Thin-Section Petrography of Concrete Aggregates: Alternative Approach for Petrographic Number Evaluation of Carbonate Aggregate Soundness

Abstract

Traditional petrographic number (PN) ratings of concrete aggregates cannot work reliably in Florida because most of the state’s rocks are too fine-grained for use of macroscopic petrographic evaluations. This study designed a thin-section approach to PN calculations, based on relevant carbonate rock components that can be seen easily in thin section. The primary lithologic features chosen to differentiate the carbonates are the types of allochems, the porosity type, the kind of cement or matrix, and any additional noncarbonate minerals. Ten sets of aggregate samples were examined (eight carbonate and two noncarbonate sets) to test the method and ranking of thin-section PN determination. Prior work with PN values showed that aggregates with PN values of less than 140 yield good field performance, 140 to 160 have fair to poor field performance, and those greater than 160 normally have poor performance. Results of the thin-section PN method found values for our samples ranging from 100 (an ideal aggregate) to 149. By using factor weights (FWs) developed, six aggregates have PNs below 140 (good), and four scored between 140 and 160 (fair to poor) and may need additional testing. This method of PN calculation is easy to learn, clearly focused on carbonate aggregates like those in Florida and surrounding regions, and relatively inexpansive, and with it, soundness predictions can be determined quickly (in approximately 1 h per sample). Correlation of thin-section PN values with field performance information has not yet been completed, however, and this must be done before predicted performance can be considered reliable.