On a welded part, colour is not decoration. It is an oxide film of a particular thickness, and the thickness records how hot the surface got and how much oxygen it was exposed to while it was hot. A fabricator reads colour the way a machinist reads a surface finish: as evidence of what the process actually did.
This is true on stainless and it is more sharply true on titanium, where the oxide does not just sit on the surface but begins to draw oxygen into the metal beneath it. That is the difference that makes colour a process control issue rather than a cosmetic one.
A freshly welded titanium part moves through a predictable sequence of colours as the surface oxidises more heavily. Pale straw and light gold indicate a thin film. Purple and blue indicate a thicker one. A dull grey or a powdery white surface indicates heavy oxidation, and at that point the metal has absorbed oxygen rather than merely grown a skin.
The useful reading is not which colour is prettiest but how far along that sequence a given area has gone. Straw on a bead and grey at the edge of the shielding tell you the gas coverage fell away before the metal had cooled. The colour is a map of where the protection stopped.
On stainless, heat tint is a chromium oxide, and forming it draws chromium out of the metal immediately below. A straw tint sits over a shallow depleted layer. A heavy blue or black scale sits over a deeper one. The colour is again a thickness reading, but the consequence is different: it is the corrosion resistance under the tint that has been reduced.
So the same visual language describes two different problems. On titanium the concern is contamination of the metal by oxygen and nitrogen. On stainless the concern is loss of chromium from the metal at the surface. Both are addressed by removing the layer, and both are prevented by better shielding and lower heat input.
An oxide film is not a harmless coating to be melted into the next pass. It melts at a different temperature from the metal, it does not wet the same way, and it introduces oxygen and other contaminants directly into the weld pool. The result is erratic arc behaviour, inclusions, and a bead that looks wrong for reasons the welder cannot see.
On titanium the problem is worse, because the oxide layer is stable and thick, and attempting to arc over it drives contamination deeper rather than clearing it. This is why multi-pass titanium work is a sequence of clean, weld, clean again, rather than a continuous run with the torch left in place.
Between passes the requirement is straightforward: the surface has to go back to bright metal before the arc returns to it. That means removing the tint from the previous bead and from the adjacent heat-affected strip, not just from the visible weld face.
The method matters. A dedicated stainless brush or a carbide burr used only on that material will do the job. A brush that has been used on carbon steel will embed iron and create a new problem in the act of solving the old one. On titanium, any abrasive that has touched another alloy is a contamination risk, and the discipline of keeping tooling separate is most of the quality control.
Once the part is finished, the reason to remove colour changes. Now it is about corrosion resistance and appearance rather than about weld quality. On stainless that means removing the depleted layer, which is pickling work and not a polishing job. On titanium it means restoring a surface that will not act as a site for further oxygen pickup in service.
This is where a lot of finishing goes wrong. Polishing a tinted weld makes it shiny and leaves the depleted metal in place under a thin, bright surface. The part looks treated and is not. A dull, uniform, correctly pickled surface is a better outcome than a mirror finish over a compromised layer.
Mechanical removal is fast and controllable and is the right first step for a heavy oxide that would exhaust a chemical bath. Chemical removal reaches into places a brush cannot, which matters inside a tube or in a tight root. Most production sequences use both, in that order, because each does something the other cannot.
The common mistake is to reach for chemical treatment first on a heavily scaled part and expect it to do the work of a grinder. The other common mistake is to finish mechanically and stop, leaving a smeared surface with embedded contamination. Removing the bulk and then cleaning the surface is the sequence that produces a part which behaves as the material was intended to.
A finished titanium part with a deliberate, even colour can be a legitimate finish, and the colour is then a design choice rather than a process record. The distinction to keep is between colour that was produced on purpose under controlled conditions and colour that happened because the shielding ran out.
For a buyer, the useful check is consistency. An even straw or an even blue across a whole assembly suggests control. A bead that is gold in the middle and grey at the ends suggests the gas coverage was marginal and the welder kept going. That pattern is worth asking about, because it usually repeats. You can see how controlled colour appears on our TIG titanium welding parts.
Colour is one of the few process variables a customer can actually see on a delivered part, which makes it worth agreeing in advance.
Read colour as evidence and it becomes a useful inspection tool that costs nothing. Read it as decoration and it becomes a way to ship a part that looks finished while the layer beneath it is the layer that matters. Our titanium exhaust system and stainless steel exhaust system builds both depend on that discipline, the material background sits under shop by material, and the same standard applies across our TIG stainless steel welding and titanium fabrication components output.
