machinery
Not a member yet
8395 research outputs found
Sort by
NEURAL NETWORK MODEL FOR PREDICTING THE EFFICIENCY OF A STEAM BOILER USING NATURAL GAS AS FUEL
No full textPredicting the performance of steam boilers is important to enable efficient operational monitoring and appropriate control strategies. The operation of such plants can be both steady-state and transient, with operating parameters changing according to the requirements of the technological process. The establishment of accurate mathematical models to predict boiler efficiency encounters a variety of problems related to the degradation of certain operating parameters and changes in steam production. Due to the inherent flexibility, adaptability, and robustness of neural network models, developers can take advantage of these benefits to create more accurate predictive models for boiler efficiency over time, that achieve comparable or even higher accuracy than conventional techniques. The aim of the neural network model presented in this paper is to predict the directly calculated boiler efficiency based on experimental data from existing steam boiler running on natural gas as fuel.
To overcome this challenge, two well-known machine learning techniques are compared: Feedforward 3-Layered Neural Network (2-5-2) with Levenberg-Marquardt algorithm and Adaptive Neuro-Fuzzy Inference Systems (ANFIS) with hybrid algorithm. The input datasets are the temperature of the generated steam and the steam production (mass flow rate). The data set is divided as follows: 70% of the data is used for training, 15% for validation and the last 15% is used for testing. The training set is used to train a machine learning model by exposing it to examples with known outcomes. The validation set is used to fine-tune the model parameters and to avoid overfitting, while the test set evaluates the model's performance on unseen data.
Error histograms and high correlation coefficients — which have been shown to be greater than 0.9 even for the test data sets — were used to evaluate the effectiveness of the presented machine-learning techniques. In addition, thorough analyzes and justifications of the results obtained are provided, offering important new perspectives for the practical implementation of these methods. In order to predict operating costs, one must be able to determine the actual efficiency of steam boilers with a reasonable degree of accuracy
A procedural and technical experimental review on material tensile and impact properties under cryogenic temperatures
No full textIn this paper, research and comparison of tensile and impact properties were conducted on several materials broadly categorized as Steel and Alloys at cryogenic temperature of 77 K or -196°C. Tensile and impact properties exhibit an inverse relationship due to the nature of metals, where if a material has high strength, its ductility will decrease, and if it has high ductility, its strength will decrease. Generally, cryogenic treatment will result in periodic strength enhancement of materials, but a significant reduction in ductility occurs when the temperature surpasses the ductile to brittle transition temperature (DBTT). However, some materials can be processed to achieve desired property advantages, such as high toughness, high ductility, or a balanced combination of ductility and toughness without significant reduction in either property
Structural integrity of tapered cylindrical shell: Study case of tower wind turbine
No full textThe present study investigates the structural integrity of a wind turbine tower structure under axial compression, described as a tapered tubular structure. Initially, the NREL model of the 5 MW-net wind turbine model was adapted and then scaled down to simplify the numerical analysis and for the convenience of future experimental study. The analysis was conducted using the Finite Element Modelling software Abaqus. To ensure the validity of the FEM modelling, the benchmarking study is conducted by referring to previously published work. The case configuration was developed by varying the material properties of the tower (high, medium, and low carbon steels) and the material properties of the tower due to the effect of the site temperature. The results obtained show that high carbon steel has the best properties for use in wind turbine structures. At -80 °C, this is the temperature condition where AH32 material has the best properties
Stances on the Xylella fastidiosa bacterium and the coronavirus in the public discourse
No full textThe paper deals with the stances on the Xylella fastidiosa (Xf) bacterium and the
coronavirus in the public discourse. Relying on Hyland’s stance model (2005), we
seek to identify the linguistic devices (hedges, boosters, attitude markers and selfmentions), used in both contexts, which helped the authors express their opinions.
The corpus comprises articles on Xf and the coronavirus collected from the internet
sources in English. The findings indicate that the most frequent stance markers
in both subcorpora include hedges and boosters. There are slight differences in
the marker frequencies, depending on the disease in question. The conclusion
discusses the role of stance markers in impacting attitudes while reporting on these
two important issues and their use in communicating tentativeness or certaint
On the possibilities of obtaining analytical solutions for gas flow of different levels of rarefaction
No full textImproving the surface quality and tribological characteristics of 3D-printed titanium parts through reactive electro-spark deposition
No full textfirst_pagesettingsOrder Article Reprints
Open AccessArticle
Improving the Surface Quality and Tribological Characteristics of 3D-Printed Titanium Parts through Reactive Electro-Spark Deposition
by Georgi Kostadinov 1ORCID,Todor Penyashki 1,*,Antonio Nikolov 2 andAleksandar Vencl 3ORCID
1
Institute of Soil Science Agrotechnologies and Plant Protection “N. Pushkarov”, Agricultural Academy, Shose Bankya Str. 7, 1331 Sofia, Bulgaria
2
Faculty of Industrial Technology, Technical University of Sofia, Kliment Ochridsky 8, 1000 Sofia, Bulgaria
3
University of Belgrade, Faculty of Mechanical Engineering, Kraljice Marije 16, 11120 Belgrade, Serbia
*
Author to whom correspondence should be addressed.
Materials 2024, 17(2), 382; https://doi.org/10.3390/ma17020382
Submission received: 8 December 2023 / Revised: 28 December 2023 / Accepted: 8 January 2024 / Published: 12 January 2024
(This article belongs to the Special Issue Advances in Metal Coatings for Wear and Corrosion Applications)
Downloadkeyboard_arrow_down Browse Figures Versions Notes
Abstract
This work presents the results of research conducted with an aim to improve the surface quality, hardness and wear resistance of titanium alloy Ti6Al4V, obtained via the laser powder bed fusion of metals (PBF-LB/M) process of additive manufacturing (AM) known as the 3D printing of metals. The 3D surfaces were coated via reactive electrospark deposition (RESD) with low-pulse energy and electrode materials of low-melting metals and multi-component hard alloys. The relationship between the electrical parameters of the RESD process and the quality, composition, structure, microhardness and wear resistance of the treated surfaces were investigated and analysed. It was found that the roughness and thickness of the resulting surface layers could be changed by changing the RESD modes within the limits of 2.5–5 µm and 8–20 µm, respectively. RESD processing allowed us to achieve two to five times lower roughness than that of titanium AM surfaces. The microhardness and wear resistance of the RESD surfaces are two to four times higher than those of the titanium substrate. Possibilities for the purposeful synthesis of new wear-resistant phases and compounds and for obtaining surface layers with predetermined thickness and roughness were established. It was shown that the subsequent reaction’s electrospark processing helped to simultaneously reduce the roughness and increase the hardness and wear resistance of the modified surfaces, and can be successfully used instead of the material-energy-labour and machine-intensive finishing treatments of the titanium surfaces obtained after 3D printing
CFD-Driven Ship Trim Optimization: Simplifying Complexity of ANN with User-Friendly Software
No full textThis paper presents a ship trim optimization strategy, employing user-friendly software tool integrating Computational Fluid Dynamics (CFD) and Artificial Neural Networks (ANNs). It evaluates parameters like brake power, shaft speed, and fuel consumption. The tool streamlines complex models into an accessible interface, providing optimized recommendations swiftly. CFD-informed ANN models ensure precise predictions. This advancement minimizes fuel usage, cuts operational costs, and promotes maritime sustainability. By bridging advanced computation and practicality, the tool enhances operational efficiency. The interdisciplinary integration of CFD, ANN, and user interface software underscores technology's role in advancing greener, cost-effective ship trim optimization for the maritime industry's future
Wet cooling of air on plate finned tube heat exchangers
No full textThe objective of this paper is to provide the reliable calculation procedure for determining heat and mass flow rates for plate finned tube heat exchangers in dehumidification regimes that can be used easily in engineering practice. For this purpose, the experiments are conducted on two heat exchangers, and datasets for six more heat exchangers are used from literature. Comprehensive database is established with total of 637 sets of data, and it gathers 365 new measurement sets and 272 sets from open literature. The new calculation procedure predicts heat transfer rate and condensate flow rate where new methodology approach (based on the porous velocity, the ratio of characteristic surfaces and hydraulic diameter) is used. Predicted results show 5–10 % deviation for heat duty and up to 20 % for mass flow rate from experimental results, which is of importance to industrial practice
KIBERNETSKO-FIZIČKI METROLOŠKI MODEL ZA INDUSTRIJU 4.0: DIGITALNI BLIZANAC I INTEROPERABILNOST
No full textU svrhu kreiranja kibernetsko-fizičkog modela (KFM) za proces planiranja inspekcije (PPI) na numerički upravljanoj mernoj mašini (NUMM), konfigurisan je merni sistem sa mernim predmetom, NUMM-om i steznim priborom. KFM sadrži dva skupa komponenti: virtuelne i fizičke. Virtuelna komponenta je virtuelni merni sistem kreiran u CAD (Computer Aided Design) okruženju. Dalje, u poglavlju se prikazuje modeliranje i simulacija virtuelnog mernog sistema baziranog na stvarnoj NUMM-i korišćenjem PTC Creo softvera. Simulacija uključuje programiranje merne putanje i generisanje DMIS (Dimensional Measuring Interface Standard) - (*.ncl) datoteke, za standardne tipove tolerancija (pravost, ravnost, itd.). Takođe, modelovane su komponente NUMM-e i definisan je celokupan sklop mašine sa kinematičkim vezama. Simulacija i generisanje izlazne DMIS datoteke u PTC Creo-u se vrši korišćenjem virtuelne NUMM-e. Koristeći generisanu DMIS datoteku, stvarna NUMM-a je programirana i korišćena je za realno merenje. Nakon toga, generisan je izveštaj o merenju. Glavni rezultat ovog poglavlja jeste modeliranje kibernetsko-fizičkog sistema baziranog na digitalnom mernom blizancu (DMB) i DMIS datoteci kao protokolu komunikacije. DMB predstavlja integraciju virtuelnog i stvarnog okruženja i stvara se u svrhu optimizacije, simulacije, komunikacije, praćenja i verifikacije procesa merenja. Ovo istraživanje prezentuje novi pristup programiranju NUMM-e u tzv. off-line režimu. S druge strane, u svrhu interoperabilnosti i komunikacije sistema, porede se i analiziraju DMIS kodovi koji potiču iz različitih softverskih aplikacija