Colour on titanium is not paint, coating or a reaction with the filler. It is a very thin oxide film that grows on the surface while the metal is hot, and its thickness decides which wavelengths cancel out and which are reflected back. A thin film reads silver or pale straw; a thicker one moves through purple and blue towards grey.
That is why titanium exhaust work colours so readily and stainless does not: the film grows quickly on titanium at temperatures that a stainless part would shrug off. The colour is a record of how hot the surface became and how much oxygen reached it.
Read from cool to hot, the sequence runs from bright silver through pale straw and gold, into purple, then blue, and finally to dull grey. Around a sound weld the ideal is a narrow band of straw or light gold beside the bead, fading quickly back to silver and clean metal.
Grey is the exception that breaks the pattern. It is not simply a hotter version of blue; it means the oxide has thickened to the point where it is powdery and the metal underneath has been contaminated rather than merely tinted.
Colour describes the surface, not the joint. A bead can be beautifully coloured and still be undercut, short of penetration, or sitting over a gap that never closed. The tint band says something about heat input and shielding; it says nothing at all about fusion.
It is also a poor guide to the back side of a tube you cannot see. The outside can be gold while the root is grey, because the two sides saw very different atmospheres and very different gas coverage.
The root is the more revealing of the two. Because the inside of a tube is harder to shield, contamination shows there first, and a root that has gone past straw into blue or grey means the purge was weak or was stopped too early. That is why the interior is checked first on every TIG titanium welding part that can be inspected.
On visible parts the same film forms on the outside, and some finishes are produced deliberately by controlling temperature rather than by adding anything to the metal. That is heat tint used as a finish, and it is common on titanium exhaust tips where the surface is the product.
Heavy oxide is not only a cosmetic problem. It is brittle, it can crack and flake, and on a thin wall the affected layer is a real fraction of the section. Where a part has to look uniform, the film is removed mechanically and the surface restored, rather than polished over.
Removing the colour without removing the contaminated layer beneath it achieves nothing useful. The colour is only the indicator; the metal under it is the actual defect, and it is the metal that has to be dealt with.
Treat colour as a fast, cheap check rather than a verdict. If a root that should come out silver comes out blue, the reaction is to look at the purge setup, then the flow, then the delay after the arc, before touching the technique or the machine.
Over time colour also becomes a useful signal during the weld itself. The operator sees the tint band widen and knows heat is accumulating, which is a cue to ease off, move on, or let the joint cool before continuing around the part.
