One wonders if any engineering calculations were done on the weld joint strength, followed by actual testing to verify the design. My philosophy would be that the joint must have so much extra strength (derated, in engineer-speak) that flaws would not result in catastrophic failure. Any design that relies on near-perfect manufacturing techniques, while using factories that cannot give such high quality, will be doomed to fail. If you must have near-perfect manufacturing quality, then you need to control that too.

It would have been fairly straight-forward to design the weld stronger as per my example above, so that welding quality is less important.

I wonder of the latest Terns are any better. Did they learn from the mistake?