Quantifying and Predicting Tensile Properties of Curcuma longa-silicone Biocomposite


  • Nurrul Amilin ZAINAL ABIDIN Universiti Teknologi MARA Pasir Gudang
  • Nasyitah OTHMAN Infineon Tehchnologies
  • Akid Hilmi ZULKEFLI Universiti Teknologi MARA
  • Jamaluddin MAHMUD Universiti Teknologi MARA https://orcid.org/0000-0002-0351-0474




hyperelastic, artificial neural network, curcuma longa-silicone biocomposite, Neo-Hookean model


This study was carried out to introduce newly developed silicone-biocomposite materials, of Curcuma longa-silicone biocomposite; and assess its tensile properties of using the Neo-Hookean hyperelastic constitutive equation. The specimens were prepared from the mix of Curcuma longa fiber and pure silicone at various fiber composition (0 wt.%, 4 wt.%, 8 wt.%, and 12 wt.%). A uniaxial tensile test was carried out by adopting the ASTM D412 testing standard. The Neo-Hookean model was employed to obtain the material constant, C1 value. Results obtained indicate that the incorporation of Curcuma longa fiber improves the stiffness of the silicone biocomposite as can be seen from the increase of the tensile modulus, while marginally decreasing its tensile strength. The material elastic constant, C1 of silicone reinforced with Curcuma longa was then predicted by using Artificial Neural Network (ANN). The regression coefficients obtained by training the neural network are satisfactory, therefore the neural network can be used for predicting the material constant, C1 of the silicone biocomposite. The prediction of ANN generates a better correlation if there are more data set and can be a good fit for predicting the unknown value.