Experimental and numerical studies related to the coupled behavior ofice mass and steel structures during accidental collisions
MetadataShow full item record
- Institutt for marin teknikk 
More and more people are attracted to go North. To design a fixed or a floating structure for Arctic conditions, to set operational limits or to specify certain safety requirements, one must rely on limited knowledge, data and experience. Several marine incidents have shown that absolute safety does not exist, including the holing and sinking of the cruise ship Explorer in 2007 and the holing of the tanker Nordvik in 2013. The concept of structural analysis in which the structure is allowed to deform inelastically due to rare and extreme actions (e.g., a collision with an ice mass) is of crucial importance in engineering. In the literature, this concept is often referred to as Accidental Limit State (ALS) design. With the ALS approach, adequate precautions against scenarios outside of ice class requirements can be established for both ships and offshore structures. Understanding ice behavior and structural performance during collisions allows us to develop predictive tools that form the core of design. Although the topic of ice-mass collisions with structures is not new, the present study is one of a few to address the experimental aspects of ice-structure collisions in which the impacted structure undergoes irreversible deformation together with ice crushing. Inelastic behavior of ice and of steel panels has been studied in experimental tests and numerical simulations. This study is a continuation of research performed in the Department of Marine Technology at the Norwegian University of Science and Technology and is a component of a larger research project on the prediction of loads exerted by various types of ice (including icebergs) on floating structures. This study provides an overview of existing methods for the analysis of ALS due to ice actions, including different ice failure/yield criteria that might be considered for ALS. It was acknowledged that ice behavior under rapid compressive loading is not fully understood and that there is a complete lack of direct, full-scale field measurements of ice forces on structures that deform plastically under ice actions. These facts served as motivation for carrying out three main types of experiments: 1. Small-scale indentation tests on confined and unconfined laboratory-grown freshwater granular and columnar S2 ice at temperatures of –10°C and –40°C. In these tests, the events that are observed in full- and medium-scale tests were reproduced with remarkable similarity. 2. Drop tests of ice blocks onto stiffened steel panels. 3. Laboratory impact tests of ice blocks onto stiffened panels in water. The tests were conducted at a scale approaching full-scale ice impact using laboratorygrown freshwater ice blocks and panels of varying rigidity. The first set of tests served as a basis for validation of a constitutive ice model. The parameters of the model were measured in experiments that are independent of the tests used to validate the accuracy of the ice model. To assess whether the model is applicable for describing ice crushing, the energy absorption from numerical simulations was compared with that found experimentally. This comparison showed that the model is, in fact, able to capture the energy absorption during ice crushing. The results of the second and third sets of experiments were also used for validation of the ice model. The study indicated that both the plastic methods of analysis and numerical simulations could be used to assess damage to the structure due to ice actions. A comparison of the structural damages from laboratory tests and numerical simulations showed that the numerical simulations slightly overestimate the maximum damage, and the difference between the estimates increases for the tests conducted in water. To improve the ice model, several modifications have been suggested and include the generalized version of a strain-based, pressure-dependent (or triaxiality-dependent) failure criterion, and a combination of a nonlinear finite element method and smooth particle hydrodynamics for modeling of the ice fragmentation and contact-pressure patterns. An interpretation of the model parameters was also suggested. An advantage of the ice model is that it is rather simple and does not require sophisticated tests for validation of the material parameters.