Materials Science Forum Vol. 1188

Abstract: This paper presents a development methodology for a metamodel-based machine learning approach for spatiotemporal prediction of temperature and solid fraction evolution of aluminum castings during cooling under low pressure sand casting (LPSC) conditions, for various pouring temperatures (T_p) ranging from 614°C to 720°C. High-fidelity finite element (FE) simulations were performed, based on a representative case study produced by the LPSC process to generate a comprehensive database, recording nodal temperatures across the casting symmetry plane at different cooling times, and for several T_p values within the studied domain. Three different machine learning (ML) algorithms were evaluated using comparative metrics (R², MAE and MSE). Among the evaluated algorithms, the artificial neural network (ANN)-based ML model was selected for its superior predictive accuracy and robustness. The accuracy of the selected ML-model was assessed by comparing predicted and FE-simulated temperature fields. The results indicate that the predicted temperature error within the cast symmetry plane remains below 1%. Furthermore, a graphical user interface (GUI) was developed to visualize the predicted casting temperature field for different T_p values not used during the learning stage, as well as the corresponding solid fraction, which is computed based on the solidification curve of the aluminum alloy AlSi₇Mg_0.3 given by the ProCAST^® database. This methodology could provide a fast, robust, and scalable framework for extending predictive models to higher-dimensional cases and diverse LPSC casting process configurations.

Abstract: The decarbonization of the aluminium industry requires a transition from fossil fuels to sustainable energy carriers. This study investigates the substitution of natural gas (NG) with hydrogen (H₂) in reverberatory furnaces, analyzing the impact on melt quality, furnace integrity and exhaust emissions. Experimental investigations were conducted in a specifically designed furnace setup combining electrical heating with a burner system capable of operating with variable fuel blends ranging from pure natural gas to 100 vol.-% hydrogen. The results demonstrate that the hydrogen content in the aluminium melt depends on the atmospheric conditions — water vapour content in the atmosphere — during the melting and heating phases. In contrast, the holding phase exhibited a quasi-static behavior with negligible further hydrogen uptake, due to the isothermal process control. Numerical simulations (CFD) revealed that admixture rate exceeding 80 vol.-% H₂ leads to significantly higher adiabatic flame temperatures. This results in the formation of local hotspots on the furnace walls and requiring the use of high-performance refractory linings. Furthermore, these thermal conditions correlated with a major increase in NO_x emissions, despite a successful reduction in CO₂ output. Considering the material quality, X-ray computed tomography (XCT) analysis indicated a marginal increase in volume porosity with higher hydrogen fractions. However, tensile testing confirmed that this porosity did not compromise the mechanical performance, as yield strength and ultimate tensile strength remained unaffected across all fuel mixtures. The study concludes that standard degassing procedures are sufficient to reduce the increased initial hydrogen load, showing that hydrogen combustion for secondary aluminium production is feasible.

Abstract: High-strength and recycling tolerable aluminum alloys make a significant contribution to weight reduction in modern lightweight construction. The advantages of aluminum alloys in terms of their low density combined with high strength can be significantly improved by the alloy composition. In contrast to the conventionally established process route, high-magnesium alloys can be produced using the twin-roll strip casting process. This allows additional process steps such as hot rolling and annealing to be drastically reduced in the economical production of near-net-shape strips, saving emissions and energy consumption. The strip casting process has already been applied to numerous aluminum alloys and enables their production, although the understanding of advanced alloys in this area is not yet fully understood because of its limited production in industry-related research due to the complexity of the process. However, transferring the high strength generated during rapid solidification into usable sheet performance remains challenging, especially at elevated Mg contents, where segregation, casting-related defects, and solute-affected recrystallization can limit ductility and processability. This study investigates the potential of a high-magnesium aluminum alloy produced by vertical strip casting. The properties of the alloy are correlated with the microstructural and mechanical characteristics and developed on the basis of an industrial reference alloy. For this purpose, an EN AW 5182 and an AlMg10 alloy were processed. The results show that high-magnesium alloys can be produced and processed using strip casting. In terms of the high-magnesium alloy, improved results can be achieved compared to the industrial EN AW 5182 alloy. Key findings: The strength of high-magnesium alloy is significantly above those of the EN-AW 5182 after strip casting enabling nearly 600 N/mm² tensile strength, but the final properties are below this potentially possible characteristic after strip casting, presumably due to non-ideal recrystallization and an insufficiently adapted process route including rolling and annealing parameters.

Abstract: High-Pressure Die Casting (HPDC) processes are often affected by complex thermo-fluid-dynamic phenomena that lead to casting defects and premature die degradation. In this study, an approach based on the Response Surface Methodology (RSM) is proposed to improve the quality of the cast part (aluminum window brackets) and extend the dies’ service life by introducing limited modification to the geometry of the die cavities.A multi-physics numerical model was initially built up to reproduce the filling and thermal behavior of the process. Infrared thermography, used to validate the numerical results, confirmed the accuracy of the model, with an average temperature error of approximately 2%. The analysis revealed that the baseline configuration (i.e. the dies’ geometry currently adopted in the industrial process) was characterized by non-negligible thermal imbalances (temperature gradients of about 50 °C and localized hot spots associated with high melt velocities), which reflected in the occurrence of flashes, metallization, and impression pad damage.New die geometries with the aim of improving the thermal uniformity while reducing the temperature gradients where investigated by varying the geometrical properties of the gating system according to a DoE-based approach. The numerical results, collected in terms of total amount of porosity in the casting critical areas, were used to train accurate metamodels that, in turns, were adopted as the starting base for a multi-objective optimization. Results from the optimization allowed to identify different scenario, each characterized by a specific geometry of the gating system able to remarkably reduce the occurrence of porosity in the cast part (up to 42% less than the current condition). The results demonstrate that the proposed methodology enables effective and sustainable optimization of HPDC processes without costly trial-and-error approaches.

Abstract: The effect of time-dependent interfacial heat transfer coefficient () within the shot sleeve and die cavity of high pressure die casting process (HPDC) has been simulated to systematically study the solidification occurring during filling. Two different-profiles have been considered with peak values of 7 kW/m²K and 12 kW/m²K for the shot sleeve, and 18 kW/m²K and 26 kW/m²K for the runner, gate, and die cavity based on the values reported in the literature. In addition, two types of gate designs were considered for plate type castings to analyze their solidification behaviour and filling velocity. Solidification typically occurs along the bottom wall of the shot sleeve, from the mid-region toward the mould-side region along the direction of pouring. At the end of filling, the solid fraction () inside the shot sleeve increases from 10 to 18% with increasing peak value for-profiles. Similarly, the solidification around the gate regions progresses rapidly above 0.4 and reduces the fluid velocity at the gate entry for profile with higher peak values. Despite the lack of consensus on the selection of value (peak value and range), this study highlights the influence profiles and gating design on solidification during filling and discusses its implications on the quality of HPDC parts.

Abstract: This article aims to review the recent advances in the laser powder bed fusion process (L-PBF) of H13 tool steel for the dies used in the high pressure die casting (HPDC) applications. The effect of processing variables is briefly reviewed for the evolution of microstructure (phase transformations, as-built microstructure and carbides precipitation), mechanical properties, and defects. The second part of the review is focused on conformal cooling applications to HPDC dies, which is critical for die life and productivity. Achieving better microstructure without defects, understanding the role of processing variables in L-PBF and their interdependencies remains the key challenge for the as-built part, while the benefits of preheating and post-heat treatments are evident. Significant benefits are realized in the applications of die inserts favoring lower die surface temperature, reduced cycle time and lubrication, and thermo-mechanical stresses. In addition, L-PBF also plays a key role in die remanufacturing where significant benefits are achieved in terms of materials savings and improved performance compared to traditional repair technologies. Overall, L-PBF offers a transformative pathway for high-performance HPDC dies; however, most investigations are trial-based. Long-term studies are needed for performance assessment and establishing failure mechanisms in production environments.

Abstract: Remanufacturing by casting of end-of-life (EoL) components can be challenging as it requires specific molds/cores, machinery and tooling. The use of 3D scanning technology and rapid casting processes can produce high-quality, efficient components. However, error propagation during the manufacturing process can significantly affect the dimensional accuracy of the final product. In this study, the dimensions of a case study part were evaluated at the main stages of the rapid hybrid low-pressure sand casting (LPSC) process, from the initial CAD model to the final casting, to identify the main causes of final 3D surface deviation. The casting design was optimized using coupled thermal and fluid-flow FE computations for two suitable casting orientations: horizontal (H) and vertical (V). After the 3D sand-mold printing process, an optical 3D scanner was used to extract surface data from each printed mold part. 3D surface deviations caused by the printing process were evaluated by comparing the individual mold components to their original CAD models using GOM Inspect Pro^® software. The final castings were also compared to the initial CAD models for both orientations to quantify the overall 3D surface deviations resulting from the rapid LPSC process chain, including 3D printing, liquid metal shrinkage and contraction during solidification and cooling. The results provide a foundation for improving dimensional accuracy of one-off replacement components produced by the hybrid LPSC process.

Abstract: Dedicated simulation software is used to create a dataset which for a particular sand-cast part and various casting parameter values provide casting quality metrics based on temperature and solidification evolution. Based on simulation data, ANNs are trained to predict the successful or failed filling of the mold (a classification problem), as well as the quality of the part through solidification time, maximum microporosity, maximum von Mises residual stress, maximum displacement of any point in the casting and total volumetric shrinkage (a regression problem). Such ANNs can provide augmented information much faster than the simulation model to the process planner. A third category of ANNs (of the regression type, too) determine the temperature evolution with time at an inaccessible point, where no thermocouple can be placed, from the measurement history at two other thermocouples close to it. This data comes from real-time monitoring during casting. Such ANNs can aid the process supervisor in a ‘digital shadow’ context. The issues associated with generalizing these predictors to become independent of specific part geometry are also discussed.

Abstract: Cleaner melt transfer is critical to the broader use of recycled aluminium alloys in high-end structural casting applications, where oxide bifilms and intermetallic inclusions, such as Fe-containing intermetallics, can significantly affect the casting's mechanical properties. In counter-gravity low- and high-pressure casting, the launder system must not only promote the sedimentation of inclusions but also deliver a stable, cleaner melt to the crucible. Prior research showed that 15° double baffles in the mid-section of the sedimentation launder at a flow rate of 100 kg·h^-1 provide high efficiency. The present work investigates the influence of baffle design at the launder-crucible interface, where the melt enters the crucible before casting. Fluid dynamic simulations were carried out at a 100 kg·h^-1 flow rate for three inlet configurations: (i) full baffle; (ii) lifted baffle; and (iii) split baffle. Inclusions of various densities and diameters were tracked. Results indicate that the full baffle, while beneficial as a benchmark and efficient, is impractical because it generates fresh oxide surfaces. The lifted baffle provided the most effective reduction in inclusions, like the full baffle setup, enhancing sedimentation and suppressing entrainment, while the split baffle showed intermediate behaviour. Moreover, the lifted configuration promoted centrifugal flow (at lower velocities, it still made a partial contribution) within the crucible, directing inclusions towards the crucible wall and the stagnation-velocity zone, and enabling the crucible itself to act as a final sedimentation stage before the counter-gravity pump extracts the melt. These results demonstrate that combining mid-launder optimisation with crucible inlet baffle design enables cleaner, more automated melt delivery, thereby strengthening the use of recycled aluminium alloys in structural casting applications.