In the manufacture of heat exchanger components for aircraft, it was found that using the pulsed TIG method to weld the inlet and return manifolds was giving excessive porosity. Consequently an alternative method was sought.
To be acceptable, any alternate method had to control porosity within the specification tolerances. It also had to maintain the weldment profile and shape produced with the original TIG welding in order to comply with airworthiness standards.
Working in aluminium alloy 6061-T6 extruded tube, with a nominal diameter of 150mm and a wall thickness 8.5mm, both beads had to have smooth surfaces and there had to be a smooth transition between the weld bead and parent metal, giving no fatigue failure points. Furthermore, minimum heat input was stipulated to avoid damage to internal components
When in operation, the heat exchanger is exposed to mechanical and sonic vibrations from a nearby oil pump and has an internal working pressure of 125bar.
Electron Beam Welding
Electron beam welding was chosen as one of the alternative joining methods to be assessed, so consequently a CVE CW606 machine was selected for the weld development trials, which were run to determine the welding parameters necessary to obtain the required weld profile.
Normal Electron Beam Welding Results
Normal electron beam welding techniques were adopted for the first trials, which proved that it was impossible to produce a positive under-bead without a resultant undercut top bead. The need for both positive top and underbeads meant that a filler metal had to be introduced into the molten weld pool.
Electron Beam Welding Using a Filler Alloy Results
An aluminium alloy filler wire containing silicon was selected and, after more trials, displayed good shape to both top and underbeads, as well as a good contour between the weld beads and the parent metal. The shape of the weldment was very close to the original TIG welds, but in order to produce the required weldment shape it proved necessary to carry out four weld passes on each joint.
Effect of Multiple Passes
The first two passes produced a bottom bead with an acceptable smooth surface and a small under cut top bead. In order to remelt approximately 50% of the full penetration weld, to de-gas and remove porosity, a third pass was needed, which left the bottom bead unchanged, and the top bead undercut but with a smoother surface finish. In the fourth pass the object was to remelt approximately 25% of the weld joint and build up the positive top bead with a smooth finish.
Effect of Wire Feed Nozzles
Initial trials in which carried out with a fixed wire feed nozzle, in which molten filler metal entered the weld pool as ‘droplets’, which solidified to produce a characteristic fish-scale effect. The feed nozzle was modified to a spring loaded design, which allowed continuous contact of the wire filler metal with the weld pool. The melted wire entered the weld pool continuously, producing a smooth regular weld bead.
Analysis of Samples Welded Using Final Parameters
All of the test pieces welded using the final parameters were subjected to visual and radiographic inspection, with the acceptance criteria AMS 2680B. The porosity observed within the weldment had a pore size and aligned discontinuity limit well within the acceptable tolerances.
Overall, electron beam welding with wire fed filler metal was able to reproduce the weld shape of the original TIG process while maintaining porosity within specified tolerances. Thus, the electron beam welding process was found to be successful in its ability to produce the heat exchanger on a commercial basis, without the problems associated with pulsed TIG welding.