Deep learning with mixup augmentation for improved pore detection during additive manufacturing.

In additive manufacturing (AM), process defects such as keyhole pores are difficult to anticipate, affecting the quality and integrity of the AM-produced materials. Hence, considerable efforts have aimed to predict these process defects by training machine learning (ML) models using passive measurements such as acoustic emissions. This work considered a dataset in which keyhole pores of a laser powder bed fusion (LPBF) experiment were identified using X-ray radiography and then registered both in space and time to acoustic measurements recorded during the LPBF experiment. Due to AM's intrinsic process controls, where a pore-forming event is relatively rare, the acoustic datasets collected during monitoring include more non-pores than pores. In other words, the dataset for ML model development is imbalanced. Moreover, this imbalanced and sparse data phenomenon remains ubiquitous across many AM monitoring schemes since training data is nontrivial to collect. Hence, we propose a machine learning approach to improve this dataset imbalance and enhance the prediction accuracy of pore-labeled data. Specifically, we investigate how data augmentation helps predict pores and non-pores better. This imbalance is improved using recent advances in data augmentation called Mixup, a weak-supervised learning method. Convolutional neural networks (CNNs) are trained on original and augmented datasets, and an appreciable increase in performance is reported when testing on five different experimental trials. When ML models are trained on original and augmented datasets, they achieve an accuracy of 95% and 99% on test datasets, respectively. We also provide information on how dataset size affects model performance. Lastly, we investigate the optimal Mixup parameters for augmentation in the context of CNN performance.

Predicting local material thickness from steady-state ultrasonic wavefield measurements using a convolutional neural network.

Acoustic steady-state excitation spatial spectroscopy (ASSESS) is a full-field, ultrasonic non-destructive evaluation (NDE) technique used to locate and characterize defects in plate-like structures. ASSESS generates a steady-state, single-tone ultrasonic excitation in a structure and a scanning laser Doppler Vibrometer (LDV) measures the resulting full-field surface velocity response. Traditional processing techniques for ASSESS data rely on wavenumber domain analysis. This paper presents the alternative use of a convolutional neural network (CNN), trained using simulated ASSESS data, to predict the local plate thickness at every pixel in the wavefield measurement directly. The defect detection accuracy of CNN-based thickness predictions are shown to improve for defects of greater size, and for defects with higher thickness reductions. The CNN demonstrates the ability to predict thickness accurately in regions where Lamb wave dispersion relations are complex or unknown, such as near the boundaries of a test specimen, so long as the CNN is trained on data that accounts for these regions. The CNN also shows generalizability to ASSESS experimental data, despite an entirely simulated training dataset.

Estimating the effective Reynolds number in implicit large-eddy simulation.

In implicit large-eddy simulation (ILES), energy-containing large scales are resolved, and physics capturing numerics are used to spatially filter out unresolved scales and to implicitly model subgrid scale effects. From an applied perspective, it is highly desirable to estimate a characteristic Reynolds number (Re)-and therefore a relevant effective viscosity-so that the impact of resolution on predicted flow quantities and their macroscopic convergence can usefully be characterized. We argue in favor of obtaining robust Re estimates away from the smallest scales of the simulated flow-where numerically controlled dissipation takes place and propose a theoretical basis and framework to determine such measures. ILES examples include forced turbulence as a steady flow case, the Taylor-Green vortex to address transition and decaying turbulence, and simulations of a laser-driven reshock experiment illustrating a fairly complex turbulence problem of current practical interest.

Buoyancy driven mixing of miscible fluids by volumetric energy deposition of microwaves.

An experiment that seeks to investigate buoyancy driven mixing of miscible fluids by microwave volumetric energy deposition is presented. The experiment involves the use of a light, non-polar fluid that initially rests on top of a heavier fluid which is more polar. Microwaves preferentially heat the polar fluid, and its density decreases due to thermal expansion. As the microwave heating continues, the density of the lower fluid eventually becomes less than that of the upper, and buoyancy driven Rayleigh-Taylor mixing ensues. The choice of fluids is crucial to the success of the experiment, and a description is given of numerous fluid combinations considered and characterized. After careful consideration, the miscible pair of toluene/tetrahydrofuran (THF) was determined as having the best potential for successful volumetric energy deposition buoyancy driven mixing. Various single fluid calibration experiments were performed to facilitate the development of a heating theory. Thereafter, results from two-fluid mixing experiments are presented that demonstrate the capability of this novel Rayleigh-Taylor driven experiment. Particular interest is paid to the onset of buoyancy driven mixing and unusual aspects of the experiment in the context of typical Rayleigh-Taylor driven mixing.