Fire Resistance of Steel Beams with Intumescent Coating – Simulation Data and Machine Learning Models

This dataset contains numerical and machine learning data generated during the study “Fire Resistance of Steel Beams with Intumescent Coating Exposed to Fire Using ANSYS and Machine Learning” (Buildings 2025, 15(13), 2334; https://doi.org/10.3390/buildings15132334).

The work investigates the fire resistance of structural steel IPE beams protected with water-based intumescent coating (IC) and subjected to ISO 834 standard fire. Numerical simulations were performed in ANSYS 16.0, using sequentially coupled thermal–structural analyses with temperature-dependent material properties. A parametric study was conducted for the entire range of standard IPE profiles (80–600), three IC thicknesses (0.4, 0.8, and 1.2 mm), three utilisation levels (0.5, 0.6, 0.7), and three beam lengths proportional to section depth.

The dataset also contains the input and output files used to train artificial neural network (ANN) models in MATLAB R2020b for predicting fire resistance time. The ANN input parameters include detailed beam geometry, IC thickness, section factor, thermal conductivity values at five temperatures, and loading factor. Output is the predicted fire resistance time until structural failure according to deflection and deflection rate criteria.

Contents:

Excel spreadsheets – Tabulated results from 486 finite element simulations, including input parameters, geometrical and mechanical properties of steel and IC (temperature-dependent), temperature evolution, mid-span deflections and deflection rates, stresses and strains during fire and fire resistance times, input and output data for ANN training before processing, processed and after normalisation.
Figures – Graphical inputs of temperature-dependent thermal and mechanical properties of steel S355, and graphical results of mid-span deflections and deflection rates in time, showing parametric effects of IC thickness, beam length, utilisation, and section factor on fire resistance.

File formats:

.xlsx – Microsoft Excel Worksheet
.xlsm – Microsoft Excel Macro-Enabled Worksheet
.tif / .jpg – Figures and diagrams
.txt – File descriptions

Usage notes:

All Excel files contain both raw and processed numerical outputs; cells cannot be modified, formulas are hidden, content can be copied.
Figures are provided in publication resolution and can be reused under the CC BY 4.0 license.

Dataset is organized into two subfolders:

01_EXCEL_FILES – Contains Excel files with tabulated data, calculations, and results relevant to the study. Each file in this folder is listed and briefly described in the accompanying Excel_Files_Description.txt.
02_FIGURES – Contains image files (figures) illustrating key results, and diagrams. Each figure is listed and described in the Figures_Description.txt file located in the same folder.

Both Excel_Files_Description.txt and Figures_Description.txt provide file names and concise summaries of their content for easier navigation and reference.