Ultrasonic welding has been widely used in joining plastic parts since it is fast, economical, and suitable for automation. It also has great potential for joining thermoplastic composite structures in the aerospace and automotive industries. For a successful industrial application of ultrasonic composite welding, it is necessary to have effective weld quality prediction technology. This paper proposes a model for weld quality prediction by establishing a correlation between ultrasonic wave transmission and welding process signatures. The signatures, welding power, and force are directly related to the weld quality. This model is used to predict the weld quality with three contact conditions and validated by experiments. The results show that the quality model performs well when a centralized and consistent contact condition is achieved. The model provides a process physics-based solution for the online weld quality prediction in ultrasonic welding of carbon fiber composite.

Energy directors (EDs) have been widely used in ultrasonic welding (UW) of polymers and polymer-based composites. They help concentrate the welding energy and localize the weld at the location where the EDs are present. However, the utilization of EDs increases manufacturing cost and time, especially for complex parts and structures. This paper presents a method for UW of carbon fiber reinforced composite without using EDs. A reusable annular clamp (called a blankholder) is used as part of the weld tool to apply a variable force (called blank holding force) on the composite sheets during the UW. The effect of the blank holding force (BHF) on the weld formation is investigated. The results show that the duration of the BHF had significant impact on the weld formation. There is a critical duration with which a localized weld can form. Suitable durations of BHF at different levels of welding energy are determined by experiments. The main function of the BHF is to create an initial melting area by improving the contact condition. The initial melting area will act as an ED to concentrate the welding energy, and therefore, promotes the formation of a localized weld.

High temperature, short welding time, and low relative motion generate high bond quality in ultrasonic metal welding (USMW). Friction is considered to be the main heat source during USMW. Hence, a comprehensive and accurate understanding of friction heating has become particularly valuable for designing USMW processes and devices. However, stick, slip, and separation states may appear alternately in the welding zone between superimposed workpieces during USMW vibrations; hence, a strong nonlinear process is created. Furthermore, the structural dynamics and the heat transfer are highly coupled because material properties depend on temperature. In this research, we propose a fast and accurate numerical methodology to calculate the friction heating through a multiphysical approach integrating a nonlinear contact model, a nonlinear structural dynamics model, and a thermal model. The harmonic balance method and the finite element method are utilized to accelerate the simulation. Several experiments were performed with aluminum and copper workpieces under different clamping forces and vibration amplitudes to confirm the presented numerical method, resulting in a good match.

This paper described the effects of prestraining and annealing on plastic anisotropy (r-value) of aluminum alloy 5182-O sheets including two prestrain paths and two annealing conditions. During the prestraining and annealing processes, r-value changed depending on prestrain paths and annealing conditions. Although there were slight changes of the normal anisotropy coefficient, $r¯$, during prestraining and annealing processes, the planar anisotropy coefficient, $Δr$, increased significantly, especially for the uniaxial prestrain condition. This could accelerate the development of earing during a sheet forming operation. Also, the corresponding sheet textures in rolling direction (RD)/TD plane after prestraining and annealing processes were observed through electron backscatter diffraction (EBSD) analysis to explain the r-value changes, where the viscoplastic self-consistent (VPSC) model was used to correlate the determined texture to measured r-values. It is found that the sheet texture also had significant changes relating to the prestrain paths and annealing conditions resulting in varied r-values.