[home] [Personal Program] [Help]

---

Session: Session 8: Ultrasonic welding
Session starts: Thursday 16 April, 09:40
Presentation starts: 09:40
Room: Main

---

David Lohuis (University of Stuttgart - Institute for aircraft design)
Stefan Carosella (University of Stuttgart - Institute for aircraft design)
Michael May (University of Stuttgart - Institute for aircraft design)
Peter Middendorf (University of Stuttgart)

Abstract:
Joining carbon fiber composites with ultrasonic welding can still be a challenge, due to unpredictable weld parameters. Parameters can vary even within the same part, emphasising the need for a prediction model, novel measurement techniques and improved control algorithms. The authors propose a capacitive measurement system for real time quality control and as a basis for future control algorithms. The method exploits the dielectric properties of the polymer matrix and its correlation to temperature and viscosity. Quantitative data can be gathered about all three parameters by setting up an electric field across the bonding area. Since the sonotrode and the ambos of the welding equipment consist of metal, both can act as the electrodes, leading to a measurement system suitable for industrial application. As a proof of concept, a prototype with an ambos as the lower and an external steel sheet as the upper electrode is tested in a robotic welding station. The experiment consists of a KUKA robot, welding equipment provided by MS Ultrasonic and an Agilent E4980A LCR-Meter. The LCR-Meter applies 500kHz AC voltage to two electrodes around the specimen to determine the amplitude and phase difference between the voltage and current signal. These differences are translated into a capacity value that changes during the weld. Specimen consist of two 25x25x3 𝑚𝑚3 carbon fiber/Elium pieces and a PMMA energy director. Capacity curves can be well reproduced and show behaviour just as expected based on polymer physics. The data indicates a capacity rise due to a superposition of the part displacement and temperature changes. Maximum temperature is assumed to be reached at maximum capacity, where the hyperbolic decline correlates to the cool down phase. With further modelling it will most likely be possible to determine quantitative changes in temperature and viscosity, leading to a more accurate understanding of the process.