A TIG torch with a gas lens cup, a ceramic nozzle close to the joint, an arc at the seam where a titanium elbow meets the section below it, and a welding helmet with the window lit from behind.
Along the pipe, a band of colour: dark blue, then a lighter blue, then a gold section, then a paler straw tone above it, with a sharp gold line where the previous joint sits.
Most people photograph titanium because of that colour. The colour is also the most useful quality record in the entire weld — and the easiest thing for a buyer to misread.
Titanium is reactive. At room temperature it is protected by a thin, stable oxide that forms instantly and stays put. Heat it and that protection becomes a problem: the metal starts absorbing oxygen and nitrogen from the air, and the oxide layer thickens and changes composition.
Different oxide thicknesses refract light differently, which is why the visible result is a progression of colours rather than one colour. That is heat tint. It is a chemical record of temperature and atmosphere, written on the surface during the weld and impossible to apply afterwards in the same way.
The practical reading, from coolest to hottest, runs roughly: pale straw, then gold, then bronze to brown, then purple, then blue, then light blue, and finally a dull grey or chalky white.
The first half of that list is normal and generally acceptable on a weld that has to be visible. The last part is not a colour at all — it is contamination.
Once the surface goes past blue into grey and then into a powdery white or pale yellow-white layer, the oxygen has not merely tinted the surface, it has diffused into the metal beneath it. That layer is hard and brittle, and it is sitting exactly where the weld is. Grinding it off looks like a fix and leaves a thinner, altered section behind.
So the question is never "does the weld look blue". The question is where the progression ends. A joint that goes straw to gold and stops is a joint with controlled heat. A joint that runs through blue into grey is a joint where the shielding failed, and the colour is telling you so in public.
Here is the uncomfortable part. The most photogenic titanium weld is not the best one.
A deep, saturated blue is a high-temperature tint. It means the trailing shield and the gas coverage let the metal stay hot in air a little longer than ideal. It is not automatically a reject — plenty of functional exhaust parts show blue — but it is not the mark of a superior weld either. The superior weld is usually the boring one: narrow straw and gold, tight band, consistent along the seam.
Which means a supplier who sells on "look at our rainbow welds" is selling you a photograph of heat input. Sometimes that heat was necessary because the wall was thick or the part was large. Sometimes it was just a hand that moved slowly.
Welding a straight run is a rhythm. Welding an elbow is a series of compromises, and you can see them in the tint.
On the outside of the bend, the torch has to reach around a convex surface. On the inside, the joint closes on itself and the torch body, the cup and the operator's knuckles all compete for the same few centimetres. Torch angle changes through the pass, and with it the angle at which the shielding gas hits the puddle.
Titanium also conducts heat poorly compared with aluminium, so heat does not spread away from the arc the way it does in other metals. It stays near the weld. On a curved joint where travel speed necessarily varies, that concentration shows up directly as a wider colour band in the places where the torch slowed down — usually the tight inside of the bend.
A uniform tint on an elbow is therefore a stronger claim than a uniform tint on a straight pipe. It says the operator kept gas coverage and travel consistent around a shape that actively resists both.
Heat tint visible on the outside is only half the record. The inside of that joint — the root, the back of the weld — was also hot, and was also in contact with whatever atmosphere was inside the pipe.
If that atmosphere was air, the root oxidised. On titanium that is worse than a cosmetic issue: an oxidised root is brittle, it can crack under vibration, and it cannot be inspected or repaired from outside. A part can look immaculate on the face and be compromised on the back.
This is why titanium welding is a purging problem as much as a torch problem. The inside of the joint has to be filled with an inert atmosphere before the arc starts and kept that way until the metal is cool enough to stop reacting. On an elbow, that is harder than on a straight tube, because the volume to purge is awkward and the shape traps air.
The outside colour tells you the torch shielding worked. It tells you nothing about the back. Only the process tells you that.
Titanium can also be coloured deliberately, by anodising — an electrochemical process where the oxide thickness is set by voltage. That produces flat, even, chosen colours, often across a whole component, and it is a finish.
Heat tint is the opposite: local, generated by welding, variable by nature, and diagnostic rather than decorative.
The two get confused constantly, including in supplier photos. If a whole exhaust system is uniformly blue with no variation near the welds, that is likely anodised or coated. If the colour concentrates in haloes around joints and fades along the tube, that is heat tint, and it is telling you about the process rather than the styling.
Follow the band. It should be narrow, it should sit close to the weld, and it should not run away down the pipe.
Look for where it ends. Straw and gold fading out is normal. Grey, chalky white or a dull matte patch at the edge of the band is contamination, not styling.
Compare the outside of the bend to the inside. A large difference in tint between the two suggests the torch slowed down where access was tight, which is where the heat went.
Ask how the back was shielded. A titanium fabricator who can describe their purging method is describing the half of the weld you cannot inspect.
Be suspicious of perfect uniformity across a whole part. That is usually a finish, not a weld.
Titanium is bought for weight and for heat tolerance, and it is welded by people who have to control an atmosphere, not just an arc. The colour on the finished part is the visible residue of that control.
Aim for the boring weld. Narrow straw and gold, consistent around the bend, with a root that was protected while it was hot. It will not photograph as well as the rainbow, and it is the one that lasts.
🎨 Colour is oxide, not paint — heat tint is a chemical record written during the weld, not a finish applied after it.
📖 The band reads as a temperature scale — straw, gold, bronze, purple, blue, and then grey where it stops being acceptable.
⚠️ Grey and chalky white are contamination — oxygen has diffused into the metal, not merely tinted the surface.
📸 The prettiest weld is not the best one — saturated blue means heat lingered in air, not superior skill.
🔄 Elbows force the compromise — torch angle, gas coverage and access all change through a curved joint.
🌡️ Titanium holds heat where it is put — poor conductivity concentrates it, which widens the tint where travel slows.
🫥 The root is the half you cannot see — an oxidised back side is brittle and cannot be repaired from outside.
💨 Purging is half the job — the inside must be inert before the arc starts and stay inert until it cools.
🔍 Anodising is a different thing — even colour across a whole part is a finish; haloes around joints are heat tint.
✅ Consistency around the bend is the real claim — a uniform tint on an elbow says more than one on a straight pipe.
🚫 Grinding off grey is not a repair — it removes the evidence and leaves an altered section behind.
🤝 Ask about the purge — it is the only question that covers the side of the weld you cannot inspect.
