Science, Technology and Innovation

Long-term changes in macrozoobenthos structure and water quality of the upper Jasiołka River (Southern Poland) in 2014, 2018, and 2025

Mariusz Klich, Karolina Falasa, Anna Matras, Amelia Michalska, Natalia Suda — Fri, 17 Apr 2026 00:00:00 +0200

The article presents the results of long-term monitoring of water quality in the upper reaches of the Jasiołka River (Southern Poland) based on macrozoobenthos are presented. Field research was conducted in May 2025 at five sampling sites, and the data obtained were compared with historical results from 2014 and 2018, as well as with data from reference Carpathian rivers: Ropa and Wisłoka. Biological methods were used for ecological assessment: the BMWP-PL biotic index and the Margalef’s diversity index. A total of 38 macroinvertebrate families were recorded in the 2025 study. Water quality assessments obtained using both indices were consistent, classifying the tested section of the Jasiołka mostly as Class I (very good quality). The long-term analysis demonstrated the stability of high water quality during the study period. The water quality of the Jasiołka is similar to that of the Ropa, but clearly better than the Wisłoka, which indicates lower anthropogenic pressure in its catchment area. Biomonitoring results correlate positively with historical physical and chemical parameter studies, which means that over a multi-year scale, the Jasiołka is not subject to pressures, or they are low.

Effect of mineral composition and chemical activation of clay raw materials on the flexural strength of ceramic bodies

Sebastian Dydo, Wojciech Panna, Jarosław Cyboroń, Radosław Madej — Mon, 25 May 2026 00:00:00 +0200

The mechanical strength of ceramic bodies in the air-dried state is one of the key technological parameters determining the course of forming, drying, and transport processes of ceramic green bodies. This study evaluates the influence of mineral composition and chemical activation of clay raw materials on the flexural strength of ceramic bodies, with particular emphasis on kaolinite, illitic clays, and smectite-rich raw materials. The effect of quartz sand addition and sodium activation (Na₂CO₃) on the mechanical properties of the specimens was also investigated.

Samples were prepared from KOC kaolin, Krakowiec clay, and bentonitic clay from the Bełchatów deposit, in both non-modified form and after sodium activation in the range of 3–10% Na₂CO₃. In the case of bentonitic bodies, quartz sand was added in amounts of 10–30%. Flexural strength was determined using the three-point bending method in the air-dried state and after drying at 105°C. In addition, the moisture content of the samples after drying was analyzed.

The obtained results demonstrated a significant influence of both mineral composition and chemical modification on the modulus of rupture values. The highest air-dried flexural strength was achieved for smectite-rich raw materials, whereas kaolin-based bodies exhibited the lowest binding capacity. Sodium activation resulted in a substantial improvement in the strength of all investigated ceramic bodies, although the effect strongly depended on the type of raw material and its particle size distribution. For bentonitic bodies containing 30% quartz sand, the highest modulus of rupture values were obtained with 10% Na₂CO₃ addition. Samples dried at 105°C exhibited higher strength values, which was attributed to closer packing of clay mineral layers and enhanced electrostatic interactions.

Insight into the thermodynamics of hydrogen sulphide adsorption through multi-temperature fit of the related isotherms

Sylwester Furmaniak, Piotr A. Gauden — Thu, 26 Jun 2025 00:00:00 +0200

Adsorption of hydrogen sulphide is a process of practical importance, among others, in the context of biogas desulphurization. Its full understanding may require knowledge about the related thermodynamic aspects. In the current study, the possibility of indirectly obtaining such data based on mathematical modelling of the temperature dependence of H₂S adsorption isotherms is examined. For this purpose, a cooperative multimolecular sorption model was selected as the theoretical isotherm equation. It was used to simultaneously fit several sets of experimental data on a multi-temperature way. Such an approach, despite certain limitations, is an effective tool for modelling hydrogen sulphide adsorption on the studied activated carbons and zeolites. Furthermore, it allows for the straightforward calculation of thermodynamic quantities characterizing the process: the isosteric enthalpy of adsorption and the entropy of the adsorbed phase. In the study, curves reflecting the changes in these functions with adsorption progress were successfully determined and analyzed, supplementing the quantitative characterization of the process.

Automatic classification of selected bird species using LSTM network and attention mechanism based on digital vocalization recordings

Marcin Skobel, Robert Wielgat — Tue, 30 Jun 2026 00:00:00 +0200

Aim: This paper focuses on the development and evaluation of a deep learning model designed for the classification of audio signals from 31 bird species. The main objective of the study was to investigate the impact of the Luong attention mechanism on the network’s ability to generalize patterns in sequential data.

Material and methods: A hybrid neural network architecture combining Long Short-Term Memory (LSTM) layers with a global attention mechanism was employed. The data preprocessing pipeline included MFCC (Mel-Frequency Cepstral Coefficients) feature extraction and the alignment of signals to a uniform time interval. The model was subjected to a threefold training and testing procedure on randomly selected data subsets, enabling a reliable assessment of result stability. Classification performance was evaluated using the following metrics: accuracy, F1-score, and mean AUC.

Results and conclusion: The results demonstrated that replacing an additional LSTM layer with an attention mechanism significantly reduces model dimensionality while simultaneously improving classification accuracy and overall performance. The proposed approach achieved a mean accuracy of 0.9787, an F1-score of 0.9493, and a mean AUC value of 0.9991. The combination of LSTM layers with an attention mechanism constitutes an effective tool for the classification of acoustic signals produced by different bird species.

The impact of internal asymmetry of induction motors on the supply network

Janusz Petryna, Tomasz Kołacz, Grzegorz Aksamit, Dawid Kara — Tue, 30 Jun 2026 00:00:00 +0200

This paper presents the impact of internal asymmetry in induction motors, such as short circuits, close circuits, stator impedance differences and rotor damage, on the supply network in the form of negative and zero-sequence currents, illustrating the thesis with examples from research and measurements. Every motor defect leaves a trace in the network. The content of these components relative to the positive sequence current, similarly to the THD, determines the quality of electricity in a given network.