Sn ANNUAL REPORT 
2. When hardening in air the average linear shrinkage of four 
Portland cement mortars after three weeks was 0.0417 per cent. on a 
prism of 10 cm. length. This change of volume ts, of course, normal and 
due to the peculiar colloidal properties of cement. 
3. The change in volume due to change in temperature cement has 
in common with all other bodies. The coefficient of expansion of Port- 
land cement concrete is from 0.0000137 to 0.00001 48. 
4. The irregular increase in volume due to abnormal conditions in 
the Portland cement is a great source of danger in using cement and 
may be due to either of the following causes: 
Excess of lime present. 
Large amount of alumina present. 
Coarse raw grinding. 
Coarse clinker grinding. 
Presence of excessive amounts of sulphates or magnesia. 
Underburning. 
Insufficient storing. 
SDR w® b F 
Such a cement is said to be inconstant in volume, and for the detec- 
tion of such materials many tests have been proposed and a great deal 
of literature exists on this point. The matter of testing cements quickly 
in this respect is far from being settled. The most accurate methods of 
determining undue increase in volume in hardening are based on direct 
measurements of the expansion and the best known instruments for this 
purpose are the Bauschinger and the Martens apparatus. 
The first is simply a micrometer which measures the length of a 
bar of cement mortar about 10 cm. long, exact to 1-200 mm. The second 
instrument measures the length of a bar either in air or water by means 
of a steel point which touches the specimen on the one hand and is 
connected with a rhombic piece of steel on the other. An increase in 
length will move the rhombic tongue, and with it a long pointer. The 
instrument virtually compares the changes in length of the test piece 
with those of the steel part, and since the two coefficients are practically 
alike ‘fluctuations in temperature have but little influence on the exactness 
of the readings. 
Accelerated Tests.—But the use of these instruments requires a 
long time, and for practical purposes where results must be had quickly 
and without the use of delicate apparatus they are not very suitable, 
especially in the hands of the untrained manipulator. For this purpose, 
nence, other tests have been employed, and of these the following are the 
most important: 
Boiling test, 3 to 6 hours. 
Steam test, 3 to 6 hours. 
Warm water test, 24 hours. 
Normal water test, 28 days. 
oe 
