F rem Page A3-1 , Page A3-2 §91 .609 , PageA3-3 §121.343 , or Page A3-5 § 121. 344a NO NO YES YES YES YES NO YES NO §135.152(a) , Appdx B (airplane) 18 parameters or c (rotorcraft) 23 arameters NO §135.1520) , Appdx F 88 parameters § 135.152(i) , Appdx F 57 parameters NO §135.152(b)(3) , Appdx o (airplane) 29 + 17 optional parameters or §135.152(b)(2) , 22 parameters per Appdx B (airplane) E (rotorcraft) 25 parameters or C (rotorcraft) NO §135.152(b)(1) , 11 parameters per Appdx B (airplane) or C (rotorcraft) 8/17/10 AC 20-141B Appendix 3 FIGURE 3-8. PART 135 FLIGHT RECORDER PARAMETER REQUIREMENTS § 135.152 A3-8
Appendix 4 APPENDIX 4. TYPICAL REASONABLENESS AND QUALITY CHECK
INSTRUCTIONS
1. General. The operator must accomplish a reasonableness and quality check of the recorded flight data to ascertain that the data is being recorded correctly and that noise and data dropouts do not interfere with the ability to interpret the recorded data. The check may be performed using data that is in electronic format or hardcopy. If a hardcopy printout is used, data traces should also be available. The check must be perf ormed using data that has been extracted in EU because octal, BCD, or hexadecimal coded data do not provide the analyst a clear understanding of how the parameters are varying and how they are correlated to each other.
2. Checklist. The analyst must use a checklist to confirm that all necessary checks have been accomplished. The checklist must refer the analyst to troubleshooting or repair procedures if a suspect parameter is identified.
3. Flight segment selection. The data to be used by the analyst should be extracted from both the takeoff and the landing phases of flight. Duri ng the cruise segment of a flight the parameters remain steady, and therefore movement of related parameters cannot be correlated. The takeoff and landing segments of flight provide the analyst an opportunity to observe data that is changing as the aircraft climbs, descends, accelerates, decelerates, and banks or turns.
Furthermore, many parameters not exercised during the cruise segment are exercised during the takeoff and landing segments. We recommend the analyst conduct the review of parameters over several flight segments.
4. Sign conventions. Each aircraft has a pre-established sign convention for the direction of movement of its flight control surfaces. It is imperative the analyst be able to confirm proper direction of movement and not just verify movement. Therefore, the sign convention should be included in the checklist or the analyst should review the assigned sign conventions before beginning the check.
5. Failed parameters. The analyst should examine the extracted data to determine if parameters normally vary in flight, e.g., flight controls, flight control surface positions, and heading, are indeed varying. Values at th eir “maximum limit” or unvarying indicate an inoperative sensor or other failure. Accelerometers tend to fail at the “maximum limit” position.
If the accelerometer trace is not moving during all segments of the flight check, check to see if it indicates maximum or minimum acceleration. An accelerometer failure indicating a mid-point value is uncommon.
6. Correlation to other parameters. The reasonableness check should include a check of the correlation between parameters that depend upon each other. For example, if roll increases, a turn is indicated and heading should begin to change soon after the increase is detected. Also, aileron position and control wheel position should have changed immediately before the roll initiates. There may even be a variation in late ral acceleration. Again, it is important the analyst confirm movement is indicated in the proper direction according to the aircraft sign convention.
Table 4-1 is provided as an aid to prepare a reasonableness checklist. It summarizes parameters in a 34-parameter DFDRS that may be expected to interact. A check mark (9) in a block indicates the parameter identified in the row and the parameter identified in the column are A4-1
Appendix 4 interdependent at some time during takeoff and climb or approach and landing. Therefore, the movement of one parameter should cause or be caused by movement in the other. The following examples show how to use Table 4-1 in developing a reasonableness checklist for each parameter.
a. Thrust reverser position reasonableness and quality check. In Table 4-1, the column labeled Thrust Reverser Positions (number 22) contains check marks in the rows labeled Airspeed, Engine Thrust, Longitudinal Acceleration, Autopilot AFCS Mode, and Air/Ground Sensing. The normal expectation is for the thrust reverser to deploy during rollout after landing.
Thus, the following checklist might be developed using the parameters identified by a check mark.
(1) Examine the thrust reverser in-transit data and the thrust reverser deployed data to determine they indicate in-transit only for a short period during the landing roll and deployed at the end of the in-transit period. The data shou ld indicate in-transit and the deployed discrete should indicate stowed near the end of the landing roll.
(2) Examine the engine thrust data during the in-transit period and immediately after the deployed indication. During the in -transit period, engine thrust should have decreased to near zero. Immediately after the thrust reverser deployed indication, the engine thrust should have increased to near the maximum indication.
(3) Examine the airspeed and longitudinal thrust data. These two parameters should be decreasing during the in-transit period and should dramatically decrease immediately after the deployed indication.
(4) Examine the autopilot mode discrete and the air/ground sensing discrete. The autopilot mode discrete should indicate the autopilot has disengaged and the air/ground sensing switch discrete should indicate that the aircraft is on the ground.
(5) Examine the remaining data for the thrust reverser discretes to ascertain that no in-transit or deployed indications appear. If intermittent indications appear, determine they are within allowable values and do not have sufficient duration to be interpreted as an actual deployment.
b. Lateral control surface position reasonableness and quality check. The column labeled Lateral Control Surface Positions contains check marks in the rows labeled Heading, Roll Attitude, Lateral Control Position, and Localizer Deviation. The lateral control surfaces are typically ailerons used in establishing the aircraft in a turn and returning the aircraft to straight flight from a turn. The lateral control surface position data may be checked along with the lateral control position data. These checks may be accomplished during the approach and landing segment.
(1) Examine the lateral control surface position trace for deviations during the initial approach segment. A large deviation would normally indicate the aircraft turning onto final approach heading. Check the la teral control position and roll attitude make a large change at the same time.
A4-2
34. Ground Speed '> '> '>I'> '> '>
33. Hydraulic pressure low
32. Angle of attack '> '> '> -- 31. Air/ground sensing '>I'> '> '> '> '> '> ,....+-
30. Master warning
29. Marker beacon I I'> I I I I I I I I I I I I I I I I I I I I I I I I'>
28. Glideslope deviation '> 27. Locali2er deviation
26. Radio altimeter '> '>I'> '> '> '>l'>I'> '>
25. Autopilot, AFCS modes '> '>
24. OAT/TAT '>
23. Ground spoiler position I I I '> I I I I I I I I'>
22. Thrust reverser positions I I I '> I I I I I I '> I I '> '>
21. Leading edge flaps/slats '>
20. Trailing edge flaps
19. Pitch trim surface positions I I I '> I I I'> I I I I I'> I '> '>
18. Lateralaccelerationl I I I l'>I l'>I I I I'>
17. Yaw control surface positions '>
16. Lateral control surface positions
15. Pitch control surface positions
14. Yaw control positions '>
13. Later al control positions
12. Pitch control positions
11. Longitudinal acceleration
10. Autopilot engagement '>
9. Engine thrust '>
8. Manual microphone keying '>
7. Roll attitude
6. Pitch attitude
5. Vertical acceleration '>
4. Heading '>
3. Airspeed '>
2. Pressure altitude
1. Time '> '> '>I'> '> '> '>I'> '> '> '> '> '> '>l'>I'> '> '> '> '> '> '>I'> '> '>l'>I'> '> '> '> '> '> '> '> "' "'" '> '> " 0 "' 0 ·.::::;: C:
C: :~ -~ :~ ·-O lfl .. 0 .. "Q. C: ,:,:, .,, ,i, a 1G, ~e~ ffi l t-g~ f ~gge ~ "'""" ~~a 2502 ~ ~t£1•; ~8:~8 c"'"'"'lei-Z3SZ3 '> '> '> '> '> '> '> '> '> '> '>l'>l'>l'>l'>I'> '> '> '> '> '> '> '> '> '> '> '>l'>l'>l'>I'> '> '>I'> '> '>I'> " "' 0 ..
E 6 ._g a:-t3;"i·5 ~~ ~~-~~ ! ..... I-.w -~ ~ ~ .... "'Cl <C _Q ii; ,i, 0 . ., """I " C C - C !:: ·-:.£! iii = ·-Ill ::i I-0 .Q .,, ,i, '> '> '> '> '> '> '> '> '> '> 3 .Q '> '> '> '> '> '> ~~ ~ . :, " ... "' ii·~ -0: it i ~I !II§ 0 ~ ~ 1:: t:; 2 .!! ~ § .,, .,, ...... C: i::, 0 .. ~ ~-~ i] ~ ~~111ir£ ~ ~;~~; c~ ........ ·ci "NM~~~~oo~ci ·NM~ ~~NM~~~~ 00 ~ ~ = ~ 0 ~ ~ ~ ~1~1~1NINININININININININIMIMIMIMIM ·-"!i:~::!-a::J~8~ ~~C!lOC~....1C!l ~~IC I C!i 8/17/10 AC 20-141B Appendix 4 Check to determine heading begins to change immediately after the lateral control (2) surface position begins to change. Heading should continue to change after the lateral control surface position returns to the zero or null value. The heading data should begin to change at a lower rate when the lateral control surface position data moves in the opposite direction. When the lateral control position is returned to zero or null, the heading data should again be constant.
Check the localizer deviation for changes. (3) Small lateral control surface position and lateral control position data changes should accompany deviations from the localizer and returns to the localizer course null.
Check the lateral control surface position data to determine there are no data dropouts (4) and there is no noise in the data. If dropouts or noise are detected, determine they are within allowable values and they would not be interpreted as an actual control surface position movement.
TABLE 4-1. PARAMETER CORRELATION A4-3
Appendix 5 APPENDIX 5. CONSIDERATIONS FOR FILTERED FLIGHT DATA
1. Purpose of this appendix. This appendix addresses the recording of filtered flight data and means of compliance with the “Filtered Flight Data” final rule published on February 19, 2010 (75 FR 7345) and effective as of April 20, 2010. The following guidance expands on terms, discusses the scope, and provides information for operators on how to comply with this rule.
2. Discussion of terms used in the regulations.
a. What is an original sensor signal?
(1) An original sensor signal is the direct electronic output from:
(a) A sensor, or
(b) A system that accepts multiple sensor inputs.
(2) An air data inertial reference unit (ADIRU), an inertial reference system (IRS), and an ADC are examples of systems capable of producing an original sensor signal. For example, an ADC accepts multiple sensor inputs from pitot-static tubes and temperature probes. These inputs are processed by the ADC to output information, such as airspeed, outside air temperature, and pressure altitude. In this ca se, we consider the original sensor signal node to be at the output of the ADC.
b. What is parameter accuracy? In 14 CFR part 121 appendix M (and appendix E in part 125 and appendix F in part 135), the third column heading label reads “Accuracy (sensor input).”
The parenthetical “sensor input” refers to the signal recorded by the DFDR that is provided by a sensor, or system as described in paragraph 2.a(1) of this appendix. The total parameter accuracy is a function of the DFDR recorded value when compared to the actual measured value at the same point in time. For example, parameter accuracy is the difference between an actual physical measurement of a flight control surface and the corresponding DFDR recorded value.
Accuracy is typically expressed as a tolerance in EU or a percentage.
c. What is dynamic condition?
(1) The term “dynamic condition” was introduced in the regulations in 1997 and is found in 14 CFR part 121 appendix M, Airplane Flight Recorder Specifications, (and appendix E in part 125 and appendix F in part 135). Until April 20, 2010 the introductory text to appendix M previously stated:
“The recorded values must meet the designated range, resolution, and accuracy requirements during dynamic and static conditions. All data recorded must be correlated in time to within one second.”
A5-1
Appendix 5
(2) Prior to the filtered flight data final rule publication on February 19, 2010, there was no clarification for the meaning of dynamic condition. Our expansion of dynamic condition was adopted from NTSB safety recommendations A-03-048 through 050. The introductory text to appendix M now states:
“The recorded values must meet the designated range, resolution, and accuracy requirements during dynamic and static conditions. Dynamic condition means the parameter is experiencing change at the maximum rate attainable, including the maximum rate of reversal. All data recorded must be correlated in time to within one second.”
(3) For example, under dynamic conditions, the maximum attainable rate of control surface movement is dependent on the flight control system design of the aircraft. When performing certification testing under dynamic conditions (see definition in appendix 6), the maximum rate attainable includes rates encountered in both normal and abnormal operating conditions. When performing continued airworthin ess testing (see functional check definition in appendix 6), it is not necessary to demonstrate the maximum rate attainable under dynamic conditions.
d. What is filtered data? As defined by 14 CFR §§ 121.346(a), 125.228(a), and
135.156(a), flight data is filtered when an original sensor signal has been changed in any way, other than changes necessary to:
(1) Accomplish analog to digital conversion of the signal,
(2) Format a digital signal to be DFDR compatible, or
(3) Eliminate a high frequency component of a signal that is outside the operational bandwidth of the sensor.
Note: Removing noise and aliased signals are industry standard techniques included under paragraph 2.d(3).
e. What is reconstructed data? Reconstructed data is derived from recorded filtered values by applying a unique algorithm to the filtered values. The final rule allows an option to reconstruct filtered data if the recorded DFDR values do not comply with the accuracy requirements in 14 CFR part 121 appendix M (or comparable appendices in parts 125 and 135).
A certificate holder must submit the procedure and the test results that demonstrate an original sensor signal value can be reconstructed from the recorded data.
3. Scope.
a. What parameters are allowed to be filtered? All parameters are allowed to be filtered.
The final rule distinguishes between the following two groups of parameters:
(1) Group 1 – Parameters (1) through (11), (18) through (41), (43) through (87), and (89) through (91). These parameters are allowed to be filtered with no further action.
A5-2
Appendix 5
(2) Group 2 – Parameters (12) through (17), (42), and (88). If these parameters are filtered, then the operator has two options. These options are discussed in paragraph 3.c(2) below.
b. How to identify if a Group 2 parameter is filtered?
(1) An operator should check its DFDR system engineering and maintenance documentation to determine those parameters that are considered by definition to be filtered.
Wiring diagrams may be necessary to trace the electronic signal from its origin to the DFDAU.
(2) If an operator does not possess the necessary information to determine if a Group 2 parameter is filtered, then the operator may choose to consult the TC, STC, or ATC holder to obtain the necessary information for its aircraft.
c. What are the options if a Group 2 parameter is identified as being filtered?
(1) After a parameter in Group 2 is identified as being filtered, the operating certificate holder must confirm that the accuracy of the parameter is in compliance with 14 CFR part 121 appendix M in both static and dynamic conditions, as specified in the appendix introductory text.
Note: Parameter 42 (power/throttle lever position) is required to be sampled one time per second. The certificate holder may show by quantitative analysis that the filtered data does not affect the parameter’s accuracy.
(2) If the operating certificate holder identifies a filtered parameter not in compliance with the accuracy requirements, the operator has two options. An operator must either redesign the DFDR system to record the original sensor signal, or demonstrate to the FAA that the filtered data can be reconstructed to be in compliance with 14 CFR part 121 appendix M (or comparable appendices in parts 125 and 135). However, if the FAA determines that the reconstruction procedures are not acceptable, then the operator must have the DFDR system redesigned.
4. Compliance times and documentation.
a. What is the compliance schedule?
(1) Operators have 18 months, no later than October 20, 2011 (reporting date), to inspect their DFDR systems and document their findings whether or not their DFDR systems record filtered data for any of the parameters included in Group 2. The operator must create a record of this determination for each airplane on its operations specifications. The operator must maintain this record as part of the correlation documentation required by 14 CFR §§ 121.344(j)(3),
125.226(j)(3), or 135.152(f)(2)(iii). If any of those recorded signals are being filtered, then the operator must identify which parameters are affected. If the parameters are not being filtered, then the operator must document its finding and is not required to take further action, unless a future change is made to its DFDR system.
(2) Operators that report filtered parameters will have two options.
A5-3
Appendix 5
(a) If an operator chooses to remove the filtering, it has to make the modifications no later than April 21, 2014 (thirty months from the last permissible reporting date).
(b) If an operator chooses to demonstrate by test and analysis that filtered data can be reconstructed, the operator has to submit its reconstruction package, including the procedures and test and analysis results, in the correlation documentation to its FAA certificate holding district office for approval no later than April 22, 2013 (18 months from the reporting date). The final year is needed for the FAA to review the documentation and verify the procedure and test results are repeatable.
(3) In all cases, compliance with the rule is required by April 21, 2014 (four years from the effective date). In no case will the subm ission of reconstruction procedures and test and analysis results be considered compliance until that submission is approved by the FAA.
Operators who choose that method of compliance are cautioned to submit their procedures and test and analysis results as early as possible in case their submissions fail to be approved and other action is needed.
b. May an operator submit data from the airframe manufacturer? Yes, operators may submit documentation from airframe manufacturers to comply with 14 CFR §§ 121.346(d)(1),
125.226(d)(1), or 135.156(d)(1). However, for all airplanes of a particular model, the operator must record the individual airplane serial number and U.S. registration number in the correlation documentation for all airplanes on its operations specifications that have identical DFDR systems.
c. How does the operator show compliance? By the last permissible reporting date (October 20, 2011), all operators must have updated correlation documentation stating which of the listed parameters in 14 CFR § 121.346(c) is filtered or not filtered by definition. This must be done for each aircraft listed in the operator’s operations specifications. There are three cases in which an operator would show compliance:
(1) At the report date, no later than October 20, 2011, the operator updates its correlation documentation to show its DFDR system does not record any filtered data for the parameters listed in 14 CFR § 121.346(c). For all airplanes of a particular model, the operator must record the individual airplane serial number and U.S. registration number in the correlation documentation for all airplanes on its operations specifications that have identical DFDR systems.
(2) An operator identifies a Group 2 parameter as being filtered in its correlation documentation and opts to reconstruct the original sensor signal values. The operator must submit the procedures and test and analysis results to reconstruct the filtered data as early as possible, no later than April 22, 2013. The data package must be the updated correlation documentation with at least the following information:
(a) Operating certificate holder name, aircraft type, and DFDR system description, A5-4
Appendix 5
(b) Name, number, and accuracy of the parameter as defined by the applicable operating rule,
(c) Actual measured data from the original source signal exercised in static and dynamic conditions,
(d) Raw DFDR data of the filtered parameter exercised in static and dynamic conditions,
(e) EU conversion algorithm,
(f) Filtered data reconstruction algorithm, and
(g) Analysis of the reconstruction algorithm applied to the raw DFDR data.
(3) An operator identifies a Group 2 parameter as being filtered and opts to redesign the DFDR system to record original sensor signal data. This case is handled through the STC process.
5. Flowchart. Figure 5-1 is a flowchart of the filtered flight data final rule. This flowchart breaks down the regulatory requirements in 14 CFR § 121.346. Figure 5-1 can be applied to the similar requirements in 14 CFR §§ 125.228 and 135.156.
A5-5
Appendix 5 FIGURE 5-1. FILTERED FLIGHT DATA FINAL RULE COMPLIANCE FLOWCHART.
START Are any NO parameters in §121.346(c) (Group 2) filtered ?
END YES Update correlation Submit STC data to documentation. See ACO for approval. See paragraph 4.c(1) paragraph 4.c(3) Do those Redesign DFDRS to filtered parameters Update correlation record unfiltered data NO meet the accuracy YES documentation. See no later than requirement under part 121 paragraph 4.c(1) April 21, 2014.
Appendix M?
NO END Can you NO the original sensor reconstruct signal value from the recorded data to meet the parameter accuracy requirement?
Maintain reconstruction procedures and test and analyses results with YES correlation documentation. See Submit reconstruction paragraph 4.c(2) procedures to FAA CHDO.
See paragraph 4.c(2) Are the reconstruction ON procedures and test and YES analysis results approved by the FAA valid and repeatable?
A5-6
Appendix 6 APPENDIX 6. DEFINITIONS
1. Applicant. A person seeking FAA approval for a DFDR installation. The approval may be a TC, amended TC, or STC. The approval may apply to a single aircraft or to multiple aircraft of a single type design.
2. Correlation. Describes the relationship between two variables. In the case of the DFDRS, the first variable is the measured or sampled data and the second is the EU converted data in the DFDR.
3. Correlation coefficient. A number that describes the degree of relation between the data being measured or sampled and the data derived from the recorder. The correlation coefficient used here is the Pearson product-moment correlation coefficient, however the coefficient of determination R2 or the sum of least squares methods are acceptable. Its value may vary from minus one to plus one. A value of positive one (1.0) indicates a perfect direct correlation. A value of zero indicates there is no correlation, or any predictive capability between the derived data (using the equation) and the raw data is purely coincidental. A value of negative one (-1.0) indicates a perfect inverse relationship between the derived value and the raw data. The absolute value of the correlation coefficient must be equal to or greater than 0.99 over the entire operating range of each mandatory parameter to accurately establish the conversion of recorded values to EU. See appendix 1 for one method, but not the only method, to determine the correlation coefficient.
4. Correlation data. The documentation required by 14 CFR § 121.344(j)(3), 121.344a(d)(3),
125.226(j)(3), and 135.152(f)(2)(iii). This documentation must show how each parameter was correlated and be maintained by the operating certificate holder.
5. Date of manufacture. Different dates of manufacture may be triggered by the varying regulatory requirements applicable to an individual aircraft. When two or more regulations apply, each with their own date of manufacture, the FAA recommends using the earliest date required as compliance for all of the regulations under which a date of manufacture is specified.
6. Dedicated sources. Any digital system interface, analog discrete interface, or analog sensor interface installed solely to produce an input signal to the DFDRS. For example, if it is necessary to install a horizontally mounted accelerometer to sense lateral acceleration, then this accelerometer is considered a dedicated sensor. Conversely, the vertical accelerometer may exist on the aircraft for another reason (vertical flight control, for instance). If the DFDR takes vertical acceleration data from such an existing accelerometer, then the vertical accelerometer is not part of the DFDRS. However, the accelero meter must meet the accuracy requirements listed in the appropriate regulatory appendix.
7. Digital flight data acquisition unit (DFDAU). An electronic device that collects, samples, reformats, and digitizes analog and digital signals representing aircraft flight data. See ARINC Characteristic 717. The equivalent of this functio n may be incorporated in other hardware that may not meet the classical definition of an ARINC 717 DFDAU.
8. Digital flight data acquisition function (DFDAF). Software performing the function of a DFDAU in a larger avionics system.
A6-1
Appendix 6
9. DFDR. A recording device using a digital method to record and store data onto a storage medium and to retrieve that data from the medium. A DFDR may be the storage device in a recording system that includes a DFDAF, DFDAU, or an FDAU. Or, it may be a stand-alone device using an internal data collection system to convert aircraft analog and discrete signals to digital form.
10. Digital flight data recorder system (DFDRS). The software, equipment, sensors, wiring, equipment racks, and other items installed in the aircraft to record flight data. It includes the following equipment items: DFDR, dedicated sensors, underwater locating device (ULD), and a DFDAF, DFDAU, or FDAU.
11. Dynamic condition. The parameter is experiencing change at the maximum rate attainable, including the maximum rate of reversal.
12. Flight data acquisition unit (FDAU). The predecessor to the DFDAU, defined by ARINC 573. It is primarily designed to accept analog and analog discrete input data, and in some cases, it will accept digital data on predefined inputs.
13. Flight data recorder (FDR). A recording device directly receiving analog signals representing various aircraft functions (e.g., vertical acceleration, heading, altitude, or airspeed) and recording those signals in digital or analog format. Older FDRs recorded the signals by scratch with a stylus on a moving oscillographic medium, typically a foil formed from steel or steel alloy. These older analog FDR installations typically conformed to ARINC Characteristic 542. The FAA now requires that DFDRs be used in the U.S. commercial fleet.
14. Functional check. A check to determine if one or more functions of a DFDRS perform within specified limits. The following identifies and defines three types of DFDRS functional checks.
a. First of type:
(1) The applicant must perform the first of type installation testing for an FAA approval (TC, amended TC, or STC). This check determines if the range, accuracy, sampling rate, and resolution of each parameter is within specified limits. During the first of type functional check, it may not be feasible to stimulate some sensors to their specified limits. In such instances it is acceptable to simulate the sensor output using suitable test equipment. This check collects correlation data in support of certification. See appendix 2 for procedures when a DFDR data correlation document does not exist for an aircraft.
(2) For data received by the recording system in a digital means, credit may be given to the source system in determining the range and resolution of the parameter. Range and resolution are functions, not necessarily of the recording system, but of the source system.
However, the range and resolution of the source system must meet the minimum requirements of the applicable appendix in 14 CFR. If the range and resolution satisfy the requirements in the applicable appendix in 14 CFR, there is no need to repeat these checks for the purpose of meeting recording system requirements.
A6-2