Search

A novel technique for processing thermal data is presented and applied to simulation as well as experimental data. Using a neural network of thermal data classification, good classification accuracies are obtained, and the resulting images exhibit very good contrast between bonded and disbonded locations. In order to minimize the preprocessing required before using the network of classification, the temperature values were directly employed to train a network using data from an on-site testing run of a commercial aircraft. Training was extremely fast, and the resulting classification also agreed reasonably well with an ultrasonic characterization of the panel. The results obtained using one sample show the partially disbonded vertical doubler. The vertical lines along the doubler correspond to the original extent of the doubler obtained using blueprints of the aircraft.

A study of nondestructive evaluation of aircraft bonded joints using the contact scanning technique is presented. Reflected signals acquired through the contacting transducer characterize the test region as being bonded or disbonded. Ultrasonic signals are attenuated more rapidly in the bonded regions when compared to disbonded regions. A peak amplitude based method and an artificial neural network are used to classify the signals. Results obtained using an artificial neural network exhibited significant insensitivity to signal variation when compared to the peak amplitude. Very good agreement is observed between results obtained using the present technique and those obtained using immersion scanning.

An account is given of the capabilities of the magnetooptic/eddy-current imager (MEI) apparatus in the case of aging aircraft structure-type flaws in 2024-T3 Al alloy plates. Attention is given to images of cyclically grown fatigue cracks from rivetted joints in fabricated lap-joint structures, electrical discharge machining notches, and corrosion spots. Although conventional eddy-current methods could have been used, the speed and ease of MEI's use in these tests is unmatched by such means. Results are displayed in real time as a test piece is scanned, furnishing easily interpreted flaw images.

The effect on the system acoustic response of variations in the adhesive thickness, coupling thickness, and paint thickness is considered. Both simulations and experimental measurements are used to characterize and classify A-scans from test regions, and to study the effects of various parameters such as paint thickness and epoxy thickness on the variations in the reflected signals. A 1D model of sound propagation in multilayered structures is used to verify the validity of the measured signals, and is also used to computationally generate signals for a class of test locations with gradually varying parameters. This approach exploits the ability of numerical simulations to provide a good understanding of the ultrasonic pulses reflected at disbonds.

A procedure for interpreting thermal images obtained during the nondestructive evaluation of aircraft bonded joints is presented. The procedure operates on time-derivative thermal images and resulted in a disbond image with disbonds highlighted. The size of the 'black clusters' in the output disbond image is a quantitative measure of disbond size. The procedure is illustrated using simulation data as well as data obtained through experimental testing of fabricated samples and aircraft panels. Good results are obtained, and, except in pathological cases, 'false calls' in the cases studied appeared only as noise in the output disbond image which was easily filtered out. The thermal detection technique coupled with an automated image interpretation capability will be a very fast and effective method for inspecting bonded joints in an aircraft structure.