XXIV. W. II. WARREN, 



U 



where u is the circumference of the cross section. 



(a + /i^) ^ 



In order to ascertain the strength of concrete cubes of various 

 proportions of cement, sand, and stone, it is necessary to prepare 

 and test the specimens under conditions which are maintained as 

 nearly uniform as possible throughout. Great care should be 

 taken in the determination of the proportions of the various 

 materials forming the concrete, and the quantity of water, so that 

 every cube tested has actually the composition intended. This is 

 accomplished by using the same cement throughout, also the same 

 sand in which the sizes of the grains are restricted by two sieves 

 of definite sizes, as for instance, passing the sand through a sieve 

 of 400 meshes per square inch, and catching it on one of 900 

 meshes per square inch. 



The broken stone used should be separated out into about three 

 or four sizes by means of suitable screens or sieves and these after- 

 wards mixed in the proportions intended. The quantity of water 

 necessary should be separately determined for each kind of con- 

 crete, and this should be weighed out and mixed, when preparing 

 the specimens in the correct proportion previously determined. 

 The mixing should preferably be performed in a machine for the 

 sake of better securing uniformity. 



The concrete should be filled into metal moulds having plane 

 and parallel faces, and carefully rammed — here also a machine 

 similar to the Bohme hammer apparatus or lever press would be 

 advantageous. The writer generally leaves the cubes in the 

 moulds for 21 hours covered with damp cloths, and 3 hours out of 

 the moulds before placing them in water. 



The testing in a modern testing machine is comparatively a 

 simple matter if the specimens have been accurately prepared, 

 having parallel plane surfaces between the compression plates of 

 the machine, the bottom one of which is provided with a ball 

 bearing. The load is gradually increased until the specimen shews 

 some signs of yielding, such as a hair crack; the load at this point 



