Casting, forging or machining from solid: choosing the right process

When a part is defined by its geometry rather than its process, the process choice becomes a cost and performance decision. Getting it wrong in either direction is expensive: too much machining on a part that should be cast, or a casting on a part whose load path demands forged grain flow.

Machining from solid

Best when: volume is low to moderate (under roughly 2,000 pieces per year), the geometry has modest complexity, or the part is needed quickly without tooling investment.

Cost structure: almost entirely machine time and material removal. A part machined from a 5 kg block to a 0.8 kg finished weight means you paid to buy, ship and cut away 4.2 kg of steel. That ratio is the number to watch.

Strength: grain flow is cut across, which is acceptable for most static applications and less so for cyclic loading.

Casting

Best when: the shape is complex, wall sections vary, or the part has internal features that would need multiple machined operations. Volume from a few hundred pieces upward justifies tooling.

Cost structure: tooling amortised over volume plus a low per-piece conversion cost. The break-even against machining typically arrives somewhere between 500 and 2,000 pieces depending on complexity.

Watch out for: porosity. A casting is not uniformly dense, and a casting supplier who will not discuss radiographic or dye-penetrant inspection results on a pressure-retaining part is telling you something.

Forging

Best when: the part carries cyclic or shock load, retains pressure, or is safety-related. Grain flow follows the profile instead of being cut across it, which is why fatigue performance is materially better than either of the other routes.

Cost structure: higher tooling than casting, but the same volume logic applies. Usually finished by machining the functional surfaces only.

A practical decision order

Ask three questions in order. Is the load cyclic or pressure-retaining? If yes, forging is the default unless there is a reason it cannot be. Is the geometry complex enough that machining would remove more than 60% of the starting stock? If yes, look hard at casting. Is annual volume under about 500 pieces? If yes, machine it, accept the material waste, and revisit when volume grows.

The mistake we see most often is a part that has been machined from solid for years because nobody revisited it after volume tripled. A process review every time volume doubles is worth more than most cost-reduction programmes.