A sheet-metal cover that lost its weld nuts in the field. The nuts were M6 projection weld nuts (DIN 929), welded onto 1.2 mm sheet. The customer torqued the cover bolts to 8 Nm. Within a few months, two of the eight nuts spun free. The customer blamed the welder. The welds looked fine — clean, round nuggets. But the nuts were welded with too much current, which burns through the sheet and leaves the nut sitting on a thin, brittle rim.
Projection weld nuts are designed to be welded with a specific current-time schedule. The three projections on the nut melt and collapse, forming a solid joint. The current must be high enough to melt the projections but not so high that it expels metal or burns a hole through the sheet. The sweet spot is narrow. At too-low current, the nut barely sticks. At too-high current, the nugget is a hollow crater. Both look passable from the outside. Only the pull test tells the difference.
What the failure actually was
The customer’s welder was a 35 kVA spot welder set by “feel.” The operator cranked the current up because the previous batch of nuts had cold welds. The result was over-welding: the nut’s projections fully melted and the sheet underneath was almost burned through. The joint’s strength came from a thin rim of melted metal, not the full projection collapse. Under vibration, the rim cracked. The nut spun.
We pulled test nuts with a torque wrench before production. At 35 kVA with the “feel” setting, the nuts failed the pull test at 12 Nm — the minimum spec was 25 Nm. After adjusting to a proper schedule (measured current and weld time), the same nuts pulled at 40 Nm without stripping. The welding schedule was the fix, not the welder’s skill.
The pull test that should be standard
Every weld nut line should pull-test nuts before production and every shift change. The test: weld a nut to a test piece, clamp the sheet, and torque the nut with a calibrated wrench until it strips or spins. Record the value. If it drops below spec, stop the line and check the schedule. The test takes 60 seconds and catches 90% of weld nut problems before they ship.
Also check the sheet thickness. Weld nuts are rated for a sheet thickness range. An M6 projection nut on 1.2 mm sheet is near the thin end — the projections have less material to weld into. On 0.8 mm sheet, use a smaller nut or a PEM-style clinch nut instead. Weld nuts on thin sheet are a design problem, not just a process problem.
When to use a clinch nut instead
If the sheet is under 1.0 mm, or the nut must be removed and re-torqued repeatedly, or the weld process isn’t well controlled, switch to a clinch nut (self-clinching). It’s pressed in, no heat, no current, no projections. The pull-out strength is higher on thin sheet than a weld nut because the clinch ring deforms the sheet around it. The downside: the sheet needs a pre-punched hole and the press applies a controlled force. For a production line with a punch press, that’s easy. For a job shop welding random parts, weld nuts are still simpler.
The inspection you can’t skip
After welding, check the nut’s thread for spatter. Projection welding near the threads can throw spatter into the thread, and the bolt won’t start. A quick thread-go gauge check on every nut is cheap insurance. The customer’s cover failure was caught by the pull test before it shipped; the spatter check catches the other failure mode.
Weld nuts fail from the wrong current-time schedule, not from a bad welder. Pull-test before production and every shift. If the nut spins under torque, check the schedule before blaming the operator. On thin sheet, clinch nuts beat weld nuts. The clean-looking weld can still be a hollow crater.
