Production of rock aggregates is an important industrial activity. Quality estimation of rock aggregates is often performed with standardized mechanical tests which are intended for testing the products, not the original rock material. In fact, conventional tests dealing with mechanical performance of aggregates (abrasion and fragmentation resistance) are laborious tasks. They are time consuming and require tough laboratory procedures. Thus, there is a need for an effective method to estimate the quality of rock aggregates in the early stages of quarry prospection. This work aims to present a non-destructive ultrasonic technique to characterize mechanical strength of carbonate rock aggregates, mainly defined with Los Angeles (L.A.) and Micro-Deval (M.D.E.) measurements. For experimentation, porosity, density, L.A., and M.D.E. coefficients were calculated for 11 carbonate rock samples. Beforehand, ultrasonic measurements were taken on rock samples using longitudinal P wave with a frequency of 55 kHz. Regression analysis indicated that L.A. and M.D.E. coefficients were linearly correlated with ultrasonic velocity. Similar results were shown for porosity and density. Artificial neural networking was performed to establish a predictive model linking porosity density and ultrasonic velocity to L.A. or M.D.E. measurements. For our knowledge, this is the first paper of a correlation between L.A. and M.D.E. coefficients with ultrasonic velocity. Results have indicated the ability of this technique to elaborate an accurate approach for prediction of mechanical performances determined with laborious experiments on rock aggregates. This paper adds to the knowledge about the wide effectiveness of ultrasonic techniques to predict the quality of aggregates and proves its efficacy in estimating their quality in the early stages of its production, in field and in time, which presents economical benefits for rock aggregate industries.
Read full abstract