One-Dimensional Viscoelastic Simulation of Ice Behaviour in Relation to Dynamic Ice Action

One-Dimensional Viscoelastic Simulation of Ice Behaviour in Relation to Dynamic Ice Action

Yazarov, M.

Stive, M.J.F. (mentor)
Høyland, K. (mentor)

Civil Engineering and Geosciences

Hydraulic Engineering

Coastal and Marine Engineering and Management (CoMEM)

2012-06-21

The goal of this work was to simulate the behaviour of ice under cyclic loading. The linear viscoelastic Kelvin, Maxwell and Burgers models were implemented in Matlab by means of The Boltzmann superposition principle. This method was verified using cases which could be represented by analytical solutions. During the extensive sensitivity analysis carried out, it was determined that the ratio n/(f?) influences the shape of the strain response curve in a creep test for all of the models considered in this study. With the knowledge of parameter n/(f?), preliminary conclusion on the result of the test can be drawn. The change of the second modulus is defined by the Kelvin unit which determines the delayed elastic deformations. The time required for these deformations to fully developed is given by 5?2. The changes in the value of the second modulus occurs only before this value is attained. However in linear viscoelastic Burgers model, this change is relatively small . This is because the Maxwell unit in the Burger model has a major influence on the shape of the stress-strain curves. Available literature was used to obtain the input parameters for Burgers model for ice. They were obtained by calibration of the model data with the experimental data presented therein by producing a best fit curve. Since this study utilized a non-linear model to replicate the experimental data, it was also used to compare the results of a linear model with a non-linear model. The region conforming to the delayed elastic deformation was not represented accurately by the linear model. Uniaxial cyclic compression test performed at UNIS was modeled using the linear viscoelastic Burgers model using the scaled down input parameters obtained through calibration with fresh ice creep test to enable their application to this case. The viscoelastic model was unable to completely agree with the findings from the test carried out at UNIS.

linear viscoelastic model
ice
dynamic ice action
cyclic loading

http://resolver.tudelft.nl/uuid:f1d2e910-6b53-4b88-8e85-ad82d1895280

Student theses

master thesis

(c) 2012 Yazarov, M.