This study aims to reveal the motion characteristics and failure mechanisms of a bergy bit, as well as the hull structural response, during interactions between local hull structures with varying stiffness and a bergy bit.

The arbitrary Lagrangian-Eulerian (ALE) method is used to conduct numerical simulations of the ship-ice collision process. Finite element models are established to simulate collisions between local hull frame structures with different stiffness levels and a bergy bit. The validity of the numerical method and model is verified by comparison with experimental results. The effects of different impact positions, impact speeds, and frame stiffness on the motion characteristics of the bergy bit and hull structural damage are then analyzed.

The results demonstrate that the proposed numerical simulation method for collision between local hull frame structures and the bergy bit can effectively reproduce the experimental process. The local impact force peak from the numerical simulation shows an error of less than 15% compared to the experimental data. Structural stiffness plays a critical role in the collision response: the local impact force generated by the elastic frame model is 26.0% lower than that of the rigid frame model. Both impact speed and impact position significantly influence the ice impact force and the damage deformation of elastic structures. When structural deformability is considered, the peak impact force and corresponding deformation are significantly reduced.

This study provides a reference for applying numerical simulation methods to the optimal design of local ice-resistant structures for polar ships.

The cohesive zone length is the sum of the cohesive element length at the failure edge and the lengths of the other cohesive elements connected to it. The cohesive zone length determines the maximum size of the mesh. Therefore, the accurate estimation of cohesive zone length and reasonable mesh division are important factors affecting calculation accuracy.

Based on several J-integral assumptions and existing research results, a modified function on length thickness ratio value is added to the original formula. The modified formula is then applied to an ice mechanics model. Based on the finite element method, a model of a double cantilever beam is established to verify the accuracy of the formula through comparison with the experimental results.

The results show that there must be at least four cohesive elements in a cohesive zone length to describe the fracture process accurately. This conclusion is also applied to the numerical simulation of a three-point bending experiment. The error of the limit load is 2.9%, and that of the fracture point is within a reasonable range.

It is concluded that the modified cohesive zone length formula is more suitable for ice materials.