Rebuilding the sun visor clip.

The problem

This little clip holds the sun visor up against the headliner. After forty-some years the plastic gets brittle, and one of mine had broken. Replacements are hard to find, but I still had an original to measure, so it became the first part I properly reverse-engineered.

A part-number question

The part number I first found in online diagrams, 74348-12010, turned out to describe a different, screw-mounted holder. Mine is a crescent-shaped clip that mounts with a push-in fir-tree pin and one screw. The closest candidate so far is the 74336 series, but I haven't confirmed it.

From photos to calipers (v1 to v5)

v1 came from online diagrams and photos only, and it modeled the wrong mechanism entirely: a screw plate with a snap-in channel. Photos can show that a feature exists. They can't show what it does.

v2 came from calipers. Seven rounds of measuring the original, 18 critical dimensions:

• the crescent-shaped body
• the C-hook that cradles the 8.1 mm visor rod, with a 7.52 mm throat
• a fir-tree push pin with 12 barbs
• a counterbored screw hole

v3 to v5 were the first prints, in PLA and PETG, held next to the original and corrected:

• the fir-tree ribs are on both sides, not one
• the hook's jaw came in about 3 mm
• the pin-to-hole spacing was corrected to a measured 9.0 mm
• bevels on the corners, with the edges that locate the part kept crisp

First fit in the car, and two deliberate changes (v5 to v8)

Aug 16: v5, printed in PLA, mounted and held the visor, but the visor had to be forced up to hook and unhook it. Measuring showed why: the gap the rod passes through on its way in was 7.73 mm, and the rod is 8.10 mm. The original shape is tighter than the rod it has to let in.

Deliberate change 1 (v6): a shorter lip at the entry turns 0.37 mm of interference into 0.62 mm of clearance. The part of the hook that actually holds the rod, the 7.52 mm throat and the 8.10 mm cradle, didn't change. That makes this a corrected part, not an exact copy, and the record says so.

Deliberate change 2 (v8): the v7 print came out with its fir-tree barbs fused into a ribbed cone. The original has 12 barbs packed into 11.2 mm, which leaves gaps of 0.23 mm, narrower than the 0.4 mm nozzle can draw. Injection molding doesn't care about that; FDM printing does. v8 has 6 barbs at the same outer diameter, so it grips the same mounting hole with fewer, properly formed steps.

ASA, and making it look right (v8 to v12)

Why ASA. A car interior bakes in the summer, and PLA starts to soften around 60 °C. ASA holds up to heat (its glass transition is around 100 °C) and to UV. It needs an enclosed printer with the chamber warmed up first, so it prints in a Prusa Core One with the vent closed.

Aug 20: v8 in orange ASA worked, but it looked boxy. v9 to v12 were about looks:

• softer edges
• a rounded front
• a clean shelf under the visor opening
• a smoother visor opening itself

The v12 rounding reaches into the spot where the rod goes in, so it needed its own test in the car instead of borrowing v8's result.

Sept 8: v12 printed in tan ASA, much closer to the headliner: about 5 g of plastic and 21 minutes of printer time. The material under it on the bed is print support, and it comes off. Every print is checked before it runs, including that no support ends up inside the screw hole.

Installed the same day. It mounts solidly, holds the visor up, and the visor hooks in and out without forcing it, which is the problem v6 set out to fix.

What's next

• Time in the car. Summer heat and sun are the real test, and I'll report back.
• Other A60s. If yours has a broken or missing clip, tell me about your car.
• The part number. Still open.