Process Data Driven Advancement of Robot-Based Continuous Ultrasonic Welding

Abstract

High volume aerospace programs require efficient, flexible and robust welding technologies. Since continuous ultrasonic welding is a highly transient process that depends on a number of processing parameters, such as weld energy, amplitude, travel speed and consolidation pressure, process monitoring is crucial for the welding result. Furthermore, it is not possible to look into the welding zone during the welding process. In order to mature this technology an in-line process monitoring has been setup that collects all relevant parameters and uses a Neural Network to predict the welding quality directly after the welding process.