Early in the study of the causes and the search for measures to correct the excessive expansion of certain combinations of aggregates and cement, the potential corrective properties of certain finely ground siliceous materials was observed.
As previously reported, the original tipoff on the potentialities of this means of correction was the performance of the high silica cement “HP” in the tests.
Subsequent measurements on the same high silica cement specimens up to 6 yr. failed to show any expansive reaction (Cement HS Table 1 and Fig. 1).
The silica addition used in this high-silica cement was a diatomaceous shale known locally as “Monterey Shale” occurring in the Miocene Period of the vast marine sedimentary deposits found along the California coast from north of the San Francisco Bay to Orange County in the south.
Subsequent tests in which this material has been substituted for varying percentages of portland cement have universally shown it to be the most effective corrective tested, when used in an amount equal to 20 per cent substitution for the cement, in highly reactive combinations. There is good reason to believe that somewhat less than 20 per cent will be found fully corrective with most reactive combinations, including the most reactive natural sands found in California.
Since 1940 studies have been made of potential pozzolans from other California sources, including samples of Marine Miocene shales from a number of locations between the San Francisco Bay region in the north and Orange County in the south.
These Marine Miocene shales have consistently given the best results of all of the siliceous materials tested, as is clearly demonstrated by the data contained in this paper.
The opaline cherts of the same geologic formation have likewise performed excellently except under certain conditions and with certain aggregates. Two and one half to five per cent substitutions of the California opaline chert No. 28038 usually slightly accelerates and magnifies the expansion in the case of the California reactive aggregates. In greater percentages, 10 per cent or more, the correction appears to be 100 per cent except that in the case of the Kaw River Kansas and Platte River Nebraska aggregates the expansive reaction of combinations of these mildly reactive aggregates and high alkali cements is materially accelerated and increased at 100 F. and 130 F. by 5 to 10 per cent of the minus 325-mesh opaline chert No. 28038 (Tables V and VI and Figs. 3 and 4). The composition of the cement likewise appears to materially influence the performance of this opaline chert type pozzolan.
While numerous reports on the California experience during the past 10 yr. have frequently included data on observations of the role of pozzolans in a reactive combination, none of the reports so far published have contained a complete summary of all work done by California along this line.
The increasing interest in the subject of blended cements, including the observations of some investigators which have cast doubt on the permanence of the correction of the expansive reaction is the occasion for this paper in which there is a complete review of all studies conducted by the Materials and Research Department of the California Division of Highways on the potential corrective properties of a number of pozzolans. The results based on tests up to seven years, appear to justify the conclusion that not only can permanent correction be anticipated with a suitable pozzolan but likewise that the usually assumed 20 per cent blend may, in certain cases, and with certain pozzolans, be materially reduced. It is clearly indicated that a 15 per cent blend of the Monterey shales is sufficient with any known California reactive aggregates.