Studies of Certain Sierran Concrete Aggregates

The purpose of this investigation has been to understand more fully the behavior of certain Sierran rocks when used as concrete aggregate. The granodioritic aggregate used in the construction of the Florence Lake Dam was selected for detailed examination. Two lines of attack to this problem have been followed. The aggregate was quantitatively separated into rock types, and these were then studied petrographically to determine the mineralogical composition, with particular emphasis being placed on the alteration products present in the rocks. The other phase of the investigation was the separation of the aggregate into its major mineral constitutents, in order to test the behavior of these isolated constituents as aggregate in mortar bars. Since the length changes of the mortar bars are to be observed over a two year period, only preliminary results can be reported at this time. On the basis of the known behavior of minerals, the only constituent present in the granodioritic aggregate which might be suspected of causing trouble is a zeolite, a hydrous sodium and/or calcium aluminum silicate. The peculiar property of base exchange whereby the sodium or calcium content of the molecule may be easily exchanged for the other element is worthy of attention in a problem as intimately connected with alkali content as are many concrete failures. The bulk percentage of this hydration product of plagioclase feldspar is certainly not more than several per cent of the aggregate. To separate the mineral constitutents of the Florence Lake Dam aggregate, about 500 pounds of fresh appearing rock was milled. Combinations of the separatory processes of sizing, tabling, magnetic concentration, and heavy liquid concentration yielded the following mineral concentrates: Potash feldspar - 97 per cent purity Quartz - plagioclase - 53 and 43 per cent respectively, yielding a combined purity of 96 per cent Biotite - 96 per cent purity Hornblende - 98 per cent purity Magnetite - 98 per cent purity Sphene, spatite, and zircon - 89, 5, and 4 per cent respectively, yielding a combined purity of 98 per cent These concentrates, mixed with different amounts of inert quartz, were used as aggregate for mortar test bars. Each aggregate mixture was fabricated with high and low alkali cements, and triplicate bars made for each condition. The mortar test bars are as follows: Refer to PDF for table All bars are curing at a temperature of 100° F. in a sealed chamber containing moisture. Length measurements to 0.001 mm. were made at weekly intervals for the first month, and then the interval increased to one month. In June 1943 the bars range in age from one to four months. So far, none of the bars made from granodioritic materials has shown excessive expansion, or even significant expansion. The highest expansion so far with these materials is about 0.04 per cent, as contrasted to an expansion of 0.22 per cent for the high alkali cement and siliceous magnesian limestone aggregate. Graphs have been made recording all expansions of the 200 bars up to June 3, 1943. RECOMMENDATIONS 1. The length measurement program should be continued as planned for the full two years. The monthly measurement interval might be increased to two months if all the expansion-age curves flatten out in time. 2. If possible, restrict the usage of granodioritic aggregates in the Big Creek area until the results of the current testing program are known. If necessary to use any of these aggregates, select only the freshest rock types, and reject the fines resulting from crushing the aggregate. 3. The behavior of the zeolite found in minor quantities at the Florence Lake Dam aggregate stock pile should be tested by using it as aggregate for mortar test bars. The only way to achieve a concentration of this mineral is by hand picking some of the remaining aggregate for rock fragments containing veinlets of this light salmon colored mineral. One of these fragments is submitted with this report. Careful chipping of the rock fragments should yield a concentrate with at least 10 per cent of the zeolite. Crush this 10 - 10 mesh and mix with different amounts of inert quartz to yield three aggregate mixtures with a zeolite content of 1, 5, and 10 per cent. Fabricate each aggregate mixture with a high and a low alkali cement. Should the zeolite prove to be the deleterious constituent of the aggregate, a quantitative determination of this mineral in future aggregate would be invaluable. The base exchange property of zeolites is utilized in water softening processes, and it might be possible to at least semi-quantitatively soften water and obtain rapid determinations of the zeolite content of a crushed rock sample. 4. Should any of the bars show excessive expansion, petrographic investigations should be made with a view to determining the mechanism of the reaction involved.