Materials get selected from a number in a table: a maximum service temperature, quoted as though it were a property of the alloy. It is not. It is a summary of how that alloy behaves in a particular environment, at a particular stress, with a particular wall thickness and cycling pattern. Change any of those and the number moves, sometimes a great deal.
What actually decides whether a part survives is which degradation mechanism is active where it sits. Scaling, internal oxidation, thermal fatigue, chloride attack, creep, and contamination cracking all have different dependencies, and the practical question is which one will get there first. Choosing an alloy without naming the mechanism is choosing by reputation.
Stainless steel protects itself with a thin chromium-rich oxide that forms instantly and reforms when damaged. That layer needs chromium to be available at the surface. At elevated temperature two things work against it: the oxide grows and can spall when it is thermally cycled, and in the presence of water vapour some oxide species become volatile, so the protective layer is removed rather than thickened.
Once the layer is being consumed faster than it can reform, the alloy is no longer stainless in any useful sense. Oxidation moves into the grain boundaries, the section thins from the inside, and the surface acquires a rough, layered scale. The part does not usually fail by melting. It fails by becoming thin, brittle and cracked.
The important detail is that the aggressive species can come from the fuel and the air rather than from the alloy. Water vapour is a combustion product and is always present in exhaust. That is why the same grade behaves differently on a engine that runs rich, or with an additive package that puts other species into the gas.
A part at steady temperature is a much easier problem than one that heats and cools. Cycling produces strain wherever expansion is restrained, and an exhaust is restrained everywhere: by its flanges, its hangers, its welds and by the temperature difference across its own wall. The strain concentrates at geometric features, and the sharpest geometric feature on most fabricated parts is the weld toe.
Thermal fatigue therefore shows up as cracking at the edge of a weld or at a bracket, not in the middle of a smooth run. It is also cumulative, so a part that survives one hard use can fail after many, and the failure looks sudden when it is nothing of the kind.
At the rear of a system the gas has cooled, and cooling has consequences that are not thermal at all. Water condenses, acidic species dissolve in it, and the result is a wet, corrosive environment sitting in the lowest part of the system. A grade chosen for its high-temperature behaviour may be the wrong choice at the back, where chloride and acid resistance matter more than scale resistance.
That is why a system can lose its rear sections first while the manifold is still sound. The two ends are not the same problem, and designing a whole system around one material assumes they are.
Titanium is chosen for a combination of properties that happens to suit exhaust work well. It has a stable, adherent oxide that does not spall the way steel scale does, so it resists the thinning mechanism described above. It expands less with heat, which reduces the strain that cycling imposes on a restrained part. And it is strong for its mass, so a thin wall can carry the load a thicker steel wall would.
Those are real advantages and they are the reason titanium exhausts exist. They are also specific. Titanium is not simply a higher-temperature version of stainless, and it does not solve a corrosion problem at the cold end, where its behaviour is good but not unlimited.
What it does bring is a set of fabrication constraints. It is reactive at temperature and has to be shielded until it has cooled, it galls when it rubs, and it is sensitive to notches. A titanium part designed like a stainless part, with the same wall and the same sharp transitions, gives away much of what the material was bought for.
The characteristic failure of a titanium weld is not a shortage of strength. It is embrittlement from oxygen, nitrogen or iron picked up while the metal was hot. A contaminated weld does not look weak; it looks fine and then cracks, because the contamination has formed a hard, brittle phase at exactly the point where stress concentrates.
That makes cleanliness and shielding part of the material specification rather than a workshop preference. It also means that a titanium part's service life is decided as much by how it was joined as by what it was made from, which is a different situation from steel, where a poor weld is often just a weaker weld.
The hottest point in an exhaust is not where the most damaging conditions are. A manifold runs hottest but is thick, well supported and usually dry. A downpipe runs slightly cooler but is thin, restrained between a turbo and a flexible section, and cycled hard. A rear section runs coolest and is wet. Each position has a different dominant mechanism, and a material that suits one may be wrong for another.
Thinking in positions rather than in a single peak temperature also explains why mixed systems exist. Using titanium where its expansion and mass matter, and a suitable stainless where wet corrosion dominates, is engineering rather than compromise.
Every statement above is about mechanisms, not limits, because the limit depends on the duty. The same alloy in the same position can last very differently between a car that does short trips and one that tows at sustained load, and between a part with a generous wall and one made as thin as the material allows.
It also stops being reliable once tuning is involved, because raising output usually raises exhaust temperature and often changes the mixture, which changes the gas composition. A material choice validated on a standard engine is not automatically valid on a modified one, and the failure will arrive as a cracked weld rather than as a warning.
The useful questions are about duty and about position.
The titanium work is on our titanium exhaust system pages, the steel equivalent under stainless steel exhaust system. Managing the heat that causes all of this is covered under exhaust heat blanket and exhaust heat shield, and the air the engine consumes, which sets much of the gas temperature, under air intakes and intercooler.
