July 30, 1917 Influence of Grading on Value of Fine A ggregate 2 73 



Ptf? CCNt SAND OVER 'ZO MESM. 



Fig. 9. — Curve of tension tests of artificially graded sand mortars. 



of 100 percent coarse is 95 percent stronger than one composed of fine 

 only, but is 35 per cent weaker than the strongest sand (70-0-30). The 

 total variation in strength is 2,505 pounds per square inch, or 100 per 

 cent from the average of 60 determinations. 



In figures 9 and 10 the results of the tension and compression tests are 

 replotted on rectangular coordinates to show the effect of the coarse sand 

 (X inch to No. 20 mesh) on the strength of the mortar. The most interest- 

 ing feature about these 

 curves, apart from 

 their regularity, is the 

 very great variation 

 in strength shown, es- 

 pecially in compres- 

 sion. Thus, without 

 considering either the 

 very coarse or very fine 

 sand, which would be 

 used rarely in actual 

 construction, the variations in strength are still large. Let it be as- 

 sumed, for instance, that a specification calls for a sand which shall 

 show from 20 to 50 per cent retained on the No. 20 sieve. The results of 

 these tests show a possible variation in compression of from 1,200 to 2,000 

 pounds per square inch for sands fulfilling this requirement. They show 

 also that much higher strength may be obtained in the mortar by the use 

 of coarser sands up to as high as 70 per cent retained on a 20-mesh sieve. 



It is realized, of course, 

 that such coarse sands 

 seldom are met with 

 in practice and, if they 

 were, would not be 

 used in ordinary work 

 where a considerable 

 amount of fine material 

 is needed to produce 

 workabiUty. It must 

 be borne in mind, how- 

 ever, that concrete-pavement construction is not ordinary work on ac- 

 count of the severity and variety of the destructive forces encountered. 

 It has been the custom to meet this condition by increasing the amount 

 of cement to about 40 per cent of the mortar in a i-to-5 mix, and, 

 while this is good practice, it seems reasonable to suppose that a still 

 greater resistance to these destructive agencies would be obtained by 

 increasing the amount of the coarse sand which takes this wear (the 

 X inch to 20-mesh material) over that which is usually considered good 



20 30 *0 30 SO 70 SO 90 100 



PER CENT 5AN0 OVtn ♦20 MtSrt. 



Fig. io. — Curve of compression tests of artificially graded sand mortars. 



