Continuing the “science” from last week, I’ve now seen how CA affects the performance of both a press-fit, and a keyed fit when the CA is applied before pressing the axle into the wheel. Because it would be great to be able to fine tune the quartering after assembly, I also picked up some Locktite 290, which is designed to wick into an assembled joint.
| Failure weights (g) | Trial 1 | Trial 2 | Trial 3 | Average | Comparison | Improvement |
| Interference + CA | 554 | 287 | 460 | 434 | 223.5 | 90% |
| Key + CA | 470 | 650 | 557 | 559 | 358.5 | 60% |
| Locktite 290 | 214 | 386 | 341 | 312 | 263 | 20% |
For comparison, I’ve copied similar strategies from the previous round of experiments. Compared to dry fit, CA improved the performance by 60-90%. Compared to CA fed by capillary action, Locktite 290 improved the performance of a straight interference fit by about 20%.
Pembroke:87 
The axle is steel, right? At least some formulations of loctite require the presence of Steel (or iron, i assume) on one (or both) of the parts to do their thing. I’m not sure if that’s true of 290.
Yes, generally they use the presence of iron.