west of West Lebanon. The entire train with the exception of the engine and tender left the rails and the derailed cars were thrown down an embankment 16 to 20 feet high on the north side of the track. Their positions after the derailment are shown in the illustrations accompanying this report. At the time of the accident the weather was cloudy, a light snow was falling and it was thawing slightly.

[¶4]  This derailment occurred on a slightly descending grade near the west end of a 2° curve leading to the south. The rail on the north side of the track was torn up for a distance of about 400 feet, but the rail on the south side remained in position. The track is laid with [Page 4] 80-pound rails 33 feet in length, on hardwood ties, about 18 to the rail. Tie plates are used on the curve.

[¶5]  Engineman Duke stated that his train was running about 35 miles per hour when he noticed the rear driver of his engine drop as if it were passing over a broken rail. He immediately applied the air brakes. The cars broke loose from the tender and rolled down the embankment. The engine and tender were brought to a stop about 700 feet beyond the broken rail.

[¶6]  Fireman Cox stated that he first noticed a jar as if the engine had struck a broken rail, and he was then thrown across the cab to the engineman's side.

[¶7]  Conductor Grant stated that the first indication he had that anything was wrong was when he felt a jar on the car in which he was riding, and immediately afterwards the train left the track. He stated that the speed limit on straight track is 50 miles per hour and at the place where the accident occurred the speed limit is 40 miles per hour. He did not know at what speed they were running when the accident occurred.

[¶8]  This investigation developed the fact that early in February 1912 tie plates had been placed under the rails on each of the ties on this curve for the purpose of strengthening the track and straightening it, the rails having begun to turn and cut into the ties. At the coroner’s inquest the evidence brought out from the men who did this work was to the effect that many of the ties in the vicinity of this accident were in a decayed condition, and that none of them was removed. The foreman in charge of the work stated that they were only in fair condition, while the men under him stated that no shims were used and that many of the tie plates were left loose, although the attention of the different foremen who were present was called to this condition. At the time the work was done and up to the time of the accident the ground was frozen so hard that the ties could not be tamped.

[¶9]  An investigation made by an engineer of the Wabash Railroad after the accident showed that there was irregularity in the surface of the track. This irregularity no doubt contributed to the failure of the rail. Arrangements were made with the Bureau of Standards of the Department of Commerce and Labor for the purpose of having this broken rail examined and the cause of its failure ascertained. This examination was conducted by Mr. James E. Howard, engineer-physicist, and his report covering the results of his examination with the illustrations accompanying the same is attached to and made a part of this report.

[¶10]  The broken rail causing this accident was an 80-pound steel rail made by the Illinois Steel Co. South Works, heat 777,799. It was rolled in June, 1903, but there is no record available as to when it [Page 5] was placed in the track. It was the only rail manufactured by the Illinois Steel Company that was found in the vicinity of the accident, the other rails were manufactured by the Carnegie Steel Company, and according to the railroad record they were laid new in August, 1904. It is probable that the rail which caused the accident was placed in the track at the same time.

[¶11]  After the accident the receiving end of the broken rail was found to be intact for a distance of 27¼[3] inches. This end remained securely fastened to the preceding rail by the splice bars. The leaving end of the fail was intact for 22 feet 9 inches. The intervening section of 6 feet 11¾ inches was broken into six pieces. All the pieces of the broken rail were sent to the Bureau of Standards for examination and test. As shown in illustration No. 5 accompanying Mr. Howard’s report, about 27 per cent of the head of the rail was worn off. The initial rupture evidently occurred under fragment No. 2, there being a seamy streak 7 9/10 inches long on the under surface at the base of the rail at that place, the fracture separating flange fragment No. 2a from a missing fragment of the inner flange. It will be noted that the rupture occurred over a tie. The receiving end of each fragment was bent downward, evidently by the blows of the wheels of the engine and cars as the different portions of the rail were broken, as shown in illustrations Nos. 8 and 9.

[¶12]  The immediate cause of the failure of this rail seems to have been the seaminess of the metal in the base of the rail. In a preceding report [1911-033] on a rail failure on the Great Northern Railroad, at Sharon, N. Dak., this same condition existed. Attention is directed to that part of Mr. Howard’s report calling attention to the initial line of rupture occurring along the seamy streak in the metal, such rail failures being numbered by thousands. Statistics show that a large majority of the broken rails reported are defective in much the same manner as this rail. In the former report mentioned, crescent-shaped flange breaks such as occurred in this rail were given particular attention. Careful track inspection should have disclosed the worn condition of the head of this rail, but the seamy streak in the base of the rail, which no doubt caused its rupture could not have been detected by inspection. This was a defect of manufacture which current specifications and tests are not adequate to discover. In view of the fact that streaks and seams in the metal are common and well-recognized causes of rail failures it would seem that definite action should be taken toward eliminating this source of danger.

Respectfully submitted,

H. W. Belnap
Chief Inspector of Safety Appliances

[Page 6]

REPORT OF THE ENGINEER-PHYSICIST

[¶13]  I have the honor to report upon the examination of a steel rail from the Wabash Railroad, which appeared to have caused the wreck of the Continental Limited, train No. 1, at Redwood curve near West Lebanon, Ind., on March 7, 1912, at about 4:42 p.m.

[¶14]  The wrecked train consisted of locomotive No. 619, Atlantic type engine, and seven cars as follows: An express car, a mail car, a combination baggage and passenger car, a day coach, two Pullman sleepers and a dining car.

[¶15]  The train was running in a westerly direction, when near the westerly end of Redwood Creek curve, a 2° curve, the seven cars of the train were derailed and thrown down an embankment on the north side of the track, the outer side of the curve. The locomotive remained on the track, passed off the curve over a trestle bridge and stopped a short distance beyond. The north rail of the track was torn up for a distance of about 400 feet.

[Page 7]
[¶16]  Two passengers were killed, and 83 passengers and employees were injured[4] . The estimated damage to the equipment was $24,100.

[¶17]  The derailment was caused apparently by the breaking of a rail on the north side of the track and located near the westerly end of the curve. The fractured rail was branded “8001 Illinois Steel Co So Works 1903”. It was an American Society of Civil Engineers section [Page 8] of rail, and originally 80 pounds weight. It was laid in 1904, on white oak ties, 17 to the rail, 33 feet in length. The track at this place had been laid and used without tie plates until the month preceding the derailment. The rails had begun to turn and cut into the ties whereupon the plates were put in, which was done when the roadbed was in a frozen condition. The roadbed continued frozen up to the time of the derailment.

[¶18]  The rail showed considerable flange wear, as indicated by Figure No. 5. It was estimated that 12 per cent of the whole section, or 27 per cent of the metal of the head, had been worn away during its period of service.

[¶19]  Data concerning the locomotive are as follows:

[¶20]  The engineer of the wrecked train testified that he applied brakes and approached Redwood curve under reduced speed, estimated at 30 to 35 miles pr hour. That felt a driver of his engine drop as he passed over the broken rail.

[¶21]  The fireman stated that he also felt the jar of the engine at the broken rail, and that he was thrown across the cab from the left to the right side, that is, toward the outside of the curve.

[Page 9]
[¶22]  A series of photographic cuts shows the appearance of the broken rail as received for examination. Figure No. 6 is a general view of the rail looking upon it from east to west. The several fragments are marked for identification from 1 to 8, inclusive. Piece No. 8 was 22 feet 9 inches long, the others were comparatively short.

[¶23]  In regard to the condition of these fragments, the heads of pieces Nos. 1, 2 and 3 were nearly straight in respect to alignment. The receiving ends of each of the other pieces were bent outward and also downward. Fragments Nos. 6 and 7 were the most distorted, the head of No. 6 being detached from the web. Doubtless fragments Nos. 4, 5, 6 and 7 were in succession detached from the forward part of the rail. Such short bends as these pieces acquired could have been formed only when each was attached to the main part of the rail. It would seem, furthermore, that the wheels which [Page 10] caused the rupture of the rail drifted their farthest over pieces Nos. 6 and 7 and the receiving end of No. 8. Not unlikely there was a buckling of the train resisted at the time these fragments were detached.

[¶24]  This outside rail of the curve during its period of service had been subjected to an outward thrust, which had caused abrasion of the outer flange of the base against the spikes. Figure No. 7 shows two such places. This abrasion, of course, occurred before the tie-plates were used.

[Page 11]

[Page 12]
[¶25]  Figure No. 8 shows fragments Nos. 1 and 4, inclusive, viewed from the gauge side of the rail. The initial fracture is believed to have been the base fracture under piece marked 2. There was a seamy streak at the under surface of the base at this place 7.8 inches long, where the fracture had its origin, separating flange fragment 2a from a missing fragment of the inner flange. This line of rupture occurred over a tie.

[¶26]  The rupture of the rail was completed, the lines of fracture at each end of piece No. 2 extending from the upper surface of the base through the web and then the head. It is probable that these two lines of fracture occurred nearly simultaneously.

[¶27]  The base fracture, between fragments 4a and 4b, appears to have been the next to develop. This line of rupture occurred over a tie and where there was a longitudinal streak or seam 11 inches long which opened but could be traced a further length of 3 inches in the base of piece No. 4, making a total length of 14 inches.

[¶28]  Next followed the upward line of rupture through the web and head between pieces Nos. 3 and 4. These three lines of rupture through the web constitute all of this class of fractures displayed by the rail. The other fractures were formed in quite a different manner.

[¶29]  The running surface of fragments 1 and 3 was decidedly concave, that of No. 2 slightly so. All other pieces were convex on the running surface of the head and with the receiving end of each bent downward.

[¶30]  Pieces Nos. 4, 5, 6 and 7 were detached by fractures having their origins at the lower edge of the head at the gauge side – that is, each of these fragments was broken by a horizontal force or one having a horizontal component.

[¶31]  Figure No. 9 shows fragments Nos. 4 to 7, inclusive, viewed from the gauge side of the rail. The receiving ends of these pieces were bent downward 0.55, 0.48, 1.45, and 0.60 inches respectively. The receiving end of the long piece, marked 8, was bent downward 0.70 inch. The positions of the ties are sketched on cuts 8 and 9.

[¶32]  Figures Nos. 10 and 11 are views of the base of the rail along the ruptured section. The initial line of rupture at 2a and the second base fracture at 4a appear on figure No. 10. The subsequent lines of rupture were of a different order. They are shown on figure No. 11.

[¶33]  A larger view of the middle of the length of piece No. 5 is shown by figure No. 12. The flaky condition of the metal of the running surface of the head is shown. This illustrates the cold-rolling effect of the wheels on the head of the rail.

[¶34]  It will be noted that this rail was very much worn, since some 27 per cent of the metal of the head had been lost by abrasion and wear. Furthermore, that there was evidence in the indented condition of the outer flange of the base that the rail had endured an outward

[Page 13]

[Page 14]

[Page 15]

[Page 16]
thrust repeatedly, which had caused the spikes to wear away and slightly notch the edge of the flange on the outside of the curve. But these indications are not thought relevant to the wreck nor directly contributory to it. The actual, immediate cause of the failure of the rail is believed to be found in the seaminess of the metal, which induced the base fracture under piece No. 2, and the fracture which immediately followed under piece No. 4.

[¶35]  From the evidence presented by the fractured rail it is believed the rupture was caused by the engine of the wrecked train in substantially the following manner.

[¶36]  The weight on the drivers caused the initial line of rupture, a base fracture, along the line of a seam, under piece marked 2. The rail was weakened by reason of the presence of the seam in the base and yielded under a load which structurally sound metal would have sustained without fracture. Base fracture at 4a followed that at 2a and was in turn followed by the fracture through the web and head between pieces Nos. 3 and 4, piece No. 2 having been entirely detached prior to these events.