5 Axis Impeller Machining for Blades, Hubs and Shrouded Impellers
This page is for design and process engineers who need impeller geometry cut in metal, not prototyped in plastic. It covers which impeller forms suit simultaneous 5 axis machining, how toolpaths are built for blades and fillets, what tolerances and surface finishes are realistic, and where 3-axis or casting is the better call.
What this guide covers
Impeller geometry, toolpath choices, materials, inspection and the limits of 5 axis impeller machiningglcncmachining.
Which impeller types actually need five axes
An impeller is a rotating disc with a hub and a set of blades that adds energy to a fluid. The blade surfaces are ruled in two directions at once: they sweep around the axis and lean along the flow path. That double curvature is what pushes the part into 5 axis territory. A 3-axis machine can reach the blade tips on an open radial impeller, but the hub fillet and the blade root stay out of reach once the blade twists.
Open radial impellers with straight blades are the easy case. Add lean, add a splitter blade, or close the flow path with a shroud and the number of reachable surfaces drops fast. Shrouded and semi-open impellers usually cannot be cut from one block at all unless the shroud is a separate piece, because the tool has to enter through the blade passage.
The practical test is simple. Look at the root fillet radius and the blade height. A fillet radius below 1.5 mm combined with a blade height over 20 mm leaves very little room for a cutter to reach the bottom of the passage without gouging the neighbouring blade. That is where 5 axis impeller machiningglcncmachining earns its cost.
Blisks and integral rotors take this further. Blades and hub are one solid piece, so there is no bolted or welded joint to relieve stress. Every blade has to be cut to the same profile, and the spacing between blades is fixed by the design. One bad toolpath shows up as an unbalanced rotor, not just a scrapped part.
How the job runs from CAD to first article
Start with the CAD model and check that the blade surfaces are clean. Imported surfaces from a CFD mesh often carry tiny gaps and overlaps that a CAM system will turn into a jumpy toolpath. Repair them first, then check wall thickness at the thinnest point of the blade.
Simulation comes next, and it is not optional here. We verify shank clearance, holder clearance and the full machine envelope, including the rotary table. A blade pass that looks fine on screen can still hit the table at the far end of travel. On our simultaneous centers the rotary table is Ø400 mm, which sets a hard limit on how much of the part can swing.
Toolpath strategy depends on the blade. For a long, thin blade we rough with a tapered barrel cutter or a bull nose and leave 0.3–0.5 mm for finishing, then finish with a ball nose in a continuous spiral from root to tip. Cutting in one direction keeps the surface marks consistent. Point milling suits slender blades; flank milling with a tapered tool suits wider, more open blades and removes more material per pass.
First article inspection closes the loop. Blade profile is checked on a CMM with a scanning head, and the root fillet gets extra attention because that is where stress concentrates. Surface finish on the flow path is measured in Ra. If a blade is out of profile, we correct the toolpath and recut before any batch run starts.
Material choice and what it does to the cut
Aluminium is the common starting point for impellers: 6061, 7075 and 2024 all cut cleanly at high spindle speed, and 7075 gives the best strength-to-weight for a compressor wheel that sees real load. The trade-off is that thin aluminium blades chatter if the toolpath leaves too much stock in one pass.
Stainless 17-4PH and 316L show up in pump and process impellers. They work-harden, so the cutter has to stay in the cut instead of rubbing. Titanium TC4 (Ti-6Al-4V) and Inconel are the difficult end of the list. Heat stays in the tool, so we slow the surface speed and use high-pressure coolant. Tool life on Inconel is measured in minutes, not hours.
Plastics and composites are a different problem. PEEK and carbon fibre impellers are usually cut for prototypes or low-volume runs. Carbon fibre dust is abrasive and needs extraction at the cut, and the finish on a machined composite surface is rougher than on metal at the same setting.
Casting is worth comparing before you commit to solid billet. For a high-volume aluminium impeller with simple blade geometry, die casting or vacuum casting can hit the shape at a lower unit cost per part, and machining is then limited to the bore, the hub face and the blade tips. The crossover point depends on blade complexity and how tight the balance spec is.
Impeller machining reference
Working ranges we hold on impeller work. Confirm the exact call on your drawing before quoting.
| Item | Typical range | Notes |
|---|---|---|
| Tolerance, blade profile | ±0.005 mm | Best held on thin blades with light finishing passes |
| Surface finish, flow path | Ra 0.8–1.6 μm | Ra 0.2–0.8 μm where polishing follows |
| Root fillet radius | 1.5 mm and up | Below this, tool reach and gouge risk rise sharply |
| Blade height, open impeller | Up to 150 mm | Depends on passage width and cutter length |
| Max part size | 4,000 mm | Larger parts on the gantry travel machines |
| Rotary table | Ø400 mm | Swings the part; check clearance in simulation |
| Materials | Aluminium, stainless, titanium, Inconel, PEEK | 17-4PH and TC4 need slower speeds |
| Batch range | 1 prototype to 10,000+ parts | No minimum order quantity |
| Inspection | 100% before shipment | CMM scan plus reports on request |
Where five axes is the wrong answer
Not every impeller should be cut on a 5-axis mill. Open impellers with straight or mildly curved blades and a generous root fillet can be turned and milled on 3-axis machines with a fixture that indexes the part. Setup time goes up, but the machine hour rate is lower.
Casting plus finish machining is often the better route once volumes pass a few thousand parts and the blade geometry is not extreme. The cast surface near the root is not as good as a machined one, so the design has to accept that or the root has to be finished separately.
Very small impellers bring their own problem. Below roughly 30 mm in diameter, tool diameter becomes the limit, not machine axes. A cutter small enough to fit the passage is too slender to cut at a useful rate, and the blade may deflect under cutting force. For those sizes, consider wire EDM for the blade profiles or a casting-and-finish route.
Highly complex shrouded impellers with narrow passages sometimes cannot be machined from one billet at all. The honest answer is to split the part into a hub and a shroud, machine both, and join them, or to change the geometry so the tool can reach. We will say so during DFM rather than quote a job we cannot hold tolerance on.
Impeller machining questions engineers ask
Can you cut a shrouded impeller from one solid billet?
Often no, and it depends on passage width and blade count. The cutter has to enter through the blade passage and still clear the shroud on both sides.
When the passage is too narrow for the smallest rigid cutter, we split the design into a hub and a shroud, machine both, then join them. We flag this during DFM, before quoting.
What file formats do you need for an impeller quote?
STEP or IGES for the solid, plus a 2D drawing with the critical tolerances, the balance requirement and the material grade. If the blade surfaces came from a CFD mesh, send the mesh too so we can compare.
Files upload securely and stay confidential. An NDA is available on request if your program needs one.
How do you balance an impeller after machining?
Machining alone does not guarantee balance. Blade-to-blade mass variation and hub runout both affect it. We control blade profile on the CMM and keep hub concentricity tight, then hand the part to your balancing step.
If you need balancing as part of the scope, say so at quote stage so it is written into the process plan rather than added later.
Which toolpath strategy suits a long, slender blade?
Point milling with a ball nose tool in a continuous spiral, cutting from root to tip in one direction. It keeps the chip load steady and avoids the marks that come from reversing direction mid-blade.
Flank milling with a tapered cutter removes material faster on wider blades, but it needs a blade surface that is close to ruled. On a strongly twisted blade, flank milling leaves mismatch at the edges.
What lead time should I plan for an impeller prototype?
Quotation and a free DFM analysis come back within 12 hours. Production can start within 24 hours of approval, and parts ship in 3–5 days.
Impeller work with several setups and CMM scanning may need longer than a simple turned part. We give the realistic date with the quote, not an optimistic one.
Do you machine impellers in titanium and Inconel?
Yes. TC4 (Ti-6Al-4V) and Inconel are both in the material list, along with 17-4PH stainless and 7075 aluminium.
Both titanium and Inconel hold heat in the cutting zone, so we reduce surface speed, use high-pressure coolant and plan for more tool changes. That shows up in the price and the schedule.
Send the impeller model and get a process answer
We review the blade geometry, the root fillet and the material, then tell you what can be cut, what needs splitting, and what it will take.
12-hour quote and free DFM100% inspection before shipmentNDA on request
