CPRR.org Header.lbi" -->
By W. H. HINSDALE.
[Reprinted from THE IRON AGE of April 28, 1892, with Emendations.]
In a recent number of The New York Sun there appeared an article having as caption, “The Broken Rail.” The article was principally a report of interviews with prominent railroad managers, and the wrecks due to broken rails were set down by them as in the list of unavoidable accidents.
President Chas. P. Clark of the New Haven Railroad is quoted as saying: “The only railroad accidents that can strictly be called unavoidable are those which result from the fracture of metals. * * * * The accidents which are caused by the breaking of metals furnish the most fruitful subject for discussion under the head of unavoidable accidents. * * * On our road all rails are tested, but some hidden flaw may escape attention, and the rail give out at a critical moment. * * * * The purpose of a railroad official may be all right, but he needs the technical knowledge which enables him to see where opportunities for improvement exist.”
Walter Katte, the chief engineer of the New York Central, said: “We make a point of inspecting the steel rails made for us at the mills. We find their exact chemical composition, and put them through a variety of physical tests before we accept them; even then there are flaws which escape the strictest scrutiny that human foresight can invent. A great strain might come just at the weak spot, and the rail give way,”
President Geo. B. Roberts of the Pennsylvania road, says: “The metal used must be of just the proper chemical composition. If the rail is too soft, it will wear away too fast, but at the same time it must be ductile enough to withstand the repeated impacts of the wheels.”
In each of the interviews referred to, the officials express an honest desire to do the best they can for the safety of the public, to say nothing of the pecuniary aspect in the interest of their respective roads. There is, however, among railroad officials, a great misapprehension as to the varying qualities of steel, and the slight causes during the course of manufacture that will seriously affect the final result.
When the gentlemen referred to above say that they inspect and test all their rails, they do not really mean anything of the kind. They do not mean that each individual rail is subjected to a rigid physical test, but that each individual rail has been merely looked at and that comparatively few of them have been subjected to any other test whatever, and if the finished rail has a presentable exterior, it is passed regardless of the hidden flaws it may contain. I do not contend that more than this can be done under the methods of inspection generally pursued by railroad officials, but that the inspection should begin at an earlier stage of the manufacture, where the flaws resulting from the methods of manufacture may be easily discerned.
As described by President Roberts of the Pennsylvania road, a proper
degree of ductility, combined with suitable hardness, are the general requirements
as to quality of the steel. These cover tensile and compressive resistance,
and excessive loss by abrasion, and these qualities being assured, if the
rail is absolutely sound, and of correct section, it would seem that little
more can be required of the manufacturer. But the question of soundness
in the finished rail not being easily determined, without spoiling it as
a rail, the necessity for an improved method of inspection on behalf of
the railroads is obvious, and a practical knowledge of the art of manufacturing
steel would undoubtedly be desirable on the part of the inspectors.
Fig. 1 represents the general appearance of a rail ingot when removed from the mold. All ingots of steel are subject to a variety of defects, one of two conditions and sometimes both may always be found in every ingot as ordinarily cast. Either a shrinkage cavity or “pipe” at the top, extending downward in its vertical axis from 15 to 25 per cent. of the total length of the ingot, or in place of this a spongy texture indicative of a poorer quality of steel, although made from equally good materials.
To go into details of the causes which produce such different results is beyond the purpose of this paper, although they are generally fairly understood by the skillful steel maker.
The sound portion of the piped ingot is generally free from seams and
cracks and produces the best results in manufacture, while the ingot of
spongy texture, if high enough in carbon for rails, is quite certain to
develop seams and cracks in working and to be of doubtful utility when
finished, and in the higher grades of steel is always rejected for first
Figure 2 is reproduced from a photograph of an ingot of high-grade tool
steel laid open through its vertical axis, showing the extent of the pipe
in that quality of steel when properly melted.
Figure 3 is from a photograph of an ingot of the same quality of tool steel material as Figure 2, but badly melted, and would be considered absolutely valueless except for scrap. If worked the product would be seamy and subject to vertical cracks, sometimes extending from side to side through the bar under the stress of working at the hammer.
A too hasty exposure to the air by stripping the mold from the ingot while its interior is still at its fullest, expansive heat, often induces exterior cracks both transverse and longitudinal which never weld in the subsequent working, and always result in great injury to the product.
President Roberts says that “the metal must be of just the proper chemical composition,” and yet rails having “just the proper chemical composition” have proven so brittle that they would not bear unloading in the usual manner from the cars, but would break like pipe stems. A lot of about 3,000 tons was made not much more than two years ago, which were as above described, and all were condemned and returned to the manufacturer. In this case the analyses of the iron as well as the steel were all that could be desired.
So eminent authority as R. W. Hunt, in a paper read before the American Institute of Mining Engineers, at Buffalo, in 1888, used the following language: “My investigations of the service of thousands of tons of rails, and the analysis of many hundreds of them, have shown the greatest variation in the wear of rails of the same section and chemical composition. This being so, there must be some physical cause. Can we find a chemical reason for rails showing ‘soft’ in wear, having the following composition?
Manganese ............... .722
“If so, why did another make, in the same track and under seemingly the same conditions, analyzing as follows, wear hard?
Sulphur ..................... .064
“I could multiply these instances to an indefinite extent. Our Bessemer friends are all right on their chemistry, and it is not in that direction that investigation is needed.”
Such incidents should serve to set aside as worthless the knowledge of the “exact chemical composition” of the rails to the railroad manager and should relegate the whole matter to some person in the interest of the railroads capable of knowing how to treat ordinary Bessemer iron, and this conclusion should be further apparent to railroad managers when they consider the very satisfactory analyses their own laboratories make from rails broken in service.
A carbon analysis from each heat, representing one from each 60 rails; a manganese determination from each tenth heat, say one from each 600 rails, and a quantitative analysis from each fiftieth heat, say a full analysis of 60 out of every 3,000 rails, measure the very dull light thrown on the chemical characteristics of the output of our large rail mills.
When nature formed the laws that govern making steel, she gave them sufficient elasticity to stretch over a wide range of so-called “practice,” with the result that between a bloom that would make a first-class rail and one that would just pull through, there is a broad field for neglect of proper manipulation and sacrifices of quality for “large output”; and yet all heats that make first-class rails are as easily and quickly made as the poorer kind, if the “technique” of the establishment is of the masterly kind that such immense operations justify in obtaining. In this connection the statement of President Clark, “that the purpose of a railroad official may be all right, but he needs the technical knowledge which enables him to see where opportunities for improvement exist,” may be criticized; for, as in all other important affairs, President Clark should employ such technical knowledge to properly inspect his rails.
A knowledge that only comes by long practice, coupled with sufficient intelligence to comprehend the phenomena of manufacture which are constantly passing under the observation of the Bessemer man, are the qualifications required to protect, in some measure more than at present, the railroad companies against rails of doubtful qualities, and such persons will be employed just as soon as railroad managers acquire enough “technical knowledge” to see the importance of it.
R. W. Hunt, in the paper before referred to, used the following language: “The mere replacing of a broken rail with a whole one may fall far short of the damage sustained by the road on whose track the accident has occurred, leaving the danger to human life out of consideration. Therefore the purchaser has a right to insist upon some precautions being taken to avoid as far as possible such disasters. And these precautions, if correct ones, are also in the interest of the makers. Steel rails are made very rapidly, and the demands of the trade necessitate that they shall be made very cheaply. The workmen are paid by the piece, and while generally making good wages they must produce a large tonnage to realize them. No matter how desirous the general management may be of producing only good work, it is very necessary that safeguards should be provided.”
No rail-makers should object to a respectful supervision of their metallurgical operations by a well-qualified representative of a leading railroad, nor hesitate to correct indifferent practice in their works when such qualified person could show that it would be beneficial, and such inspection and suggestion need not embarrass operations nor increase cost. A little sound knowledge, at the right moment in the right place, is just what such inspection would comprehend.
In the interview with Walter Katte of the New York Central road, he says: “We find the exact chemical composition of the rails.” As will have been observed from the foregoing, it does not follow that this supposed knowledge gives the railroad official any definite information as to the quality of his rails. Usually the proportion of carbon will give him a general idea as to their hardness, but “the exact chemical combinations” should be left to a competent inspector and the manufacturer, and with more attention paid to the physical conditions developed during the working of the steel, he will soon learn that there is as much difference between the product of a sound bloom and one that is cracked and full of seams as there is between clear and straight-grained oak and knotty and twisted hemlock. To be sure, both blooms are steel, and so are both the others timber, and the comparison is fair.
Undoubtedly the best point of observation for determining the quality of the future rail is at the blooming mill, and it does not require a technical knowledge of steel manufacture to see the external defects of the future rail fully exhibited in the bloom. Specifications should therefore include inspection of the blooms, and the rejection of such as show defects in rolling. It is not unusual to see blooms develop in rolling many irregular transverse surface cracks, while under the powerful action of the finishing rolls these cracks are reduced in size with the reduction of the bloom, and by the oxidizing effect of the air are obscured from ordinary observation, but they still exist in the rail in no less proportion than they existed before in the bloom, and the integrity of the rail is thereby impaired.
The remedy for this condition of the bloom rests with the manufacturer, and as President Clark intimates, the railroad manager may not have the technical knowledge required to suggest the remedy, he may reasonably content himself by insisting that a remedy shall be applied. The transverse cracks occur at various points along the surface, and often extend an inch or more beneath. Exposure to the air quickly produces an oxidized surface or scale covering the walls of the cracks, which effectually prevents their welding in the subsequent working, and although attenuated in the process of reduction in the finishing rolls, the cracks before plainly visible in the bloom are still present in the rail, and these incipient breaks invite disaster under any unusual stress, and not infrequently these cracks are enlarged to the point of failure by the vibrations due to passing trains, It is true that in some mills the worst blooms are subjected to a process of trimming which consists in gouging out a portion of steel on each side of some of the larger cracks, but this is only a partial remedy at best, and if not properly done often induces other faults in the surface of the rail.
It is therefore a reasonable conclusion that it is hazardous to use a bloom that requires this surface trimming, as it indicates a general condition of the steel which unfits it for No. 1 rails, and all such blooms should be rejected.
The manufacturer will say that this trimming process is performed to improve the quality of his “seconds,” but that large numbers of such blooms are passed by the inspectors as No. 1 rails will be only too obvious by an hour of observation at the blooming mill. Some portion of these defects may be due to lack of care in blowing the heats, and a too hasty manipulation in not permitting the chemical reaction to take place in the converter, but continuing it in the molds after pouring.
With some manufacturers this slow reaction within the mold is purposely induced, to the palpable in jury of the product, by reducing the percentage of manganese for the purpose of reducing the extent of the pipe, This is a prolific cause of cracks in the blooms.
The reduction in percentage of manganese and the consequent slow reaction
in the mold also frequently causes the molten steel to rise lo or 12 inches
in the mold during its cooling, and results in a spongy texture of the
ingot, with a shrinkage cavity commencing at about one-fifth the length
of the ingot from the top. As the shrinkage takes place and the cavity
enlarges a vacuum therein would be the result, but that soon the pressure
of the atmosphere ruptures the soft spongy steel, and the air passing into
the cavity causes the formation of scale on the walls, which prevents their
welding together in the rolling, and thus the phenomena of piped rails
having apparently sound ends result.
Fig. 4 was reproduced from a photograph of an ingot broken to show the
conditions above described. Fig. 5 is from a photograph of a broken ingot
showing no sponginess at the top, but with a sound surface, and by removal
of the piped portion would have produced rails of an excellent quality.
Fig. 6 shows a section of rail long in use, made from the sound portion
of an ingot. Fig. 7 was taken from a section of rail made from the upper
or piped end of an ingot, and shows that the use of such is hazardous.
Fig. 8 was taken from a section of rail which broke in service, causing a wreck. No. 9 illustrates the appearance of a bloom having a perfect exterior. Nos. 10, 11 and 12 illustrate the appearance of an imperfect bloom before and after the trimming process above mentioned.
The United States Government specifications require 30 per cent. of every ingot destined for Government work to be rejected and removed from the top, and the railroad managers of to-day should not be satisfied with a less careful inspection of such important materials as their rails. Hunt's specifications for rails require 12 per cent. to be removed from the top of the ingot, but this was doubtless a compromise, as Mr. Hunt and the manufacturers understand that 20 per cent. would be nearer the safety mark. Mr. Hunt's moderate requirements not being complied with, anything that will pass the cursory examination of the inspector, after the rail is finished, is accepted as sound, till the very practical test in the track brands it otherwise, and a “mysterious dispensation of Providence” follows, such as occurred recently at Lima, Ohio. The evil is one of growing importance in these days of hundred-ton locomotives, and heavy cars running at a speed of more than a mile a minute.
One of the unpleasant things to contemplate in this connection is that there are many such rails still being distributed throughout the railroads of the United States, to be heard from day after day as the accounts of broken-rail wrecks are reported. Yesterday it was Findlay, Ohio, with a record of two lives lost and many injured. November 11th, 1891, the defective rail was found at Hornellsville, New York—vestibule train, No. 8, the finest train that runs over the Erie road; the engineer was killed and a dozen persons hurt. January 15th, 1892, the defective rail was found at Jonesville, Minnesota; two passengers killed outright and twenty-three injured. January 1st, Hamilton, Minnesota, reported a defective rail broken; result, four sleeping cars “ditched” and piled up in confusion. And the same day at DeSoto. Mo., a train wrecked by a defective rail, a sleeping car turned over on its side and five persons injured. Beloit and Urbana, Ohio, each recently furnished on the same date trains wrecked by defective rails. February 18th New York reported a wreck from a defective rail in the tunnel, when fortunately no lives were lost. A more recent broken-rail wreck occurred in the yard near the Grand Central station.
The foregoing includes but a portion of the disasters from this cause alone within the limits of four consecutive months.
Heavier rails are being rapidly substituted for the old and lighter models, but this will not prove to be a sufficient safeguard against breakage unless the quality shall be improved, and while this substitution is going on, it will be a great pity and source of lasting evil if the old methods of determining their quality is not improved upon, when we consider that the size of the defects will continue to be proportion to size of the rails.
The new rapid process of rolling and finishing direct without reheating the blooms is quite likely to increase the existing evils on account of the lower and uneven temperatures at which the blooms will arrive at the finishing rolls.
“They are getting as good as they pay for” has been heard more than once from the lips of a manufacturer, while admitting that he could make better rails, and this remark, doubtless, indicates the direction competition has turned the thoughts of manufacturers, but inspectors with technical knowledge and long experience in the manufacture of rails can fairly arbitrate matters of divided responsibility between the railroads and the rail makers as to the quality of rails that shall be laid in American tracks, and when railroad officials confront the long list of killed and maimed in broken-rail wrecks, they must feel a personal responsibility if they neglect any safeguards.
Broken-rail wrecks should not carry with them the verdict of “Nobody to blame,” for the contention cannot be sustained that all such are unavoidable.
W. R. HINSDALE,
800 Broad Street, Newark, N. J.
Courtesy Bruce C. Cooper Collection.