AH36의 GISSMO 손상 모델 구성 및 낙하 실험 기반 검증 연구

Abstract

In ship structural analysis, evaluating damage is essential to ensure the safety and reliability of vessels under various operational conditions. Ships are often subjected to unpredictable and complex loading scenarios, necessitating advanced methodologies for accurately predicting material failure behavior. The GISSMO (Generalized Incremental Stress State-dependent Model) is widely used in various engineering fields—including automotive crash analysis, metal forming, and extrusion—to predict material damage and fracture. The GISSMO model was calibrated for AH36 steel specimens with different central geometries to represent a range of stress states. Full-field strain distributions obtained via Digital Image Correlation (DIC) were used to extract stress–strain curves, which served as input for finite element models developed using ANSYS/LS-DYNA. GISSMO parameters were then optimized using LS-OPT to ensure agreement with the experimental results. In addition, high-speed tensile tests were conducted to assess strain-rate sensitivity and further refine the model’s predictive capability under dynamic conditions. The calibrated model was validated through both numerical and experimental drop tests. Compared to a conventional fracture strain model, the GISSMO-based simulation showed significantly better agreement with experimental results, particularly in predicting fracture initiation and damage propagation. These findings demonstrate the potential of GISSMO as a reliable damage model for ship collision simulations involving complex and high-energy impact conditions.