Use of analysis of variance and linear regression as prediction tool for mechanical performance of SBR

Abstract

Statistical methodology analysis of variance (ANOVA) and linear regression were applied to predict the mechanical performance of the styrene‐butadiene rubber (SBR) containing a residue of the elastomer originated from shoe sole cuttings. A basic formulation was developed to which three different contents of SBR residue were added. Compositions were also prepared with three different concentrations of silica to compare the performance of the filler, as well as to verify the existence of any interaction between the components. Using of ANOVA have permitted the validation of the results of the mechanical properties stress rupture, elongation and tearing resistance. Linear regression analysis demonstrated that silica and SBR residue concentrations have an effect on the mechanical properties and this relationship can be expressed through behaviour models, which can be represented graphically. It was possible to predict a mechanical behaviour model including intermediate concentrations not experimentally measured within the observed limits of the sample.