Denial of Petition for Reconsideration-(NTSB-AAR-81-8)

In accordance with the Safety Board ^ts rules (49 CFR Part 845), the Safety Board has considered the Petition for Reconsideration of its analysis, findings, and probable cause in the aviation accident involving a Trans World Airline, Inc., (TWA) Boeing 727-31, N84OTW, near Saginaw, Michigan, on April 4, 1979.

On June 9, 1981, the Safety Board determined that the accident occurred after the airplane entered an uncontrolled maneuver at 39,000 feet pressure altitude while near Saginaw, Michigan. The airplane descended to about 5,000 feet in 63 seconds before the flightcrew regained control and made an emergency landing at Metropolitan Airport, Detroit, Michigan. The Safety Board’s analysis of the evidence indicated that the uncontrolled maneuver began after the leading edge slats on the right wing were retracted and the airplane’s No. 7 leading edge slat remained in the extended or a partially extended position. The isolated extension of the No. 7 leading edge slat resulted in a roll which led to a reduction in the airplane’s lateral control margin to zero or less. The loss of lateral control was the result of a combination of the extended slat, mach number, angle of attack, and sideslip. The airplane entered a descending right spiral, and control of the airplane was regained only after the No. 7 slat was torn from the airplane.

The Safety Board’s investigation revealed no evidence of irregularity, malfunction, or failure of the airplane’s flight control, autopilot, hydraulic, or flap systems that might have caused or contributed to a lateral control problem. Further, there was no evidence of any combination of failures or malfunctions in the airplane’s flight control system that would have caused an unscheduled extension of the No. 7 leading edge slat by itself.

In adopting its report, the Safety Board determined that the probable cause of the accident was the isolation of the No. 7 leading edge slat in the full or a partially extended position after an extension of the Nos. 2, 3, 6, and 7 leading edge slats and the subsequent retraction of the Nos. 2, 3, and 6 slats, and the captain’s untimely flight control inputs to counter the roll resulting from the slat asymmetry. Contributing to the cause was a preexisting misalignment of the No. 7 slat which, when combined With the cruise condition airloads, interfered with the retraction of that slat. After eliminating all probable individual or combined mechanical failures or malfunctions which could lead to slat extension, the Safety Board determined that the extension of the slats was the result of the flightcrew’s manipulation of the flap/slat controls. Contributing to the captain’s untimely use of the flight controls was distraction due probably to his efforts to rectify the source of the control problem. 1/

The following discussion addresses allegedly new evidence the Air Line Pilots Association (Petitioner) submitted to the Safety Board which Petitioner states negates the portion of the probable cause related to the flightcrew’s manipulation of the flap/slat controls as the reason for the extension of the slats. Petitioner contends that it is likely there was a mechanical failure of a Ronson slat actuator and that the Boeing Commercial Airplane Company had knowledge of such mechanical failures yet did not inform the Safety Board. Further, Petitioner contends that the problems highlighted in a Federal Aviation Administration Airworthiness Directive regarding a slat actuator manufactured by Decoto, which is interchangeable with the Ronson actuator, were encountered with the Ronson actuator on the TWA flight.

In support of the petition, Petitioner provided a coordination sheet, dated April 26, 1979, which was prepared by the Boeing Commercial Airplane Company, as proof that there were six cases of cracking in the piston seal ring groove of the Ronson actuator. The Safety Board investigated the development of the coordination sheet and determined that it was prepared by a Boeing staff member after the TWA accident.

The coordination sheet by its own terms was an information brief prepared in response to an April 19, 1979, memorandum which requested the status of certain items. The coordination sheet was not a technical paper, nor was it coordinated with the technical staff of Boeing or product Integrity personnel responsible for monitoring Service Difficulty Reports and service reports for the Ronson actuator. The coordination sheet was completed and returned to the Boeing technical staff as a historical document for its use in the subject accident investigation. The basis for the paragraph describing six cracked actuators was established to have been a telex, dated October 6, 1975, from a Boeing field representative assigned to Lufthansa German Airlines. This telex reported that Lufthansa had “scrapped six Ronson P/N 101095-5 actuator rods due to cracks in the piston seal groove.” (See attachment 1.)

On October 7, 1975, the Boeing field service representative who had sent the October 6, 1975, telex sent a telex correcting the original telex. (See attachment 2.) The second telex stated that an incorrect part number had been quoted in the first telex. The second telex specifically reported that, “All six cracked pistons discussed in refs /A/ and /B/ were part of LE edge flap actuators, P/N 65-17818-14 and —20, and not leading edge slat actuators. Thus comments in reference /B/ on SB-27—29 do not apply.” The October 7, 1975, telex apparently was overlooked when the April 26, 1979, coordination sheet was prepared, with the result that incorrect information was included in the coordination sheet.

A Boeing spokesman stated that the Boeing technical staff and the Boeing accident investigation staff were aware of the error in the coordination sheet, and that an incorrect part number had been reported. They were aware of the October 7, 1975 telex, which corrected the earlier report that there had been failures in the Ronson slat actuator. The coordination sheet was not corrected immediately since it was an information document. The coordination sheet was revised on June, 2, 1982, after it was produced in litigation and again on November 12, 1982. In any event, however, the best evidence regarding Boeing’s records as to whether such failures had occurred is the two telexes.

During the accident investigation, the Safety Board requested from the Boeing Commercial Airplane Company all records concerning fractures in the Ronson actuator. It responded that there had been no transverse fractures of the main piston in the operating history of the Boeing 727, and this was the basis for the statement in the analysis section of the accident report that “. . . in 16 years of service history and over 36 million flight hours, such fractures of a slat actuator piston have never occurred.” The evidence is, notwithstanding the statement to the contrary in the coordination sheet, that the report was correct and that it continues to be correct. The Board concludes also that the Boeing Commercial Airplane Company did not withhold any pertinent information concerning mechanical deficiencies in the Ronson actuator piston.

Petitioner submitted a letter to the Safety Board on June 24, 1983, wherein the basis for the original petition was restated. Additionally, three other issues were raised in the letter. The first was a report of. an uncommanded extension of a No. 7 slat on a Boeing 727-100 on August 28, 1982; the second involved problems with the Decoto slat actuator piston and the controllability of the Boeing 727 with the No. 7 slat extended; and the third issue was the validity of using flight data recorder (FDR) data to calculate drag.

The Safety Board investigated the August 28, 1982, incident of the uncommanded extension of a No. 7 slat on a Boeing 727-100. The Safety Board obtained the maintenance records on the airplane, and interviewed the captain of the flight. The maintenance records indicate that a problem with the No. 7 slat was first entered in the airplane logbook on August 25, 1982. The entry stated: “No. 7 slat sticks on retraction and will not come up unless aircraft slowed to low speed.” The slat actuator was replaced and was operationally checked.

On August 28, 1982, the incident referenced in petitioner’s supplemental letter occurred. The log book entry stated: “No. 7 slat came out in flight going through 4,000 ft. with trailing flaps up.” The No. 7 slat actuator was again replaced and the operational check indicated the system was functional. A teardown of two of the three actuators removed from the airplane, including the actuator which was removed on August 28, 1982, indicated that there were no discrepancies of the actuator. The actuator teardowns were performed by Northwest Airlines.

The airline stated that the No. 7 slat problem on the airplane was written up in the . logbook a total of six times in 9 days. Three different slat actuators were installed in the airplane during that period. The company concluded that the problem was not with the actuator, and subsequently replaced the entire No. 7 slat system. A spokesman for the company stated that the replacement of the system resolved the problem, since no further incidents developed with the slat.

The interviews with the captain and the first officer revealed that the incident occurred after takeoff. While climbing through 2,500 feet, the flaps were placed in the up position and the airplane was accelerated to 250 knots. As the airplane accelerated and climbed through 4,000 feet, a right roll and airframe buffet began. According to the flightcrew, at that time the flap/slat indicator light on the instrument panel indicated an in-transit condition. The flight engineer’s leading edge flap/slat test panel indicated that the No. 7 slat was down and locked. The No. 7 slat remained extended, and the flight was continued from Tulsa, Oklahoma, to Dayton, Ohio, where the airplane was landed Without incident .

The Safety Board’s investigation of this incident indicates that the incident on August 28, 1982, involving the No. 7 slat was not an isolated event, but one of several similar incidents involving the airplane. Additionally, the problem always manifested itself with the No. 7 slat not retracting after takeoff, or retracting in an abnormally slow manner. There were no instances of an independent extension of the No. 7 slat. The Safety Board believes, therefore, that the August 28, 1982, incident cited by Petitioner involved the failure of the No. 7 slat to retract after takeoff, and not from an uncommanded slat extension after all the slats had been retracted. This conclusion is supported. by the slat-problem history of the particular airplane, the phase of flight in which the incident occurred, and the absence of cracks, fractures, or any other defects in the actuators removed from the airplane.

With regard to Petitioner’s concerns about the information provided in attachment 5 to their letter of June 24, 1983, the Safety Board believes that the corrective action required by Airworthiness Directive 82-22-01 to modify the Decoto—manufactured Boeing 727 leading edge slat actuators was the appropriate remedy to preclude any possibility of an unscheduled extension of a single leading edge slat that might cause a serious lateral control problem in the B-727 airplane. The “cracks” in the Decoto actuators originated in the extend lock key grooves in the actuator piston. Cracks were found in the fillet radii of these grooves in several actuator pistons removed from service for overhaul. Tests of three production pistons showed that cracks in the fillet radii were detectable by magnaflux inspection at 100,000 cycles and that performance degradation was apparent at 127,000, 171,000, and 191,000 cycles, respectively. The piston with 127,000 cycles fractured transversely through the extend lock key grooves at 142,000 cycles, which is the equivalent of 71,000 flight hours at an average of 1 hour per flight. Since the recommended actuator overhaul limit is 18,000 hours, it appears that a complete fracture of the piston in the extend lock key grooves of an unmodified Decoto piston is not likely before detection either at overhaul or through performance degradation. Further, there is no evidence at all that a similar problem has arisen with the Ronson actuator.

Moreover, a fracture of the actuator piston, caused by the problem identified with the Decoto actuators, did not occur on the accident airplane (N84OTW), because such a fracture would have left more than two—thirds of the piston locked in the actuator cylinder. In fact none of the piston was found in the aft portion of the cylinder that remained on the airplane. Therefore, the Safety Board does not believe that the problem associated with the Decoto actuators is germane to the accident involving N84OTW.

The issue raised in Petitioner’s June 24, 1983, submission concerning the validity of using FDR data to calculate drag also was addressed in the August 25, 1983, supplement to the original petition. We have chosen to address the questions concerning FDR data and drag calculations as they were framed in the latter letter.

The first issue raised by Petitioner in the August 25, 1983, supplement involved a restatement of the previous submissions that an individual slat on the accident airplane could extend as a “result of a single failure. This issue is addressed in the preceding discussion of the alleged six cracked Ronson actuators referenced in the Boeing coordination sheet, and the August 28, 1982, uncommanded No. 7 slat extension on a Boeing 727-100. A second part of this issue was that the physical evidence resulting from the extension of a single slat would be consistent with the physical evidence of the accident investigation. This subject is discussed. as part of the two final issues raised in the August 25, 1983, supplement which are alleged errors in the Safety Board’s analysis of (1) airspeed trace comparison and drag calculations, and (2) vertical acceleration frequency analysis. Finally, Petitioner made several statements in the “Summary” section of the August 25, 1983, supplement which warrant a response from the Safety Board.

Petitioner challenges the methods used by the Safety Board for, airspeed comparison and drag calculations. Petitioner suggests that any similarity between airspeed traces for the accident and flight test airplanes is coincidental, and that other configurations and conditions would produce similar airspeed traces. Petitioner rejects the validity of the Safety Board’s drag calculations and the resultant conclusion because the point selected for drag measurement occurred during a “dynamically unstable c.ondition.”

The method chosen by the Safety Board to calculate the increase in drag for N84OTW and the flight tests flown on October 2, 1980, was derived for quasi-steady-state maneuvers (climbs, descents, accelerations, and decelerations) flown at or near ig with a fixed engine throttle position. Drag is determined from the following equation derived from a force balance diagram:

The equation is derived by assuming that the total energy of the airplane consists of potential energy in the form of altitude and kinetic energy in the form of speed, or some combination thereof. This method was used for the drag calculations because of the maneuvers flown, the parameters available on the FDR, and the timeframes available to make the calculations.

If altitude is held constant by an auto ₁gilot altitude hold mode, the change in drag (AD) is reflected by the change in airspeed (ΔV/Δt). Therefore, a direct relationship exists between an increase in drag and an associated decrease in airspeed if the net thrust and altitude remain constant, as was the case with the accident airplane.

The drag calculation for the accident flight was made at FDR time 23:32, which was 11 seconds after initiation of the leading edge slat extension (9 seconds after the onset of buffet). The decrease in airspeed was 0.5 KIAS/second and the rate of descent was7.5 feet/second. The drag increase indicated by the deceleration was calculated as7,030 lbs, and the drag increase resulting from. the rate of descent was 1,230 lbs, for atotal drag increase of 8,260 lbs. The total increase in airplane drag coefficient (ΔC_D) wascalculated to be 0.0283, indicating that the airplane drag had essentially doubled from itsvalue before the onset of buffet.

Of the 21 flight conditions/configurations tested during the Boeing flight test on October 2, 1980, Boeing test conditions 011 and 012 proved most representative of the performance changes accompanying the configuration change that occurred on the accident airplane — these were, extension of 2º of trailing edge flaps followed by extension of the Nos. 2, 3, 6, and 7 leading edge slats. For purposes of a comparison with drag calculations for the accident airplane, a calculation of the drag increase for Boeing flight test condition 011 was made at FDR time 3:46, about 12 seconds after initiation of leading edge slat extension, where the rate of decrease in airspeed was 0.5 KlAS/second and the rate of descent was 1.5 ft/sec. The drag increase indicated by the deceleration was calculated as 7,100 lbs, and the drag increase indicated by the rate of descent as 270 lbs, for a total drag increase of 7,370 lbs. The total - increase in (ΔC_D) was calculated to be 0.0274, within 3 percent of the value calculated for the accident airplane .

Flight tests could not be conducted for the condition where the No. 7 slat was extended individually. However, a ΔC_D was calculated for this condition. The calculation was made by assuming that the drag associated with extension of the No. 7 slat alone would be about 1/4 of the drag associated with extension of slats Nos. 2, 3, 6, and 7. Therefore, the ΔC_D assumed for the extension of the No. 7 slat alone was 0.0274/4 or 0.0069. The drag associated with full autopilot lateral control authority plus the associated rudder deflection and sideslip was 0.0038. This value was supplied by the Boeing Qommercial Airplane Company based on flight test data. Consequently, the total ΔC_D for a condition with only the No. 7 slat extended was 0.0107. The deceleration which the airplane would experience for this configuration change, plus the control deflection and sideslip at the accident airplane conditions (including 450 ft/min rate of descent) was calculated to be 0.14 KIAS/second, or 28 percent of the deceleration which was indicated on the accident airplane FDR trace. As a result, the Safety Board concluded that extension of the No. 7 slat alone would not have produced the changes in airspeed and altitude indicated on the accident airplane’s FDR, since 72 percent of the known deceleration would not be accounted for.

Flight tests conducted during the accident investigation indicated that the increase in drag coefficient for the accident airplane was within 3 percent of the increase in the drag coefficient for the flight test airplane when the latter was configured with its Nos. 2, 3, 6, and 7 slats extended. Additionally, the computed increase in drag coefficient for extension of the No. 7 slat alone (plus increases related to lateral control deflection rudder deflection, and sideslip—at the maximum authority of the autopilot to produce) was less than half of the increased drag coefficient of either the accident airplane or the flight test airplane. Consequently, the Safety Board’s accident report concluded that the accident airplane’s increase in drag coefficienf was the result of a configuration identical to that of the flight test airplane, i.e., the Nos. 2, 3, 6, and 7 slats were extended. This conclusion was stated on page 27 of the accident report:

Nos. 2, 3, 6, and 7 slats during flight tests compares almost exactly with the airspeed decrease experienced by Flight 841 following initial oscillation of its g-trace, which under constant thrust and 1.O-g flight conditions can only be attributed to similar drag producing configurations, the Safety Board concludes that the Nos. 2, 3, 6, and 7 slats were extended as a consequence of flightcrew action.

The foregoing calculations and analyses show that in its analysis of the accident the Safety Board did not make incorrect assumptions about the relationship between the production of drag and airspeed changes. The Safety Board’s conclusion that the configuration of the two airplanes was identical is based on a recognized and valid analytical method of calculating drag changes, and on data of sufficient accuracy to make the necessary calculations with a high degree of confidence.

Further, although similarities between airspeed traces can result from different . configurations, there is absolutely no evidence to indicate that other drag—producing devices, such as landing gear, speed brakes, a substantial amount of trailing edge wing flap, or any combination thereof, were extended on the accident airplane during initial entry into its spiral dive. Finally, when other drag-producing devices/conditions related to an extension of the No. 7 slat alone are accounted for, the increase in drag coefficient is less than half (about 38 percent) of the increase experienced by the accident airplane during the timeframe of interest.

In addition to disputing the Safety Board’s calculations concerning the relationship between drag production and airspeed changes of similarly configured airplanes, Petitioner challenged the conclusions reached by the Safety Board that were based on the comparison of two flight tests (011 and 012). Petitioner provided graphic data from which they concluded that the comparison of flight test data between the two flight tests does not support the Safety Board’s assumption that similar speed decreases imply identical airplane configuration. Petitioner states that the Safety Board’s assumption was not valid because the data used related to a time when the airplane was not in a stabilized flight condition .

The Safety Board’s analysis of Petitioner’s graphic data indicates that Petitioner failed to account for the difference in airplane response to the extension of the Nos. 2, 3, 6, and 7 slats with the autopilot engaged (test condition 011) and without the autopilot engaged (test condition 012). As a result, Petitioner’s graphic data are an inaccurate comparison of the changes in the two airspeed traces. Condition 011 was conducted with the autopilot altitude-hold mode engaged, and the autopilot was to be used to maintain a constant altitude using the pitch axis control. However, the autopilot had not been properly set — the test switch had been inadvertently left in the HOLD position — and it did not perform as intended in the test protocol. The airplane pitched up slightly, but then returned to the trim attitude. The pitch attitude and angle of attack osciflated slightly above and below the trim values. Altitude oscillated slightly, about 40 feet above and below the selected altitude.

Condition 012 was conducted with the altitude hold disengaged so that the airplane’s pitch attitude, angle—of-attack, and altitude were not controlled. As a result, after all the leading edge slats were extended, the airplane pitched up abruptly 6.5 º and climbed. The large pitchup and climb resulted in a greater deceleration than occurred during condition 011. However, following the pitch-up, when both the deceleration and rate of climb were used in the equation, the total increase in airplane drag coefficient (ΔC _D)was calculated to be 0.0301, which was within 6 percent of the value calculated for the accident airplane.

Consequently, the comparison between the flight test condition 011 FDR and the accident airplane FDR was valid because the autopilot was engaged in altitude hold for both cases. Also, the equation was correct since it was derived for quasi-steady-state conditions where the airspeed and altitude were changing.

Petitioner is correct in asserting that the Boeing Commercial Airplane Company did not reaccomplish the flight tests to verify the data produced in test condition 011 after the SP 150 autopilot test switch was found mispositioned. However, simulations were accomplished to resolve satisfactorily the different response characteristics of the SP 50 and SP 150 autopilots, and Petitioner was provided with the results of the simulations.

Petitioner posed an additional question concerning drag calculations having concluded that the accident airplane was in a 10º bank at 23:29 FDR time and that thereafter the bank angle increased rapidly. Petitioner’s question was: What caused the 100 bank initially if not an asymmetric slat condition?

The answer to Petitioner’s question is related to the fact that each slat extends independently . and at a slightly different rate. Consequently, a small amount of asymmetry usually occurs while the slats are being extended. When the slats are extended within the normal flight envelope, the asymmetry causes no problems. However, when the slats ar.e extended outside the normal envelope at high Mach numbers, as during the flight tests and the accident sequence, a slight rolloff can occur. Flight test conditions 011 and 012 where the Nos. 2, 3, 6, and 7 slats were extended confirm this. The flight test airplane rolled about 5º right wing down even though the autopilot heading hold mode for condition 011 and the test pilot in test condition 012 were attempting to maintain wings level. . Therefore, contrary to Petitioner’s assertion that the 100 bank at 28:29 FDR time supports a conclusion that a single slat extended, the Board believes that this evidence strengthens the conclusion that a single slat did not extend, because based on the Boeing simulations for extension of the No. 7 slat alone,, with the autopilot engaged, the bank angle at FDR time 23:29 would have been at least 50º right wing down. Furthermore, the FDR data indicate that the roll rate did not start to increase rapidly until after 23:30, which is compatible with the increasing slat asymmetry that developed after the Nos. 2, 3, and 6 slats began to retract.

Petitioner further challenged the validity of the Safety Board’s drag calculations and airspeed trace comparisons by stating that “the FDR is a crude instrument at best and was never intended for the purpose of precise drag measurements, especially under dynamically unstable conditions.” The petition also states that, due to lag in FDR pitotstatic systems, accurate readings are not possible under dynamic flight conditions, so no comparison could be made between the accident airplane and the test airplane.

The purpose of the FDR data comparisons—airspeed and rate of descent comparisons and the associated drag calculations—was to explore the magnitude of the changes caused by the extension of four leading edge slats and the changes caused by the extension of one slat by itself. These were established to be substantially different, the former being very large and the latter relatively smaller. The Safety Board believes that its objective was accomplished and that the FDR data can be used for this Purpose with a high degree of confidence .

The effect of lag in the FDR pitot-static systems was minimized by using similar B-727-100 airplanes which incorporated the same pitot—static systems and the same make/model FDR’s. The Safety Board’s drag calculations considered the possibility of lag by making the calculations at the point in time where the rate of change in airspeed and altitude reached constant values. This point occurred at essentially the same elapsed time for both the accident and flight test airplanes. Since the calculations for both the accident and flight test airplanes showed that the drag had doubled within a short time after buffet onset and that the value was far in excess of what would have resulted by the individual extension of the No. 7 slat, the effects of lag, as well as any other source of small inaccuracies within the FDR, would not have been significant.

In response to Petitioner’s challenges of the accuracy of the FDR, the Safety Board points out .that, throughout the investigation, the .Air Line Pilots Association (ALPA) repeatedly requested that flight tests be conducted to acquire FDR data for comparison with the FDR data from the accident airplane (see the letter dated November 9, 1979, from ALPA’s President and the letter dated March 31, 1980, from ALPA’s accident coordinator). The Safety Board also sought flight tests, and it prevailed upon the Boeing Commercial Airplane Company to conduct flight tests. The Safety Board believes that at this point Petitioner is not in a .position to assert that the data cannot be compared, and that in essence, the flight tests were not necessary.

Petitioner contends that the analysis of vertical acceleration frequency is unsubstantiated and is not a valid basis for the conclusion of the accident report “that the high frequency g-trace oscillations associated with airframe buffet on the flight test aircraft’s FDR and the accident . aircraft’s FDR were identical at a frequency of . 6 cycles/second and an amplitude of +0.05g.” Petitioner states that the conclusion is invalid because the frequency response capability of the FDR was exceeded and because a frequency of 6 cycles/second could be induced using other configurations.

The Safety Board’s analysis of buffet characteristics using the vertical acceleration trace of the FDR involved a comparison of the buffet amplitude as well as the buffet frequency. At the beginning of the analysis, the Safety Board acknowledged there were limitations to the use of the FDR frequency readings. Nevertheless, when the flight test vertical acceleration FDR traces were compared with similar traces from the sensitive flight test instruments, the buffet amplitude recorded by the test flight airplane FDR was found to correlate closely with the test flight instrument—produced data. The FDR traces correlated even when the frequency limit of 6 cycles/second was exceeded. This correlation was confirmed by FDR frequency response tests conducted by TWA.

The configuration changes made during the flight ‘tests (trailing edge flaps,, leading edge flaps, slats, and spoiler extensions) indicated that each change produced a significantly different buffet as determined by the amplitude of the vertical acceleration oscillations. The only airframe buffet identical in both amplitude and frequency to that recorded on the accident airplane’s FDR resiulted from the extension of the Nos. 2, 3, 6, and 7 slats.

The vertical acceleration was essentially equal to the load factor (lift/weight) at the cruise angle—of—attack before the extension of the Nos. 2, 3, 6, and 7 leading edge slats. However, in the first 2 seconds following the extension of the leading edge slats, the load factor decreased to about 0.97 for flight test condition 011. The load factor for the same timeframe for the accident airplane was 0.96. These data translate to a loss of lift of about 4,000 to 5,000 lbs in a short time, since extension of the leading edge slats results in a loss of lift under the specific flight conditions. The Safety Board concluded that, since the extension of the No. 7 slat alone could not produce the same rapid loss of lift as that produced by the extension of four slats, the accident airplane had four slats extended.

slats are compared with the accident airplane’s traces, they indicate virtually identical buffet characteristics, increases in drag, and loss of lift. AU of these conditions could not have occurred simultaneously as the result of extending any other aerodynamic device(s). Moreover, the bank angles calculated from the accident airplane’s FDR traces correspond with the extension of four slats followed by retraction of three slats; the FDR traces do not correspond with bank angles associated with the extension of the No. 7 slat alone.

Accordingly, the Safety Board finds Petitioner’s challenges to its airplane drag calculations and FDR trace correlations without merit. Contrary to Petitioner?s assertion, FDR information has been used extensively to calculate aerodynamic performance under dynamic conditions, such as wind shear. Consequently, there is nothing unique about calculating changes in drag over a given period of time when thrust is a known constant.

Petitioner states in the summary section of the August 25, 1983, supplement that all of the Safety Board’s investigative efforts were designed to show that a configuration (Nos. 2, 3, 6, and 7 slats extended) existed at the initiation of the accident sequence. To the contrary, the greatest portion of the Safety Bos.rd’s investigative efforts were expended in an attempt to identify any condition which might have permitted or caused the independent extension of a single slat. All factors singularly and in combination that possibly could have caused the independent extension of the No. 7 slat were considered and systematically eliminated based on systems tests, system fault analyses, aerodynamic data, and the physical evidence. The one possible failure mode—transverse fracture of the slat actuator piston in the retract lock key grooves--was eliminated as extremely improbable because of the high safety margin associated with the piston design, service history, and flight test data. In fact, it was not until late 1980—18 months after the accident—that flight test data established conclusively that there had been an initial extension of four slats rather than a single slat.

Petitioner also asserted in the summary section of the August 25, 1983, letter that it had submitted to the Safety Board a hypothetical piston failure mode which subsequently became the exact failure mode that caused the FAA to issue an Airworthiness Directive against the Decoto—manufactured slat actuators. . Petitioner’s assertion is not correct. The piston failure mode hypothesized by Petitioner involved a transverse fracture of the piston in the retract lock key grooves, whereas the failure mode that prompted the Airworthiness Directive involved potential fracture in the fillet radii of the extend lock key grooves. The two grooves a.re on opposite ends of the piston and are subject to different forces and stresses. Identical stress conditions do not occur at each end of the piston. Consequently, the area through the retract lock key grooves is not subjected to similar tensile stresses that might produce cracking or fracture.

Finally, Petitioner asserted in the summary section that the Board labored under a preconception that the flightcrew was attempting to obtain an airplane configuration with 2˚ of trailing edge flaps extended without extension of the Nos. 2, 3, 6, and 7 slats even though there was no reason for It to do so, because the flight tests (conducted 18 months after the accident) demonstrated that such a configuration did not produce an increase in performance. Such an assertion presumes that the flightcrew either was aware of the results of the flight tests conducted 18 months after the accident or that it had preaccident experience with such a configuration. Be that as it may, the Safety Board did not adopt any scenario that might explain why the flightcrew might desire such a configuration, although the Board was aware of the widespread belief in the pilot community at that time that such a configuration would enhance performance, and ofother reasons that might motivate a flightcrew to configure the airplane in such a manner. Such reasons include a reduction in trim angle of attack and pitch attitude as demonstrated in flight tests. Consequently, the lack of an increase in performance . demonstrated during the flight tests is irrelevant even if it is accepted as providing compelling support for an assertion that there was no motivation for the flightcrew to configure the airplane with 2º of trailing edge flap extended. .

As set forth in 49 CFR 845.41, changes in reports in response to requests for reconsideration of probable cause will be made only if they are based on new evidence or on a showing that the Board’s findings as to the facts, conditions, and circumstances of the accident are erroneous. Petitioner has provided no new evidence in either its petition or the supplements thereto to establish a valid possibility of a mechanical failure of the Ronson actuator piston on the accident airplane. Finally, Petitioner has provided no information to show that the Board’s findings as to the facts, conditions, or circumstances of the accident are erroneous, or that the Board’s conclusions regarding the probable cause of the accident are incorrect.