No abstract

In this article, the authors propose using a neural network to optimize the aerodynamic drag of a body in a gas flow. The use of such optimization methods using evolutionary algorithms makes it possible to obtain a mathematical model that describes the shape of the aerodynamic profile, which allows the operator to easily change parameters. It is shown that the value of the drag force for a body of optimized shape is lower than the values for bodies of rotation and known profiles taken for comparison. As bodies, we can consider parts of engineering structures in the air flow, parts of vehicles, and bodies located in the air flow in the duct, in particular in the duct of the photo separator. The latter is important from the point of view of neural network photo separation of seeds and grain.

No abstract

No abstract

The paper presents the results of the study of possibility using a body with an optimized shape with a minimum aerodynamic drag force as a heat sink in a convective gas flow for cooling machine parts. The relevance of cooling or heat sink systems for mechanical engineering is not in doubt. Maintaining the optimum thermal state of machine parts within the specified limits is necessary to prevent heating, which can cause violations of normal operating conditions, and, as a result, increased wear, jamming and breakage of parts. To ensure the normal operation of the system, it is necessary to cool the parts contacting the hot gases, removing heat from them to the atmosphere directly, or using an intermediate body (water, low-freezing liquid). Heat sink is an aerodynamic process, a process of heat mass transfer of a substance, and aerodynamic characteristics and, in particular, the nature of streamlining of bodies of the simplest forms are of great scientific and practical interest. Bodies of complex shape can always be represented as a combination of simpler ones, for which it is easy to explore and calculate the flow paths, on the basis of which analytical methods for calculating aerodynamic characteristics are compiled. A computational experiment was conducted in the software product Ansys Fluent. The conditions of the experiment (comparison of heat removal bodies) during numerical modeling are the preservation of constants: the volume and shape of the working zone; distances from the sources, drains and centers of bodies; aerodynamic flow rates; masses of compared bodies; thermal power of the source and other parameters in addition to the shape of the surface. The resulting optimized shape of the body coincides with the current lines, which is the main advantage, since no separation from the surface is observed when streamlining the flow. Thus, the entire surface area will be the effective surface area of the heatsink as opposed to other known shapes of bodies, whereby the temperature of the heat loaded element placed in the center of the heatsink will decrease.