AI Summary & Key Takeaways
AI InsightIt was observed that there is a highcorrelation between experimental and predicted values obtained from the proposed GEPmodel.
📌 3 Key Takeaways
Abstract
This study aimed to propose a prediction model for estimation of strength of concretes withvarious cements and mixture proportions. The strength of the samples produced with threedifferent types of cement at different rates of water-to-cement ratios and cement richness wereinvestigated experimentally and evaluated statistically. Three type of cement possessing 28-day strengths of 32.5, 42.5, and 52.5 MPa was used in the production of concretes. Theconcretes were produced at cement richness values of 300, 400, and 500 kg/m3 and w/c ratesat changing levels within the interval of between 0.3 and 0.6. By this way, combined influencesof cement strength, amount of cement and w/c ratio was experimentally investigated. Totally36 mixes were cast then the compressive strength values were examined after specified moistcuring periods (7 and 28 day). A statistical study were conducted on the experimental resultsand the significances of the cement strength, w/c values and amount of cement on thecompressive strength of the concretes were assessed. Another crucial focus of the current paperis to generate an explicit expression to predict the compressive strength of the concretestackled with the current study. To derive an explicit formula for estimation, a soft computingmethod called gene expression programming (GEP) was benefited. The GEP model was alsocompared with a less complicated estimation model developed by multi linear regressionmethod. The results revealed that compressive strength of the samples were significantlyinfluenced by cement type and aggregate-to-cement ratio. It was observed that there is a highcorrelation between experimental and predicted values obtained from the proposed GEPmodel.