Ti;sTS 131 



ing. At the end of 52 days tests were made on four pipe, of which 

 two had been cured by sprinkling for 30 days, and two had been 

 stacked after 10 days. Through a mistake a top sand box 14 inches 

 wide was used during all of the tests of this series. The average 

 results are as follows : 



The first group of four pipe, age 7 days, broke at an average of 

 2437 pounds. This is about half of the strength which the pipe 

 should attain ultimately. The second group, after 28 days curing, 

 broke at 3644 pounds, an increase of about 50 percent over the 

 strength at 7 days. From a consideration of the 14-inch pipe in 

 Table XUl a further increase in strength could be expected if the 

 curing were continued. 



The third group, four pipe, were cured in the usual way. The 

 average breaking load was 5248 pounds. 'iMie next group were cured 

 in the usual way, but w^ere immersed in water for five hours just be- 

 fore testing. The efifect of the soaking was to reduce the strength 

 32 percent. 



The last four tests included two pipe that w-ere cured 10 days and 

 two that were supposed to have been cured 30 days. The two 

 former had a slightly higher average strength, but the number 

 tested was so small that no conclusions can be drawn. 



The resistance of cement pipe to crushing depends to a large 

 extent on the moisture content of the pipe. It is useless to standard- 

 ize the details of testing and disregard the condition of the test 

 specimens. The object of testing is not to obtain the maximum 

 possible strength, but to ascertain the strength under actual work- 

 ing conditions. Cement pipe is used to convey water, and after 

 laying it is certain to become saturated. The strength of dry pipe, 

 therefore, is artificial and abnormal, and does not measure the 

 strength of the pipe when buried in a trench. At the present time 

 specifications for pipe testing do not stipulate the condition of the 

 pipe. The following clause should be added to cover this feature. 

 All test specimens shall be tested when wet. Preferably 

 they shall be kept thoroughly wet for at least a week prior to 

 testing. If it is necessary to test dry pipe on short notice, they 

 shall be immersed in water for at least four hours before testing. 

 Pipe tests at the University of Arizona hereafter will be made 

 with the pipe in a thoroughly wet condition. 



LOADS ON PIPE IN DITCHKS AND DESK'.N OF PIPE LINES 



The external loads which are borne by a pipe line are due to the 

 backfilling in the trench and any superimposed loads, such as the 



