Baltimore, MD, USA
N8986B
BOMBARDIER INC CL 600 2B19
The No. 2 engine fan blade airfoil separation was caused by a low cycle fatigue crack that originated at the forward blade pin hole attachment tang due to insufficient residual stress (hoop stress) levels in the Ti-8-1-1 blade material. The fan blade had undergone a hot form repair following a bird strike incident in December 2007 to restore the blade profile. During the hot form repair process the blade was placed in a ceramic die and heated to 1150-1200oF to restore the blade geometry. The hot form process relieved the compressive residual stress layer in the fan blade and the layer was not properly restored with a lance peen/shot peen step as called out in the GE Standard Repair Manual. There was visual evidence of 100% lance peen/shot peen coverage on the fan blade attachment pin holes surfaces that was identified during the materials lab examination, but the peen marks were most likely formed during the original fan blade manufacturing process. When the fan blade airfoil separated during the incident it was contained by the engine fan case containment system. The fan blade impact with the fan case containment system resulted in a bulge in the Kevlar wrap that damaged the attaching hardware at the fan cowl flanges, and the cowl subsequently separated when exposed to the jet stream. When the fan cowls separated from the engine they impacted the thrust reverser translating cowl and core cowl and resulted in gouging and surface buckling. The fan blade airfoil separation caused a fan imbalance that led to high No. 2 engine vibration levels. The high vibration levels caused the engine transition case to LPT flange separation, AGB housing fracture, and starter damage.
HISTORY OF FLIGHT On January 23, 2017, at about 1830 eastern standard time (EST), an Endeavor Air, Bombardier CRJ200, N8986B, equipped with two General Electric (GE) CF34-3B1 turbofan engines, experienced a No. 2 (right) engine fan blade separation during initial climb from Baltimore-Washington International Airport (BWI), Baltimore, Maryland. The crew reported hearing a loud bang from the back of the airplane, immediately followed by increased vibration levels, a No. 2 engine turbine inlet temperature warning indication, and a No. 2 engine thrust reverser warning indication. The crew declared an emergency, shut down the No. 2 engine, and returned to BWI where they made an uneventful single engine landing. There were no reported injuries to the passengers or crew. The flight was operated in accordance with 14 Code of Federal Regulations Part 121 and was a regularly scheduled flight from BWI to Cincinnati/Northern Kentucky International Airport (CVG), Covington, Kentucky. DAMAGE TO THE AIRPLANE The No. 2 engine fan blade separation did not result in any damage to the airplane fuselage or right wing. The No. 2 engine fan cowl halves separated in flight and were not recovered. The No. 2 engine thrust reverser translating cowls and core cowls both exhibited impact damage including material gouging and minor surface buckling. The damage to the thrust reverser translating and core cowl surfaces was radially in towards the engine centerline and there was no evidence of engine uncontainment. TEST AND RESEARCH Engine Examination A visual examination of the No. 2 engine was conducted in a maintenance hangar at BWI following the incident. One fan blade foil was separated at the blade root pin hole attachment. The fan blade airfoil and the inboard half of the forward fan blade pin hole attachment tang were embedded in the fan case honeycomb and Kevlar containment layers in-line with the fan plane of rotation, between the 5 and 6 o'clock positions. All remaining fan blades were intact but exhibited leading edge impact damage and material tearing most concentrated along the blade tips. All the low pressure turbine (LPT)-to-transition case flange bolts were fractured and there was a 7/16 inch gap between the case flanges. The flange separation allowed the LPT case and jet pipe to rotate 30 degrees in the clockwise direction relative to the rest of the engine. The No. 2 engine accessory gearbox (AGB) housing was fractured near the starter and a piece of the housing was separated and missing. Two of the three AGB attachment points had failed. The engine starter housing was fractured downstream of the air valve. Metallurgy The separated fan blade airfoil and forward blade pin hole attachment tang fragment were removed from the fan case and sent to the NTSB material laboratory in Washington D.C. for analysis. The analysis concluded that the forward and middle blade tang fracture surfaces had cracks that originated from the inner diameter pin hole surface and the crack features were consistent with low cycle fatigue (LCF). No nicks, dents, or corrosion were observed near the crack origins. A fatigue striation count estimated that the crack grew over 2,560 cycles between initiation and fan blade airfoil separation. The material composition and blade geometry were consistent with drawing specifications for Ti-8-1-1. During the manufacturing process and after hot form repair, the fan blade root is shot peened/lance peened to form a compressive residual stress layer that improves blade wear characteristics and reduces the likelihood of crack initiation and propagation. There was visual evidence of 100% lance peen/shot peen coverage on all recovered fan blade pin hole surfaces. A residual stress analysis of the separated fan blade pin hole surfaces identified residual stress levels that were well below the values measured on other blades installed on the incident engine as well as multiple fan blades returned to GE from other fleet operators. ADDITIONAL INFORMATION Fan Blade Serial Number KGACA251 History A combination of manufacturing records, shop repair records, and engine logbooks were used to compile the history of the separated fan blade, serial number KGACA251. The fan blade was manufactured in December 2010 by Tect-Utica using Allvac Ti-8-1-1 two inch bar stock material. The blade entered service in April 2001 and remained in the same engine until December 2007, when it was one of four blades that were damaged by a bird strike event. Standard Aero completed field repairs on the engine, and three of the four blades were determined to be beyond repair limits and were scrapped. Blade serial number KGACA251 was sent to the Airfoil Technologies International- United Kingdom (ATI-UK) for repair. ATI-UK performed a hot form repair on the blade, and it was recertified in February 2008. In April 2008 the fan blade was installed in another engine but did not enter service before being removed and installed in the incident engine in June 2008. At the time of installation, the blade had 4,234 cycles since new. In June 2012 the engine entered the Standard Aero shop and the incident fan blade underwent: "Major Blend, Polish Forward Tang, and Apply Molydag and Molykote Coating." While in the shop, an eddy current inspection of the fan blade root was completed to comply with service bulletin 72-0106. The fan blade had a total of 10,736 cycles when it entered the shop for repair. No additional findings were reported for the engine or fan blade. The fan blade had accumulated a total of 18,063 cycles at the time of the separation event. Fan Blade Hot Form Repair When the incident fan blade entered the ATI-UK shop in December 2007, a hot form repair was completed to restore the blade profile in accordance with the GE Repair Manual- Fan Rotor Assembly- Fan Blades- Creep Hot Form Repair procedures. The procedures outline blade cleaning and protection requirements in preparation for reforming the blade profile with a ceramic die that is heated to 1150-1200oF at a pressure of 30-40 psi. The blade remains in the die at the specified temperature and pressure for one to two hours. After removal from the ceramic die, the blade root is lance peened/shot penned to restore the compressive layer to pre-repair residual stress level. Corrective Actions Following the incident, GE released an all operators wire (AOW) 17CF34-003 on February 23, 2017 to notify CF34-3 operators of the fan blade separation event. AOW 17CF34-004 was released on March 21, 2017 and requested that operators send GE the results from fan blade pin hole eddy current inspections recommended in SB 72-0106 (Regional Fleet) and SB 72-0091 (Business Jet Fleet). GE originally released the eddy current inspection service bulletins in January 2001 and recommended that operators perform an eddy current inspection of the fan blade pin hole attachment of all CF34-3 fan blades prior to 18,000 cycles since new with a re-inspection every 3,000 cycles. The repair records for all CF34-3 fan blades that underwent hot form repair at ATI-UK were reviewed and GE released AOW 17-CF34-008 on August 11, 2017 to notify operators of upcoming SB's that contained the initial list of suspect fan blade serial numbers that did not have documentation to verify the blades were shot peened/lance peened as part of the hot form repair process in accordance with GE repair specifications. SB 72-0314 (Regional Fleet) and SB 72-0306 (Business Jet Fleet) were released as category 2 bulletins on September 27, 2017 with the list of suspect fan blades and a recommendation to inspect all CF34-3 engines to determine if any of the suspect fan blades were installed. If a suspect blade was identified, the operator had the option to return the blade to GE for scrap or complete a fan blade bushing repair in accordance with a cold expansion procedure approved by GE. The bushing repair re-establishes the residual stress layer in the fan blade pin holes. GE released SB 72-0316 (Regional Jet Fleet) and SB 72-0308 (Business Jet Fleet) as category 2 bulletins on January 4, 2018 to request that operators review all available CF34-3 fan blade repair records to determine if any additional fan blades were repaired at ATI-UK that were not on the original serial number list.
A Bombardier CRJ200 No. 2 engine failure due to the separation of a fan blade airfoil at a low cycle fatigue crack that originated at the forward fan blade pin hole attachment. The separated fan blade had previously undergone a hot form repair and both metallurgical analysis and repair records indicated that the lance/shot peen step required to restore sufficient residual stress levels to the blade material was not completed.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
