P. A. Fleming, D. M. Wilkes, G. E. Cook, D. DeLapp, T. Bloodworth, D. Lammlein, Welding Automation Laboratory Vanderbilt University, Nashville, TN; T. J. Lienert, M. Bement, Los Alamos National Laboratory, Los Alamos, NM; A. M. Strauss, Vanderbilt University, Nashville, TN; T. Prater, Vanderbilt University, Clarksville, TN
Friction Stir Welding (FSW) is a relatively new and effective joining technique which is becoming used more and more in a number of industries including: aerospace, automotive, locomotive and maritime. This is largely due to the advantages FSW has over other techniques including high weld quality, low power and heat and no fumes or spatter. There are several disadvantages to the process however, including the reliance of correct weld parameter selections and pin tool design, as well as the condition of said tool. If the tool was to wear unknown to the operator, the system might produce bad welds, even if the weld parameters are correct. In this paper we investigate techniques based on acoustic emissions, thermography and force detection in order to automatically detect tool wear. The tool selected is of a material chosen to wear at a rate faster than a normal tool would.
Summary: Friction Stir Welding (FSW) is a relatively new and effective joining technique which was developed by The Welding Institute in 1991. FSW
is used more and more in a number of industries
including aerospace, automotive, locomotive, and maritime.
FSW has a number of advantages over older welding techniques including no consumables,
low heat, no porosity or spatter, and good mechanical properties. However, it also has some
challenges which need to be overcome. One of these challenges is the automatic detection
of tool wear.
The cylindrical tool pin used in FSW affects the quality of welds both by its geometry and by
its condition. If a tool were to wear without detection, the process could start to produce
poorer quality welds until the wear is discovered.
In our work, we have devised an experiment where the FSW tool is fashioned out of a metal
which will wear quicker than the typical steel tool would have. The tool is used repeatedly
in an FSW process while acoustic emissions are recorded as well as force signals from
a dynomometer and thermal imagery. The condition of the tool is also recorded after each
run. Finally, at the end of the experiment, metallographic and mechanical tests are done
on each sample.
The goal of this work is to discover a relationship between one or more of the signals collected and the tool condition and weld quality. All of these signals are available in process, so this work would aid in preventing assembly lines from losing quality and wasting both time and money by discovering tool wear as it happens.
