Investigating the Ultrasonic Inspection defect detection capability in Glass Fiber Reinforced Plastic (GFRP)

Attenuation of ultrasonic waves is caused by absorption and scattering. It results in a loss of sound amplitude as the wave travels through a material and is one of the key limitations affecting the performance of an ultrasonic inspection. In this case study we intend to look at the effects of wave attenuation on the defect detection capability in GFRP at a fixed frequency.

GFRP is an ultrasonically attenuative material, however, owing to its light weight and good mechanical properties, it is widely used in aerospace, marine and automotive components, wind turbine blades and many other areas.

The purpose of this case study is to show how attenuation changes the ability to find flaws in GFRP. This will help you choose the right probe frequency when planning an inspection.

Simulation Set-Up

The transducer used is a 0.5 MHz single conventional transducer operating in pulse-echo mode. No wedge is used in this simulation and the probe is placed in direct contact with the GFRP specimen. The specimen has a longitudinal wave velocity of 4310 m/s, a shear wave velocity 2024 m/s and a density of 2100 kg/m3. The plate is 50 mm thick and 200 mm wide.

A 2-D simulation using Salvus was set up to run for 40 microseconds with and without an attenuation coefficient of 100 dB/m (@ 0.5 MHz) and the results were judged on the detectability of a 2 mm Flat Bottomed Hole (FBH), placed 5 mm from the backwall of the specimen.

The figure below shows the numerical mesh created for the simulation and the FBH was created by removing elements from it.

Results

This video shows the reflection of the 2mm reflector with no attenuation in the part:

We ran a total of 4 cases, with and without attenuation and with and without the FBH. The cases without the FBH were run to clearly see the indication from the FBH.

The compute time for the 4 cases was less than 25 seconds using a standard laptop (2024 MacBook Air), below are the A-Scan results.

The first A-scan shows the simulations run with and without the 2mm FBH, both with no attenuation.

The results are as expected. The signal from the FBH is clearly seen before the expected backwall signal and the backwall signal from the simulation with the FBH is reduced, due to the presence of the FBH.

The Second A-Scan shows the effect of attenuation with and without the FBH. Compared to the A-Scan in the first case, you can see that the peak signal amplitude, from the backwall reflection is a factor of 3 larger in the A-scan without attenuation.

The third A-Scan shows the received signal of the FBH with and without attenuation.

Finally, we adjusted the attenuation to a level where it becomes barely detectable to see the signal from the FBH. The graph indicates that with the attenuation coefficient increased to 200 dB/m (@ 0.5 MHz), the amplitude received by the reflector is reduced further by 40%, hardly indistinguishable from the noise.

Conclusion

Through Salvus simulations, ultrasonic inspection capability can be evaluated and optimized in a reduced time frame. This simulation allows for accurate allowances for material attenuation to determine inspection performance.

Future studies could include adding the noise from specific equipment. This could be performed in two ways: either artificially adding noise to the signal or by calibrating the simulation against real test data.