Why Two Aluminum Casting Quotes for the Same Part Can Differ by 60%

A client sent the same STEP file to four foundries last spring. The quotes came back at £4.20, £5.05, £6.10 and £6.80 per piece. Same part, same alloy, same annual volume of 12,000.

He assumed three of them were padding. Actually all four were honest. They were just sitting on different machines, quoting different scrap rates, and two of them had silently redesigned his runner system.

Here's what sits behind that spread, and how to read it.

The tonnage bracket problem

Clamp force is the first thing that moves a quote, and it moves in steps rather than smoothly.

You need enough clamping tonnage to hold the die shut against the metal trying to push it open. Roughly, that's your part's projected area plus the runner and overflows, multiplied by cavity pressure. Land at 780 tons of requirement and you're running on an 800-ton machine. Land at 810 and you're on a 1250, because nothing exists in between.

That jump is expensive. A 1250-ton machine costs more per hour than an 800, and it isn't a small increment.

Which explains something buyers find suspicious. One supplier quotes low because their 800-ton press is a genuine fit. Another quotes high because their smallest suitable machine is a 1250 and they'd be running your small part on an oversized press. Neither is lying. The second one just isn't the right shop for this job.

So ask early: what tonnage will this run on, and how many of those do you have? A shop with one 800-ton press and a full order book will quote defensively. A shop with four will not.

Alloy naming is a genuine trap

This one causes real confusion on cross-border projects, and it costs people money.

A European engineer writes EN AC-46100 on the drawing. An Asian supplier reads it, supplies ADC12, and both parties think they've agreed. They roughly have — those designations broadly correspond. But the compositional windows overlap rather than match exactly, particularly on iron and zinc.

Iron content matters more than you'd expect. Too little and the casting solders to the die face, wrecking cycle time and eventually the tool surface. Too much and the alloy gets brittle. The usual working band sits around 0.8 to 1.1%, and a melt built mostly from reclaimed scrap can drift outside it.

None of this is a reason to avoid ADC12 or Asian Aluminum Die Casting suppliers. It's a reason to specify the composition limits directly on your drawing rather than trusting a designation to carry them, and to ask for the spectrographic certificate per heat rather than a generic datasheet.

Same principle applies to zinc casting, implicitly. Zamak 3 and Zamak 5 look interchangeable on paper until the plating adhesion comes back different.

Runner yield — the number nobody quotes explicitly

Here's a cost line most buyers never see broken out.

When aluminum is injected, you're not just filling the part. You're filling the runner system, the biscuit at the shot sleeve, and the overflow tabs that catch the first cold slug. All of that solidifies too. All of it gets trimmed off and thrown back in the furnace.

Yield — part weight divided by total shot weight — typically lands between 50% and 70% on aluminum work. A 400-gram part might need a 650-gram shot.

Remelting recovers the metal but not the energy, and it isn't free. Every remelt cycle also burns off some magnesium and picks up a little oxide, so foundries can't run 100% reclaim indefinitely.

Where this hits your quote: a supplier who has designed a tight, efficient gating system quotes a lower metal cost than one who's added generous runners to guarantee fill. The second approach is safer for them and more expensive for you. If a quote seems high, ask what shot weight they've assumed. The answer is revealing, and honest shops will tell you.

Scrap allowance, and why nobody advertises it

Every foundry builds an expected scrap rate into the price. On straightforward parts it might be 2 to 3%. On thin-wall work with a tight leak spec it can reach 8 to 12%.

That number reflects genuine process risk. A part with a demanding cosmetic surface, a difficult flow path, or a helium leak requirement will fail more often, and someone pays for the failures.

The problem arises when a supplier quotes an optimistic scrap rate to win the job, then discovers reality at production. What follows is either a price renegotiation or a quality argument, and both cost more than the original difference would have.

Worth asking directly: what scrap rate have you assumed, and what drives it on this specific part? A foundry that answers with a number and a reason has modeled your job. One that says "we don't really have scrap" has not.

What actually makes a part expensive

Setting aside the market, these are the levers inside your own drawing.

Wall thickness is the big one, and it works in both directions. Thick walls mean more metal, longer solidification, slower cycles. But go too thin and fill becomes marginal, scrap climbs, and you've traded metal cost for reject cost. Somewhere around 2.0 to 3.0 mm is comfortable for general aluminium casting products.

Side actions each add tooling money, cycle seconds and a maintenance point. Converting an undercut to a drilling operation after casting is often cheaper across the whole program, even though it adds a machining step.

Tolerance callouts get applied far too liberally. As-cast aluminium casting parts hold roughly ±0.13 mm. Anything tighter forces machining. I've seen drawings where ±0.05 mm was specified on a clearance hole, which added a fixture, a setup and two minutes of cycle time to every part for no functional reason at all.

Cosmetic class does the same thing quietly. Class A visible surfaces need slower fills, better die maintenance and higher scrap allowance. Specify it where it's seen and nowhere else.

Cavity count is straight arithmetic. Two cavities roughly halve machine cost per piece and add 60 to 70% to the tool. Whether that pays depends entirely on volume, so run it against a realistic forecast.

Distributed sourcing, and what it hides

Online platforms and broker networks have made it easy to get an aluminum casting quote in 24 hours from a partner you'll never meet. That's genuinely useful. It's also worth understanding what you're trading.

When something goes wrong at shot 4,000 — and eventually something does, usually die soldering or a dimension drifting as the cavity wears — you want a direct line to the person who cut the tool. A well-run network provides that. A thin one gives you an account manager relaying messages across a timezone gap, and each round trip costs a day.

Ask who owns the tool and whether it can be transferred. Get it in writing before the tooling PO, not after. Tool ownership disputes are the single most common way a casting program gets stuck, and they're entirely preventable with one clause.

Also ask where the tool physically lives. "We have 260 partners" is a capacity statement, not an answer about your specific die.

Things worth confirming before you commit

• Whether the quote includes trimming, deburring, machining and finishing, or just the raw casting. This alone explains a lot of price gaps.
• What leak rate they test to, in mbar·L/s, and whether testing is 100% or sampled. "Pressure tight" isn't a specification.
• Whether T5 or T6 is assumed. Conventional die castings blister above 500 °C because entrapped gas expands, so T6 usually means moving to permanent mold in A356 or paying for vacuum-assisted high-integrity casting. A supplier who agrees to T6 on a standard die casting without qualifying it hasn't thought it through.
• Whether the price includes first article inspection and a dimensional report, or whether that's extra.
• Their machine list. Not "we have modern equipment" — the actual tonnage range, because that tells you whether your part fits comfortably or awkwardly.

Where aluminum castings end up

Electronics and energy work absorbs die cast housings, heat sinks and enclosures in volume. Fluid power takes valve bodies, pump housings and manifolds, where pressure tightness drives everything. Automotive and EV pulls motor housings, battery trays and structural nodes, increasingly at tonnages that didn't exist commercially ten years ago. Plumbing runs cast metal pipe fittings and flanges, and most any pipe fittings manufacturer will have aluminum, brass and zinc cells running in parallel. Industrial machinery mixes castings with custom metal fabrication constantly.

That last combination is worth sitting with. Casting and structural fabrication aren't competitors. A cast node consolidates what would otherwise be six welded plates, but the frame around it still gets cut, formed and welded. A supplier who offers casting services and custom fabrication together can move that boundary to wherever total cost lands lowest — and it's rarely where either specialist alone would have drawn it.

Back to those four quotes. My client went with the £5.05 — not the cheapest. That supplier had returned the model with a note proposing a parting line 11 mm from where he'd drawn it, which killed a slide and took roughly £7,000 off the tool.

The £4.20 quote came with no comments at all. Which, once you've been through this a few times, stops looking like a bargain and starts looking like a shop that hasn't opened the file yet.