Fire Resistance of Steel Beams with Intumescent Coating – Simulation Data and Machine Learning Models
Description
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.
Files
Steps to reproduce
To reproduce the analyses and simulations presented in 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), follow these steps: 1. Software Requirements: ANSYS 16.0: Use ANSYS for finite element analysis (FEA) simulations. MATLAB R2020b: Utilize MATLAB for data processing and machine learning model development. 2. ANSYS Simulation and Analysis: Perform FEA simulations in ANSYS to model the behavior of steel beams under fire conditions. Use DesignModeler, a CAD module within the Ansys Workbench platform, to develop parametric geometry model of the IPE steel beam. Use Engineering Data to define temperature-dependent material properties of steel and intumescent coating. Use ANSYS Mechanical to develop Transient Thermal and Static Structural systems, assign materials, define boundary conditions, loads, mesh, input and output parameters. Define Simulation Parameters using Parameter Set. Obtain simulation results from ANSYS, including temperature distributions, stress-strain responses, and displacements at specific points. Verify the model based on results provided in the database. 3. ANN Model Development: Data Integration: Integrate the simulation results with the machine learning models to enhance prediction accuracy. Normalization: Normalize the data to ensure consistency and improve the performance of machine learning algorithms. Feature Selection: Identify and select relevant features that influence the fire resistance of steel beams. Machine Learning Algorithms: Implement machine learning algorithms in MATLAB to predict the fire resistance of steel beams. Model Training: Train the models using the preprocessed simulation data. Validation: Validate the models by comparing their predictions with experimental results. By following these steps, researchers can replicate the study's findings and further explore the fire resistance characteristics of steel beams with intumescent coatings.