Failure analysis of a 30kHz ultrasonic welding transducer

Ultrasonic welding is used across variety of industries to join together thermoplastic materials. During ultrasonic welding, high frequency mechanical vibrations and compressive pressure are applied to the plastic materials. This creates intermolecular friction within the plastic which raises the temperature high enough to reach the melting point. During high volume production in manufacturing environments, the ultrasonic welding system component which fails most often is the transducer, which is responsible for creating the vibration. The transducer converts electrical energy into mechanical energy by means of a polarized piezoelectric PZT (lead zirconate titanate) polycrystalline ceramic material using high frequency voltage. The transducer is composed of round PZT disks attached to a titanium machined body by means of a central bolt. The bolt creates compressive pre-stress which prevents any tensile stresses within the brittle crystals during vibration and ensures perfect coupling between all components. The piezoelectric PZT crystals behave according to linear coupled electrical and mechanical equations. The transducer assembly is vibrated near the parallel resonant frequency to maximize efficiency and amplitude output. The purpose of this thesis was to analyze 30 kHz multilayer PZT transducer failures caused by excessive amplitude (stress) and recommend ways to increase the lifespan. The 30 kHz transducer design being studied is produced by Herrmann Ultraschalltechnic GmbH and is currently used in production facilities around the world. Based upon warranty and field data, this size transducer fails twice as often as the 20 kHz and 35 kHz transducers. The analysis was based upon studying overall converter design practices and creating a Finite Element model to understand the stress magnitude and distribution to predict the point of failure. Laboratory experiments were then performed under high amplitude conditions to increase the stress levels until failure occurred. The Finite Element model predicted the current design should be capable of withstanding up to 208% of nominal production level amplitudes (stress) before failure due to uncoupling between the ceramic crystals and the adjacent conducting plates. During the laboratory experiments the failures occurred between 179%-244%, however the failure mode did not correlate with the failure predicted by the Finite Element model. All of the failures during the experiment resulted from electrical arcing at the inner diameter of the bottom ceramic disks. Furthermore, these laboratory failure modes were not consistent with those normally observed in production environments where fracture and shifting of crystals are the dominant modes. We can conclude that the electrical load used in the experiments did not accurately represent production welding conditions, which means it could not be used as a comparison against the Finite Element model. In order to increase the lifespan of the transducer additional testing must be performed. First and foremost would be new tests which accurately reproduce production welding conditions. Then these results could be tested against the Finite Element model. Once this was completed, if the models accuracy was verified, improvements could be made and tested with the Finite Element model to try and increase the lifespan.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Mechanical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2011

Author dc:creator

• Hampton, Christopher P.

Contributors dc:contributor

• Hsia, K. Jimmy

• Ultrasonics
• welding
• failure analysis
• lead zirconate titanate (PZT)
• transducer

Statement dc:rights

• Copyright 2010 Christopher Hampton

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/18653

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/18653