Establishment of co-relation factor for strain hardening ratio and carbon equivalent of mostly used 16mm diameter locally manufactured rebars

Abstract

The construction industry in Sri Lanka relies extensively on locally manufactured 16mm diameter rebars for structural applications. Ensuring the mechanical reliability of these rebars is essential for safe and durable construction. This thesis, titled “Establishment of Predictive Models for Strain Hardening Ratio Based on Carbon Equivalent in Locally Manufactured 16mm Rebars,” investigates the relationship between the Tensile-to-Yield strength ratio (T/Y) and Carbon Equivalent (CE), a key metric derived from chemical composition. The research began by tracing the historical evolution of CE as a predictor of mechanical behavior, highlighting its established role in estimating ductility and strain-hardening characteristics. Using tensile test data and CE values calculated via BS 4449:2015, a linear regression analysis confirms a statistically significant relationship between CE and T/Y, consistent with prior metallurgical findings. This validation reinforces CE as a reliable indicator of mechanical performance in rebars. Building on this foundation, the study introduces a binomial logistic regression model that classifies rebar compliance based solely on CE. By defining a ductility threshold (T/Y ≥ 1.15), the model enables binary classification of rebar quality, offering a practical, non-destructive method for compliance screening. The model demonstrates high classification accuracy and strong ROC performance, making it suitable for integration into QA/QC workflows. The outcomes of this research provide the construction sector with a dual-layered analytical framework: one that confirms the CE–T/Y relationship through regression, and another that operationalizes CE as a predictive tool for quality assurance. This approach enhances traceability, efficiency, and decision-making in material selection. Future research is recommended to explore the influence of manufacturing processes, environmental exposure, and broader diameter ranges on model robustness.