[ooKER & M'KERGow] THERMAL CHANGE TO TENSION 17 



In the first three experiments the load was applied at the rate of 

 20,000 pounds in 10 seconds, while in the last three the rate was one- 

 half this, as the former rate of loading caused some lag in the readings. 

 In all cases the value of k was approximately -003, and this value was 

 used for obtaining the true reading. The specimen finally broke at 

 a stress of 11,800 pounds per square inch. All the results of Table 

 IV. are plotted on Figure 4, and from this figure it will be seen that 



?f- 



Trte^rrncLLToifie (Xurvcs 

 I Lead loo-Zooo bTu/nds 



TTT ffla t4/mu/T, xj defljt diioTv/ilJrees uj/rot^ 





7 Ti/mje. Trwnu&S o 



bOOO fOOO itOOO 



the cement behaved as an approximately elastic solid, for the thermal 

 stress curves, when corrected, are nearly straight lines, and, moreover, 

 their maximum ordinates are proportional to the stress; this is more 

 clearly seen in the Figure curve VII, where the maximum ordinates are 

 plotted with reference to the loads, and the resulting curves give a 

 straight line very nearly. 



The other half of the same bar was subsequently used for obtain- 

 ing a stress-strain curve. The observations are given in Table V., and 

 the stress strain curve is plotted Curve VIII to compare with the cor- 

 rected thermal load curve obtained previously, and the linear relation 

 is very apparent. 



Impulsive Stress 



It is a matter of common observation that if a tension member be 

 suddenly loaded in a testing machine, the readings of an extensometer 

 applied to the test piece will fluctuate between wide limits, owing to 



Sec. III.. 19()4. 2 



