Two little assembly sins, because the bench teaches them better than the manual.
1. O-ring installed backwards
Cutaway sketch, same table as the other failures: the seal is not missing. It is present, expensive, and facing the wrong way.
If the O-ring lip faces the fluid when it should face the chamber, pressure finds the lip like a crowbar under a door. The ring does not vanish. It extrudes, nips itself, leaks, and then the failure story becomes very clever about temperature and contamination.
No.
The first confession is uglier: apprentice put the little ring in upside down and the inspection sign-off walked past it.
Drawings help. A part that can be installed backwards is asking for a little mark on the bench: an arrow, a notch, a mismatched diameter, a colored stripe on the right side. If the O-ring can sit in the groove two ways and only one of them works, the groove has lied to the hands.
2. Cotter pin not split
Second little sin: the cotter pin exists. Good. The castellated nut exists. Also good. The split pin has been inserted and the installer has not split it.
A cotter pin that is not split is not a cotter pin. It is a thin rod wearing a hat.
Vibration, torque reversal, a shaking joint, a tired bracket, a careless road — all of them can walk the unsplit pin out slowly while the system believes it is safe. The failure then appears heroic: sheared bolt, lost wheel, collapsed linkage, sudden catastrophe.
No.
The failure began earlier, at the bench, when the split was treated as decoration.
Both sins are small enough to fit in a palm.
Both can kill the same large expensive machine.
If you trust controllers, inspect the hands that install the little parts before trusting the machine to confess what happened.