Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “melt pool”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
NASA NTRS1993

Core formation by giant impacts: Conditions for intact melt region formation

Among the many effects of high-speed, giant impacts is widescale melting that can potentially trigger catastrophic core formation. If the projectile is sufficiently large, the melt pools to form an intact melt region. The dense phase then segregates from the melt, forming a density anomoly at the melt region's base. If the anomoly produces a differential stress larger than a certain minimum, it overcomes the mantle's long-term elastic strength and rapidly forms a core. It was previously shown that giant impacts effectively trigger core formation in silicate bodies by the time they grow to the mass of Mercury and in icy bodies by the time they grow larger than Triton. In order for this process to be viable, an intact melt region must be formed. Conditions under which this occurs is examined in more detail than previously published.

Tonks, W. B.↗
NASA NTRS2023

CALPHAD Models to Guide Refractory Alloys Additive Manufacturing: In-Situ Compounds Formation, Nanoparticles, and Impurities Considerations

This technical memorandum (TM) seeks to model and discuss aspects of nanoparticle and micrometer-scale particle or ‘microparticle’ inclusions in refractory metal alloy manufacturing, via a currently used additive manufacturing (AM) method that results in equivalent wt.% addition of dissolved elements in the melt pool composition and subsequent formation of nanometer-scale MC and MO inclusions. Additionally, the authors begin to consider the effects of oxygen (O), a ubiquitous impurity, in AM; and seek to understand a combined optimization of alloying, nano- and micro strengthening and refinement, elemental additions, and even in-situ compounds and alloys species formation. To gain insights, this TM focuses on a pair of refractory alloys currently of interest to NASA: tungsten-rhenium-tantalum carbide (with O impurities) (W-Re-TaC(-O)) and tungsten-rhenium-tantalum-carbon (with O impurities) (W-Re-Ta-C(-O)), which are processed via powder bed fusion (PBF)-SLM.

ICME integrated computational materials engineerin↗
DOE OSTI2025

Massively parallel phase-field simulations targeting exascale

The interface thickness in the phase-field (PF) method limits its simulation scales. Consequently, large-scale PF simulations become prohibitively expensive for resolving the extremely fine microstructures that typically form during rapid solidification processing. This challenge is significant in predicting microstructure evolution in metal additive manufacturing and has been identified by the United States Department of Energy’s Exascale Computing Project. Here, to address this, we develop a multi-GPU and MPI-based massively parallel simulation code, utilizing state-of-the-art algorithms, software, and libraries, for large-scale three-dimensional (3D) PF simulations. We report the first GPU-parallel PF simulations on Frontier (currently the second TOP500 exascale cluster) and Summit machines, taking dendritic growth as an example problem. We evaluate the parallel performance of our implementation using scaling studies with more than 24 000 GPUs (among the largest known computations to date) and the acceleration performance using large-scale simulations of dendritic growth in 3D. Finally, massively parallel GPUs in these supercomputers enabled the first coupled multiscale simulations of laser melting and subsequent dendritic solidification on the scale of a full melt-pool, demonstrating the feasibility of performing PF simulations with a point total over 2 billion grid points within an acceptable time.

Exascale↗
NASA NTRS2012

Modeling of Melt Growth During Carbothermal Processing of Lunar Regolith

The carbothermal processing of lunar regolith has been proposed as a means to produce carbon monoxide and ultimately oxygen to support human exploration of the moon. In this process, gaseous methane is pyrolyzed as it flows over the hot surface of a molten zone of lunar regolith and is converted to carbon and hydrogen. Carbon gets deposited on the surface of the melt, and mixes and reacts with the metal oxides in it to produce carbon monoxide that bubbles out of the melt. Carbon monoxide is further processed in other reactors downstream to ultimately produce oxygen. The amount of oxygen produced crucially depends on the amount of regolith that is molten. In this paper we develop a model of the heat transfer in carbothermal processing. Regolith in a suitable container is heated by a heat flux at its surface such as by continuously shining a beam of solar energy or a laser on it. The regolith on the surface absorbs the energy and its temperature rises until it attains the melting point. The energy from the heat flux is then used for the latent heat necessary to change phase from solid to liquid, after which the temperature continues to rise. Thus a small melt pool appears under the heated zone shortly after the heat flux is turned on. As time progresses, the pool absorbs more heat and supplies the energy required to melt more of the regolith, and the size of the molten zone increases. Ultimately, a steady-state is achieved when the heat flux absorbed by the melt is balanced by radiative losses from the surface. In this paper, we model the melting and the growth of the melt zone with time in a bed of regolith when a portion of its surface is subjected to a constant heat flux. The heat flux is assumed to impinge on a circular area. Our model is based on an axisymmetric three-dimensional variation of the temperature field in the domain. Heat transfer occurs only by conduction, and effects of convective heat transport are assumed negligible. Radiative heat loss from the surface of the melt and the regolith to the surroundings is permitted. We perform numerical computations to determine the shape and the mass of the melt at steady state and its time evolution. We first neglect the volume change upon melting, and subsequently perform calculations including it. Predictions from our model are compared to test data to determine the effective thermal conductivities of the regolith and the melt that are compatible with the data

Balasubramaniam, R.↗
NASA NTRS1985

Impact melt products of chondritic material

Experimental data concerning impact melting processes in chondritic material are reviewed. It is shown that a large variety of objects in chondritic meteorites could have formed as a result of impact melting, including: shock veins; metal-troilite mixtures; metal and sulfide nodules; melt pockets and vugs. The type of object produced in an impact melt is related to the interaction of the shock waves with the particular target rock. It is suggested that various iron meteorites (including groups IAB, IIICD, and IIE, as well as several ungrouped irons) were formed from individual melt pools in chondritic regoliths. The small-scale structure of impact-melted metallic Fe,Ni and troilite in Weston (H chondrite regolith breccia) is illustrated in a photograph.

Rubin, A. E.↗
DOE OSTI2025

Effects of shielding gas and spot reduction on humping suppression in high-speed laser welding

High-speed laser welding is a crucial process to enhance the production rate of fuel cell fabrication. However, exceeding a critical welding speed leads to defect formation such as humps and root concavities, which degrade weld quality and mechanical performance. This study employed shielding gas and spot reduction approaches to suppress these defects. The relationships between individual defect types and their influence on weld strength were systematically examined. Moreover, in situ high-speed synchrotron X-ray imaging was employed to comprehend the mechanisms of humping mitigation by shielding gas through analysis of instantaneous keyhole and melt pool dynamics. The observations reveal that shielding gas mitigates humping by shortening the molten pool through enhanced surface heat convection and by altering surface tension to prevent melt flow collisions. The effects of shielding gas and spot reduction were further incorporated into the dimensionless humping index (π h ) that describes the humping formation tendency. These findings provide mechanistic insights and quantitative guidance for defect control strategies in high-speed laser welding.

08 HYDROGEN↗
NASA NTRS2005

Portable Electron-Beam Free-Form Fabrication System

A portable electron-beam free-form fabrication (EB F3) system, now undergoing development, is intended to afford a capability for manufacturing metal parts in nearly net sizes and shapes. Although the development effort is oriented toward the eventual use of systems like this one to supply spare metal parts aboard spacecraft in flight, the basic system design could also be adapted to terrestrial applications in which there are requirements to supply spare parts on demand at locations remote from warehouses and conventional manufacturing facilities. Prior systems that have been considered for satisfying the same requirements (including prior free-form fabrication systems) are not easily portable because of their bulk and massive size. The mechanical properties of the components that such systems produce are often inferior to the mechanical properties of the corresponding original, conventionally fabricated components. In addition, the prior systems are not efficient in the utilization of energy and of feedstock. In contrast, the present developmental system is designed to be sufficiently compact and lightweight to be easily portable, to utilize both energy and material more efficiently, and to produce components that have mechanical properties approximating those of the corresponding original components. The developmental EB F3 system will include a vacuum chamber and associated vacuum pumps, an electron-beam gun and an associated power supply, a multiaxis positioning subsystem, a precise wire feeder, and an instrumentation system for monitoring and control. The electron-beam gun, positioning subsystem, and wire feeder will be located inside the vacuum chamber (see figure). The electron beam gun and the wire feeder will be mounted in fixed positions inside the domed upper portion of the vacuum chamber. The positioning subsystem and ports for the vacuum pumps will be located on a base that could be dropped down to provide full access to the interior of the chamber when not under vacuum. During operation, wire will be fed to a fixed location, entering the melted pool created by the electron beam. Heated by the electron beam, the wire will melt and fuse to either the substrate or with the previously deposited metal wire fused on top of the positioning table. Based on a computer aided design (CAD) model and controlled by a computer, the positioning subsystem

Watson, J. Kevin↗
DOE OSTI2025

The effect of blue and infrared laser melting frequency on oxide morphology in 304L

Continuous Wave Laser Beam (LB) melting offers control over the localised heating and cooling of melt pools for welding, brazing, additive manufacturing and solidification. Research into laserliquid metal interactions has primarily focused on the heat and mass transport under large thermal gradients imposed by the localised melting conditions. Recent commercial availability of blue (450 nm) wavelength lasers has enabled the printing of material with poor IR (1064) nm attenuation like Cu. However, little research has been conducted into varying input laser frequency and scan rate at fixed absorption to understand the effects of laser light on surface oxide formation and the underlying liquid metal thermodynamic response. Here, this article conducts laser melting of 304L under Blue (450 nm) and Infrared (IR, 1064 nm) laser frequencies and examines their impact on laser oxide thickness, chemistry and coloration. We find that laser frequencies induce changes in the oxide layer thickness and chemistry that cannot be explained using conventional thermal absorption shifts and fluid dynamics. We suggest that light coupling may have thermodynamic implications for the chemical potential of the liquid metal, which may drive the observed phase behaviour. Identifying mechanisms for the tuning of the chemical potential via laser frequency control could enable the ability to control solidification microstructure and atomic ordering in laser welding and additive manufacturing processes.

304L↗
DOE OSTI2025

Beam Shaping and Oscillation for Metals Additive Manufacturing

Additive Manufacturing (AM) is pivotal for next-generation industrial manufacturing but is hampered by process control challenges leading to unwanted defects and sub-optimal material properties. Our research builds on prior work that utilized beam oscillation techniques to mitigate these issues (Wu et al., 2023). We focus on investigating the influence of beam oscillation frequency on melt pool dynamics, utilizing in-situ synchrotron material characterization from the ESRF (Chen et al., 2026). Preliminary findings indicated that beam oscillation results in unique fluid flow patterns and enhanced melting stability. These findings suggested the possibility of optimized AM process control, including minimized defects and tailored thermal histories. We proposed further research to explore the effects of beam oscillation frequencies comprehensively, using the in-situ characterization capabilities at the ESRF. The goal was to quantify this relationship, offering a pathway to more precise AM process control and optimization. Our research is expected to offer new scientific insights into the fluid mechanics and thermal behaviour in AM, contributing to its broader applicability and commercial viability.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Assessment of critical flaw sizes and crack driving forces during additive manufacturing of metallic materials

Additive manufacturing (AM) of complex engineering components is often plagued by a high susceptibility to cracking, particularly in high-strength metallic materials. While alloy design efforts have made progress in mitigating solidification defects, there remains a need for mechanistic guidelines to predict susceptibility to solid-state cracking. To address this gap, driving forces for the growth of melt pool cracks are calculated across a wide range of alloys using an efficient computational framework. Calculations are coupled with rapid single track laser experiments to elucidate trends in cracking from laser melting. The analyses conducted here highlight the important role of material properties in susceptibility to cracking, notably fracture toughness and elastic modulus. An important finding is that residual stresses that are limited in magnitude to the yield stress of the material are likely insufficient to drive cracking during cooling. Furthermore, the implications of these results are discussed in the context of alloy design for AM and residual stress accumulation during AM.

36 MATERIALS SCIENCE↗
NASA NTRS1999

Formation of a Martian Pyroxenite: A Comparative Study of the Nakhlite Meteorites and Theo's Flow

The unusual composition of the nakhlites, a group of pyroxenitic martian meteorites with young ages, presents an opportunity to learn about nonbasaltic magmatic activity on another planet. However, the limited number of these meteorites makes unraveling their history difficult. A promising terrestrial analog for the formation of the nakhlites is Theo's Flow in Ontario, Canada. This atypical, 120 m-thick flow differentiated in place, forming distinct layered lithologies of peridotite, pyroxenite, and gabbro. Theo's pyroxenite and the nakhlites share strikingly similar petrographies, with concentrated euhedral to subhedral augite grains set in a plagioclase-rich matrix. These two suites of rocks also share specific petrologic features, mineral and whole-rock compositional features, and size and spatial distributions of cumulus grains. The numerous similarities suggest that the nakhlites formed by a similar mechanism in a surface lava flow or shallow intrusion. Their formation could have involved settling of crystals in a phenocryst-laden flow or in situ nucleation and growth of pyroxenes in an ultramafic lava flow. The latter case is more likely and requires steady-state nucleation and growth of clusters of pyroxene grains (and olivine in the nakhlites), circulating in a strongly convecting melt pool, followed by settling and continued growth in a thickening cumulate pile. Trapped pockets of intercumulus liquid in the pile gradually evolved, finally growing Fe-enriched rims on cumulus grains. With sufficient evolution, the melt reached plagioclase supersaturation, causing rapid growth of plagioclase sprays and late-stage mesostasis growth.

Friedman, R. C.↗
NASA NTRSDate not supplied

Gaseous By-Product of Thermal Vacuum Processing of Lunar Highland Simulant

Scientists at Kennedy Space Center are advancing technologies to achieve oxygen extraction on the Lunar surface. One such technology is Molten Regolith Electrolysis (MRE) in which regolith simulant is melted under high temperatures to perform oxygen extraction through electrolysis of the melt pool. A study was conducted to understand the molten formation and proper-ties of Lunar Highlands Simulants (LHS-1) under high vacuum environments ranging between 10-6 Torr. These studies include investigating the regolith melt behavior and by-product gas analysis systems to prepare for pilot plant development and operations. The off-gassing rates and pressure build-up, gaseous compounds, and material compatibility may be a risk to MRE systems. The KSC findings are reported here for inclusion into future mis-sion architecture and operational planning. The four stages of the experiment were vacuum, resistive heating, regolith melt and gas detection. The regolith mass of 70 g was melted in a 40 cm tall x 50 cm diameter vacuum chamber at temperature increase rate 47 ℃ /min and ramp rate of 1 ampere/3 min up to an 18-amp maximum. The test is conducted for about 60 minutes and the constituent gases produced during heating of regolith is monitored with a re-sidual gas analyzer. During the molten formation, primary off-gassing volatiles included water vapor (18 amu), a peak at 44 amu (FeO2 – expected), and atomic oxygen (16 amu), among a series of other compounds that could create compatibility issues such as magnesium, chlorine, and silicon oxide. These results add to the knowledge required for successful oxygen extraction on the moon.

Molten Regolith Electrolysis↗
DOE OSTI2024

Effect of cold-cap-bottom shear stress on primary foam stability

Here, we analyze the heat transfer between the melt pool and the layer of reacting materials called cold cap, which is the determining factor for the glass production rate in Joule-heated air-bubbled nuclear waste glass melters. We consider the effects of bubbling rate, melt viscosity, melt thermal conductivity, conversion kinetics, and shear rate at the cold-cap bottom, and develop simple relationships for the cold-cap bottom temperature and glass production rate. We show that the melt convection enhances the heat transfer to the cold cap by suppressing the thermal boundary layer and primary foam, while the conversion kinetics opposes these effects. Additionally, we mention how dissolving silica particles affects the primary foam stability and discuss the limitations of the presented approach for vigorously foaming feeds.

36 MATERIALS SCIENCE↗
NASA NTRS2001

Magnetic Damping of g-Jitter Induced Double-Diffusive Convection

This paper describes a numerical study of the g-jitter driven double diffusive convective flows, thermal and concentration distributions in binary alloy melt systems subject to an external magnetic field. The study is based on the finite element solution of transient magnetohydrodynamic equations governing the momentum, thermal and solutal transport in the melt pool. Numerical simulations are conducted using the synthesized single- and multi- frequency g-jitter as well as the real g-jitter data taken during space flights with or without an applied magnetic field. It is found that for the conditions studied, the main melt flow follows approximately a lineal- superposition of velocity components induced by individual g-jitter components, regardless of whether a magnetic field exists or not. The flow field is characterized by a recirculating double diffusive convection loop oscillating in time with a defined frequency equal to that of the driving g-jitter force. An applied magnetic field has little effect on the oscillating recirculating pattern, except around the moment in time when the flow reverses its direction. The field has no effect on the oscillation period, but it changes the phase angle. It is very effective in suppressing the flow intensity and produces a notable reduction of the solutal striation and time fluctuations in the melt. For a given magnetic field strength, the magnetic damping effect is more pronounced on the velocity associated with the largest g-jitter component present and/or the g-jitter spiking peaks. A stronger magnetic field is more effective in suppressing the melt convection and also is more helpful in bringing the convection in phase with the g-jitter driving force. The applied field is particularly useful in suppressing the effect of real g-jitter spikes on both flow and solutal distributions. With appropriately selected magnetic fields, the convective flows caused by g-jitter can be reduced sufficiently and diffusion dominant. solutal transport in the melt is possible.

Shu, Y.↗
DOE OSTI2025

Computational Fluid Dynamics Simulations of Glass Vitrification Refractory Coupon Tests

The Waste Treatment and Immobilization Plant (WTP) at the Hanford site is nearing the start of the Direct-Feed Low-Activity Waste (DFLAW) operations. DFLAW is destined to convert a pretreated low activity waste portion of the 56 million gallons of tank waste into a stable solid glass. In the subsequent decade completion of the high-level waste (HLW) facility is anticipated. Sustained operational missions of both LAW and HLW melter facilities are expected over multiple decades. In high-temperature glass melters, the refractory lining corrodes over time, which could potentially be an issue for longer term operations, this refractory corrosion is higher at the level of the glass-air interface due to surface tension driven flow. The glass viscosity, melt pool temperature, and glass chemical composition can impact the rate at which the refractory corrodes. This rate is important to quantify for the various waste glasses to be produced at the WTP since the integrity of the refractory should not be a limiting factor affecting the lifetime of the melter. To this end, a series of glasses representative of the first batches of waste glass produced by the WTP will be melted in small-scale crucibles with Monofrax® K-3 coupons inserted. The corrosion of the K-3 will be measured in the melt and at the meltline (or neckline). A model for the corrosion rate will be constructed and implemented into a previously developed framework for a computational fluid dynamics (CFD) model of the full-scale WTP. To assist with experimental design and validate the implementation of the model in the full-scale melter, CFD simulations of the small-scale crucible tests were performed. The bubbling that occurs in the small-scale crucible is initially validated here with a model that uses silicone oil at room temperature. The viscosity of the oil ranges from 1 to 100 Pa•s, which corresponds to operating glass pool temperatures near 1150 °C down to idling temperatures near 950 °C. The simulation results show good agreement with the bubble sizes that form during experiments. CFD modeling of the crucible setup was used to determine bubbling characteristics to match the range of near-wall velocities expected in the full-scale WTP. This study presents the initial CFD modeling results, corrosion testing plan, and some preliminary corrosion samples with an outline for the next steps for the development of the corrosion model.

Abboud, Alexander W. [Idaho National Lab]↗
DOE OSTI2025

Effect of laser melt schedule on the microstructure of additively manufactured IN718 Superalloy

Laser powder bed fusion (L-PBF) has enabled the fabrication of geometrically complex metallic structures and components that are challenging to producing using conventional manufacturing approaches. The site-specific and far from equilibrium thermal conditions of L-PBF offer the potential to facilitate multi-length scale design of structure and properties across the atomic-through macro-levels. However, L-PBF systems face scalability challenges due to throughput constraints. Laser rotary powder bed fusion (L-RPBF) systems are being investigated as a solution to enhance the deposition rates compared to conventional L-PBF. Rotary systems also offer additional flexibility for controlling the time structure of melting through laser interleaving on alternating layers. Here, in this study, IN718 test samples were printed using single-laser or interleaved dual-laser configuration in a L-RPBF system to investigates the effect of process settings and melt-interleaving on as-fabricated microstructure. The microstructural evolution, such as grain size and crystallographic texture, was assessed by determining variations in the melt-pool shapes. Laser interleaving leads to a reduction in average grain size compared to single laser by ∼ 40 % at high power (400 W) and by ∼36 % at medium power (370 W). Results presented here identify key challenge for obtaining uniform microstructures and barriers for the broader adoption of high-deposition rate L-RPBF.

Dual-laser↗
NASA NTRS2015

Multiscale and Multiphysics Modeling of Additive Manufacturing of Advanced Materials

The objective of this proposed project is to research and develop a prediction tool for advanced additive manufacturing (AAM) processes for advanced materials and develop experimental methods to provide fundamental properties and establish validation data. Aircraft structures and engines demand materials that are stronger, useable at much higher temperatures, provide less acoustic transmission, and enable more aeroelastic tailoring than those currently used. Significant improvements in properties can only be achieved by processing the materials under nonequilibrium conditions, such as AAM processes. AAM processes encompass a class of processes that use a focused heat source to create a melt pool on a substrate. Examples include Electron Beam Freeform Fabrication and Direct Metal Deposition. These types of additive processes enable fabrication of parts directly from CAD drawings. To achieve the desired material properties and geometries of the final structure, assessing the impact of process parameters and predicting optimized conditions with numerical modeling as an effective prediction tool is necessary. The targets for the processing are multiple and at different spatial scales, and the physical phenomena associated occur in multiphysics and multiscale. In this project, the research work has been developed to model AAM processes in a multiscale and multiphysics approach. A macroscale model was developed to investigate the residual stresses and distortion in AAM processes. A sequentially coupled, thermomechanical, finite element model was developed and validated experimentally. The results showed the temperature distribution, residual stress, and deformation within the formed deposits and substrates. A mesoscale model was developed to include heat transfer, phase change with mushy zone, incompressible free surface flow, solute redistribution, and surface tension. Because of excessive computing time needed, a parallel computing approach was also tested. In addition, after investigating various methods, a Smoothed Particle Hydrodynamics Model (SPH Model) was developed to model wire feeding process. Its computational efficiency and simple architecture makes it more robust and flexible than other models. More research on material properties may be needed to realistically model the AAM processes. A microscale model was developed to investigate heterogeneous nucleation, dendritic grain growth, epitaxial growth of columnar grains, columnar-to-equiaxed transition, grain transport in melt, and other properties. The orientations of the columnar grains were almost perpendicular to the laser motion's direction. Compared to the similar studies in the literature, the multiple grain morphology modeling result is in the same order of magnitude as optical morphologies in the experiment. Experimental work was conducted to validate different models. An infrared camera was incorporated as a process monitoring and validating tool to identify the solidus and mushy zones during deposition. The images were successfully processed to identify these regions. This research project has investigated multiscale and multiphysics of the complex AAM processes thus leading to advanced understanding of these processes. The project has also developed several modeling tools and experimental validation tools that will be very critical in the future of AAM process qualification and certification.

Liou, Frank↗
DOE OSTI2026

In-situ sensor monitoring of multi-class gas porosity formation in laser powder bed fusion using convolutional neural network

In-situ monitoring of defect formation remains a significant challenge in the laser powder bed fusion (LPBF) process. Recent advances have enabled real-time defect detection with machine learning and in-situ sensing technologies; however, most studies focus on binary classification of keyhole pores, limiting nuanced multi-class pore differentiation and formation mechanisms. This work introduces a multi-class pore detection framework (no pore, small pores < 15 µm, and large pores > 15 µm) by leveraging photodiode sensor data alongside high-fidelity synchrotron X-ray imaging. The 15 µm threshold is selected to distinguish between two fundamentally different defect mechanisms, following the physical size-mechanism boundary established by prior high-resolution synchrotron X-ray characterization of Al6061 LPBF. Distinguishing these classes is critical because large keyhole pores are structurally detrimental, whereas small gas pores are often benign, requiring different process control strategies. Thermal emission monitoring data collected simultaneously with high-speed X-ray imaging at the Stanford Synchrotron Radiation Lightsource (SSRL), are correlated with subsurface melt pool dynamics to establish ground truth. Continuous Wavelet Transform (CWT) with optimized parameters converts the photodiode time-series signals into time–frequency images, facilitating feature extraction. Convolutional Neural Networks (CNN) are then applied for real-time multi-class pore classification in an average inference time of 1 ms per signal window. It achieves 79% accuracy and an Area Under the Receiver Operating Characteristic curve (AUC ROC) score of 0.89 with five-fold cross-validation. The results demonstrate that coupling CWT-based feature engineering with CNN architecture enables reliable multi-class pore detection in Al6061 builds using affordable in-situ sensors. This approach advances scalable and affordable quality assurance in additive manufacturing by moving beyond binary defect detection toward more nuanced classification of porosity mechanisms with in-situ sensors and machine learning.

Laser powder bed fusion, Multi-class pores, In-sit↗