Mathematical modeling for predication of strength of remixed concrete.

Abstract

This paper deals with the methodology, related to application of mathematical model for reuse of partially set old concrete by adding fresh concrete to form serviceable mix by considering their time lags and blend ratios. As Compared to the strength of the freshly prepared concretes the preset concrete obviously gives the reduction in strength. This reduction is further possible to be minimized to a certain extent on blending some quantity of a relatively fresh mix to the existing quantity of the preset mix. In the statistical methods of concrete mix design in civil engineering the most frequently used models are Scheffe’s and Osadebe’s models, which are quite suitable for concrete mix optimization but are greatly limited as a predetermined number of experiments must be carried out in order to formulate them and they can only be applied for mix ratios that fall within the predetermined observation points. Ibearugbulem’s regression model has been formulated as a new model to take care of these inherent problems in Scheffe’s and Osadebe’s. Some modifications were made to obtain the new model. This new model has been tested on concrete cubes for different mix ratios for 28 days compressive strengths. The Fisher f-test shows that the values of compressive cube strength predicted by the new regression model are very close to those from the experiment strength values, with f-value of 3.44 at 95% confidence level. Hence this new model of regression is useful in concrete mix desig