Search

The article describes the computational complex designed for large-scale mathematical modeling and optimization of intra-quarter thermal regimes heating networks. Developed software allows to manage all users of thermal energy and carry out its accounting. On its basis, a strictly analytical system of targeted budget subsidies to individual consumers of thermal energy can be built, taking into account the social norms of housing, the estimated and actual power of heating devices, the location of the apartment in the house and the state of its enclosing structures. The article concludes about the effectiveness of the proposed mathematical modeling method in the organization and effective management of ensuring energy saving in the design of heating networks.

The study presents an analysis and the possibility of applying machine learning methods to automate the technological process of drinking water deodorization. One of the most actual and popular machine learning methods today is the use of artificial neural networks. Possible neural network structures, activation functions and the number of neural connections were described. The relevance of the introduction of these methods into the system of water management is described.

The designs of modern vehicles and aviation can be traced to a constant increase in the range of parts that have complexly shaped free-form surfaces. Production of such quality parts requires new methods of automated mechanical engineering and particularly program operated multi-purpose five-coordinate machine processing. The paper is dedicated to dynamical modelling of surface treatment of materials by spherical cylindrical milling cutter. The simulation results are the basis for the formation of optimal conditions for machining free-form surfaces and determining combinations of tilt and advance angles and their optimization for positioning the tool for end milling of surfaces on five-axis CNC machines.

This paper presents carburizing the surface of 20 low carbon steel using electric arc and graphite prior. A carbon black solution was prepared with graphite powder and sodium silicate in water. A detailed analysis of the phase structure and the distribution profile of the sample hardness after plasma treatment were given. The hardened layer consists of three different zones: 1 – the cemented layer (thin white zone) on the surface, 2 – heat-affected zone (darkly etching structure), 3 – the base metal. The experimental result shows that the various microstructures and micro-hardness profiles were produced depending on the type of graphite coating (percentage of liquid glass) and processing parameters. The experiment proved that the optimum content of liquid glass in graphite coating is 50–87.5%. If the amount of liquid glass is less than 50%, adhesion to metal is insufficient. If liquid glass content is more than 87.5%, carburization of a metal surface does not occur. A mixture of the eutectic lamellar structure, martensite and austenite was obtained by using graphite prior with 67% sodium silicate and the levels of the hardness layer increased to around 1000 HV. The thickness of the cemented layer formed on the surface was around 200 μm. It is hoped that this plasma surface carburizing treatment could improve the tribological resistance properties.