Ultrasonic welding (USW), a solid state joining technology, has been used to bond aluminum alloys commonly used in the automotive industry. Bonding occurs due to USW's high frequency (∼20 kHz) in-plane vibration of sample interfaces while being held under moderate clamp pressure normal to the plane of vibration. Vibration and clamp pressure are transmitted into bond formation via contact with a weld-tip. To better understand how weld-tip geometry affected bond formation, experiments were conducted to quantify how tip geometry influenced plastic deformation characteristics between fully welded coupons of 0.9mm thick AA6111-T4 aluminum alloy. Weld-interface microstructure features were documented by optical microscopy and features quantified in a 19 point matrix. Correlation between microstructure features, such as rolling-wakes, and resulting weld bond strengths of more than 3.0kN is made. Weld zone microstructure features appear to result from deformation at and severe migration of the original weld interface during USW. To confirm this hypothesis, intrinsic and extrinsic markers were employed to monitor weld interface deformation characteristics. Various physical and analytical techniques were used in conjunction with these markers to show that joining of "like" and "dislike" aluminum samples is achieved through mechanical mixing of mating interfaces and not by elemental diffusion. It is also hypothesized that severe deformation of the original interface would result in areas of high residual strain within a formed weld zone. To investigate this and the influence that tip geometry may have on residual strain, fully welded samples were annealed at 500°C for a controlled period of time and recovery, recrystallization and grain growth characteristics were monitored. In all welds, initial recrystallization and grain growth occurred at the outer ends of weld zones and along weld interfaces where the most turbulent mixing and grain size reduction was observed. Similarity in how all welds responded to annealing indicates that the tip geometries investigated had little influence on resulting weld formation. This claim is further supported by lap-shear failure load data for welds made with these tips being within statistical error of each other.