Date of Award
Open Access Dissertation
The objective of this research is to devise new methods and tools to generate real time awareness of the material state of composite and metallic structures through ultrasonic nondestructive evaluation (NDE) and structural health monitoring (SHM) at its very early stage of failure. To device new methodology it is also important to verify the method through virtual experiments and hence computational NDE is getting popular in the recent years. In this thesis, while experimental methodology is developed to understand the material state at its early stage of failure, a new peridynamic based Peri-Elastodynamic (PED) computational method is also developed for virtual NDE and SHM experiments. In the experimental part, material state awareness through precursor damage quantification is proposed for composite materials and in the predictive part modelling of ultrasonic wave propagation in the engineered materials is developed. Symbiotic information fusion between the Guided Coda Wave Interferometry (CWI) and Quantitative Ultrasonic Image Correlation (QUIC) was devised for the awareness and the quantification of the precursor damage state in composites. The proposed research work is divided into two major parts a) Experimental and b) Computational.
a) Experimental: In composite materials, the precursor damages (for example matrix cracking, microcracks, voids, fiber micro-buckling, local fiber breakage, local debonding, etc.) are insensitive to the low-frequency ultrasonic NDE or Structural Health Monitoring (SHM) (~100–~500 kHz) methods. Overcoming this barrier, an online method using the
later part of the guided wave signal, which is often neglected is proposed for the precursor damage quantification. Although the first-arrival wave packets that contain the fundamental guided Lamb wave modes are unaltered, the following part of the wave packets however carry significant information about the precursor events with predictable phase shifts. The Cross-correlation and Taylor-series-based modified CWI technique is proposed to quantify the stretch parameter to compensate the phase shifts in the coda wave as a result of precursor damage in composites. The results are thoroughly validated with newly formulated high frequency (>~25MHz) QUIC method. The proposed process is validated and verified with American Society of Testing of Materials (ASTM) standards woven composite-fiber-reinforced-laminate specimens (CFRP). Both online CWI and offline QUIC was performed to prove the feasibility and reliability of the proposed precursor damage quantification process. Visual proof of the precursor events is provided from the digital micro optical microscopy and scanning electron microscopy. Additionally, acoustic-nonlinearity of analysis Lamb wave propagation was employed to investigate, stress-relaxation phenomena in composites. Fatigue loading on composite specimens followed by relaxation experiments were conducted to examine influence of damage and relaxation on acoustic-nonlinearity. It was observed that the stress-relaxation in composite is primarily coupled with the second-order nonlinearity parameters derived from the Lamb wave modes. Furthermore, these parameters were found inherently associated with the remaining strength of the composites. Results from the nonlinear analysis were found to be in good agreement with those obtained from CWI analysis.
In the near future, it is expected that the structure, structural component or individual material states could be digitally certified for their future missions by including a predictive tool in a “Digital Twin” software fusing the information from experimental finding. This thesis contributes to this concept and the information obtained from experimental NDE discussed above can be utilized by a predictive tool to predict accurate material behavior as well as NDE or SHM sensor signals off-line, simultaneously. Considering multiple advantages of peridynamic based approach in incorporating experimental data and damage modelling capability over tradition approaches, newly devised Peri-Elastodynamic (PED) is discussed in the following paragraph to simulate the three-dimensional (3D) Lamb wave modes in materials for the first time.
b) Computational: PED is a nonlocal meshless approach which is a scale-independent generalized technique to visualize the acoustic and ultrasonic waves in plate-like structures. Characteristics of the fundamental Lamb wave modes are simulated in a plate-like structure with a surface mounted piezoelectric (PZT) transducer which is actuated from the top surface. In addition, guided ultrasonic wave modes were also simulated in a damaged plate. the PED results were validated with the experimental results which shows that the newly developed method is more accurate and computationally cheaper than the FEM to be used for computational NDE and SHM. PED was also extended to investigate the wave-damage interaction with damage (e.g., a crack) in the plate. The accuracy of the proposed technique herein is confirmed by performing the error analysis on symmetric and anti-symmetric Lamb wave modes compared to the experimental results for both pristine and damaged plate
Patra, S.(2018). Ultrasonic Analysis and Tools for Quantitative Material State Awarness of Engineered Materials. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/5057