SM490 steel is a structural steel that complies with JIS G3106 and is widely used in bridges and civil engineering structures due to its combination of strength, ductility, and good weldability. Before being applied to more complex analyses, the finite element model (FEM) for this material must be compared with experimental data to determine its reliability. This study developed an FEM model using Abaqus to simulate the true stress-true strain curve of SM490 steel under monotonic tensile loading and quantitatively evaluated the results against secondary experimental data. Validation was performed at nine points within the strain range of 2.151–7.450% using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Normalized RMSE (NRMSE), and the coefficient of determination (R²). The evaluation results yielded an MAE of 12.96 MPa, an RMSE of 17.17 MPa, a MAPE of 3.06%, an NRMSE of 3.14%, and an R² of 0.910, indicating good agreement between the numerical and experimental results. The largest error of 8.66% occurred at a strain of 2.279%, specifically in the transition region from the yield plateau to strain hardening. Conversely, the best agreement, with a MAPE of 0.90%, was obtained in the strain range of 5.0–7.5%. These results indicate that the FEM model used adequately represents the strain-hardening behavior of SM490, although evaluations in the elastic and yield plateau regions still require numerical results with higher resolution.
Copyrights © 2026