The first reports to 911 happened only seconds after the crash, with cries tinged with shock and terror. “Plane fall”, then fire engulfing the sky was the witness account given by several drivers on Louisville Grade Lane, while others, caught behind the wall of fire, begged for escape. It landed on the building. nowhere to go was the cry heard below the smoke, where one caller stood, reporting the incident. Within those early moments, the scope of the disaster unfolding had already been assessed: Fourteen were dead, twenty-three injured, while destruction sprawled through warehouses, roads, and a petroleum recycling unit.
But underlying the human aspect, experts swiftly homed in on the apparent failure of the mechanical systems that manifested during the crunch moment of the plane’s flight. According to the National Transportation Safety Board’s preliminary investigation, the left engine and the pylon assembly came loose from the McDonnell Douglas MD-11F aircraft during the moment of takeoff rotation. This event was caught on camera, where the GE CF6 engine went soaring over the wing, spewing flames, before plummeting alongside the runway. A secondary fire broke out at the pylon installation area of the wing.
Flight data analysis also illustrated the freighter did not climb beyond an elevation of approximately 30 feet above ground level before the left main landing gear impacted the roof of the UPS warehouse. This series of impacts lasted for approximately 3,000 feet and included the scattering of debris and the ignition of several structures. Reports from the control tower personnel included normal takeoff acceleration and an abnormal flight path and left rotation indicative of asymmetric thrust and lift.
The MD-11 engine pylons are seated below the wing and attached to a wing clevis by forward and aft lugs, with a spherical bearing. Materials analysis conducted by the NTSB found that the MD-11 left pylon aft mount lugs were compromised by fatigue cracks on both sides of the outboard and inboard lugs, while assessments of overstress failures were also made. There were also indications of sudden final fracture. The outer race of the spherical bearing bifurcated; its ball component is exposed. There were also parts of the attachment components attached to the wing debris found near the runway site. The right engine pylon had also failed.
According to experts, this kind of fatigue crack takes time to develop through several flights until finally, the strength is weakened, beyond which the force becomes more than the structure can handle. “It finally gets to a point where the force overcomes what the structure can withstand and that’s a point of failure,” John Cox, an American safety expert, explained. Anthony Brickhouse identified this as “a major clue” and emphasized the fact that this was just one of the things that shouldn’t have escaped maintenance.
The maintenance records indicated that it had accumulated a total of 92,992 hours and 21,043 cycles. The last detailed inspection for a 72-month inspection of the left side aft mount on the pylon was done in October 2021, while lubrication of thrust links and spherical bearings was done in October 2025. Special detailed inspections based on cycle thresholds had not yet been due. The plane had recently undergone some repair work at San Antonio, taking a six-week downtime before resuming operations.
This crash has similarities to the 1979 American Airlines Flight 191 disaster, in which the McDonnell Douglas DC-10 lost its left engine and pylon on takeoff from Chicago O’Hare Airport. In that incident, the loss of the pylon also caused the removal of the leading edge of the wing, resulting in an uncontrollable roll. As mentioned earlier, the MD-11 aircraft’s wing design is relationally similar to the DC-10 wing design.
In the aftermath of UPS flight 2976, UPS immediately grounded its entire MD-11 fleet on November 7 because of recommendations from Boeing. The FAA also developed airworthiness directives that initially targeted MD-11 and MD-11F planes and later DC-10 planes, as all three are similar in model design. Fortunately, authorities have managed to retrieve the cockpit voice recorder with more than two hours of high-quality recordings and the flight data recorder, which has 63 hours of data from 24 flights.
The NTSB is currently investigating metallurgical failure origins, models of load at the pylon wing interface, and adequacy of inspection intervals. Boeing is pursuing additional models and testing, with fleet-wide assessments of pylon hardware still underway. The final report, expected in 18 to 24 months, will detail likely causes and safety actions, not blame.
