Technical Knowledge
- Industrial Member Reports

Insights into Friction Stir Welding

TWI Technology Briefing 991/2011

By K A Beamish and M J Russell

Background

Process management for friction stir welding (FSW) is a developing area. A range of monitoring and process management approaches are currently applied by industrial users, as described in the first report from this project (Blignault, 2008). An entry-level system is capable of monitoring the core process variables of tool rotation speed, tool position (in x, y, and z), and welding speed. Force measurement has been of interest for some time and maintenance of a set z-force provides one of the principal methods of process control, the other (dominant) method being control of z-axis position. Machine manufacturers are now integrating enhanced process monitoring systems into their equipment. Advanced monitoring and control of FSW has resulted in the emergence of in-process monitoring as a potential quality control method.

This Core Research Programme (CRP) project is focused on the measurement and interpretation of multiple process parameters such as tool forces, temperature and torque by means of a precision in-situ monitoring system. The data generated is compared with that obtained from basic and enhanced process monitoring systems. Force, torque and temperature traces can be used to establish a reference against which subsequent welds can be compared. Characteristic FSW flaws were intentionally generated to demonstrate the identification of potential quality issues. In future such data may be used as part of Quality Control procedures and could ultimately be incorporated into machine control systems. An assessment of two representative tool designs, the MX-Triflute™ and the MX-Triflat™ has been made over a range of welding conditions. The in-situ monitoring system has been used to identify the characteristic force footprints of the individual design elements of these tools which gives a quantitative basis for future tool development.

Objectives

Investigate the action of representative FSW tool designs and to assess the influence of individual design features.
Investigate and demonstrate the potential for applying an integrated FSW process monitoring system as an on-line Quality Assurance tool.

Experimental Approach

Two FSW tool designs, an MX-Triflute™ and an MX-Triflat™, have been separated into their component forms. The MX-Triflute™ comprises a plain tapered cone, a thread feature, and a Triflute™ feature. The MX-Triflat™ comprises a plain tapered cone, a thread feature and a Triflat™ feature. The weld signature of the individual elements was investigated separately using FSW tools of each elemental form. Representative weld flaws including root flaws and voids were also generated through appropriate adjustment of welding parameters and tool profiles. The quality of the welds produced was assessed by a combination of visual examination, metallographic examination and bend testing where appropriate. All welds were made in 6mm thickness AA5083-O aluminium alloy.

The welds produced in the current programme have been assessed by three separate monitoring systems. The first, generic, weld monitoring system is as-supplied by the machine manufacturer. The second system has alternative torque and down force measurement methods and the capability to record temperature. The third, the Artemis advanced monitoring system, applies separate techniques to record torque, down force, traverse force and tool bending forces, it also records temperature. The Artemis system has been built during this CRP project.

Results and Discussion

The capabilities of the three systems are compared and recommendations are made regarding the selection of appropriate weld quality indicators. The Artemis system has also provided an insight into the working mechanisms of various tool designs. The relationship between probe features, process data and weld microstructure is discussed.

The current work has examined weld signatures of sound welds and of those containing flaws generated as a result of controlled alterations to tool geometry, workpiece geometry or welding conditions. Some process variations resulted in identifiable changes to several of the monitored process responses (torque, down force, tool temperature, traverse force and tool bending forces), others were characterised by changes in one process response only. The magnitude of parameter variation that results from an identifiable weld flaw is noted and recommendations are made regarding appropriate weld quality indicators for industrial users.

The Artemis system generates polar plots of tool bending forces. There is a correlation between the outline of the polar plot of tool bending forces (the force footprint) and the presence of a weld flaw. The polar plot of tool bending forces corresponding to a good quality weld has a smooth outline. The presence of volumetric flaws (voids) causes the outline to become disrupted and irregular.

Main Conclusions

An advanced weld monitoring system, Artemis, has been designed and commissioned which is capable of the accurate determination the bending forces acting around the circumference of the tool. The following conclusions have been reached from this work:

The Artemis advanced monitoring system can detect the presence of weld flaws, including volumetric flaws (voids), via the polar plot of tool bending forces.
A basic weld monitoring system is adequate to provide a qualitative indication of weld quality issues. Spindle torque and tool temperature are recommended as primary weld condition indicators for the majority of industrial applications.
A greater understanding of the role of different tool components has been established.

Recommendations

The benefits of using the Artemis monitoring system in process Quality Control (QC), and for process development should be fully assessed.
More work is required to assess the repeatability of polar plot outlines and thus determine if slight variations are significant.
A full assessment is required to determine appropriate weld condition indicators and establish the effectiveness of the Artemis system for welds made under force control.