The Aging Composite Airframe by JCHalpin JCH Consultants Inc Briefing for CMH-17 Damage Tolerance Working Group, Oct 2015; & FAA/Bombardier/TCCA/EASA/Industry Composite Transport Damage Tolerance and Maintenance Workshop (Montreal, Quebec) September 2015

INTRODUCTION; Both fail-safe and slow crack growth design concepts can (and have been) defeated by durability-related fatigue and environmentally assisted cracking in aging in CIVL & MIL metallic airframes. Durability-related cracking manifests itself in the literature as the onset of wide spread fatigue damage, WFD. Composite airframes accumulate other types of damage over time; a potential for service induced Wide Spread Damage, WSD. Techniques are needed for assessing operational limits for composite and metallic aircraft structural applications relative to the onset of WFD (durability-related fatigue cracking) and service induced WSD that can defeat the structure’s ability to carry its residual strength requirement. Figure 1 illustrates the different perspectives regarding the difference between the WFD and WSD terminology. Figure1. The Effect of Defects Distribution in Structural Integrity Planning; Ref [1]. The distribution of defect sizes in any given structure can be considered to consist of a composite of the several distributions shown in Figure 1. The material as received from the vendor will contain very small flaws or defects such as inclusions, cracks, porosity and surface pits, scratches, disbonds, delaminations, surface cuts or scratches and machine marks. These inherent material flaws are considerably below the detection capability of the non-destructive inspection (NDI) and should be sufficiently small to not grow appreciably in service. A distribution of larger defects can exist as a result of the fabrication process or as large inherent flaws. The production quality control process is designed to detect and eliminate as many of these defects as possible but those that are not detected may propagate due to fatigue mechanisms during service. The largest defect size that could remain undetected in the newly fabricated structure after the final inspection is designated as ao. This defect dimension provides the starting point for damage growth projections that demonstrate adequate service life or the necessity for an in-service inspection and/or residual strength estimates. The largest damage size that can remain undetected after an inspection is designated as aNDI and becomes the initial damage size for a usage period. The ”defects introduced in service” are the result of discrete source events, as described in AC25.571 and other sources. An accumulation of accidental damage can also compromise both fail-safe and slow-growth design concepts. The purpose of this discussion is to provide a historical perspective for the evolution of certification/substantion concepts for composite airframes. This understanding is fundamental as community works to identify and document the conditions that determine operational limits composite and hybrid (metallic and composite) airframes. This understanding will assist in development of potential updates to AC 25.571 and AC 20-107.

the our for the

SENSITIVITY OF LAMINATED COMPOSITES differs from the relatively homogeneous metallics used in airframes. The graphite (and boron) fibers have demonstrated fatigue resistance over the past 50 Page 1 of 13

years. When these reinforcement fibers are used in thin oriented tapes or lamina, the lamina are fatigue resistant in the fiber direction. A laminated construction is built-up from plies of oriented fiber lamina that respond to in-plane loadings. They can be designed to control in-plane stiffness, dimensional change to environments and fatigue load conditions. These laminates are known as “fiber dominated.” When these laminates are penetrated with open or filled holes they experience strength reductions more sever that a typical metallic alloy. The composite laminate strength reduction is consistent with the elastic stress concentration for an anisotropic solid. Impacts from discrete sources, transverse to the lamination plane induces internal damage, similar to a local manufacturing or load-induced delamination. Impact damage can result in a comparable loss of laminate strength equivalent to an “open hole,” Figure 2. The layers in a “laminate” are bonded together by the basic resin inherent in the individual layers, the lamina. The strength and stiffness properties perpendicular to the lamination plane(s) limited by the resin bonding to about 10 to 15% of the notched in-plane properties. A review Ref [2] of out of plane structural failures identified them to be the result of high interlaminar tensile and shear stresses relative to low interlaminar strengths. Figure 2. Damage Sensitivity of Laminated Composite Systems: Primary is in-plane Loading Notch effects and secondary Induced-out-of-plane Loading Effects. These out of plane stresses and subsequent failures result either directly from the application of out-ofplane loads or indirectly as a result of laminate geometry under in-plane loads or environmentally induced dimensional changes. Examples of these loadings are: - Indirect stresses in laminate corner radii, - Indirect stresses due to thickness changes, - Indirect stresses due to panel buckling deformations, - Direct stresses due to fuel pressure loads, or - Indirect stresses due to irregular load paths. Interlaminar damage growth is a potential location for fatigue. The “static notch” effect limits in plane loading for graphite composite laminates. This strength reduction due to open holes and impact damage can be appreciated by a review of “material allowable” data for a contemporary material system, Figure 3.

Figure 3. Typical In-plane Strength Scaling Effect for Notch & Compression After Impact, CAI for Hexcel 8552 IMF7, Ref [3] The laminate material property data was generated with FAA oversight through an FAA Special Project Number SP4652WI-Q and meet the requirements of NCAMP Standard Operating Procedure NSP 100 and MIL-HDBK-17-1F—Composite Materials Handbook for Polymer Matrix Composites. The notional test environments are: - CTD, Cold Temperature Dry (−65°F) Page 2 of 13

- RTD, Room Temperature Dry (75°F) - ETD, Elevated Temperature Dry (250°F) - ETW, Elevated Temperature Wet (250°F) Open hole in compression loading, OHC, at RTD and ETW have similar strength reductions as compression after impact, CAI, at RTD. For strength critical locations these values are used to establish the Design Ultimate Loading, DUL, conditions, typically open hole in compression or Compression after Impact. This means that the Maximum Fatigue Spectrum loads are between ¼ to 1/5 th of the nominal undamaged static strength. A variety of laboratory damage growth studies have demonstrated that stable slow damage growth is minimized at typical in-plane Design Limit Load, DLL, conditions. Figure 4 drawn from Ref [4] illustrates the need for high in-plane fatigue stress levels (more than 0.75 of the compression strength after impact in static – CAI – in this case) to obtain damage growth. Figure 4. C-scan measured delaminated area in repeated compression on a T800H/F665-2 composite The T800/F655-2 material was impacted at 6 joules. Composite laminates exposed to low-velocity impact may sustain extensive internal damage, delamination and resin cracking, without visual signs of damage on the impacted surface, Figure 2. This internal damage can potentially cause significant reduction in the strength of the laminate. The 6 joules (4 ft-lbs) impact is a typical damage threat level that can be monitored by visual inspection for barely visible impact damage (BVID). This impact energy level is typical of CIVIL transport airframes events. Constant amplitude fatigue testing was performed at various ratios of compression-compression testing, R = 10 after impact (CAI). High constant amplitude stress values (>0.75 CAI) are required to obtain damage extension. A nearly nogrowth behavior is observed below 0.75 CAI. Three impact threat distributions; low (4 ft-lbs), medium (6 ft-lbs) and high (15 ft-lbs) were defined in Ref [5] based on fleet damage surveys. The medium threat is utilized in MIL damage tolerance design requirements as a conservative estimate of impacts received by structural areas exposed to both operational and maintenance induced impact damage, JSSG-2006 Ref [6.a]. FULL SCALE FATIGUE TESTS (FSFT) AND DAMAGE TOLERANCE TEST (METALLICS & COMPOSITES); in November 1958 the (USAF) initiated the Aircraft Structural Integrity Program (ASIP) using a “ safe life” probabilistic approach with reliance upon the results of a laboratory test of a full-scale airframe with a safe life ”Scatter Factor (SF)” for the substantiation of metallic airframes. At that time, commercial aircraft as well as some military transport aircraft were designed to the commercial aviation regulation, CAR 4b.270, fail-safe requirements that supplemented CAR 4b.316, Ref [6.b]. Manufactures had the choice of either “safe life,” or ”fail safe”. CAR 4b.316 relied on safelife approach (safety-by-retirement) that retired the structure before the fatigue life is exhausted (e.g., setting life limits based on “safe-life” fatigue test duration determined by a Scatter Factor, (SF). CAR 4b.270 fail-safe approach relied on obvious detection of fatigue damage, and load path redundancy to avoid catastrophic failures. The fail-safe approach was considered superior to safe-life and easier to implement– No Full-Scale Fatigue Testing Required. Preferred strategy for majority of civil transport category airplanes certified in '60s and '70s was Fail-safe without a requirement for a fatigue test -- it was simpler to implement. At that time fail-safe certified airplanes where considered to have an Page 3 of 13

indefinite life. The early USAF design concept implemented the ”Safe-life“ option since most combat aircraft were designed with many single load path structures. Early composites applications for removal structure (empennage, flight control surfaces and doors) where also certified using a “safe Life” SF = 90% probability with a 95% confidence. At the beginning of the 1970’s metallic airframes in both the MIL and CIVIL aircraft fleets experienced significant cracking and fatigue damage. The United States Air Force (USAF) developed a damage tolerance philosophy to help eliminate the type of structural failures and cracking problems that had been encountered on various military aircraft. The Air Force review of structural failures had revealed that the safe life philosophy did not protect against designs that were intolerant to defects that could be introduced during manufacturing or during in-service use, Figure 1. As the 1970’s USAF Damage Tolerance preparatory work was evolving the then emerging F-16 structural requirements implemented Damage Tolerance policy. A composite skin bonded on to a titanium cruciform internal member was proposed for the pivoting horizontal stabilizer. The USAF was looking for a method to monitor, and truncate, manufacturing defects including the quality of structural adhesive bonding in a primary load path. NDI technology and analysis of damage growth at bonded interfaces was NOT sufficiently mature at that time. Proof test integration was selected to assure initial quality and a useful operational life capability. In support of this approach the Wearout Model (Halpin et al 1973, Ref [7]) using the power-law damage growth model was formulated to demonstrate adequate residual strength for an operational life interval: - Retention of residual strength experiencing operational usage (a fatigue test), and - Proof-test truncation using the power-law damage growth model to assure a minimum predictable life before inspection and maintenance. This procedure was implemented for the F-16A models. Multiple proof test failures in the vicinity of the DLL resulted in a redesign of the horizontal stabilizer. 1972 USAF formally made the damage tolerance approach a part of ASIP with the publication of MIL STD-1530A implemented originally through MIL-A-83444, Airplane Damage Tolerance Requirements. The Air Force now implements damage tolerant design through the recommended practices of the Department of Defense Joint Services Specification Guide, JSSG-2006 (1998) Ref [8];  Single load path protecting against structural failure using the damage tolerance (DT) concept of slow crack growth and DT-based inspections.  Fail-safe design combined with damage tolerance analysis and FSFT tests protect structural safety of civilian and military transport aircraft.  Primary focus was on metallic structure.  Implemented a 2 Life Time FSFT fatigue test + addition test time for metallic and/or composite DT demonstrations.  Safe-life approach was de-emphasized. The evolution of the damage tolerance was also taking place in the CIVIL transport community. In 1978 AC 25.571-1 implements damage tolerance. Fail-safe design options were the predominant approach in the 1960s and 1970s. The Level-of –Validity LOV concept was introduced into the CIVIL guidance in 2010. In the CIVIL Transport guidance for composites a “No-growth,” Damage Tolerance – Safe Life fatigue test and scatter factor was introduced in 1984, AC 20-107A. As the USAF was preparing for the B-2 bomber and the US Navy was evaluating fighter aircraft applications impact damage assessments where made to understand the accidental impact threat Page 4 of 13

environment. Field surveys of low-velocity impact damage where performed initially from four different in-service aircraft types (F-4, F-111, A-10, and F-18), Ref [5] and Ref [9]. The methodology employed in these assessments is summarized in Figure 5. Figure 5. Logic for Low-Velocity Impact Damage Surveys, Assessments and Design Criterion, (1980 – 2000). The basic assumption behind the surveys is that the sources of damage, the impact damage threats, induced during aircraft operations and/or induced during maintenance is independent of a specific vehicle. That is the concept for the term discrete threats. They are typical of the working environment for airframes. The consequence of the damage does depend on the specific material characteristics and construction details. The interpretation, translation, of the observed impact dents, etc. into design criterion for composite structure is summarized in Figure 5. Similar studies where undertaken in the CIVIL fleets. The output of these studies have provided the basis for the both CIVIL and MIL requirements for Airworthiness and the Continuing Airworthiness inspection and maintenance, the MSG-3 protocol, Ref [10] The ability to quantify the basic damage and material defects is fundamental to the implementation of a Damage Tolerance policy. Returning to Figure 1, examples of the damage and inherent defect surveys are illustrated in Figure 6. Figure 6. Teardown Inspections of Metallic Airframe Fatigue Test Articles, Damage Surveys of Inservice Aircraft and Airframe Structural Components Removed from Aircraft for Cause Provide Basic Data For DT, WFD & WSD, Refs [11], [12] and [13]. The magnitude of the risk for a structural failure in a fleet of aircraft is heavily influenced by the sizes of the growing crack or defect population at a given location and the capability of the nondestructive inspection (NDI) system used to detect defects or cracks at that location. For Metallic airframes the Equivalent Flaw Size (EFS) distribution is the description of the population of cracks that are representative of a critical location in the structure at a given time. This data is normally obtained during teardown inspections of either fatigue test articles or in-service aircraft. Fracture mechanics can be used to translate the cracks found to time zero to obtain the Equivalent Initial Flaw Size (EIFS) distribution or to a common flight hour that minimizes translation to obtain the EFS distribution, Ref [11] and Ref [14]. The results illustrated in Figure 6 demonstrate the significant differences in the EFS that have been characterized for several USAF aircraft constructed from different alloys and the importance of establishing the proper distribution for the aircraft that is analyzed for WFD. A similar characterization is evolving for carbon fiber - epoxy composites in terms of inspectable dent depth, refs [5,9,12] as illustrated in Figure 6. The NDI reliability for both visual and instrument detection is qualified to 90% probability with a 95% confidence of detection. This criterion Ref [8] is utilized for both composite and metallic airframe structure Probability of Detection, POD. The reason for this criterion is economic. It was judged that the sample size required for a more conservative reliability would require sample sizes of such a magnitude that it would not be sustainable in practice. The quantification of a protocol for accidental damage and defection detection provides the basis for the relationship between damage size, damage awareness and residual strength in AC20-107B, Fig 3. Page 5 of 13

The Early Debates For The Damage Tolerance FSFT And Inspection Criterion are important for an understanding of the current design and substantion approaches. The MIL ASIP implemented a deterministic approach using “Rogue Flaws” to define inspection intervals, supplemented with fracture mechanics (measurable defect, growth and residual Life prediction) and the formal implementation of the FSFT requirement. The legacy of the “safe life” approach was debated for application in the damage tolerance concept: - What should be the duration for the FSFT, a Design Service Goal (DSG) x Scatter Factor (SF)? - What should be the Scatter Factor, the “Safe Life” reliability Scatter Factor, SF, (99%prob/95conf)? - What is the inspection reliability, a POD of 90%prob/95% confidence of detection (practical implementation)? - Should a Damage Tolerance SF = 2 (with reoccurring in-service inspection at ½ of the residual life) be employed versus the reliability based “safe life” SF? In the 1970 it was understood that there are three major limitations, Ref [8] associated with the POD inspection criterion, characterization: 1) The choice of particular POD and confidence limits was made on a rather arbitrary economic basis. For example, 90/95 values were selected for JSSG-2006 recommended crack sizes even though there is no real interest in a crack length that is detected only 90 percent of the time. Rather, 90/95 limits were selected because higher POD or confidence limit values would have required much larger sample sizes in the demonstration programs for the analysis methods being used. The 95 percent confidence limit was assumed to provide the required degree of conservatism. 2) A POD limit defect size is not a single, uniquely defined number but, rather, is a statistical or random quantity. Any particular POD estimate is only one realization from a conceptually large number of repeats of a demonstration program. 3) The POD characterization is not related to the size of defects (cracks) that may be present in the structure after an inspection. To calculate the probability of missing a large defect or crack requires knowledge of the POD for all cracks sizes and the distribution of the sizes of the defects or cracks in the population of structural details being inspected. Given the economic decision to implement the POD reliability of 90/95% a strategy was required to assure a conservative application for the damage tolerance approach. Limitation 2 above was resolver with a multiple reoccurring inspection strategy, figure 7. Figure 7 Illustration of an Early Rational for the (SF)DT = 2 A required period of unrepaired service usage was selected as two service lifetimes demonstrated with a FSFT. The Scatter Factor of two was selected to cover various uncertainties associated with damage &/or crack growth during service usage, variability in material properties, manufacturing quality and inspection reliability. Reoccurring in-service inspections at increments of ½ demonstrated & expected life would accommodate variations in inspection reliability. The safety goal is to achieve equal or better that a 99% reliability for a FSFT demonstration. The 2 nd column is a notional in-service inspection schedule based on a 1 DSG FSFT. The 3 rd column is a notional in-service inspection schedule based on a 2 DSG FSFT. The 4th column is the notional probability of missing a defect of a defined size. It Page 6 of 13

became clear that a safety management system based on inspection required the FSFT duration of 2 service lives. The selection of a DT SF of 2 for both the duration of the FSFT and the in-service inspection approach is essential for the implementation of a Slow Damage Growth Damage Tolerance concept – Safety by Inspection. This change to a DT SF =2 changed the initial 1970’s concept of a modified Safe Life concept with a SF reliability of 90%prob/95%conf that was developed in the Wear-out model, Ref [7]. There was resistance in the Composite community to Damage Tolerance concepts. In addition, for valid reasons, the US Navy wanted to retain a “safe Life” approach as they believed that the DT inspection concepts where not practical for operations on-board aircraft carriers. The US Navy prefers “safe Life” minimizing aircraft carrier workload. Technology advocates within the USAF did not support the implementation of the damage tolerance concepts preferring a “reliability” approach. In 1981 Sendecky, Ref [15] removed power law damage growth from 1973 Wearout model Ref [7] to blunt the adoption of damage growth modeling for composite and bonded structure. Whithead et al in the 19861997 time frame developed an approach to implement a Safe-life reliability demonstration for composite structures. This approach was implemented using the AC 25.571 Safe-life fatigue demonstration requirements with an appropriate reliability scatter factor. Safe-life composite material Scatter factors Ref [16] for Interlaminar fatigue sensitivity for: - A level reliability, (SF)A ≅ 42, or - B level reliability, (SF)B ≅5. Because a FSFT of 42 DSG’s is impractical the Safe Life B level reliability was selected for substantion. Having accepted a “Safe-life” design and substantion approach the challenge was to find a way to perform the FSFT in on a sensible schedule at an affordable cost. The safe-life commitment generated the need for accelerating FSFT using Load enhancement (LEF) see CMH-17F. A simplified LEF approach using damage growth power law data for resins and adhesives is available in Ref 16. Figure 8 is a notional description of current inspection approaches, MIL and CIVIL, for composite airframes. This figure illustrates an integration of the AC20-107B Damage Tolerance guidance with the MSG-3 protocol. This protocol uses the MSG-3 damage tolerance concepts with emphases on monitoring accidental service induced damage.

Figure 8 Inspection Intervals are a Function of Damage Severity and Inspection Capabilities (Notional) * For MIL transports Nearly No Damage Growth (&/or slow crack growth structure), the required period of unrepaired service usage is two service usage lifetimes, modified for Fail-Safe structure. Transport aircraft use the A (FH/cycles), B (TBD months), C (TBD yrs.) & D (TBD yrs.) check system to phase the DT inspections [per (ATA) Maintenance Steering Group’s MSG-3 guidance].

The composite and/or hybrid (metallic and composite) airframe has available 3 basic approaches: - Single load path protecting against structural failure using the damage tolerance (DT) concept of slow crack growth or “nearly-no-growth” and DT-based inspections;  FSFT with SF ≅2  Typical (average) fleet usage for the FSFT  Common inspection MSG-3 damage tolerance approach Page 7 of 13

-

Single load path protecting against structural failure using a reliability based “safe-life” “nearly-no-growth” damage tolerance;  FSFT with SF ≅B-level reliability  Load enhancement to accelerate test (interlaminar damage growth)  Average to Aggressive (USNav) usage  Inspection strategy – hybrid safe life and MSG-3 monitoring accidental damage

- Fail-safe design combined with slow crack growth damage tolerance analysis and inspection concept;  2 Fail-safe options  Multiple load path  Damage arrest  FSFT with SF ≅2  Typical (average) fleet usage  Common inspection damage tolerance MSG-3 approach There continues to be competition between the different design and substantiation concepts: - “No damage growth” but do not call it safe life versus “nearly no damage growth” as a subset of damage tolerance slow damage growth, - The “requirement“ for a safe life LEF, and the relationship to FSFT duration and inspection intervals, - Tolerance to “wide area damage” but don’t use the term “fail-safe,”and - MSG-3 inspection intervals and concept focus is a WSD perspective; Ref [10] sect 2-4-2 Scheduled Structural Maintenance verses the WFD civil guidance. ESTABLISHING OPERATIONAL LIFE LIMITS FOR USAF AIRFRAMES AND FAIL-SAFETY; the MIL requirements have partially resolved the tolerance to “wide area damage” and the use of the term “fail-safe,” When the MIL system formally introduced damage tolerance requirements in July of 1974 the policy was to allow the use of either fail-safe or slow crack growth (damage) design concepts. The primary focus was on the slow crack growth concept since most combat aircraft were designed with many single load path structures. At that time, commercial aircraft as well as some military transport aircraft were designed to the commercial aviation regulation CAR 4b.270, fail-safe requirements. In the 1970’s there were 2 catastrophic failures of “fail-safe” airplanes: B748 wing separation (1976) and B707 horizontal stabilizer separation (1977) Ref 6.b. The shortcomings of this requirement were recognized, it didn’t focus on the inspection for defects, the retention of adequate residual strength or address the issues of continuing damage in adjacent structure, including safe periods of unrepaired usage after a load path failure. Fail-safety can be jeopardized later in life by the onset of WSD – WFD is a subset of WSD. Fail-safe design combined with Slow Damage (Nearly No) Growth fault tolerance is the preferred concept since it provides large damage capability and is the only concept that protects against all forms of damage an aircraft may encounter in its lifetime. These include fatigue cracking, corrosion, accidental damage, delamination, disbonding, manufacturing defects, discrete source damage, environmentally age induced and maintenance-induced damage. Because of the difficulties of implementing the fail-safe structure damage tolerant design Page 8 of 13

requirements aircraft manufacturers used (and received approval for using) the single load path structure damage tolerant design requirements for these structural elements. The unintended consequence of this action was that the airframe manufacturers no longer designed their structures to be fail-safe. Three structural bulletin where issued in 2011 to address this topic Ref [17]. The purpose of Ref [17.a] is to revise the fail-safe requirements to encouraged fail- safe design and certification of future Air Force aircraft. Additionally, this structural bulletin provides the criterion for the determination of fail-safe operational life limits for fail-safe design concepts. Requirements for slow crack growth designs have also been revised and are provided in Ref [17.b]. Fail-safe assessments of current aircraft to identify those Safety of Flight locations that have inherent fail-safe capability are covered in the structural bulletin Ref [17.c]. Fail-safe assessment requirements for current aircraft are used to validate the operational life limits from the results of teardown inspections of high usage aircraft and, if necessary, to impose flight restrictions (including possible grounding) until the structure is modified or replaced, or the aircraft is retired. This is the preferred USAF method for establishing Operational Life Limits. The MIL requirements have partially resolved the concerns for tolerance to “wide area damage” in metallic structures. A potential “wide area damage” composite challenge is a discrete source damage event next to adjacent structure damaged by BVID and the ability to sustain the load transfer process. That is the difference and the challenge in the WSD versus a WFD perspective. ESTABLISHING AN OPERATIONAL LIMIT FOR COMPOSITE; AC 120-104 Establishing and Implementing Limit Of Validity (LOV) to Prevent Widespread Fatigue Damage, WFD, is intended to protecting residual strength from gradual material deterioration subjected to operational usage & environments. It outlines a process to achieve this objective, Figure 9. The LOV is the period of time (in flight cycles, flight hours, or both), up to which it has been demonstrated that WFD is unlikely to occur in an airplane’s structure by virtue of its inherent design characteristics and any required maintenance actions. In the Figure 9 side comments there is some uncertainty regarding the Structural Modification Point and Inspection Start Point indicated by the question mark symbols. Figure 9 MSD/MED Residual Strength Curve (Including WFD Inspections) MSD is multiple site damage and MED is multiple element damage due to wide area cracking in a metallic airframe. In this context, the questions raised are valid for all structures, however the AC 120-104 guidance is limited to a specific mode of damage, through thickness cracking for metallic structure, not the WSD concept for other sources of damage as denoted in Figure 1. It is correct to ask when should, or how to recognize, a Structural Modification or Replacement Point for either a composite or hybrid airframe. Stated another way, what is the equivalent process for WSD risk for a Composite or Hybrid airframe structure? There are options for assessing Operational Limits and WSD Inspection Thresholds for Affected Components either Composite or Metallic Construction by focusing the AC 120-104 Options using the MSG-3 Age Exploration Approach Ref [10]. Would additional Full Scale Fatigue Testing provide additional WSD information concerning the Environmental and Corrosion Durability issues experienced in bonded metallic and/or composite structure and structural elements? How would an additional Full Scale Fatigue Test Page 9 of 13

develop an understanding concerning the accumulation of Accidental Damage over time? OR Would an alternative aging lead-the-fleet approach supplementing the initial FSFT be more effective involving: a Supplemental Structural Inspection Program, SSI, (to include in-service impact damage, High Energy Wide Area Blunt Impact with ground equipment, repair accumulation for composites and age related environmental damage); a revalidation of service loads and the usage fatigue loads spectrum to access if the “nearly no growth” damage tolerance has been compromised; and Lead-the-fleet teardown of several aging airframes? Would the laboratory FSFT be a better option than teardown inspections looking for degradation and damage not readily evident and potentially would not be captured by existing operational inspections? The lead-the-fleet approach could be implemented with a modification of AC 120-93, “Damage Tolerance Inspections for Repairs and Alterations” for Bonded Repairs & 14 CFR Part 26, Subpart E Aging Airplane Safety—Damage Tolerance Data for Repairs and Alterations (§ 26.43 — Holders of and applicants for type certificates—Repairs.). The Scheduled Maintenance Development guidance of the MSG-3 Maintenance Steering Group [10] was first published in 1980. In the 1993 revision procedures to determine the appropriate scheduled maintenance requirements for COMPOSITE STRUCTURE were introduced. The essential elements are: - Top-down approach focusing on ‘consequences of failure’ - Common sources of accidental damage for metallic & composites material structures. - Recognition of ‘damage tolerance rules’ and Supplemental Inspection Programs, (SIP’s) Section 2.4, “Aircraft Structural Maintenance Program Development,” develops scheduled maintenance tasks based on susceptibility to any form of damage, and the degree of difficulty involved in detecting such damage. Requirements for detecting; Accidental Damage (AD), Environmental Deterioration (ED), Fatigue Damage (FD), and procedures for preventing and/or controlling corrosion (CPCP) are defined. Inspection Thresholds and repeat inspection intervals (DT or Safe Life approaches) and Zonal inspection guidance and a rating system for accidental and other damage (Structural Significant Item, SSI). Section 2.4.3.1.a “Damage Sources” identifies Accidental Damage (AD), which is characterized by the occurrence of a random discrete event that may reduce the inherent level of residual strength. Sources of such damage include ground and cargo handling equipment, foreign objects, erosion from rain, hail, lightning, runway debris, spillage, freezing, thawing, etc., and those resulting from human error during aircraft manufacture, operation or maintenance that are not included in other damage sources. The same sources of accidental damage as those considered for metallic materials are to be considered for nonmetallic material such as composites. The consequence of damage may not be readily apparent and may include internal damage, e.g., disbonding or delamination. Large size accidental damage, such as that caused by engine disintegration, bird strike or major collision with ground equipment, will be readily detectable and no maintenance task assessment is required. Environmental Deterioration (ED) analysis for non-metallic SSI is no longer limited to corrosion. It is more generic recognizing that deterioration could occur after a period of time due to specific environmental factors. Page 10 of 13

The Age Exploration program task is a systematic evaluation of structural items based on analysis of collected information from in-service experience. It documents and verifies the individual item's resistance to a deterioration process with respect to increasing age. Airframes are constructed from a combination of materials that are provided by a variety of suppliers utilizing different processing technologies. While there will be some generic conclusions, most likely operational limits will be specific to individual airframe model designs. The Airworthines Limitations Section is the natural repository for model specific operational limitations. SUMMARY Several modes of Discrete Source Damage, Accidental Damage & Repairs will accumulate in number, and probably in magnitude over time as an airframe ages. There is a consistent record of a gradual increase in loads over a typical airframe life (≈ 1.07% per year MGTOW, MIL Transports) - this may suggest that a verification of the “nearly no-growth” capability is appropriate. Many of the “durability” associated Damage Tolerance topics would be well managed under the LOV concept provided the emphasis on the metallic WFD is put into a WSD perspective. There is a potential interlaminar fatigue sensitivity - that may increase as multiple impact events & repairs accumulate over time combined with a gradual increase in the operational loads. This sensitivity is potentially more subtitle than the through thickness cracking of metallics. Environmental Deterioration (ED) for non-metallic Structural Significant Items SSI is not limited to corrosion. It is more generic recognizing that deterioration could occur after a period of time due to specific environmental factors. An Age Exploration program can verify resistance to a deterioration process with respect to increasing age Encourage a Fleet Leader Age Exploration program that would include Damage Surveys, Bonded Repairs & Tear-down Inspections to understand and document aging composite airframe events affecting WSD Operational Limits and age related Environmental Deterioration. The function of teardown inspections is to look for degradation and damage not readily evident and potentially would not be captured by existing inspections, as defined in inspection schedules either defined by Certification or MRB (Maintenance Review Board) activities. These assessment activities could be implemented through Part 26 & MSG-3 tasking. REFERENCE Ref 1; E.K. Walker, J.C. Kevel, and J.E. Rhodes (1979), "Design for Continuing Structural Integrity. “Structural Integrity Technology,” ASME, NY. Ref 2; Paul, P C, Staff, C R, Sanger, K B, Mahler, M A, Kan, H P, and Kautx, E F, “Analysis and Test Techniques for Composite Structures Subject to Out-of-Plane Loads,” Composite Materials Testing and Design (Tenth Volume), ASTM STP 1120, Glenn C Grimes, Ed, ASTM, Philadelphia, 1992, pp. 238-252; Kedward, K. T., Wilson, R. S. and Mclean, S. K., Flexure of Simply Curved Composite Shapes, Composites. Volume 20. Number 6. November 1989. Ref 3; Clarkson, E, Hexcel 8552 IM7 Unidirectional Prepreg – Qualification Statistical Analysis Report, NCP-RP-2009-028 Rev N/C, (June 2011); and Marlett K, Hexcel 8552 IM7 Unidirectional Page 11 of 13

Prepreg - Qualification Material Property Data Report, CAM-RP-2009-015 (April 2011); National Institute for Aviation Research, Wichita State University. Ref 4. Whitehead, R.S. et al.; Northrop Corporation - Certification Testing Methodology for Composite Structure” Vol. 1: Data Analysis and Vol. 2: Methodology Development; Report NADC-87042-60; 1986 and Rouchon J, How, Over the Past 30 Years, “Part 25” Composite Structures Have Been Coping With Metal Minded Fatigue and Damage Tolerance Requirements, 24th ICAF Symposium, Naples, Italy, May 2007 Ref 5. Kan, H.P., Cordero, R, and Whitehead, R.S; Advanced Certification Methodology for Composite Structures, DOT/FAA/AR-96/111, April 1997 Ref 6.a) JSSG 2006, Aircraft Structures, Oct 1998; and MIL-HDBK-516C, Airworthiness Certification Criteria, Dec 2014; b) Fuselage Structural Integrity Forum –Historical Perspective of Fatigue Requirements, Sept. 201; http://www.ntsb.gov/news/events/Documents/fuselage_forum-1.2%20FAA-Panel%201.2-Final.pdf Ref 7. Halpin, J.C., Jerina, K. L., Johnson, T. A., Characterization of Composites for the Purpose of Reliability Evaluation, AFML-TR-77-289, WPAFB, Ohio, 1972 and in Analysis of Test Methods for High Modulus Fibers and Composites, ASTM STP 521, American Society for Testing Materials, 1973, pp. 5-64, Halpin JC, Johnson T A, Waddups M E. Kinetic fracture models and structural reliability. Int. J Fract. Mech. 1972; 8: 465–8. Ref 8, Miedlar P C, Berens A P, Gunderson A, and Gallagher J P, AFRL-VA-WP-TR-2003-3002, Analysis And Support Initiative for Structural Technology (ASIST), Nov. 2002 Ref 9, Labor, J.D., Service/Maintainability of Advanced Composite Structures, Report No. AFFDLTR-78-155, November 1978; Labor, J.D., Design Criteria Guidelines for the Service/Maintainability of Advanced Composite Structures, Report No. AFFDL-TR-78-156, November 1978; Butler, B.M., Wing Fuselage Critical Component, Preliminary Design (Northrop), Report No AFFDL-TR-78-174, March 1979; Fiscus, I.B., and Watson, D.C., Study of Impact Damage to B-1B Weapon Bay Doors, Report No. AFWAL-TR-84-4153, December 1984; Wood, H.A., and Bon, T.J., Advanced Composites Supportability Working Group Findings and Recommendations, Report No ASD(ENF)-TR-83-5017, January 1984; US Navy Survey of Damage and Defects in Advanced Composite Components in Service, Report No. TTCP/HTP-3, September 1985; Stone, R.H., Repair Techniques for Graphite/Epoxy Structures, NASA Report No CF 159056, January 1983; Cook, T.N., Adami, M.G., DiGenova, R.R., and Maass, D.P., 1985; Stone, R.H., Repair Techniques for Graphite/Epoxy Structures, NASA Report No CF 159056, January 1983; Cook, T.N., Adami, M.G., DiGenova, R.R., and Maass, D.P., Advanced Structures Maintenance Concepts, Report No. USAAVRADCOM-TR-80-D-16, June 1980; Ramkumar, R.L., Composite Impact Damage Susceptibility, Report No.

NADC-79068-60, January 1981. Ref 10. ATA MSG-3 Volume 1 (Fixed Wing Aircraft), Operator / Manufacturer Scheduled Maintenance Development, Revision 2013.1 Ref 11. Gallagher J P, Babish C A, & Malas J C, Damage Tolerant Risk Analysis Techniques for Page 12 of 13

Evaluating the Structural Integrity of Aircraft Structures, ICF 11, May 2005; Tiffany C. F., Gallagher J. P., and Babish C. A., IV, Threats To Aircraft Structural Safety, Including A Compendium Of Selected Structural Accidents / Incidents, ASC-TR-2010-5002 Ref 12. Composites @ Airbus; Damage Tolerance Methodology, FAA Workshop DT & Maintenance; Chicago, July 2006 and Ratier L. and Fualdes C., Airbus Composite Fatigue and Damage Tolerance Certification Experiences, 2015 September FAA/Bombardier /TCCA/EASA/Industry Composite Transport Damage Tolerance and Maintenance Workshop (Montreal, Quebec) Ref 13. Schmidt H J, Schmidt-Brandecker B, Tober G, Design of Modern Aircraft Structure and the Role of NDI, NDT.net - June 1999, Vol. 4 No. 6 Ref 14. Johnson W S, The History, Logic and Uses Of The Equivalent Initial Flaw Size Approach To Total Fatigue Life Prediction, Procedia Engineering 2 (2010) 47–58. Ref 15 Sendeckyj GP. Fitting Models To Composite Materials Fatigue Data. Test Methods And Design Allowables For Fibrous Composites, In: Chamis CC (ed) ASTM STP 734, Philadelphia, PA: American Society for Testing and Materials, 1981, 245-260. Ref 16 Halpin J C, ”Review of Damage Growth in Resins, Adhesive & Composite” CMH-17 Meeting, Salt Lake City, March 15- 18, 2015 Ref 17 a) JSSG-2006 Structures Bulletin, EN-SB-08-001, Revision A, Revised Damage Tolerance Requirements and Determination of Fail-Safety Life Limits for Fail-Safe Metallic Structures, 18 March 2011; b) JSSG-2006 Structures Bulletin, EN-SB-08-002, Revision A, Revised Damage Tolerance Requirements and Determination of Operational Life Limits for Slow Crack Growth Metallic Structures, 18 March 2011; c) JSSG-2006 Structures Bulletin, EN-SB-08-003, Revision A, Fail-Safe Assessments of Current Aircraft, 18 March 2011

Page 13 of 13