Akademik Veri Yönetim Sistemi
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, cilt.104, sa.2, ss.233-242, 2010 (SCI-Expanded)
The ready mixed concrete industry has been developing rapidly as a result of technological improvements, especially for skyscrapers and high rise structures. Concrete design is a hard and expensive job, moreover it takes too much time. Concrete must conform to some special conditions such as durability and strength during its life cycle. The selection of an appropriate chemical admixture is a very important criterion to achieve the desired specifications for concrete. In this study, the effects of admixture components and admixture dosage on the response variables were examined using mixture experiments. Mixture experiments were conducted on mortar for the optimization of the chemical admixture for concrete. They were performed on a much smaller scale in the laboratory. Computer-generated D-optimal design was used to obtain mixture data due to the additional constraints on some of the component proportions. Mixture data was modeled employing second-degree Scheffe polynomials. A graphical approach based on trace plots was suggested for examining the effect of the admixture dosage. For the simultaneous optimization of multiple responses, a desirability function approach was used and the optimum component proportions were determined in each dosage level. The target was to reach an optimum point by obtaining maximum compressive strength and maximum water reduction for concrete with minimum cost of the admixture. The cost calculation was made for the optimum component proportions and, according to the objective, the most appropriate component combination was chosen. The results were validated by confirmation experiments and prediction intervals for mortar, and by further experiments for concrete. (C) 2010 Elsevier B.V. All rights reserved.