8D Problem Solving

A: A feels conservative but it's targeted containment: you protect the customer and preserve evidence while D5 is still open. B looks tidy from a documentation mindset, but it bakes an unapproved change into the system and erases the signal you're trying to study. C is a classic first-instinct move; it quiets scrap today while driving thread yield and joint fatigue tomorrow. D sounds safety-first, yet 8D containment isn't a production shutdown unless there's an immediate hazard; here it just burns launch without learning.

A: A catches the documentation break: D2 must tie to a controlled characteristic. B sounds like something you've heard on the shop floor, but MMC bonus doesn't create a flatness requirement. C is tempting when you're annoyed with paperwork, yet the GA is still a contractual reference here. D mixes functional thinking with GD&T; they address different error modes and aren't swappable.

A: B follows the chain: one bad out of 5,000 is 0.02%, and sampling 10% gives a similar escape order. A flips the logic by treating sample rate as defect rate. C feels cynical but ignores basic probability. D shows real math effort yet invents a cycle basis that isn't stated, a common D3 mistake.

A: A removes human variability and closes the cause. B feels responsible but relies on memory and compliance. C improves detection, not prevention, so it's still containment. D reduces symptoms on paper while the physics stay the same, a DFMEA red flag.