Appendix 5 Figure 5-7 Aft Pressure Dome Outer Ring and Dome Web Splices (MSD/MED) Typical outer ring splices
- Outer ring splice -Web splices Type and possible location of MSD/MED
• MSD/MED-outer ring splice
• Attachment profiles - at fastener rows and/or in
• MED-web splices
• Bulkhead skin and/or splice plates -at critical fastener rows Legend:
F fastener R radius Service or test experience of factors that influence MSD and/or MED (examples)
• Corrosion
• High stresses - combined tension and compression
• High induced bending in radius
• Inadequate finish in radius - surface roughness Figure 5-8 Skin Splice at Aft Pressure Bulkhead (MSD) Pressurized skin
• Unpressurized skin Skin cracking *T" frame ~ Bulkhead web Service or test experience of factors that influence MSD and MED (examples) Type and possible location of MSD and MED • Shell discontinuous induced bending
• MSD - skin at end fastener holes stresses
• High load transfer at fastener A5-5
Appendix 5 Figure 5-9 Abrupt Changes in Web or Skin Thickness-Pressurized or Unpressurized Structure (MSD/MED)
- Edge support member Web or skin Typical cracking . Radius
• Milled
• Chem-milled Bonded doubler Type and possible location of MSD and MED Abrupt change in stiffness
• Milled radius
• Chem-milled radius
• Bonded doubler
• Fastener row at edge support members Edge member support structure
• Edge member - in radius areas Service or test experience of factors that influence MSD and MED Pressure structure
• High bending stresses at edge support due to pressure Non-pressure structure
• Structural deflections cause high stresses at edge supports Figure 5-10 Window Surround Structure (MSD, MED) Window surround structure Type and possible location of MSD/MED
• MSD - skin at attachment to window surround structure
• MED-repeated details in reinforcement of window cutouts or in window corners Service or test experience of factors that influence MSD and/or MED (examples)
• High load transfer A5-6
Appendix 5 Figure 5-11 Over-Wing Fuselage Attachments (MED)
- Upper wing surface
- Typical fuselage attachments Type and possible location of MSD/MED
• MED-repeated details in overwing fuselage attachments Service or test experience of factors that influence MSD and/or MED (examples)
• Manufacturing defect - prestress
• Induced deflections Figure 5-12 Latches and Hinges of Non-Plug Doors (MSD/MED) -- Latch hook Door Hinge lobe Hinge pin Lubrication passage View A Attach bolts View B Type and possible location of MSD/MED Service or test experience of factors that
• MSD - piano hinge influence MSD and/or MED (examples)
• At hinge fastener attachment row • Bending stresses due to fuselage elongation
• In fillet radius • High local stress
• Emanating from hole in lobes • Fretting
• MED- latches
• In multiple latch hooks
• At lube channel of latch spool
• At spool bracket attach bolts (also corrosion) A5-7
Appendix 5 Figure 5-13 Skin at Runout of Large Doubler (MSD) Fuselage, Wing, or Empennage 000000
••••• Skin Type and possible location of MSD/MED
• MSD-cracks initiated at multiple critical fastener holes in skin at runout of doubler Skin doubler Service or test experience of factors that influence MSD and/or MED (examples)
• High load transfer - high local stress Figure 5-14 Wing or Empennage Chordwise Splices (MSD/MED) Front Spar Typical skin and stringer splice Rib Fitting - - Skin panel Splice plate Stringer → Chordwise joints Type and possible location of MSD/MED
• MSD-skin and/or splice plate
• Chordwise critical fastener rows
• MED-stringer runout of fitting
• Fatigue-critical fastener holes at stringer and/or fitting Service or test experience of factors that influence MSD and/or MED (examples)
• High load transfer
• Local bending A5-8
Appendix 5 Figure 5-15 Rib-to-Skin Attachments (MSD/MED) Typical shear tie Typical skin cracking Skin Typical - Stringer cracking 0000
- Rib web Type and possible location of MSD and MED
• MSD- critical fasteners in skin along rib attachments
• MED -critical rib feet in multiple stringer bays (particularly for empennage under sonic fatigue) Service or test experience of factors that influence MSD and MED (examples)
• Manufacturing defect - prestress due to assembly sequence
• Sonic fatigue (empennage) Figure 5-16 Typical Wing and Empennage Construction (MSD/MED) Riveted Skin and Stringer Construction (MSD & MED) Integrally Stiffened Skins (MSD) Drain hole crack Fastener • (a) attachment (b) to rib Cracks Cracks Root rib, tank Crack (e) end, etc. • Root rib, tank end, etc.
Cracks (C) Skin Inherent fail safe and crack stopper Do not have inherent crack stopper characteristics characteristics of riveted skin and
• MSD-chordwise cracks link up at stringer construction
a) Rib attachment holes • MSD-Chordwise cracks link up at
• MED - d) Rib attachment holes
b) Drain or vent holes e) Drain or vent holes
c) Stiffener run-outs at f) Stringer run-outs at root rib or tank root rib or tank end rib end rib
• MED- becomes MSD A5-9
Appendix 6 A6-1 APPENDIX 6
WIDESPREAD FATIGUE DAMAGE EVALUATION This appendix tells you how to perform a WFD evaluation as part of the overall process for establishing an LOV. The WFD evaluation may be simple or very complex. The situation will vary from model to model and area to area on any given airplane structural configuration. The evaluation of the identified WFD-susceptible structure has two objectives—
• To predict when WFD is likely to occur.
• To establish additional maintenance actions, as necessary, to ensure continued safe operation of the airplane.
The information contained in this appendix was derived in part from the 1999 and 2003 ARAC Working Group reports referenced in Appendix 1.
THE EVALUATION The WFD evaluation of susceptible structure is a process for determining when WFD is likely to occur in the structure, what maintenance actions might be necessary to prevent it from occurring before the LOV, and when those maintenance actions should be begun. This evaluation should be based on service history, test data, fatigue analyses, damage-tolerance analyses, teardown inspections of high-usage airplanes, or any combination of these.
1. Predicting When WFD is Likely to Occur.
a. Characterization of Events Leading to WFD. The fatigue process that leads to WFD is shown in Figure 6-1. This figure is applicable both to damage that occurs in multiple sites (multiple site damage (MSD)) and damage that occurs in similar structure at more than one location (multiple element damage (MED)). For any susceptible structural area, it is not a question of whether WFD will occur—but when it will occur. In Figure 6-1, the “when” is illustrated by the line titled “WFD (average behavior),” which is the point when, without intervention, 50 % of the airplanes in a fleet would have experienced WFD in the considered area. (Note that the probability density function for flight cycles or flight hours to WFD has been depicted for reference.) The WFD process includes this phase of crack initiation and a crack growth phase. During the crack initiation phase, which generally spans a long period of time, there is little or no change in the basic strength capability of the structure. The actual residual strength curve depicted in Figure 6-1 is flat, and equal to the strength of the structure in its pristine state. However, at some time after the first small cracks start to grow, residual strength begins to degrade. Crack growth continues until the capability of the structure degrades to the point of the minimum strength required by § 25.571(b). In this context, the line in Figure 6-1 called WFD(average behavior) represents a point when, without intervention, 50 % of the airplanes in a fleet fall below the minimum strength requirements of § 25.571(b).
Appendix 6 Figure 6-1 MSD/MED Residual Strength Curve NOTE: No Scale Implied. PROBABILITY DENSITY For Definition Use Only FUNCTION FOR CYCLES TO
WFD CONDITION
REQUIRED ACTUAL
RESIDUAL RESIDUAL
STRENGTH STRENGTH
MSD/MED
CRACK MSD/MED
INITIATION CRACK GROWTH
CRACK LENGTH
FLIGHT CYCLES/HOURS WFD (AVERAGE BEHAVIOR)
MSD - MULTIPLE SITE DAMAGE
MED - MULTIPLE ELEMENT DAMAGE A6-2
Appendix 6 A6-3
b. Widespread Fatigue Damage (average behavior). WFD average behavior should be estimated for each susceptible structural area. Such an estimate may be based on full-scale fatigue test evidence and analyses at a minimum, and if available, in-service history. In making this estimate, consider the following:
(1) A complete review of the in-service history of the susceptible areas (including loads, mission profiles, environment, and operational statistics of the fleet, stated as a number of accumulated flight cycles or flight hours).
(2) Significant production variants (material, design, assembly method, and any other change that might affect the fatigue performance of the detail).
(3) Relevant full-scale and component fatigue test data. (See Appendix 2 of AC 25.571-1 (as revised) for guidelines on full-scale fatigue test evidence, including addressing cracks found during testing.)
(4) Relevant data from teardown inspections of full-scale and component fatigue test data, including teardown inspections of high-time airplanes.
(5) Any fractographic analysis available.
The evaluation of test results for the reliable prediction of the time when WFD might occur in each susceptible area should include appropriate test-to-structure factors (see definition in Appendix 2 of this AC). The WFD evaluation may be analytically determined, supported by test or in-service information.
c. Cracking Pattern. Regardless of whether the assessment of WFD average behavior is based on in-service data, full-scale fatigue test evidence, analyses; or a combination of any of these, the following should be considered:
(1) Initial Cracking Scenario. This is the size and extent of multiple location cracking expected at the initiation of MSD or MED. To determine the time to MSD or MED initiation you will need empirical data or an assumption of the crack locations and sequence, plus a fatigue evaluation. Alternatively, analysis can be based on either:
(a) The distribution of equivalent initial flaws as determined from the regression to zero cycles of flaws found during fatigue testing, teardown inspections, or both.
(b) The distribution of fatigue cracking determined from relevant fatigue testing or service experience, or both.
(2) Final Cracking Scenario. This is an estimate of the size and extent of multiple location cracking that could cause residual strength to fall to the minimum required level (WFD condition), as shown in Figure 6-1. There are techniques for 3-D elastic-plastic analysis of such problems; but there are also several alternative test and analysis approaches that provide acceptable estimates. One such approach is to define the final cracking scenario as not yet critical—a point in time before the cracking has developed into the WFD condition. An example of an MSD problem would be the occurrence of the
Appendix 6 A6-4 first crack link-up in a row of fastener holes. An example of an MED problem would be simultaneous cracking of two or three adjacent structural elements. Using a sub-critical scenario reduces the complexity of the analysis and, in many cases, will not greatly reduce the estimate of the total time to WFD(average behavior) because the majority of the total time is generally associated with crack initiation.
(3) Crack Growth. Progression of the crack distributions from the initial cracking scenario to the final cracking scenario should be developed. You can develop crack growth predictions in two ways—
(a) Analytically: typically based on linear elastic fracture mechanics; or
(b) Empirically: from test or service fractographic data.
(4) Differences Between MSD and MED. We expect details of the approach used to characterize events leading up to WFD to be different. The differences will depend on whether you are considering MSD or MED. This is especially true for crack interaction.
(a) Crack Interaction. MSD has the potential for strong crack interaction, and the effect of multiple cracks on each other needs to be addressed. MED, in most cases, does not have the same potential for strong crack interaction. The differences between interaction effects for MSD and MED are illustrated in Figure 6-2.
Figure 6-2 Difference Between MSD and MED Interaction Effects
Appendix 6 A6-5
(b) MSD and MED Interaction. Some areas of an airplane are potentially susceptible to both MSD and MED. Simultaneous occurrence of MSD and MED is possible, even though it’s not common. A comparison of ISPs or modification start points might indicate the possibility of this occurring. If so, your evaluation should consider interaction between MSD and MED.
(5) Multiple Site Damage (MSD). When you are assessing MSD, certain assumptions or methods may have a greater impact than others on the final outcome of the WFD evaluation.8 The following items were found to have such an impact:
8 “Recommendations for Regulatory Action to Prevent Widespread Fatigue Damage in the Commercial Airplane Fleet,” Revision A, dated June 29, 1999 (a report of the Airworthiness Assurance Working Group for the Aviation Rulemaking Advisory Committee’s Transport Aircraft and Engine Issues Group).
(a ) The flaw sizes assumed at initiation of the crack growth phase of analysis.
(b) Material properties used (static, fatigue, fracture mechanics).
(c) Ligament9 failure criteria.
9 Ligament is the material between discontinuities in a given structure. Types of discontinuities include holes, cutouts, or cracks.
(d ) Crack growth equations used.
(e) Statistics used to evaluate fatigue behavior of the structure (e.g., time to crack initiation).
(f) Methods of determining the structural modification point (SMP).
(g) Detectable flaw size assumed.
(h) Initial distribution of flaws.
(i) Factors used to determine lower-bound behavior as opposed to mean behavior.
(6) Multiple Element Damage (MED). When considering MED in your evaluation, you need not consider interaction between cracks in different elements. You should consider the following:
(a) There is not a high probability that, after a crack initiates in an element, a second crack will initiate in an element right next to it. If this does happen, however, the consequences to the overall structure may be severe. This is because having two structural members fail right next to each other can completely negate any ability of the structure to tolerate additional damage. Consequently, when performing your evaluation, you should make conservative assumptions and assume failures to be adjacent to each other.
(b) When an element fails completely, the load that has to be redistributed onto the non-failed structure can be large and can have a significant impact on evaluation
Appendix 6 A6-6 results. You should therefore consider the effects of load redistribution and include it in your evaluation.
2. Establishing Maintenance Actions.
a. SMP and ISP. Fatigue damage is the gradual deterioration of a material subjected to repeated loads. This gradual deterioration is a function of use and can be statistically quantified. The term WFD is used, and can be statistically quantified, at the end of the deterioration process—when the structure is no longer able to carry the residual strength loads. As depicted in Figure 6-1, WFD can never be absolutely precluded because there is always some probability, no matter how small, that it will occur. Therefore, modifying or replacing structure at a pre-determined, analytically-derived time stated in flight cycles or flight hours, minimizes the probability of having WFD in the fleet. Modification or replacement is the most reliable method for precluding WFD. The point at which a modification is undertaken is referred to as the “structural modification point (SMP)” and it is illustrated in Figure 6-3. The SMP is generally a fraction of the number representing the point in time when WFD(average behavior) will occur, and should result in the same reliability as a successful two-lifetime fatigue test. This level of reliability for setting the SMP is acceptable if inspections for MSD or MED are shown to be effective in detecting cracks. If the inspections are effective, they must be implemented before the SMP. The implementation times for these inspections are known as the “inspection start points (ISP).” Repeat inspections are usually necessary to maintain this effectiveness in detecting cracks. If inspections for MSD or MED are not effective in detecting cracks, then SMP should be set at the time of the ISP. For the purposes of this AC, an inspection is effective if, when performed by properly trained maintenance personnel, it will readily detect the damage in question.10 The SMP should minimize the extent of cracking in the susceptible structural area in a fleet of affected airplanes. In fact, if this point is appropriately determined, a high percentage of airplanes would not have any MSD or MED by the time the SMP is reached.
10 The cracking identified in airworthiness directive (AD) 2002-07-09 is an example of the type of cracking that MSD inspections are effective in detecting. These cracks grow from the fastener holes in the lower row of the lower skin panel in such a way that the cracking is readily detectable using non-destructive inspection methods. The cracking identified in AD 2015-21-06 is an example of places where MSD inspections are not effective. These cracks grow in the outer surface and between the fastener holes in the lower row of the lower skin panel in such a way that the cracking is not readily detectable using non-destructive inspection methods. Modification is the only option to address this type of cracking.
(1) Structural Modification Point (SMP). The SMP should be established as a point in time when structure should be modified or replaced to prevent WFD from occurring. This is typically established by—
• Calculating when WFD would first occur in the structure (predicted using the WFD(average behavior)).
• Setting a time before the predicted occurrence of WFD to perform modifications or replacements that will prevent it.
Appendix 6 A6-7
(a) As an example, the SMP may be determined by dividing the number representing the timing of when WFD(average behavior) will occur by a factor of 2 if there are effective inspections, or by a factor of 3 if inspections are not effective.
(b) A DAH may find that the SMP for a particular structural area has been exceeded by one or more airplanes in the fleet. In that case, the DAH should evaluate those high-time airplanes to determine their structural condition and notify the airworthiness authorities and propose appropriate maintenance actions specific to those airplanes.
(2) Inspection Start Point (ISP). If an inspection is determined to be effective, you will need to establish when those inspections should start. This point is illustrated in Figure 6-1. The start point is determined through a statistical analysis of crack initiation based on fatigue testing, teardown, or in-service experience of similar structure. The ISP is assumed to be equivalent to a lower-bound value with a specific probability in the statistical distribution of cracking events. Alternatively, you may establish an ISP by applying appropriate factors (e.g., dividing by a factor of 3) to the number representing WFD average behavior.
(3) Inspection Interval. The interval between inspections depends on the detectable crack size and the probability that it will be detected with the specific inspection method.
(4) Adjustment of SMP. The SMP may be extended or reduced, based on the following:
(a) Extension of SMP. You can extend the SMP if you can show freedom from WFD up to the new SMP. The tasks required to extend the SMP may include any or all of the following:
1 Additional fatigue or residual strength tests, or both, on a full- scale airplane structure or a full-scale component followed by detailed inspections and analyses.
2 Fatigue tests of new structure or structure from in- service airplanes on a smaller scale than full component tests (i.e., sub-component or panel tests, or both). If a sub-component test is used, the SMP would be extended only for that sub-component.
3 Teardown inspections (destructive) on structural components that have been removed from service.
4 Teardown inspections (non- destructive) accomplished by selected, limited disassembly and subsequent reassembly of specific areas of high-time airplanes.
5 Analysis of in- service data (e.g., inspections) from a statistically significant number of airplanes.