Follow-Up Fixturing System

Wing and fuselage structures are joined with bolts and rivets, so a joint is only as good as its hole. And the parts are thin: long against their thickness and low in stiffness, they deform and vibrate under the tool. The tool knocks against the part, the holes come out poor, tools break, and parts get scrapped.

The usual answer is a fixture shaped to the part that supports a large area of it — stiff, but made for one part and very expensive. The alternative is to support the part only where the drill is, and to move that support with the spindle: a follow-up fixture.

This one has two halves: a feed bolted to the spindle mount pushes a support plate onto the part, and vacuum cups on that plate grip it close to the hole.

An earlier follow-up fixture drove its support from one side of the spindle, and resisted torque and vibration poorly. That is where my part of this one starts.

The feed cannot sit on the drilling axis — the spindle is already there. So it sits off to the side, at an offset e, and the axial force at the drilling point reaches it as a moment: M = F·e.

An electric cylinder is built to push along its own axis. Put a moment on it and it no longer holds position accurately — and holding position is the whole job of a feed.

So instead of one actuator beside the spindle there are two, placed symmetrically about the drilling point, each an electric cylinder driven by a closed-loop stepper. Each takes F/2 at the same offset on opposite sides: the two moments are equal and opposite, and cancel. What is left on each cylinder is axial force, which is what it was built to carry.

Checked in FEA

I ran the feed module through a static analysis in SolidWorks Simulation at an 800 N limit load. The largest displacement anywhere in the module came out at 1.768 × 10⁻³ mm — about 0.0018 mm.

What the patent protects

The layout is in a Chinese invention patent, CN119319467B, granted in October 2025. Claim 1 includes this layout. In the words of the granted claim, the feed modules come “in two groups, distributed symmetrically on both sides of the spindle mount (5), in a symmetric layout overall” (my translation).

A vacuum cup on this fixture has two jobs that want opposite materials. To seal, it has to be soft: a hard cup on a curved part touches only some of its rim, and the vacuum never forms. To locate, it has to be hard: a soft cup seals well but deforms and vibrates under load, so it cannot serve as a datum for precise machining.

The rigid-rib cup stops asking one part to do both. The soft cup still does the sealing; a metal rib inside it does the locating. The rib is a disc that sits inside the cup’s bowl, with a stem through the cup’s center hole and a clamping piece behind it, so load goes through the rib into the fixture body instead of through the soft cup. An air passage runs through the rib and its stem, so the vacuum is drawn through it.

Same load, both designs

Under the same 100 N axial load in SolidWorks static FEA, the plain cup’s largest displacement is 2.042 mm. With the rib it is 7.003 × 10⁻⁴ mm — about 0.0007 mm, more than three orders of magnitude less.

Two simulations, not a test. What they show is the point of the design: the rib, not the soft cup, is now what the part is located against.

A contact face that can be changed

Making the rib the datum has a cost: it is hard, it presses directly on the part, and hard on a softer surface leaves marks. So the rib’s contact face is a quick-change insert, swapped to suit the material of the part being drilled.

Filed

The cup is the subject of a Chinese invention patent application, CN120650317A, published in September 2025 and under substantive examination. I am co-inventor 4 of 9. The quick-change insert is not part of the application.