The unusual amount of new bridge construc tion required since 1890 caused a general re vision of the standard specifications for bridges, the effect of which was to include the results of recent studies and experiment, and to elim inate some of the minor and unessential items formerly prescribed.
Meanwhile another movement was in prog ress. Experience having shown the great ad vantage of more uniformity in various details and standards relating to the manufacture of bridges both in reducing the cost and the time required for the shop work, an effort was be gun to secure more uniformity in the require ments for the production and tests of steel, which is the metal now exclusively employed in bridges.
With greater uniformity in the physical, chemical and other requirements for steel, as determined by standard tests, the unit stresses to be prescribed for the design of bridges will naturally approach to a corresponding uniform ity. To what extent this is desirable may be inferred from the fact that the application of several of the leading specifications to the de sign of a railroad bridge under a given live load yields results which may vary by an amount ranging from zero to 25 per cent of the total weight.
In the revision of specifications a decided tendency was observed to simplify the design by making an allowance for impact, vibration, etc., by adding certain percentages to the live load according to some well-defined system. It needs but relatively little experience in making comparative designs of bridges under the same loading, to show the advantage of this method over that in which the allowance is made in the unit stress according to any of the systems formerly adopted in such a case. Not only arc the necessary computations greatly simplified, but the same degree of security is obtained in every detail of the connections as in the princi pal members which compose the structure.
In 1907-08 a series of experiments were made under the auspices of the American Rail way Engineering Association to determine the proper percentage of the live load to be added for the effect of impact. Experimental trains were operated over bridges at different speeds and the actual stresses in different members of the bridge trusses were measured by special instruments designed for the purpose. The results were published in 1911 in the 'Proceed ings of the Association.)
In 1914 a series of large-size tests of steel columns was begun by the United States Bu reau of Standards in co-operation with several of the national engineering societies to deter mine accurately the variations in the strength of built-up columns due to the form of cross section, the manner in which the structural shapes are connected, the ratio of slenderness, the quality of material, etc. A number of tests have also been made by the bureau for consulting engineers and bridge companies in which different alloy steels were used The latter test columns were one-half of the actual sizes as designed for several long-span bridges.
The American Railway Engineering Associa tion in 1903 made the first successful attempt to standardize American railroad bridge specifica tions. These specifications, revised at inter vals, are now recognized as standard and are generally used not only as the basis for the design of new metal bridges, but also as a basis for determining the strength of bridges con structed under other specifications.
A . movement which has done much good since 1890 and promises more for the future is that of the organization of bridge departments by the railroad companies. The great economy of making one design rather than to ask a num ber of bridge companies to make an equal num ber of designs, of whiCh all but one are wasted, is the first advantage; but another of even greater significance in the development of bridge construction is that which arises from the designs being made by those who observe the bridges in the conditions of service and who will naturally devote closer study to every de tail than is possible under the former usual con ditions. The larger number of responsible designers also leads to the introduction of more new details to be submitted to the test of service, which will indicate those worthy of adoption in later designs. In order to save time and labor and secure greater unformity in the design of the smaller bridges, some of the rail roads prepare standard plans for spans varying by small distances. For the most important structures consulting bridge engineers are more frequently employed thau formerly, when so much dependence was placed upon comparative designs made by the bridge companies.