Machining Castings vs Machining from Bar: Tolerance and Cost Reality
A casting saves money only when its volume is high enough to amortize the tooling and its shape removes enough machining to pay for the porosity risk; below that crossover, machining from bar stock is cheaper, faster, and dimensionally safer. The rule of thumb: cast above a few thousand pieces per year for a near-net shape, machine from bar below it.
Every machined metal part starts as either wrought stock — bar, plate, or tube that was rolled or forged — or as a casting that was poured into a mold. The choice between them shapes your tolerance, your lead time, your porosity risk, and your unit cost. Buyers often assume casting is automatically cheaper because it is "near net shape." Sometimes it is. Often, by the time you add tooling amortization, machining allowances, and porosity scrap, bar stock was the better deal. This guide lays out the real trade with numbers.
What Casting Actually Buys You
The argument for casting is material and machining removal. A part that is 80% air — a pump housing, a valve body, a bracket with heavy ribs — wastes most of its bar stock as chips when machined from solid. Casting that shape near-net means the foundry makes the waste, not your machining hours, and the CNC work is reduced to the functional faces, bores, and threads. That logic is strongest for complex hollow geometries in aluminum, iron, and zinc, and for parts where the as-cast surface is acceptable on non-critical areas.
| Route | Tooling cost | Typical minimum sensible volume | Lead time to first parts |
|---|---|---|---|
| Machined from bar | None | 1 piece | Days |
| Sand casting + machine | Low–moderate | ~50–500 pieces | Weeks |
| Investment casting + machine | Moderate | ~500–2,000 pieces | Weeks |
| Die casting + machine | High (die is expensive) | ~2,000–10,000+ pieces | Weeks–months |
| Forging + machine | High | Volume-dependent | Weeks |
The crossover is not only about volume — it is about how much of the part is actually machined away. A near-net casting that still needs every face machined saves little. A casting that leaves only six critical features to cut saves a lot. When you compare quotes, compare machined-to-print cost both ways, not raw material weight.
Tolerance Reality: As-Cast vs As-Machined
Here is where most casting decisions go wrong: the as-cast tolerance is loose, and no CNC machine fixes a casting that is already outside its machining allowance. As-cast dimensional capability varies wildly by process — sand casting holds only rough linear tolerances, die casting holds much better but still looser than machining, and both can shift part to part with die wear, mold shift, and shrinkage variation. Critical features must be machined, which means the casting needs enough stock on those features, located from datums the machinist can trust.
| Process | Typical as-cast linear tolerance | Typical surface finish | Machining stock on critical faces |
|---|---|---|---|
| Sand casting (aluminum/iron) | ±0.5 to ±1.5 mm | Rough, 12.5–25 µm Ra | 1.5–3 mm per face |
| Investment casting | ±0.1 to ±0.3 mm typical | Good, 3–6 µm Ra | 0.8–1.5 mm per face |
| Die casting (aluminum/zinc) | ±0.1 to ±0.3 mm | Good, 2–5 µm Ra | 0.5–1 mm typical, thin walls need care |
| Machined from bar | Machined tolerance, ±0.005–0.05 mm | Excellent | None — part is the stock |
Read the last column carefully: casting does not remove the need for tight machining; it just decides how much material the machinist has to work with. And a casting with too little stock on a feature that shifted is scrap before the CNC program even runs. That is why casting suppliers must control the same datums the machinist will use, and why first-article machining of castings always includes a stock check.
Porosity: The Risk That Machining Reveals
Castings contain porosity — gas holes and shrinkage voids — and machining is exactly what exposes it. Cut into a die casting and you can open pores that were hiding under the surface skin; the part that looked perfect in the foundry shows pinholes on a machined sealing face. Pressure-tight parts (valve bodies, pump housings, manifolds) are the classic victims: a porous casting that leaks is scrap after all the machining cost is already spent. Bar stock and forgings, being wrought, do not have this failure mode — their internal structure is dense and directionally worked.
| Application risk | Casting concern | Bar stock concern |
|---|---|---|
| Pressure-tight bore or face | Porosity opens during machining, leaks | None — dense structure |
| Thin machined wall | Pores break through, wall leaks | Fine if stock is adequate |
| Fatigue-loaded feature | Porosity and grain structure limit life | Wrought grain is directional, usually better |
| Cosmetic machined surface | Pin holes visible after finish | Clean, consistent |
The mitigation is process control, not avoidance: foundries reduce porosity with gating design, mold fill control, and sometimes vacuum or squeeze assist; buyers reduce risk by specifying density or x-ray requirements only where the application genuinely needs them, and by checking the first machined batch for exposed porosity before committing to volume. Die casters also use the term "machining stock vs. casting skin" deliberately — the dense skin of a die casting is an asset, and machining it all off can expose the more porous interior. Our guide to finishing and post-machining steps covers the same logic one operation later: each subsequent step reveals what the previous one hid.
When Bar Stock Is Simply Better
Below the volume crossover, and for any part where the geometry does not waste material, machining from bar is the stronger choice on every axis that matters. Lead time is days instead of weeks because there is no tooling and no foundry queue. Dimensional behavior is predictable because wrought stock has consistent structure and no mold shift. Tolerance is a machining question, not a casting question. And changes to the design cost nothing — edit the program, not the die. Prototypes and low-volume production of almost anything should start from bar unless the shape is so wasteful that material cost alone justifies a casting.
Bar stock also wins on mechanical properties in many cases. Wrought material is worked and has directional grain; cast material is isotropic but coarser-grained and weaker in the same alloy. A 6061-T6 machined from bar and a 6061-T6 machined from a casting are not the same metal — the bar version is typically stronger and more consistent. When the design is strength- or fatigue-critical, wrought stock is the safer default, and a casting is chosen only when the shape economics are overwhelming.
The Decision Framework That Works
Run the numbers both ways on any part over roughly 500 pieces a year. Estimate machined-from-bar cost from the envelope size and cycle time. Estimate casting cost from tooling amortization, casting price, machining stock, and a porosity scrap allowance — and be honest about the scrap allowance, because porous parts fail at the end of the line where they cost the most. Our CNC machining cost guide gives the per-part cost logic that feeds both estimates.
Then apply three filters. Volume: below about 1,000–2,000 pieces per year, bar almost always wins unless the shape wastes extreme material. Geometry: if the part is a block with holes, bar; if it is a hollow housing or ribbed bracket where casting removes 60%+ of the machining, casting deserves a real quote. Risk: if the part is pressure-tight, fatigue-critical, or cosmetically machined on large surfaces, weight the porosity and structure risk against the saving.
What a Good Machined-Casting Quote Looks Like
When you do go the casting route, the quote should separate the stages: casting supplier, machining, and inspection. Ask who owns the datum scheme between foundry and machine shop — the same feature should locate the casting in both. Ask what stock the casting carries on each machined face and how the foundry guarantees it. Ask how porosity is handled: inspection method, acceptance criteria, and what happens to a batch that exposes pores mid-run. And ask for the first machined articles to be checked for stock and porosity before full production, not after.
A floor that machines castings routinely — locating cast surfaces, checking stock, and confirming datum transfer on CNC mills and turning centers — will answer these questions without hesitation. We machine both routes at BQUQ: bar-stock parts for speed and stability, and customer-supplied or locally sourced castings where the volume justifies them, with the stock and porosity checks built into the precision components workflow. Send the drawing with the volume and the application, and the recommendation will come back as a route, not a slogan.
Email sc@bquq.com or WhatsApp +86 137 1315 7787 with your PDF/DXF/STEP file. An engineer reviews it and replies with price, lead time and DFM notes on working days.
Frequently Asked Questions
Q: Is it cheaper to machine from bar stock or from a casting?
A: Below roughly 1,000–2,000 pieces per year, bar stock is usually cheaper because there is no tooling cost and lead time is days. Above that, a near-net casting can win if its shape removes significant machining — but only after tooling amortization and porosity scrap are included.
Q: What tolerances can you hold machining a casting?
A: As-cast tolerances are loose — sand castings typically ±0.5 to ±1.5 mm, die and investment castings ±0.1 to ±0.3 mm. Machined features then hold normal CNC tolerances, down to ±0.005 mm, provided the casting carries enough stock and the datums transfer cleanly.
Q: Why do machined castings sometimes leak?
A: Porosity. Castings contain gas and shrinkage voids that machining exposes when it cuts through the dense surface skin into the porous interior. Pressure-tight parts need porosity control, inspection, and sometimes impregnation — risks that bar stock does not carry.
Q: When should I avoid castings entirely?
A: For pressure-tight, fatigue-critical, or cosmetically machined parts, and for any volume below the tooling crossover. Wrought bar stock is denser, stronger, and more consistent; casting is only worth its risks when the shape economics clearly beat machining from solid.
Q: How much machining stock should a casting carry?
A: Typically 0.5–1.5 mm per machined face for die and investment castings and 1.5–3 mm for sand castings, depending on feature size and as-cast tolerance. Too little stock and a shifted casting scrap at the machine; too much and the casting's cost advantage evaporates.
Authored by the BQUQ Engineering Team. BQUQ is an ISO9001-certified source factory in Dongguan, China, running CNC machining, metal stamping, custom springs, heat sink and collet lines under one roof. Send drawings to sc@bquq.com or WhatsApp +86 13713157787 for a quote within 12 working hours. www.bquq.com
