Richard W. Neu

Predicting and quantifying pores in laser powder bed fusion (LPBF) remains limited by models tailored to narrow process regimes and by conventional thermal sensing that is highly sensitive to emissivity, both of which fail under realistic process variability. Here, we introduce a process-agnostic, context-aware deep learning framework for quantitative, multi-metric pore characterization, enabled by our patented single-camera two-wavelength imaging pyrometry (STWIP) system. This novel high-speed (up to 100,000 frames per second (fps), with 22,500 fps demonstrated in the present study), high-resolution diagnostic uniquely captures spatiotemporal melt pool thermodynamics and morphology with physical fidelity, overcoming emissivity and alignment challenges inherent to conventional monitoring. By integrating frequency-domain and temporal representations across adjacent layers and laser scans, the optimized model achieves 88% accuracy, 85% precision, and 84% recall for pore-presence detection, with 83% mean accuracy for pore count and size and 72% for porosity ratio across unseen test sets. Performance remains robust, albeit slightly lower, on entirely new builds. These results are remarkably strong given the numerically small ground-truth pore metrics and the extreme data imbalance and process variability of realistic LPBF, and they substantially exceed prior qualitative or regime-specific approaches. Our method enables robust, quantitative, and potentially transferable defect prediction across regimes, advancing real-time quality assurance in metal additive manufacturing.

ABSTRACT Creep–fatigue crack growth (CFCG) experiments were performed at 850°C and 950°C on nominally 12.70‐mm‐thick compact‐type specimens of alloy 247LC‐DS, a Ni‐base superalloy. The loading cycles consisted of trapezoidal waveforms with hold times of 2 and 20 min at the maximum force. The crack plane was oriented normally to the solidification direction. The fracture surfaces and the crack path‐microstructure interactions were characterized. Competing time‐ and cycle‐dependent mechanisms contributed to crack growth during these tests. Time‐dependent crack growth rates, , at 850°C and 950°C correlated with for 2‐min and 20‐min hold durations. These CFCG rates were also comparable to the creep crack growth (CCG) rates previously reported. Crack growth rates showed significant variability under all test conditions. This scatter was attributed to the formation of microcracks connecting into secondary cracks in the interdendritic regions aligned with the solidification direction. Material models representing the CCG and CFCG data are described.

Abstract This article examines the effects of both surface and bulk residual stresses on fatigue properties, employing various test methodologies to quantify the effects of process-induced residual stresses on microstructure and fatigue behavior. It provides key definitions and distinctions relevant to crack formation, growth, and arrest under cyclic loading conditions. The article presents the types of residual stress and their relevance to fatigue properties. It explains the objective of a fatigue test program. The article describes the influence of surface roughness on fatigue properties. It provides an overview on specimen-preparation methods and their impact on fatigue. The article also presents intrinsic fatigue properties for additive manufacturing.

Abstract This study examines the low-cycle fatigue (LCF) and creep–fatigue (CF) behavior of wrought 316L and 316H stainless steels to develop acceptance criteria for accelerated testing of additively manufactured (AM) stainless steels. New LCF and CF data were generated between 550 and 700 °C, focusing on the impact of temperature, control mode, and hold time on fatigue life. Life assessment methods, including time fraction (TF), ductility exhaustion (DE), and stress-modified DE (SMDE), were evaluated for their applicability to nuclear code cases. Results highlight that conventional accelerated CF tests often lead to fatigue-dominated failures due to insufficient hold times. DE and SMDE models correlated more effectively with experimental data than TF, particularly when non-damaging viscous strains were excluded. Testing laser powder-directed energy deposition (LP-DED) 316H revealed non-conservative life predictions across all models, contrasting with the conservative predictions for wrought 316L and 316H, despite comparable LCF and creep properties. These findings underscore the need to refine accelerated CF test protocols to better capture damage mechanisms in AM materials.

ABSTRACT Where AI systems are increasingly and rapidly impacting engineering, science, and our daily lives, progress in AI safety for physical infrastructures is lagging. Most of the research and educational programs on AI safety do not consider that, in today's connected world, safety and security in physical infrastructures are increasingly entangled. This technical note sheds light, for the first time, on how computer science and engineering communities, for example, mechanical and civil, can collaborate on addressing AI safety issues in the physical infrastructures and the mutual benefits of this collaboration. We offer examples of how probabilistic views of engineers on safety can contribute to quantifying critical parameters such as “threshold” and “safety buffer” in the AI safety models, developed by the world‐leading computer scientists. We also offer examples of how novel AI and machine learning tools, for example, do ‐operator, a mathematical operator for intervention (vs. conditioning); do ‐calculus, machinery of causal calculus; and physics‐informed neural networks with a small number of samples can help fatigue and fracture research. We envision AI safety as a process, not an object, and contribute to realizing this vision by initiating a collaborative and interdisciplinary approach in establishing this process.

Abstract Ni-base superalloys used for hot section hardware of gas turbine systems experience thermomechanical fatigue (TMF), creep, and environmental degradation. The blades and vanes of industrial gas turbines (IGTs) are made from superalloys that are either directionally-solidified (DS) or cast as single crystals (SX). Consequently, designing and evaluating these alloys is complex since life depends on the crystallographic orientation in addition to the complexities related to the thermomechanical cycling and the extent of hold times at elevated temperature. Comparisons between the more complex TMF tests and simpler isothermal low cycle fatigue (LCF) tests with hold times as cyclic test methods for qualifying alternative repair, rejuvenation, and heat-treatment procedures are discussed. Using the extensive set of DS and SX data gathered from the open literature, a probabilistic physics-guided neural network is developed and trained to estimate life considering the influence of crystallographic orientation, temperature, and several other cycling and loading parameters.

Abstract Creep crack growth experiments were performed at 750°C, 850°C, and 950°C on nominally 3 and 12.7 mm thick compact type specimens of alloy 247LC‐DS, a Ni‐base superalloy used for hot‐section gas turbine blades. The primary crack plane was transverse to the solidification direction. The crack path–microstructure interaction was characterized. Crack growth occurred in a creep‐ductile manner and data analyses utilized time‐dependent fracture mechanics. No measurable crack growth occurred at 750°C. Cracks grew by formation, growth, and coalescence of cavities on interdendritic carbides in both the primary crack plane and normal to said plane at 850°C and 950°C. The variability in the crack growth rate was higher in thicker specimens, but the mean creep crack growth rate versus C t relationship in 247LC‐DS was neither sensitive to test temperature ≥850°C nor specimen thickness. Quantitative relationships between da/dt and C t were derived for mean, upper, and lower bound creep crack growth rate trends.

Abstract Thermomechanical fatigue (TMF) crack growth of Alloy 718 when the cycle includes long high‐temperature tensile dwells is often accelerated because sustained‐load crack growth occurs during the dwells and a temperature affected zone (TAZ) develops ahead of a crack tip that reduces fatigue crack growth resistance upon application of a cycle after the long dwell. This paper presents new TMF and sustained‐load tests that aim to study and characterize the TAZ development and growth ahead of semi‐elliptical and single‐edge cracks. The influence of constraint on crack growth is assessed. This investigation focuses on a coarse‐grained (CG) Alloy 718 at 650°C. By comparing the crack growth behavior between a fine‐grained and CG microstructure after a long tensile dwell, a modification of the TAZ growth prediction model that includes the influence of grain size is proposed. Using this improved model for TAZ growth and acceleration, the TMF crack growth prediction of surface cracks is improved.

Finite element analyses were performed for varying crack tip or notch geometries and temperatures in C(T) specimens of CM247LC-DS Ni-base superalloy used in gas turbine blades to understand the accumulation of creep strains at notches. The constitutive models used considered elastic-primary-secondary creep deformation behavior. Simulations were performed under 2D plane stress, 2D plane strain, and 3D conditions for temperatures of 750°C, 850°C, and 950°C. The estimated load-line displacements as a function of time from finite element analyses and from analytical expressions were validated using results from experiments conducted on notched and fatigue precracked C(T) specimens. Differences in the development of the creep zone size due to the notch tip geometry diminish in time under plane stress conditions, but they do not for plane strain conditions. Over short periods of time, small-scale creep plays a significant role in the evolution of the notch tip creep zones. The 3D analysis results show that creep zones grow at a faster rate on the surface of the specimen than in the mid-thickness region, and these differences are exacerbated as the notch root radius decreases. This work can potentially be the basis for formulating a theory for creep crack formation from notches in CM247LC-DS alloys.

Abstract Predicting the life of thermomechanical fatigue (TMF) is challenging because there are several parameters describing the mechanical and thermal cycles including dwell periods within the cycle that can impact life. The relationships between these TMF history parameters and fatigue life are not always clear. This paper explores the use of a neural network (NN) with a probabilistic physics‐guided architecture to learn these relationships and predict the cycles to failure for a wide range of possible creep‐fatigue and thermomechanical fatigue histories. Using inputs of strain range, maximum and minimum temperature, the phasing between the thermal and mechanical cycles, cycling frequency, and dwell time in either tension or compression, the model predicts both the mean fatigue life and its confidence intervals (CI). The model is demonstrated from a comprehensive set of data on single‐crystal and directionally solidified Ni‐base superalloys extracted from the literature. The model is evaluated in several ways to determine the success of learning the relationship between the applied TMF histories and cycles to failure.