A kissing bond — a zero-volume interfacial disbond held in compressive contact so that it transmits load while appearing well bonded to conventional ultrasound — is among the hardest adhesive-joint defects to inspect. Yan, Drinkwater and Neild (2009) showed that contact-type kissing bonds, made by compressively loading a disbonded adhesive layer, exhibit a large, load-dependent ultrasonic nonlinearity measurable through second-harmonic through-transmission, but their study covered only this defect class at low loads, left the residual system nonlinearity undecomposed, did not address couplant variability or scanning implementation, and stopped short of any quantitative procedure that fuses linear and nonlinear ultrasonic indications into a bond-condition diagnosis. The open problem is to establish and validate a complete quantitative characterization procedure: one that discriminates kissing-bond classes (including contaminant-layer weak bonds that may add little nonlinearity), carries an explicit budget of the adhesive/instrument/coupling/transducer background nonlinearity, is implementable as a robust scanning measurement, and demonstrably relates measured linear and nonlinear ultrasonic observables to bond condition beyond the low-load regime. No published procedure is known that meets these jointly linked requirements, so the core problem remains open.

Background

Adhesive joints are structural connections formed by a polymer adhesive layer between adherends (e.g. aluminum or composite parts), widely used in aerospace and automotive structures. Their most evasive defect is the kissing bond: two adhered surfaces that are physically disbonded but pressed together so tightly that the interface transmits compressive load and reflects little ultrasound. Because ordinary pulse-echo testing keys on reflections from air gaps, a kissing bond is essentially invisible to conventional linear ultrasonic inspection, which is why it is a long-standing concern for safety-critical bonded structures.

The physical signature that makes kissing bonds detectable at all is contact acoustic nonlinearity: when an ultrasound wave crosses an interface whose stiffness varies with pressure — as for two rough or clapping (alternately opening and closing) surfaces in partial contact — the wave is distorted and the interface generates harmonics of the transmitted frequency. Quantitatively, the accumulated second-harmonic amplitude A2A_2 relative to the squared fundamental amplitude A12A_1^2 defines an apparent nonlinearity parameter, often written β′\beta', which grows with the strength of the interface nonlinearity. Linear ultrasonic indicators — reflection coefficient and interfacial stiffness inferred from the frequency-dependent transmission of the bond line — measure how much the interface is compliant but not whether it is genuinely adhered, which is precisely the ambiguity of a kissing bond.

In the study underlying this problem, Yan, Drinkwater and Neild (NDT&E International, 2009) transmitted tone bursts through aluminum lap joints containing a fully disbonded adhesive layer loaded in compression, so that the disbond behaved as a contact-type kissing bond of adjustable pressure. They extracted the apparent nonlinearity parameter across loads and frequencies, suppressed instrument-generated harmonics using the transmitting transducer's natural filtering, calibrated against reference specimens, and compared the load dependence of the measured nonlinearity with a pressure-dependent interfacial-stiffness model. Their findings: the kissing-bond nonlinearity is large at low contact pressure and falls steeply as compression closes the interface, the extracted nonlinearity parameter is approximately amplitude-independent, and known background sources (water coupling path, instrumentation, receiver transducer, aluminum adherends, adhesive material) were each bounded in isolation and found small relative to the low-load kissing-bond signal, but the residual was not attributed where the contrast is weakest.

The study was a laboratory demonstration on one defect class, not an inspection method. Beyond it, several issues remain unresolved on the path to making nonlinear ultrasonics a quantitative industrial bond-condition tool: (i) the specimens were contact-type kissing bonds made by compressing a disbond, whereas real weak bonds also arise from surface contaminants (oils, release agents) that may leave a stiff interface adding little or no nonlinearity — the published definition of a kissing bond explicitly includes coupling through a very thin contaminant layer, but this class was not tested — so whether the same measurement strategy can discriminate such classes is unknown; (ii) the demonstration did not decompose the residual system nonlinearity among adhesive material, instrumentation, coupling, and transducers, and such a decomposition (or a bounded accounting of it) is needed to state the achievable detection threshold; (iii) scanning implementation requires controlling couplant repeatability and the path-length-dependent nonlinearity added by water coupling, which the fixed laboratory arrangement did not address; and (iv) no quantitative procedure existed to combine the nonlinear measurement with linear ultrasonic information (reflection, interface stiffness, attenuation) into a joint interpretation of bond condition, outside the low-load regime where the nonlinear contrast happens to be strongest.

Problem Statement

Establish, and validate on realistic adhesive joints, a quantitative ultrasonic procedure for characterizing kissing-bond condition that goes beyond the single-class, low-load laboratory demonstration of Yan, Drinkwater and Neild (2009) and jointly meets the following requirements:

  1. Class coverage and discrimination: the procedure must give a measurable indication of bond condition not only for contact-type kissing bonds (disbonds loaded in compression, where nonlinearity is strong at low load) but also for other kissing-bond classes, in particular contaminant-layer weak bonds whose interfaces may be stiff and add little or no nonlinearity — i.e. it must distinguish, quantitatively or by a defensible decision rule, these classes from each other and from well-bonded material.
  2. Nonlinearity budget: the measured nonlinearity must be accompanied by a decomposition or bounded accounting of the residual system nonlinearity — contributions of adhesive material, instrumentation, coupling, and transducers — sufficient to state a detection/quantification threshold valid at the loads where the kissing-bond contrast is weakest.
  3. Robust implementation: the measurement must be shown to work as a repeatable scanning configuration, with coupling variability (liquid couplant variability, path-length-dependent water nonlinearity) controlled or designed out.
  4. Joint interpretation: linear ultrasonic observables (reflection coefficient, interface stiffness, attenuation) and nonlinear observables (harmonic generation, apparent nonlinearity parameter) must be combined into a single quantitative interpretation of bond condition, validated against independent evidence of bond quality (e.g. destructive strength tests) beyond the low-load regime where the nonlinear contrast is strongest.

References

  1. Dawei Yan, Bruce W. Drinkwater, Simon A. Neild, "Measurement of the ultrasonic nonlinearity of kissing bonds in adhesive joints", NDT&E International 42(5):459–466 (2009). DOI: 10.1016/j.ndteint.2009.02.002, https://doi.org/10.1016/j.ndteint.2009.02.002
  2. Igor Solodov, Mike Kornely, Jens Philipp, Elisabeth Stammen, Klaus Dilger, Marc Kreutzbruck, "Linear vs nonlinear ultrasonic testing of kissing bonds in adhesive joints", Ultrasonics 132:106967 (2023). DOI: 10.1016/j.ultras.2023.106967, https://doi.org/10.1016/j.ultras.2023.106967
  3. Latifa Attar, Mounsif Ech Cherif El Kettani, Damien Leduc, Mihai Valentin Predoi, Jocelyne Galy, "Detection of kissing bond type defects and evaluation of the bonding quality in metal/adhesive/composite structures by a wavenumber-frequency insensitive SH mode", NDT&E International 137:102841 (2023). DOI: 10.1016/j.ndteint.2023.102841, https://doi.org/10.1016/j.ndteint.2023.102841
  4. C. Jeenjitkaew, F. J. Guild, "The analysis of kissing bonds in adhesive joints", International Journal of Adhesion and Adhesives 75:101–107 (2017). DOI: 10.1016/j.ijadhadh.2017.02.019, https://doi.org/10.1016/j.ijadhadh.2017.02.019
  5. Bengisu Yılmaz, Elena Jasiūnienė, "Advanced ultrasonic NDT for weak bond detection in composite-adhesive bonded structures", International Journal of Adhesion and Adhesives 102:102675 (2020). DOI: 10.1016/j.ijadhadh.2020.102675, https://doi.org/10.1016/j.ijadhadh.2020.102675
  6. Do-Kyung Pyun, Lucas W. Koester, Daniel J. Barnard, Leonard J. Bond, "Nonlinear Ultrasonic Inspection of the Effect of Contaminants on Material Properties of Epoxy-Adhesive", Journal of Nondestructive Evaluation 41:75 (2022). DOI: 10.1007/s10921-022-00904-y, https://doi.org/10.1007/s10921-022-00904-y
  7. Do-Kyung Pyun, Daniel J. Barnard, Leonard J. Bond, "Nondestructive characterization of contaminant-induced material softening in epoxy polymers using nonlinear ultrasonic measurements", NDT&E International 153:103364 (2025). DOI: 10.1016/j.ndteint.2025.103364, https://doi.org/10.1016/j.ndteint.2025.103364
  8. Wujun Zhu, Yanxun Xiang, Menghui Jiang, Taotao Ding, Zhiqin Qian, Fu-Zhen Xuan, "Baseline-free localisation of kissing bonds in bonded structures using nonlinear Lamb wave responses under dual-amplitude excitations", Nondestructive Testing and Evaluation (2026). DOI: 10.1080/10589759.2026.2693717, https://doi.org/10.1080/10589759.2026.2693717
  9. Shengbo Shan, Zheng Li, Qingzhao Kong, Li Cheng, "Macro and micro delamination detection in adhesive structures: A synthesized linear and nonlinear air-coupled ultrasonic technique", Ultrasonics 154:107675 (2025). DOI: 10.1016/j.ultras.2025.107675, https://doi.org/10.1016/j.ultras.2025.107675
  10. Jingpin Jiao, Zhiqiang Li, Li Li, Guanghai Li, Xinyuan Lu, "Nondestructive Evaluation of Adhesive Joints Using Nonlinear Non-collinear Wave Mixing Technique", Journal of Nondestructive Evaluation 44:4 (2024, online first; print issue 2025). DOI: 10.1007/s10921-024-01142-0, https://doi.org/10.1007/s10921-024-01142-0
  11. E. Davoodi, M. S. Safizadeh, "Detection of kissing bond defect in composite patches bonded to an aluminium plate using the vibrothermography technique", Insight – Non-Destructive Testing and Condition Monitoring 67(12):741–750 (2025). DOI: 10.1784/insi.2025.67.12.741, https://doi.org/10.1784/insi.2025.67.12.741
  12. Colin L. Williams, Matthew H. Lear, Parisa Shokouhi, "A review of the microstructural contributions to the acoustic nonlinearity parameter measured with longitudinal and Rayleigh wave second harmonic generation in metals", NDT&E International 142:103027 (2024). DOI: 10.1016/j.ndteint.2023.103027, https://doi.org/10.1016/j.ndteint.2023.103027
  13. Robert L. Crane, John Hart-Smith, John Newman, "Nondestructive inspection of adhesive bonded joints", chapter in Adhesive Bonding (Woodhead Publishing / Elsevier, 2021), pp. 215–256. DOI: 10.1016/B978-0-12-819954-1.00008-3, https://doi.org/10.1016/B978-0-12-819954-1.00008-3