A technique of sample decomposition and rhodium determination in the environment using catalytic method from sulfarsazene oxidation with periodate is developed. The accuracy of the results is confirmed by an independent procedure of rhodium determination using mass spectrometry with inductively coupled plasma. The content of rhodium in the samples of street dust (Moscow) ranges within 0.01 - 0.04 g/ton depending on the place of sampling. The standard deviation of the repeatability of the results ofrhodium determination is below 0.05.

Principle of operation of gas-sensing cells containing ion-selective electrodes for determination of gases or corresponding ions in solutions is described. Prospects of using cells in ionometric analysis of various objects are demonstrated. Analytical characteristics of gas-sensing cells, different types of gas-permeable membranes, compositions of internal electrolyte are discussed. An original gas sensing cell with ammonium-selective electrode for ammonium-ion determination in water is developed. The developed cell containing air space as a gas-permeable membrane is durable and rather simple in design. Conditions of ammonia static gas extraction and ionometric determination are specified. Statistical data processing proved high selectivity, correctness and repeatability of the developed procedure of ammonium-ion determination in mineralized water of complex composition.

High sensitive and selective comprehensive method of determining polycyclic aromatic hydrocarbons (PAH) and their chloro - and nitro-derivatives in soil is presented. The method consists in their extraction from the soil, removal of the compounds interfering with the determination, chromatographic separation of PAH into three groups of compounds, concentration and chromato-mass-spectrometric determination. The determinable concentrations of PAH chlorine- and nitro-derivatives in soil range within 0.1 - 100 pg/kg. The developed method used in analysis of soils of different functional zones of Moscow revealed the areas with excessive concentrations of benzo[a]pyrene (66% of analyzed samples) and provided determination of spatial distribution of PAH in Moscow.

A computer-oriented mathematical model of the austenite decomposition in alloyed steel in conditions of welding thermal cycle is presented. It is shown that each diffusion phase transformation can be described using four parameters: two limits of the temperature range of transformation and two parameters of the Avraami equation. Martensitic transformation can be described using three parameters. A procedure for determination of all required time-temperature parameters of the model obtaining from the chemical composition of steel is described. Regression models of the temperature intervals and kinetic parameters of ferrite-pearlite, bainite and martensite transformations are obtained on the basis of literature data. It is shown that shortcomings of existing visual methods of analysis of the final phase composition of steels are the reasons for inaccuracy and inconsistency of published transformation diagrams. The hardness of the sample which correlates with a number of other mechanical properties of the material is thus chosen as the most objective and reproducible criterion of final phase composition. To check the developed model we compare experimental data and the results of hardness calculation for 140 alloyed steels after their cooling at different values of the cooling rate.

Standard GOST-regulated method is used to develop a procedure of testing stainless steels (the basic material of operating equipment of nuclear power plants) for inter-crystallite corrosion resistance (ICCR). The methodology includes the methods of surface preparation and environment control, the choice of anode etching parameters and conditions of metallographic evaluation of ICCR. The reliability of the results is confirmed by the data of comparative testing of 08KH18N10T steel specimens after sensitizing and austenite conditioning according to the developed standard (AMU GOST 6032-2003) procedures. The proposed approach to ICCR assessing directly upon viewing the surface is advantages compared to the method of “replicas.” The developed procedure is used to study corrosion of the internal surfaces of the main circulating pumps MCP-195-M at Kalinin and Balakovo NPP to analyze the possibility of extending their lifetime after a 20-year period of their service.

Different methods of x-ray diffraction data analysis are compared in a case study of mechanically alloyed powder steels Fe_95,5C_4,5, Fe_90,6Si₅C_4,4, Fe_92,5Cr₃C_4,5 (Fe - 4.3 at.% (1 wt.%) C alloyed with Si and Cr). It is shown that conventional techniques of x-ray analysis, e.g., Williamson-Hall method, are not suitable for studying the structure of nanostructured steels. The grain and dislocation structures of those materials are studied using CWMP full-pro-file method. The average size of the blocks of coherent scattering, their distribution in size, dislocation density and their characteristics are studied as a function of the type of alloying elements and temperature of annealing. A bimodal grain structure is formed upon annealing of mechanically alloyed steels due to selective growth of grains. The dislocation density and evolution of the dislocation structure upon annealing strongly depend on the type of alloying element.

A spectroscopic method of steel structure control is developed. Atomic emission spectroscopy (AES) is used to study the effect of the material structure on the analytical signal in analysis of Kh6VF die tool steel. Analysis of a series of samples taken from the same specimen of a saw blade revealed statistically different results obtained for the same sample before and after 20-min heat treatment at a temperature of 800°C and subsequent cooling in the furnace. The analysis is carried out on an AtomComp 81 and ARL 4460 devices with excitation by the high-voltage spark in equal conditions. The ratio ofthe line intensity of the analyte to that of the main component of the sample (Fe₂₅₉₉) is used in calculations. Assessment of the reliability of differences is performed using contrast of means.

Testing methods of cluster analysis and classification of the recorded arrays of pulses of acoustic emission (AE) is used to detect crack-induced AE signals in the brittle layer of tense-sensitive coating and also to reveal different phases upon transition from the regular to avalanche process of damage accumulation in the physical simulation model. Such phases of damage accumulation are simulated for the process of glass granulate cone formation and then detected using AE control system synchronized with a camera of high speed movie recording.

Method of kinetic indentation at a micro level is used to determine the mechanical properties of ferrite in steel 10. Indentation is carried out within a micro range at the following parameters: the upper limit of the test load is 2 N, depth limit of indentation is 9 μm, and the curvature radius R of a tetrahedral Vickers indenter is 0.22 μm. Indentation load varied from 10 to 2000 mN and exposure time at a constant load from 10 to 60 sec. The dependences of the mechanical characteristics obtained upon kinetic indentation - the modulus of elasticity, Vickers hardness and kinetic hardness - on the indentation load and hold time under the load are derived. Study is carried out to improve the methodology of kinetic indentation, reduce the divergence of the results and reveal the factors affecting determination of the mechanical properties.

Durability of the structural element up to the crack onset depends on the local stresses in the notch area, whereas subsequent survivability of the element prior to final fracture is determined by stress intensity factor (SIF). It is advisable to obtain local stresses in the notch and SIF using common analytical solutions. In this regard, we derived the expressions for SIF in case of small and deep annular cracks and consider them as very sharp notches in conditions of tension, bending or torsion of a rod proceeding from G. Neuber classical solutions for small and deep notches. Dimensional classification of small and deep annular cracks is presented. Estimation of the errors revealed good agreement between the proposed relationships and corresponding data of finite element method. Keywords: specimen; shallow and deep annular cracks; stress concentration factors; estimation of errors.

The mathematical theory of classification is rather diverse, it contains a large number of approaches, models, methods, and algorithms. We distinguish three main results: the best method of diagnosis (discriminant analysis), an adequate indicator of the quality of algorithm of discriminant analysis, and the statement about stopping after a finite number of steps of iterative algorithms of cluster analysis. Proceeding from the Neyman - Pearson Lemma it is shown that the optimal method of diagnosis exists and is expressed through the probability density function corresponding to the classes. If the probability density is unknown, the non-parametric estimators of the training samples should be used. “The probability (or share) the correct classification (diagnosis)” is often used as a quality indicator of the diagnostic algorithm - the more the better. Is shown that a widespread use of this indicator is unreasonable and the other “predictive power” indicator obtained by conversion in the model of linear discriminant analysis is offered. Stopping after a finite number of steps of iterative algorithms of cluster analysis is demonstrated on an example of k-means. In our opinion, the results make a fundamental contribution to the theory of classification and each expert should be informed about the obtained conclusions for further developing and application of the theory of classification.

It is shown that in the method ofNMR¹H ofhigh resolution for wide groups of signals at a low noise level, regardless their overlapping, the consecutive measurements performed during short periods oftime are not independent. Getting data that can be used for statistical processing requires spectra recording with an interval of more than 1 h. For overlapping groups of signals only of the distributions of the measurement errors H_β è H_γ are compatible with the hypothesis of normality. For other variables and for the error of measuring integrated intensity of the narrow signal allocated against the background of the “hump,”, the hypothesis of normality is rejected at a significance level less than 0.01.