were completed March 31, 1931. The tests were made with trains of varying lengths up to 150 cars.

Following these investigations the director of research prepared a complete and comprehensive report, including the details of each test made. This report was contained in 60 large volumes of texts and charts and was carefully studied and analyzed by the committee on brakes and brake equipment of the association. The information obtained from these tests was used in the development of specifications for power brakes for freight cars which are now standard of the association.

Following these tests the air-brake companies proceeded with the development of a brake in commercial form that would embody all of the functions found desirable for a freight-car brake. This work resulted in the production of the freight air brake meeting the specifications adopted by the association, which is designated as the AB brake. These AB brakes were then submitted to further road tests and found to be acceptable and to meet the specifications, after which rules were enacted by the American Railway Association requiring all new cars built after September 1, 1933, to be equipped with this type of air brake. Subsequently, rules were adopted requiring this type of air brake to be applied to all cars rebuilt on or after August 1, 1937, and to be applied to all freight cars in interchange service by January 1, 1945.

Due to the depression, and later to the war effort, it has not been possible to meet this goal. Approximately 50 percent of all interchange freight cars will be equipped with such improved brakes by January 1, 1945, and cars are currently being equipped as fast as manufacturers are able to obtains materials for their manufacture.

Research and investigation in connection with power brakes for railway equipment, however, is an ever-continuing subject. The air-brake companies, in cooperation with the railroads, have developed air-brake equipment for the highspeed streamlined passenger trains, and improvements are being continually studied.

Coupler research.

The question of an automatic means of coupling freight cars was one of the earliest problems confronting railroads, both from the standpoint of safety and from the standpoint of economical and efficient operation.

In the early days of railroading cars were small and light in weight and trains were short, and a simple link and pin was used to couple cars. This required the men to go between cars to couple and uncouple them, and resulted in many accidents and injuries to persons, and was slow. From the very first, inventors began to devise means to make coupling automatic. In 1885 the Master Car Builders' Association authorized a public trial of the devices then available. At that time there were in existence some 3,100 patents covering such devices.. This trial was held in September 1855. at Buffalo, N. Y., 42 couplers being tested. This represented but a very small portion of the couplers considered, as those devoid of merit were not admitted to test. As a result of these tests, in 1887 one type of coupler was adopted as standard. One of the requirements was that the link and pin could also be coupled with the automatic coupler. The type selected was the present type of coupler with swinging knuckle; the knuckle was slotted to permit coupling with the existing link-and-pin couplers.

As cars were becoming heavier and trains longer, the strength of the couplers soon became an important factor. Further elaborate tests were made in 1892 to develop standards and limits, at the Watertown Arsenal and at the Pennsylvania Railroad testing laboratory at Altoona, Pa. Drop tests were made at Altoona and pulling tests at Watertown Arsenal. In 1894 a much heavier droptest machine was constructed and a further series of elaborate tests conducted at Altoona, Pa.

Tests and research were continued until 1899 when the committee on Master Car Builders couplers presented its complete report covering specifications for automatic couplers, designs, contour, gages to insure interchangeability, test, and service gages. The committee also recommended a standard drop test machine and code of tests. These recommendations were adopted by the railroads as recommended practice. The committee continued its tests and research and specifications were further revised and advanced to standard in 1905.

The coupler was the subject of continual research and tests and specifications were revised from time to time. In 1911 it was decided that instead of having a number of couplers that would interchange one with another, but requiring the carrying in stock of a multitude of different repair parts, there should be but one

standard coupler with all of its parts standardized and interchangeable. The committee on couplers of the Master Car Builders' Association was instructed to design such a coupler and was authorized to invite the coupler manufacturers to join with them in this work.

The general conditions and difficulties existing with couplers in general service at that time were investigated on many of the railroads throughout the United States, followed by static and dynamic tests on those designs of couplers to ascertain their strength. This resulted in formulating specifications on design, operation, and strength which would be desirable in the standard coupler. These tests were conducted throughout the years 1911 to 1916, using the testing facilities of the Pennsylvania Railroad, Baltimore & Ohio Railroad, National Malleable & Steel Castings Co., American Steel Foundries, Baldwin Locomotive Works, and the Pressed Steel Car Co. to facilitate the work. Coincident with the laboratory tests to improve and reinforce couplers, road tests were conducted. The couplers were applied to locomotive tenders, and to a lot of 50-ton coal cars operating between the coal mines and tidewater where they received extremely hard service and where they could be under constant observation.

As a result of this investigation the type D coupler was developed and adopted as standard for all American railways in 1918. The service of the type D coupler was watched closely and all defects occuring were studied and as a result the type E coupler was developed. This coupler was given extensive laboratory tests by the committee on couplers and draft gears in cooperation with the coupler manufacturers. Specifications for the type coupler were adopted in 1930 as recommended practice and advanced to standard in 1931.

Through all of these developments the various types of couplers have been kept interchangeable so they would couple one with another and, in the case of the type D coupler, the parts are interchangeable with the parts of the type E coupler.

With the construction of high-speed, light-weight passenger equipment, and the desire to eliminate shocks, jerks, noise, etc., that occur in handling passenger trains, attributable to slack in contour of couplers, there developed a demand for a tight-lock coupler, that, is a coupler in which there would be no slack action in the contour lines. To meet these demands, the committee on couplers and draft gears requested the coupler manufacturers to study the problem and develop a coupler that would adequately meet these requirements. In response to this, there were a number of individual designs developed by the coupler manufacturers. On account of the additional expense of developing and maintaining several designs, the manufacturers were requested to pool their individual ideas and develop a single standard that would adequately meet the requirements for such a coupler.

The general prerequisites of the design were that the tight-lock coupler should be free from slack, capable of coupling with existing American Railway Association standard couplers and equivalent in strength to the standard E coupler. Coupler manufacturers worked diligently on this problem and developed a coupler that conformed with these prescribed requirements. Following extensive laboratory and road tests, the tight-lock coupler was adopted as recommended practice in 1931 and, following further development, was adopted as standard in 1937.

The service of these couplers is being constantly followed and improvements are adopted as they are developed. The tight-lock coupler also serves as an efficient antitelescoping device.

Draft-gear research.

One of the first problems to confront the railroads was the absorption of shocks between cars in road and switching service. This problem was actively before the Master Car Builders' Association upon its organization in 1867 and committees have been constantly studying the problem since that time.

When cars were only 10 to 20 tons capacity and trains were short, a fairly satisfactory arrangement was worked out using common helical springs behind the coupler and a buffer in front of the coupler. This subject has been one of continuous research from the very first and the records of the former Master Car Builders' and American Railway Master Mechanics' Associations and the present Mechanical Division of the Association of American Railroads, contain numerous reports of research and test.

The first friction-draft gear was invented by the late George Westinghouse and the first sets were placed in service about 1897. Extensive laboratory and service tests of this device were made in 1898 by the committee on improved

freight car buffers of the Master Car Builders' Association. During the years 1900 to 1902, inclusive, extensive draft-gear tests were conducted, including both road and laboratory tests. The laboratory tests were conducted in the laboratory of the Pennsylvania Railroad at Altoona, Pa., and at Purdue University. Manufacturers cooperated in these tests. Tests were continued throughout the years, together with experimental work carried on by the individual railroads and the manufacturers, with improved draft devices, both of the friction and heavy spring types. During this development of draft gears, each of the draft-gear manufacturers built draft-gear laboratories and conducted a large amount of draft-gear research, the results of which were made available to the association committee.

In connection with this improved testing and research, slow-motion pictures were used as they became available.

During the period of Government operation of railroads in 1918 to 1920, the committee on couplers and draft gears cooperated with the Inspection and Test Section of the United States Railroad Administration in a series of draft-gear tests, including (a) 9,000-pound drop tests using a solid anvil, (b) static test, (c) car-impact tests. The car-impact tests were made at the gravity testing plant of the T. H. Symington Co. at Rochester, N. Y.

In December 1924 the committee recommended designing, building, and housing a 27,000-pound drop-test machine for testing of draft gears, together with the necessary recording apparatus. From the information to be obtained with this testing equipment, it was proposed to prepare suitable specifications for the purpose of eventually restricting the use of draft gears to those that are known to meet the prescribed standard of efficiency and, through the experiments to be conducted, obtain such facts and information as will assist in the development of draft gears generally. This recommendation was approved and arrangements made to purchase testing machine and devices and a laboratory was built at Purdue University, La Fayette, Ind. The laboratory building was completed and the testing machinery installed early in 1927.

Arrangements were made to test all available draft gears, the tests to be under the direct supervision of the dean of engineering of the university and under the general supervision of the committee on couplers and draft gears.

Ten manufacturers submitted their gears to test. The tests started in 1927 and were concluded in 1929. Complete report showing the result of these tests was published by the committee, May 1929.

The results of these tests were carefully studied by the committee, the manufacturers, and the railroads, and specifications for approved draft gears for freight service were prepared and adopted in 1931.

In accordance with the terms of these specifications and rules which have been adopted by the railroads only approved draft gears may be applied new after January 1, 1934. Arrangements are made whereby the manufacturer must submit his device for approval test at the laboratory at Purdue University and receive certificate of approval if the gear meets the requirements of the standard specifications. Throughout the years 1930 and 1931, the laboratory was used by the manufacturers in development and research work. During 1932, the laboratory was busy making approval tests of draft gears in accordance with the specifications. The laboratory has also been used to develop improvement in draft gears and revisions of the specifications for draft gears.

From September 11, 1930, to December 30, 1930, an elaborate series of impact tests were conducted by the committee on couplers and draft gears on the car impact test plant of the Symington Co. at Depew, N. Y., to determine the effect of recoil in draft gears.

The draft gear testing laboratory at Purdue University is continuously being used in connection with development work of the manufacturers and conducting of approval tests, check tests, etc., by the association under the direction of the committee on couplers and draft gears. An appropriation is made by the Association of American Railroads annually to carry on this work.

During the past few years, approved and certified draft gears have been removed after approximately 5 years' service and subjected to laboratory tests to determine if the gears as sold to the railroads conformed to the capacity and characteristics of the gears as approved and certified under the specifications of the association. Also, as improvements in existing draft gears are developed, the improved gears are submitted for test at the laboratory of the association and, where approved, revised certificates of approval are granted.

This is an ever-continuing activity and the laboratory is practically continually in use, either making tests for the association or carrying on development work for the draft-gear manufacturers.

In addition, this same laboratory is used for conducting other tests requiring drop testing. The drop-testing machine installed in this draft-gear laboratory is equipped with two falling weights, one weighing 27,000 pounds and the other 9,000 pounds. Tests requiring the use of this equipment, other than draft gears, are also conducted at this laboratory, including closure test of helical springs, etc.

The equipment for testing hand brakes for railway cars is also located at this same laboratory.

Brake shoes.

The present specifications for brake shoes are the result of long and exhaustive tests. The results of early tests were reviewed by a special committee of the former Master Car Builders' Association in 1878. Other tests were made from time to time between 1889 and 1893. In 1891 and 1892 some valuable and reliable experiments were conducted by the Pennsylvania Railroad. In 1893 the former Master Car Builders' Association appointed a special committee on laboratory tests of brake shoes. This committee, in cooperation with the engineers of the Westinghouse Air Brake Co., designed and had built a testing machine and recording instruments. The instrument companies and brake-shoe manufacturers also cooperated fully. This machine was first installed at the plant of the Westinghouse Air Brake Co. at Wilmerding, Pa., and was transferred to Purdue University, Lafayette, Ind., in 1898. The machine was improved in 1906. Elaborate tests were made on this machine to determine (a) wearing qualities of various types of brake shoes (b) effect of the shoe upon the wheel (c) coefficient of friction with both cast iron and steel wheels. These tests were completed in 1910 and the standard specifications for brake shoes revised based on the information secured. At the present time all new types of brake shoes must be tested by the association before their use is authorized. One or more of the brake-shoe manufacturers have built similar test machines and tests of brake shoes of varying composition are being made frequently. Lubrication, including design of boxes, wedges, bearings, and methods.

Studies in car lubrication are continually being made for the purpose of reducing costs and improving service by preventing delays to cars and trains caused by lubrication failures. Many of these studies are in progress at the present time and include experiments with new methods of lubrication and new designs of journal boxes, journal bearings, wedges, etc.

In 1942, in connection with the effort to conserve scarce metals required in the manufacture of railway car journal bearings, the association instituted a project of journal-bearing research. This project is carried on under the direction of the mechanical engineer and a subcommittee consisting of representatives from the committee on specifications for materials and the committee on lubrication of cars and locomotives.

For the purpose of this investigation, arrangements were made to utilize the laboratory of the Railway Service & Supply Corporation, at Indianapolis, Ind., the association reimbursing that company for its out-of-pocket expense in connection with the use of its facilities.

As a result of this work, and for the war emergency, it was found that considerable changes could be made in the design of journal bearings resulting in a large saving in copper and tin, critical metals used in their manufacture. The work performed at the laboratory is supplemented by carefully observed roadservice tests.

In addition to the work carried on at the laboratory, investigations were being continuously conducted on the roads represented on the committee in charge of these tests and in these roads' laboratories. The results of these supplementary tests were made available to the members of the association through published reports.

Another project of importance is the reclamation and renovating of lubricating oil and waste. The majority of railroads have established central plants employing the latest methods for the reclamation of lubricating oil and waste packing or have contracted this service with one or more companies devoting all of their activities to this purpose. As a result of this research, lubrication failures of freight cars have been reduced to a very low point and car-miles per failure have increased to the greatest on record.

The general committee of the mechanical division, American Association of Railroads, has authorized a study of lubricants for roller bearings, the laboratory work to be conducted at the laboratory of the Railway Service & Supply Corporation at Indianapolis, Ind., where all conditions encountered in North

America as to temperature can be duplicated. This research study will be conducted under the direction of the mechanical engineer of the division and the committee on lubrication of cars and locomotives.

Reclamation.

Continual progress is being made and experiments and tests are continually being conducted in the reclamation and reuse of many important details of cars and locomotives, including couplers, coupler knuckles, coupler yokes, cast-steel truck side frames, springs, etc. Through this work large savings in railroad operation are made possible.

During the year 1943 an investigation was conducted of reclaimed car parts, including couplers, yokes, etc., reclaimed by the fusion-welding process. These tests were under the direction of a special subcommittee appointed for this purpose. The parts to be tested were reclaimed at reclamation plants of representative railroads and tested at the laboratory of the National Malleable & Steel Castings Co. at Sharon, Pa.

These tests were for the purpose of considering revisions of interchange rule 23 covering permissible welding of car parts and to conserve critical materials needed for the war effort.

Trucks and truck sides.

In 1912 the railroads, through the former Master Car Builders' Association, adopted specifications for cast-steel truck side frames. These specifications were revised in 1914, 1915, and 1919 as the result of research and study. These specifications were last revised in 1929.

These specifications, as they stand at present, are the result of a very extensive research program carried out by the committee on car construction during the years 1925, 1926, and 1927. All types of side frames were tested under the direction of a special subcommittee as follows:

(a) On the dynamic testing machine designed and built for the T. H. Symington Co. at their laboratory, Baltimore, Md.

(b) On the dynamic testing machine designed and built for the American Steel Foundries at their Granite City, III., plant.

(c) On the 600,000-pound tensile testing machine of the Baltimore & Ohio Railroad at their laboratory at Mount Clare shops, Baltimore, Md.

This research program involved the testing to destruction of a considerable number of 40- and 50-ton cast-steel side frames. The scope of the tests was as follows:

Fatigue tests on Symington side frame testing machine.

Fatigue tests on American Steel Foundries testing machines.

Static tests on Baltimore & Ohio Railroad Co.'s tension-testing machine
of 600,000 pounds capacity.

Tension tests from coupons attached to frames, made by the Baltimore &
Ohio Railroad test bureau.

Chemical analyses from tension-test coupons, made by the Baltimore & Ohio test bureau.

Photomicrographs of annealing coupons attached to side frames, made by the Baltimore & Ohio test bureau.

The first paragraph of interchange rule 3 provides that any car of an untried type, whether built new, altered, or changed shall not be accepted in interchange, nor accepted from car owner, until its size, capacity, and design shall have been approved by the transportation and mechanical divisions of the Association of American Railroads.

Under this provision new designs of trucks and side frames, before being applied to cars, are submitted to the association for approval and truck bolsters are given dynamic test and truck side frames both dynamic and static tests in accordance with a specified procedure developed by the committee on car construction.

These tests are conducted under the supervision of a representative of a special subcommittee and static tests are conducted wherever facilities are most readily available, dynamic test being conducted either at the dynamic testing machine of the Symington-Gould Corporation at Depew, N. Y., or the machine of the American Steel Foundries located at Granite City, Ill.

The testing of these frames and bolsters is at the expense of the manufacturers submitting same for approval, except that the expense of the association's observers is paid by the Association of American Railroads.