Light metal foundries consume large amounts of energy and rely mainly on gas-fired melting, which is inefficient and promotes porosity caused by dissolved hydrogen. This pilot study examines local induction melting combined with ultrasonic melt treatment (UST) as a decentralized and electrified alternative. AlSi12 was cast with varied mold temperature, casting temperature, UST duration and remelted fraction. Porosity was characterized by optical microscopy in 2D and, for a subset of specimens, by X-ray computed tomography in 3D, and the mechanical behavior by tensile testing. An experimentally benchmarked simulation was used to estimate the energy demand for larger batch sizes. Including the energy required for cooling, the measured laboratory efficiency was 7.9% to 14.9%, and the simulation predicts 41.3% for 100 kg, or 48% with a preheated crucible, against 53.1% derived for industrial induction furnaces. Specimens treated with UST showed a lower mean porosity, 0.56% against 2.14% in optical microscopy and 0.153% against 0.367% in computed tomography, where the pore count was 73% lower at almost unchanged pore size. Their mean strain at fracture was 45.5% higher and their mean toughness 53.2% higher, whereas the 4.4% higher mean tensile strength lies within the scatter of the measurements.
Accurate wind flow prediction is essential for various applications, including the placement of wind turbines and a multitude of environmental assessments. Traditionally this can be achieved by using time-consuming computational fluid dynamics (CFD) simulations on reanalysis data. This study explores the performance of an autoencoder (AE) and a variational autoencoder (VAE) in approximating downscaled wind speed and direction using real-world reanalysis data and reference geo- and vegetation data. The AE model was trained for 2000 epochs and demonstrates the ability to replicate wind patterns with a mean absolute error (MAE) of approximately −0.9. However, the AE model exhibited a consistent underestimation of wind speeds and a directional shift of approximately 10 degrees compared to CFD reference simulations. The VAE model produced visually improved results, capturing complex wind flow structures more accurately than the AE model. It mainly achieves better local accuracy and a reduced variance of the results. The overall result suggests that while autoencoders can approximate wind flow patterns, challenges remain in capturing the full variability of wind speeds and directions with sufficient precision. The study highlights the importance of balancing reconstruction accuracy and latent space regularization in VAE models. Future work should focus on optimizing model architecture and training strategies to enhance accuracy, prediction reliability and generalizability across diverse wind conditions and various locations.
Reflection silencers are installed in the exhaust system of stationary combustion engines to attenuate low-frequency noise by means of destructive interference. The acoustic properties of mufflers are experimentally determined by the standard two-load method, which only considers measurements without mean flow. In real engine operation, however, exhaust mass flow is always present. Measurements are significantly more complex and expensive if fluid flow is taken into account, which is why the available data is limited. Thus, the impact of mean flow on the attenuation of silencers is not clearly known yet. This work contributes to the state of the art by quantifying the influence of the Mach number on the transmission loss of double-tuned straight-through mufflers based on reproducible, noise corrected measurement results that include uncertainties. A frequency range between 20 Hz and 891 Hz is investigated at eleven different Mach numbers between 0 and 0.1 under ambient conditions. It is found that resonance peaks diminish with increasing Mach number, while other frequencies remain unaffected by mean flow. These findings can be transferred to operating conditions of stationary combustion engines and other exhaust systems. The experimental data will serve as a basis for the validation of analytical and numerical models in subsequent work.