Kashifuji Gear Cutting: 7 Ways to Slash Production Costs

Most gear cutting quotes are built from cycle time. That number is easy to measure and easy to argue about. It is also rarely the biggest line item. On a typical spiral bevel or hypoid gear, the hours spent on fixturing, indicating the blank, and proving the first article often exceed the hours of actual cutting. Every extra setup adds its own stack of error. Runout that was 0.01 mm in the chuck can become 0.03 mm after a second op, and that error lands directly in the tooth contact pattern.

This is where a simultaneous five-axis machine earns its place. With a Ø400 mm rotary table and a trunnion, the tooth form and the reference bore can be cut in one setup. The blank stays clamped, so the relationship between the bore, the face, and the tooth flank is fixed by the machine, not by the operator. On gearbox housings and drivetrain parts we see setup count drop from three to one, and the first-article scrap that goes with each re-clamp disappears with it.

There is a limit. Five-axis is not free money. If your part is a simple spur gear with one bore and one keyway, a three-axis machine with a dedicated fixture will beat it on cost and on cycle time. Single-setup five-axis pays back when the part has intersecting features, an angled tooth form, or a tight bore-to-flank relationship that is hard to hold across two ops. Judge it by the number of datums, not by the machine brochure.

Tool life on gear cutting tools is driven by chip load per tooth, not by surface speed alone. When a hob or a cutter starts to break down, the reflex is to slow the spindle. That usually makes things worse. Lower speed with the same feed per tooth raises the specific cutting force and rubs the edge instead of cutting. A better move is to keep the chipload and reduce the depth, or split the roughing and finishing passes so the finishing tool only removes the last 0.3–0.5 mm.

Coolant is the parameter most shops set once and never revisit. On hardened stainless and titanium, the failure mode is heat at the tool tip, and the fix is directed high-pressure coolant aimed at the contact zone, not a flood that never reaches it. On aluminium and brass, the same high pressure is wasted. Coolant strategy should change with the material group, and the nozzle position should be checked at the start of every run, not just at setup.

There is a tradeoff worth naming. High-pressure through-tool coolant consumes more pump power and more maintenance. On short runs it is hard to justify. On long runs of 17-4PH or Ti-6Al-4V, it typically pays back within the first batch because tool changes and re-cuts drop. We decide per part number, not per shop policy.

Chip evacuation is part of the same decision. Deep tooth spaces trap chips, and a recut chip is the fastest way to chip a carbide edge. Air blast plus a short, controlled coolant pulse often works better than continuous flood.

• 1
  Keep the chipload, reduce the depthSlowing the spindle with the same feed per tooth rubs the edge and shortens tool life.
• 2
  Aim the coolantHigh pressure at the contact zone for stainless and titanium; flood is enough for aluminium and brass.
• 3
  Split rough and finishLeave 0.3–0.5 mm for the finishing pass so the finish tool sees a clean, light cut.
• 4
  Do not recut chipsAir blast plus a short coolant pulse clears deep tooth spaces better than continuous flood.

A common cost leak is heat treating at the wrong point in the sequence. If a gear is cut soft, then carburized and hardened, the tooth flank distorts. The distortion has to be removed, and that means a grinding or hard-finishing step with its own setup, its own wheel cost, and its own scrap risk. Some of that is unavoidable on case-hardened gears. Much of it is not.

Where the design allows, cutting the tooth form after hardening with a carbide or CBN tool removes the finishing setup entirely. That works well for through-hardened steels in the 45–55 HRC range and for smaller modules. Above that hardness, tool cost and edge life turn the decision the other way, and grinding is the cheaper route. The point is to make this a deliberate choice from the drawing, not a default that is repeated because it was done that way on the last job.

Distortion also depends on how the blank is held during heat treat. A gear that is quenched while lying flat and unsupported will move differently from one that is fixtured or press-quenched. If the hardening shop is not told what the flank tolerance is, they cannot help hold it. Send the tooth tolerance with the heat treat order.

Material selection changes gear cutting cost more than most engineers expect. Free-machining grades such as 303 stainless and C36000 brass cut cleanly and hold a good flank finish, but they are weaker and are not suitable for loaded gears. Alloy steels like 4140 and 4340 cut well in the annealed state and harden predictably, which keeps the finishing step short. Titanium and Inconel cut hot, wear tools fast, and need the coolant and speed strategy adjusted from the first pass.

The practical rule is to match the material to the load case, then accept the machining cost that follows. Choosing a harder material to gain a small safety margin often adds more cost in tooling and finishing than the margin is worth. If the load case is not clear, it is worth resolving before the material is ordered.

Inspection is the other half. On gear work, checking the tooth form at the end of the run is too late. We monitor the first article against the drawing, then hold a sampling interval through the run, and inspect 100% before shipment. Reports are available on request. This is not a quality slogan. It is what keeps a batch from being re-cut, and re-cutting a hardened gear usually means starting over.

• 1
  Free-machining grades303 and C36000 cut fast and finish well, but are not for loaded teeth.
• 2
  Alloy steels4140 and 4340 machine well annealed and harden predictably.
• 3
  Titanium and InconelPlan for high-pressure coolant and shorter tool life from the first pass.
• 4
  Inspection timingFirst article, sampling through the run, then 100% before shipment.

The last cost lever is the earliest one. A DFM review before the first chip is cut catches the features that force extra setups, tight tolerances that do not need to be tight, and tooth forms that could be cut on a smaller machine. On gear work, small drawing changes have large effects. Relaxing a non-functional bore tolerance from ±0.01 mm to ±0.02 mm can remove a finishing op. Adding a relief groove can let a tool exit cleanly instead of being backed out.

We run this review as part of the quotation. It takes hours, not weeks, and it is where most of the savings on a gear job are actually found. Once the process is set and the first article is approved, the cost is largely fixed. The window to change it is before the order.

None of these seven items is exotic. Most of them are decisions that get made by default because no one asked the question. Ask the question at quoting time and the cost per good tooth drops without touching the machine.