Crankshaft CNC Machining
How a crankshaft goes from a billet or forging to a balanced, ground shaft. This page covers journal geometry, material behavior, tolerance budgets and the point where crankshaft CNC machining stops being the right process.
What crankshaft CNC machining has to hold
A crankshaft turns reciprocating piston motion into rotation, so every dimension that matters is referenced to the main journal centerline. The crank pin centerline sits at the stroke radius from it, and that offset distance is the single most important number on the drawing. Get the offset wrong by 0.02 mm and the compression ratio of that cylinder moves.
The load path runs through the fillet where the pin meets the web. That radius is not a cosmetic blend. It controls how stress spreads into the web, and it is also where a tool has the least room to reach. A tight fillet plus a deep web means a tool holder that may not fit, which decides the process before any tolerance is discussed.
Counterweights do the opposite job: they cancel the rotating mass of the pin and web so the shaft does not shake the block apart at speed. That is why a crankshaft is never a turned part only. Roughing removes the bulk, then the pin centerlines are established, then the journals are finished to size and roundness.
Roundness and cylindricity usually matter more than the diameter callout. A journal 0.01 mm under nominal will still run if it is round; a journal on nominal but 0.015 mm out of round will wipe a bearing shell. Measure both, and measure them at temperature, because a shaft pulled hot from a grinder reads small.
Billet, forging or casting: what the stock decides
Material drives the whole plan. Ductile cast iron and nodular iron cast near net shape, so machining is mostly journal cleanup and hole work. Forged 4340 or 4140 arrives with a grain flow that follows the webs, which is why high-load and racing shafts are forged rather than cut from plate. Billet 4340 gives freedom in web shape and counterweight profile, but you pay in stock removal and in the risk of exposing inclusions.
Alloy steel cuts cleanly at 28–32 HRC with coated carbide and air-blast or high-pressure coolant. Above 38 HRC the tool life drops fast, and the journal finish starts to depend on the last pass rather than the tool grade. Nitrided or induction-hardened shafts are normally ground after heat treatment, not milled to final size.
Aluminum crankshafts exist for small two-stroke and model engines, and 7075 or 2024 works there because the loads are low and weight dominates. It is the wrong choice for a four-stroke production engine. Titanium and Inconel show up in special projects; both need slower speeds, sharper edges and more patience on the finishing passes.
Stock condition also sets your datum strategy. A casting with draft and parting-line flash cannot be located off the raw surface. We find the main journal centerline first, then everything else is measured from it. Skipping that step is the most common reason a first article comes back out of position.
How the operations are sequenced
Sequence matters more than any single operation. The usual route is: saw or turn the stock to length, face and center both ends, turn the main journals as a straight shaft, then offset the work to turn each pin. On a mill-turn or five-axis center the pin and web can be cut in one setup with a Ø400 mm rotary table indexing the crank angle.
Five-axis positioning removes the re-fixturing error that builds up when a pin is turned in a second setup. That error is small on a 200 mm shaft and large on a 1,500 mm one. We hold ±0.005 mm on features we can reach in a single setup; the same feature split across two fixtures is a different conversation.
Drilling the oil passages comes after the journals are established, because the hole exits must land on the fillet where the bearing feeds. Cross-drilling from the main to the pin at the wrong angle starves the rod bearing. Deburr both ends of every passage and verify with a bore scope or a flow check.
Grinding is the last metal-cutting step, not a finishing extra. Journal diameter, roundness and surface finish are set by grinding after heat treatment and after any straightening. Polishing follows grinding only to lower Ra, not to fix size. A polished journal that is still out of round will fail just as fast.
Tolerance, finish and balance in one budget
Journal diameter, roundness, taper, fillet radius and surface finish all compete for the same tolerance stack. Typical engine work sits at ±0.005 mm on journal diameter with roundness inside 0.005 mm and taper under 0.008 mm over the journal length. Surface finish lands at Ra 0.2–0.8 μm on bearing journals; Ra 0.8–1.6 μm is acceptable on non-bearing surfaces.
Ra is not a target on its own. A journal at Ra 0.2 μm with a torn or smeared surface will still damage a bearing, because the peaks point the wrong way. We look at the bearing contact pattern after assembly, not just the profilometer trace. Crosshatch or circumferential lay has to match the drawing note.
Balance is a separate measurement, taken after grinding. Dynamic balance at the running speed of the engine, not at shop rpm, is what removes the vibration that wears main bearings. Correction is done by drilling the counterweights, so leave stock there if the drawing does not already allow for it.
Finally, straightness. A long shaft can pass every local check and still be bent 0.05 mm over its length. Check it between centers after grinding, and again after any handling or shipping. Straightening after final grind changes the residual stress and can move the journals later.
Which process fits which crankshaft
Use this to pick a route before quoting.
| Shaft type | Best route | Why |
|---|---|---|
| One-off prototype, simple pin layout | 3-axis mill + turn | Cheapest setup, easy to re-cut |
| Prototype with offset pins, tight fillet | 5-axis or mill-turn | One setup holds pin-to-main position |
| Production forged steel shaft | Turn, mill, heat treat, grind | Grain flow kept, journals sized after HT |
| Cast iron small engine shaft | Cast near net + mill journals | Little stock removal, low tool wear |
| Hardened shaft above 45 HRC | Grind after heat treatment | Milling cannot hold size at that hardness |
| Long shaft over 1,500 mm | Turn + grind between centers | Fixture error compounds on long spans |
| Aluminum model engine shaft | 3-axis mill, high speed | Low load, weight is the constraint |
The trade-off in one line
If the pin layout is simple and the volume is one or two parts, turn and mill it on a 3-axis machine and accept a second setup. If pin-to-main position, fillet radius or balance is critical, run it on a five-axis or mill-turn center in one setup and grind the journals afterward.
Questions we get on crankshaft work
Can a crankshaft be fully machined without grinding?
For low-speed or model engine shafts, yes. Milling and turning can hold ±0.005 mm and reach Ra 0.8–1.6 μm if the material is not hardened.
For any shaft that runs at production engine speed or carries a bearing shell under load, the journals are ground. Grinding is what delivers roundness and the surface lay a bearing needs.
How do you hold the pin centerline position?
We establish the main journal centerline first, then index the crank angle with a rotary table or a five-axis move. The pin offset is measured from that reference, not from the raw stock surface.
On a second-setup part the error adds up. That is why a pin that is 0.01 mm out on a short prototype can be 0.05 mm out on a long shaft.
What surface finish should a rod journal have?
Ra 0.2–0.8 μm is the usual band for a loaded bearing journal. Below that, the risk is a smeared or torn surface that looks good on the gauge and fails in service.
Check the lay direction against the drawing. A circumferential finish and a crosshatch finish behave differently in the oil film.
Do you balance the shaft before shipping?
Dynamic balance is measured after grinding, at the speed the engine will run. Correction is done by removing material from the counterweights.
If the drawing does not leave correction stock on the counterweights, tell us before machining so we can plan for it.
Which materials do you machine most often for crankshafts?
Forged 4340 and 4140, ductile and nodular cast iron, and 1045 for lower-load shafts. 7075 and 2024 appear in small engine work.
We also run 17-4PH and titanium on special projects, with slower cutting data and more finishing passes.
Can you work from a sample instead of a drawing?
Yes. We can reverse-engineer the journal sizes, pin offset and fillet radii from a physical shaft and send the model back for confirmation before cutting.
That route is common for restoration and discontinued engine parts where no drawing exists.
Send the drawing, get a route and a price
Upload the shaft model and we come back with a process route, tolerance check and quotation. DFM analysis is included.
Quote in 12 hours100% inspectionNo MOQNDA on request
More from GreatLight
We publish setup notes, tooling trials and inspection data from the factory floor.
