A stainless bolt that will not turn is often described as cross-threaded, and usually it is not. Cross-threading happens as the nut goes on. Galling happens later, under load, and it produces a joint that behaves as though it had been welded. The two failure modes look similar from outside and are completely different inside.
Galling is adhesive wear. Metal from one thread surface transfers to the other, tears and cold-welds. Once that has happened the interface is no longer a thread. It is a rough, torn, work-hardened sleeve, and turning harder does not free it. It shears the fastener or strips the thread.
Austenitic stainless has a thin, brittle passive film over a soft and very ductile bulk. Under pressure the film ruptures easily and does not regenerate fast enough to keep the two surfaces apart. Clean metal then meets clean metal, and clean metal of the same alloy in the same condition has a strong tendency to bond.
Carbon steel behaves differently in practice because its surface scale and its lower ductility at the interface change what happens under load. The stainless surface also work-hardens as it deforms, so the material that has just been torn becomes harder and rougher, which makes the next pass worse rather than better.
Galling needs two conditions at once: contact pressure high enough to break the film, and relative movement. A static joint under steady load may survive indefinitely. The same joint tightened, loosened and tightened again is being asked to slide under pressure repeatedly, and each cycle removes more of the barrier.
This is why hardware that has been apart several times is more likely to seize than factory-fresh hardware. It is also why a nut run down a long length of thread at speed is more at risk than one turned slowly over a short engagement. Distance travelled under load matters as much as the load itself.
Thread class and clearance decide how much load lands on how few flanks. A tight fit concentrates pressure on a small contact area and a slightly looser fit spreads it. Coarse pitches put more material behind each flank and tolerate contamination better, while fine pitches offer more flank area but far less tolerance for a burr.
Rolled threads generally leave a smoother, harder surface with better grain flow than cut threads, and a smoother flank breaks the film less readily. Chasing a damaged thread with a die removes material and leaves a torn surface, which is one of the least helpful things you can do to hardware that is already marginal.
Speed is the most underrated variable. An impact driver delivers torque in pulses, and each pulse is a small sliding event under high pressure. On stainless hardware that is close to a description of how to cause galling. Steady hand turning at low speed is far kinder to the interface.
Alignment matters just as much. A nut being pulled sideways by a misaligned bracket loads a fraction of the thread and rubs the rest. Bolting a flange up without checking that the faces sit parallel is a reliable way to create a joint that will not come apart later, and the damage is done before anyone notices.
An anti-seize compound works by putting a soft, shearable film between the surfaces so that the film fails before the metal does. Solid lubricants of the molybdenum disulphide type are effective for this because they keep working under high pressure where a liquid film would simply be squeezed out of the contact.
The limits are practical. A lubricant applied once at assembly does not survive a hot exhaust flange for long, and hardware that has been through many heat cycles is often dry and lightly scaled inside the thread. Re-applying at every reassembly is the only version of this that works, and it is the step most often skipped.
Two stainless parts of the same alloy are the worst pairing for galling, because there is no difference in hardness or structure to interrupt the bond. Pairing stainless with a different alloy, or with a plated nut, changes the interface enough to reduce the tendency, at the cost of introducing a galvanic couple that has to be thought about on its own terms.
Water is the other half of the story. Hardware that sits wet, or in a crevice that never dries, corrodes inside the thread and then seizes for a different reason. A galled thread is welded. A corroded thread is packed with oxide that is harder than the parent metal. Both end with a fastener that has to be cut out.
The design side is about reducing how many times the joint has to move. Studs threaded into a fixed component and a nut turned on the free end concentrate the wear on a replaceable part. Through-bolts with a nut and washer at each side do the opposite, spreading the problem across hardware that is harder to replace.
On the assembly side, cleanliness before lubrication, a deburred thread end, a chamfer to start the nut square and a deliberate slow first turn are all cheap. None of them appear on a drawing. All of them decide whether the part comes apart in service. The kind of stainless hardware and welded assemblies we produce is illustrated under TIG stainless steel welding.
Fasteners are usually the last line of a specification and the first thing to cause a problem in service. These are the points worth settling early.
Galling is one of the few failure modes where the fix is mostly behaviour rather than material. Slower assembly, a shearable film on the thread and a joint designed to be taken apart will outperform a harder fastener almost every time. Our stainless steel exhaust system work and our exhaust pipe range both depend on hardware still behaving years later, the material options sit under shop by material, and the same reasoning appears in our titanium fabrication components and TIG titanium welding parts.
