■■6-16 PROCEEDIXGS OF SECTION H. 



•strain, to square root of strain, or to any function of strain. The 

 experiment then to make is to subject it to repetitions of stress to 

 ascertanr whether there is any sucli law. The result will certainly be 

 disappointing, for it has always been found that permanent deforma- 

 tions take place after each of the first few applications of load. 

 Fortunately, however, if enough applications are made, the stress 

 strain diagram settles down to practically a straight line, so that 

 Hooke's Law expresses the facts very closely uji to the stresses to 

 which the concrete luxs been loaded. Were this not so^ concrete could 

 not b>e called an elastic material, and engineers would hardly be 

 justified in using it in structures intended to be permanent. It may 

 bo argued that timber is an imperfectly elastic material, and is never- 

 theless used without hesitation in first-class structures. Concrete, 

 liowever, resembles in its crystalline structure metal much more than 

 a fibrous organic material like timber. Possibly, too, experiment 

 would prove that repeated applications of loading develop a true 

 elasticity in timber; at all events, timber girders in railway bridges 

 •cany heavy loads for many years without visible increase in the per- 

 manent set brought about by the first few loadings. 



In the November, 190'8, number of '' Concrete" I noticed a 

 •description of tests of an important wharf at Bi-ocklebank, Liverpool. 



The following is a quotation from it : — '" The wharf was designed 

 for a working superload of 6§ cwt. per square foot, the test load being 

 specified at 10 cwt. per squai'e foot. The resulting deflection in the 

 main transverse l)eaiii was | in., and the set immediately after 

 removal of load ^ in. only. In the four secondary beams the deflec- 

 tions Avere ^, ^•\, ^, and ^ in. respectively, . and the corresponding 

 amounts of set on removal of the load 3^, -;J^, nil, and -/jrin. At the 

 middle of the two deck panels the deiiections were ^^^ in. and j in., 

 the coiTesponding set on removal of load being -j-',. and -^'., in. respec- 

 tively." 



These tests show that a considerable permanent set, larger pro- 

 portionately than we would care for in a steel or even in a timber 

 structure, may be expected in reinforced concrete beams when loaded 

 for the first time, but oiu- knowledge of the elastic properties then 

 acquired by the concrete justifies the belief that no further set will 

 take ])lace unless a heavier load is imposed. 



It is interesting- to compare this with the behaviour of steel when 

 ^.stressed beyond its static elastic limit : it is well known that by doing 

 so the elastic limit may be raised almost up to breaking point, and 

 that so long as the stress never goes below a certain limit, it may be 

 ai)plied innumeraljle times up to the new elastic limit without fear of 

 failure. 



It seems then that in concrete we do every day wliat no engineer 

 would ordinarily think of doing in steel, and that is use it beyond its 

 >;tatic elastic limit. 



As regards the value of the coefficient of elasticity either in 

 tension oi- in compression, it depends upon tlie amount to which the' 

 concrete has l)een stressed becoming less as the stress increases. 



Many antliorities claim that if a curve be plotted having stresses 

 i\K abscissa' and modidi of elasticity as ordinates. it will take the 

 i'f)rm of a ]>arabola of the second degree. 



