Process and Alloy Are One Decision: Why Half of Aluminium Casting Drawings Specify Something That Cannot Be Made
The Quotation That Comes Back as a Question
A drawing arrives at a foundry specifying 6061-T6, high pressure Die Cast, 2 mm wall, 500 pieces per year. Every individual requirement is reasonable. The combination does not exist.
6061 is a wrought alloy — it hot-tears in a steel die. T6 is generally incompatible with high-pressure casting. A 2 mm wall rules out the processes that would accept T6. And 500 pieces per year does not amortise die tooling under any arithmetic.
This is not an unusual drawing. It is a common one, produced by an engineer who selected each parameter correctly in isolation. The error is structural: process and alloy were treated as two independent dropdown menus when they are a single coupled decision, and wall thickness and volume constrain both.
| Alloy / Material | HPDC | Gravity | Sand | Investment |
|---|---|---|---|---|
| A380 / ADC12 | ✅ Standard | ⚠️ Poor feeding | ⚠️ Rarely used | ❌ Not used |
| A356 / LM25 | ⚠️ Limited | ✅ Standard | ✅ Standard | ✅ Standard |
| A413 / ADC12 | ✅ Excellent | ⚠️ Possible | ⚠️ Possible | ❌ Not used |
| A360 | ✅ Standard | ⚠️ Possible | ❌ Rare | ❌ Not used |
| 6061 / 6063 (wrought) | ❌ Hot tears | ❌ Hot tears | ❌ Hot tears | ❌ Not viable |
| ZAMAK 3 / 5 (zinc) | ✅ Hot chamber | ❌ Not used | ❌ Not used | ❌ Not used |
Three cells in this table cause most of the rejected drawings.
Casting alloys carry 7–13% silicon because silicon does two things: it raises fluidity so metal reaches thin sections before freezing, and it narrows the freezing range so the alloy solidifies over a short temperature window rather than staying mushy.
6061 carries roughly 0.4–0.8% silicon. In a mould it has poor fluidity and a wide freezing range, which produces hot tearing — the casting cracks as it contracts while still partly liquid, restrained by the mould.
This is not a capability gap at particular aluminum die casting suppliers. It is the alloy's metallurgy. If a drawing requires 6061-T6 properties, the honest routes are Cnc Machining from billet, or aluminum fabrication from extruded profile. Any supplier who quotes 6061 as a casting is either misreading the drawing or planning a silent substitution to A356 — which is a different material with different properties.
A380 is engineered for high-pressure injection. Its wide freezing range is manageable when 1,000 bar of intensification pressure is feeding solidification shrinkage. Under gravity, with only the head of metal providing pressure, the same freezing range produces interdendritic shrinkage porosity throughout the section.
Gravity and sand routes therefore use A356 or LM25 — narrower freezing range, better natural feeding, and heat treatable.
Investment casting exists to deliver fine detail and near-net shape in alloys that are difficult to cast otherwise — stainless steels, titanium, nickel-based alloys. For aluminium, HPDC produces better dimensional consistency at a fraction of the piece cost above modest volumes. Aluminium investment casting appears mainly for aerospace components in A356/A357 where T6 properties plus intricate geometry are both mandatory and volumes are low.
During high-pressure die casting, metal enters the cavity at 30–60 m/s. That velocity is fundamentally turbulent — the metal front fragments and folds, entraining atmospheric air and vaporised die lubricant. Those gases end up as small, dispersed, high-pressure pores distributed through the casting.
While the part stays near room temperature, the pores are inert. They reduce fatigue strength modestly, but they cause no visible problem.
T6 requires a solution soak at 500–540°C for several hours. At that temperature, two things happen simultaneously: the trapped gas expands sharply according to gas law, and the surrounding aluminium softens toward its solidus. The pressurised gas lifts the softened skin outward, producing surface blisters — dome-shaped raised defects that are unrepairable and scrap the part.
| Requirement | Viable Route |
|---|---|
| Thin wall + high volume + as-cast properties | HPDC in A380 / A413, T5 ageing only |
| T6 properties + moderate volume | Gravity or low-pressure in A356 / LM25 |
| T6 properties + thin wall | Vacuum-assisted HPDC in A356, with section limits |
| T6 properties + intricate geometry + low volume | Investment casting in A356 / A357 |
T5 is the standard HPDC treatment — artificial ageing only, with no solution soak, so no gas expansion. T5 improves hardness and dimensional stability but delivers nowhere near T6 yield strength.
The engineering conversation worth having: does the design genuinely need 200+ MPa yield, or was T6 copied from a machined-part drawing? If the requirement is real, the process must change. If it was inherited by habit, as-cast A380 at roughly 160 MPa yield may be entirely adequate — and dramatically cheaper.
The measurable indicator is secondary dendrite arm spacing (SDAS) — the distance between the side branches of the solidification dendrites. Faster cooling produces finer SDAS, and finer SDAS produces higher strength, better elongation and better fatigue life.
| Process | Cooling Rate | Typical SDAS | Relative Strength |
|---|---|---|---|
| Sand casting | Slowest | 50–100 µm | Lowest |
| Investment casting | Slow | 40–80 µm | Low |
| Gravity permanent mould | Moderate | 25–50 µm | Medium |
| High pressure die casting | Fastest | 8–20 µm | Highest as-cast |
Two implications buyers should carry into specification.
First, a datasheet figure is not a part property. Published values assume a separately cast test bar with a defined cooling rate. Your part has variable section thicknesses, so a 3 mm wall and a 10 mm boss in the same casting have measurably different local properties. Where a specific strength is critical, specify testing on material cut from the actual part, not from a test bar.
Second, HPDC has the finest as-cast microstructure of any route. Its skin — the rapidly chilled outer 0.3–0.5 mm — is denser and stronger than the core. Machining that skin away on a functional surface removes the strongest material in the section and exposes the more porous core. Where a die cast housing has a critical stressed face, minimising machining allowance on it is a genuine design decision, not a cost detail.
Volume usually gets discussed first in sourcing conversations. Wall thickness is the harder constraint, because it eliminates options outright.
| Nominal Wall | Available Processes |
|---|---|
| 0.4–1.0 mm | Zinc casting only (hot chamber) |
| 1.2–3.0 mm | HPDC only |
| 3.0–5.0 mm | HPDC, gravity, investment |
| 5.0–12.0 mm | Gravity, sand, investment |
| Above 12 mm | Sand casting |
A part designed at 2 mm nominal wall has already selected high-pressure die casting, regardless of whether annual volume justifies the tooling. If volume does not justify it, the geometry must be redesigned to thicker sections for gravity casting — which changes weight, changes cooling performance, and may change the product envelope.
This is why wall thickness belongs in the first design review, not the sourcing review. By the time a drawing reaches procurement, the process has usually been chosen implicitly and irreversibly.
Two further couplings worth knowing:
Thin walls conflict with T6. The processes that hold thin walls (HPDC) are the ones that cannot take T6. Any specification demanding both a 2 mm wall and 200 MPa yield needs re-examination at concept stage.
Zinc reaches where aluminium cannot. Below roughly 1.2 mm, aluminium is not viable and ZAMAK is the only production option. For small components under 500 g where weight is not a driver, zinc also brings 500,000+ shot tool life against aluminium's 200,000–350,000, faster cycles, and a plating-ready surface.
Working in this order prevents the mismatch:
- 1Fix the mechanical requirement first— actual yield and fatigue needed, not inherited from a previous drawing
- 2Determine whether T6 is genuinely required— this single answer eliminates either HPDC or thin walls
- 3Set nominal wall thickness— this narrows the process list to one or two options
- 4Check volume against tooling amortisation— if the process indicated by wall thickness is not affordable at your volume, redesign the geometry now
- 5Select the alloy from those compatible with the surviving process— not from a general alloy table
- 6Confirm cosmetic and corrosion requirements against that alloy— A380 will not anodise bright; A360 resists salt better
- 7Then dimension and tolerance the drawing
Most rejected drawings result from running this sequence backwards — dimensioning first, then choosing an alloy from a materials handbook, then discovering at RFQ that the combination is not manufacturable.
Where casting is the wrong answer entirely: below roughly 200 units with tight tolerances, CNC machining from billet avoids tooling and delivers in one to two weeks. For brackets, frames and panels built from flat stock, custom fabrication requires no tooling at all. Many assemblies are best split — a cast body carrying the complex sealed geometry, with fabricated mountings attached — which also removes side actions from the die and reduces tooling cost.
No. 6061 is a wrought alloy containing only 0.4–0.8% silicon, giving it poor fluidity and a wide freezing range. In a mould it hot-tears, cracking as it contracts while still partly liquid. Casting alloys carry 7–13% silicon precisely to avoid this. If a drawing requires 6061-T6 properties, the viable routes are CNC machining from billet or fabrication from extruded profile. A supplier quoting 6061 as a casting is either misreading the drawing or intending to substitute A356, which has different properties.
Because of trapped gas. HPDC fills at 30–60 m/s, which is inherently turbulent and entrains air and vaporised die lubricant as dispersed high-pressure pores. T6 requires a solution soak at 500–540°C, where that gas expands sharply while the surrounding aluminium softens — lifting the skin into unrepairable surface blisters. HPDC parts are therefore supplied as-cast or T5 aged. For genuine T6 properties, use gravity or low-pressure casting in A356, or vacuum-assisted HPDC with section thickness limits.
A380 and A413 are engineered for high-pressure die casting and perform poorly under gravity, where there is no intensification pressure to feed their wide freezing range. A356 and LM25 are the standard choices for gravity, sand and investment casting, and they are heat treatable to T6. A360 is a die casting alloy chosen when corrosion resistance outranks strength. Wrought alloys such as 6061 and 6063 are not castable by any route. Zinc ZAMAK alloys run only in hot-chamber die casting machines.
Cooling rate, measured as secondary dendrite arm spacing. Faster solidification produces finer dendrite spacing and therefore higher strength, better elongation and better fatigue life. Sand casting cools slowest at 50–100 µm SDAS; high pressure die casting cools fastest at 8–20 µm. This also varies within a single part, because a 3 mm wall and a 10 mm boss cool at different rates. Where a specific strength is critical, specify testing on material cut from the actual casting rather than from a separately poured test bar.
Zinc die casting reaches 0.4–1.0 mm. Aluminium high pressure die casting holds 1.2–3.0 mm reliably. Gravity permanent mould needs 3–5 mm minimum. Sand casting needs 5–6 mm and suits sections above 12 mm. Wall thickness is a harder constraint than volume because it eliminates processes outright — a part designed at 2 mm nominal has already selected HPDC, whether or not the annual quantity justifies the tooling investment.
Minimise it where you can. The rapidly chilled outer 0.3–0.5 mm of a die casting is denser and stronger than the core, so machining a functional surface removes the best material and exposes more porous metal underneath. On critical stressed faces, sealing faces and areas that will be leak tested, keeping machining allowance minimal is a design decision with real mechanical consequences, not just a cost saving.
Fix the true mechanical requirement first, then decide whether T6 is genuinely needed, then set nominal wall thickness — those three answers usually leave only one viable process. Check that process against your volume and tooling budget, and redesign the geometry now if it does not fit. Only then select the alloy from those compatible with the surviving process, confirm it against cosmetic and corrosion requirements, and dimension the drawing last. Most unmanufacturable drawings result from running this sequence in reverse.