The American Society for Testing and Materials (ASTM) is an independent organization devoted to the development of standards.

American Society for Testing and Materials

A review of recent research on mechanics of multifunctional composite materials and structures

The paper provides a state of the art review of guided wave based structural health monitoring (SHM). First, the fundamental concepts of guided wave propagation and its implementation for SHM is explained. Following sections present the different modeling schemes adopted, developments in the area of transducers for generation, and sensing of wave, signal processing and imaging technique, statistical and machine learning schemes for feature extraction. Next, a section is presented on the recent advancements in nonlinear guided wave for SHM. This is followed by section on Rayleigh and SH waves. Next is a section on real-life implementation of guided wave for industrial problems. The paper, though briefly talks about the early development for completeness,. is primarily focussed on the recent progress made in the last decade. The paper ends by discussing and highlighting the future directions and open areas of research in guided wave based SHM.

https://iopscience.iop.org/article/10.1088/0964-1726/25/5/053001/pdf

Guided wave based structural health monitoring: A review

Electrospun nanofiber reinforced composites: a review

Data fusion and machine learning for industrial prognosis: Trends and perspectives towards Industry 4.0

Convolution Hierarchical Deep-Learning Neural Network Tensor Decomposition (C-HiDeNN-TD) for high-resolution topology optimization

A split spectrum processing (SSP) method is proposed to accurately determine the time-of-flight (ToF) of damage-scattered waves by comparing the instantaneous amplitude variation degree (IAVD) of a wave signal captured from a damage case with that from the benchmark. The fundamental symmetrical (S0) mode in aluminum plates without and with a notch is assessed. The efficiency of the proposed SSP method
and Hilbert transform in determining the ToF of damage-scattered S0 mode is evaluated for damage identification when the wave signals are severely contaminated by noise. Broadband noise can overwhelm damage-scattered wave signals in the time domain, and the Hilbert transform is only competent fordetermining the ToF of damage-scattered S0 mode in a noise-free condition. However, the calibrated IAVD of the captured wave signal is minimally affected by noise, and the proposed SSP method is capable of determining the ToF of damage-scattered S0 mode accurately even though the captured wave signal is severely
contaminated by broadband noise, leading to the successful identification of damage (within an error on the order of the damage size) using a triangulation algorithm.

A split spectrum processing of noise-contaminated wave signals for damage identification

Convolution hierarchical deep-learning neural network (C-HiDeNN) with graphics processing unit (GPU) acceleration

The Spires and roughness elements are most widely used to simulate the atmospheric boundary layer in wind tunnel experiments and have successfully simulated the profiles of wind speed and turbulence intensity for various terrains. With the further development of wind engineering, it is helpful to understand the mechanism of this technique so as to accurately simulate the power spectral density and integral scale of turbulence. Some tests show that spires behave as vortex generators by flow separation on the edges of their windward plate, meanwhile, their linear decrease of blockage with height brings about a corresponding linear wind profile downstream. Roughness elements physically simulate the roughness on earth surface and adjust the profiles of wind speed and turbulence intensity in wind tunnel. However, it is obvious that the profiles of simulated power spectra and integral scales with height are contrary to those in natural atmospheric boundary layer. Based on the knowledge of the mechanism, an improved spire shape has been proposed to simulate the required turbulence profiles for rather large scaled model tests.

Discussion on the simulation of atmospheric boundary layer with spires and roughness elements in wind tunnels

Nonlinear guided waves are sensitive to small-scale fatigue damage that may hardly be identified by traditional
techniques. A characterization method for fatigue damage is established based on nonlinear Lamb waves in conjunction
with the use of a piezoelectric sensor network. Theories on nonlinear Lamb waves for damage detection are first
introduced briefly. Then, the ineffectiveness of using pure frequency-domain information of nonlinear wave signals for
locating damage is discussed. With a revisit to traditional gross-damage localization techniques based on the time of
flight, the idea of using temporal signal features of nonlinear Lamb waves to locate fatigue damage is introduced. This
process involves a time-frequency analysis that enables the damage-induced nonlinear signal features, which are either
undiscernible in the original time history or uninformative in the frequency spectrum, to be revealed. Subsequently, a
finite element modeling technique is employed, accounting for various sources of nonlinearities in a fatigued medium. A
piezoelectric sensor network is configured to actively generate and acquire probing Lamb waves that involve damageinduced
nonlinear features. A probability-based diagnostic imaging algorithm is further proposed, presenting results in
diagnostic images intuitively. The approach is experimentally verified on a fatigue-damaged aluminum plate, showing
reasonably good accuracy. Compared to existing nonlinear ultrasonics-based inspection techniques, this approach uses a
permanently attached sensor network that well accommodates automated online health monitoring; more significantly, it
utilizes time-domain information of higher-order harmonics from time-frequency analysis, and demonstrates a great
potential for quantitative characterization of small-scale damage with improved localization accuracy.

Ye Lu

Papers

Convolution Hierarchical Deep-Learning Neural Network Tensor Decomposition (C-HiDeNN-TD) for high-resolution topology optimization

A split spectrum processing of noise-contaminated wave signals for damage identification

Convolution hierarchical deep-learning neural network (C-HiDeNN) with graphics processing unit (GPU) acceleration

Discussion on the simulation of atmospheric boundary layer with spires and roughness elements in wind tunnels

Fatigue damage localization using time-domain features extracted from nonlinear lamb waves