The Deadly Tail-Strike Strikes Again
Above and beyond the aging of airframes and normal wear and tear on structural components, there are valid concerns with accumulated and hidden misrepaired damage. The absolute importance of structural integrity in the pressure vessel has been graphically demonstrated by the final accident report on China Airlines Flight CI-611 that broke apart over the sea between Taiwan and Hong Kong in May 2002.
The investigating body, the Aviation Safety Council (ASC) of Taiwan, in a report released on Feb. 25, blamed cracks in the plane’s fuselage as well as poor maintenance for the crash that killed all 206 passengers and 19 crew on board.
The ASC concluded after a two-year investigation that a roughly six-foot crack in the fuselage caused the in-flight breakup. The accident aircraft had suffered a tailstrike during landing in Hong Kong 22 years earlier, and a temporary repair was apparently the only repair done to the damage. ASC could find no evidence of a permanent repair, even though after such a strike, it would have been required by manufacturer Boeing [BA].
ASC found evidence of heavy corrosion beneath the doubler that was installed to repair the tailstrike. Although ASC — Taiwan’s equivalent to the U.S. National Transportation Safety Board — could not confirm exactly how the fuselage came apart, it criticized the air carrier for not finding the damage to the aging aircraft in regular checkups over its 23-year lifespan.
CAL officials, however, rebuked the ASC’s report and questioned its findings. “Since the section of the aircraft that is suspected of causing the crash was not found, the information is not conclusive enough to determine the exact cause of the accident,” CAL said in a statement. “The ASC said that they have found 75 percent of the aircraft, but they actually collected only 50 percent of the damaged Section 46, the key part needed to determine the cause of the crash,” said CAL spokesman Roger Hen.
The air carrier also objected to ASC’s claim that maintenance oversight was to blame. “Our checkups are conducted in compliance with the SRM [structural repair manual]. Boeing stated clearly that if repairs are done in accordance with the SRM, there is no need to report it to the aircraft manufacturer,” Hen said.
According to Boeing’s Repair Assessment Program (RAP), that aircraft should have undergone inspections upon reaching 22,000 flights. Yet the accident occurred when the jet completed its 21,398th flight, a few landings short of the maintenance program’s maximum threshold.
The Taiwan Civil Aeronautics Administration (CAA) also slammed the ASC report, accusing the agency of failing to find the actual cause of the crash. “How could the ASC jump hastily to its conclusions when half of the [section in question] has yet to be found?” asked CAA Director-General Billy Chang.
Important wreckage never recovered
Flight CI-611, a Boeing 747-200 (Reg: B18255), crashed into the Taiwan Straits 23 nautical miles NW of Makung, Penghu Islands off Taiwan, Republic of China (ROC) on May 25, 2002 (ASW, June 30, 2003). The aircraft was flying from Taipei to Hong Kong (HKG), and radar data indicated that the aircraft experienced an in-flight structural breakup at 35,000 feet.
The last “paint” of CI-611 received from Makung radar was at 07:28:03UTC, at 34,900 feet. The aft fuselage section was recovered earliest along the track, the upper forward fuselage next and the engines were furthest along the track. The cockpit voice recorder (CVR) and flight data recorder (FDR) were recovered 25 days after impact in 50 to 70 meters (roughly 200 feet) of water. During the debris retrieval period, three typhoons disrupted the effort. Salvage was conducted in four phases with the final phase involving a formation trawl for pieces embedded in the sandy bottom. The entire wreckage recovery operation lasted five months, recovered approximately 1,500 pieces, and 78 bodies. However, a large portion of the important aft fuselage (Section 46) remains unrecovered.
The FDR stopped recording at 07:27:58.9 UTC (at 34,573 feet), and the last transponder beacon reply was received at 07:28:03 UTC (34,900 feet, by Makung site). FDR data from other aircraft on the Hong Kong route disclosed that the wind profiles had been fairly slack all the way up to CI-611’s cruise height. There was no carbon remains found on any of the recovered bodies or their clothes. No signs of burning or blast damage were found. Injuries were consistent with an explosive decompression. The condition and position of the hinge and door mechanisms indicated that the forward cargo door did not open prior to the airplane’s breakup. The aft cargo door was retrieved in three major segments. The upper portion of the door was recovered with the hinge intact and the actuators in the closed position.
Examination of the hinge, latches, and door mechanisms indicate that the aft cargo door did not open prior to breakup. The Section 46 structure was distributed over a large area, extending more than four miles East to West. The investigators’ comprehension of the progressive breakup process was aided by paint transfer to and from cabin structures (such as seats) and the empennage assembly (such as fin and horizontal stabilizer leading edges), as well as wreckage distribution patterns.
The 22-year-old tailstrike
On Feb. 7, 1980, the accident aircraft suffered a tailstrike during landing in Hong Kong. A preliminary inspection at Hong Kong found abrasion damage on the bottom skin of the fuselage tail between STA 2080 and 2160, and between STA 2578 and 2658. The aft drain mast was missing. The left outflow valve door inboard corner was partially cut. There was no structural repair conducted at HKG. The damage assessment or evaluation report of the specific damage at HKG was unavailable. According to the CAL flight engineer report, the aircraft was ferried back to Taipei. CAL could not provide the aircraft release information according to the Aircraft Flight Operation Procedures of the Taiwan Civil Aeronautics Administration (CAA) in 1976, which state, “Article 45 — A maintenance release shall be completed and signed off by a certified mechanic and the personnel shall certify that the maintenance work performed has been completed satisfactorily with the Maintenance Control Manual.”
Regarding the temporary repair to the tailstrike, the Boeing Field Service Representative (FSR) at Taipei advised Boeing that China Airlines had accomplished a temporary repair consisting of temporary aluminum skin patches. He further advised that China Airlines intended to complete a skin replacement or external patch permanent repair per the structural repair manual (SRM) at a later date (but within four months). ASC could find no record that indicated Boeing was advised that the permanent repair had been completed.
Crash problem area also in the aft fuselage
The recovered portions of the semi-monocoque structure (skins/frames/stringers) in Section 46 were arranged on a reconstruction jig to assist in evaluating the fractures and deformations of the panels. A field examination was conducted on the fracture faces. Item 640 — the 640th piece of wreckage recovered — was found to have flat-fracture surfaces (indicative of slow-growth fatigue mechanisms) on the skin adjacent to an external repair doubler. The doubler measured about two by 10 feet. No slow-growth mechanisms were noted on the remaining skin segments in Section 46. Many of the rivet shanks attaching the doubler were projecting beyond the approved stem length (i.e., the substandard riveting being over- or under-driven).
The fracture directions show the crack progressed under the belly of the airplane and then continued forward along stringer S-50R. The crack then progressed upward at approximately STA 1900. The direction of the fracture propagation was based on hole-to-hole cracking patterns, chevron marks, and branching cracks. The entire empennage separated from Section 46 forward of the aft pressure bulkhead at STA 2360. Separation of the empennage structure resulted from a combination of impact from the Section 46 structure and insufficient remaining Section 46 structure to support the weight and loads of the empennage. The empennage impacted the water relatively intact.
All engine pylon fuse-pins were found intact with engines impacting the water still attached to their associated pylon structures. Further metallurgical field examination of recovered structural items of the fuselage Sections 41, 42, 44 and 46, as well as the outboard and center wing sections, and keel beam were carried out. All fractures examined were characterized as being due to various modes of ultimate ductile separation (i.e., instantaneous tension, shear, compression, bending). No other evidence of any slow-growth cracking mechanisms or pre-existing corrosion was observed.
A Scanning Electron Microscope (SEM) determined the extent of fatigue cracking. In most instances, the fatigue initiated at the scored skin edge next to the doubler and progressed inboard through the direction of skin thickness. Under the SEM, the fracture morphology was clearly indicated by the orientation of striations in the fracture surface. Macroscopic photographs of the underlying surface beneath the doubler indicated that many scratches existed on the faying surface of adjacent fuselage skin.
The scratches were covered with paint. There was evidence that an exterior surface-finish paint had penetrated and cured between the aircraft skin and the doubler — indicating how loose the doubler was. This was further confirmed by SEM evidence of relative movement between opposing sides of the fatigue cracks beneath the doubler. Cracking of the paint around the rivets and filiform corrosion starting there was an indication that the patch had been “working” for some time. The paths of the fatigue cracks were very straight and always followed the track of scratches along a direction parallel to the stringers (i.e., consonant with the 1980 tailscrape direction).
The permanent ‘temporary’ patch doubler
This evidence seemed to indicate that the temporary patch had become the permanent patch. If a permanent patch had been carried out per the Boeing repair guidelines the doubler would have been much larger and the underlying stress-raiser scratches would have been “dressed out.” There would also have been evidence of two sets of rivets (i.e., the temporary and the later permanent doubler’s riveting outside that perimeter). In addition, no paperwork or records of any kind covering the replacement of the temporary patch could be produced.
There was evidence of heavy corrosion beneath the doubler. Corrosion penetrated completely through the skin thickness between S-49L and S-50L. The examination determined that, with a few minor exceptions, all of the fatigue cracks initiated from longitudinal scratches on the faying surface of the skin with the doubler (original exterior surface of skin) and from multiple origins. Cyclic cabin pressure is the prevailing driving force for cracking at this detail, so each striation seen is considered to represent the microscopic crack advancement during one flight cycle of the airplane. The striations can be compared to the whorls of a fingerprint growing outward in the direction of crack propagation.
Undetected beneath the doubler, the cracking was a relentless march toward eventual and inevitable failure. Eddy Current inspection of the accident aircraft could not detect the ineffective 1980 structural repair and the fatigue cracks that were developing under the repair doubler. Additionally, the total time over which the fatigue cracks propagated through the skin thickness could not be determined. Failure to clean the bilges during major overhauls (to enable proper visual examination of the inner skin) was a major contributing factor. CAL’s visual inspections include no magnification or light enhancement.
No record of any permanent repair
The ASC could not obtain any other engineering process records regarding the permanent repair of this specific area, i.e., a complete description of the nature and location of the damage; drawings/diagrams depicting the size and shape of the repair; applicable engineering guidance and maintenance instructions; work cards containing complete description of the steps to remove and repair the damage and the inspector’s signoffs. Also, this particular repair was not listed in the major repair records. CAL told ASC that it considered the accident aircraft’s tailstrike structure repair in 1980 as “not a major repair” and therefore did not require notification to Boeing.
However, this was in conflict with CAL’s own 1975 Maintenance Manual, which defined a major repair as any to the fuselage skin extending more than 6″ in any direction. Nevertheless CAL later claimed, “A permanent repair was conducted on May 23 through 26, 1980.” The Taiwan Civil Aeronautics Administration (CAA) had no aviation safety inspection system in place in 1980, thus there was no regulatory oversight of CAL maintenance. According to the CAA, Taiwan is not an ICAO contracting state. Therefore, ICAO does not assess its aviation safety. In this case, the U.S. Federal Aviation Administration (FAA) conducts the International Aviation Safety Assessment (IASA) on behalf of ICAO. Officially, however, the CAA and the FAA have no obligation to each other.
CAA aging aircraft program
The CAA Principal Maintenance Inspector for CAL stated that the CAA would search the FAA’s or Boeing’s Web site to gather aging aircraft information. As for the Repair Assessment Program (RAP), the CAA originally obtained the information from China Airlines. After the accident, the CAA issued an Airworthiness Directive (AD 2002-09-02, Repair Assessment for Pressurized Fuselages). In addition, the CAA issued an Advisory Circular (AC120-020, Damage Tolerance Assessment of Repairs to Pressurized Fuselages) suggesting operators adopt the FAA-approved repair assessment guidelines for the fuselage pressure boundary to be part of their maintenance program.
The last structural failure breakup of a large passenger jet that was attributable to fatigue cracking in a prior repair was Japan Airlines‘ JAL123 on Aug. 12, 1985. In that accident, 520 died due to improper repairs by Boeing to the aft pressure bulkhead after a landing tailstrike eight years earlier. FAA’s Airworthiness Assurance Working Group (AAWG) conducted two surveys covering some 1,051 repairs on 65 aircraft that had been retired. About 40 percent of the repairs were adequate but the remaining 60 percent required additional supplemental inspections, according to a December 1996 report on the surveys. FAA Advisory Circular 120-73, “Damage Tolerance Assessment of Repairs to Pressurized Fuselages,” was issued on Dec. 14, 2000, and states that no operator could operate nominated aircraft (including Boeing 747-200 models) beyond a certain number of flight cycles or May 25, 2001, whichever occurred later, unless its operations specifications had been revised to reference repair assessment guidelines that were then incorporated in its maintenance program. For Boeing 747s, the flight cycle implementation time was 15,000 cycles. As of its last flight, the China Airlines accident aircraft had accumulated 64,810 flight hours and 21,398 flight cycles.
ASC’s findings
In its Interim Safety Recommendation of March 21, 2003, (“Aircraft Pressure Vessel Structure Repair Alert”) the ASC concluded:
“An improperly treated scratch on the aircraft pressure vessel skin, especially if covered under a repair doubler, could be hidden damage that might develop into fatigue cracking, eventually causing structure failure.”
Its Final Report’s bottom line is that, based on the recordings of CVR and FDR, radar data, the dado panel open-close positions, the wreckage distribution, and the wreckage examinations, the in-flight breakup of CI-611, as it approached its cruising altitude, was highly likely due to a structural failure in the aft lower lobe of the fuselage. It also found that “the permanent repair of the tailstrike was not accomplished in accordance with the Boeing SRM, in that the area of damaged skin in Section 46 was not removed (trimmed) and the repair doubler did not extend sufficiently beyond the damaged area to restore the structural strength.”
The main fatigue crack in combination with the Multiple Site Damage (MSD) were of sufficient magnitude and distribution to facilitate the local linking of the fatigue cracks so as to produce a continuous crack within a two-bay region (40 inches). Analysis further indicated that during the application of normal operational loads the residual strength of the fuselage would be compromised with a continuous crack of 58 inches or longer length. Although the ASC could not determine the length of cracking prior to the accident flight, the ASC believes that the extent of hoop-wise fretting marks found on the doubler, and the regularly spaced striation marks and deformed cladding found on the fracture surface, suggest that a continuous crack of at least 71 inches (about six feet) in length, considered long enough to cause structural separation of the fuselage, was present before the in- flight breakup of the aircraft.
Findings related to risk
The first Corrosion Prevention and Control Program (CPCP) inspection for the accident aircraft was in November 1993, making the second CPCP inspection of the lower lobe fuselage due in November 1997. CAL inspected that area 13 months later than the required four-year interval. In order to fit into the CAL maintenance schedule’s computer control system, CAL estimated the average flight time or flight cycles for each aircraft and scheduled the calendar year based inspection. Reduced aircraft utilization led to the dates of the flight hour inspections being postponed, thus the corresponding CPCP inspection dates were passed. CAL’s oversight and surveillance programs did not detect the missed inspections.
According to maintenance records, starting from November 1997, the accident aircraft had a total of 29 CPCP inspection items that were not accomplished in accordance with the CAL Maint Program and the Boeing 747 Aging Airplane Corrosion Prevention & Control Program. The aircraft had therefore been operated with unresolved safety deficiencies from November 1997 onward. The CPCP scheduling deficiencies in the CAL maintenance inspection practices were not identified by the CAA audits. The determination of the implementation of the maximum flight cycles before the Repair Assessment Program was based primarily on fatigue testing of a production aircraft structure (skin, lap joints, etc.) and did not take into account the variation in standards of repair, maintenance, workmanship and follow-up inspections that exist among air carriers.
Examination of photographs of the item 640 repair doubler on the accident aircraft, which was taken in November 2001 during CAL’s structural patch survey for the Repair Assessment Program, revealed traces of staining on the aft lower lobe fuselage around STA 2100. This was an indication of probable hidden structural damage beneath the doubler. Note the discussion about this in Air Safety Week, June 30, 2003. The staining was likely residue from leftover coffee and tea poured out by flight attendants in the aft galley. There was a communication problem between Boeing and CAL regarding the tail strike repair in 1980. The Boeing FSR would have seen the scratches on the underside of the aircraft. However, the opportunity to provide expert advice on a critical repair appears to have been lost, as there are no records to show that the FSR had a role in providing advice on the permanent repair (Boeing has three FSR’s stationed with CAL at Chang Kai Shek Airport. FSR’s work with the operator only in an advisory capacity).
Conclusion
Like JAL123, China Airlines Flight CI-611 demonstrated that an event of another era can come back to haunt a new generation of operators. The absolute importance of structural integrity in the pressure vessel has now been graphically demonstrated. However, we do not know how many other repaired airframes were found approaching the same parlous state. CAL claimed that the 1980 tailstrike was “not a major repair” but evidently it was. So how can one differentiate between this tailstrike event and the numerous “ramp-rash” events occurring with monotonous regularity? Are they all not major repairs? Any repair to a pressurized hull has the potential for injecting a permanent weakness and stress-raiser.
Like tire blowouts, tailstrikes are almost inevitable. Why don’t manufacturers provide prophylactic tailstrike doublers in the area of striking when the aircraft is manufactured? Are we missing the point that an oleo’d tailwheel or tail shock-bumper should be installed so that airframes don’t have to soak up tailstrikes or scrapes? How will the composite fuselage of the B787 change the face of absorbed shock-loading fatigue for such events? The uninvolved FSR on site with the operator seems to have been a vital link in this accident chain. Nevertheless, the important lessons of CI-611 seem to have been missed by CAL as demonstrated by its (and the CAA’s) aggressive refutation of the ASC’s findings:
It would seem to be a case (for denial-quenching) that strongly reaffirms the need for cockpit cameras — yet it neatly coincides and conflicts with the FAA’s recent finding that cameras are not required to complete the CVR/FDR/CCTV (close circuit television) air safety data triangle.
New 747 AD developments
As a footnote, it is also relevant that the FAA issued Emergency AD 2005-04-51 on Feb. 17, 2005. It covers repetitive inspections for large cracks of the frame structure and skin in the forward fuselage Section 41. This is a well known area for cracking.
The whole concept of “doublers” instead of skin replacement may need re-examination. It is ironic that as we enter the dawn of a new era of composite pressurized hulls with the B787, it is only just becoming apparent how critical slow-growth fatigue damage is for high specification metals. Is there a similar hard lesson awaiting in the field of composite structures? But in the area of present concern for metal fatigue and aging airplanes it would appear that we are only just now scratching the surface. The ASC report is available at http://www.asc.gov.tw/acd_files/CI611_Report_English_VOL_1.pdf and http://www.asc.gov.tw/acd_files/CI611_Report_English_VOL_2.pdf.