Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Mechanical and Aerospace Engineering

Committee Chair(s)

Brent E. Stucker


Brent E. Stucker


Thomas Fronk


Leijun Li


David Geller


David Britt


A dislocation density-based constitutive model has been developed and implemented into a crystal plasticity quasi-static finite element framework. This approach captures the statistical evolution of dislocation structures and grain fragmentation at the bonding interface when sufficient boundary conditions pertaining to the Ultrasonic Consolidation (UC) process are prescribed.

The hardening is incorporated using statistically stored and geometrically necessary dislocation densities (SSDs and GNDs), which are dislocation analogs of isotropic and kinematic hardening, respectively. Since the macroscopic global boundary conditions during UC involves cyclic sinosuidal simple shear loading along with constant normal pressure, the cross slip mechanism has been included in the evolution equation for SSDs. The inclusion of cross slip promotes slip irreversibility, dislocation storage, and hence, cyclic hardening during the UC. The GND considers strain-gradient and thus renders the model size-dependent. The model is calibrated using experimental data from published refereed literature for simple shear deformation of single crystalline pure aluminum alloy and uniaxial tension of polycrystalline Aluminum 3003-H18 alloy.

The model also incorporates various local and global effects such as (1) friction, (2) thermal softening, (3) acoustic softening, (4) surface texture of the sonotrode and initial mating surfaces, and (6) presence of oxide-scale at the mating surfaces, which further contribute significantly specifically to the grain substructure evolution at the interface and to the anisotropic bulk deformation away from the interface during UC in general. The model results have been predicted for Al-3003 alloy undergoing UC. A good agreement between the experimental and simulated results has been observed for the evolution of linear weld density and anisotropic global strengths macroscopically. Similarly, microscopic observations such as embrittlement due to grain substructure evolution at the UC interface have been also demonstrated by the simulation.

In conclusion, the model was able to predict the effects of macroscopic global boundary conditions on bond quality. It has been found that the normal pressure enhances good bonding characteristics at the interface, though beyond a certain magnitude enhances dynamic failure. Similarly, lower oscillation amplitudes result in a lower rate of gap closure, whereas higher oscillation amplitude results in an enhanced rate of gap relaxation at the interface. Henceforth, good bonding characteristics between the constituent foils are found at an optimum oscillation amplitude. A similar analogy is also true for weld speed where the longitudinal locations behind the sonotrode rip open when higher weld speeds are implemented in the UC machine, leading to lower linear weld density and poor bonding characteristics.




This work made publicly available electronically on September 1, 2011.