Study of Compressive Strength of Palm Kernel Shell Concrete Under Various Curing Methods and Predictive Modeling using Machine Learning

Authors

• Hyginus Obinna Ozioko, Department of Civil Engineering, Michael Okpara University of Agriculture, Umudike, Abia State, NigeriaFollow
• Emmanuel Ebube Eze, Department of Civil Engineering, Michael Okpara University of Agriculture, Umudike, Abia State, NigeriaFollow
• Edward Ingio Adah, Department of Civil and Environmental Engineering, University of Calabar, Calabar, Cross River State, NigeriaFollow

Keywords

Palm kernel shell, Compressive strength, Machine learning, Adaptive neuro-fuzzy inference system (ANFIS), Artificial neural networks (ANN) and gene expression programming

Document Type

Article

Abstract

This study evaluates the compressive strength of Palm Kernel Shell Concrete (PKSC) under different curing methods and develops machine learning models for strength prediction. Concrete mixes containing 0–100% PKS replacement of coarse aggregate were prepared at a constant water-cement ratio of 0.5 and cured by immersion, sprinkling, wet hessian, and open-air methods. Compressive strength was measured at 7, 14, 21, and 28 days. Increasing PKS content reduced slump from 82 mm to 18 mm, oven-dry density from 2390 kg/m³ to 1430 kg/m³, and initial setting time from 108 min to 76 min. Immersion curing produced the highest compressive strength across all mixes, with the control mix achieving 34.9 MPa at 28 days compared with 25.7 MPa under open-air curing. The 40% PKS mix achieved 21.7 MPa at 28 days under immersion curing, satisfying the minimum strength criterion for structural lightweight concrete. The predictive models showed strong performance, with R² values of 0.9965 for ANFIS, 0.99701 for ANN, and 0.9462 for GEP testing. The findings indicate that PKS can be used as a sustainable lightweight aggregate when replacement level and curing method are properly controlled.

Recommended Citation

Ozioko, Hyginus Obinna; Eze, Emmanuel Ebube; and Adah, Edward Ingio (2026) "Study of Compressive Strength of Palm Kernel Shell Concrete Under Various Curing Methods and Predictive Modeling using Machine Learning," Engineering and Technology Journal: Vol. 44: Iss. 5, Article 6.
DOI: https://doi.org/10.30684/2412-0758.1567

DOI

10.30684/2412-0758.1567

First Page

115

Last Page

152