This picture (which I posted 8 years ago) is from the wreckage of flight 587 after taking off from JKF international airport.

The NTSB concluded the loss of the tail in flight was due to improper rudder inputs by the pilot who was trying to fly through the turbulence of a Japan Airlines 747.

A delamination in the tail was repaired prior to delivery from the manufacturer. I believe this repair can be seen in the photo. There are three rows of what appear to be "through bolts" which effectively 'dam' the stresses at the first row. These bolts prevent stresses from reaching the thickest part of the "attachment point"- the point intended to carry the load. For whatever reason, I believe the NTSB lied about the excessive rudder inputs causing the separation. The tail probably broke loose prior to the "excessive rudder inputs" causing the pilot to correct for a partial loss of the control surface. Leading to increasing oscillations- the tail wagging the dog.

Now there are questions about flight 447. The tail has been shown to have broken off in flight.

"Without black box data, what exactly caused the tail to come off may never be known. "

And:

"On May 12, 1997, another A300 experienced an in-flight upset, a series of rolls and stalls, severe rudder deflections and high structural loading on an approach to Miami. The aircraft landed safely and intact and was cleared for flight following an inspection using accepted techniques, but Airbus engineers requested a more thorough inspection of the composite tail. When that inspection occurred in 2002, the composite material had delaminated so severely that the $1.9 million tail was scrapped."

If the Airbus can not fly in turbulence, it has absolutely no business in the air.