Your Aluminium Casting Cost Was Decided Before You Requested a Quote

The 70% You Already Spent

By the time a drawing leaves engineering and reaches procurement, roughly 70% of the component's lifetime cost is already committed. Not spent — committed. The money leaves the business later, but the decisions that determine how much will leave were made when someone chose a wall thickness, positioned a boss, and set a tolerance.

Procurement then negotiates over the remaining margin. A skilled buyer might extract 8–12% from a supplier through competitive tension. A single wall-thickness change made three weeks earlier could have removed 30%.

This is why the most valuable conversation in any custom Die Casting programme happens before the RFQ goes out, not after the quotes come back. What follows are the specific design decisions that move aluminium casting cost — and the numbers attached to each.

Where Cost Actually Gets Locked In

The cost curve behaves counter-inituitively. Spending is back-loaded — tooling invoices and piece-price payments land late. Commitment is front-loaded.

A wall thickness reduced from 4 mm to 2.8 mm at concept stage costs nothing to implement and removes roughly 25% of the part's aluminium mass, shortens cycle time by 15–20%, and reduces the shrinkage risk that drives scrap. The identical change requested after the die is cut requires cavity welding and re-machining, or a new insert.

Wall Thickness: The Single Highest-Leverage Variable

Wall thickness drives four cost components simultaneously, which is why it outranks every other design decision.

Material mass. Aluminium is priced per kilogram. A 20% wall reduction across a part removes close to 20% of raw material cost directly.

Cycle time. Solidification time scales approximately with the square of section thickness — Chvorinov's rule. Halving a wall does not halve the cooling time; it cuts it to roughly a quarter. Since machine time is the dominant conversion cost in Aluminum Die Casting, this is where the real saving sits.

Scrap rate. Thick sections solidify last, shrink most, and generate the porosity that fails leak tests and X-ray acceptance.

Machine size. Thinner walls mean lower shot weight and smaller projected area, which can move a part down a clamp-force tier — from a 900-ton machine to a 600-ton machine — with a corresponding drop in hourly rate.

Target Wall Thickness by Process

The instinct to add thickness "for strength" is usually wrong in cast metal design. A 2.5 mm wall with a properly proportioned rib is stiffer, lighter, cheaper and far less porosity-prone than a 5 mm plain wall. Stiffness scales with the cube of section depth — a rib gives you depth without mass.

Uniformity Matters More Than Absolute Value

A part with walls uniformly at 3 mm casts more reliably than one alternating between 2 mm and 5 mm. Where a transition is unavoidable, taper it over a length of at least three times the thickness difference. Abrupt steps create the isolated thermal mass that gates cannot feed once they freeze.

Draft, Fillets and Bosses: Small Rules, Large Consequences

Draft Angle

Draft is the taper that allows the part to release from the die. Insufficient draft causes drag marks, ejector push-through, and in severe cases the casting hanging up and damaging the die surface.

Internal surfaces need more draft than external because the casting shrinks onto internal cores as it cools, increasing extraction friction. This asymmetry surprises designers coming from injection moulding, where the differential is smaller.

Fillet Radius

Sharp internal corners do three destructive things at once: they concentrate stress in the finished part, they concentrate heat during solidification (creating a hot spot that shrinks), and they concentrate thermal fatigue stress in the die steel, where they become the origin point for heat-checking cracks that end die life prematurely.

Specify an internal fillet radius of at least 1× the adjoining wall thickness. External corners take 1.5×. A zero-radius corner on a drawing is almost always an oversight rather than a requirement, and it is worth challenging every time.

Boss Design

Bosses are where isolated thermal mass hides. A solid boss with an outer diameter more than 2.5× the adjoining wall thickness will shrink internally, producing either sub-surface porosity or a visible sink mark on the opposite face.

Three corrections, in order of preference:

1. Core out the boss centre — a cored hole removes the thermal mass entirely and gives you a pilot for tapping
2. Add connecting ribs — three or four radial ribs break up the mass and improve feeding
3. Reduce boss diameter — often the specified diameter was chosen for convenience, not function

Parting Line: Negotiate It, Do Not Inherit It

The parting line determines which surfaces carry a witness mark, which features need side actions, and how much flash requires removal. It is one of the few design decisions that is genuinely collaborative — the designer knows which surfaces are cosmetic or functional; the tooling engineer knows which split minimises complexity.

Discuss it at DFM. Once the die is cut, moving the parting line means rebuilding the cavity.

Side actions (slides and lifters) form undercuts that a simple two-part die cannot produce. Each one adds tooling cost, adds cycle time (the slide must retract before ejection), and adds a maintenance point that will eventually cause downtime. Before accepting a side action, ask whether the feature can be:

• Re-oriented to align with the draw direction
• Moved to the parting line where it forms naturally
• Machined post-cast, which is frequently cheaper than a slide across the tool's full life
• Eliminated by rethinking how the part mounts

A single avoided side action typically pays for several hours of DFM engineering time.

Tolerance Discipline: Where Drawings Quietly Add Cost

Every tightened tolerance either forces a machining operation or narrows the accepted process window. Both raise cost — the first directly, the second through scrap.

The most common and most expensive drawing error is applying a blanket tolerance block to every dimension. On a part with 60 dimensions, perhaps six are genuinely critical to function — bearing bores, sealing faces, threaded ports, mounting datums. The remaining 54 can run at general casting tolerance per NADCA Product Standards or BS EN ISO 8062 with no functional consequence.

Datum strategy matters too. Establish datums on cast features that are stable and repeatable — never on a parting-line surface, where flash and die wear introduce variation. Poor datum selection creates measurement disputes that consume months of engineering time and produce no improvement in the actual part.

Designing Around Defects Instead of Inspecting for Them

The pattern is consistent: the design fix is free and permanent, the production fix is recurring and imperfect. This is the strongest argument for running solidification simulation — MAGMA, ProCAST or Flow-3D — during design review. Simulation predicts hot spots, air entrapment zones and last-to-freeze regions before steel is cut, converting expensive physical trial iterations into cheap virtual ones. A supplier who runs simulation only after a failed T1 trial is using it as a diagnostic tool. Used properly, it is a design tool.

Choosing the Cheapest Adequate Process

Cost engineering also means questioning whether casting is the right route at all.

Fabrication as a genuine alternative. For brackets, frames, enclosure panels and chassis built from flat and bent stock, custom fabrication eliminates tooling entirely and delivers in 1–3 weeks. Two constraints apply to structural fabrication in aluminium: welded 6061-T6 loses roughly 40% of its yield strength in the heat-affected zone, so load calculations must use the annealed allowable; and weld sequence with fixturing determines whether the assembly holds flatness.

Hybrid designs frequently win. A die cast housing for the complex sealed body joined to fabricated brackets for the mounting interface often costs less than forcing either process to do the whole job. The mounting geometry that would demand three side actions in a die becomes two bends and a weld in aluminum fabrication.

Material route matters as much as process. If a drawing specifies 6061-T6, that part cannot be cast — 6061 hot-tears in a steel die. It must be machined from billet or built from extruded profile. Conversely, specifying A380 for a part that will be anodised bright silver guarantees disappointment: its 3–4% copper anodises to a mottled dark grey. These are aluminum alloy processing manufacturing constraints, and they belong in the design review, not the corrective action report.