Table of content
- Why method selection matters for composite aerostructures
- Ultrasonic testing: pulse-echo, through-transmission, and bond testers
- Active thermography
- Shearography
- Tap testing (coin tapping)
- Comparison by defect type
- Common mistakes
- Key takeaways
- FAQs
- References
- Further reading
Why method selection matters for composite aerostructures
Composite aircraft structures fail in ways metals don’t, and no single NDE method detects every one of those failure modes well. NASA’s own survey work for its Advanced Composites Project found that traditional metal-inspection NDE methods are “sometimes not adequate, or require substantial modification,” for composites, and that the defect types of concern — delamination, disbond, porosity, impact damage, moisture ingress — are “significantly different from metals.” Picking the wrong method, or relying on one method alone, can mean a real defect goes undetected.
This guide compares four methods aerospace NDT engineers actually use on composite aerostructures — ultrasonic testing (UT), active thermography, shearography, and tap testing — against the defect types that matter most: delamination, disbond (especially in honeycomb/sandwich structures), porosity/voids, impact damage, and moisture ingress. Source: NASA, “NDE Methods and Capabilities Handbook” (NASA/TM-2020-220568).
Ultrasonic testing: pulse-echo, through-transmission, and bond testers
UT is not one technique but a family, and which variant you use matters as much as choosing UT at all.
Pulse-echo (single-sided). Effective for delamination in laminates when the transducer/crystal size is matched to the expected defect size — a common rule of thumb is crystal diameter approximately equal to defect diameter (e.g., a 0.25-inch crystal for a 0.25-inch delamination). Can flag gross surface porosity through attenuation against a specification limit. Key limitation: the single-sided sound path means a near-surface defect can block or attenuate sound to material below it, so pulse-echo can miss stacked or deeper defects.
Through-transmission UT. Effective for disbond, unbond, foreign material, and core crush in bonded and sandwich structures, inspecting both the laminate and the bondline simultaneously. Typical operating frequency is 1–5 MHz, because composites attenuate sound strongly. Requires access to both sides of the part, which isn’t always practical on an assembled structure.
Bond testers (low-frequency UT: resonance, mechanical impedance analysis, pitch-catch). Resonance testing detects delamination and foreign material in laminates but needs a couplant and struggles with porosity. Mechanical impedance analysis (MIA) and pitch-catch methods work well for unbonds and core crush without a couplant, from a single side, down to roughly 0.5-inch defects.
Moisture ingress — a documented difficulty. In-service F/A-18 carbon/epoxy-skin, aluminum-honeycomb rudders failed from moisture-degraded skin-core bonds. Standard manual through-transmission UT flags disbond as an amplitude drop, but published research on this case found that water within the honeycomb cells provides significant sound transmission on its own, which can mask a disbond signal. Researchers combined thermography and neutron radiography to locate the water, then applied time-base analysis to through-transmission UT A-scans to try to separate the honeycomb-wall signal from the water-transmitted signal — described in that research only as showing potential, not as a validated, field-proven inspection method.
Phased-array UT (PAUT). A vendor-sourced blog citing FAA-funded testing reports probability-of-detection figures (POD[90/95] ≈ 0.91 inch for delamination in 12–32-ply CFRP with inspection aids, degrading to ≈ 1.75 inch for freehand scanning without coverage aids; ≈ 0.78 inch for disbond under focused go/no-go inspection). These figures are vendor-cited and have not been independently verified against primary FAA data, so they should be treated as illustrative rather than confirmed. A separate, more traceable program does exist — the FAA’s Airworthiness Assurance NDI Validation Center (AANC), operated by Sandia National Laboratories — cited in CompositesWorld’s “Assessing conventional and advanced NDI for composite aircraft” with different figures: a 90% detection chance for flaws 1.7–3 inches in diameter on 9-ply carbon/epoxy skin over Nomex honeycomb, and advanced methods (including phased array) reducing POD figures by 66–72% relative to conventional methods. These AANC-traceable figures don’t match the vendor-blog figures above and are reported here as a separate, better-sourced data point rather than a confirmation of them.
Governing standards. ASTM E2533 (standard guide for nondestructive examination of polymer matrix composites) and ASTM E3370 (covering a 0.5–20 MHz UT frequency range for aerospace composites) govern UT of composites generally; personnel qualification follows NAS 410 / EN 4179. Airframe-specific procedures (e.g., Boeing DPS documents, Airbus AITM documents) layer additional requirements on top of these.
Active thermography
Active thermography detects defects by watching how an applied heat stimulus flows through the material — a defect with different thermal diffusivity than sound material shows up as a contrast anomaly. For aerospace composites, flash (pulsed) thermography is governed by ASTM E2582, “Standard Practice for Infrared Flash Thermography of Composite Panels and Repair Patches Used in Aerospace Applications.”
Penetration is the key limiting factor. CFRP thermal penetration for flash thermography is typically limited to roughly 1.5–2.0 mm, which makes the technique best suited to thin-to-medium laminates; subsurface damage deeper than that limit can be missed entirely, regardless of how good the camera is.
Disbond detection. A disbond or void typically shows up as a high-contrast thermal anomaly with a diffusion pattern consistent with a gap in material continuity — effective for near-surface disbonds in sandwich/bonded structures, subject to the same depth ceiling.
Moisture ingress. This is a separately studied application: lock-in (modulated) thermography has been evaluated specifically for water-ingress detection in honeycomb sandwich structures, using binarization processing to distinguish water-filled cells from dry ones. A related study on flax/basalt hybrid composites (a different, thermoplastic natural-fiber material system, not an aerospace thermoset) found pulsed thermography effective to roughly 4 mm depth with some sensitivity to weave-pattern noise — useful supporting evidence for thermography’s general depth-limitation pattern, but not a direct aerospace-CFRP performance claim.
Cycle-time claims. Some vendor-linked sources report active thermography inspecting 8–10x faster than UT for equivalent coverage area; this figure comes from a vendor-affiliated user report, not an independent standard or study, and should be treated as a vendor claim rather than a verified industry figure.
Shearography
Mechanism. Shearography is a laser-interferometric, full-field, non-contact method. A self-referenced shearing device (commonly a modified Michelson shearing arrangement) lets a digital camera record speckle patterns before and after the part is loaded (making active shearography a stimulated method, not a passive one); subtracting the two images yields a “shearogram” whose fringe pattern marks a phase change corresponding to surface-strain gradients — shearography measures the first-order derivative of surface displacement, not displacement directly. Source: ASNT, “Overview of Digital Shearography for NDT.”
General capability claim. ASNT’s own overview states that internal discontinuities generate surface strain concentrations under loading that digital shearography can detect, and notes that conventional X-ray and UT “perform poorly on some composite materials” while shearography “has gained wide industrial acceptance for NDT of aircraft structures, particularly composite structures.” This is a general capability claim in ASNT’s own published material, not a quantified, defect-by-defect validation study.
Named vendor source. DEKRA (which has a stated cooperation with Dantec Dynamics, a shearography equipment manufacturer) lists disbonds, delaminations, cracked or crushed core, kissing bonds, wrinkling, fluid ingress, porosity, cracks, repair defects, and barely visible impact damage (BVID) among the defects shearography can detect in aerostructures under load. DEKRA does not publish sensitivity or depth thresholds for these claims. No independent, non-vendor quantitative validation of shearography’s detection limits on composite aerostructures has been published. Readers should treat the specific defect list as a vendor capability claim, not an independently confirmed performance specification.
Tap testing (coin tapping)
What it can do. Tap testing can detect near-surface disbonds and unbonds in bonded parts, to roughly four plies deep at most — a dull or dead tone where a defect is present, versus a sharp, consistent tone over sound material.
What it cannot do. Industry technical sources state plainly that tap testing should not be recommended for laminate delamination. It is not accurate for sizing a defect, since the tone transition is blurred rather than sharp, and it is not suitable for deep defects or complex multi-material assemblies. Any tap-test finding needs confirmation by UT or radiographic inspection before being acted on.
No governing standard exists. No FAA advisory circular, SAE ARP, or ASTM standard specifically governs tap-test procedure or limits on composite aerostructures. FAA AC 20-107B, the core composite-aircraft-structure advisory circular, does not mention tap testing, coin tapping, UT, thermography, or shearography by name anywhere in its text. It instead requires only that “repeatable and reliable non-destructive inspection techniques must be established” for bonded joints, generically, and separately states that bond-inspection NDI technology “has not been reliably demonstrated at a production scale to date.” AC 20-107B also notes that visual inspection remains “the predominant damage detection method used in the field,” and that “certain processing defects cannot be reliably detected at completion of the repair (e.g., weak bonds).” Tap testing’s reputation in industry literature as subjective and operator-dependent is not contradicted by a governing standard — there simply isn’t one naming it specifically.
POD figures. A vendor blog citing “FAA-funded testing” reports a bond/disbond POD[90/95] figure of approximately 0.78 inch for focused tap testing — the same figure attributed elsewhere in that source to focused UT inspection. This has not been independently verified against primary FAA data and should be treated as unconfirmed. A separately traceable figure from the FAA’s AANC program (operated by Sandia National Laboratories, cited in CompositesWorld) puts manual tap testing’s practical detection limit at roughly 8–10 plies — a different, better-sourced figure than the vendor-blog number above, and not a confirmation of it.
Comparison by defect type
NASA’s NDE Methods and Capabilities Handbook rates NDE techniques against flaw types using an Effective / Marginally Effective / Ineffective scale (its Table 3.0-1), built from round-robin testing across 88 fabricated composite standards. The handbook covers UT, laser/guided-wave UT, thermography, radiography, computed tomography, shearography, visual inspection, and tap testing — all four methods this article compares. The table below uses the handbook’s own ratings directly for UT, thermography, shearography, and tap testing, against its own defect-type columns (Porosity, Delaminations, FOD/Inclusions, Fiber Defects, Impact Damage), mapped to this article’s defect categories; disbond and moisture ingress aren’t separate columns in the handbook’s matrix, so those two rows draw on the other sources cited in each method’s section above and are less rigorously sourced than the top five rows.
Defect type | UT (pulse-echo) | Active thermography (single-sided flash) | Shearography | Tap testing |
|---|---|---|---|---|
Porosity | Effective | Marginally effective | Ineffective | Ineffective |
Delaminations | Effective | Effective | Effective | Marginally effective |
FOD/Inclusions | Effective | Effective | Effective | Ineffective |
Fiber defects | Effective | Effective | Effective | Ineffective |
Impact damage | Effective | Effective | Effective | Ineffective |
Disbond (sandwich/honeycomb)* | Effective (through-transmission UT) | Effective near-surface, same depth ceiling | Effective per vendor-sourced claims (DEKRA); no independent validation found | Effective to ~4 plies deep; not accurate for sizing |
Moisture ingress* | Documented difficulty: water transmission can mask disbond signal in honeycomb | Studied specifically via lock-in thermography for honeycomb water detection | No sourced capability data found | Not suited |
*Disbond and moisture ingress are not separate columns in NASA’s matrix; these two rows draw on the vendor and literature sources discussed in each method’s section above.
This table is a starting point for narrowing candidate methods, not a substitute for consulting the primary standards and the NASA handbook’s actual rating matrix directly for a specific inspection program.
Common mistakes
- Using tap testing on laminate delamination. Industry sources state this explicitly should not be done; tap testing is suited to near-surface disbond detection in bonded structures, not laminate delamination.
- Treating a vendor-cited POD figure as an independently confirmed FAA result. Several PAUT and tap-test POD figures circulating in secondary literature cite “FAA-funded testing” but have not been independently traced to primary FAA data; verify before relying on them for a qualification decision.
- Assuming a single UT variant covers all defect types. Pulse-echo, through-transmission, and bond testers (resonance, MIA, pitch-catch) have genuinely different strengths and access requirements — choosing “UT” without specifying the variant is choosing an incomplete answer.
- Citing a standard for tap testing that doesn’t exist. No FAA AC, SAE ARP, or ASTM standard specifically governs tap-test procedure on composites; state this gap honestly rather than implying regulatory coverage that isn’t there.
- Treating shearography’s vendor-published defect list as independently validated. The most detailed aerostructure defect-detection claims for shearography come from a single named vendor (DEKRA/Dantec); no independent quantitative validation was found.
Key takeaways
- No single method covers every composite aerostructure defect type well — UT, active thermography, shearography, and tap testing each have genuinely different strengths and blind spots.
- UT splits into pulse-echo, through-transmission, and bond-tester variants with meaningfully different capabilities; specify the variant, not just “UT.”
- Active thermography’s main constraint is penetration depth (roughly 1.5–2.0mm in CFRP), which limits it to thin-to-medium laminates and near-surface defects.
- Shearography’s aerostructure-specific defect-detection claims come primarily from named-vendor sources (DEKRA/Dantec); no independent quantitative validation was found.
- Tap testing has no governing FAA, SAE, or ASTM standard, and industry sources explicitly advise against using it for laminate delamination.
- Moisture ingress in honeycomb sandwich structures is a documented weak point for conventional through-transmission UT, since water transmits sound in a way that can mask disbond signals.
- NASA’s NDE Methods and Capabilities Handbook (NASA/TM-2020-220568) is a real, publicly available comparison framework covering UT, thermography, shearography and tap testing among other methods.
FAQs
Which method is best for composite delamination? Pulse-echo UT (with transducer size matched to the expected defect size) is well-established for laminate delamination; active thermography is effective for near-surface delamination within its ~1.5–2.0mm CFRP penetration limit. Tap testing should not be used for laminate delamination, per industry sources.
Can tap testing be relied on for honeycomb disbond? It can detect near-surface disbonds to roughly four plies deep, but it is not accurate for sizing and should be followed by UT or radiographic confirmation — and no governing standard specifically covers its use on composites.
Is shearography validated for aerospace use the way UT and thermography are? ASNT’s own technical overview describes shearography’s mechanism and notes wide industrial acceptance for composite aircraft structures, but the specific defect-detection claims available come from a single named vendor, not independent validation studies — treat vendor specifications accordingly.
Why does moisture ingress challenge conventional UT in honeycomb structures? Water inside honeycomb cells transmits sound on its own, which can mask the amplitude drop that through-transmission UT normally uses to flag a disbond — a documented difficulty in published research on in-service composite rudders.
Is there an official standard comparing these four methods by defect type? Yes — NASA’s NDE Methods and Capabilities Handbook (NASA/TM-2020-220568) provides a real, publicly available comparison framework (Table 3.0-1) that rates UT, laser/guided-wave UT, thermography, radiography, computed tomography, shearography, visual inspection, and tap testing against porosity, delaminations, FOD/inclusions, fiber defects, and impact damage.
References
- Howell, Patricia A. (NASA Langley Research Center), “NDE Methods and Capabilities Handbook — Resources for the Composites NDE Community,” NASA/TM-2020-220568, presented at NDT in Aerospace 2020 — the comparison-framework backbone for this article’s comparison table, covering UT, laser/guided-wave UT, thermography, radiography, computed tomography, shearography, visual inspection, and tap testing against porosity, delaminations, FOD/inclusions, fiber defects, and impact damage, based on 88 round-robin-tested standards.
- Aviation Pros, “Nondestructive Testing of Aircraft Composites” — UT variant capabilities (pulse-echo, through-transmission, bond testers) and tap-testing limitations.
- Ultragarsas/Ultrasound journal, “Examination of F/A-18 honeycomb composite rudders for disbond due to water using through-transmission ultrasonics” — the moisture-ingress/UT masking case study.
- ASTM International, E2582, “Standard Practice for Infrared Flash Thermography of Composite Panels and Repair Patches Used in Aerospace Applications” — governing standard for active (flash) thermography of composites.
- ASTM International, E2533 (guide for NDE of polymer matrix composites) and E3370 (UT frequency range for aerospace composites) — governing UT standards.
- ASNT, “Overview of Digital Shearography for NDT” — shearography mechanism and general capability description.
- DEKRA (using Dantec Dynamics equipment), shearographic testing page — named-vendor defect-detection claims for aerostructures, cited explicitly as vendor-sourced.
- FAA, Advisory Circular 20-107B (Composite Aircraft Structure) — confirms the absence of method-specific NDI guidance, including for tap testing.
- PMC8952369, “Binarization Mechanism Evaluation for Water Ingress Detectability in Honeycomb Sandwich Structure Using Lock-In Thermography” — moisture-ingress-specific thermography study.
Further reading
- Telops Radia: Quantitative Infrared Thermography for NDT — camera hardware detail for the active-thermography method covered above.
- LVI Behavior in Flax/Basalt Hybrid Composites: IRT and Terahertz Spectroscopy — supporting case study for thermography’s depth-limitation pattern (different material system, not aerospace CFRP).
- Aerospace NDT: Methods, Standards and Certification — the hub this method-selection guide supports.
- Active Thermography for Composites: Pulsed and Lock-in — full depth on the pulsed/lock-in thermography method referenced above.
- Industrial CT for Additive Manufacturing Parts — a volumetric method for AM composite/metal aerostructure parts not covered in this comparison.




There are no comments yet!