The article dwells upon the method for building a diagnostic model of a complex diagnostic object on the example of an electric train’s pneumatic system. The diagnostic model is built by formalizing the verbal description of the object using fuzzy logic.
Schelkanov A.V. Building diagnostic models of complex technical systems (on the example of pneumatic system of electric trains) // Dinamika Sistem, mechanismov i machin. - 2016. - №1, v.1. - p.127-132
One of the most promising trends in developing systems for vibration control, monitoring and automatic diagnostics of power units’ condition is shaft sensors application. However, turbine rotor vibration reliability criteria required for such purpose are not developed in full measure. The list of standards includes only specifications for vibration control systems and GOSTs for estimation of vibration rates at one side of a bearing journal.
The purpose of the research is to establish criteria for estimation of rotor condition and vibration reliability in operation, using full shaft sensor system as required by GOST R 55263-2012.
In this paper we formulated all main criteria necessary for condition assessment in modern systems of automatic diagnostics. We suggested threshold values for parameters used for determining the condition of shafting with bearing journals.
Kumenko A.I., Kuzminyh N.Y.,Timin A.V. Development of turbine rotor vibration reliability criteria // Dinamika Sistem, mechanismov i machin. - 2016. - №1, v.1. - p.58-64
Diagnostics of certain defects and failures of machines and mechanisms by vibration-acoustic signal parameters requires assessment of the measured signals’ peak values. Taking into account stochastic nature of such signals, the assessment of measuring errors and signals’ peak values considering distribution laws of their momentary values is an urgent task. This paper is aimed at developing a method for measuring errors assessment and calculation of vibration-acoustic signals’ peak values. As the results of analyzing a mathematical system for relative RMS random error which may occur in estimating statistical characteristics of processes there are mathematical relationships used for calculating such estimations for various distribution laws of momentary signal values. In particular, a task on determining a functional connection between asymptotic estimation of quantile sampling estimate and discreet values in a sampling (a signal’s sampling rate) is solved, allowing to justify and verify precision of measuring tools, as well as to evaluate accuracy of measurements of VA signals’ peak values.
Kostyukov V.N., Boychenko S.N., Kudryavtseva I.S., Naumenko A.P. Methodical basis of peak value assessment for vibration-acoustic signal parameters // Dinamika Sistem, mechanismov i machin. - 2016. - №1, v.1. - p.51-58
Some diagnostics problems of complex rotor systems for power turbine units are considered. A list of the main problems that must be solved for the diagnostics of technical condition of shafting on journal bearings is offered. One of the most difficult task – diagnostics of a residual deflection of separate rotors in the shaft line is also considered.
The multiple-bearing test rotor stand with journal bearings for the analysis of physical processes and development of a number of diagnostic signs of the defects measured automatically with use of the shaft encoder complete system according to GOST P 55263-2012 is offered.
The stand consisting of two rotors and a driving engine connected by rigid couplings allows to simulate and vary all main defects of production and assembly of rotors and bearings. The key geometrical and technical parameters of rotors are specified. The main dynamic characteristics of the rotor stand and results of simulation of a rotor deflection with use of the special software are provided.
Kumenko A.I., Kostyukov V.N., Suminov I.A., Kuzminykh N.I. Mathematical and physical simulation of diagnostics problem on the test rotor stand // Proceedings of CM2016/MFPT2016
It is proposed a modern concept and a set of methods for monitoring of the technical condition of system “turboset's shafting – journal bearings – support units”. Technology is developed as an integral part of domestic multipurpose COMPACS® system®. The system input data covers: - design, technological, mounting and operating parameters of the turbine unit; - absolute vibration of supports and control points of the shafting line; - relative displacement and vibration in the rotor journals according to GOST R 55263-2012. Also, the system includes as precomputed: - mounting line of the shafting in cold state; - for each journal bearing, static and dynamic properties of the bearing oil film in domain of possible motion (DPM are given; - static and dynamic properties of supports; - matrix of influence of displacements of supports on the forces in supports of the shafting.
While in operation, the methods allow to determine static and dynamic loads on the supports of turboset, to determine misalignment of supports and to evaluate corrective alignments of the supports and centering of coupling halves. Also, stress in the rotor pins, welded joints and coupling bolts are computed.. We suggest an automatic conclusion on compliance of the technical condition of turboset with safety levels, identification of source of malfunction and recommendations for necessary measures. The work covers some procedures, including ways of determination of stiffness of supports, a method of determination of above mentioned matrix of influence and a method of inverse problem solution – determination of misalignments of the supports based on determination of position of rotor journals in the bearing bores.
We have shown the results of static characteristic modeling by an example of turbine unit 1000 MW for atomic power stations.
Kostyukov V.N., Kumenko A.I., Kuzminykh N.Y., Timin A.V. Development of methodological support for turbine unit shaft defect and stress monitoring systems at thermal power stations and atomic power stations with application of shaft displacement sensors // Proceedings of CM2016/MFPT2016
The article considers the possibility of using the statistical characteristics – both numeric and functional - for the vibration processes assessment. The research revealed that the usage of an area under curve of module of characteristic function of the instantaneous values of vibroacoustic signals for different assemblies and mechanisms as a diagnostic sign and forming the vector of diagnostic signs on its basis make possible to suggest a fundamentally new diagnostic parameter that is invariant to the design of the machine and the parameter’s limiting values. The parameter application allows to increase the reliability of diagnostic results and, at the same time, to simplify the diagnostic algorithms.
Boychenko S.N., Kostyukov V.N., Naumenko A.P., Kudryavtseva I.S. Formation of diagnostic signs vector on the basis of a characteristic function of a vibro-acoustic signal // Proceedings of CM2016/MFPT2016
In the last few years a common means for reliability control at refineries has become a risk management system which should provide optimization of maintenance and repair costs while increasing the turnaround interval of process plants. However, a large amount of data required for risk calculation, as well as significant problems connected with unevenness of the data, lack of standards for their collection, subjectiveness of manual data input, does not allow to consider the risk management systems to be complete enough for the increase of turnaround interval and providing required minimal risk level. The main parameter for risk assessment is the probability of equipment failure. Considering the possibilities and properties of the modern condition monitoring systems it becomes possible to propose real-time calculation of process plant failure probability and operation risk.
The paper considers a statistical approach to the assessment of equipment failure probability according to the data of permanently installed condition monitoring systems. It gives the results of research on statistical parameters of machinery operation time and failure probabilities.
Usage of condition monitoring systems for calculation and monitoring of machinery operation risk allows improving the efficiency of implementation of condition monitoring means and turnaround interval of process plants.
Kostyukov V.N., Kostyukov A.V., Boychenko S.N. Assessment of failure probability and operation risks of process plants according to data of condition monitoring systems // Proceedings of CM2016/MFPT2016
To visualize the physical processes that occur in the journal bearings of the shafting of power generating turbosets, a technique for preliminary calculation of a set of characteristics of the journal bearings in the domain of possible movements (DPM) of the rotor journals is proposed. The technique is based on interpolation of the oil film characteristics and is designed for use in real-time diagnostic system COMPACS®.
According to this technique, for each journal bearing, the domain of possible movement of the shaft journal is computed, then triangulation of the area is performed, and the corresponding mesh is constructed. At each node of the mesh, all characteristics of the journal bearing required by the diagnostic system are calculated.
Via shaft-position sensors, the system measures—in the online mode—the instantaneous location of the shaft journal in the bearing and determines the averaged static position of the journals (the pivoting vector). Afterwards, continuous interpolation in the triangulation domain is performed, which allows the real-time calculation of the static and dynamic forces that act on the rotor journal, the flow rate and the temperature of the lubricant, and power friction losses. Use of the proposed method on a running turboset enables diagnosing the technical condition of the shafting support system and promptly identifying the defects that determine the vibrational state and the overall reliability of the turboset. The authors report a number of examples of constructing the DPM and computing the basic static characteristics for elliptical journal bearings typical of large-scale power turbosets. To illustrate the interpolation method, the traditional approach to calculation of bearing properties is applied. This approach is based on a Reynolds two-dimensional isothermal equation that accounts for the mobility of the boundary of the oil film continuity.
Kumenko A.I., Kuzminyh N.Y.,Timin A.V. Calculation and Interpolation of the Characteristics of the Hydrodynamic Journal Bearings in the Domain of Possible Movements of the Rotor Journals // Thermal Engineering, 2016, Vol.63, No. 10, pp. 699-706
A perfection concept is proposed for automated control diagnostics systems used at power plants with application of shaft sensors. It is proposed a systematic approach to solve a number of practical problems related with safety and performance of power generating turbosets. Diagnostics of multi-rotor turboset technical condition are significantly improved due to "complete" set of shaft motion sensors, applied at both sides of each journal bearing. There have been given diagnostics signs of defects measured with shaft sensors and suggested some extra criteria of reliability of the shafting and supports.
To illustrate the method let us compute correcting alignments of fragment of shafting. Using simplified 2-rotor 4-support turboset, it is shown that: corrective alignments of supports, recommended by manufacturers of turboset and generator, may be not sufficient to compensate misalignments of supports, caused by thermal deformations of turboset support system. Those corrective misalignments should be computed individually for each turboset, using real thermal deformations and real clearances of journal bearings.
Kumenko A.I., Kuzminyh N.Y.,Timin A.V. Shaft sensor based on modeling diagnostic signs of power unit defects // Procedia Engineering. - 2016. - № 152. - P.531 – 539
The present work describes the structure and operational concept of the system for machinery condition monitoring using adaptive measurements. The work reveals the results of adaptive measurements procedure of the system illustrated by trends of centrifugal pump vibration signs. The work objective is to increase validity of machinery condition monitoring results through providing high spectral resolution. The spectral resolution is essential for separating defects in the vibration signal spectrum and retaining high speed of vibration measuring. This is required in order to eliminate the omission of defects with minimal expenses on hardware.