[coKER A m'kergow] THERMAT- change TO TENSION 



21 



where c = 1 is slightly too great, since in ohtaining it we assume an 

 instantaneous loading. 





If it be assumed that the value of c is linearly proportional to the 

 time, then, since it is unity and one-half in the extreme cases, the value 

 of c in the present case would be "94. 



We have, however, preferred to retain the value unity, since the 

 assumption of a linear law may not be warranted. It is, however, 

 worthy of note that if the value, -94, be used for the correction factor, 

 the maximum value in the first experiment becomes 10 -47, as compared 

 with 10-14 in the second experiment. In all these experiments we found 

 the differences were never greater than four per cent. 



We may assume therefore that any impulsive load effect which 

 shows a greater difference than this, as compared with a steady loading, 

 is not due to the correction factors employed. Although a number of 

 experiments were made on both tension and compression members, it 

 will be sufficient to quote one or two in order to indicate the general 

 character of the results. 



In an experiment upon a steel tension specimen, 2-00 by 0'38 inches 

 in section, the load was first applied gradually, and then very quickly. 

 The data obtained are shown in Table VIIL, and this show that with 

 the gradual loading the maximum corrected value of the deflections 

 was 9 •?4; when, however, the same loading was effected in four seconds. 



