(3) The damage tolerance characteristics can be shown analytically by reliable or conservative methods such as the following:
By demonstrating quantitative relationships with structure already verified as damage tolerant;
(11) By demonstrating that the damage would be detected before it reaches the value for residual strength evaluation; or
(iii) By demonstrating that the repeated loads and limit load stresses do not exceed those of previously verified designs of similar configuration, materials, and inspectability.
(4) The maximum extent of immediately obvious damage from discrete sources should be determined and the remaining structure shown to have static strength for the maximum load (considered as ultimate load) expected during the completion of the flight.
Normally, this would be an analytical assessment. In the case of uncontained engine failures, the fragments and paths to be considered should be consistent with those used in showing compliance with § 25.903(d)(1) of the FAR, and with typical damage experienced in service.
(Refer to AC 20-128, "Design Considerations for Minimizing Hazards Caused by Uncontained Turbine Engine and Auxiliary Power Unit Rotor and Fan Blade Failures.") Inspection.
(1) Detection of damage before it becomes critical is the ultimate control in ensuring the damage tolerance characteristics of the structure. For this reason, Amendment 25- 54 revised § 25.571 to require that the applicant establish inspections or other procedures, as necessary, to prevent catastrophic failure from accidental, environmental, or fatigue damage, and include those inspections and procedures in the Airworthiness Limitations Section of the "Instructions for Continued Airworthiness" required by 14 CFR § 25.1529 (see also App. H to part 25).
(2) Due to the inherent, complex interactions of the many parameters affecting damage tolerance, such as operating practices, environmental effects, load sequence on crack growth, and variations in inspection methods, related operational experience should be taken into account in establishing inspection procedures. Additionally, careful consideration should be given to the practicability of inspection procedures.
(3) Comparison with past successful practice is the primary means by which inspections or other procedures for accidental and environmental damage are substantiated. This is generally accomplished through the Maintenance Review Board activity for a new model large transport airplane. Although such inspections or other procedures, as necessary to prevent catastrophic failure of an airplane, are required to be included in the Airworthiness Limitations Section of the "Instructions for Continued Airworthiness," subsequent individual operator
experience may indicate that different inspections or other procedures are justified for that individual operator. Sections 43.16 and 91.403(c) provide a means for FAA approval of alternatives to the airworthiness limitations. In reviewing such proposed alternatives, the FAA will evaluate them using the methods described in this AC, or other acceptable methods proposed by the TC holder and operators, to ensure that the objectives of § 25.571 will continue to be met.
k. Threshold for Inspections.
(1) Where it can be shown by observation, analysis, and/or test that a load path failure in multiple load path "fail-safe" structure, or partial failure in crack arrest "fail-safe" structure, will be detected and repaired during the normal maintenance, inspection, or operation of an airplane prior to failure of the remaining structure, the thresholds can be established using either:
(i) Fatigue analysis and tests with an appropriate scatter factor; or
(ii) Slow crack growth analyses and tests, based on appropriate initial manufacturing damage.
(2) For single load path structure and for multiple load path and crack arrest "fail-safe" structure, where it cannot be demonstrated that load path failure, partial failure, or crack arrest will be detected and repaired during the normal maintenance, inspection, or operation of an airplane prior to failure of the remaining structure, the thresholds should be established based on crack growth analyses and/or tests, assuming the structure contains an initial flaw of the maximum probable size that could exist as a result of manufacturing or service induced damage.
FATIGUE EVALUATION.
General. The evaluation of structure under the following fatigue (safe-life) strength evaluation methods is intended to ensure that catastrophic fatigue failure, as a result of the repeated loads of variable magnitude expected in service, will be avoided throughout the structure's operational life. Under these methods, the fatigue life of the structure should be determined. The evaluation should include the following:
Estimating or measuring the expected loading spectra for the structure;
(2) Conducting a structural analysis, including consideration of the stress concentration effects;
(3) Performing fatigue testing of structure that cannot be related to a test background to establish response to the typical loading spectrum expected in service;
(4) Determining reliable replacement times by interpreting the loading history, variable load analyses, fatigue test data, service experience, and fatigue analyses;
(5) Evaluating the possibility of fatigue initiation from sources such as corrosion, stress corrosion, disbonding, accidental damage and manufacturing defects based on a review of the design, quality control and past service experience; and
(6) Providing necessary maintenance programs and replacements times to the operators. The maintenance program should be included in the Instructions for Continued Airworthiness in accordance with Section 25.1529.
Scatter Factor for Safe-life Determination. In the interpretation of fatigue analyses and test data, the effect of variability should, under § 25.571(c), be accounted for by an appropriate scatter factor. In this process it is appropriate that the applicant justify the scatter factor chosen for any safe-life part. The following guidance is provided (see Figure 1):
HAVE THE CRITERIA OF § 7.b. (3) BEEN MET:
-- Service and Test Experience of Similar Components § 7.0. (3)(1) -- QA System Ensuring Fatigue Scatter Lies within Certain Limits § 7.b. (3) (il) .. Representativeness of Test Specimen § 7.b.(3)(il)
ALL CRITERA SOME CRITERA
MET MISSED USE BSF, =3.0] USE BSF, 23.0 § 7.b. (3) § 7.b.(5) PAVE THE ELEMENTS OF §7.6.(4) BEEN ACCOUNTED FOR IN DESIGN:
-- Fatigue Scatter to Account for P-99% and C-95% § 7.b.(4)(1)
- Spectrum Severity § 7.b. (4) (ii)
EST, DETERM NEDIFROM ANALYSIS AND TEST § 7.b.(5)(i)
- Required Level of Safety -- Number of Specimens Tested -- Representativeness of Test
- Fatigue Scatter to Account for P-99% and C-95% --Type of Repeated Loads Test -- Accuracy of the Test Load ALL ELEMENTS SOME ELEMENTS Spectrum MET MISSED -- Spectrum Severity -- Service Environmental Conditions Minimum Value ≥ 3.0 § 7.b.(5) (ii) Adjust BSF, for Resolution of Test Loads to Actual Loads § 7.b. (6) ADJUST BSF, for: ADJUST BSF, for: SAFELIFES -- Number of Specimen -- Fatigue Scatter § 7.b. (4)(i) TEST CYCLES/SCATTER FACTOR* Tested § 7.b.(4) (iii) -- Spectrum Severity § 7.b. (4) (ii)
- Resolution of Test Loads -- Number of Specimen Tested to Actual Loads § 7.b. (6) § 7b. (4)(1) -- Resolution of Test Loads to Actual Loads § 7.b. (6) *Scatter Factor = BSF, × Adjustment
SAFE LIFES TEST CYCLES/SCATTER FACTOR* Figure 1. Safe-Life Determination 16
(1) The base scatter factors applicable to test results are: BSF1= 3.0, and BSF2 = (see paragraph 7b(5) of this AC). If the applicant can meet the requirements of paragraph 7b(3) of this AC, he may use BSF1 or, at his option, BSF2.
(2) The base scatter factor, BSF1, is associated with test results of one representative test specimen.
(3) Justification for use of BSF1. BSF1 may only be used if the following criteria are met:
Understanding of load paths and failure modes. Service and test experience of similar in-service components that were designed using similar design criteria and methods should demonstrate that the load paths and potential failure modes of the components are well understood.
(ii) Control of design, material, and manufacturing process quality.
The applicant should demonstrate that his quality system (e.g., design, process control, and material standards) ensures the scatter in fatigue properties is controlled, and that the design of the fatigue critical areas of the part account for the material scatter.
(iii) Representativeness of the test specimen.
(A) The test article should be full scale (component or subcomponent) and represent that portion of the production aircraft requiring test. All differences between the test article and production article should be accounted for either by analysis supported by test evidence or by testing itself.
(B) Construction details, such as bracket attachments, clips, etc., should be accounted for, even though the items themselves may be non-load bearing.
(C) Points of load application and reaction should accurately reflect those of the aircraft, ensure correct behavior of the test article, and guard against uncharacteristic failures.
(D) Systems used to protect the structure against environmental degradation can have a negative effect on fatigue life and therefore should be included as part of the test article.
(4) Adjustments to base scatter factor BSF1. Having satisfied the criteria of paragraph 7b(3), justifying the use of BSF1, the base value of 3.0 should be adjusted to account for the following considerations, as necessary, where not wholly taken into account by design analysis. As a result of the adjustments, the final scatter factor may be less than, equal to, or greater than 3.0.
(i) Material fatigue scatter. Material properties should be investigated up to a 99% probability of survival and a 95% level of confidence.
Spectrum severity. Test load spectrum should be derived based on a spectrum sensitive analysis accounting for variations in both utilization (i.e. aircraft weight, cg etc.) and occurrences/size of loads. The test loads spectrum applied to the structure should be demonstrated to be conservative when compared to the usage expected in service.
(i11) Number of representative test specimens. Well established statistical methods should be used that associate the number of items tested with the distribution chosen, to obtain an adjustment to the base scatter factor.
(5) If the applicant cannot satisfy the intent of all of paragraph 7b(3) of this AC, BSF2 should be used.
(i) The applicant should propose scatter factor BSF2 based on careful consideration of the following issues: the required level of safety, the number of representative test specimens, how representative the test is, expected fatigue scatter, type of repeated load test, the accuracy of the test loads spectrum, spectrum severity, and the expected service environmental conditions.
In no case should the value of BSF2 be less than 3.0.
Resolution of test loadings to actual loadings. The applicant may use a number of different approaches to reduce both the number of load cycles and number of test setups required. These include, but are not limited to, spectrum blocking (e.g., a change in the spectrum load sequence to reduce the total number of test setups); high load clipping (e.g., the reduction of the highest spectrum loads to a level such that the beneficial effects of compression yield are reduced or eliminated); and low load truncation (e.g., the removal of non-damaging load cycles to simplify the spectrum). Due to the modifications to the flight-by-flight loading sequence caused by these changes, the applicant should propose either analytical or empirical approaches to quantify an adjustment to the number of test cycles which represents the difference between the test spectrum and assumed flight-by-flight spectrum. In addition, an adjustment to the number of test cycles may be justified by raising or lowering the test load levels, as long as appropriate data supports the applicant's position. Other effects to be considered are different failure locations, different response to fretting conditions, temperature effects, etc.
analytical approach should use well established methods or be supported by test evidence.
c. Replacement times. Replacement times should be established for parts with established safe-lives and should, under § 25.571(a)(3), be included in the "Instructions for Continued Airworthiness" required under § 25.1529. These replacement times can be extended 18
if additional data indicate an extension is warranted. Important factors that should be considered for such extensions include, but are not limited to, the following:
Comparison of original evaluation with service experience.
(2) Recorded load and stress data. Recorded load and stress data entails instrumenting airplanes in service to obtain a representative sampling of actual loads and stresses experienced. The data to be measured include airspeed, altitude, and load factor versus time data; or airspeed, altitude, and strain ranges versus time data; or similar data. The data, obtained by instrumenting airplanes in service, provide a basis for correlating the estimated loading spectrum with the actual service experience.
(3) Additional analyses and tests. If test data and analyses based on repeated load tests of additional specimens are obtained, a re-evaluation of the established safe-life can be made.
(4) Tests of parts removed from service. Repeated load tests of replaced parts can be utilized to reevaluate the established safe-life. The tests should closely simulate service loading conditions. Repeated load testing of parts removed from service is especially useful where recorded load data obtained in service are available, since the actual loading experienced by the part prior to replacement is known.
(S) Repair or rework of the structure. In some cases, repair or rework of the structure can gain further life.
d. Type design developments and changes. For design developments or design changes involving structural configurations similar to those of a design already shown to comply with the applicable provisions of § 25.571(c), it might be possible to evaluate the variations in critical portions of the structure on a comparative basis. Typical examples would be redesign of the wing structure for increased loads, and the introduction in pressurized cabins of cutouts having different locations or different shapes, or both. This evaluation should involve analysis of the predicted stresses of the redesigned primary structure and correlation of the analysis with the analytical and test results used in showing compliance of the original design with § 25.571(c).
e. Environmental effects such as temperature and humidity should be considered in the damage tolerance and fatigue analysis and should be demonstrated through suitable testing.
8. DISCRETE SOURCE DAMAGE.
a. General. The purpose of this section is to establish FAA guidelines for consistent selection of load conditions for residual strength substantiation in showing compliance with § 25.571(e), Damage tolerance (discrete source) evaluation. The intent of these guidelines is to define, with a satisfactory level of confidence, load conditions that will not be exceeded on the flight during which the specified incident of § 25.571(e) occurs. In defining these load 19
conditions, consideration has been given to the expected damage to the airplane, the anticipated response of the pilot at the time of the incident, and the actions of the pilot to avoid severe load environments for the remainder of the flight consistent with his knowledge that the airplane may be in a damaged state. With these considerations in mind, the following ultimate loading conditions should be used to establish residual strength of the damaged structure.
The maximum extent of immediately obvious damage from discrete sources (§ 25.571(e)) should be determined and the remaining structure shown, with an acceptable level of confidence, to have static strength for the maximum load (considered as ultimate load) expected during completion of the flight.
c. The ultimate loading conditions should not be less than those developed from the following conditions:
At the time of the incident:
(i) The maximum normal operating differential pressure, multiplied by a 1.1 factor, plus the expected external aerodynamic pressures during 1g level flight, combined with 1g flight loads.
(i1) The airplane, assumed to be in 1g level flight, should be shown to be able to survive any maneuver or any other flight path deviation caused by the specified incident of § 25.571(e), taking into account any likely damage to the flight controls and pilot normal corrective action.
(2) Following the incident:
(i) Seventy percent (70%) limit flight maneuver loads and, separately, 40 percent of the limit gust velocity (vertical and lateral) at the specified speeds, each combined with the maximum appropriate cabin differential pressure (including the expected external aerodynamic pressure).
(ii) The airplane must be shown by analysis to be free from flutter up to V/M, with any change in structural stiffness resulting from the incident.
Konald T. Waj Ronald T. Wojnar Manager, Transport Airplane Directorate Aircraft Certification Service 20