Ridge Regression Analysis Application on Iterative FR-SCC Mix Design to Predict Compressive Strength and Slump Flow Parameters

Abstract

This study addresses the development and predictive modeling of polypropylene fiber-reinforced self-compacting concrete (FR-SCC) as a strengthening material for reinforced concrete structures. In jacketing applications, compatibility between the existing substrate and the new layer is critical: concretes with similar compressive strength and modulus of elasticity transfer load more uniformly and reduce the risk of delamination. At the same time, SCC must achieve adequate flow since vibration cannot be applied during placement; poor workability may compromise compaction, bonding, and durability especially when the fiber types like the polypropylene fibers are implemented in the mix design owing to hydrophilic surface behavior. Thus, proportioning SCC with both target compressive strength and sufficient workability is essential. A total of 29 SCC trial iterations were performed, systematically adjusting binder content, water-to-binder ratio, superplasticizer dosage, and aggregate proportions to achieve self-compaction and target strengths with and without 6mm polypropylene fibers. In parallel, a transparent predictive framework was implemented in Excel. The model employs z-score standardization, a physics-guided polynomial expansion, ridge regression, and guideline systems to avoid extrapolations. It achieved high accuracy for compressive strength (R2 ≈ 0.90) and reasonable performance for slump flow (R2 ≈ 0.66). These findings confirm that the proposed ridge regression approach can effectively capture both strength and workability trends, offering a practical tool for optimizing FR-SCC mix design.