Method improves ultrasonic inspection of complex parts

Ultrasonic waves can find faults in adhesive bonds holding composite parts together and the research from Penn will let engineers select the best frequencies to detect adhesive failures in hard-to-reach places.

Different ultrasonic modes work best for different materials and configurations using the right one will locate more flaws with higher precision, according to the researchers. The selection process could save time and effort for engineers who perform maintenance on complex structures made from composite materials.

Adhesive bonds are better for attaching composite parts than nuts and bolts, which pierce and weaken structural integrity, but heavy operation can crack the glue, damaging the bond’s effectiveness. Ultrasonic waves let engineers examine bonded regions non-destructively.

‘This technique is very widely used in aerospace engineering because those structures require a very high reliability,’ said Baiyang Ren, postgraduate in engineering science and mechanics.

For bonded regions located between easily accessible, wide surfaces, obtaining a clean readout is straightforward as ultrasonic waves pass through the bonded intersection to a receiver on the other side without interference. However, for bonds between inaccessible, irregularly angled surfaces, the ultrasonic wave will convert into a different mode by the time it enters and travels through the bonded region, and this new wave mode may not be as sensitive to adhesive flaws.

‘Usually…the mode conversion at this region is a blind thing for people,’ Ren said in a statement. ’We want to know [the mode conversion] so that we know how to inspect that bond.’

There are thousands of ultrasonic modes and for the best results engineers have to choose one that will convert into the actual mode they want, but identifying the right one for a specific project through trial and error can be time-consuming.

Researchers devised a selection process to determine the best combination of ultrasonic modes for a given material by identifying the criteria for optimal frequencies and eliminating possibilities through a series of models and calculations.

‘We want the mode to travel fast, travel longer, and be sensitive to this bond and still be received by some particular receiver,’ said Ren. ‘Each criteria will filter out some of the modes. After several steps, what things you have left are the modes that you want to try.’

To evaluate the efficacy of the selection process, researchers tested two optimal modes on pieces of carbon-fibre-reinforced polymer, glued together with planned defects created by inserting Teflon into the bonded region.

They then compared the results with a readout produced by one of the discarded frequencies. The two modes predicted to be successful detected flaws at a much finer resolution than the discarded one.

The results of the US government-funded project are available in the International Journal of Adhesion and Adhesives.