CNC Machining vs Die Casting: Volume Is the Whole Argument

Two different economic logics

The two processes are not variations on a theme. They are different ways of putting a shape into metal, and the economics are opposite.

Die casting melts an alloy and injects it under pressure into a hardened steel die. The die is the expensive part — a substantial piece of tooling, machined to tolerance, heat treated, and specific to one component. Once it exists, parts come out of it in seconds with very little labour, so the unit cost is low and stays low across a large quantity. The die cost is spread across every part produced from it.

CNC machining starts from a solid block, bar or casting and cuts the shape out of it. There is no die; the investment is in fixtures, programming and setup, which is small relative to a die. Every part then carries its own cycle time, tooling wear and operator attention. The unit cost is higher and does not fall much with quantity once the setup is amortised — see CNC machining for how that route is quoted.

So the comparison is not “which makes a better part”. It is:

• Die casting — high fixed cost, low variable cost
• Machining — low fixed cost, high variable cost

Everything else follows from that, including where the two cross.

	Die casting	CNC machining
Starting point	Molten alloy injected into a die	Solid stock
Tooling	Hardened steel die — substantial	Fixtures and programs only
Unit cost at low volume	High — die not amortised	Lower
Unit cost at high volume	Low — die amortised	Higher — cycle time per part
As-supplied tolerance	Loose	Tight
Draft angle required	Yes, on all vertical walls	No
Wall thickness	Must be uniform	Any
Internal porosity	Inherent to the process	None
Material options	Castable alloys only	Any grade, wrought or cast
Surface finish as supplied	Good for a casting	Machined
Design change cost	Die modification or remake	Program change
Lead time to first part	Long — the die has to be made	Short
Best quantity	High	One-off to medium
Typical parts	Housings, enclosures, brackets at volume	Everything else

Two rows in that table are the ones buyers most often overlook: the draft angle and the material. Both are consequences of the process, and both change the design rather than just the price.

The crossover, and why there is no universal number

There is a volume above which die casting is clearly cheaper and below which machining is clearly cheaper. Where that volume sits is different for every part, and it is set by four things.

Part size. A larger part needs a larger die, and die cost scales with the size of the tooling. It also needs a bigger machine to run it.

Part complexity. A complex casting that removes a great deal of machining earns more at a lower volume, because the alternative — cutting all those features from solid — is expensive per part. A simple casting that saves only a little machining needs a higher volume to justify the die.

How much machining it removes. This is the calculation that is usually done badly. The relevant saving is not the cost of the casting against the cost of the block; it is the total machined cost of the part that the casting replaces. If the cast form removes half the machining time, that saving applies to every part and pays back the die.

How much machining is still needed. Almost every die casting still has critical features machined. If the casting still needs most of the machining operations, the saving is much smaller than the raw comparison of material costs suggests.

The practical approach is not a rule of thumb but two quotations at your actual quantity — one for the machined route, one for the cast-and-machined route including the tooling — and a check of the crossover on either side of it. That is a routine piece of work rather than a special study.

One further point is worth stating, because it is where die casting programmes most often go wrong: the die cost is recovered over the life of the programme, but the programme length is an estimate. Where a product line is cancelled early, the saving never materialises and the die is a sunk cost that a machined route would not have carried. Where the volume is highly uncertain, the machining route keeps the option open.

What die casting changes about the design

Die casting is not simply “machining, but moulded”. It imposes four design constraints that a machined part does not have, and all four have to be handled before the die is cut.

Draft angle. Every vertical wall needs a slight taper so the casting can be ejected from the die without dragging or tearing. That affects mating faces, wall heights, and how parts sit against each other. It is normally a degree or two per side, and it is a design fact rather than a process preference.

Uniform wall thickness. Molten metal shrinks as it solidifies, and thick sections shrink more and later than thin ones. The result is shrinkage porosity and sink marks where the thickness changes abruptly. Good casting design keeps walls as uniform as possible and transitions between thicknesses gradually. A machined part has no such constraint.

Porosity. Gas and shrinkage porosity are inherent to the process and cannot be entirely eliminated. A casting may be sound on the outside and porous inside, which affects pressure-tightness, fatigue performance and how the part behaves if a critical surface is machined into it. Where a part has to hold pressure or a critical surface has to be machined deep, the porosity matters and should be discussed rather than assumed.

Radii and fillets. Sharp internal corners are not a casting feature; generous radii are what let the metal flow and the part release. A machined part can have a sharp internal corner, up to the limit of the cutter radius — a casting generally cannot.

None of these are defects. They are the terms the process works on, and a part designed for casting from the start is a better casting than a machined design converted afterwards.

The alloy changes, and so does the finish

This is the point that most often gets missed in the comparison, and it is not about cost.

A die-cast part is not made of 6061. Casting alloys are chosen for how they behave as a liquid — how well they fill a die, how they release, how they solidify — and those are different alloys from the wrought grades that are machined from bar and plate. The mechanical properties differ: cast alloys are generally less ductile, may be stronger or weaker depending on the alloy and section, and carry porosity that a wrought part does not have at all.

The consequence is direct: a part designed and proved in machined 6061 does not become the same part in die-cast aluminium. Moving the route means re-checking the mechanical behaviour, not just the price. Where the part is load-bearing, that check is not a formality.

Anodizing behaves differently too. Die-casting alloys contain high levels of silicon and copper, and those elements produce a darker, less uniform anodic layer than the same process on 6061. For a part where the anodized appearance matters, this is a specification risk that should be resolved on a sample before the die is committed — it cannot be corrected afterwards. Where appearance matters and the part has to be cast, the usual answer is a painted or powder-coated finish rather than anodizing. Our anodizing page covers what the process does on the alloys it suits.

Neither point affects machinability. Cast alloys machine well, often better than some wrought grades, and the machining operation on a casting is routine. What changes is the material the part is ultimately made of.

The hybrid route, which is what most programmes do

The framing of “machining or casting” implies a choice between two finished processes. The industrial reality is that most cast parts are also machined, and the interesting decision is not which process to use but how much of the part each one does.

The standard production route for a housing or enclosure looks like this:

1. Die cast the part to near-net shape, with the general form, the ribs, the bosses and the outer profile produced by the casting
2. Machine the features that have to be accurate — bores, bearing seats, sealing faces, mating faces, threaded holes, and any dimension held to a tight tolerance
3. Finish, often by coating or painting rather than anodizing

This is why the comparison is really a question about how much machining the casting removes. A casting that comes out of the die looking like the finished part, needing only a few faces skimmed, is a large saving. A casting that still needs every face machined, every hole drilled and every bore bored has removed very little and paid for a die to do it.

Two practical implications follow:

The machining content is the variable you are optimising. Design the casting so that the features requiring accuracy are few and accessible, and the saving grows. Design it so that the tolerances force machining everywhere, and the die buys you a shape rather than a saving.

The as-cast tolerance has to be understood before the machining is planned. A casting is located in the fixture by its cast surfaces, and the as-cast variation determines how much stock the machining operation has to remove to clean up. Where the cast tolerance is wide, the machining allowance must be wide too, and that affects both the cycle time and the minimum wall thickness the design can support.

For parts that need accuracy on several faces, the machining side of the hybrid route benefits from the same setup-reduction logic that applies to any machined part, which is covered in the comparison of three-axis, four-axis and five-axis machining.

Where each process lives

Die casting suits:

• Housings and enclosures at volume — see enclosures and housings and die casting
• Brackets and mounting components where the shape is complex and the tolerances are not
• Parts whose form removes a great deal of machining
• Components where a smooth as-cast surface is acceptable, possibly with a coating
• Programmes with a known, long production life

CNC machining suits:

• Prototypes and first articles, where the design is still moving
• Low and medium volumes, where a die would not be recovered
• Parts with tight tolerances throughout
• Parts with sharp internal corners, undercuts or no draft
• Any material outside the castable alloys — see materials
• Parts where the lead time matters more than the unit cost

How the two routes are quoted side by side

Where the quantity sits near the crossover, the useful answer is two prices rather than one, and it helps to know what goes into each.

The machined route is quoted from the finished geometry: the stock, the number of setups, the cycle time, the tolerance and the finish. The tooling line is small — fixtures and programs — and it is usually absorbed rather than quoted separately. The price per part falls with quantity as the setup is spread, but it flattens out once that has happened, because the cycle time is per part and does not go away.

The cast-and-machined route has two separate lines. The casting carries the die and the casting operation, and the machining carries only the features that have to be accurate. The die is a one-off and the casting cost per part is low, so the combined price falls steeply with quantity at first and then flattens at a much lower level than the machined route.

The two curves therefore cross, and the useful information for a decision is not a single price but where the crossing is relative to your volume — how far above or below it you sit, and how sensitive the answer is to the volume being wrong. A programme at twice the crossover volume is a comfortable die-casting case. One at the crossover is a genuine judgement call, and the deciding factor is usually how certain the volume forecast is rather than the arithmetic itself.

Two further lines belong in the comparison and are often left out. First, the cost of a design change after the die is cut, which is fast and cheap on the machined route and slow and expensive on the cast route. Second, the lead time to first part, which is short for machining and long for casting because the die has to be made. Where a programme needs parts in weeks rather than months, that alone can settle the route regardless of unit cost.

Where the decision is close, sending both the drawing and the volume forecast is what makes a useful answer possible — and where the programme is at an early stage, the honest recommendation is often to machine the first batch and revisit the die once the volume is real. The quality documents page covers what ships with either route.

When neither route is the right answer

The part is small, simple and needed in enormous volume. For very high volumes of simple parts, other processes — stamping, cold heading, or dedicated forming — can beat both, because they have even lower unit costs. The comparison of cold heading and turning covers where that line sits for fasteners.

The part has to be welded. Casting alloys are generally not welding grades, and a fabricated assembly is built from wrought material.

The part has to be anodized to an appearance standard. The cast alloy does not take the finish the way a wrought grade does, and the answer is usually a different material rather than a different finish.

The part needs internal features that a die cannot form. Undercuts and internal cavities that cannot be drafted out of the die either need a more complex die with slides, or a different process.

The callout that belongs on the drawing

Instead of	Write
Cast or machined	State the quantity — it decides the route more than the geometry does
Die cast part	State which features must be machined after casting
Die cast, tight tolerance	Say which dimensions are as-cast and which are machined
Die cast, anodized	Revisit — the cast alloy does not anodize like 6061; check a sample
Cast, load-bearing	State the load and the porosity requirement — the alloy changed
Cast, design not final	Revisit — a die change is slow and costly

The two lines worth adding to any enquiry where the route is open are the quantity and the list of features that have to be accurate. The first decides whether a die can be justified; the second decides how much the casting actually saves.

Both routes are quoted against the drawing, and where the quantity sits near the crossover the useful answer is two prices rather than one. Send the drawing with the quantity, the material and the features that must be machined, and the quote comes back within 24 hours with DFM feedback. Orders start at 50 pieces per style and material, samples from a single piece in 5–10 days, and every batch ships with a material certificate tracing the heat plus pre-shipment photos.