Journal of building engineering | 2021
Soft computing techniques: Systematic multiscale models to predict the compressive strength of HVFA concrete based on mix proportions and curing times
Abstract
Abstract Advances in technology and environmental issues allow the building industry to use ever more high-performance engineered materials. In this study, the hardness of concrete mixtures with high-volume fly ash (HVFA) has been evaluated and modeled for the LEED (Leadership for Energy and Environmental Design). High-performance building materials may have greater strength, ductility, external factor resistance, more environmentally sustainable construction, and lower cost than conventional building materials. To overcome the mentioned matter, this study aims to establish systematic multiscale models to predict the compressive strength of concrete mixes containing a high volume of fly ash (HVFA) and to be used by the construction industry with no theoretical restrictions. For that purpose, a wide experimental data (a total of 450 tested HVFA concrete mixes) from different academic research studies have been statically analyzed and modeled. For that purpose, Linear, Nonlinear Regressions, Multi-logistic Regression, M5P-tree, and Artificial Neural Network (ANN) technical approaches were used for the qualifications. In the modeling process, most relevant parameters affecting the strength of concrete, i.e. fly ash (class C and F) incorporation ratio (0–80% of cement s mass), water-to-binder ratio (0.27–0.58), and gravel, sand, cement contents and curing ages (3–365 days). According to the correlation coefficient (R) and the root mean square error, the compressive strength of HVFA concrete can be well predicted in terms of w/b, fly ash, cement, sand, and gravel densities, and curing time using various simulation techniques. Among the used approaches and based on the training data set, the model made based on the ANN, M5P-tree, and Non-linear regression models seem to be the most reliable models. The results of this study suggest that the M5Ptree-based model is performing better than other applied models using training and testing datasets. The maximum and minimum percentage of error between the actual test results and the outcome of the prediction using MLR, LR, M5P-tree, and ANN were 0.03–43%, 0.03–54%, 0.04–33%, and 0.03–41% respectively. Based on the outcomes from the models and statistical assessments such as coefficient of determination (R2), mean absolute error (MAE) and the root mean square error (RMSE), the models M5P-tree, ANN, and MLR respectively were predicted the compressive strength of the HVFA concrete very well with a high value of R2 and low values of MAE and RMSE based on the comparison with experimental data. The sensitivity investigation concludes that the curing time is the most dominating parameter for the prediction of the compressive strength of HVFA concrete with this data set.