Under the sponsorship of the National Science Foundation's Industry/University Cooperative Research Center at ISU, an effort was initiated in 2015 to repackage existing research-grade software into user-friendly tools for the rapid estimation of signal-to-noise ratio (SNR) for ultrasonic inspections of metals. The software combines: (1) a Python-based graphical user interface for specifying an inspection scenario and displaying results; and (2) a Fortran-based engine for computing defect signals and backscattered grain noise characteristics. The later makes use the Thompson-Gray measurement model for the response from an internal defect, and the Thompson-Margetan independent scatterer model for backscattered grain noise. This paper, the third in the series [1-2], provides an overview of the ongoing modeling effort with emphasis on recent developments. These include the ability to: (1) treat microstructures where grain size, shape and tilt relative to the incident sound direction can all vary with depth; and (2) simulate C-scans of defect signals in the presence of backscattered grain noise. The simulation software can now treat both normal and oblique-incidence immersion inspections of curved metal components. Both longitudinal and shear-wave inspections are treated. The model transducer can either be planar, spherically-focused, or bi-cylindrically-focused. A calibration (or reference) signal is required and is used to deduce the measurement system efficiency function. This can be "invented" by the software using center frequency and bandwidth information specified by the user, or, alternatively, a measured calibration signal can be used. Defect types include flat-bottomed-hole reference reflectors, and spherical pores and inclusions. Simulation outputs include estimated defect signal amplitudes, root-mean-square values of grain noise amplitudes, and SNR as functions of the depth of the defect within the metal component. At any particular depth, the user can view a simulated A-, B-, and C-scans displaying the superimposed defect and grain-noise waveforms. The realistic grain noise signals used in the A-scans are generated from a set of measured "universal" noise signals whose strengths and spectral characteristics are altered to match predicted noise characteristics for the simulation at hand.
At Iowa State University's Center for Nondestructive Evaluation (ISU CNDE), the use of models to simulate ultrasonic inspections has played a key role in R&D efforts for over 30 years. To this end a series of wave propagation models, flaw response models, and microstructural backscatter models have been developed to address inspection problems of interest. One use of the combined models is the estimation of signal-to-noise ratios (S/N) in circumstances where backscatter from the microstructure (grain noise) acts to mask sonic echoes from internal defects. Such S/N models have been used in the past to address questions of inspection optimization and reliability. Under the sponsorship of the National Science Foundation's Industry/University Cooperative Research Center at ISU, an effort was recently initiated to improve existing research-grade software by adding graphical user interface (GUI) to become user friendly tools for the rapid estimation of S/N for ultrasonic inspections of metals. The software combines: (1) a Python-based GUI for specifying an inspection scenario and displaying results; and (2) a Fortran-based engine for computing defect signal and backscattered grain noise characteristics. The latter makes use of several models including: the Multi-Gaussian Beam Model for computing sonic fields radiated by commercial transducers; the Thompson-Gray Model for the response from an internal defect; the Independent Scatterer Model for backscattered grain noise; and the Stanke-Kino Unified Model for attenuation. The initial emphasis was on reformulating the research-grade code into a suitable modular form, adding the graphical user interface and performing computations rapidly and robustly. Thus the initial inspection problem being addressed is relatively simple. A normal-incidence pulse/echo immersion inspection is simulated for a curved metal component having a non-uniform microstructure, specifically an equiaxed, untextured microstructure in which the average grain size may vary with depth. The defect may be a flat-bottomed-hole reference reflector, a spherical void or a spherical inclusion. In future generations of the software, microstructures and defect types will be generalized and oblique incidence inspections will be treated as well. This paper provides an overview of the modeling approach and presents illustrative results output by the first-generation software.
This paper describes a research effort to model the geometry and ultrasonic responses of naturally occurring titanium hard-alpha defects. The hard-alpha defects were first geometrically reconstructed based on metallographs obtained from destructive sectioning. Using the reconstruction data, ultrasonic models were then developed to predict the defects’ responses in specific experimental configurations. The predicted responses were subsequently compared with the experimental observations, and error analyses were performed on the predictions to examine their accuracy, consistency as well as other underlying inter-relationships.
Recently, computer models of inspection processes are emerging as important simulation tools in many aspects of applications of nondestructive evaluation. This work reports one such application to the design of test samples for studying the detection of hardalpha inclusions in titanium alloys. The infrequent occurrence of natural hard-alpha defects has necessitated extensive use of synthetic-hard-alpha (SHA) test samples. However, in manufacturing these titanium test samples, the traditional "trial and error" practice has proven to be costly and time-consuming. The current goal of designing a block simulating a titanium billet containing numerous SHA defects, with properties approximating those of natural hard-alpha defects, presents an even more complex challenge.
The ability of terahertz (THz) electromagnetic waves to penetrate a wide range of materials gives potential for diverse applications in nondestructive evaluation, biomed, and agriculture and there has been rapid expanding both in its use. One possible application is in relation to corn breeding, specifically when the doubled haploid method is used as a process that greatly speeds up plant breeding, and this requires seed sorting. Haploid kernels are induced in corn plants in order to decrease the time to reach homozygous genetic corn lines. These haploid kernels must be separated from the surrounding diploid kernels; presently this is labor intensive and performed using visual markers. This current work represents a proof of concept study which sought to determine if haploid classification can be automated using terahertz time domain spectroscopy (THz-TDS) with data analysis paired with a machine learning algorithm, such as a probabilistic neural network (PNN). In this work, a THz-TDS system was used to collect time domain waveforms from a sample of mixed haploid and diploid corn kernels. Effects of variabilities in beam focus and kernel geometry were reduced by taking multiple scans at different heights. The waveform data were then transformed to the frequency domain and further classified by PNN with a training set random subsampling technique. Leave-one-out and K-folds cross-validation procedures were used to train the model. The preliminary results show promise yielding an average classification rate of 75 percent correct by 5-fold cross-validation.
Recently, terahertz ray imaging has emerged as one of the most promising new NDE techniques, and new systems are being developed for applications. In this work, we conducted system calibration on a new time‐domain spectroscopy system, and then utilized this system to characterize glass‐fiber composite plates and polyimide resin disks. Extensive experimental measurements in thru‐transmission mode were made to map out the T‐ray beam pattern in free space as well as to scan these two test materials. Material properties such as index of fraction and absorption coefficient are of the primary interest. Both results were shown in good agreement with known data. Using these characterized material properties, we also demonstrated accurate modeling of the T‐ray signal propagating through the polyimide resin disk.
'%3e%3cg id='b'%3e%3cpath id='a' stroke='%23000' stroke-width='2' d='M-6-25H6m-12 3H6m-12 3H6'/%3e%3cuse xlink:href='%23a' y='44'/%3e%3c/g%3e%3cpath stroke='%23fff' d='M0 17v10'/%3e%3ccircle r='12' fill='%23cd2e3a'/%3e%3cpath fill='%230047a0' d='M0-12A6 6 0 0 0 0 0a6 6 0 0 1 0 12 12 12 0 0 1 0-24z'/%3e%3c/g%3e%3cg transform='rotate(-123.69)'%3e%3cuse xlink:href='%23b'/%3e%3cpath stroke='%23fff' d='M0-23.5v3M0 17v3.5m0 3v3'/%3e%3c/g%3e%3c/svg%3e)
