Increasing levels of transportation and exploratory activities in the High North increase the signiﬁcance of ice-capable ship designs, and the demand for them. This demand covers a wide range of ship types; such as tugs, vessels for search and rescue (SAR), patrol boats, military vessels, cruise ships, and merchant ships. Both the economically driven preference for operations in the Arctic over operations in a warmer climate, and the safety of the operations, require adequate performance prediction methods. The capability of model-scale ice and its availability and advantages in handling compared to sea ice spurred to the decision to investigate its material behavior to develop a numerical model. This model serves as a corner-stone towards a numerical ice tank and provides insight into the mechanical behavior of model-scale ice. Therefore, systematic ice property tests were conducted with the model-scale ice of Aalto University to deﬁne the material behavior. The model-scale ice is ﬁne grained (FG) and doped with ethanol to artiﬁcially weaken the material. The experiments investigate the behavior until failure in tension, compression and bending. Furthermore, the elastic modulus is determined by ice sheet deﬂection experiments and the grain-size is measured. The stress plane that is investigated is orthogonal to the vertical (thickness) coordinate and is the same as the one in which stresses occur when ships interact with ice. On the basis of the experiments, the mechanics and the constitution of the model-scale ice are investigated to deﬁne a suitable material model and its parameters. It was found that a damage based elasto-plastic material model represents the behavior of the Aalto model-scale ice well. The numerical model accounts explicitly for ﬂaws in the model-scale ice, comprised of voids ﬁlled with liquid and air, which are randomly distributed. It is found that the random distribution of ﬂaws enables the reproduction of the variation in experimentally observed failure patterns and aﬀects the response forces. Furthermore, the cantilever beam bending experiments and their simulation reveal that the gradual change of ice properties over thickness has to be modeled to represent the experimentally measured axial stiﬀness and ﬂexural stiﬀness in the same model. Ultimately, the model-scale ice is demonstrated to be a functionally graded material which is capable of representing tensile, compressive and ﬂexural failure modes.Additionally, the development of the numerical model of the Aalto model-scale ice provides a deeper insight into its mechanical deformation processes. The material behavior that is found reveals that Cauchy similitude in scaling cannot be applied, because the model-scale ice of Aalto University is on micro scale not a purely elastic material. Consequently, model-scale ice consumes more energy prior to bending failure than a material complying Cauchy similitude.
|Tila||Julkaistu - 2016|
|OKM-julkaisutyyppi||G5 Tohtorinväitöskirja (artikkeli)|