ARCHIVES
VOL. 12, ISSUE 2 (2026)
AI-assisted performance-based fire design of reinforced concrete structures: Machine learning prediction of residual capacity and Eurocode-aligned simplified design charts
Authors
Arogo Eyaramuonan Charles
Abstract
Performance-based fire design of reinforced concrete (RC) structures
requires accurate prediction of residual capacity following elevated
temperature exposure. Traditional design approaches rely on prescriptive
tabulated data and simplified analytical formulations provided in Eurocode 2
Part 1-2 [1], which may not fully capture complex thermo-mechanical
degradation mechanisms. This study presents an AI-assisted framework for
predicting post-fire residual capacity of RC materials and members using
machine learning (ML). A curated database of experimental fire tests is used to
train predictive models for residual compressive strength, reinforcement yield
strength, and member resistance ratios. Model predictions are benchmarked
against Eurocode 2 fire provisions [1] and standard fire exposure
conditions consistent with ISO 834 [2]. Interpretable ML techniques
are employed to quantify parameter influence, and simplified AI-based design
charts are developed for engineering application. Results demonstrate that ML-based
predictions reduce estimation error by up to 35% compared with simplified
Eurocode approaches while maintaining conservative safety envelopes through
quantile-based design curves. The framework provides a computationally
efficient and reliability-aware tool for performance-based fire design and
post-fire assessment.
Download
Pages:8-10
How to cite this article:
Arogo Eyaramuonan Charles "AI-assisted performance-based fire design of reinforced concrete structures: Machine learning prediction of residual capacity and Eurocode-aligned simplified design charts". International Journal of Research in Advanced Engineering and Technology, Vol 12, Issue 2, 2026, Pages 8-10
Download Author Certificate
Please enter the email address corresponding to this article submission to download your certificate.