علوم و تکنولوژی پلیمر (Apr 2022)

Development of a New Model Based on Ogden-Roxburgh Model for the Prediction of the Stress-Softening Behavior of Carbon Black-Filled Rubber Compounds

  • Mir Hamid Reza Ghoreishy,
  • Foroud Abbassi-Sourki

DOI
https://doi.org/10.22063/jipst.2022.3130.2142
Journal volume & issue
Vol. 35, no. 1
pp. 69 – 82

Abstract

Read online

Hypothesis: The aim of this study was to propose a modified model for the prediction of a stress softening behavior (Mullins effect) in carbon black filled rubber compounds. A new equation was suggested for the calculation of the damage variable in the classical Ogden-Roxburgh model based on a previously developed kinetic equation. The parameters of the new model were assumed dependent on the first principal strain. The developed model was verified by comparison of the model predictions with experimental dataMethods: Four rubber compounds based on S-SBR and E-SBR reinforced by 40 and 60 phr carbon blacks were prepared and cured into rubber sheets. The rubber test specimens (ASTM D412 C) were cut and subjected to cyclic tensile tests at an extension rate of 500 mm/min. In order to show the stress softening behavior, three cycles were selected in a way that the maximum stretch at each cycle was increased consecutively. The volumetric tests were also carried out to determine the bulk modulus and Poisson's ratio. The finite element models of the mentioned tests were created for Abaqus code. The new model was implemented into Abaqus through a user-defined subroutine developed specifically for this research. An optimization algorithm developed in Isight code was employed to determine the parameters of themodel for the prepared compoundsFindings: Comparing the predicted force versus time and force versus displacement with their corresponding experimentally measured data and goodness of fitting for new model and classical Ogden-Roxburgh model revealed that the developed model has higher capability and accuracy in prediction of the mechanical behavior of the rubber compounds. Comparing the ratio of the computed errors between two models showed that the new model has higher accuracy with an average of 38%. Moreover it is found that there are good correlations between variation of the model parameters with rubber grades and filler contents.

Keywords