Porosity limits the performance of laser-based directed energy deposited (DED-LB) aluminum alloys. In this study, in situ synchrotron X-ray imaging captured the complete lifecycle of porosity including nucleation, growth, coalescence, and upward motion during DED-LB of a recycled and repurposed aluminum alloy. The observations confirm that the formation of porosity is primarily induced by hydrogen. A high hydrogen concentration in the repurposed powder (~500 ppm), combined with the sharp drop in hydrogen solubility from the liquid aluminum to solid state (<1 ppm), drives the formation of pores with a diameter over 300 μm. Direct observations of bubble initiation and arrest enable the estimation of a critical cooling rate of ~764 K s⁻¹, which restricts bubble escape. It was further discovered that an in-situ laser remelting strategy promotes efficient bubble escape by maintaining buoyancy and reducing drag. These insights establish a pathway for mitigating the hydrogen-induced porosity in additive manufacturing.

Laser powder bed fusion (LPBF) additive manufacturing (AM) process has been used in production of metal components in industries such as aerospace, automotive, and defense. However, challenges including porosity formation and microstructure inconsistency can compromise part performance and reliability. This study presents an adapted Cellular Automaton (CA) model for predicting microstructure evolution and hydrogen porosity formation in an Al-10 wt.%Si alloy during LPBF process. Key developments to the CA framework include the incorporation of a melt pool temperature profile, a non-equilibrium grain growth mechanism accounting for high cooling rates in AM, and a refined hydrogen pore nucleation criterion. The model captures the microstructure transition from fine equiaxed grains at the edge of the melt pool to elongated and finally large equiaxed grains at the melt pool center. Two different types of porosity morphology are simulated and validated: spherical and inter-granular. Sizes and distributions of pores and grains are compared with experimental results, showing good agreement. Future work will be aimed at extending the model to multi-layer and multi-beam configurations for broader applicability in large-scale manufacturing scenarios.

Externally solidified crystals (ESCs) can easily form during high pressure die casting (HPDC) of aluminum alloys if the process is not properly controlled, leading to significant reduction of casting quality and performance. This study explores seven machine learning (ML) models, including decision tree, random forest, logistic regression, neural network, K-nearest neighbors (KNN), support vector machine (SVM), and naïve Bayes classifier. The random forest and classification tree models showed the highest accuracy of 95

High-pressure die casting (HPDC) is extensively utilized in the automotive industry because it can efficiently manufacture complex components. But the components produced can suffer from porosity defects that deteriorate the subsequent mechanical properties of the castings. To address this problem, the highly integrated die casting (HIDC) process was developed and tested to significantly suppress porosity defects. However, the mechanism of defect formation and the evolution law of flow and solidification of the HIDC process have not been fully elucidated. A refined three-dimensional model with dynamic mesh was developed to simulate the entire transient HIDC process of ADC12 aluminum alloy. The model encompasses low-pressure filling, depressurization, high-pressure injection, and solidification. Experimental data were used to define the inlet velocity and piston position curves as initial conditions. Additionally, the influence of liquid fraction on the wall heat transfer coefficient was considered. The simulation results agree well with the experimental data in terms of porosity area and defect distributions. The evolution of defect formation and solidification fraction in the HIDC process was demonstrated. In addition, the effects of filling percentage and slow injection speed were discussed. It was found that a filling percentage of 60% is the optimal value for balancing gas inclusions and stratified solidification defects, which is in good agreement with the experimental results. Increasing the slow injection speed from 0.09 m/s to 0.16 m/s mitigated gas entrapment caused by melt backflow. Consequently, the predicted porosity decreased from 7.2% to 0.29%.

This study systematically investigates the effects of heat treatment on the fatigue behavior of Mg-9Al-0.4Zn1.5Sn alloy under both room temperature and an elevated temperature of 150 degrees C. Fatigue cracks mainly nucleate at casting defects (porosity, shrinkage cavities, old oxides) and surface grains. At the elevated temperature, oxidation accelerates at crack tips under cyclic loading, where brittle oxides act as stress concentrators and secondary initiation sites, promoting crack branching and growth. Fine precipitates, residual phases at grain boundaries, and Mg2Sn phase obstruct dislocation glide, leading to dislocation pileups and higher resistance to plastic flow and increased stress amplitude-known as cyclic hardening. The appearance of curved and linear dislocations after fatigue at the elevated temperature indicates thermally activated processes. Non-basal slip improves fatigue resistance by promoting uniform deformation and delaying localized damage.

The demand for lightweight, high-efficiency components in electric vehicles (EVs) highlights the critical need for reliable Al-Cu joints with superior electrical and thermal conductivity. While diffusion bonding has emerged as a promising approach, interfacial impurities and voids often degrade joint quality and conductivity. Conventional manual polishing was initially employed to prepare Cu and Al surfaces; however, this method proved insufficient in consistently removing oxides and contaminants, leading to non-uniform bonding. In addition, the larger surface area of the samples made traditional polishing impractical, further motivating the use of electropolishing. To overcome these limitations, we introduce electropolishing pretreatment to achieve cleaner, void-free interfaces. Electropolishing effectively dissolves surface asperities and contaminants, enabling intimate atomic contact during bonding and minimizing the formation of brittle intermetallic phases. A systematic investigation of bonding parameters was conducted using a custom-designed graphite clamping system. Microstructural analyses reveal that advanced polishing plays a pivotal role in producing uniform, impurity-free interfaces, resulting in reduced intermetallic thickness, improved bonding strength, and enhanced current-carrying capability. This study demonstrates the clear advantages of electropolishing over conventional polishing and establishes a scalable pathway to manufacture high-performance conductive joints for next-generation EV motor and power distribution systems.

Low-cycle fatigue behavior of a T6-treated Mg–9Al–1Sn–T6 alloy was investigated at both room temperature and an elevated temperature (ET) of 423 K (150 °C). The alloy exhibited continuous work hardening at room temperature, while it showed cyclic saturation at 150 °C. When the strain amplitude reached 0.6 pct, the stress amplitude of the Mg–9Al–1Sn–T6 alloy at 150 °C initially increased slightly, followed by a gradual decrease. Fatigue deformation of the alloy is predominantly governed by dislocation slip. The presence of both curved and linear dislocations in the microstructure at 150 °C indicates that high temperature activates nonbasal slip systems. Precipitates and residual phases impede dislocation movement, leading to an increase in stress amplitude. Fatigue cracks primarily originate from casting defects and surface grains, regardless of testing temperature. At 150 °C, oxides formed during crack propagation, acting as new initiation sites and promoting crack growth, thereby decreasing the fatigue life.

We present DendriX, an open and extensible scientific software platform for two-dimensional (2D) and three-dimensional (3D) simulation of lithium dendrite growth governed by strongly coupled nonlinear multiphysics equations. DendriX integrates block-structured adaptive mesh refinement with flexible numerical solver strategies to improve computational efficiency. The software employs finite-volume discretization, supports both explicit and implicit time integration, and allows user-selectable solver configurations to balance accuracy, stability, and computational performance. Combined with hybrid MPI–OpenMP parallelization, DendriX enables scalable and efficient high-resolution 3D simulations and achieves improved computational efficiency compared with widely used commercial multiphysics software in representative benchmark cases. Beyond lithium dendrite growth, DendriX provides a maintainable and extendable framework applicable to a broad range of phase-field and multiphysics problems requiring adaptive resolution and high-performance computing.

Much of the aluminum shape castings for critical automotive structures are made of high-carbon-footprint primary aluminum alloys via high-pressure die casting. With the rapid growth of mega and giga aluminum castings for electric vehicles and the need for high integrity cast products, there is a dire need to reduce carbon footprint and energy usage. Advances and developments in sustainable shape casting of aluminum automotives structures are reviewed. Specifically, (i) a newly developed low-carbon-footprint cast aluminum alloy that is made of 100

Aluminum high-pressure die castings (HPDC) are widely used in the automotive and other industries to achieve lightweight components with high productivity. However, the formation of externally solidified crystals (ESCs) during HPDC process can significantly reduce the mechanical performance, particularly the elongation and fatigue, of these cast parts. ESCs can be classified into two main types: Type I, consisting of large α-Al dendrites, and Type II, characterized by large crystals with fine dendrites that exhibit a clear boundary with the matrix. This study investigates the formation mechanisms of these two types of ESCs during the HPDC process. The effects of various process parameters on the formation and movement of ESCs were analyzed through high-pressure die casting trials, computer simulations, and water analog experiments. The investigation suggests that both types of ESCs start within the shot sleeve. Type I ESCs form and float within the melt, while Type II ESCs develop along the shot sleeve wall and plunger tip post-pouring. Increasing the melt temperature was found to reduce the formation of Type II ESCs. Both types of ESCs are carried into the die cavity during the filling process. Their distribution is closely related to the fast shot speed and the turbulence of the molten aluminum. Notably, a reduced fast shot speed was found to significantly decrease the number of ESCs transported into the die cavity, especially Type II ESCs. This reduction led to a considerable improvement in the mechanical performance of the cast components, particularly in terms of elongation. These findings on ESC formation as related to process parameters provide important guidance to achieve high-performance castings in industrial applications.

The emergence of machine learning (ML) techniques has significantly improved the accuracy and efficiency in materials characterization. This paper reviews the application of ML algorithms in microstructural analysis and defect detection processes of aluminum castings. By leveraging ML methods, multiple ML models were trained to automatically identify and classify different types of casting defects and microstructural features. Advanced image processing techniques, combined with convolutional neural networks (CNNs), enable the detection of casting defects such as shrinkage porosity, oxides and multiscale microstructure features (i.e., eutectic phases and secondary dendrite arm spacing of aluminum). This study highlights the advantages of the developed ML models in the accuracy and reduction of measurement time in the lab and reducing the reliance on manual analysis and subjective judgment. The findings emphasize the significant impact of ML techniques on metallurgical research and industrial applications, enhancing the reliability and performance of material analysis tools.

Iron (Fe) is considered an impurity for aluminum alloy AA6061, and its content can increase significantly due to variations in the feedstock, especially when scrap aluminum is added in alloy production. The influence of Fe content on phase formation and hot tearing susceptibility (HTS) in AA6061 during solidification was investigated. A weighted average hot tearing index was applied to evaluate the HTS in AA6061 alloys with 0.3 wt pct Fe and 0.7 wt pct Fe. It was found that the higher Fe content reduced HTS, with the weighted average hot tearing index decreasing from 3 in AA6061-0.3Fe to 1.3 in AA6061-0.7Fe. Increasing Fe content from 0.3 to 0.7 pct resulted in an increased volume fraction and refined particle size of Fe-rich intermetallic phases. Meanwhile, the formed β-Al5FeSi phase transferred into α-Al8Fe2Si that distributed in the eutectic region, which provided strong bonding and additional feeding in inter-dendritic regions, mitigating crack initiation. Tensile test results showed that increasing Fe content from 0.3 to 0.7 pct led to a yield strength improvement of 10 pct while maintaining similar ductility. The results elucidate the effect of Fe-rich intermetallics on the HTS in aluminum alloys and offer a new insight in mitigating hot tearing, especially for recycled aluminum alloys.

Lightweighting in the automotive industry has driven the emergence of large aluminum castings for body structures, often referred to as "giga castings." The increasing use of aluminum giga castings in critical structures requires improved quality, with more reliable and quantifiable performance in both safety and durability. Aluminum casting processing is very complex and involves many competing mechanisms, multi-physics phenomena, and potentially large uncertainties. One of the most effective ways to optimize the design and manufacturing processes of aluminum giga castings to achieve the desirable mechanical properties is through the development and exploitation of robust and accurate multi-scale computational material models. This paper reports on an integrated computational materials engineering (ICME) approach for through-process modeling of local tensile properties of an aluminum giga casting using GM virtual cast component development (VCCD) tools.

Image-based machine learning (ML) methods are increasingly transforming the field of materials science, offering powerful tools for automatic analysis of microstructures and failure mechanisms. This paper provides an overview of the latest advancements in ML techniques applied to materials microstructure and failure analysis, with a particular focus on the automatic detection of porosity and oxide defects and microstructure features such as dendritic arms and eutectic phase in aluminum casting. By leveraging image-based data, such as metallographic and fractographic images, ML models can identify patterns that are difficult to detect through conventional methods. The integration of convolutional neural networks (CNNs) and advanced image processing algorithms not only accelerates the analysis process but also improves accuracy by reducing subjectivity in interpretation. Key studies and applications are further reviewed to highlight the benefits, challenges, and future directions of using ML in material failure analysis through image processing.

Automotive structure lightweighting is becoming increasingly important especially for electric vehicles to improve driving range. Aluminum shape castings offer effective lightweighting solutions because of their low density and high flexibility in product design and structural integration in comparison with steel parts. The aluminum casting processes are very complex and challenging, particularly for large integrated multi-functional castings. This requires not only good alloy selection and casting design but also optimal casting and heat treatment process control to achieve the desired material properties and product performance. One of the most effective ways to develop high integrity aluminum castings is through exploitation of virtual casting computational tools. This paper reports the advancement and implementation of multi-scale computational tools for virtual cast component development (VCCD) at GM using an integrated computational materials engineering (ICME) approach from an atom to auto. The VCCD tools have been extensively used for robust design of aluminum shape castings.

Ultra-large aluminum shape castings have been increasingly used in automotive vehicles, particularly in electric vehicles for light-weighting and vehicle manufacturing cost reduction. As most of them are structural components subject to both quasi-static, dynamic and cyclic loading, the quality and quantifiable performance of the ultra-large aluminum shape castings is critical to their success in both design and manufacturing. This paper briefly reviews some application examples of ultra-large aluminum castings in automotive industry and outlines their advantages and benefits. Factors affecting quality, microstructure and mechanical properties of ultra-large aluminum castings are evaluated and discussed as aluminum shape casting processing is very complex and often involves many competing mechanisms, multi-physics phenomena, and potentially large uncertainties that significantly influence the casting quality and performance. Metallurgical analysis and mechanical property assessment of an ultra-large aluminum shape casting are presented. Challenges are highlighted and suggestions are made for robust design and manufacturing of ultra-large aluminum castings.

This paper evaluates the influence of contact pressure on the surface contact state, fretting damage degradation mechanism, crack initiation and propagation characteristics, and fretting fatigue life of A319-T6 cast aluminum alloy in fretting fatigue test using bridge-type fretting pad. For a certain range of contact pressure, the fretting fatigue life exhibits a trend of increasing and subsequently decreasing with increasing contact pressure. With finite element analysis, the fretting contact region can be divided into two characteristic zones according to the dynamic evolution of the contact status within one fatigue cycle. The contact pressure affects both the distribution of the two characteristic zones as well as the relative slip range, which results in distinct fretting damage mechanisms. The nucleation characteristics and propagation paths of fatigue cracks under three typical contact pressures were analyzed considering the effects of fretting damage and heterogeneous microstructures of the alloy, accounting for the contact pressure dependency of the fretting fatigue life. The dynamic evolution of the contact status affected by contact pressure is considered.The fretting damage evolution with the variation of contact pressure is evaluated.The nucleation characteristics and propagation paths of fatigue cracks are elaborated.The contact pressure dependency of the fretting fatigue life is summarized.

Aluminum castings are increasingly being used in automotive powertrain and structural applications for vehicle lightweighting. Cast aluminum parts often need to be machined and joined to steel components in multi-material systems. Bimetallic components produced by casting aluminum over steel substrate are advantageous due to the elimination of machining and traditional dissimilar material joining processes. However, forming a strong metallurgical bonding during overcasting process has been challenging. In this paper, the effects of processing conditions such as substrate surface preparation, pre-heating temperature, melt conditions as well as casting design were studied by investigating the bimetallic interfaces for aluminum A319 casting alloy and a high alloy steel substrate in a sand-casting process. Zn and Al-based substrate coatings were evaluated. The experimental results suggest a metallurgical bond between aluminum and steel can be formed during the sand-casting process via high temperature diffusion with or without coatings. Al-based substrate coating prior to overcasting results in the formation of a continuous layer of intermetallics at the bimetallic interface, with large Al–Si–Fe intermetallics penetrating the cast aluminum.

The present study systematically evaluated the effects of a newly developed Al-1.25 wt