A genetic algorithm-based calibration procedure for hysteresis loops in timber structures

Hysteresis loops of timber structures often exhibit complex behaviors such as pinching and stiffness degradation, which are challenging to model accurately. This study proposes an automated calibration procedure for hysteresis loops of timber components and systems using the HystereticSM material model in OpenSees combined with a genetic algorithm (GA). The HystereticSM model was selected for its ability to capture pinching, strength, and stiffness degradation while requiring relatively few input parameters, thereby simplifying the calibration process. In the proposed framework, GA optimization is employed with energy dissipation as the objective function to efficiently match experimental hysteresis behavior. To demonstrate the robustness of the proposed procedure, it was applied to the calibration of timber connections fabricated with three types of fasteners (timber rivet, bolt, and nail) and a light wood-frame shear wall. Results showed that the calibrated models reproduced the experimental hysteresis loops with small errors in both peak force and energy dissipation, confirming the effectiveness of the approach across different hysteresis characteristics. In addition, practical guidelines for parameter initialization and recommendations for the calibration process are provided, supporting broader application of the framework. These findings highlight the potential of the proposed procedure as a standardized and efficient approach for calibrating hysteresis models of timber structures and establish a foundation for extending it to other different hysteresis characteristics in future work. ● Developed a novel Genetic Algorithm (GA)-based procedure to automate calibration of hysteresis loops in timber components. ● Demonstrated that the HystereticSM model accurately captures pinching, strength, and stiffness degradation. ● Achieved good agreement with experimental results for rivet, bolt, nail connections, and a light wood-frame shear wall. ● Confirmed that GA with energy dissipation as the objective function provides efficient and reliable optimization. ● Provided practical guidelines for initializing parameters, enabling broader application to diverse hysteresis behaviors.