Influence of dynamic load and temperature on guided wave ultrasonic damage detection in thin plates

Long-thin metallic materials are essentially used in constructing structures of high economic importance, but their service life is shortened by damage such as cracks, corrosion, cavities, notches, and dents. Damage is an inevitable condition of metallic structures over time and, when not detected, could result in a catastrophic breakdown. In the past decades, high interest has been developed in using the guided wave ultrasonic technique (GWUT) to monitor the health of structures and detect damage due to its long-distance coverage potential with little attenuation and cost-effectiveness. Most guided wave ultrasonic studies have focused on detecting and characterising empty cracks or notches. Limited literature is available to explain the behaviour of guided waves while travelling in thin plates exposed to damage filled with debris, which is more likely possible in long-thin structures such as pipelines for oil, water or gas transportation. Debris- filled damage leads to corrosion processes, particularly inducing pitting corrosion. This form of corrosion is localised and difficult to detect. It has contributed to many structural failures, particularly in oil and gas pipelines. Hence, early detection and characterisation of this form of damage is vital to avert catastrophic failure. This study explored the detection of damage filled with different proportions of debris in thin plates using guided wave ultrasonic techniques. The captured response signals underwent analysis through various signal-processing methods in MATLAB. Additionally, the research examined how temperature variations and low-frequency vibrations impact the guided wave responses, aiming to simulate the effect of environmental operation conditions. Through the analysis, an empirical model was developed to predict debris-filled damage and differentiate it from empty damage and the health state of the structure. The predictive model has an average error of about 1.34. Also, the analysis revealed that cross- correlation of the detrended response and reference signals could demonstrate a quick way to visualise and spot debris-filled damage in the structure. Additionally, a model called Olisa-Khan low-vibration mitigation architecture (Olisa-Khan LMA) was created to counteract the severe effects of varying low- frequency vibrations and improve the performance of the damage detection technique. The average percentage deviation of the model response signal and static response signal was about 1.64 %, suggesting the two signals are very close. The slight deviation could be attributed to the signal loss due to clipping and imperfection in the system. In characterising debris that filled the damage, an excitation signal with a central frequency of 80KHz was found optimal because the deviation of each state of damage differs from the other and decreases from an empty case to a debris-filled case and continues as fluid-filled viscosity increases. The study's merit cannot be overemphasised as it establishes models, especially for predicting novel damage of debris-filled and characterising different debris that filled the damage even in severe environmental operation conditions. Hence, the study would be useful for continuously monitoring long-thin structures of high economic values for possible damage detection and characterisation.