Flux does two jobs: it dissolves the oxide that would otherwise prevent wetting, and it shields the heated joint from the air long enough for the filler to flow. Without it, most brazing fillers will not wet steel, stainless or aluminium, however clean the surface appears.
It is a process aid, not part of the finished joint. That distinction is where most of the trouble starts, because a flux that has done its job is still present, still chemically active, and now sitting on a surface that will spend years in service.
After brazing, the residue is a glassy or powdery deposit around the joint, often extending further than expected because it spreads as it heats. Some of it sits on the outside where it can be seen. Much of it sits inside a tube, behind a fitting, or in the root of a lap joint.
The visible portion is the easy part. Residue that flowed into a gap or crept along a capillary path can remain trapped after any amount of external cleaning, and it is that fraction which causes problems much later.
Many brazing fluxes are aggressive by design, because breaking a stable oxide demands it. Left in place, they attract moisture and hold it against the metal, and the result is pitting or crevice attack along the joint line. On stainless the attack can be sharply localised.
There is a mechanical effect as well. A hard glassy residue on a sealing face prevents gaskets from seating, and residue inside a tube becomes a site for deposit build-up. Neither shows up at the bench, and both show up in service.
Inside a closed assembly, behind a flange, under a bracket, in the thread of a fitting, and in the narrow annulus of a lap joint. Anywhere the cleaning water cannot reach freely is a place the flux will stay, and the tighter the joint, the harder it is to flush.
On aluminium the problem is more acute because the same oxide that makes flux necessary also traps residue against the surface. Our TIG aluminium welding process avoids the issue by using an inert shield instead of a chemical one, which is one reason it is preferred for closed parts.
Clean while the part is still warm if the filler allows, then again once cool. Warm water, a stiff brush and repeated flushing work better than a single wash, and the flush should be directed through the joint rather than over it. Dry thoroughly afterwards, since trapped water is its own corrosion problem.
Do not assume a residue is harmless because it is thin. A white bloom around a joint after a few weeks in service is usually flux that was never removed, not a new fault. If the part is going into a painted or coated finish, remove the residue before any surface preparation, or the coating will fail over it.
Where a joint can be made by fusion in an inert atmosphere, no flux is needed and none of this applies. That is the usual reason a fabricator chooses TIG stainless steel welding for closed assemblies and for parts that cannot be flushed afterwards.
If brazing is the right process, choose the mildest flux that will do the job and design the part so the joint can be cleaned. A joint that cannot be reached for cleaning should not be brazed with an aggressive flux at all, however convenient it looks on the bench.
