Galvanized sheet is the bread and butter of duct work, and welding it is where otherwise competent shops quietly lose quality and time. The problem is chemistry, not skill. Zinc boils at about 907 degrees Celsius, and a steel weld pool runs over 1400. So the moment an arc or a resistance electrode touches galvanized sheet, the zinc at the joint vaporizes. That vapor has three destinations, and all three cause trouble: it blows through the molten steel and leaves porosity, it condenses on surrounding surfaces as fine white fume, and it attacks anything in its way, including your electrodes and your lungs. Once you understand those three destinations, every practical adjustment below follows logically.
Resistance spot welding is the workhorse for joining galvanized sheet, on duct plugs, hanger brackets, corner reinforcements, and small assemblies. A dedicated spot welder set up for coated sheet makes the schedule repeatable. The good news is that zinc actually helps the process in one narrow way: it lowers contact resistance, so galvanized sheets weld at slightly lower resistance heating than bare steel. The bad news is everything else. The zinc layer at the faying surface melts first, spreads, and then the nugget has to form through it, which calls for more force and more current than the same job in bare mild steel.
For 0.6 plus 0.6 millimeter galvanized, a working schedule that holds up in production is around 9 to 10 kiloamps for 10 to 14 cycles at 50 hertz, with electrode force of 2.0 to 2.5 kilonewtons. Compared to bare steel of the same gauge, that is roughly 15 percent more current and 20 percent more force. Nugget diameter should land near 4 millimeters, and the only acceptance test worth anything is the chisel and peel: break a sample weld, look at the button, and measure it. A button that tears out of one sheet at full diameter is a weld. A flatty that separates at the interface is a cosmetic mark on two sheets.
Zinc is brutal on copper electrodes. The vapor alloys with the copper face, forming brass, which softens the tip, mushrooms it, and then contact area grows and current density falls until the welds go cold. In production on galvanized, expect to dress or change tips every 200 to 400 welds depending on coating thickness and force. Shops that skip a dressing schedule chase a moving target: the schedule was set when tips were fresh, so early welds are good and the last hundred are junk. Automated tip dressers on production spot welders pay for themselves within months for this reason alone. Keep a log of weld counts per dressing cycle and you will see your own number within a week.
When the job needs a continuous weld, MIG on galvanized demands respect for the zinc boil. The practical adjustments are simple to say and annoying to keep: slow the travel speed a little, use a slight push angle, and run a short-arc or pulsed transfer at the low end of the parameter window so the pool is not hotter than it needs to be. For 0.8 to 1.2 millimeter galvanized sheet, something near 90 to 120 amps with 0.9 millimeter wire is a sane starting window. Where the design allows, grinding or shearing the zinc back 10 to 15 millimeters from the joint line removes the porosity source entirely, and you repair the coating afterward.
Porosity shows up as bubbles in the bead surface and, worse, voids inside it. On duct work that mostly matters for structural attachments rather than airtightness, because nobody seam-welds galvanized duct for sealing; the seams are mechanical locks. If a weld must be pressure-tight, the zinc should be removed at the joint first, no debate. There is no wire, gas, or trick that makes zinc vapor friendly to a molten pool.
Zinc fume is the reason metal fume fever exists, and every welder who has spent a day on galvanized without extraction knows the night that follows: chills, fever, feeling like flu, gone by morning. It is not a rite of passage, it is a dose response, and chronic exposure has longer consequences. Local exhaust at the torch or gun is the fix that actually works; general room ventilation dilutes what the welder personally breathes. For spot welding cells, a fume arm at the gun plus tip extraction on high-volume lines keeps exposure sane. I have audited shops where the plasma table had beautiful downdraft and the spot welder sat in still air, and that is backwards: the resistance welder pushes vapor off the joint with force and it goes straight up past the operator's face.
Every weld burns the zinc around it. A spot weld damages a disc of coating maybe 15 to 20 millimeters across; a MIG bead strips its full width plus heat-affected margins. Left bare, that steel rusts at the first humidity cycle, and on duct installed in coastal or humid markets it rusts fast. The repair is not complicated, but it has to happen: wire brush the fume and spatter, then apply a zinc-rich cold galvanizing compound or zinc primer over the bare area. Two thin coats beat one thick one. Crews that keep an aerosol zinc spray at every welding station actually do it; crews that keep one can in the office do not.
If the duct gets painted or powder coated afterward, the zinc coating becomes an asset and a liability at once. Fresh galvanized surfaces are smooth and slightly reactive, and powder coats on top of them fail by outgassing and by poor adhesion unless the surface is prepared. The practical sequence that works: degrease, light sweep blast or chemical pretreatment, and for powder a pre-bake around 180 to 200 degrees Celsius for 10 to 15 minutes to drive volatiles out of the coating pores before the actual cure. Skip the pre-bake on freshly galvanized sheet and pinholes will decorate your paint line like acne. Aged galvanized that has weathered a month or two sometimes powder coats easier than fresh sheet, which surprises people until they think about what the zinc layer is still doing chemically on day one.
| Method | Zinc tolerance | Typical use on duct | Watch for |
|---|---|---|---|
| Spot welding | welds through coating with higher force and current | plugs, brackets, corner reinforcement | tip dressing every 200 to 400 welds |
| MIG short-arc | possible, more spatter and porosity | structural attachments, thicker hardware | strip zinc at joint if gas-tight needed |
| Lock seam mechanical | unaffected by coating | duct body seams, Pittsburgh locks | none, this is why duct is locked not welded |
| Flange plus fastener | unaffected | TDF and TDC connections, formed on a TDF flange forming machine | cut edges are bare steel, seal or paint them |
| Brazing | poor, zinc contaminates braze | rare on galvanized duct | avoid unless coating removed |
The table explains the structure of the whole industry: duct bodies are mechanically locked because locking ignores the coating, and welding is reserved for brackets and hardware where the joint earns its complications. The same logic applies to reinforcement beads rolled into the panel by a duct beading machine, which stiffens sheet without touching the coating at all.
Not all galvanizing is equal. Hot-dip galvanized coil with a zinc layer around 275 grams per square meter total is the standard for HVAC sheet and gives the best corrosion life. Electro-galvanized is thinner, smoother, and welds a little cleaner, but it rusts sooner at scratches and cut edges. Galvannealed, with its iron-zinc alloy surface, spot welds more consistently than pure zinc and paints without sweep blasting, at some cost in bare corrosion resistance. If your product ships to the Gulf or Southeast Asia, coating mass matters more than weld convenience; if it ships into a paint-and-powder workflow, galvannealed deserves a look.
Zinc vapor forces its way through the forming nugget and leaves voids. Raise electrode force about 20 percent and current about 15 percent versus bare steel, dress tips on a fixed cycle, and verify with a chisel-and-peel button test rather than trusting the weld timer.
Every 200 to 400 welds in typical production on 0.6 to 1.0 millimeter sheet, sooner with heavy coating mass or high force. Log weld counts and watch for cold welds late in the cycle; that pattern means your dressing interval has slipped past the tip life.
For structural MIG welds, grinding back 10 to 15 millimeters improves quality and fume levels noticeably. For any weld that must be pressure-tight, yes, remove the coating at the joint. For spot welding, the coating stays; the schedule and force handle it. Always repair the coating afterward with a zinc-rich spray.
Outgassing. Volatiles escape from the zinc layer pores during cure. The fix is surface pretreatment plus a pre-bake at 180 to 200 degrees Celsius before powder application. Shops that skip the pre-bake on fresh galvanized see pinholes; shops that bake first do not.