Date of Award

8-2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Mechanical Engineering

Advisor

Grujicic, Mica

Committee Member

Joseph , Paul

Committee Member

Omar , Mohammed

Committee Member

Singh , Rajendra

Abstract

High strength aluminum alloys with superior blast/ballistic resistance against armor piercing (AP) threats and with high vehicle light-weighing potential are being increasingly used as military-vehicle armor. Due to the complex structure of these vehicles, they are commonly constructed through joining (mainly welding) of the individual components. Unfortunately, these alloys are not very amenable to conventional fusion based welding technologies (e.g. Gas Metal Arc Welding (GMAW)) and in-order to obtain high-quality welds, solid-state joining technologies such as Friction Stir Welding (FSW) have to be employed. However, since FSW is a relatively new and fairly complex joining technology, its introduction into advanced military vehicle underbody structures is not straight forward and entails a comprehensive multi-prong approach which addresses concurrently and interactively all the aspects associated with the components/vehicle-underbody design, fabrication and testing. One such approach is developed and applied in the present work. The approach consists of a number of well-defined steps taking place concurrently and relies on two-way interactions between various steps. In the present work, two of these steps are analyzed in great detail: (a) Friction Stir Welding process modeling; and (b) Development and parameterization of material models for the different weld-zones.
Within the FSW process modeling, interactions between the rotating and advancing pin-shaped tool (terminated at one end with a circular-cylindrical shoulder) with the clamped welding-plates and the associated material and heat transport are studied computationally using a fully-coupled thermo-mechanical finite-element analysis. To surmount potential numerical problems associated with extensive mesh distortions/entanglement, an Arbitrary Lagrangian Eulerian (ALE) formulation was used which enabled adaptive re-meshing (to ensure the continuing presence of a high-quality mesh) while allowing full tracking of the material free surfaces/interfaces. To demonstrate the utility of the present computational approach, the analysis is applied to the aluminum-alloy grades, AA5083 (a solid-solution strengthened and strain-hardened/stabilized Al-Mg alloy) and AA2139 (a precipitation hardened quaternary Al-Cu-Mg-Ag alloy). Both of these alloys are currently being used in military-vehicle hull structural and armor systems. In the case of non-age-hardenable AA5083, the dominant microstructure evolution processes taking place during FSW are extensive plastic deformation and dynamic re crystallization of highly-deformed material subjected to elevated temperatures approaching the melting temperature. In the case of AA2139, in addition to plastic deformation and dynamic recrystallization, precipitates coarsening, over-aging, dissolution and re-precipitation had to be also considered. To account for the competition between plastic-deformation controlled strengthening and dynamic-recrystallization induced softening phenomena during the FSW process, the original Johnson-Cook strain- and strain-rate hardening and temperature-softening material strength model is modified using the available recrystallization-kinetics experimental data. Lastly, the computational results obtained in the present work are compared with their experimental counterparts available in the open literature. This comparison revealed that general trends regarding spatial distribution and temporal evolutions of various material-state quantities and their dependence on the FSW process parameters are reasonably well predicted by the present computational approach.
The introduction of newer joining technologies like the so-called Friction Stir Welding (FSW) into automotive engineering entails the knowledge of the joint-material microstructure and properties. Since, the development of vehicles (including military vehicles capable of surviving blast and ballistic impacts) nowadays involves extensive use of the computational engineering analyses (CEA), robust high-fidelity material models are needed for the FSW joints. A two-level material-homogenization procedure is proposed and utilized in the present work in-order to help manage computational cost and computer storage requirements for such CEAs. The method utilizes experimental (microstructure, micro-hardness, tensile testing and X-ray diffraction) data to construct: (a) the material model for each weld zone; and (b) the material model for the entire weld. The procedure is validated by comparing its predictions with the available experimental results and with the predictions of more-detailed but more costly computational analyses.

Share

COinS