As new materials are developed and further concerns on green alternatives and serviceability arise, understanding material behavior during the entire span of their lifetime becomes crucial to engineering applications. Moreover, many problems display a significant dependence to time and loading effects which, by varying across multiple time scales, require material models that incorporate these effects into any valid characterization and prediction. This dissertation aims at proposing a new approach to analyze and predict viscoelastic materials that deteriorate during multiple loading conditions. The model is constructed from mechanical and mathematical basis while satisfying physical laws.
In this work, the proposed constitutive law is used for the analysis of the mechanical properties of ice. The mechanical behavior, biaxial envelop and multiple loading types demonstrate the validity of the model when compared to experimental results and other ice models available in the literature. A rigorous calibration scheme for the viscoelastic and damage parameters is also presented.
Moreover, as material deterioration or damage is modeled in standard Finite Elements software, it is commonly known that computational results can be dependent on the spatial discretization or mesh. That is, damage zone and energy dissipation are dependent on the selection of the mesh yielding a disappearing damage zone and energy dissipation upon refinement. This non-physical behavior is corrected by the novel regularization approach proposed in this document, which introduces a length scale of the material and produces results that are no longer sensitive to the mesh selection.
The nonlocal damage model is finally used in the analysis of asphalt concrete viscoelastic behavior and cracking prediction. As presented in the ice case, a rigorous calibration approach is presented first followed by the validation to experimental data available in the literature under different loading conditions.
The coupled viscoelastic and damage model is compared to other model and their Finite Elements implementations are highlighted in terms of computational efficiency. A nonlinear coupled system for solving this problem is programmed as a User Element in a commercial Finite Element analysis software.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8MS458Q |
| Date | January 2017 |
| Creators | Londono Lozano, Juan Guillermo |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.002 seconds