Brazing joins two pieces by melting a filler that flows into the gap between them while the parent metal stays solid. The filler is chosen to melt below the point where the parent would be affected, so the joint forms by wetting and capillary action rather than by fusion. Nothing about the base metal is deliberately melted and re-formed, which is the whole point of the process.
That distinction sounds academic until a part is thin, already finished, or carrying a fitting that cannot survive a molten pool. In those cases a brazed joint is not a shortcut around welding. It is a different mechanism, chosen because the alternatives damage something that has to be kept.
The filler melts, the parent does not. Heat is applied until the filler reaches its working range, and capillary action then pulls the liquid along the joint gap. The parent reaches a temperature high enough to accept wetting but stays well below its own melting point, so grain structure and wall thickness are largely untouched.
This is also why brazing tolerates dissimilar metals. Two materials with different melting points and different expansion behaviour can be joined by a third that wets both, without either one having to form a shared melt pool. That freedom is unavailable to fusion welding, where one side usually has to give.
A thin wall is where fusion welding gets awkward. Arc energy that is comfortable on a thick flange will blow through a light tube, and the repair for a blown hole is often worse than the original problem. Brazing spreads its heat over a wider area and lets the filler find the gap instead of chasing a puddle.
Fittings, bosses, small brackets and sealed assemblies benefit most. The same logic covers work on parts that are already plated, polished or painted nearby, since a localised heat source can be kept away from finished surfaces. Our TIG stainless steel welding work sits at the other end of that decision, and the two are frequently weighed against each other on the same bench.
A brazed joint is generally weaker than the parent metal in shear and much weaker in tension across the joint face, so it should not be the primary structural connection on a part that flexes. It also softens as temperature climbs, which matters on anything near a hot section.
Vibration is the other problem. A joint that relies on a thin film of filler in a narrow gap can fatigue if the assembly moves, and an exhaust moves constantly. Where the part carries load, transmits engine motion, or runs hot, fusion remains the honest choice.
A sound brazed joint shows a small, even fillet at the edge and a continuous line where the filler met the parent. Gaps in that line, a grainy dull surface, or filler that balled up instead of flowing all point to the same family of causes: contamination, insufficient heat, or a gap too wide for capillary action to bridge.
Cut one open when the part matters. A joint that looks acceptable from outside but shows voids on section was never joined properly, only held by surface tension at the mouth. Destructive inspection of a scrap sample is the cheapest test available and the one most often skipped.
Ask three questions in order. Does the joint carry structural load? Will it run hot? Can the parent survive a molten pool at that wall thickness? If the answers are no, no and no, brazing is a legitimate method and often the tidier one. If any answer flips, fusion wins.
The decision should also account for what happens after the part leaves the bench. A joint that cannot be inspected, cannot be reworked, and cannot be re-brazed without disassembly is a liability regardless of how neatly it was made. Our fabricated components are specified around that service life rather than around the joining method alone.
