I think that the gun managed for so many years would seem to indicate that the braze connection was within an acceptable range, which would include the temperature to perform the braze. I don't think these incidents are design or execution problems, but more likely exceeding an intended limit or changes that had subsequently happened.