The Casting Shot Is 40% of Your Part Cost. Here's the Other 60%.
A buyer once asked me why his casting quote had gone up 30% when the part hadn't changed. It had, though. He'd added two threaded inserts and specified a bead-blast finish. The shot was identical. Everything after it wasn't.
That conversation gets repeated across the industry constantly, because the phrase "Die Casting quote" implies the price is about casting. It mostly isn't.
Where the money actually goes
Pull apart a typical unit price on a mid-complexity aluminium casting product and the shot itself — melt, inject, solidify, eject — accounts for something like 40%. The remaining 60% happens after the part is already out of the die.
Those percentages shift with part complexity, and they're indicative rather than universal. But the shape holds: machining is usually the single largest post-cast item, and finishing plus assembly together often rival the casting itself.
Which explains something buyers find baffling. Two shops quote the same part. One is 35% higher. Nobody is lying — one of them read the drawing and priced eight operations, the other priced a casting and will discover the rest later.
Gate trim, and the thing nobody photographs
The part that comes out of the die isn't your part. It's your part plus a runner system, a biscuit, and a row of overflow tabs — sometimes 40% more metal than the finished component.
Trimming that off happens in a trim die on a press, which is a second piece of tooling most quotes bundle silently into the die cost. Ask whether it's included. On a complex part it can be 15 to 25% of the tooling bill on its own.
What survives the trim is flash along the parting line and a witness mark where the gate was. Both need dealing with. Vibratory deburring in a tumbling media bowl handles the general case at low cost per part, but it rounds every edge equally — including ones you wanted sharp. Manual deburring keeps control and costs real labour. Robotic deburring sits between them and only pays back at volume.
The design lever here: put the gate somewhere the witness mark doesn't matter. That's a conversation to have during DFM, not after the first article shows a scar on a cosmetic face.
Machining is where setups, not cuts, cost money
Twenty percent of unit cost, give or take, and the intuition about what drives it is usually wrong.
It isn't cutting time. On aluminum, metal removal is fast. What costs money is every time the part gets unclamped, reoriented, refixtured and re-probed. Each setup adds handling labour, a fixture, and a tolerance stack-up between features machined in different orientations.
So the question isn't "how much machining does this part need" but "how many setups". A housing with machined features on four faces needs four operations unless somebody designs it otherwise. Move those features onto one or two planes and you've cut the machining cost substantially without changing a single tolerance.
Two more things worth knowing. Cast bosses should sit slightly proud where they'll be faced, so a light skim cleans them rather than a heavy cut fighting the hardened as-cast skin. And machining into a subsurface void is the most expensive failure mode in the whole chain, because you've added all your machining value before discovering the part was scrap — which is the real argument for vacuum assist on anything that gets heavily machined.
Tapping, and why aluminum threads fail
Small line item, disproportionate warranty exposure.
Aluminum is soft. A tapped thread in A380 strips at a fraction of the torque a steel thread takes, and if the assembly will be opened and closed repeatedly — a service cover, a battery hatch — the thread degrades every cycle.
Three routes. Direct tapping is cheapest and fine for permanent, low-torque joints. Thread inserts, whether helical coil or press-in, add cost per hole but give you steel threads in an aluminum part. Self-tapping screws into a cast pilot boss work for plastics-style assembly at low torque.
The mistake I see: specifying direct tapping on a part that gets serviced annually, then discovering the failure in year two across the installed base. If a joint will be opened more than a handful of times, insert it.
Surface finishing is a materials decision, not a colour choice
This is where high-silicon Die Casting Alloys bite people.
A380 contains around 8-9% silicon, which is what makes it castable. Anodize it and the silicon doesn't convert to oxide the way aluminum does — it stays as dark particles in the film. The result is grey, slightly mottled, and nothing like the clean anodized finish on a machined 6061 part. Customers who approved a colour chip based on a wrought sample are routinely disappointed.
Workable options: Powder Coating gives excellent coverage and colour control and doesn't care about silicon; chromate conversion under MIL-DTL-5541 gives corrosion protection and paint adhesion cheaply; bead blasting produces an even matte that hides minor cast texture. Anodizing is fine when you accept a grey industrial tone rather than a decorative one.
Zinc Casting is the opposite case — it plates beautifully, chrome included, which is exactly why perfume caps and decorative hardware are zinc rather than aluminum despite the weight.
The material question, extended
Most articles stop at aluminum versus zinc. Magnesium deserves a mention because it's the only one that beats aluminum on weight.
AZ91D runs about 1.8 g/cm³ — a third lighter than aluminum — and holds thinner walls, which is why laptop chassis and camera bodies use it. The costs are real though: shorter tool life, tighter process control, and a melt that requires cover gas because magnesium oxidises aggressively. Fewer shops run it, and those that do charge for the capability.
Rough allocation: aluminum for thermal performance, corrosion resistance and general housings; zinc for fine detail, thin walls, long tool life and genuine plating; magnesium when grams decide the product and the budget tolerates it.
Why in-house matters more than it sounds
Here's the operational argument for a supplier who runs casting, machining and finishing under one roof.
Each transfer between vendors adds transport days, an incoming inspection, packaging, and — most expensively — a place for accountability to disappear. When a machined casting fails a leak test, the caster says the machining opened a subsurface void and the machinist says the casting was porous. Both can be right. Neither pays.
A single-source shop internalises that argument. There's also a real scheduling gain: a part that moves between four vendors accumulates queue time at each, and queue time usually exceeds process time by a wide margin.
The counter-argument is genuine capability. A shop that casts well and machines adequately may not beat a specialist machinist. Worth asking what's actually in-house versus subcontracted, because "one-stop" appears on a great many websites and means different things behind each one.
What to send with an RFQ
Suppliers price what they can see. Incomplete information gets priced pessimistically, or optimistically and then revised.
Send the 3D model plus a 2D drawing showing which dimensions are critical — not every dimension toleranced, which forces machining everywhere. State the alloy and whether equivalents are acceptable. Specify the finish including the standard reference, not just a colour. Say which threads are inserted. Give realistic annual volume and the batch pattern, because 12,000 a year in monthly releases prices differently from 12,000 once. And state your test requirement — leak rate in mbar·L/s if it's a pressure part, since that single line changes vacuum requirements, alloy choice and inspection cost.
Then ask what the quote covers. Casting only, or trimmed, machined, finished and packed? Most price gaps between suppliers dissolve once you align that boundary.
Where the chain lands
Electronics runs die cast housings and heat sinks — cast, faced, tapped, chromated. Fluid power takes valve bodies through cast, multi-axis machining and leak test, where every hidden void surfaces late and expensively. Automotive pulls structural fabrication nodes and motor housings with full traceability under IATF discipline. Plumbing runs cast metal pipe fittings and flanges; a pipe fittings manufacturer typically keeps aluminum, brass and zinc cells with separate melt records and its own thread-gauging cell.
Industrial machinery mixes castings with custom metal fabrication as standard, and the finishing chain is where the two disciplines actually converge. A cast node still needs machining, coating and assembly — the same operations a fabricated weldment needs. Which is why a supplier offering both casting services and custom fabrication can move the cast-versus-fabricate boundary based on total processed cost rather than on which department is busy.
The buyer with the 30% increase ended up dropping one insert and switching from bead blast to as-cast on the hidden faces. Price came back most of the way down.
Nothing about the casting changed. It never was the casting.