328 



the grain; IV., compression along the grain; V., compres- 

 sion across the grain. 



I. — Beam Experiments. 



The timbers used for these tests were a little over 6 ft. 

 long, the depths of the beams ranging from 3-J- to 5J in. 

 In most cases the supports were placed uniformly 70 in. 

 apart, and the beams were loaded at the middle so that 

 the deflection was produced at the rate of J in. per minute. 

 On one side of the beam at the middle a vertical engraved 

 scale of polished nickel was tacked. Across the front of this 

 was stretched a fine thread, stretched with a piece of elas- 

 tic between two small nails placed half-way up the beam, di- 

 rectly over each support. As the polished nickel acted as a 

 mirror the position of the thread on the scale could be read 

 with great accuracy. 



The amount of deflection produced by a load on a tim- 

 ber beam depends upon the time that the load remains on 

 the beam, even when the load is only a small proportion of 

 that which is required to break the beam. As a conse- 



quence the modulus of elasticity deduced from experiments 

 in a testing-machine, where the load is increased gradually 

 and continuously, is likely to be much higher than should 

 be reckoned on in practice when computing the deflection of 

 a beam under a load that is likely to remain on for a con- 

 siderable time. The curve shown in fig. 1 illustrates the re- 

 sult of an experiment on a beam of Sugar Gum, taken from 

 a tree twenty-nine years old. The beam was 3 J in. wide, 

 3| in. deep, with a span of 70 in. Its ultimate breaking 

 load in the centre was 4,180 lb. This beam was loaded with 

 1,960 Tb., and the deflection was '67 in. This deflection was 

 now kept constant, and the load required to produce it was 

 measured at intervals. At the end of six minutes the load 

 had dropped down to 1,900 ft). In one hour it was only 1,830 

 lb. In twenty-four hours it was down to 1,615 lb., and at 

 the end of three days, when the specimen had to be removed 

 from the machine to make room for other experiments, the 

 load required to produce this deflection of '67 in. was only 

 1.510 lb. Similarly, when a constant load is allowed to re- 

 main on a beam for any considerable time, the deflection 

 produced continually increases. So in using the ordinary 

 tabulated values of the modulus of elasticity for computing 

 the amount of deflection of a beam under practical loadings, 

 allowance must be made for this time effect, and the actual 

 value used for the modulus of elasticity should not be more 

 than half the usual tabulated value as deduced from the or- 

 dinary tests. 



