Five Things That Decide a Titanium Weld Before the Arc Starts

By Felicia November 3rd, 2026 34 views
Five Things That Decide a Titanium Weld Before the Arc Starts



What the frame actually contains

A titanium elbow, already welded, banded with colour. Below it a second tube, held vertically in a fixture, its open end sealed by a green plug. A TIG torch with a pink ceramic cup sits just off the joint, not yet fired. A clamp-on instrument is attached at the joint, with a lead running away to the right. Behind, a shelf of finished parts, all of them coloured.

Count what is in the picture. One finished weld, and five separate pieces of preparation standing around a weld that has not happened yet. The photograph is mostly about the ten minutes before.


The colour is the output, not the input

The banding is worth one paragraph and no more, because it answers a different question than the one this article is about.

What that colour records is the thermal history of the metal after the arc passed — how hot it got, how long it stayed hot, and what was in the gas around it while it cooled. It is a report on the process, written after the fact. It tells you something went right or wrong; it does not tell you what to do about it, because by the time you can see it, the joint is finished.

Everything that could still be changed is in the setup. So the useful question about this photograph is not what the blue means. It is what had to be true for the arc to be switched on at all.


One: the edge has to be bright metal

Titanium does not have a mill scale that burns off under the arc. It has an oxide that is stable, hard, and melts at a temperature far above the metal underneath, which means it does not dissolve into the puddle and disappear. It sits there, and the puddle forms around it.

That is why every cut edge on a titanium joint has to be mechanically cleaned back to bright metal before it is welded. A sawn edge, a laser-cut edge with its recast layer and its tint, an edge that has been handled and left with a fingerprint on it — all of them are something the arc has to weld through rather than to.

This is the first of the five and it is the one most often done by habit rather than by rule. The test is not whether the edge looks clean. It is whether anything has been done to it since it was cut.


Two: the gap has to be closed, because there is nothing to fill it

A titanium joint in this kind of work is usually made without filler, or with a minimum of it. That removes the welder's traditional escape route.

On steel, a poor fit is recoverable: feed more rod, wash the puddle across the gap, and the joint fills. On titanium, a gap means the arc has to travel further to reach both faces, the puddle loses its support, and the root either sags through or pulls back. Adding filler to cover it puts more heat into a joint that is already struggling, and titanium does not tolerate extra heat gracefully.

So the real tolerance on a titanium joint is set at the saw and at the prep bench, not at the torch. The welder inherits the fit. If the fit is open, no amount of skill converts it into a sound joint — the best available outcome is a joint that looks closed.


Three: the plug is a volume problem, not a flow problem

The green plug in the tube end is doing something that is easy to misunderstand.

Purging is usually described as blowing argon into the part until the air is gone. That is the right idea and the wrong mental model, because the difficulty is not getting gas in, it is getting the volume down to something that can be displaced in a reasonable time and then held while the metal cools. A long tube open at both ends is a large volume with a slow exchange. A plug placed close behind the joint reduces that volume to a short column, which is why the plug is positioned where it is rather than at the far end.

The second half of the plug's job is the part nobody plans for: the metal stays dangerously reactive well after the arc is out. Shielding has to continue through the cooling period, not stop when the noise stops. A purge that is removed too early protects the weld while it is being made and abandons it while it is still hot enough to absorb what is around it.

And a plug that leaks is worse than no plug, because it produces a confident reading and an unshielded root. The failure is silent by construction.


Four: the fixture is both a heat path and a contamination path

The fixture holding that tube is not neutral. Every point where it touches the workpiece does two things at once.

It conducts heat out, which is usually what you want: a chill close to the joint shortens the time the metal spends in the reactive range and pulls the colour back toward the acceptable end of the scale. But the same chill also changes how the joint cools and contracts, which is a distortion question, and a fixture that grips hard enough to hold a shape is also a fixture that is imposing one.

And it is a contamination path. Titanium picks up iron and other elements from anything that touches it while it is hot. A fixture, a clamp, a hand tool that also gets used on carbon steel is a source, and the contamination it leaves does not announce itself at the time — it shows up as a joint that behaves differently from the one next to it, or as a corrosion site months later. The usual answer is fixtures and tooling dedicated to titanium, or lined at the contact points.


Five: the number is worth more than the memory

The instrument clamped at the joint is the most quietly important object in the frame.

Whatever it is logging — temperature at the joint, current in the circuit — the principle it represents is the same. The variables that decide a titanium weld are not visible in the finished part. Purge quality is invisible. Interpass temperature is invisible once the part has cooled. The current that was actually used, as opposed to the current someone intended to use, is invisible. Colour is the only one that leaves a mark, and it is the least precise.

So the only way to control any of them is to record them while they are happening, on that part, at that joint. A shop that measures its purge and its interpass has a process. A shop that judges the result by how the finished weld looks has a habit, and habits drift.


What "clean" has to mean here

One more thing the frame implies without showing it: the handling rule.

Titanium contamination is invisible until it is not. A part that was welded on a bench that also carries steel offcuts, wiped with a rag that has been everywhere, or handled with gloves that were last used on a different job, will weld and will look reasonable, and will not be what the drawing asked for.

That is why the cleanliness requirement in titanium work is a rule about the whole route — separate consumables, separate abrasives, gloves that only do this job — rather than a cleaning step scheduled before welding. A cleaning step implies the part was allowed to get dirty, which means the contamination happened somewhere and nobody owns it.


What a buyer can ask

Ask what happens to the cut edge between cutting and welding. Ask how the fit is held, and what the shop does when it is not held. Ask where the purge plug goes and how the purge is verified — by a reading or by the look of the finished weld. Ask whether the fixtures are dedicated to titanium. Ask what is recorded per joint and whether you can see a record from a job like yours.

None of those questions are about welding technique. That is the point: the technique is the last five percent, and the five things above are the rest.


The finished elbow in that photograph is the part you can see. The green plug, the clamp, the instrument and the dressed edges are the part you cannot, and they were all decided before the torch was lit. A titanium weld is not judged at the arc. It is set up there.


Key Features

🧽 The edge must be bright metal — titanium oxide does not burn off under the arc, so every cut edge is dressed back before it is welded.

📐 The gap cannot be filled — with no filler to bridge it, a poor fit has no recovery, so the tolerance is set at the saw, not at the torch.

🟢 The plug is a volume tool — placed close behind the joint to shrink the volume that has to be displaced, so the purge actually reaches the root.

🌬️ Shielding continues after the arc — titanium stays reactive while it cools, so the gas has to stay on well past the noise stopping.

⚙️ The fixture is a heat path — a chill near the joint shortens the time in the reactive range and changes how the joint contracts.

🚫 And a contamination path — fixtures and tools that also touch carbon steel can leave iron on titanium, which does not show at the time.

📊 Recording beats remembering — purge, interpass and actual current are invisible in the finished part, so they are logged per joint.

🔍 Verify purge by reading, not by colour — colour is the post-mortem; a measured gas reading at the exit is the control.

🧤 Cleanliness is a route rule — dedicated consumables, dedicated abrasives and gloves that only do this job, not a cleaning step at the end.

🧲 Earth return affects the arc — where the return is taken from changes how the arc behaves, so it is chosen rather than assumed.

🧾 The setup is the process — five decisions before the arc, seconds of arc, and a part that reports on itself afterwards.

🤝 Questions worth asking — edge preparation, how fit is held, where the plug goes, whether fixtures are dedicated, and what is recorded.

Flat Sheet to Muffler — The Cut Is Decided Before the Beam Fires
Previous
Flat Sheet to Muffler — The Cut Is Decided Before the Beam Fires
Read More
Fupower Exhaust S-Bend TIG Welding | Mandrel Bent Pipe for Smooth Exhaust Flow
Next
Fupower Exhaust S-Bend TIG Welding | Mandrel Bent Pipe for Smooth Exhaust Flow
Read More