Hole bottom plane machining: how the sealing face of an aviation valve body is actually produced

• 1
  Key takeaways
• 2
  What the bottom plane of a valve hole actually does
• 3
  Hole bottom plane machining: what happens at the cutting edge
• 4
  Why the spindle and fixture decide the result
• 5
  Milling, grinding, and lapping: which fits which face
• 6
  When hole bottom plane machining is the wrong process
• 7
  How we set up hole bottom plane machining
• 8
  Process fit for hole bottom plane machining

Inside an aviation valve body, the bottom plane is the annular face at the end of a bore. A poppet, a ball, or a spool lands on it and stops flow. The face is usually narrow, often 1 to 3 mm wide, and sits at the bottom of a bore that may be 20 to 40 mm deep. Tool reach matters as much as the cutting edge.

Sealing is a contact problem. The valve element touches the face along a line. If that line is not continuous, fluid finds a path. Flatness of 0.005 mm across the face and a surface finish of Ra 0.2–0.8 μm keep the contact line closed under pressure. A rough face leaves micro-channels that open as pressure rises.

The functional requirement is rarely a diameter. It is the relationship between the bottom plane and the bore axis. If the face is not square to the axis, the valve element contacts on one side only. That single-side contact is what causes weeping at low pressure and chatter at high flow.

Engineers sometimes specify the face as a simple depth dimension. That misses the point. Depth sets where the valve sits. Squareness and flatness set whether it seals. Two parts can share the same depth and behave completely differently.

At the bottom of a bore, a flat end mill or a back-spotting tool cuts on its end teeth only. The effective cutting speed at the center of the tool drops to near zero. That dead zone pushes material instead of shearing it, which raises cutting force and leaves a smeared center if the tool dwells.

For a face 12 mm wide, a 10 mm end mill has to sweep. A 16 mm tool can cover it in one pass but may not fit the bore. This is why tool selection is a compromise between reach, stiffness, and the number of passes. More passes mean more chances to leave a step.

Interrupted cuts are common. Valve bodies have cross-drilled ports that break into the bore. When the end teeth cross a port, the load drops to zero and then spikes. Carbide handles this better than high-speed steel, but the edge still chips if feed per tooth is too high. A feed of 0.03–0.05 mm per tooth on aluminium is a reasonable starting point.

Heat is the other issue. Aluminium valve bodies conduct heat away quickly, so the face may look clean while the subsurface is stressed. Stainless and titanium hold heat at the edge, which accelerates flank wear. On 17-4PH or Ti-6Al-4V, expect to change inserts more often than the cycle time suggests.

A bottom plane is cut at the end of a long reach, so every error upstream shows up on the face. Spindle tilt of a few thousandths of a degree produces a wedge across the face. On a 12 mm face, 0.005° of tilt is about 1 μm of height difference from one side to the other. That is enough to show on a blueing check.

Thermal drift matters on long cycles. As the spindle warms, the tool center moves. If the bottom plane is the last feature cut, it inherits that drift. Cutting the face early, then returning for a light finishing pass, often gives a flatter result than one deep pass at the end.

The fixture has to hold the part without distorting the bore. Clamping force on a thin-wall valve body can ovalize the bore by several micrometres. When the clamp releases, the face springs back and the measured flatness no longer matches the machined condition. Light clamping plus a settled roughing pass avoids most of this.

On a five-axis machine, the rotary table adds one more error source. Table runout and backlash appear as a tilt at the tool tip. For faces under 0.005 mm flatness, we check the table with a test bar before the run, not after.

Milling with a rigid end mill suits most aluminium and stainless valve bodies. It holds flatness around 0.01 mm and Ra 0.8–1.6 μm without a second setup. If the drawing calls for Ra 0.4 μm or better, milling alone usually falls short.

Grinding with a small wheel or a motorized end cap improves flatness and finish. It also introduces burn risk on thin sections and can load the wheel when cutting aluminium. Grinding works best on hardened steel and on faces where the tolerance is tighter than 0.005 mm.

Lapping is a hand or semi-hand process. It can reach very fine finishes on small faces, but it depends on operator feel and is hard to control on a production run. We treat lapping as a repair step, not a primary process.

The choice follows the material and the finish callout, not the drawing title. A 7075 aluminium body at Ra 0.8 μm is a milling job. A 440C stainless seat at Ra 0.2 μm is a grinding job. Mixing them up adds cost without improving the seal.

If the face sits at the bottom of a bore narrower than 6 mm and deeper than 5 times its diameter, a rotating tool cannot reach it with enough stiffness. The face will be tapered or chattered. In that case, a formed tool or an EDM electrode is the better route.

If the material is a soft aluminium alloy and the drawing asks for Ra 0.2 μm, milling will smear rather than cut. The surface may measure fine but fail a tape test. Grinding or a diamond-turned insert is needed.

If the valve body is a one-off prototype with no sealing requirement, spending time on a ground face adds cost for no benefit. A milled face at Ra 1.6 μm is enough for fit checks and flow testing.

If the face is interrupted by more than two ports, the cutting edge sees repeated impacts. Below 0.01 mm flatness, this becomes a tool-life problem, not a machining problem. Consider EDM or a redesign that moves the ports away from the sealing band.