Title: Statistical Comparison of Properties of Manufactured Sand-to-Mortar Durability and Scaling Resistance

Author(s): P. P. Hudec and G. Agistalis

Publication: Symposium Paper

Volume: 192

Issue:

Appears on pages(s): 693-704

Keywords: aggregates; expansion; freeze-thaw durability; mortars; sands; scaling; statistical analysis

DOI: 10.14359/5779

Date: 4/1/2000

Abstract:
The results tests on rock aggregate and manufactured sand from Paleozoic carbonate rocks from quarries in SW Ontario were compared those of mortars containing the manufactured sand. The aggregate tests included petrographic analysis, water absorption and adsorption, linear expansion under various conditions, analysis, water absorption and adsorption, linear expansion under various conditions, thermal expansion, insoluble residue content, micro-Deval abrasion loss, freeze-thaw loss, and rate of settlement of -.075 mm. (-#200) fraction. The mortar tests consisted of drying shrinkage, water absorption and adsorption, linear expansion under various conditions, thermal expansion, and scaling and freeze-thaw loss. Multivariate statistical techniques (factor, D-cluster, tree, and stepwise regression analysis) were used to compare and group the properties of aggregates and mortars. Factor analysis showed that the results could be grouped into four factors: (1) Durability factor, (2) Porosity factor, (3) Thermal and (4) Isothermal Expansion Factors. The first two factors were found to be the most encompassing, and grouped the most significant test for aggregate and mortar frost resistance. Stepwise regression predictive models of mortar resistance to salt scaling were developed, based on results of simpler tests on aggregates. K-cluster analysis successfully classifies the aggregates and mortars made from them into good and poor categories. The tree analysis provides the passing limits that can be applied to aggregate tests of any defined group of aggregates.