Stud welding for sheet metal: specification, placement and inspection

A cover panel arrives with eight M6 studs, all visually sound, all with a full flash ring. Two of them spin under a nut at 6 N·m on the customer's line. The visual check passed because the weld was complete at the surface; what failed was fusion into the sheet, and the stud had been welded onto a spot where the base did not sit flat.

Stud welding for sheet metal is decided by three things before a single stud is welded: the sheet thickness, the distance from the weld to an edge or a bend, and the length the stud must have after the plunge. Get those wrong and no amount of gun adjustment saves the part, because the failure is in the geometry rather than the setting.

This article covers both commercial stud processes, the placement rules that come out of the sheet rather than the stud, the two tests that actually detect a bad weld, and the drawing fields that stop the argument at goods-in. The bending and welded assemblies hub holds the rest of this cluster.

Stud welding for sheet metal: which process suits which thickness

Drawn-arc stud welding is the process from about 2 mm thickness upward, and capacitor discharge covers 0.5–1.5 mm. The dividing line is the size of the molten pool, not the stud.

In drawn-arc welding the gun lifts the stud 1.5–3 mm off the plate, strikes an arc, and holds it for 0.15–0.7 s while a pool forms in the plate and a film melts on the stud base. The stud is then plunged into the pool, displacing metal into the flash ring that forms the visible collar, and held for a fraction of a second while it solidifies. A ceramic ferrule around the base contains the pool and shapes the collar.

That pool is about 1 mm deep on a 6 mm stud at 550 A. On 2 mm sheet the remaining ligament is thin but intact. On 1 mm sheet there is no ligament left and the arc burns through, which is why the process has a practical floor around 1.5 mm.

Capacitor discharge welding stores energy in a capacitor bank and dumps it in a few milliseconds through a small tip on the stud base. The pool is shallow, the heat-affected zone is narrow, and the sheet behind the weld stays cool enough to hold its shape. The trade is stud size: CD is practical up to about 8 mm, and above that the energy required starts to defeat the point of the process.

Neither process needs a hole, a nut insert or access to the back face, which is why studs appear on panels that are closed on three sides.

Stud size against sheet thickness

The table below comes from two calculations. Stud capacity is the proof load of a grade 4.8 stud: 310 MPa on the stress area, 36.6 mm² for M8. Sheet capacity is tear-out of a plug around the weld, taken as the base diameter plus twice the thickness, sheared at 270 MPa, which is 0.6 × 450 MPa for CRCA with a tensile strength of 450 MPa.

Read the table as a warning about small sheet, not about large studs. The number in the fourth column is the thickness at which the joint is equally likely to fail in the stud or in the sheet. Above it, the stud is the weak link and a stronger stud buys strength. Below it, the sheet is the weak link, and moving from grade 4.8 to grade 8.8 changes nothing because the panel tears out first.

That has a design consequence. On a 1 mm cover that needs an M8 fixing, the answer is not a bigger stud or a longer one; it is a doubler plate, a formed rib at the stud position, or a change to a rivet nut or a weld nut fitted from the open side.

Currents and times in the table are indicative ranges for CRCA. The gun setting for a specific stud and sheet combination comes from the stud manufacturer's data sheet and from the procedure qualification, and it changes with the coating on the sheet, the earth return path and the condition of the ferrule.

Placement rules that come out of the sheet

Edge distance and spacing

A stud too close to a sheet edge tears the edge out rather than pulling a plug, and the panel fails at a lower load than the calculation above predicts. Keep the stud centre at least 1.5 stud diameters from the sheet edge; for M8 that is 12 mm, and for M12 it is 18 mm.

Spacing matters for a different reason. Two studs welded closer than about 3 diameters share a heat-affected zone, and on thin sheet the second weld draws current through the first one's hot metal. Arc blow pushes the second arc to one side, the collar is asymmetric, and the panel bows between the two studs. Space them 3 diameters apart and, on sheet below 1.5 mm, let the panel cool between welds rather than running a row back to back.

Flat seating

The stud base has to sit flat on the sheet within about 0.2 mm across its diameter. On a flat laser-cut panel that is automatic. On a rolled surface, the inside of a formed corner or a panel with a raised bead under the stud position, it is not, and the arc strikes on the high side. The result is a weld around two thirds of the circumference and a stud that leans 3–5° as it cools.

If a stud has to sit on a curved surface, specify a flanged base or a stud designed for it, or add a flat pad to the part. Adding a 0.5 mm coining step in the press tool is cheaper than a weld repair on every unit.

Distance from bends and holes

Weld after bending, not before. A stud welded 8 mm from a bend line sits inside the deformation zone, so the bend pulls the stud off vertical and the flash ring cracks on the tension side. Keep the stud centre at least the inside radius plus twice the thickness from the bend line, and remember that the tolerance on the bend position moves the stud with it.

The same logic applies to holes and laser-cut edges. A stud 5 mm from a 10 mm hole shares the hole's stress concentration, and the panel tears between the two under load. Keep 2 diameters from a hole edge, which for M8 means 16 mm from the edge of the hole, not from its centre.

Position tolerance

A stud position is a feature of the assembly, not of the panel. Set ±1 mm as the default, and tighten to ±0.5 mm only where the mating part needs it, because ±0.5 mm means a fixture holding the gun rather than a hand-held weld. Ganesh holds ±0.5 mm as the typical tolerance for sheet metal features, tighter on critical features by agreement, and a stud pattern that lands inside that needs the fixture designed into the job from the start.

Inspection: two tests, not one

Visual inspection catches half the failures and misses the other half. A stud with a complete, even flash ring, no undercut and no visible porosity is probably sound. A stud with a cold weld — the arc ran but fusion into the sheet was incomplete — often looks exactly the same.

The bend test is the fastest destructive check. Bend the stud 30° with a pipe over it, in one direction, and look at the weld. A sound weld deforms plastically; the stud bends and the collar stays attached. A cold weld breaks at the interface with little or no deformation, and the base shows a smooth area where fusion did not occur. Test the first stud of every shift and one per fifty thereafter, and record the result on the control plan.

The torque test is the check for a threaded stud, because a stud that passes a bend test can still spin under a nut if fusion is partial and the load path is torsional. Apply the specified proof torque with a calibrated wrench and hold it for a few seconds. The stud either holds or it turns. The proof torque comes from the stud grade and diameter, and the acceptance value belongs in the inspection plan rather than on the drawing.

The pull test quantifies what the other two reveal. Pull to failure in a tensile machine and record the load and the failure mode: stud fracture, weld fracture or sheet tear-out. The failure mode tells you which number in the table above you are actually sitting on, and it is the number a customer's quality engineer will ask for.

Judgement criteria for the weld itself come from the quality levels in ISO 5817, applied to the collar and the fusion boundary, and the absence of cracks in the heat-affected zone. ISO 14555 sets out the procedure qualification that proves the gun setting is right before production starts, and the welder qualification that proves the operator can repeat it.

What Ganesh runs, and where the stud process sits

Welding at Ganesh runs on three CLOOS robotic MIG cells and more than 30 manual MIG/MAG, spot and SPM stations, with welding fixtures and checking fixtures designed and built in the in-house tool room. The welded assemblies capability describes the route, and the welded enclosure article covers the panel work that most stud positions land on.

Stud welding is a separate process from MIG, with its own procedure qualification and its own inspection record. It is planned as its own operation, after forming and before pretreatment, so that the stud passes through the seven-tank line with the panel and a failed stud is found before the part is painted rather than after. What has to travel with the part is the qualification for that stud and sheet combination, kept with the job file next to the first article, which is the same traceability route used for the welded assemblies described in the welding inspection article.

The plant runs ISO 9001:2015 and ISO 14001:2015, works to a 0.8–40 mm thickness range and offers first article inspection and PPAP documentation. A stud pattern tighter than ±1 mm is a fixture decision rather than a marking-out one, and that decision belongs at quotation.

What to put on the drawing

Six fields stop most stud problems before the part is made.

• Stud standard and material, for example ISO 13918 type PT, grade 4.8, with the thread and the coating.
• Length after welding, with a tolerance, plus the thread length that must remain usable.
• Weld process where it matters: drawn arc above 1.5 mm, capacitor discharge below.
• Position tolerance for each stud, relative to the datum scheme used for the mating part.
• Symbol per ISO 2553, so the stud and the MIG welds are read from one drawing rather than a note that contradicts a view.
• Finish around the stud, stating whether the thread is masked, coated or left bare, because a coated thread is a second operation and a cost that arrives later.

Add the edge and bend distances only if your design sits near the limit. A stud pattern in the middle of a panel needs no note; a stud 10 mm from a flange needs one, and it should say so at the view rather than in a general note nobody reads.

When a stud is the wrong fastening

A stud needs one clean, flat face and access for the gun. Four cases break that, and the honest answer is a different fastener.

A closed box with a stud pattern inside cannot be welded after the box is closed, and welding before closing traps the gun in a cavity with no earth return. A stud on the curved outer face of a canopy louvre has no flat seating. A stud on 0.6 mm aluminium is outside what either process does well, and the aluminium assembly guidance in the spot welding article covers the alternatives that suit thin aluminium better.

And a stud carrying a load that varies in service, on a panel that flexes, is a fatigue joint with a stress concentration at the collar. If the joint carries more than a fixing load, a bracket welded or bolted through a doubler is the more predictable detail, and the quotation route can compare the two before the drawing is released.

Send the drawing with the stud positions, the sheet thickness and the length after welding, and the engineering team will check each position against the bend lines, the edge distances and the fixture before quoting.