Fatigue crack detection using dual laser induced nonlinear ultrasonic modulation

We present a technique for microcrack modeling in the finite element framework, and numerically investigate the occurrence of nonlinear wave modulation. Typically, fatigue cracks are initiated and developed when structures are exposed to repeated loading; the crack widths of the fatigue cracks are extremely small in the early development stage. As the fatigue cracks grow by combining and coalescing, the overall size increases. Enlarged cracks undermine the safety of the structure. Therefore, fatigue crack detection is very important to ensure the integrity of structures. Although the nonlinear ultrasonic wave modulation technique has been widely used due to its high detecting sensitivity, the basic principle is not fully understood. To reveal the mechanism of nonlinear wave modulation, the movements of the crack surfaces are calculated through numerical simulation. The shape of the crack surface can determine the intensity of the wave modulation. In this study, we investigate the variation of the crack widths due to fatigue failure using microscopic imaging of real fatigue cracks, and use these images to create realistic models of the fatigue cracks.

Reference-free fatigue crack detection using nonlinear ultrasonic modulation under various temperature and loading conditions

For a metallic component, prediction of a remaining useful life (RUL) is of extreme importance to prevent a sudden failure of the structure. If a fatigue crack exists, the nonlinear interaction of a low frequency pumping wave (LF) and a high frequency probing wave (HF) generates additional responses at the frequencies equal to the sum and difference of the input frequencies. In this study, the nonlinear ultrasonic modulation is used for fatigue crack prognosis of the lifting-lug that connects a tension cable. Nonlinear ultrasonic modulation components are used to define a fatigue index. The functional relationship between fatigue index and the number of loading cycles is established based on Paris–Erdogan theories. Then, by curve fitting of the established function to the fatigue index values obtained at various loading cycles, the RUL corresponding to a fatigue index is estimated from the fitted function. The proposed RUL estimation method is associated with the following advantages: (1) the fatigue index is defined as a function of nonlinear ultrasonic modulation components, (2) the functional relationship between fatigue index and loading cycles is established for lifting-lug structure, and (3) online remaining useful life is predicted and updated using sequentially measured ultrasonic signals. To validate the proposed method, the lifting-lug specimen is used and subjected to cyclic loading. The experiment is conducted using surface-mounted piezoelectric sensors for ultrasonic signal excitation and measurement. The experimental results confirm that the proposed method can estimate the RUL successfully.

By generating ultrasonic waves at two different frequencies onto a cracked structure, modulations due to crack-induced nonlinearity can be observed in the corresponding ultrasonic response. This nonlinear wave modulation phenomenon has been widely studied and proven capable of detecting a fatigue crack at a very early stage. However, under field conditions, other exogenous vibrations exist and the modulation components can be buried under ambient noises, making it difficult to extract the modulation components simply by using a spectral density function. In this study, the nonlinear modulation components in the ultrasonic response were extracted using a spectral correlation function (the double Fourier transform) with respect to time and time lag of a signal’s autocorrelation. Using spectral correlation, noise or interference, which is spectrally overlapped with the nonlinear modulation components in the ultrasonic response, can be effectively removed or reduced. Only the nonlinear modulation components are accentuated at specific coordinates of the spectral correlation plot. A damage feature is defined by comparing the spectral correlation value between nonlinear modulation components with other spectral correlation values among randomly selected frequencies. Then, by analyzing the statistical characteristics of the multiple damage feature values obtained from different input frequency combinations, fatigue cracks can be detected without relying on baseline data obtained from the pristine condition of the target structure. In the end, an experimental test was conducted on aluminum plates with a real fatigue crack and the test signals were contaminated by simulated noises with varying signal-to-noise ratios. The results validated the proposed technique.

Reference-free fatigue crack detection using deep long short-term memory network (DLSTM) and nonlinear ultrasonic modulation

This paper presents a fatigue crack detection technique using changes of linear ultrasonic features caused by structural nonlinearity. We theoretically and experimentally demonstrate that a nonlinear mechanism produced by a fatigue crack not only generates nonlinear modulations or harmonics, but also reduces the amplitude of the linear responses because the energy in the linear responses is partially transmitted to the harmonic or modulation components. The experimental results show that the amplitude reduction of the linear responses is more sensitive to a fatigue crack than the amplitude increase of the harmonics and modulation components. In addition, a baseline-free fatigue crack detection technique is proposed considering the additional amplitude reduction of the linear components due to nonlinear ultrasonic modulation.

This paper theoretically and experimentally demonstrates that (1) a localized nonlinear mechanism such as a fatigue crack not only generates nonlinear modulations or harmonics, but also reduces the amplitudes of the linear responses at the excitation frequencies, (2) the energy in the linear responses is partially transmitted to the nonlinear components in the presence of structural nonlinearity, and (3) the amplitude reduction of the linear responses is larger than the amplitude increase of the nonlinear components. In addition, a baseline-free fatigue crack detection technique is proposed considering the additional amplitude reduction of the linear components due to nonlinear ultrasonic modulation.

Rotating shafts in drop lifts of manufacturing facilities are susceptible to fatigue cracks as they are under repetitive heavy loading and high speed spins. However, it is challenging to use conventional contact transducers to monitor these shafts as they are continuously spinning with a high speed. In this study, a noncontact crack detection technique for a rotating shaft is proposed using air-coupled transducers (ACTs). (1) Low frequency (LF) and high frequency (HF) sinusoidal inputs are simultaneously applied to a shaft using two ACTs, respectively. A fatigue crack can provide a mechanism for nonlinear ultrasonic modulation and create spectral sidebands at the modulation frequencies, which are the sum and difference of the two input frequencies Then LF and HF inputs are independently applied to the shaft using each ACT. These three ultrasonic responses are measured using another ACT. (2) The damage index (DI) is defined as the energy of the first sideband components, which corresponding to the frequency sum and difference between HF and LF inputs. (3) Steps 1 and 2 are repeated with various combinations of HF and LF inputs. Crack existence is detected through an outlier analysis of the DIs. The effectiveness of the proposed technique is investigated using a steel shaft with a real fatigue crack.

Nonlinear vibro-acoustic modulation for microcrack detection of steel strands based on S-transform bispectrum

Noncontact detection of fatigue cracks by laser nonlinear wave modulation spectroscopy (LNWMS)

Fatigue crack and its precursor often serves as a nonlinear source, and the nonlinear ultrasonic features created by a fatigue crack have a much higher sensitivity compared with linear features. This paper presents a fatigue crack visualization technique based on noncontact laser ultrasonics and state space techniques. Under a broadband laser pulse excitation, defect nonlinearity exhibits modulation at multiple frequency peaks in a spectral plot due to interactions among various input frequency components of the broadband input. These modulations are weak and hardly discernable in both the frequency and time domains. In order to detect the nonlinear changes caused by fatigue cracks in the time domain, a state space attractor is reconstructed using a single laser pulse response and its geometrical deviations from the baseline data obtained from the pristine condition of a target structure are computed. Through scanning tests using a Q-switched Nd:YAG laser and laser Doppler vibrometer (LDV), the proposed method can be used for visualizing fatigue cracks in metallic plates.

Rotating shafts operating under complex loading conditions are susceptible to develop fatigue cracks at the localized stress concentration points. Fatigue cracks are the prime sources for catastrophic failures. Therefore, it is vital and challenging for the early fatigue crack detection to prevent costly machinery breakdowns. A rotor test rig with a facility of introducing stochastic loads on to the rotating shaft is designed and developed to induce fatigue crack on the shaft is utilized for this study. Accelerated fatigue tests are performed, along by introducing stochastic loads, on a shaft specimen seeded with circumferential V-notch configuration. The methodology for early crack detection is based on analyzing the steady-state vibration data acquired, till the shaft specimen develops fatigue crack, from various locations of the test rig using different sensors like miniature accelerometers, wireless telemetry strain gauge, and laser vibrometer. In the present study, signal processing tools like continuous wavelet transforms (CWT), Hilbert–Huang transform (HHT), and short-time Fourier transform (STFT) are employed for the early fatigue crack detection. Wavelet packet (WP) analysis is used to decompose the experimental vibration response data into narrow frequency band signals before the processing of CWT, HHT, and STFT analyses. In addition, these techniques are compared to evaluate their detection performance when employed directly on raw data and preprocessed data using WP analysis. Finally, this paper provides a detailed description and comprehensive selection of time-frequency techniques for the early fatigue crack detection.

Photoacoustics (PA) is a promising technology for medical imaging due to its superior contrast and resolution. This technique can be implemented either in time domain by powerful nanosecond laser pulses or in frequency domain (FD) by low power intensity-modulated continuous wave (CW) chirped laser irradiation (“the photoacoustic radar”). In our previous work [1,2] FD-PA imaging with linear frequency-swept optical modulation was implemented and the application of various signal processing methods to obtain depth-selective imaging was demonstrated. Later on we compared the FD and the pulsed techniques with respect to the maximum detectable depth using a dual-mode PA system [3]. Comparable depths were found with the FD technique exhibiting significantly superior signal-to-noise ratio (SNR). Very recently [4] we demonstrated that by combining the additional imaging channel that CW method possess due to its phase information with the cross-correlation amplitude results, the contrast and resolution of the image can be improved. In this study the contrast of images produced with both techniques is compared for various depths of a scattering medium. Experiments were implemented using two ultrasonic transducers with different center frequencies of 3.5 and 0.5 MHz. The samples consisted of two black rubber squares of areas 2x2 mm 2 and 4x4 mm 2 located at the same depth. These specifications were devised to enhance the comparison to the limits of the system, in other words, as we are trying to determine how small and how deep we can detect an absorber in a highly scattering medium, by comparing the contrast of the two rival methods. The wavelength of both pulsed and CW lasers utilized was 1064 nm and laser energies were controlled under the safety standard limits [5]. Since the CW method provides the capability to control the waveform shape and its bandwidth (BW), it is essential to examine the optimal BW of the frequency-modulated chirps to maximize the signal-to-noise ratio (SNR). It is demonstrated theoretically and experimentally that the optimal frequency BW does not depend exclusively on the transfer function of the ultrasonic transducer and its -6 dB BW, but also on the acoustic and optical properties of the tissue and lesion. To be specific, other than the transducer’s BW, acoustic attenuation of the medium and optical absorption and speed of sound in the lesion are key factors. Some results are demonstrated in figures 1 and 2 by utilizing the high frequency transducer. Figure 1.a shows the cross-correlation amplitude in the PA radar method generated in the middle of a black rubber square of 4x4 mm 2 in 19 mm depth of a 0.47% intralipid solution. Figure 1.b shows the transient response of the same sample to a pulsed laser. Comparison of these signal traces, demonstrates the higher SNR achieved by the CW method. Figures 2.a and b are the complete images produced at a specific depth corresponding to the surface of the black rubber samples. These images are constructed by complete scanning of a 15x15 mm 2 area using the CW and pulsed methods. It demonstrates that at the ultimate depth for the high frequency transducer, the pulsed laser achieved better detectivity. Nevertheless by employing the additional phase channel it is demonstrated that the contrast and resolution of the CW method is superior for shallower absorbers in the case of high frequency transducer. Next experiment exhibits the results of imaging a sample (absorption coefficient: 2 cm

ABSTRACTFor the fatigue crack embedded in metallic components, the non-collinear transverse wave mixing technique is applied to detect them and image the fatigue damage around crack. When two non-collinear transverse waves with same frequency meet in components, a resonance longitudinal wave can be generated if the resonance condition is satisfied; the mixing nonlinear parameter is applied to evaluate the nonlinearity of detection positions. The non-collinear wave mixing experiments are conducted on components with different length of fatigue crack. By changing the relative position of transducers, different detection positions can be controlled. According to the distribution of mixing nonlinear parameter at vertical direction, the length of fatigue crack at vertical direction can be measured. Furthermore, the mixing nonlinear parameters of different detection positions around fatigue crack are measured, the image of mixing nonlinear parameter can visualise the damage around fatigue crack which is very easy to assess the fatigue damage of components.

The impact of pulsed versus continuous wave (cw) laser excitation on the photophysical properties of single quantum dots (QDs) has been investigated in an experiment in which all macroscopic variables are identical except the nature of laser excitation. Pulsed excitation exaggerates the effects of photobleaching, results in a lower probability of long ON fluorescence blinking events, and leads to shorter fluorescence lifetimes with respect to cw excitation at the same wavelength and average intensity. Spectral wandering, biexciton quantum yields, and power law exponents that describe fluorescence blinking are largely insensitive to the nature of laser excitation. These results explicitly illustrate important similarities and differences in fluorescence dynamics between pulsed and cw excitation, enabling more meaningful comparisons between literature reports and aiding in the design of new experiments to mitigate possible influences of high photon flux on QDs.