High amplitude, low frequency, vibrational parallel manipulation and assembly of mechanical devices

Rimai, Benjamin E. Ph.D., Purdue University, August 2013. High Amplitude, Low Frequency, Vibrational Parallel Manipulation and Assembly of Mechanical Devices. Major Professor: Raymond J. Cipra, School of Mechanical Engineering. Assembly of micro-scale and millimeter-scale devices poses unique challenges. Typical assembly and part-handling methods become infeasible for small components because of limitations on the accuracy of macro-scale robots and because of the adhesive forces which cause the components to adhere to robot end-effectors and other tools. Strategies for manipulating small components without direct grip-based manipulation have been previously investigated in the electrical engineering and micro electromechanical systems fields. The work presented in this dissertation followed work which demonstrated that groups of micro-scale components placed on a vibratory bowl feeder can be successfully separated into a stream of single items in a process hereafter called vibrational manipulation. This dissertation presents work that has studied the vibrational manipulation process experimentally, developed and validated a spatial simulation which was used to model the vibrational manipulation of many parts in parallel, developed and analyzed methods for assembling devices using vibrational manipulation, and demonstrated the vibrational parallel assembly of a four-bar mechanism. In order to transform vibrational manipulation into an assembly process, a better understanding of the vibrational manipulation process was first required. Experimental characterization of the vibrational manipulation occurring in a vibratory bowl feeder for 170µm diameter metallic pins was conducted. Effects of pin length and vibrational frequency and amplitude on the distribution of pin capture times were analyzed. Concurrently, magnetic fields were found to improve the mean capture times by as much as 80%. A spatial simulation, which uses rigid body dynamics, impulse based contact resolution, and spherical meshing to approximate complicated component geometry, was developed to model the vibrational manipulation and vibrational assembly processes. Very good agreement was observed between the simulation and the aforementioned experiments. The simulation was used to predict the behavior of the simple vibrational assembly of a washer onto an axle. The process was also examined experimentally and, once again, very good agreement was observed between the simulation and experiments. The simulation was finally used to predict the effects of component geometry, operating conditions, material properties, and other characteristics on the performance of vibrational parallel assembly. Strategies for using the simulation to rapidly improve the design of components for vibrational parallel assembly were developed. Finally, the design of a four-bar mechanism was improved for vibrational parallel assembly as a demonstration of the design improvement process. Vibrational parallel manipulation and assembly appears to be a very feasible and high performance alternative to typical macro-scale assembly techniques when the components to be manipulated or the devices to be assembled are in the millimeter or sub-millimeter size ranges