Search

Acoustic devices for determining the elasticity modulus based on the measurement of the samples frequency resonant oscillation due to the sample exposure to acoustic waves with consistently changed frequencies. Objective: Development of an algorithm for increasing the hardness of materials due to magnetic resonance imaging. Materials and methods: The paper shows the possibility of using as a uniform flux to influence the volume of the material of the magnetic field formed by powerful permanent magnets. The process of influencing the volume of material of the experimental samples was that the effect of a uniform magnetic flux permeating the sample is initiated in a result of resonant oscillations of the sample caused by broadband exposure of equal amplitude using a “white noise” generator and a piezoelectric emitter. Results: Treatment of samples of materials placed in a uniform magnetic field, resonant polyfrequency vibrations with nanoscale amplitude in the range of 20…80 nm, allows you to change the viscosity of the material, the modulus of elasticity of the material and the hardness of material samples to improve the performance of these materials . Conclusions: Nanoscale amplitudes of natural oscillations of objects of complex shape in energy fields, which include uniform magnetic fields, can correct the physical and mechanical properties of materials of such objects in order to achieve their identity or add strictly defined properties. [ABSTRACT FROM AUTHOR]

The article considers methods of non-destructive testing based on various physical laws and phenomena. The possibility of creating a new topical tool for obtaining a wide range of data of mechanical engineering products such as shape, size and location in space is considered. It is proposed to use sound diagnostics using a high-frequency broadband signal to capture the frequency characteristics of the object. The purpose of the study is to develop a method of non-contact measurement of mechanical engineering products on several grounds. With the help of vibroacoustic diagnostics and the method of quantitative control, the distribution of the entire volume of products was 100 pieces. on two parties: the main and control, quantitative parameters of each unit of a product are removed. A signal from 0 to 20,000 Hz was applied by means of a frequency generator. The frequency response of each sample was recorded in the Spectrum Analiyser program. Estimation of the deviation of the product size and its frequency spectrum was performed in the NeuroPro 0.25 software. The created neural network allows is predicted in real time values of several quantitative signs irrespective of their nature. A working model for collecting statistical data for the efficient operation of the neural network is obtained. The developed technique allows detecting the configuration of products on the basis of indirect measurements through the frequency spectrum. This technique can be used to diagnose parts by geometric features, physical properties, defects. This requires an increase in input data for neural network training. With a sufficient selection of parts with different defects of the neural network on the acoustic frequency characteristics will be able to divide the parts into groups of worthy and unworthy on various grounds. [ABSTRACT FROM AUTHOR]

The paper deals with modern methods of acoustic diagnostics of parts. Using of acoustic waves in the diagnosis of threaded connections makes it possible to inspect assemblies, bolted connections during assembly, as well as to almost instantly to identify defective products, incorrect dimensions and other factors of rejection based on the acoustic frequency response. With a large variety of threaded parts, a significant variation in pitch and diametrical dimensions, the use of acoustic waves represents a new innovative approach to non-destructive testing of units in the assembly of mechanical engineering products. Purpose: Development of an algorithm for diagnostics of threaded connections when assembling units using acoustic waves. To control the details of the node, an acoustic method was used to take the frequency response with the construction of a neural network. To create a neural network, all dimensions of the parts were measured, acoustic diagnostics were carried out to remove frequency characteristics, and the obtained data were entered into the NeuroPro program. To obtain the frequency characteristics of the nodes, a device with two sensors was used. One of them worked for reception, the second sent a constant signal at the frequency of white noise. Each bolt assembly was modified by screwing the bolts to the nut 10º. The measurement was carried out with screwing the bolts in the positions: 0º, 10º, 20º. Using a neural network, data on the degree of tightening of the bolts of the node were obtained and divided into categories. The most significant of them have been identified. The experiment showed that it is enough to have 2 frequency filters in order to determine the dimensions with an accuracy of 0.03 mm. Building a neural network of parts and a prefabricated unit will allow to almost instantly identifying defective products, incorrect dimensions and other factors of rejection based on the acoustic frequency response. This technique can be used to control units, bolted connections during assembly. [ABSTRACT FROM AUTHOR]