Casting or CNC from billet — where is the break-even?

This is the most common conversion we get asked about, and the most common source of wrong decisions in both directions.

The starting point is two cost structures that look entirely different:

Machining from billet: no tooling cost. High cost per part, which barely falls with volume. You pay in machine time, and the machine time is the same for part number 5,000 as for part number 1.

Casting: high tooling cost up front. Low cost per part. The cycle is seconds, not minutes.

The two lines cross at one specific point. The whole discussion is about where.

The arithmetic

Call the tooling cost T, the cost per machined part C₁, and the price per casting C₂.

N = T ÷ (C₁ − C₂)

N is the number of parts before the tool has paid for itself. Below N, machining is cheapest. Above N, casting is cheapest, and the gap widens with every part.

So far, simple. It is the three things people forget to include that decide whether the answer comes out right.

Forgotten thing 1: material waste

A machined part starts as a block. Everything that is not part becomes swarf.

On a typical bracket with pockets, ribs and through holes, swarf can account for a large share of the starting material. You pay for the whole block and sell the swarf for a fraction.

A casting starts as roughly the shape it will end up as. Runners and overflows go back into the furnace and are remelted — not free, but considerably better than selling swarf.

When the metal price rises — LME aluminium has risen sharply over the past year — this moves the break-even point downwards. Material waste costs more in a high-price market, and casting wins earlier than it used to.

Forgotten thing 2: the whole life of the series

Many people reject casting because they calculate on the first order.

If the product will live for ten years, ten years of volume belongs in the calculation. A part you make 3,000 of per year is a part you make 30,000 of — and the difference between those two figures is often the difference between no and yes.

Include spare parts as well. A part that has to be available for years after the main series has finished is a part where the tool has to remain in place anyway.

Forgotten thing 3: the geometry changes

This is the most important, and the one least often included.

When you cast, you are not drawing the same part. You are drawing a better part.

A machined bracket is limited by what a cutter can reach. Every pocket needs a floor radius matching the cutter. Internal cavities are difficult or impossible. Ribs cost machine time, so the designer only adds them where they are strictly necessary.

A cast bracket can have a thin shell with ribs where the bending moment is, closed volumes, cored holes, and metal concentrated exactly where the stress is high.

The result is often a part that is both lighter and stronger than the machined one. That is not a cost saving — it is a product improvement, and it belongs in the assessment.

Five cases where machining wins at any volume

We recommend machining when:

1. The part will never be made again. If this is a one-off requirement, the tooling cost is wasted no matter how many parts are in that single run.

2. The geometry is too simple. A round disc with a hole in the middle has nothing to gain from casting. Turning is both faster and cheaper.

3. The tolerance requirement is tighter than casting can give, across the whole part. If almost every dimension is critical, almost everything has to be machined anyway — and you have then paid for a tool without avoiding the machining.

4. The material has to be something we do not cast. Stainless steel, titanium, copper alloys. We cast aluminium and zinc.

5. The design is not finished. If the geometry will change three more times, a tool is premature. Machine the prototypes, lock the design, then cast.

The middle route people overlook

It is not either or.

The most common and most profitable solution for a part with a few critical dimensions is to cast the blank and machine only the faces that require it.

The part is cast with machining allowance on the critical faces and finished geometry everywhere else. Then only what has to be tighter than casting delivers is milled or turned.

You get the low cost per part of casting on 90 per cent of the geometry, and the precision of machining on the 10 per cent that matters.

This is what we do most of. One of our product examples is a zinc die casting that is threaded and made assembly-ready here — casting where casting is best, machining where machining is best, in the same building.

The other five methods

Casting and machining from billet are not the only options, and the calculation above is not complete until you have weighed the rest. Five other methods compete for the same part, and each wins in its own corner.

Injection moulding in plastic is the strongest competitor when the part does not have to conduct heat, does not have to conduct current and does not carry a load for years. The tool is cheaper, the cycle shorter and the part lighter. If it has to take temperature, UV over time or a thread tightened repeatedly, the picture reverses.

MIM — metal injection moulding is metal's answer to injection moulding, and it wins on small, complicated parts in high volume — the kind that fit in a palm. Above that size the raw material cost dominates and casting takes over.

Metal 3D printing has no tooling cost at all, which makes it right for prototypes and for runs of a few dozen parts. Cost per part barely falls with volume, so casting overtakes it early.

Laser cutting and press braking make the part from sheet. That is fast and tool-free, but it starts from something flat: every bend and every weld is an operation and a link in a tolerance chain. Brackets with two or three planes are often made from sheet until the number of operations makes casting cheaper.

Stamping is the series form of sheet work, and it beats everything else on price when the part is thin, flat and made in hundreds of thousands. It cannot vary wall thickness, and it cannot make a housing.

The rule that ties them together: the more geometry the part has in three directions, the earlier casting wins.

What to send us

To give you a concrete break-even point we need:

• The drawing or the 3D model. STEP is best.
• What you pay per part today, if the part is already being made.
• Annual volume and expected product life.
• Which dimensions are actually critical — not all of them, only those that determine the function.

We then calculate N together with you, and we say so honestly if the answer is that you should carry on machining. That happens regularly, and it is a better conversation than selling you a tool you will never pay off.