GeoBusinessIQGeoBusinessIQ

Pouring a shape or cutting one: how a metal part gets its form

A metal component either arrives close to its final shape and gets finished, or starts as a solid block and has everything unnecessary cut away. That single difference propagates into pattern costs, achievable tolerances, the kind of defects you will be hunting, and how much of your purchased metal ends up as swarf. Most production parts eventually use both processes, so the practical question is how much shape should arrive from the mould before a cutter touches it.

Comparison criteria

Criteria are stated explicitly and neither option is declared a winner: which one fits depends on the constraint that binds hardest in your operation.

CriterionCasting: metal poured or injected into a mould cavityMachining: the shape cut from solid stock
How the shape is createdLiquid metal takes the form of a cavity and solidifies, so complex external and internal shape costs little more than simple shape.A cutter removes everything that is not the part, so every feature costs time on a machine and every pocket costs more than a plain face.
Material converted into chipsModest. Runners, risers and machining allowance are the losses, and foundry returns are usually remelted rather than discarded.Substantial on complex parts, and the loss is worst exactly where material is most expensive, such as titanium and high-nickel alloys.
Tooling commitment before the first partA pattern, core boxes or a die must be made, sampled and corrected, and that outlay is only recovered across repeated orders.No dedicated tooling beyond fixtures and standard cutters, so a new part starts as a program and a piece of stock.
How cost behaves as quantity growsDrops steeply once the pattern is paid for, because pouring is fast and the incremental cost is metal, energy and finishing.Declines gently through improved fixturing and optimised cycles, then holds, since each part still needs its own cutting time.
Tolerance and surface straight from the processCoarser and dependent on method, with allowance left on functional faces for a later operation; sand, investment and die routes differ widely.Tight and controllable, limited by the machine, the fixture and thermal effects rather than by shrinkage or mould movement.
Internal integrity of the metalAs-cast structure with the possibility of porosity, shrinkage cavities and inclusions, which is why pressure or fatigue duty brings extra verification.Inherits the condition of wrought stock, with a continuous worked structure and a mill certificate covering what was supplied.
How defects are foundOften invisible from outside, so radiography, dye penetrant or pressure testing is needed to reveal what is inside the section.Mostly dimensional and visible: tool wear, deflection, chatter marks, a mis-set datum, all detectable with measurement.
Absorbing a design changeRequires pattern or die alteration and a fresh sampling cycle, which puts the change in a toolroom queue.A program edit and possibly a new fixture, so changes can be running within days on the same machine.

Choose Casting: metal poured or injected into a mould cavity when

  • The component has substantial internal volume or cored passages that would otherwise become chips
  • Orders repeat often enough for a pattern or die to be recovered across them
  • The alloy is available as a foundry grade with established pouring behaviour
  • The external form would otherwise demand many separate setups on a machine tool

Choose Machining: the shape cut from solid stock when

  • Quantities are low, or the geometry is still being revised between builds
  • Required tolerance or surface condition sits beyond what any as-cast surface offers
  • The duty demands metal free of internal discontinuities, such as a fatigue-critical joint
  • The specified alloy and temper exist as bar, plate or forging but not as a castable grade

Almost every part uses both, and the argument is about the allowance

Very few cast components ship without a cutter touching them. Mounting faces, bores, sealing surfaces and threaded holes are machined because no casting process holds those to the required condition. The genuine engineering decision is therefore how much shape the foundry delivers and how much stock is left for the machinist. Too little allowance and a slight mould shift produces parts that clean up short. Too much and you have paid for metal that becomes swarf plus the machine hours to remove it. That allowance is negotiated between the foundry and the machine shop early, and getting it wrong is a common cause of scrap that only becomes visible after several operations have been paid for.

The two processes fail in completely different places

A machined part goes wrong in ways you can measure: a worn cutter drifts a diameter, a fixture lets a thin wall deflect, a datum gets picked up on the wrong face. Inspection finds those with gauges, and the loss is confined to the parts made since the last check. Casting defects hide inside the section. Porosity, shrinkage and inclusions can survive every visual and dimensional check, then appear when a bore is opened up, when a part is pressure tested, or in service. That difference should drive where inspection money goes: dimensional verification for the cut features, and a volumetric or leak-based method wherever the casting carries load or contains fluid.

Pattern lead time is a programme item, not a purchasing detail

Buyers new to castings often treat the pattern as part of the piece price and are surprised when it dictates the launch date. Pattern or die manufacture, first sampling, dimensional correction and a repeat sample can consume more calendar time than everything else in the introduction, and a correction loop after the first pour is normal rather than a sign of failure. Machined parts have no equivalent, which is why prototype and pre-production quantities are so often cut from solid even when the production intent is casting. Plan the transition explicitly, including who pays for the pattern, where it is stored, who maintains it and what happens to it if the supplier relationship ends.

Frequently asked questions

Can a cast part be finished to the same tolerance as one cut from bar?
On the machined features, yes, because those surfaces are produced the same way in both cases. The difference sits on the as-cast surfaces you leave alone and in how the part behaves during cutting: residual stress released as material comes off can move a casting in ways a homogeneous billet does not. Thin-walled castings in particular can distort after roughing, which is why stress relief and a two-stage machining sequence are common on demanding parts.
How does order quantity change the comparison?
It changes which cost dominates. At low quantities, the pattern or die outweighs everything and cutting from solid usually comes out lower on total cost even with heavy material loss. As quantities repeat, the tooling spreads out and the fast per-part conversion of a foundry route takes over, particularly for shapes that need many machining setups. The crossover is specific to the geometry, so run the arithmetic for the actual part rather than relying on a general rule about quantity.
Why do castings get rejected once machining has already started?
Because opening a surface exposes what was inside. Porosity or a shrink cavity sitting below the as-cast skin becomes visible when a face is faced off or a bore is bored, and the part is scrapped after value has already been added to it. That is why demanding applications specify volumetric inspection at the foundry rather than relying on downstream discovery, and why the commercial agreement should be explicit about who bears the machining cost invested in a casting that turns out to be defective.

Data limitations

  • Manufacturing figures are operator-supplied inputs, not market data. GeoBusinessIQ holds no factory costs, production volumes, yields, cycle times, tooling prices or capacity data and does not estimate them — every result reflects only the figures you enter.
  • No manufacturer, supplier, vendor or factory is recommended, rated or ranked anywhere in this cluster, and no directory of them is published. Selection material describes how to run your own assessment; the assessment itself remains yours.

Explore the graph

Sources

  • National Institute of Standards and Technology NIST (accessed )
    Covers: Measurement science, manufacturing technology research, cybersecurity frameworks, and industrial standards support.
    Does not cover: Certification of products, endorsement of vendors, or costs for any specific implementation.
    Why it matters: A United States federal research institute whose public material covers measurement, manufacturing technology and control-system security.
    Review cadence: annual
  • United Nations Industrial Development Organization UNIDO (accessed )
    Covers: Industrial development analysis, industrial statistics methodology, and manufacturing capability programmes across member states.
    Does not cover: Company-level data, factory costs, supplier information, or real-time production statistics.
    Why it matters: The United Nations agency for industrial development; used for structural framing of how manufacturing sectors develop, never for point figures.
    Review cadence: annual

Educational and operational information only — not legal, engineering, safety, customs, tax, or financial advice. Requirements vary by jurisdiction, product, process, and contract; confirm with the relevant authority or a qualified professional before acting.

Last updated: