Regression models applied in the characterization of lightweight mortar with EVA and piassava fibers

ABSTRACT

The civil construction has great potential for recycling other industries residues using these materials as aggregates. The EVA residue (Ethylene Vinyl Acetate) has been used in lightweight mortar production using piassava vegetable fibers (Attalea funifera Martius) as reinforcement. In this work, we build mathematical models to simulate the mechanical properties as compressive strength, tensile strength and modulus of elasticity for mortar with unconventional aggregates of EVA and piassava fibers. We used a polynomial regression model with three explanatory variables to describe each mechanical property, using as independent variables the EVA and fiber percentages in the mixture and also the water/cement ratio. The goodness of fit was measured through hypothesis tests, and statistical criteria’s as Shapiro-Wilk, Durbin-Watson, Breusch-Pagan, the coefficient of determination (R²), and the PRESS statistic. The results allow us to affirm the constructed models reproduce the material behavior. Using the proposed models and simulation techniques, it was found that the piassava fibers can be used as reinforcement elements in the tensile strength and modulus of elasticity for mortars. In addition, we calculate an optimum ratio between aggregates to obtain maximum resistance.

Keywords
Regression Models; Multivariate Analysis; Mechanical Strength; Lightweight Mortar