Application of the nested anova model to evaluate the uniformity of ready-mixed concrete

Abstract

The objective of this research was to apply and validate the Nested Anova Model as an analysis tool of the uniformity of ready-mixed concrete production. In theory this method allows the decomposition of the total variance ( σ² total) of the concrete strength, or any other property to be evaluated, as the sum of its components. In the present study the following components were considered: “batch-to-batch”, “within-batch” and “within-test or laboratory” ( σ² total =σ² batch-to-batch + σ² within-batch +σ² within-test), being the batch the basic unit of production. This variance decomposition allowed assessing the contribution of each component to the total variance of the studied property, in the same unit of measurement and in relative terms. The method was successfully applied to an actual ready-mixed concrete production and distribution study case. It was concluded that the main source of concrete strength variability was the within-batch variation (51.5%; normal delivery condition without the addition of water on site). The batch-to-batch variability resulted to be the second most important contributor to the strength variation (41.1%). The variability attributable to the control laboratory (or within-test deviation estimated based on differences in strengths of companion test cubes made from the same sample of concrete) resulted to be the smallest contributor to the strength variation (7.4%). The results clearly show that the main cause of concrete strength variability was a deficient homogenization of the concrete during the mixing process in the mixer trucks, and therefore there is a great opportunity for improvement of the company