Titanium is not difficult to melt. It is difficult to keep clean. Once the metal is hot it takes oxygen and nitrogen out of any atmosphere it can reach, and that pickup does not sit on the surface as scale the way it does on steel. It dissolves into the metal and takes ductility with it.
On the face of a weld the shielding cup looks after that. On the inside of a tube, nothing is shielding anything unless gas is deliberately put there. The back side of a titanium joint is the same alloy, reaches a similar temperature, and is exposed to whatever air happens to be trapped inside the pipe.
An argon back-purge does one job: it displaces the air inside the tube before the arc starts, and keeps displacing it until the metal is cool enough to stop reacting. Argon is heavier than air, so a slow, steady flow will fill a tube and hold a layer in it when both ends are controlled.
Flow rate matters more than the total volume of gas used. Too little and the pocket never forms, so the joint looks clean on the outside and dull grey inside. Too much and the gas turns turbulent, dragging room air in behind it, while also cooling the puddle faster than the operator expects.
Most tube work uses a purge dam at each end of the section, tape over any branch or flange opening, a feed line at one end and a vent at the other. Gas has to move through the tube rather than merely sit in it, and leaving the line open for a while before the first arc lets the pocket settle.
On thin-wall tube the purge also has to keep running after the weld stops. This is where a great deal of contamination happens: the joint is finished, the gas is shut off immediately, and the still-warm root oxidises in the seconds that follow. A delay before shutting down is not optional on titanium.
The inside of a titanium weld is the honest report on how the purge went. A bright silver or light straw root means the atmosphere held. A dull matte grey, or a white powdery deposit, means oxygen was present, and the heavier the deposit the deeper the affected zone runs.
From the outside some of this can be inferred from the way the bead sits and from the colour of the face, but that is a guess. Wherever the interior can be seen, through an open end or a flange, it should be looked at before the part moves on. Hidden contamination does not improve later.
Not every joint allows a clean purge. A small closed fitting, a collector with several branches, or a repair on a part still mounted on a vehicle all make it hard to fill the volume. In those cases the approach changes rather than the standard: seal what can be sealed, purge what can be purged, and reduce heat input so the back side has less to react with.
For parts that are welded and then finished, such as the hand-built pieces in our TIG titanium welding parts range, the purge is built into the fixture instead of being improvised. A fixture that clamps and seals at the same time gives a more repeatable purge than tape applied fresh to every part.
Purge discipline is mostly habit. Hose used for argon only and never shared with compressed air, dams and tape kept clean, and a set sequence for each joint type so the delay after the arc is not left to memory. Small things, but they decide whether a root comes out silver or grey.
It also helps to treat the purge as part of the joint design rather than a bench trick. Where a design leaves a tube open at both ends, or holds back a flange until the root is complete, the fabricator gets a better result than any amount of improvisation. The same thinking runs through the titanium fabrication components we build.
