EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Prediction of Compressive Strength of Rice Husk Ash Blended Concrete Using Non-Linear Regression Models

Abdulrahman Garba, Ahmad Sani, Salisu Abdullahi Dalhat

Journal of Systematic, Evaluation and Diversity Engineering · 2026

DOI: 10.70382/ajsede.v13i5.045 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Systematic, Evaluation and Diversity Engineering, 2026. Verfasst von Abdulrahman Garba und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Ordinary Portland Cement (OPC) production is a major contributor to global CO₂ emissions, motivating interest in supplementary cementitious materials such as Rice Husk Ash (RHA), a silica-rich agricultural by-product with pozzolanic properties. This study investigated the effect of RHA as a partial cement replacement (2.5%, 5%, 12%, and 15% by weight) on the compressive strength of concrete cured for 7 to 28 days, and compared the predictive performance of three non-linear regression models — Spline, Polynomial, and Physics-Based — against experimental results. X-ray fluorescence analysis confirmed a silica content of 85.00% in the RHA used, consistent with a reactive natural pozzolan. The highest average compressive strength (39.09 N/mm²) was recorded at 5% RHA replacement after 21 days of curing, while higher replacement levels produced lower strengths. The Spline and Polynomial models achieved identical coefficients of determination (R² = 0.8310), with the Spline model recording lower mean absolute error (6.0425) and mean absolute percentage error (18.52%) than the Polynomial model. The Physics-Based model recorded the lowest R² (0.5946) despite the lowest bias. Because the four experimental mixes were not varied independently across replacement level, curing age, and water-cement ratio, and because the dataset was limited to four points, these findings are presented as a preliminary comparative assessment rather than a validated predictive framework. The Spline model showed the most balanced performance among the three approaches. Further work with an independently varied, larger dataset is recommended to confirm these findings and support field application of RHA in sustainable concrete production.

Zitieren

APA

Garba, A., Sani, A., & Dalhat, S. A. (2026). Prediction of Compressive Strength of Rice Husk Ash Blended Concrete Using Non-Linear Regression Models. Journal of Systematic, Evaluation and Diversity Engineering. https://doi.org/10.70382/ajsede.v13i5.045

Vancouver

Garba A, Sani A, Dalhat SA. Prediction of Compressive Strength of Rice Husk Ash Blended Concrete Using Non-Linear Regression Models. Journal of Systematic, Evaluation and Diversity Engineering. 2026. doi:10.70382/ajsede.v13i5.045

BibTeX
@article{Garba2026,
  author = {Garba, Abdulrahman and Sani, Ahmad and Dalhat, Salisu Abdullahi},
  title = {Prediction of Compressive Strength of Rice Husk Ash Blended Concrete Using Non-Linear Regression Models},
  journal = {Journal of Systematic, Evaluation and Diversity Engineering},
  year = {2026},
  doi = {10.70382/ajsede.v13i5.045}
}
Kostenlos starten →