Aluminium 5083 Hatch Covers: Porosity, Soot and Crater Cracks
A workshop is building twenty aluminium hatch covers from 6 mm 5083-H111 plate, with fillet and butt welds made by MIG using 1.2 mm 5356 wire and argon. The roller door is kept open for ventilation, and the shop is beside a busy yard. The first four covers show heavy pinhole porosity along the fillets, a black sooty smut on and beside the welds, and short cracks at the ends of several runs where the welder stopped. Gas flow is set at 18 l/min, the plates are wiped with a rag and cutting fluid is still visible on some edges, and the same wire brush is used on both the aluminium bench and the mild steel bench. The customer needs the covers in six days and the aluminium plate is expensive, so scrapping the four covers is not attractive.
1. Rank the likely causes of the porosity in this shop and set out the cleaning and process changes you would enforce before cover 5 is started.
Read the worked answer
Ranked causes.
- Hydrocarbon and oxide contamination on the faying surfaces. Cutting fluid, marker pen, adhesive residue and the naturally occurring aluminium oxide (melting point around 2070 °C against 660 °C for the metal) all generate hydrogen and entrapment in the pool. This is the dominant cause here, since fluid is visibly present.
- Contaminated wire brush. The same brush used on mild steel transfers embedded steel particles into the aluminium, producing porosity, smut and a corrosion cell. This is a classic, wholly avoidable workshop error.
- Gas shielding disturbed by drafts. At 18 l/min with a roller door open, the argon shroud is blown away or becomes turbulent and entrains air. Flow should be roughly 12–14 l/min with a windbreak.
- Moisture - condensation on cold plate, a damp spool, or a leaking gas line drawing in moist air.
- Technique - excessive stick-out, a low push angle, or long arc lengthening the path for air to reach the pool.
Enforced changes.
- Degrease with acetone or an approved aluminium cleaner, then wire brush with a stainless brush reserved for aluminium only, then degrease again, and weld within about an hour of cleaning.
- Store wire in a dry cabinet; discard corroded spools. Check hoses and torch O-rings.
- Close or screen the door, fit a windbreak, and set flow to 12–14 l/min with a flow meter at the torch.
- Push technique at 10–15° travel angle, short arc, minimal stick-out.
- Run a test coupon and cut it to check porosity before recommitting the plate.
2. Explain the crater cracks at the end of the weld runs and why filler choice matters on 5083. What would happen if someone substituted 4043 wire to save time, and what would you specify instead?
Read the worked answer
Why crater cracks form. A crater is the last metal to solidify, and it solidifies while the surrounding weld is already contracting. The shrinkage strain concentrates in a short, thin section with a concave profile, and if the pool is simply switched off, that strain tears the crater. Fixes are procedural: use the set's crater-fill or burn-back function, reduce current progressively over the last 10–15 mm, run onto a run-off tab or into the corner, and use a backstep sequence so run terminations land in already-welded or non-critical regions.
Why filler choice matters on 5083. 5083 is a non-heat-treatable Al-Mg alloy with roughly 4.0–4.9% magnesium and a narrow freezing range; matching composition fillers (5356 with about 5% Mg, or 5183 with about 4.8% Mg plus tighter impurity limits) give a weld metal that solidifies with the parent and retains strength and corrosion resistance. Filler selection on aluminium is about balancing crack sensitivity against strength and ductility, not just about matching numbers.
If 4043 were substituted. 4043 is an Al-5% Si filler. In a 5083 weld it produces a dissimilar Al-Mg-Si weld metal: magnesium from the parent combines to form Mg2Si, which reduces ductility and toughness, and the weld metal loses strength relative to 5356. Corrosion resistance in a marine or washdown environment also suffers. On top of that, 4043 anodises to a different colour (it goes dark grey/black), which is immediately visible on a finished hatch cover.
Specification. 5356 for general work; 5183 where higher deposited strength and better performance in a marine environment are called for, or where the finished part is to be anodised and a colour match is required.
3. The four completed covers are dimensionally acceptable but porous. Describe how you would assess and repair them, and what you would change in the fabrication sequence to limit distortion given aluminium's properties.
Read the worked answer
Assessing the covers. Porosity on aluminium is not judged by eye alone. Cut a section or take a radiograph across the worst run and compare against the acceptance criteria in ISO 10042 (the aluminium equivalent of ISO 5817) at the level the drawing calls for. Isolated surface pinholes that clean up within the specified allowance can be dressed; continuous interconnected porosity, or porosity in the throat of a fillet, is a rejectable defect that must be removed.
Repair. Gouge or grind out the porous region back to sound metal, extending 25 mm beyond each end, re-clean with the dedicated stainless brush and solvent, then re-weld with the corrected parameters and a crater-fill finish. Aluminium loses strength permanently if overheated, so do not try to burn out porosity with extra current. Where several repair passes would be needed on the same area, replace the section instead - repeated rework is a false economy and the customer's inspector will see it.
Distortion control. Aluminium has about three times the thermal expansion of steel and much higher thermal conductivity, so contraction is large but spreads over a wide area. To control it:
- Fit-up accurately and tack frequently; poor gaps force operators to fill volume and put in heat.
- Use the minimum number of passes with the highest practical travel speed.
- Backstep and balance welds about the neutral axis, alternating sides of the frame.
- Clamp or jig the panel to a flat bed or use strongbacks, and let it cool in the jig.
- Keep interpass temperature controlled; do not stack hot covers on top of one another.
- Remove the black smut afterwards with the dedicated brush and hot soapy water so the oxide layer can re-form and the surface is fit for painting or anodising.