Modern structures on aircraft make increasing use of large-scale composite structures. Quantitative damage monitoring for composites, including damage occurrence, number, localization, and size estimation, will help reduce maintenance costs, improve fleet management efficiency through condition-based maintenance, and potentially more rapidly enable new material systems and structural concepts by integrating health monitoring into the design itself. With the advantage of easily interpretable, intuitive, and accurate imaging result, the delay-and-sum imaging algorithm is frequently researched and applied to damage monitoring. However, when it is applied to multidamage monitoring in large-scale composites, the consumed time and computation resource are too much for pixel value calculation. Besides, due to the material anisotropy and existence of bolt holes and stiffeners in the researched complex carbon fiber composite laminate, propagation mechanism of Lamb wave is quite complicated, which makes the conventional localization result not accurate by delay-and-sum algorithm according to the point with pixel peak value. What is more, the imaging quality is deteriorated with concurrence of multiple damages, and thus, quantitative damage information cannot be extracted. Hence, the damage index merging algorithm is introduced for quick damage identification and damage merging in every subarea divided by piezoelectric sensor array. The delay-and-sum imaging algorithm is performed afterward only in subareas identified with damages, which significantly improves the efficiency of damage imaging for large-scale composites. The nonlinear normalization of pixel values compensates the deterioration of multidamage imaging quality. Then, a weighted average algorithm is introduced for more accurate localization. A further damage size level estimation is realized with probabilities by extracting the image pixel peak value. Experiments with six damages simultaneously on the complex carbon fiber laminate verify the effectiveness of the proposed quantitative multidamage monitoring algorithm with localization error below 2 cm and correct damage number and size level estimation.

Bonded structures are frequently adopted in structural connections and are highly prone to degradation or decrease of interfacial strength due to adhesive aging, poor quality of surface preparation, as well as the exposure to harsh environment and external loading. This study addresses the establishment of a framework in which a modally selective ultrasonic guided wave is used for disbond identification and sizing. In this framework, the propagating and evanescent modes of ultrasonic guided waves are first obtained, followed by the excitability analysis for each ultrasonic guided wave propagating mode, providing a theoretical basis for effective wave excitation in the experiment. Then the interaction of ultrasonic guided wave with disbond is interrogated analytically using a method combining semi-analytical finite element and normal mode expansion, whereby wave transmission, wave reflection, and mode conversion can be calculated quantitatively. Taking all these aspects into account, mode 11 at around 3.85 MHz features a high propagation velocity, large mode excitability, and increasing amplitude drop with the enlargement of disbond size, and is thus selected for disbond detection. Both numerical and experimental validations are performed, in which disbonds of different lengths from 10 to 40 mm are examined, and the results well corroborate the effectiveness of the proposed framework for ultrasonic guided wave–based disbond detection.