KEY TAKEAWAYS

✓
CNC machining wins on tight tolerances, complex internal geometry, and low-to-mid volumes where tooling cost would make casting prohibitive.

✓
Casting porosity is a real failure mode — for pressure-critical or fatigue-loaded parts, machined billet is the safer default.

✓
Lead time heavily favors machining: no tooling, no mold qualification, no first-article casting approval cycle.

✓
Material choice is broader in machining — you are not locked into alloys that cast well; you can run the alloy the application actually demands.

✓
Hybrid approaches exist — rough castings finish-machined to tolerance — but only make sense at higher volumes with stable designs.

Engineers default to casting for a simple reason: unit cost looks lower on a spreadsheet. But that math only holds at volume, with a frozen design, and when the part geometry and material properties that casting imposes are actually acceptable. In aerospace, defense, medical, and high-performance industrial work, those conditions frequently don’t all hold at once.

The real cost of a wrong process decision shows up later — in secondary machining to clean up draft angles, in rejected lots traced to subsurface porosity, in six-week mold lead times that stall a program. Understanding where machining genuinely outperforms casting isn’t academic. It directly affects part performance, program risk, and total cost of ownership.

This article is not an argument that casting is bad. Sand casting, die casting, and investment casting are excellent processes when applied correctly. The goal here is to give engineers a rigorous framework for identifying the cases where machining is the right answer from the start — not a fallback when casting fails.

Casting tolerances are process-limited. Investment casting, the most precise casting method, typically holds tolerances in the range of plus or minus 0.010 inch on small features — and that degrades with part size and geometry complexity. Die casting can do better in some dimensions but introduces its own variation from die wear, thermal cycling, and ejection forces. Sand casting is rougher still, commonly plus or minus 0.030 inch or worse.

CNC machining operates in an entirely different regime. A competent shop running proper fixturing holds plus or minus 0.001 inch routinely, and critical features can be held to 0.0005 inch or tighter with process control. Surface finish from a milling or turning operation is typically 32 to 125 Ra microinch as-machined, with grinding and honing bringing critical surfaces into single-digit Ra territory. No casting process approaches that without secondary machining.

The practical consequence is straightforward: any part with functional mating surfaces, sealing faces, bearing journals, or precision bores will require machining regardless of how it was formed. If machining is mandatory anyway, the question becomes whether the casting step is adding value or just adding cost and lead time. For low-to-mid volumes, it usually isn’t.

Rule of thumb: if more than 40% of a casting’s surfaces require post-machining to reach functional tolerances, you should be running the part from billet and skipping the casting entirely.

Casting is fundamentally a fill-and-solidify process. That means every internal feature has to be producible with a core, and every external surface needs draft for ejection or pattern removal. Complex internal passages, undercuts, thin walls with inconsistent section thickness, and features that intersect in three-dimensional space are all difficult or impossible to cast reliably.

CNC machining — particularly 4-axis and 5-axis simultaneous machining — accesses geometry that casting cannot. Deep narrow slots, cross-drilled intersecting passages, precisely located hole patterns on compound angles, and pockets with tight corner radii are standard CNC work. EDM extends this further into geometries that even rotating cutters cannot reach. The constraint in machining is cutter access and fixturing, not the geometry of a mold cavity.

This matters most in fluid-handling components, manifolds, housings with integrated cooling or lubrication passages, and structural parts with lightweighting pockets. A hydraulic manifold with intersecting cross-drilled passages and O-ring face seal ports is a machining job — full stop. Attempting to cast that geometry introduces core shift risk, porosity at junctions, and almost certainly requires finish machining of every port face anyway. The machined-from-billet approach produces a dimensionally consistent, fully inspectable part with no hidden internal defects.

Porosity is the silent failure mode in casting. Shrinkage porosity, gas porosity, and cold shuts are inherent risks in any solidification process. Investment casting and die casting have improved dramatically with process controls, but neither eliminates the risk entirely. In safety-critical applications — pressure vessels, structural flight hardware, rotating machinery — subsurface porosity that passes visual inspection can initiate fatigue cracks under cyclic loading.

Machined billet stock is wrought material. It has been worked — rolled, forged, or extruded — which refines grain structure, closes voids, and produces consistent mechanical properties throughout the cross-section. ASTM and AMS material certifications for billet give engineers a documented, traceable material baseline. Castings have their own material specs, but the localized properties within a given casting can vary in ways that billet simply does not.

For AS9100-controlled programs in aerospace and defense, the traceability requirements alone often tip the decision toward machined billet. Every piece of stock has a heat number, a cert, and a documented property range. Nimble’s certified partner network sources material with full traceability and maintains documentation through CMM inspection records — which matters when a customer’s receiving inspection or a DCMA audit is asking for objective evidence. The material integrity advantage of machined billet is not marginal in high-reliability applications — it is fundamental.

Warning: Impregnation processes can seal casting porosity for pressure applications, but they do not restore mechanical properties or eliminate the fatigue crack initiation risk in dynamically loaded parts. Do not treat impregnation as a blanket fix for structural porosity.

The standard argument for casting is unit cost at volume. That argument is valid — but the volume threshold where casting wins is higher than most engineers assume, and it shifts significantly based on part complexity and design stability. Die casting tooling for an aluminum part runs $15,000 to $80,000 or more depending on complexity. Investment casting tooling is lower, typically $3,000 to $20,000, but the per-piece cost is higher. Sand casting has minimal tooling cost but rough output that requires heavy post-machining.

For machined aluminum parts, a credible rule of thumb is that casting becomes cost-competitive somewhere between 500 and 2,000 pieces per year — but only if the design is frozen. Every engineering change after tooling is cut costs money and time. Machining from billet has essentially zero tooling cost: update the CAM program and you are running the new revision. For programs still in development, machining is almost always the right call economically even at volumes that seem to favor casting on a unit-cost basis.

Total cost of ownership includes tooling amortization, first-article qualification, change order costs, scrap rates, and lead time. A casting that costs $18 per piece versus a machined part at $45 per piece looks obvious — until you factor in $40,000 in tooling, an eight-week first-article cycle, and three engineering changes in the first production year. Run the full model before committing to a process.

Casting lead times are driven by tooling and qualification, not production. A new die casting tool takes six to fourteen weeks to fabricate, validate, and qualify for production. Investment casting patterns take four to ten weeks. Once tooling is ready, first-article inspection and approval add more time. For programs with schedule pressure — which describes most of them — this is a serious constraint.

CNC machining from stock requires no tooling. Raw material for common alloys like 6061 aluminum, 303 stainless, and 4140 steel is available from domestic distributors in days. A competent shop can turn around first articles on machined parts in one to three weeks for most geometries. Nimble’s certified partner network provides 24-hour quoting so engineers know quickly whether a machined approach is viable before committing to a process path.

This lead time gap is especially critical in three scenarios: prototype and development builds where design iteration is expected, bridge production while casting tooling is being made, and low-rate initial production (LRIP) for programs that have not yet established their production volume. In each case, machining gets hardware into test or into customers’ hands faster, with lower financial commitment and maximum flexibility for design changes.

Casting imposes material constraints that engineers often accept without questioning them. Die casting works well with aluminum alloys like A380 and A383, zinc alloys, and some magnesium alloys — these alloys have good fluidity and solidification behavior. Investment casting opens up stainless steels, tool steels, and some superalloys. But the alloy the process prefers and the alloy the application demands are not always the same thing.

Machining eliminates this constraint entirely. Need 7075-T651 aluminum for strength-to-weight ratio? It machines beautifully but casts poorly. Need 17-4 PH stainless in H900 condition for a combination of corrosion resistance and high strength? That is a standard machining material — not a practical casting alloy. Need titanium 6Al-4V for a weight-critical aerospace bracket? Machined from billet or forging, with full AMS 4928 or equivalent certification, is the standard approach.

The application should drive the alloy selection; the alloy should not be compromised to fit a process preference. When the ideal material for a part’s functional requirements is one that does not cast well, machining is not just preferable — it may be the only credible option. Engineers working through a free DFM review can surface these material-process compatibility issues before they become production problems.

Choosing machining over casting should be a deliberate, documented decision based on functional requirements — not a default. The following factors, weighted together, determine when machining wins.

• Tolerance requirements below plus or minus 0.010 inch on functional features: machining wins automatically — casting cannot hold it without post-machining anyway.
• Volume below 500 pieces per year with any design uncertainty: tooling investment for casting is rarely justified.
• Complex internal geometry or undercuts that casting cannot produce without significant core complexity and scrap risk.
• Material requirements that specify a wrought alloy not available in a reliable casting form.
• Structural or pressure-critical applications where porosity is an unacceptable failure risk and material traceability is mandatory.
• Schedule requirements that cannot absorb a six-to-fourteen week tooling lead time.

If three or more of these conditions apply to a given part, machining from billet is almost certainly the right answer. If only one applies — particularly if it is the volume threshold and volumes are clearly above 2,000 per year with a stable design — casting deserves serious evaluation.

The hybrid approach — rough casting finish-machined to tolerance — is valid but carries both processes’ risks and costs. It makes the most sense at higher volumes with very stable designs where the casting provides significant material savings on a large, complex part. For most development programs and many production programs in aerospace and defense, machined billet is the lower-risk, more flexible, and ultimately more cost-effective choice when the full program cost is evaluated honestly.

Key insight: ‘Design for casting’ and ‘design for machining’ are fundamentally different activities. Draft angles, parting lines, and wall thickness rules for casting will constrain your geometry in ways that may compromise function. If you are machining the part, design it for machining — you will get better parts.

READY TO SOURCE?

Table of Contents

Related Services

Get a Quote in 24h

Upload your drawing and get a detailed quote with free DFM review.