4.1 The goal of the NDT is to detect defects that have been implicated in the failure of the COPV metal liner, or have led to leakage, loss of contents, injury, death, or mission, or a combination thereof. Liner defects detected by NDT that require special attention by the cognizant engineering organization include through cracks, part-through cracks, liner buckling, pitting, thinning, and corrosion under the influence of cyclic loading, sustained loading, temperature cycling, mechanical impact and other intended or unintended service conditions.

Note 3: Liners made from stainless steel and nickel-based alloys exhibit a higher damage resistance to impact than those made from aluminum.

Note 4: Safe life is the goal for any COPV so that a through crack in the liner will not develop during the service life.

Note 5: The use a material with good fatigue and slow crack growth characteristics is important. For example, nickel-based alloys are better than precipitation-hardened stainless steel. Aluminum also has good ductility and crack resistance.

4.2 The COPVs covered in this guide consist of a metallic liner overwrapped with high-strength fibers embedded in polymeric matrix resin (typically a thermoset). Metallic liners may be spun formed from a deep drawn/extruded monolithic blank or may be fabricated by welding formed components. Designers often seek to minimize the liner thickness in the interest of weight reduction. COPV liner materials used can be aluminum alloys, titanium alloys, nickel-chromium alloys, and stainless steels, impermeable polymer liner such as high density polyethylene, or integrated composite materials. Fiber materials can be carbon, aramid, glass, PBO, metals, or hybrids (two or more types of fiber). Matrix resins include epoxies, cyanate esters, polyurethanes, phenolic resins, polyimides (including bismaleimides), polyamides and other high performance polymers. Common bond line adhesives are generally epoxies (FM-73, West 105, and Epon 862) or urethanes with thicknesses ranging from 0.13 mm (0.005 in.) to 0.38 mm (0.015 in.). Metal liner and composite overwrap materials requirements are found in ANSI/AIAA S-080 and ANSI/AIAA S-081, respectively. Pictures of representative COPVs are shown in Guide E2981.

4.3 The operative failure modes COPV metal liners and metal PVs, in approximate order of likelihood, are: (a) fatigue cracking, (b) buckling, (c) corrosion, (d) environmental cracking, and (e) overload.

Note 6: For launch vehicles and satellites, the strong drive to reduce weight has pushed designers to adopt COPVs with thinner metal liners. Unfortunately, this configuration is more susceptible to liner buckling. Therefore, as a precursor to liner fatigue, attention should be paid to liner buckling.

4.4 Per MIL-HDBK-340, the primary intended function of COPVs as discussed in this guide will be to store pressurized gases and fluids where one or more of the following apply:

4.4.1 Contains stored energy of 19 310 J (14 240 ft-lbf) or greater based on adiabatic expansion of a perfect gas.

4.4.2 Contains a gas or liquid that would endanger personnel or equipment or create a mishap (accident) if released.

4.4.3 Experiences a design limit pressure greater than 690 kPa (100 psi).

4.5 Per NASA-STD-(I)-5019, COPVs should comply with the latest revision of ANSI/AIAA S-081. The following requirements also apply when implementing S-081:

4.5.1 Maximum Design Pressure (MDP) should be substituted for all references to Maximum Expected Operating Pressure (MEOP) in S-081.

4.5.2 COPVs shall have a minimum of 0.999 probability of no stress rupture failure of the composite shell during the service life.

Note 7: For other aerospace applications, the cognizant engineering organization should select the appropriate probability of survival, for example, 0.99, 0.999, 0.9999, etc., depending on the anticipated failure mode, damage tolerance, safety factor, or consequence of failure, or a combination thereof. For example, a probability of survival of 0.99 means that on average, 1 in 100 COPVs will fail. COPVs exhibiting catastrophic failure modes (BBL composite shell stress rupture versus LBB liner leak), lower damage tolerance (cylindrical versus spherical vessels), lower safety factor, and high consequence of failure will be subject to more rigorous NDT.

4.6 Application of the NDT procedures discussed in this standard is intended to reduce the likelihood of liner failure, commonly denoted leak before burst (LBB), characterized by leakage and loss of the pressurized commodity, thus mitigating or eliminating the attendant risks associated with loss of the pressurized commodity, and possibly mission.

4.6.1 NDT is done on fracture-critical parts such as COPVs to establish that a low probability of preexisting flaws is present in the hardware.

4.6.2 Per the discretion of the cognizant engineering organization, NDT for fracture control of COPVs should follow additional general and detailed guidance described in MIL-HDBK-6870, NASA-STD-5019, MSFC-RQMT-3479, or ECSS-E-30-01A, or a combination thereof, not covered in this guide.

4.6.3 Hardware that is proof tested as part of its acceptance (that is, not screening for specific flaws) should receive post-proof NDT at critical welds and other critical locations.

4.7 Discontinuity Types—Specific discontinuity types are associated with the particular processing, fabrication and service history of the COPV. COPV composite overwraps can have a myriad of possible discontinuity types, with varying degrees of importance in terms of effect on performance (see 4.7 in Guide E2981). As for discontinuities in the metallic liner, the primary concern from an NDT perspective is to detect discontinuities that can develop cracks or reduce residual strength of the liner below the levels required, within the context of the life cycle. Therefore, discontinuities should be categorized as follows:

4.7.1 Inherent material discontinuities: inclusions, grain boundaries, etc., detected during (a) and (b) of 5.5.

Note 8: Inherent material discontinuities are generally much smaller than the damage-tolerance limit size. Any design that does not satisfy this statement should be revised. Quality control procedures in place in the manufacturing process should eliminate any source materials that do not satisfy specifications.

4.7.2 Manufacturing-induced discontinuities: caused by welding, machining, heat treatment, etc., detected during (b) and (c) of 5.5.

Note 9: Manufacturing-induced discontinuities depend on the manufacturing process, and can include machining marks, improper heat treatment, and weld-related discontinuities such as lack of fusion, porosity, inclusions, zones of local material embrittlement, shrinkage, and cracking.

4.7.3 Service-induced discontinuities: fatigue, corrosion, stress corrosion cracking, wear, accidental damage, etc. detected during (d) and (e) of 5.5 (after the COPV has been installed). In these cases, NDT should either be made on a “remove and inspect” or “in-situ” basis depending on the procedure and equipment used.

4.8 A conservative damage-tolerance life assessment is made by assuming the existence of a crack-like discontinuity or system of discontinuities, and determining the maximum size or other characteristic of this discontinuity(s) that can exist at the time the vessel is placed into service but not progress to failure under the expected service conditions. This then defines the dimensions or other characteristics of the crack or crack-like discontinuity or system of crack-like discontinuities that should be detected by NDT.

Note 10: Welding or machining may result in non-crack like flaws/imperfections/conditions that may be important, and NDT choices for these flaws/imperfections/conditions may be different than for crack-like ones.

4.9 Acceptance Criteria—Determination about whether a COPV meets acceptance criteria and is suitable for aerospace service should be made by the cognizant engineering organization. When examinations are performed in accordance with this guide, the engineering drawing, specification, purchase order, or contract should indicate the acceptance criteria.

4.9.1 Accept/reject criteria should consist of a listing of the expected kinds of imperfections and the rejection level for each.

4.9.2 The classification of the articles under test into zones for various accept/reject criteria should be determined from contractual documents.

4.9.3 Rejection of COPVs—If the type, size, or quantities of defects are found to be outside the allowable limits specified by the drawing, purchase order, or contract, the composite article should be separated from acceptable articles, appropriately identified as discrepant, and submitted for material review by the cognizant engineering organization, and given one of the following dispositions; (1) acceptable as is, (2) subject to further rework or repair to make the materials or component acceptable, or (3) scrapped (made permanently unusable) when required by contractual documents.

4.9.4 Acceptance criteria and interpretation of result should be defined in requirements documents prior to performing the examination. Advance agreement should be reached between the purchaser and supplier regarding the interpretation of the results of the examinations. All discontinuities having signals that exceed the rejection level as defined by the process requirements documents should be rejected unless it is determined from the part drawing that the rejectable discontinuities will not remain in the finished part.

4.10 Certification of PVs—ANSI/AIAA S-080 defines the approach for design, analysis, and certification of metallic PVs.

4.11 Certification of COPVs—ANSI/AIAA S-081 defines the approach for design, analysis, and certification of COPVs, while ANSI/AIAA S-080 defines the approach for design, analysis, and certification of PVs. More specifically, the PV or COPV thin-walled metal liner should exhibit a leak before burst (LBB) failure mode or shall possess adequate damage tolerance life (safe-life), or both, depending on criticality and whether the application is for a hazardous or nonhazardous fluid. Consequently, the NDT procedure should detect any discontinuity that can cause burst at expected operating conditions during the life of the COPV. The Damage-Tolerance Life requires that any discontinuity present in the liner will not grow to failure during the expected life of the COPV. Fracture mechanics assessment of crack growth is the typical approach used for setting limits on the sizes of discontinuities that can safely exist. This establishes the defect criteria: all discontinuities equal to or larger than the minimum size or have J-integral or other applicable fracture mechanics-based criteria that will result in failure of the vessel within the expected service life are classified as defects and should be addressed by the cognizant engineering organization.

4.11.1 Design Requirements—COPV design requirements related to the metallic liner are given in ANSI/AIAA S-080. The key requirement is the stipulation that the PV or COPV thin-walled metal liner should exhibit an LBB failure mode or should possess adequate damage tolerance life (safe-life), or both. The overwrap design should be such that, if the liner develops a leak, the composite will allow the leaking fluid (liquid or gas) to pass through it so that there will be no risk of composite rupture.

4.12 Probability of Detection (POD)—Detailed instruction for assessing the reliability of NDT data using POD of a complex structure such as a COPV is beyond the scope of this guide. Therefore, only general guidance is provided. More detailed instruction for assessing the capability of an NDT procedure in terms of the POD as a function of flaw size, a, can be found in MIL-HDBK-1823. The statistical precision of the estimated POD(a) function (Fig. 1) depends on the number of examination sites with targets, the size of the targets at the examination sites, and the basic nature of the examination result (hit/miss or magnitude of signal response).

FIG. 1 Probability of Detection as a Function of Flaw Size, POD(a), Showing the Location of the Smallest Detectable Flaw and a₉₀ (Left); POD(a) With Confidence Bounds Added and Showing the Location of a_90/95 (Right)

4.12.1 Given that a_90/95 has become a de facto design criterion, it is important to estimate the 90^th percentile of the POD(a) function more precisely than lower parts of the curve. This can be accomplished by placing more targets in the region of the a₉₀ value but with a range of sizes so the entire curve can still be estimated.

Note 11: a_90/95 for a metallic liner and generation of a POD(a) function is predicated on the assumption that critical initial flaw size (CIFS) for a liner of a given thickness can be detected with a capability of 90/95 (90 percent probability of detection at a 95 percent confidence level). This is problematic for COPVs with very thin metallic liners where the CIFS will be smaller than the minimum detectable flaw sizes given in Table 1 in NASA-STD-5009. At this limit of detection (CIFS < a_90/95), a_90/95 will have no validity for a thin-walled COPV.

4.12.2 NASA-STD-5009 defines typical limits of NDT capability for a wide range of NDT procedures and applications. Given the defect criteria established by the Damage-Tolerance Life requirements and the potential discontinuities to be detected, NASA-STD-5009 can be used to select NDT procedures that are likely to achieve the required examination capability.

Note 12: NDT of fracture critical hardware should detect the initial crack sizes used in the damage tolerance fracture analyses with a capability of 90/95. The minimum detectable crack sizes for the standard NDT procedures shown in Table 1 of NASA-STD-5009 meet the 90/95 capability requirement. The crack size data in Table 1 of NASA-STD-5009 are based principally on an NDT capability study that was conducted on flat, fatigue-cracked 2219-T87 aluminum panels early in the Space Shuttle program. Although many other similar capability studies and tests have been conducted since, none have universal application, neither individually or in combination. Conducting an ideal NDT capability demonstration where all of the variables are tested is obviously unmanageable and impractical.

4.12.3 Aspect Ratio and Equivalent Area Considerations—Current standards governing aerospace metallic pressure vessels (ANSI/AIAA S-080) and COPV liners (ANSI/AIAA S-081) require that fracture analysis be performed to determine the CIFS for cracks having an aspect ratio ranging from 0.1 to 0.5. However, there is insufficient data to support the approach of testing at only one aspect ratio and then using an equivalent area approach to extend the results to the required range of aspect ratios (1-9).20 Accordingly, POD testing on metallic COPV liners should be performed at the bounds of the required range of crack aspect ratios.

Note 13: Caution: To minimize mass, designers of aerospace systems are reducing the wall thickness for metallic pressure vessels and COPV liners. This reduction in wall thickness produces higher net section stresses, for a given internal pressure, resulting in smaller CIFS. These smaller crack sizes approach the limitations of current NDT. Failure to adequately demonstrate the capabilities of a given NDT procedure over the required range of crack aspect ratios may lead to the failure to detect a critical flaw resulting in a catastrophic tank failure.

4.12.4 To provide reasonable precision in the estimates of the POD(a) function, experience suggests that the specimen test set contain at least 60 targeted sites if the system provides only a binary, hit/miss response and at least 40 targeted sites if the system provides a quantitative target response, â. These numbers are minimums.

4.12.5 For purposes of POD studies, the NDT procedure should be classified into one of three categories:

4.12.5.1 Those which produce only qualitative information as to the presence or absence of a flaw, that is, hit/miss data,

4.12.5.2 Those which also provide some quantitative measure of the size of the target (for example, flaw or crack), that is, â versus a data, and

4.12.5.3 Those which produce visual images of the target and its surroundings.

4.12.6 Detailed POD Guidance—For detailed guidance on how to conduct a POD study, including system definition and control, calibration, noise, demonstration design, demonstration tests, data analysis, presentation of results, retesting, and process control plan, consult MIL-HDBK-1823.

4.12.6.1 For detailed guidance on how to conduct a POD study for ET, PT, and UT, consult MIL-HDBK-1823, Appendices A through D, respectively.

4.12.6.2 For detailed test program guidance; specimen design, fabrication, documentation, and maintenance; statistical analysis of NDT data; model-assisted determination of POD; special topics; and related documents, consult MIL-HDBK-1823, Appendices E through J, respectively.

4.13 NDT Data Reliability—MIL-HDBK-1823 provides nonbinding guidance for estimating the detection capability of NDT procedures for examining either new or in-service hardware for which a measure of NDT reliability is needed. Specific guidance is given in MIL-HDBK-1823 for ET, PT, and UT. MIL-HDBK-1823 may be used for other NDT procedures, such as RT or Profilometry, provided they provide either a quantitative signal, â, or a binary response, hit/miss. Because the purpose is to relate POD with target size (or any other meaningful feature like chemical composition), “size” (or feature characteristic) should be explicitly defined and be unambiguously measurable, that is, other targets having similar sizes will produce similar output from the NDT equipment. This is especially important for amorphous targets like corrosion damage or buried inclusions with a significant chemical reaction zone. Other literature on NDT data reliability is given elsewhere (2-7).

Note 14: AE as generally practiced does not yield the size of a flaw in a metallic liner of a COPV; however, can be used for accept-reject of COPVs (see Section 7 in both this guide and Guide E2981).

4.14 Further Guidance—Additional guidance for fracture control is provided in other governmental documents (NASA-STD-5003, SSP 30558, SSP 52005, NSTS 1700.7B), and non-government documents (NTIAC-DB-97-02, NTIAC-TA-00-01).

1.1 This guide discusses current and potential nondestructive testing (NDT) procedures for finding indications of discontinuities in thin-walled metallic liners in filament-wound pressure vessels, also known as composite overwrapped pressure vessels (COPVs). In general, these vessels have metallic liner thicknesses less than 2.3 mm (0.090 in.), and fiber loadings in the composite overwrap greater than 60 percent by weight. In COPVs, the composite overwrap thickness will be of the order of 2.0 mm (0.080 in.) for smaller vessels, and up to 20 mm (0.80 in.) for larger ones.

1.2 This guide focuses on COPVs with nonload sharing metallic liners used at ambient temperature, which most closely represents a Compressed Gas Association (CGA) Type III metal-lined COPV. However, it also has relevance to (1) monolithic metallic pressure vessels (PVs) (CGA Type I), and (2) metal-lined hoop-wrapped COPVs (CGA Type II).

1.3 The vessels covered by this guide are used in aerospace applications; therefore, examination requirements for discontinuities and inspection points will in general be different and more stringent than for vessels used in non-aerospace applications.

1.4 This guide applies to (1) low pressure COPVs and PVs used for storing aerospace media at maximum allowable working pressures (MAWPs) up to 3.5 MPa (500 psia) and volumes up to 2000 L (70 ft³), and (2) high pressure COPVs used for storing compressed gases at MAWPs up to 70 MPa (10  000 psia) and volumes down to 8 L (500 in.³). Internal vacuum storage or exposure is not considered appropriate for any vessel size.

Note 1: Some vessels are evacuated during filling operations, requiring the tank to withstand external (atmospheric) pressure.

1.5 The metallic liners under consideration include, but are not limited to, ones made from aluminum alloys, titanium alloys, nickel-based alloys, and stainless steels. In the case of COPVs, the composites through which the NDT interrogation should be made after overwrapping include, but are not limited to, various polymer matrix resins (for example, epoxies, cyanate esters, polyurethanes, phenolic resins, polyimides (including bismaleimides), polyamides) with continuous fiber reinforcement (for example, carbon, aramid, glass, or poly-(phenylenebenzobisoxazole) (PBO)).

1.6 This guide describes the application of established NDT procedures; namely, Acoustic Emission (AE, Section 7), Eddy Current Testing (ET, Section 8), Laser Profilometry (LP, Section 9), Leak Testing (LT, Section 10), Penetrant Testing (PT, Section 11), and Radiographic Testing (RT, Section 12). These procedures can be used by cognizant engineering organizations for detecting and evaluating flaws, defects, and accumulated damage in metallic PVs, the bare metallic liner of COPVs before overwrapping, and the metallic liner of new and in-service COPVs.

1.7 All methods discussed in this guide (AE, ET, LP, LT, PT, and RT) are performed on the metallic liner of COPVs before or after overwrapping and structural cure. The same methods may also be performed on metal PVs. For NDT procedures for detecting discontinuities in the composite overwrap in filament wound pressure vessels; namely, AE, ET, Shearography Testing (ST), RT, Ultrasonic Testing (UT) and Visual Testing (VT); consult Guide E2981.

1.8 Due to difficulties associated with inspecting thin-walled metallic COPV liners through composite overwraps, and the availability of the NDE methods listed in 1.6 to inspect COPV liners before overwrapping and metal PVs, ultrasonic testing (UT) is not addressed in this standard. UT may still be performed as agreed upon between the supplier and customer. Ultrasonic requirements may utilize Practice E2375 as applicable based upon the specific liner application and metal thickness. Alternate ultrasonic inspection methods such as Lamb wave, surface wave, shear wave, reflector plate, etc. may be established and documented per agreed upon contractual requirements. The test requirements should be developed in conjunction with the specific criteria defined by engineering analysis.

1.9 In general, AE and PT are performed on the PV or the bare metallic liner of a COPV before overwrapping (in the case of COPVs, AE is done before overwrapping to minimize interference from the composite overwrap). ET, LT, and RT are performed on the PV, bare metallic liner of a COPV before overwrapping, or on the as-manufactured COPV. LP is performed on the inner and outer surfaces of the PV, or on the inner surface of the COPV liner both before and after overwrapping. Furthermore, AE and RT are well suited for evaluating the weld integrity of welded PVs and COPV liners.

1.10 Wherever possible, the NDT procedures described should be sensitive enough to detect critical flaw sizes of the order of 1.3 mm (0.050 in.) length with a 2:1 aspect ratio.

Note 2: Liners often fail due to improper welding resulting in initiation and growth of multiple small discontinuities of the order of 0.050 mm (0.002 in.) length. These will form a macro-flaw of 1-mm (0.040-in.) length only at higher stress levels.

1.11 For NDT procedures that detect discontinuities in the composite overwrap of filament-wound pressure vessels (namely, AE, ET, shearography, thermography, UT and visual examination), consult Guide E2981.

1.12 In the case of COPVs which are impact damage sensitive and require implementation of a damage control plan, emphasis is placed on NDT procedures that are sensitive to detecting damage in the metallic liner caused by impacts at energy levels which may or may not leave any visible indication on the COPV composite surface.

1.13 This guide does not specify accept/reject criteria (4.10) used in procurement or used as a means for approving PVs or COPVs for service. Any acceptance criteria provided herein are given mainly for purposes of refinement and further elaboration of the procedures described in the guide. Project or original equipment manufacturer (OEM) specific accept/reject criteria should be used when available and take precedence over any acceptance criteria contained in this document.

1.14 This guide references established ASTM test methods that have a foundation of experience and that yield a numerical result, and newer procedures that have yet to be validated which are better categorized as qualitative guidelines and practices. The latter are included to promote research and later elaboration in this guide as methods of the former type.

1.15 To ensure proper use of the referenced standard documents, there are recognized NDT specialists that are certified according to industry and company NDT specifications. It is recommended that an NDT specialist be a part of any thin-walled metallic component design, quality assurance, in-service maintenance, or damage examination.

1.16 Units—The values stated in metric units are to be regarded as the standard. The English units given in parentheses are provided for information only.

1.17 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility of the user of this standard to establish appropriate safety, health, and environmental practices and determine the applicability of regulatory limitations prior to use.

1.18 This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.