Scholarlink

This study investigates the mechanical performance and predictive modeling of concrete composites reinforced with shredded Polypropylene (PP) bioplastic and coir fibers. 121 concrete samples were prepared with varying PP content (0.9%, 1.0%, 1.1%) and coir fiber content (0.4%, 0.5%, 0.6%) and subjected to a 21-day curing period. ASTM D790 standards evaluated flexural strength. An Artificial Neural Network (ANN) model was also developed using the R “nnet” package to predict flexural strength outcomes. The final ANN structure, consisting of nine hidden neurons, demonstrated high predictive accuracy based on R2, Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). Garson’s Algorithm was applied to determine the relative importance of each input parameter, showing that PP by weight contributes approximately 53.49% and coir fiber by weight contributes about 46.51% to the overall prediction of stress (MPa). Multiple Linear Regression (MLR) analysis revealed that both PP and coir fiber are statistically significant predictors of flexural stress (p < 0.001), with PP exhibiting a positive influence (β = 906.94) and coir fiber showing an adverse effect (β = −2057.78) on strength performance. These results establish a data-driven foundation for optimizing mix designs using sustainable materials and confirm that PP bioplastic combined with coir fiber presents a viable alternative reinforcement for non-structural and lightweight concrete applications. Nonetheless, limitations regarding experimental scale, material consistency, and laboratory conditions highlight the need for further research and real-world validation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Polypropylene (PP) Bioplastic with Coir Fibers as a Reinforcement in Concrete Composites Using Artificial Neural Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nathalie Faye H. Almazan,
  • Blete Mae S. Dante,
  • Mitzi S. Velos,
  • Donna Ville Gante

摘要

This study investigates the mechanical performance and predictive modeling of concrete composites reinforced with shredded Polypropylene (PP) bioplastic and coir fibers. 121 concrete samples were prepared with varying PP content (0.9%, 1.0%, 1.1%) and coir fiber content (0.4%, 0.5%, 0.6%) and subjected to a 21-day curing period. ASTM D790 standards evaluated flexural strength. An Artificial Neural Network (ANN) model was also developed using the R “nnet” package to predict flexural strength outcomes. The final ANN structure, consisting of nine hidden neurons, demonstrated high predictive accuracy based on R2, Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). Garson’s Algorithm was applied to determine the relative importance of each input parameter, showing that PP by weight contributes approximately 53.49% and coir fiber by weight contributes about 46.51% to the overall prediction of stress (MPa). Multiple Linear Regression (MLR) analysis revealed that both PP and coir fiber are statistically significant predictors of flexural stress (p < 0.001), with PP exhibiting a positive influence (β = 906.94) and coir fiber showing an adverse effect (β = −2057.78) on strength performance. These results establish a data-driven foundation for optimizing mix designs using sustainable materials and confirm that PP bioplastic combined with coir fiber presents a viable alternative reinforcement for non-structural and lightweight concrete applications. Nonetheless, limitations regarding experimental scale, material consistency, and laboratory conditions highlight the need for further research and real-world validation.