BEICK ROADS. 7 
specimen brick are subjected to destructive influences very similar to 
those encountered in actual service, and the results obtained, there- 
fore, indicate very closely the effect which traffic may be expected to 
produce on a pavement constructed of similar brick. The methods 
of making the test, of which there were formerly a great many, have 
undergone repeated changes in order that service conditions may be 
more nearly approached, and also in an effort to bring about uni- 
formity, so that the results obtained may be of the greatest possible 
scientific value. The method which is now proposed by the sub- 
committee on paving brick of the American Society for Testing- 
Materials may be briefly described as follows: 
The apparatus necessary for making the test, ordinarily called 
the rattler, consists of a 14-sided barrel of regular polygonal cross 
section supported on a suitable frame and fitted with the necessary 
driving mechanism. The staves, each of which forms a side of the 
barrel, are made of 6-inch 15.5-pound structural steel channels 27J 
inches long. These staves are double bolted to the cast-iron heads 
of the barrel, which are provided with slotted flanges for holding 
the bolts. Cast-iron wear plates are bolted to the inside of the 
barrel heads. The inside diameter of the barrel is 28§ inches. 
In this barrel is placed what is known as the abrasive charge. 
This charge consists of two sizes of cast-iron spheres having respec- 
tive diameters of 3§ inches and 1J inches and weighing, respectively, 
7.5 pounds and 0.95 pound when new. Ten of the larger spheres are 
used, and the number of the smaller spheres is made such that the 
weight of the entire charge will approximate 300 pounds. The indi- 
vidual larger spheres are discarded whenever their weight falls to 
7 pounds or less and the smaller spheres when they become sufficiently 
worn by usage to pass through a circular opening having a diameter 
of If inches. 
The test is made by placing a charge of 10 dry brick in the barrel, 
together with the abrasive charge, and then revolving the barrel 1,800 
times. The number of revolutions per minute is not permitted to fall 
below 29J nor to exceed 30J, and the operation is made continuous 
from start to finish. 
The results of the test are reckoned in terms of the loss in weight 
sustained by the brick, and this loss is expressed as a percentage of 
the original weight of the brick tested. In determining the loss in 
weight, no piece of brick which weighs less than 1 pound is considered 
as having withstood the test. 
Good paving brick will ordinarily lose from 18 per cent to 24 per 
cent of their original weight in the rattler test, and specifications con- 
cerning this loss should be prepared with a view to the character of 
the traffic for which the pavement is designed. 
