Pilot Debrief

BOEING 737-824 near Denver, CO — 2019-12-23

Final reportDCA20LA047
Sign in to save
Date
2019-12-23
Location
Denver, CO, USA
Airport
DEN
Aircraft
BOEING 737-824
Registration
N87513
Category
Airplane
Highest injury
None
Fatalities
0
Phase of flight
Landing

Probable cause

Maintenance personnel’s excessive grinding of the left main landing gear’s aft trunnion pin during its initial overhaul, which caused heat damage to the base metal and led to the fatigue crack that caused the pin to fail during the accident flight. Contributing to the accident was the failure of maintenance personnel to detect the excessive grinding during the initial overhaul and the fatigue crack during the subsequent overhaul.

Contributing factors

Contributing to the accident was the failure of maintenance personnel to detect the excessive grinding during the initial overhaul and the fatigue crack during the subsequent overhaul.

NTSB narrative

This accident occurred when the left main landing gear of a United Airlines Boeing 737-800 collapsed during the landing roll. Both flight crewmembers reported that the airplane had touched down smoothly on the runway centerline, but the first officer reported that he then felt the airplane “shudder” and tilt left wing down. The first officer further reported that, as the airplane continued to decelerate, he struggled to keep the airplane on the runway centerline. Visual meteorological conditions and a headwind were present when the airplane landed, so the weather did not play a role in the accident circumstances. Postaccident examination found that the aft trunnion pin in the left main landing gear failed during the landing due to a fatigue crack. The crack, which had grown to a depth of 0.154 inches, was large enough that stress concentration at the crack tip (from loads during the landing) caused the pin to fracture, resulting in the collapse of the left main landing gear. The fatigue crack initiated from a small intergranular region just below the external chromium electroplated layer. The size of the intergranular region was about 0.011 inches deep and 0.074 inches wide. Multiple fatigue cracks had initiated from this intergranular region. These individual cracks coalesced and propagated inward, as shown by ratchet marks and fatigue striations. Etching showed that the intergranular region where the fatigue crack initiated was located along an area exhibiting a darker visual contrast. This characteristic was consistent with over tempering and an area of localized exposure to higher temperatures relative to the alloy steel material of the pin outside this area. The most likely cause of this elevated heat input was excessive grinding performed during maintenance overhaul of the pin. Any grinding operation introduces the risk of a local microstructure change, but over-tempering indicates hard or excessive grinding involving temperatures that are high enough to change the steel material’s microstructure. According to United Airlines, the pin had accumulated 23,535 landing cycles since entering service in November 1998. The first overhaul occurred in May 2008 when the pin had accumulated 10,613 cycles; the second overhaul occurred in December 2017 when the pin had accumulated 21,226 cycles. A fatigue crack analysis showed that the thumbnail crack had been present for at least 6,225 landing cycles. Between the time of the last overhaul (December 2017) and the accident, the pin had accumulated 2,309 landing cycles, and the pin had accumulated 12,922 landing cycles between the earlier overhaul (May 2008) and the accident. Therefore, the crack was present before the December 2017 overhaul but was likely not present before the May 2008 overhaul. According to the work order for the last trunnion pin overhaul, which was completed on December 28, 2017, the chromium layer had not been stripped from the trunnion pin before the cadmium electroplating process. Postaccident testing performed at Boeing on the accident pin demonstrated that the underlying cracks could not be detected by magnetic particle inspection when the chromium electroplating layer was present. Fluorescent penetrant inspection was also performed with no indications noted. The earlier overhaul, which was completed on May 23, 2008, included grinding of the trunnion pin surface, first to remove the old chromium electroplating layer and then after the new chromium layer had been electroplated on the trunnion pin. Because grinding steps were performed during this overhaul, the over-tempering of the pin most likely occurred at that time. The investigation determined that the bare trunnion pin was not inspected after the chromium electroplating layer was stripped. According to information from the May 2008 overhaul provider, the company performs a visual inspection, temper etch, and magnetic particle inspection after stripping and machining the pin (before shot peening). The investigation could not determine why these inspections were not performed on the pin. A temper etch on the bare pin should have revealed an area of excessive grinding, which would have prevented the part from progressing through the overhaul and being placed back into service. In summary, the grinding operation on the trunnion pin that occurred during the May 2008 overhaul created heat damage and areas of over-tempering to the base alloy steel material. No inspection was performed to detect the excessive grinding, so the trunnion pin was returned to service with the crack undetected under the chromium electroplating layer, and the crack continued to grow for about 4,150 cycles, progressing to a depth of 0.105 inches. In addition, elevated temperatures during the baking steps of the May 2008 overhaul led to the initial heat tinting oxidation, which caused the initial discoloration of a portion of the crack. The nondestructive inspections of the pin during the subsequent overhaul in December 2017 revealed no indications of base metal cracking under the external chromium layer, and the overhaul processes subjected the pin to multiple elevated temperature exposures. These exposures induced additional heat tinting oxidation on the crack surface that had grown since the first overhaul. The pin was returned to service, and the fatigue crack propagated over 2,309 flight cycles (after the December 2017 overhaul) until the final fracture occurred during the accident landing due to ductile separation. The final depth of the fatigue crack was 0.154 inches. This portion of the slow growth region did not exhibit discoloration or heat tinting because there had been no elevated temperatures exposures after the December 2017 overhaul.

Analysis

Primary failure mode
Mechanical failure
First missed decision gate
Overhaul of trunnion pin not performed as per guidelines before the accident.

NTSB coding

Evidence available

  • 5 docket documents
View NTSB final reportView NTSB docket

Docket documents5

Related mishaps