Welding stainless steel: filler, gas, distortion
Stainless is not difficult so much as different. It conducts heat badly, expands more than carbon steel, and its corrosion resistance depends on a surface layer that welding attacks.
Three physical facts that drive everything
Low thermal conductivity — about a third that of carbon steel. Heat stays where you put it, so the same settings that suit mild steel will overheat stainless. Expect to weld at roughly 10–20 % lower current.
High thermal expansion — around 1.5 times carbon steel. Combined with the poor conductivity, that means distortion, and it means the sequence and the clamping matter more than on mild steel.
The chromium oxide layer. What makes the steel stainless is a self-healing chromium oxide film. Heat, contamination with carbon steel, and oxidation during welding all damage it, and where it is damaged the steel rusts like anything else.
Filler and gas
Filler is chosen to leave a small amount of delta ferrite in the deposit, which prevents solidification cracking: 308L for 304, 316L for 316, 309L for stainless-to-carbon joints. The L means low carbon, which matters for the next section.
Gas: argon with 2 % CO₂ or 2 % oxygen for MAG, kept deliberately lean because carbon picked up from CO₂ reduces corrosion resistance. Pure argon for TIG, or argon with a few per cent hydrogen on austenitic grades to raise the heat and brighten the surface. The gas calculator gives the designation to order by.
Sensitisation — the invisible damage
Held between roughly 450 and 850 °C, chromium in austenitic stainless combines with carbon and precipitates as chromium carbide at the grain boundaries. The metal immediately around those boundaries is left short of chromium, and it corrodes preferentially — intergranular corrosion, which shows up months later as cracking alongside the weld.
Three defences, and they are used together. Use L grades with carbon below 0.03 %, so there is little carbon to react. Use stabilised grades — 321 with titanium, 347 with niobium — which tie the carbon up. And keep the time in that temperature range short: low heat input, interpass temperature below about 150 °C, no unnecessary weaving.
Cleanliness and cross-contamination
Stainless must not share tools with carbon steel. A wire brush that has touched mild steel, a grinding disc used on structural work, or a bench where carbon steel was cut all embed iron particles in the surface, and those particles rust and stain the stainless around them.
Dedicated stainless brushes and discs, a separate bench or a clean cover, and gloves. It sounds fussy until the first job comes back with rust spots on material that "cannot rust".
Managing distortion
Tack more often than on carbon steel — every 50 mm rather than every 100. Weld in short runs, alternating ends. Use clamps and backing bars, which also draw heat away. And accept that on thin stainless sheet some straightening afterwards is normal rather than a sign of poor technique — the same measures that work on thin sheet generally apply, with less heat still.
Frequently asked questions
Why 308L filler on 304 rather than a matching one?
Because 308L deposits a few per cent of delta ferrite, and that ferrite is what stops hot cracking down the centreline of the weld. A perfectly matching fully austenitic deposit would be more prone to solidification cracking, not less. The same logic gives 316L for 316 and 309L for joining stainless to carbon steel.
What are the rainbow colours next to the weld?
Oxide, and it means the chromium oxide layer that makes the steel stainless has been damaged. Light straw is usually acceptable; blue and grey are not, because corrosion resistance is reduced there. Remove them by pickling or mechanical cleaning, then passivate.
Does stainless need a back purge?
On pipe and on anything where the root side matters, yes. An unpurged root oxidises into a grey, sugary surface that has lost its corrosion resistance and often its integrity. On open fillets in general fabrication, no.
Updated: