TESTS ON CEMENT 



213 



Fig. 1 (f) shows the radial cracks that mark the first stages 

 of disintegration. Such cracks should never occur with 

 good material. They are signs of real failure, and cement 

 showing them should never be used. 



Fig. 2 shows three pats that, 

 for different reasons, have left 

 the glass plate on which they 

 were made. The disk shown in 

 (a) left the plate because of 

 lack of adhesion; the one in (6), 

 through contraction; and the one 

 in (c), through expansion. The 

 condition illustrated in (a) is 

 never dangerous in either air or 

 ^ water; that in (c) is only danger- 



FIG. 2 ous when existing in a marked 



degree ; and that in (6) hardly ever occurs in water, but in 

 air it often indicates dangerous properties. Air pats that 

 develop the curvature shown in (6) generally disintegrate 

 later. A curvature of about | in. in a 3-in. pat can be 

 considered to be about the limit of safety. 



The normal pat tests are the only absolutely fair and 

 accurate methods of testing cements for soundness, but the 

 serious objection to them lies in the fact that frequently 

 several months or even years elapse before failure in the 

 cement so tested becomes apparent. To overcome this 

 difficulty the accelerated tests have been devised. These 

 tests are intended to produce in a few hours results that 

 require months in the normal tests. 



Many forms of accelerated tests have been devised. At 

 present, however, the only tests employed commercially 

 are the boiling test and the steam test. 



The boiling test is made by forming specimens of neat- 

 cement paste into pats, such as are employed for the normal 

 tests, or preferably into balls about li in. in diameter. 

 The specimens are allowed to remain in moist air for 24 hr. 

 and are then tested. 



The form of apparatus used for the boiling test is shown 

 in Fig. 3. It consists of a copper tank that is heated by a 



