Journal of Scientific & Industrial Research Vol. 65, November 2006, pp. 894-899 A new approach to determination of compressive strength of fly ash concrete using fuzzy logic Tayfun Uygunolu and Osman Ünal Afyon Kocatepe University, Technical Education Faculty, Turkey Received 27 September 2005; accepted 05 July 2006 In this study, determination of effect of fly ash (FA) content on the compressive strength of concrete depending on water/cement ratio and concrete age was investigated by use of fuzzy logic (FL) approach. In the approach of modelling with FL, compressive strength values of various sample of concrete that produced by replacement of cement by F class of FA by ratio of 0 (control), 10%, 20% and 30% were used. Water/binder ratio of these concrete samples was varied between 0.27-0.60 in six different values. Experimental compressive strength values of the concrete specimens at 3, 7, 28, 90, 180 and 365 days compared with FL values obtained using the fuzzy sets. Optimum FA content and water/binder ratio for the best compressive strength for early age and hardened concrete can be obtained with FL. Keywords: Fuzzy logic, Fly ash, Concrete, Compressive strength IPC Code: B01F3/80 Introduction Mineral admixtures, mostly pozzolanic materials, in cement and concrete include economic benefits and environmentally safe recycling of industrial and other waste by-products. Unlike chemical admixtures, they are used in relatively large amounts as replacement of cement with fine aggregate in concrete 1 . Fly ash (FA) is a pozzolanic material widely used as a cement replacement to produce high-performance concrete and high-volume FA concrete. Benefits, in terms of the durability of high-strength concrete, are well established and the use of FA is now considered commonplace 2-4 . FA is generally finer than Portland cement and ASTM specification C 618 defines two broad types of FA (Class F and Class C) related to the type of burned coal. Class F fly ash (CaO < 6%), designated as low calcium ashes, is not self-hardening but generally exhibits pozzolanic properties 1 . At early ages, the strength of concrete containing high volumes of FA as a partial cement replacement is lower than that of control concrete; mainly due to slow pozzolanic reaction of FA 5 . One of the important properties needed for successful application of these materials in structures is compressive strength (CS) 6 . Actual CS of concrete is unknown during the early life of structure 7 . Current empirical equations for estimating CS are based on tests of concrete without supplementary cementitious materials. The more one knows about the concrete composition versus strength relationship, the better one can understand the nature of concrete and how to optimize the concrete mixture 8,9 . The main contribution of FA to concrete, amongst others, is to the workability and the reduction of temperature rise in fresh concrete as well as to the durability and long term strength development of hardened concrete 10 . On the use of optimum FA in the concrete for the best CS, Öner et al 11 investigated the strength development of concrete containing FA and the optimum percent of FA in concrete. Han et al 12 predicted CS of FA concrete by apparent activation energy function. Papadakis & Tsimas 13 an empirical formulation for the estimation of CS of FA concrete depending on concrete age and FA content has been developed 14 . Sebastiá et al 15 developed a neural network based method to predict the unconfined CS of FA-cement mixtures. This study presents new approach to determine CS of FA concrete using fuzzy logic (FL). Experimental Study A fuzzy logic system (FLS), which implements a nonlinear mapping between its inputs and outputs (Fig. 1), was used. Experimental variables were water- binder ratio (R), concrete age (Y) and FA Replacement __________ *Author for correspondence Tel: +90 272 228 13 11 / 345; Fax: +90 272 2281319 E-mail: uygunoglu@aku.edu.tr