DEVELOPMENT OF RAIL TESTING

From the earliest days of railroading, maintenance of the right-of-way has been one of the greatest challenges of the operating companies. The greatest single factor of this problem is the prevention of service failures of rail in track.

The causes of rail failures are many. Often the effects are serious - destruction of property and the injury or death of crews and passengers. The early iron strap rail frequently broke loose from its wooden stringers and curled up in "snakeheads" which tore through the wooden floors of the old cars. Cast iron rails crystallized and cracked under the strain of severe climatic changes or sudden shock loads. Bessemer steel rail also showed a tendency toward brittleness in cold weather. The present day steel rails manufactured by the basic oxygen and/or electric furnace process, although vastly superior to the older types of rail in both strength and wearing quality, still develop defects. Under today's heavy loads and high speeds these defects will cause rails to fail in service unless inspected regularly.

Many external defects and visible indications of internal defects are overlooked in visual inspections. With a Sands mirror an inspector can visually inspect approximately one mile of rail per day. However, even this type of inspection, acknowledged as the most efficient method of visual inspection, cannot detect those internal separations of the steel within the rail head known as Transverse Defects, and frequently overlooks vertical split heads.

The transverse fissure first came into prominence as an outstanding cause of rail failures as a result of a derailment at Manchester, N.Y. in 1911 in which 29 persons were killed and over 60 seriously injured. In the investigation following the accident, Dr. Howard of the U.S. Bureau of Safety identified a broken rail as the cause of the derailment. A study of the rail revealed a defect which was entirely internal (which Dr. Howard termed "transverse fissure") and which was definitely established as the cause of the rail failure. A number of other railroads began private investigations to determine the prevalence of transverse fissures in their rails. The results of these investigations showed that transverse fissures were wide-spread.

As early as 1877, a patent, No. 189,858 "Mode of Detecting Defects in Railroad Rails", showed the effort to detect defects. This invention called for the energizing of the rail with flux, then picking up variations in residual magnetism.

The railroads of the country requested the U.S. Bureau of Standards in 1912 to undertake a thorough investigation of the prevalence and cause of transverse fissures and to aid in the development of a method of inspection which would accurately locate and measure the size of such hidden defects in rails.

In 1915 the Bureau of Standards began a series of experiments culminating in the development of magnetic testing equipment for locating and measuring transverse fissures in rails. In this method a magnetizing solenoid was passed along the rail setting up a flux. Any leakage of the flux was detected with searching coils which were attached to a sensitive voltmeter. The principle of this apparatus was sound in theory and successful in laboratory tests but in actual tests on track it was found that the equipment was unable to differentiate between actual defects in rails and the strains caused by slipping wheels, surface irregularities, cold working by car wheels, etc. Two railroads and one of the major steel companies made further unsuccessful attempts to adapt this method to field testing.

In 1923 Dr. Elmer A. Sperry, a noted inventor and founder of the various Sperry enterprises, started to develop and build an inspection car that would detect the presence of transverse fissures in the rails while traveling along the track. While Dr. Sperry was engaged in this work, another serious derailment near Victoria, Mississippi , on October 27, 1925 caused the death of 21 persons and the injury of over 100 passengers. Again, the cause of the accident was found to be a broken rail caused by a transverse fissure.

Dr. Sperry contracted with the American Railway Association in August 1927 to build a detector car for the AAR and, in addition, to supply a rail testing service to the railroads. This car energized the rail with current and measured variations in potential drop by means of a pair of contacts. While the laboratory tests were satisfactory, the actual conditions in track prevented satisfactory operation. The searching unit, located between the brushes, depended upon contact with the rail. The average conditions of the surface of the rail, - dirt, oxide, and scale on the rail head, prevented continuous contact of the searching units with the rail and caused many false indications. Extensive research was continued to develop a practical means of cleaning the rail before testing, but no solution was found and the method was abandoned.

Far from being discouraged by this failure, Dr. Sperry proceeded, as in the case of so many of his other inventions, with the development of an entirely new principle for the detection of hidden defects in rails, - the induction method of testing.

In the induction method, a heavy electric current of low voltage was passed through the rail, setting up a magnetic field around the rail. A pair of searching coils was suspended at a constant distance above (but not in contact with) the surface of the rail to detect any deflection or variation in the magnetic field caused by fissures within the rail. When these coils passed into the changing magnetic field around the defect a current would be induced within the coils. This current was then amplified to actuate a series of recording pens. The induction principle, greatly perfected and refined, is still the basic principle of Sperry Detector Cars.

Sperry Rail Service then began an extensive long-term program of providing a reliable testing service for all railroads. Sperry Rail Service was organized to fill the great need for efficient equipment for testing rails in tracks. To overcome the limitations of the early detector car required extensive laboratory research and development combined with practical experience in testing rail in track. Any such long term program could best be carried out by an organization which leased its services and equipment to the railroads rather than sold its detector cars outright. There would be little lag between the development of new testing techniques and their application in rail testing. This policy, established in 1928, has worked out to the definite advantage of the railroads served by Sperry since it has expedited Sperry's continuing program of development. Statistics published in the Sperry Railer show the remarkable progress made by Sperry over this period of years and confirm the advantage of the railroads of such a policy.

In recent years, this policy has resulted in the addition of ultrasonic detection equipment to the Sperry fleet of detector cars to further improve the efficiency of Sperry Detector Cars. Ultrasonic rail testing was first offered in 1949 with the equipment mounted on a motor car and a hand inspection made at each joint. Within eleven years the Ultrasonic inspection equipment was automated to the point where it could effectively assist the induction systems already in operation. Ultrasonic equipment for the detection of defects in the rail head was added to the detector cars in 1960 and ultrasonic equipment for the detection of bolt hole and web defects was added in 1961. A specially designed all-ultrasonic detector car was first put into operation by Sperry on October 8, 1959, on the New York City subway system.

Although the Sperry detection equipment was originally designed to detect the transverse fissure, it also detects a large number of other types of defects which render rail unsafe. Such defects, along with the transverse fissure, are fully described in this manual.

In the course of these early investigations of the cause, occurrence, and detection of transverse fissures, the inspection methods employed at the rolling mills came under the scrutiny of the AAR and the steel companies, starting in 1930. Rails were found to be rejected at the mill for serious surface defects, but internal structural weaknesses in the steel, which were potential defects, escaped detection.

When these potentially defective rails are placed in service, dangerous defects often develop under traffic. Production factors which contribute to the manufacture of inferior rails have been minimized through refinement in mill practice, but invisible internal defects still occur in the latest types of rail.

At the present time there are millions of rails in tracks which were rolled prior to the development of the more modern production techniques. In both new and old rail the extent of the growth of microscopic irregularities is still of great concern. The location and size of these irregularities in a rail determines the capacity of that rail to withstand the stresses imposed upon it in service.

TYPES OF DEFECTS IN TRACK TODAY

An attempt to solve the transverse fissure problem by elimination of the prime cause (shatter cracks or hydrogen flakes) resulted in the development of the "control-cooled" process of rail manufacturers and led to the general adoption of the process in 1936 - 1938 by American steel mills. The transverse fissure problem was essentially solved by the control cooling processes and the vacuum degassing process subsequently used by some manufacturers in recent years. However, isolated hydrogen problems resulting in the occasional development of true transverse fissures have been known to occur even into the 1990's.

While the initial goal of eliminating the true transverse fissure by means of mill controls has been successful to the extent of reducing the number of such defects to an insignificant minimum, these methods of rail manufacture have not been successful (nor was it originally intended to do so) in eliminating or reducing the incidence of rail defects other than the transverse fissure. Efforts are currently being made by the manufacturers of steel rails to further reduce the incidence of internal rail defects by improving the "cleanliness" of the rail steel. While it is too early to ascertain the total effect of these new processes, there is evidence that some improvements have been made.

Other transverse defects which are still found in all rail in significant numbers are detail fractures, engine burn fractures, and occasional compound fissures. Progressive transverse defects will also develop underneath engine burns or rail welds that have been resurfaced by welding when the area being rebuilt is not properly cleaned or cooled during the process.

Longitudinal defects such as vertical split heads, horizontal split heads, head and web separations, split webs and piped webs are still discovered in considerable numbers. Rail that has been re-laid or otherwise disturbed after years of service in one position is particularly susceptible to the development of longitudinal defects due to altered stresses and changed loading patterns. In some cases, it has been noted that rail loading changes which result from contour grinding have aggravated certain dormant rail conditions such as internal head stringers or inclusions, causing rapid and unpredictable progression of longitudinal rail defects.

With more continuous welded rail being laid each year, defects at rail welds, both factory welds and thermit process field welds, are becoming more frequent. Rail weld defects usually progress from inclusions or slag entrapments at the weld interface although some longitudinal weld defects in the mid-web area may result from residual stresses in the weld area or stress risers resulting from improper removal of upset metal. In the mid 1960's approximately 10% of the mileage tested by Sperry Rail Service was continuous welded rail. By the mid 1970's approximately 25% of the mileage tested was CWR and by the mid 1990's approximately 77% of the mileage tested was CWR.

Total rail joint defects, both bolt hole breaks and rail end head and web separations, are of course less numerous than in previous years due to increasing amounts of welded rail in track. However, even in welded rail, insulated joints, heel joints and other special works are still susceptible to joint failures. The density of rail joint failures in typical bolted rail remains the same as in previous years.

 Learn About Sperry Rail