Metal additive manufacturing (AM) has started to overshadow traditional manufacturing practices thanks to its ability to produce complex, high-performance and application-customized components. However, AM process parameters have not been optimized, leading to inconsistencies and imperfections such as cracks and pores in parts, as well as deviations from the original design. Nondestructive evaluation (NDE) methods used for part qualification such as x-ray computed tomography (CT) and conventional ultrasonic testing (UT) have limitations in their abilities. X-ray CT is costly, hazardous, and offers limited resolution for larger components while many UT methods have limited applicability for inspection of parts with complex geometries or rough surfaces. Here, we conduct an integrated numerical and experimental study to investigate the feasibility of resonance ultrasound spectroscopy (RUS) as an alternative NDE method to inspect complex AM lattice structures with a varying number of missing struts. The most encouraging results are obtained when test samples have traction-free boundary conditions. The results of numerical simulations including eigenfrequency and frequency domain analyses are promising, indicating that the pristine and defective lattice samples should theoretically be distinguishable. In addition, given a reference intact sample, characterizing the extent of the defect in terms of the number of missing struts appears feasible. We introduce a similarity metric to compare the spectra after being locally normalized. However, the experimental results are not as conclusive. Although pristine and defective lattices may be distinguished for some cases, the number of missing struts cannot be inferred. The discrepancies between the numerical and experimental results are likely due to our simplified assumptions about material properties in numerical simulations and/or the presence of other unaccounted defects and heterogeneities in test samples.

Additive manufacturing (AM) is becoming increasing popular owing to its ability to manufacture geometrically complex parts and produce customer-designed parts faster than traditional machining. One of the challenges of creating high quality AM parts is that the AM process often produces defects that are difficult to detect. A number of techniques have been used to evaluate the quality of AM parts, such as traditional ultrasonic testing and x-ray micro computed tomography (micro-CT) scans. These methods are not ideal, as traditional ultrasonic testing can require multiple tests to evaluate the entire part, while micro-CT has difficulty detecting small defects in large parts. Resonance-based ultrasonic methods have the advantage of only requiring one testing configuration to evaluate the entire part. Nonlinear resonance ultrasound spectroscopy (NRUS) is a resonance-based nondestructive testing (NDT) technique for material characterization that is especially sensitive to small-scale imperfections such as microscopic cracks. Previous NRUS tests have shown correlations between the parameters measured by NRUS and the fatigue life (fatigue endurance) of a small set of samples, indicating the potential of NRUS for evaluating the build quality of AM parts as related to their performance. However, these measurements on AM metals show large variability due to the experimental setup used. Typical NRUS tests involve bonding the sample to an excitation source that induces vibration in the sample. Unfortunately, the bonding introduces artifacts in the measurements leading to the observed large measurement variability. In this study, we seek to evaluate the use of non-contact excitation sources for NRUS testing with the goal of improving the measurement repeatability. We compare the NRUS measurements using contact and non-contact excitations on wrought and AM 316L stainless steel samples with several different heat treatments. This study suggests the improved repeatability of linear resonance frequency measurements when using an air-coupled transducer. However, the intensity of resulting excitations is not sufficient for NRUS measurements, which require higher excitation voltages. We propose two additional approaches for non-contact NRUS measurements: one using a high-power laser and the other using an air cavity.

The goal of this research is to find new noninvasive methods to certify the quality of safety-critical additively manufactured (AM) metallic parts for use in industries such as aerospace and defense. Additive manufacturing facilitates rapid prototyping, building, and repairing of custom components with increased agility, production rate, and reduced waste. A recognized barrier to the wide adoption of additive manufacturing is the lack of new approaches for AM part qualification. Our research objective is to exploit the material's linear and nonlinear ultrasonic response - which represents the measurable changes and distortion in elastic waves encountering macroscopic and microscopic defects - to establish links between microstructure and macroscale mechanical properties of AM metals. We measure linear and nonlinear ultrasonic parameters for a series of AM and wrought 316L grade stainless steel samples and compare the obtained parameters against mechanical properties of the samples measured on corresponding coupons. The samples are heat-treated to different temperatures to induce microstructural changes which alter their mechanical properties and ultrasonic response. Two sets of specimens are manufactured, one from the additive manufacturing method Laser Powder Bed Fusion (L-PBF), and the second from a traditional wrought method. Using the nonlinear ultrasonic method of Second Harmonic Generation (SHG), the acoustic nonlinearity parameter is estimated. SHG has been shown to offer a highly sensitive response to microstructural heterogeneities such as dislocations and grain boundaries. A linear ultrasonic parameter, wave speed, is also recorded with pulse-echo testing. Alongside these ultrasonic measurements, mechanical testing parameters including elastic moduli and yield strength are evaluated for the specimens. To accompany the experimental testing, a series of numerical simulations were conducted using commercial finite-element software to study the effects of randomly distributed heterogeneities on wave distortion in a controlled environment. In these simulations, randomly generated heterogeneities are scattered throughout a 2D plate with materials properties different from the bulk material. Ultrasonic wave propagation is simulated within this heterogeneous medium to investigate the effects of the heterogeneities' elastic properties, geometry, and distribution on ultrasonic signals, including distortion measured in terms of higher harmonic generation (HHG). Experimental results indicate correlations between the nonlinearity parameter and both ultimate tensile strength and yield strength, where nonlinearity generally decreases as these mechanical parameters increase, particularly in the AM samples. We hypothesize that microstructural changes in grain size and distribution through the heat treatment process influence these trends in measured nonlinearity. Additionally, substructures at even smaller length scales, such as nanoscale precipitates and dislocations affect the ultrasonic and mechanical behavior. Measurements of elastic moduli and total elongation do not exhibit trends with the nonlinearity parameter. The linear parameter, wave speed, does not correlate well with the mechanical parameters, which is attributed to its lack of sensitivity to detect changes in microscopic features. These results show promising evidence for the feasibility of AM parts qualification using nondestructive nonlinear ultrasonic testing. Results of the simulations indicate that changes in heterogeneity size, volume fraction, and material property deviations from the bulk material affect HHG to varying degrees. As expected, heterogeneities of smaller sizes and volume fractions have a less significant effect. However, at increasingly large values, changes in HHG are more pronounced, and material density and stiffness deviations from the bulk material are shown to have a larger effect on HHG. Future work includes continuing nonlinear ultrasonic testing, as well as comparing results to nonlinear resonant ultrasound spectroscopy (NRUS). New geometries and materials will be tested to expand the dataset. Microstructures will be imaged using scanning and transmission electron microscopy (SEM, TEM) and evaluate our hypotheses, and further complexity in numerical simulations will be implemented to isolate microstructural features and explore their effects on material behavior.

Nondestructive evaluation (NDE) procedures are needed for materials processing, as well as for post-process materials testing. They play important roles in product design, analysis of service-life expectancy, manufacturing and quality control of manufactured products. They are also essential to on-line monitoring of the integrity of structural elements and complex systems. Rational accept and reject criteria should be based on NDE tests. Critical safety, efficiency and operational features of large-scale structures depend on adequate NDE capabilities. The lectures presented in this volume are concerned with quantitative ultrasonic NDE. They present fundamental concepts and basic theory, as well as applications to the detection of cracks and the evaluation of material properties. The following topics are discussed: basic wave propagation theory for ultrasonic NDE; piezoelectric transducers, EMATS and ultrasonic spectroscopy; laser-based ultrasonics; acoustoelasticity; ultrasound in solids with porosity, microcracking and polycrystalline structuring; the determination of mechanical properties of composite materials; inverse problems and imaging.

The use of mechanical resonances to test properties of materials is perhaps older than the industrial revolution. Early documented cases of British railroad engineers tapping the wheels of a train and using the sound to detect cracks perhaps mark the first real use of resonances to test the integrity of high-performance alloys. Attempts were made in the following years to understand the resonances of solids mathematically, based on the shape and composition. But Nobel Laureate Lord Rayleigh best summarized the state of affairs in 1894, stating {open_quotes}the problem has, for the most part, resisted attack{close_quotes}. More recently, modern computers and electronics have enabled Anderson and co-workers with their work on minerals, and our work at Los Alamos on new materials and manufactured components to advance the use of resonances to a precision non-destructive testing tool that makes anisotropic modulus measurements, defect detection and geometry error detection routine. The result is that resonances can achieve the highest absolute accuracy for any dynamic modulus measurement technique, can be used on the smallest samples, and can also enable detection of errors in certain classes of precision manufactured components faster and more accurately than any other technique.

This paper describes resonant ultrasound spectroscopy (RUS) as a powerful and established technique for measuring elastic constants of a material with general anisotropy. The first step of this technique consists of extracting resonance frequencies and damping from the vibrational frequency spectrum measured on a sample with free boundary conditions. An inversion technique is then used to retrieve the elastic tensor from the measured resonance frequencies. As originally developed, RUS has been mostly applicable to (i) materials with small damping such that the resonances of the sample are well separated and (ii) samples with simple geometries for which analytical solutions exist. In this paper, these limitations are addressed with a new RUS approach adapted to materials with high damping and samples of arbitrary geometry. Resonances are extracted by fitting a sum of exponentially damped sinusoids to the measured frequency spectrum. The inversion of the elastic tensor is achieved with a genetic algorithm, which allows searching for a global minimum within a discrete and relatively wide solution space. First, the accuracy of the proposed approach is evaluated against numerical data simulated for samples with isotropic symmetry and transversely isotropic symmetry. Subsequently, the applicability of the approach is demonstrated using experimental data collected on a composite structure consisting of a cylindrical sample of Berea sandstone glued to a large piezoelectric disk. In the proposed experiments, RUS is further enhanced by the use of a 3-D laser vibrometer allowing the visualization of most of the modes in the frequency band studied.