Rivets Pulling Through an Aluminium Canopy After One Week

Forming causes:

• Bend radius too tight for the temper. A 1.5 mm inside radius on 1.5 mm sheet is 1.0t. That is workable for soft (O) aluminium bent across the grain, but H14 is half-hard and work-hardens further as it bends. The usual guidance is a minimum inside radius of about 1–1.5t across the grain and 2–3t along the grain for H14 - and 1.0t is at or beyond the tight end.
• Bend direction. The folds run across the full 12 m length. If the sheet's rolling (grain) direction also runs that way, every fold is along the grain - the worst case, giving orange-peeling on the outside of the bend and initiating cracks that then run.
• Tooling carried over from galvanised work. Galvanised mild steel tolerates a 1t radius; aluminium does not. Same tool, different material, different result.

Fixing causes:

• Edge distance. A 4.8 mm rivet at 6 mm from the edge gives roughly 1.25d. The usual rule is at least 2d from the edge to the hole centre (about 10 mm here) and about 2.5d pitch. The 100 mm pitch is generous, but the edge distance is far too small, so the sheet tears out to the edge under load and thermal movement.
• Punched, un-deburred holes. Punching leaves a ring of work-hardened, micro-cracked material round every hole; combined with a hole only 6 mm from a fold, that is a ready-made crack starter.
• Over-fixing. Rivets at 100 mm centres along a 12 m pan rigidly restrain the sheet. Aluminium expands about 23 × 10⁻⁶ per °C, so a 12 m pan moves roughly 11 mm over a 40 °C swing. Fixed solid at both ends, that movement has to go somewhere - it goes into the holes.

Diagnosis: elongation plus edge tear-out plus cracking at the fold indicates both a forming problem and a fixing detail problem; treating only one will not stop it.

It is galvanic corrosion, and it is entirely predictable from the materials list. Zinc-plated steel pop rivets through aluminium sheet, wetted by rain, form a corrosion cell. Aluminium sits low in the galvanic series relative to steel and zinc, so it becomes the anode and corrodes around each rivet; the white product is aluminium corrosion. The mild steel angle behind the sheet makes it worse wherever water sits between the two metals.

Aggravating factors here:

• The joint traps water between the pan and the angle, and there is no drainage or ventilation.
• Un-deburred holes hold moisture and break any passive oxide film.
• The canopy is external and will stay wet for long periods.

Changes:

1. Isolate the dissimilar metals. A separation layer between the aluminium and the steel angle - EPDM tape, a mastic strip, or a coated separator. Do not rely on paint on the steel; it will be damaged during fixing.
2. Change the fastener material. Aluminium or A4 stainless rivets are far better than zinc-plated steel. Stainless against aluminium is still a couple, so seal the rivet and the hole with a sealant as well, and never use plain steel washers.
3. Seal the holes during fixing so water cannot sit in the annulus.
4. Detail for drainage and ventilation: break the contact line so water runs off instead of being trapped, and avoid closing the pan tight to the frame along its full length.
5. Protect the steel frame properly - galvanising or a maintained coating system - and accept that the coating is the second line of defence, not the first.

Note the timescale: one week to visible corrosion product is fast, which suggests water is pooling rather than merely wetting. Fix the drainage detail at the same time or the material changes will only slow it down.

Revised detail:

• Material condition: specify O temper and allow the pans to be formed and then finished, or retain H14 but open the bend radius to at least 3 mm inside (2t) and, where possible, orient the fold across the grain. Confirm the grain direction with the supplier on the delivery paperwork rather than assuming.
• Holes: drill, do not punch, and deburr both faces. Increase edge distance to at least 10 mm from the edge to hole centre (2d), keep pitch around 25–40 mm locally at the ends and no tighter than about 2.5d generally - 100 mm pitch is fine as a maximum but must not be combined with rigid end fixing.
• Thermal movement: fix the pan rigidly at one point per pan - the centre or one end - and use clearance or slotted holes everywhere else, with sealing washers. Allow about 11 mm of movement per 12 m run over a 40 °C swing. Better still, use a proprietary clip or standing-seam system designed for the movement rather than a rivet pattern.
• Isolation: separation tape between aluminium and steel angle; stainless or aluminium fasteners with sealant.
• Drainage: set the pans to a fall, break the trapped contact line, and let water leave rather than sit.
• Watertightness: sealant alone will not survive 11 mm of movement; the watertightness has to come from the geometry of the upstand and laps, not from a bead of mastic.

Sequencing: the frame is already fixed, so measure the actual frame before remaking pans; do not assume the frame went in to the drawing any more than the duct would. Remake the two failed pans rather than patching them - a pan with a crack out to the edge has already told you its load path is wrong.