103. An over-limed or highly limed cement is usually slower setting than an over-clayed one. Among natural cements, those of the aluminous variety are usually quick setting. Other things being equal, a well-burned Portland cement will be slower setting than an underburned sample. It is not certain that such is the case for all natural cements, though it probably is true of most of them. It has been said that underburned cements owe their quick setting to their porosity, but the formation of different compounds in the higher temperature may also account for the difference.

104. The effect of the age of cement on its time of setting is very marked, but varies widely with different samples. The idea that cements invariably become slower setting by storage is a false one. The origin of this error may be found in the fact that by the time cement has reached its destination, it has usually passed through the earlier and more rapid changes in characteristics. Dr. Erdmenger1 has stated that some Portland cements become slower setting, while some set more rapidly as a result of storage. Dr. Tomei made experiments on several Portland cements2 which show that they generally become quicker setting at first (from one to four months after grinding), and then become gradually slower setting, until at the end of a year they set in about the same length of time as when fresh. The writer has seen this trait exhibited very plainly by samples of Portland cement of American manufacture, but has not noticed it in natural cements. Table 18 gives the results of some tests on the effect of aeration on the time of setting of five samples of natural cement from the same factory.

Table 18. Time Of Setting Of Five Samples Of Natural Cement As Affected By-Aeration

1 "Notes on Concrete," by John Newman, p. 11.

2 Trans. A. S. C. E.. Vol. xxx, p. 12.

105. The coarse particles in a cement retard the setting because they are inert. Either fine grinding or sifting will doubtless hasten the rate of setting, but, as has been stated above, the detection of changes in the rate is difficult. Table 11, § 81, gives the results of a few tests on this subject.

106. Addition Of Salts

The time of setting of a cement is sometimes regulated at the factory by addition of sulphate of lime to the finished product. Such additions are admitted to the extent of two per cent, by the regulations of the Association of German Portland Cement Makers, and are now quite generally made by American Portland cement manufacturers. Table 19 gives the results of a few experiments on the effect of plaster of Paris on the time of setting of several cements.

Table 19. Effect Of Plaster Paris On Time Of Setting

It is seen that small percentages retard the initial setting in a marked degree, the maximum effect usually being given by 2 per cent, of the plaster. Larger percentages tend to make the cement quicker setting again, so that with 6 to 10 per cent, added, the cement may begin to set quicker than without the addition of plaster. The final set (time to bear one pound wire) does not appear to be thus hastened by large percentages. This might be considered to indicate that the hastening of the initial set is caused by plaster of Paris taking up the water from the cement and obtaining sufficient hardness to bear the light wire.

The probable explanation of the action of a small amount of sulphate of lime in retarding the setting is that suggested by M. Candlot,1 namely, that the aluminate of lime, to which is due the initial setting, dissolves less readily in a solution of sulphate of lime than in pure water. If the aluminate does not commence to hydrate until the silicate of lime has set, the subsequent combination of the sulphate and aluminate may cause the mortar to disintegrate.

107. Solutions of common salt have been found to retard the setting, but when a large percentage of salt is used, it sometimes forms a crust on the top which may resist a light wire and thus make the paste appear to be quicker setting. Sea water generally retards the setting somewhat more than solutions of common salt, probably on account of the magnesian salts present, but M. Candlot says that cements to which sulphate of lime has been added set more rapidly when gaged with sea water than when gaged with fresh water.

The effect of calcium chloride on the setting of cements is entered into in detail in M. Candlot's treatise on " Cements and Hydraulic Limes," and may be summarized as follows: A weak solution of calcium chloride renders Portland cement slower setting because the aluminate of lime dissolves more slowly in such a solution than in pure water. On the other hand, the aluminate dissolves rapidly in a concentrated solution of calcium chloride, and therefore such a solution hastens the setting of Portland cement. Aluminous cements, i.e., cements containing a very high percentage of alumina, are not appreciably affected by gaging with a comparatively weak solution of calcium chloride on account of the large excess of aluminate of lime present; and on the other hand, cements containing no alumina are not affected, as in such cements the hardening is due to the silicate of lime. A weak solution of the chloride hastens the hydration of the free lime, and therefore a cement which contains a dangerous percentage of the latter may be made sound by gaging with such a solution, as the lime may thus be hydrated before the cement sets. The chloride of calcium test for soundness is based on the supposition that the free lime may be hydrated by the action of the chloride soon after the setting of the cement, and thus the expansive action be hastened.

1 "Ciments et Chaux Hydrauliques," par E. Candlot.

The effect of sugar on the time of setting does not seem to be well known, but it is said 1 that the presence of saccharine matter may either accelerate or retard the setting of the cement, depending on the amount of sugar present, the character of the cement and the amount of water used.