17 Apr 2018 - N772SW 737-700 In-Flight Uncontained Engine Failure
On 17 Apr 2018 a Southwest 737-700, N772SW (27880/601), experienced an engine failure while climbing out of New York La Guardia at 10:43 a.m. about 20 minutes after takeoff, as the aircraft was passing through 32,500 feet. A fan blade was thrown from the #1 engine into the fuselage smashing a passenger window at row 14, killing a passenger who was partially sucked out of the aperture.
The first NTSB briefing describes how multiple aural alerts and warnings sounded on the flight deck. The two pilots donned oxygen masks and reported to air traffic control that they had a Number 1 engine fire, were operating on a single engine and were initiating an emergency descent. “Because they were concerned with potential aircraft controllability issues, they elected to land the airplane with flaps 5 instead of the normal flap setting for a Boeing 737, which would be either flaps 30 or flaps 40,”
The engine fan blades had accumulated over 32,000 engine cycles since new and 10,712 cycles since the last overhaul.
NTSB Chairman Robert Sumwalt said in a statement on 19 Nov 2019: "The accident demonstrates that a fan blade can fail and release differently than that observed during engine certification testing and accounted for in air-frame structural analyses. It is important to go beyond routine examination of fan blades; the structural integrity of the engine nacelle components for various air-frame and engine combinations need to be ensured."
The NTSB said the engine failure was caused by a broken fan blade, and the board said the FAA should require Boeing to determine the fan blade impact location or locations on the engine fan case and redesign the structure to minimize the potential of a catastrophic failure.
Boeing is working on a design enhancement “that would fully address the safety recommendation from the NTSB. Once approved by the FAA, that design change will be implemented in the existing NG fleet.”
The final report was issued on 13 Dec 2019
*** Updated 23 Nov 2020 ***
Contained or Uncontained Engine Failure?
NTSB chairman Robert Sumwalt described the incident as “an apparent in-flight engine failure,” but declined to use the term uncontained engine failure. “I don’t want to sound bureaucratic, but ‘uncontained engine failure’ connotates a very specific thing,” Sumwalt said. “The engine is designed not to have an uncontained engine failure. There are protection rings around the engine to keep shrapnel from coming out. Even though we believe that there were parts coming out of this engine, it may not have been in that section of the engine that technically would qualify this as an uncontained engine failure.”
Previous Similar Incident
On 27 Aug 2016 a Southwest 737-700, N766SW, suffered an apparently similar incident when a fan blade seperated from the fan disk because of fatigue cracking. In that incident debris from the CFM56-7B engine inlet damaged the aircraft’s fuselage, wing and empennage. Details of this incident are available here.
The AD / SB Trail
March 2017 - As a result of the 27 Aug 2016 incident to N766SW, CFM issued a service bulletin (revised in June 2017) that recommended one-time ultrasonic inspection of CFM56-7 fan blades with more than 15,000 cycles since a shop visit “as soon as possible”.
August 2017 - the FAA released a proposed AD, 2017-NE-11-AD, that would require engines with more than 15,000 cycles-in-service since their last engine shop visit to undergo ultrasonic inspection of certain fan blades within six months of the rule’s effective date.
Some airlines requested longer compliance times for this proposed AD. SWA asked for 18 months and AA asked for 20 months. CFM asked for 12 months.
The FAA says the document had been working through the lengthy federal regulatory review process for months, and that it had been weeks away from becoming an AD even before the 17 April Southwest incident.
March 2018 - European Aviation Safety Agency (EASA) issued a comparable AD, 2018-0071, that became effective 2 Apr 2018 requiring ultrasonic inspection of each affected fan blade within nine months.
19 April 2018 - The FAA has now said that it will issue an AD for the CFM56 within the next two weeks. The directive will require “an ultrasonic inspection of fan blades when they reach a certain number of takeoffs and landings. Any blades that fail the inspection will have to be replaced,” the FAA said. This AD will finalize the actions in the AD that was proposed in Aug 2017.
20 April 2018 - CFM issue CFM56-7B S/B 72-1033 recommending that fleet operators perform ultrasonic fan-blade inspections “within the next 20 days” on any engines with more than 30,000 cycles. CFM also recommends repetitive inspections.
Within the compliance time as defined in Table 1 of this S/B (reprinted below), and, thereafter, at intervals not exceeding 3 000 Engine Cycles, accomplish an ultrasonic inspection of each affected fan blade in accordance with the instructions of the S/B.
According to CFM, there are 14,000 CFM56-7 in operation and the engine has accumulated 350 million flight hours with few problems.
20 April 2018 - EASA adopted CFM56-7B S/B 72-1033 into an AD 2018-0093-E, superseding AD 2018-0071 issued in March 2018. The inspection requirements are exactly as per the CFM SB.
20 April 2018 - The FAA issues AD 2018-09-51 which only partially adopts the CFM SB. It only mandates one-time inspections on engines with more than 30,000 cycles. The FAA says in their AD that "We are considering further rulemaking to address these differences."
2 May 2018 - The FAA issues AD 2018-09-10 that requires initial and repetitive ultrasonic inspections (USI) or eddy current inspections (ECI) of the concave and convex sides of the CFM56-7B fan blade dovetail to detect cracking and replacement of any blades found cracked. The initial inspection on each fan blade before the fan blade accumulates 20,000 cycles since new, or within 113 days from the effective date of this AD (May 14, 2018), whichever occurs later. Thereafter, repeat this inspection no later than 3,000 cycles since the last inspection.
The NTSB Investigation
On April 17, 2018, at 1103 eastern daylight time, Southwest Airlines flight 1380, a Boeing 737700, N772SW, experienced a failure of the left CFM International CFM-56-7B engine and loss of engine inlet and cowling during climb about flight level 320. Fragments from the engine inlet and cowling struck the wing and fuselage, resulting in a rapid depressurization after the loss of one passenger window. The flight crew conducted an emergency descent and diverted into Philadelphia International Airport (KPHL), Philadelphia, Pennsylvania. Of the 144 passengers and five crewmembers onboard, one passenger received fatal injuries and eight passengers received minor injuries. The airplane sustained substantial damage. The regularly scheduled domestic passenger flight was operating under Title 14 Code of Federal Regulations Part 121 from LaGuardia Airport (KLGA), Queens, New York, to Dallas Love Field (KDAL), Dallas, Texas.
The NTSB launched a go-team consisting of an investigator-in-charge from the major investigations division and specialists in powerplants, structures, survival factors and operations. Specialists in meteorology, maintenance records, air traffic control, flight recorders, and materials supported the investigation from other locations. Chairman Robert Sumwalt accompanied the team.
Parties to the investigation include the Federal Aviation Administration, Southwest Airlines, GE Aviation, Boeing, the Aircraft Mechanics Fraternal Association, the Southwest Airlines Pilots Association, Transport Workers Union Local 556, and UTC Aerospace Systems. The CFM-56 engine is a joint product of GE Aviation and Safran Aircraft Engines of France; therefore the French Bureau d'Enquêtes et d'Analyses pour la sécurité de l'aviation civile appointed an accredited representative supported by technical advisors from Safran and the European Aviation Safety Agency.
Initial examination of the airplane revealed that the majority of the inlet cowl was missing, including the entire outer barrel, the aft bulkhead, and the inner barrel forward of the containment ring. The inlet cowl containment ring was intact but exhibited numerous impact witness marks. Examination of the fan case revealed no through-hole fragment exit penetrations; however, it did exhibit a breach hole that corresponded to one of the fan blade impact marks and fan case tearing. (See photo below)
Damage to cowl - inboard Photo: NTSB
The No.13 fan blade had separated at the root; the dovetail remained installed in the fan disk. Examination of the No. 13 fan blade dovetail exhibited features consistent with metal fatigue initiating at the convex side near the leading edge. Two pieces of fan blade No. 13 were recovered within the engine between the fan blades and the outlet guide vanes. One piece was part of the blade airfoil root that mated with the dovetail that remained in the fan disk; it was about 12 inches spanwise and full width and weighed about 6.825 pounds. The other piece, identified as another part of the airfoil, measured about 2 inches spanwise, appeared to be full width, was twisted, and weighed about 0.650 pound. All the remaining fan blades exhibited a combination of trailing edge airfoil hard body impact damage, trailing edge tears, and missing material. Some also exhibited airfoil leading edge tip curl or distortion. After the general in-situ engine inspection was completed, the remaining fan blades were removed from the fan disk and an ultrasonic inspection was performed consistent with CFM International Service Bulletin 721033. No cracks were identified on the remaining blades.
The No. 13 fan blade was examined further at the NTSB Materials Laboratory; Figure 2 shows a portion of the blade in detail. Fatigue fracture features emanated from multiple origins at the convex side and were centered about 0.568 inch aft of the leading edge face of the dovetail and were located 0.610 inch outboard of the root end face. The origin area was located outboard of the dovetail contact face coating, and the visual condition of the coating appeared uniform with no evidence of spalls or disbonding. The fatigue region extended up to 0.483 inch deep through the thickness of the dovetail and was 2.232 inches long at the convex surface. Six crack arrest lines (not including the fatigue boundary) were observed within the fatigue region. The fracture surface was further examined using a scanning electron microscope, and striations consistent with low-cycle fatigue crack growth were observed.
The accident engine fan blades had accumulated more than 32,000 engine cycles since new. Maintenance records indicated the accident engine fan blades had been periodically lubricated as required per the Boeing 737-600/700/800/900 Aircraft Maintenance Manual.
According to maintenance records, the fan blades from the accident engine were last overhauled 10,712 engine cycles before the accident. At the time of the last blade overhaul (November 2012), blades were inspected using visual and fluorescent penetrant inspections. After an August 27, 2016, accident in Pensacola, Florida, in which a fan blade fractured, eddy current inspections were incorporated into the overhaul process requirements.
In the time since the fan blade overhaul, the accident engine fan blade dovetails had been lubricated 6 times. At the time each of these fan blade lubrications occurred, the fan blade dovetail was visually inspected as required for the fan blades installed in the accident engine.
The NTSB materials group is working to estimate the number of cycles associated with fatigue crack initiation and propagation in the No. 13 fan blade and to evaluate the effectiveness of inspection methods used to detect these cracks.
On April 20, 2018, CFM International issued Service Bulletin 72-1033 applicable to CFM International CFM 56-7B-series engines recommending ultrasonic inspections of all fan blades on engines that have accumulated 20,000 engine cycles and subsequently at intervals not to exceed 3,000 engine cycles.
On April 20, 2018, the FAA issued emergency AD (EAD) 2018-09-15 based on the CFM International service bulletin. The EAD required CFM56-7B engine fleet fan blade inspections for engines with 30,000 or greater cycles. The EAD required that within 20 days of issuance that all CFM56-7B engine fan blade configurations to be ultrasonically inspected for cracks per the instructions provided in CFM International SB 72-1033, and, if any crack indications were found, the affected fan blade must be removed from service before further flight. On the same day, EASA also issued EAD 2018-0093E (superseding EASA AD 2018-0071) that required the same ultrasonic fan blade inspections to be performed.
The remainder of the accident airplane’s airframe exhibited significant impact damage to the leading edge of the left wing, left side of the fuselage, and left horizontal stabilizer. (See figure 3.) A large gouge impact mark, consistent in shape to a recovered portion of fan cowl and latching mechanism, was adjacent to the row 14 window (see figure 4; the window was entirely missing. No window, airplane structure, or engine material was found inside the cabin.
Three flight attendants were assigned to the flight, and an additional SWA employee was in a jumpseat in the cabin. During interviews, the flight attendants and the employee reported that they heard a loud sound and experienced vibration. The oxygen masks automatically deployed in the cabin. The flight attendants retrieved portable oxygen bottles and began moving through the cabin to calm passengers and assist them with their masks. As they moved toward the mid-cabin, they found the passenger in row 14 partially out of the window and attempted to pull her into the cabin. Two male passengers helped and were able to bring the passenger in.
During interviews, the flight crew stated the climbout from LaGuardia was normal with no indications of any problems; the first officer was the pilot flying and the captain was the pilot monitoring. They reported experiencing a sudden change in cabin pressure, aircraft yaw, cockpit alarms, and a “gray puff of smoke.” They donned their oxygen masks, and the first officer began a descent. Flight data recorder data showed that the left engine parameters all dropped simultaneously, vibration increased, and, within 5 seconds, the cabin altitude alert activated. The FDR also indicated that the airplane rolled left to about 40 degrees before the flight crew was able counter the roll with control inputs. The flight crew reported that the airplane exhibited handling difficulties throughout the remainder of the flight. The captain took over flying duties and the first officer began running emergency checklists. The captain requested a diversion from the air traffic controller; she first requested the nearest airport but quickly decided on Philadelphia. The controller provided vectors to the airport with no delay. The flight crew reported initial communications difficulties because of the loud sounds, distraction, and wearing masks, but, as the airplane descended, the communications improved. The captain initially was planning on a long final approach to make sure they completed all the checklists, but when they learned of the passenger injuries, she decided to shorten the approach and expedite landing.
A cockpit voice recorder (CVR) group was convened and has completed a draft transcript of the entire event. The CVR transcript will be released when the public pocket is opened. Additional information will be released as warranted.
The Final report
The National Transportation Safety Board (NTSB) determines that the probable cause of this accident was a low-cycle fatigue crack in the dovetail of fan blade No. 13, which resulted in the fan blade separating in flight and impacting the engine fan case at a location that was critical to the structural integrity and performance of the fan cowl structure. This impact led to the in-flight separation of fan cowl components, including the inboard fan cowl aft latch keeper, which struck the fuselage near a cabin window and caused the window to depart from the airplane, the cabin to rapidly depressurize, and the passenger fatality.
1. None of the following were factors in this accident: (1) flight crew qualifications, which were in accordance with US regulations; (2) flight crew medical conditions; (3) the airworthiness of the airplane before the left engine failure occurred; and (4) Southwest Airlines’ maintenance of the airplane.
2. The low-cycle fatigue crack in the fan blade dovetail initiated because of higher-than-expected dovetail stresses under normal operating loads, and this crack was most likely not detectable during the fluorescent penetrant inspection at the time of the fan blade set’s last overhaul and subsequent visual inspections at the time of fan blade relubrications.
3. The requirement to perform an eddy current inspection at the time of fan blade overhaul and an ultrasonic inspection at the time of blade relubrication should enable cracked fan blades in CFM56-7B engines to be detected and removed from service before the cracks reach a critical size and the blades fracture.
4. The fan blade fragments that traveled forward of the fan case, along with the displacement wave created by the fan blade’s impact with the fan case, caused damage that compromised the structural integrity of the inlet and caused portions of the inlet to depart from the airplane.
5. Portions of the fan cowl departed the airplane because (1) the impact of the separated fan blade with the fan case imparted significant loads into the fan cowl through the radial restraint fitting and (2) the associated stresses in the fan cowl structure exceeded the residual strength of the fan cowl, causing its failure.
6. The impact of the inboard fan cowl aft latch keeper with the fuselage near the cabin window adjacent to seat 14A caused the window to depart the airplane, the rapid depressurization of the cabin, and the passenger fatality.
7. This accident demonstrated the susceptibility of the fan cowl installed on Boeing 737 next-generation-series airplanes to a fan-blade-out impact location near the radial restraint fitting and the effects of such an impact on the structural integrity of the fan cowl.
8. Given the results of CFM’s engine fan-blade-out (FBO) containment certification tests and Boeing’s subsequent structural analyses of the effects of an FBO event on the airframe, the post-FBO events that occurred during this accident could not have been predicted.
9. The structural analysis modeling tools that currently exist to analyze a fan-blade-out (FBO) event and predict the subsequent engine and airframe damage will allow airplane manufacturers to better understand the interaction of the engine and airframe during an FBO event and the response of the inlet, fan cowl, and associated structures in the airplane’s normal operating envelope.
10. Performing required checklists according to standard operating procedures is a critical part of safe flight operations. However, given the emergency situation aboard this flight, the flight crew’s performance of most, but not all, of the items on the Engine Fire or Engine Severe Damage or Separation non-normal checklist and the nonperformance of the three other relevant non-normal checklists allowed the crew to appropriately balance the procedural requirement of executing checklists with the high workload associated with maintaining airplane control and accomplishing a safe and timely descent and landing.
11. The flight crew’s decision to land at Philadelphia International Airport was appropriate given the airplane’s location at the time of the emergency, the circumstances of the emergency, and the airport’s multiple runways and aircraft rescue and firefighting capabilities.
12. Although not a factor in the outcome of this accident, the flight attendants should have been properly restrained in their assigned jumpseats in case an emergency evacuation after landing was necessary.
13. Federal Aviation Administration guidance addressing options for reseating passengers if an in-flight loss of seating capacity were to occur would help air carriers implement procedures to address this situation.
Seven safety recommendations were released as result of the investigation.
The final report is available here: