It has been shown that the sample with the highest kinetic characteristics can be determined from the calculation and comparison of the values of the diffusion coefficients of americium-241 and uranium on the example of their sorption from a model solution of liquid radioactive waste using various experimental samples of solid-phase extractant based on TODGA.

   The diffusion coefficients of americium-241 and uranium during their sorption have been calculated for three experimental samples of solid-phase extractant based on TODGA. The dependence of the reaction rate on the diffusion coefficients has been estimated. An increased sorption rate for americium-241 in comparison with uranium has been calculated from the dependence of the diffusion coefficients on the characteristics of the experimental samples of solid-phase extractant based on TODGA. Prototype no. 2 of solid-phase extractant based on TODGA with the highest kinetic characteristics has been determined from the calculation of the diffusion coefficients of americium-241 and uranium. The possibility of reducing the volume and costs of sorption experiments with preliminary determination of diffusion coefficients has been shown.

   The replacement of old-type fuel with the new one in the existing RBT reactors, which use spent fuel from the SM reactor converted to uranium-rich fuel after refurbishment, has been discussed. The procedure to convert the reactor to a new fuel is thoroughly described since it is not provided in the reactor regular operation and is not regulated by safety documents. The reactor core loading with new and old fuels in quantitative and percentage terms at the beginning of 2019 is demonstrated. It is shown that the calculated efficiency of RBT control rods does not exceed the designed limits. Attention is paid to the training of personnel engaged in the conversion to the new fuel. Conclusions are made about the safe operation of a new-type fuel, correctness of the selected procedure, and the further possibility of converting the RBT reactors to the new fuel.

   Molten salt reactors (MSRs) represent advances in safety, sustainability, proliferation resistance, and economics. Therefore, an MSR has been chosen as one of the promising reactors for next generation by the Generation IV International Forum. The MSR has been designed to operate on the basis of the Th/233U fuel cycle. Hence, 233U does not exist in nature; it is required to investigate commercially available fuel materials to replace 233U in starting fuel. We have analyzed the fuel cycle and neutron performance of the SD-TMSR with three different fissile materials loading at startup: low-enriched uranium (LEU) (19.79 %), reactor-grade Pu, and 233U. Two different feed mechanisms have been applied. Moreover, the MSR burnup routine within SERPENT-2 has been utilized to simulate the online reprocessing and refueling in the SD-TMSR. In conclusion, the continuous flow of Pu reactor-grade offers the transition to a thorium fuel cycle within a relatively short time (≈ 4.5 yr) compared to 26 yr for 233U startup fuel.

   Nonlinear partial differential equations with variable delay of pantograph type are studied. These equations, in addition to the unknown function u = u(x, t), also contain functions with stretching of one or several independent variables of the form u(px, t), u(x, qt), or u(px, qt), where p and q are the scaling parameters (0 < p < 1, 0 < q < 1). Exact solutions of various classes of such equations are described for the first time. Examples of nonlinear partial differential equations with variable delay of pantograph type that allow self-similar solutions are given (note that partial differential equations with constant delay do not have self-similar solutions). Additive, multiplicative, and generalized separable solutions, as well as solutions of a more complex form, are obtained. Special attention is paid to nonlinear partial differential equations of pantograph type of a fairly general form that contain arbitrary functions. In total, more than 40 nonlinear equations with variable delay of pantograph type, admitting exact solutions, are considered. It is shown that some equations can be generalized to the case of delay, which arbitrarily depends on time. The described equations and their exact solutions can be used to formulate test problems designed to check the adequacy and assess the accuracy of numerical and approximate analytical methods for solving the corresponding nonlinear initial-boundary value problems for partial differential equations with variable delay of pantograph type.

   The version of the Ginzburg–Landau variational equation occurring in condensed matter physics which is called the ψ⁶-model is studied. As in the case of the ψ⁴-model, the corresponding equation is studied together with periodic boundary conditions. Conditions for the existence of single-mode equilibrium states are obtained for a periodic boundary value problem. An answer is given to the question of their stability in the sense of Lyapunov’s definition. In the case of a change in stability by single-mode equilibrium states, local bifurcations near these solutions are studied. It is shown that subcritical (rigid) bifurcations are characteristic of the problem under consideration. Bifurcation analysis involves the methods of the theory of dynamical systems with infinite-dimensional phase space and, first of all, the method of integral manifolds and the method of Poincaré normal forms. It is shown that two-dimensional invariant manifolds formed by spatially inhomogeneous solutions other than single-mode ones bifurcate from single-mode equilibrium states. For such solutions, asymptotic formulas are obtained.

   Analytical properties of solutions of four families of three-dimensional autonomous conservative systems with one or three quadratic nonlinearities are investigated. Conservative systems of the first family have one quadratic nonlinearity and three linear components. Systems of the second family are conservative systems that have one quadratic nonlinearity, two linear components, and one constant. Conservative systems of the third family contain three quadratic nonlinearities and one linear component. Systems of the fourth family include conservative systems with three quadratic nonlinearities and one constant term. The absence of chaos in these systems is their common qualitative property. To analyze the solutions of the considered systems, Painlevé test, as well as the reduction of the systems to their equivalent equations of the second or third order and the comparison of the latter with the known nonlinear P-type equations, has been used. The systems whose general solutions have the Painlevé properties are highlighted. The solutions of such systems are expressed in terms of elliptic functions or solutions of the first Painlevé equation. It is shown that the systems whose general solutions contain moving critical points include those in which one of the component has no moving singular points at all. The systems considered in this work belong to a class of seven families of conservative systems (with the total number of members in the right parts equal to 4). The analytical properties of solutions of the systems of the other three families will be studied elsewhere.

   The results of modeling the AND and OR elements as components of the majority element are presented when switching inputs and simultaneously collecting charge from the particle track. The simulation is performed using 3D TCAD physical models of CMOS transistors according to the design rule of 65 nm bulk technology with shallow trench isolation of transistor groups for tracks with linear energy transfer of 60 MeV cm2/mg. It is found that the beginning of switching elements AND and OR at the inputs hardly affects the transient processes of the formation of noise pulses at the output of the element when collecting the charge from the track. The noise pulses are shifted in time by a time interval equal to the time offset of the track relative to the time of switching element inputs. Collection of the charge from the track leads either to switching an element in advance of the input signals changing or to an additional switching delay. The widths of the noise pulses at the outputs of the AND and OR elements are distributed in the range from 12 to 345 ps. At the same time, the duration of the noise pulse remains almost unchanged for the each specific track input point into the common area of the transistor location regardless of the time of formation of the track.

   The prototype of a multichannel application specific integrated circuit (ASIC) designed at the National Research Nuclear University MEPhI as a front-end part of the gamma-detection systems utilizing “slow” scintillators such as NaI(Tl) or CsI(Tl) readout by Silicon Photomultipliers has been tested. The measurements have been carried on using a specially developed electronic test-board and a software package written in LABVIEW. The gamma-detector module includes a NaI(Tl) scintillator crystal and a matrix of SiPM PM6600 (KETEK) as a photosensor part. The system energy resolution (FWHM) has been measured to be (16.9 ± 0.9) % for 59.5 keV gamma rays. For comparison, the same measurements have been replicated using the analog part of the multichannel ASIC MAROC3 with observed energy resolution (27.7 ± 0.9) %.

   The levelized cost of electricity for nuclear power plants with different types of nuclear reactors has been predicted taking into account the uncertainty in the cost data on the construction and operation of reactor facilities, as well as the uncertainty in the cost data on fuel cycle products and services. Three types of thermal reactors operating in the once-through fuel cycle are considered (two options of spent fuel management are analyzed including the long-term storage of spent fuel at the away-from-reactor storage facility and final geological disposal of spent fuel), as well as four types of fast reactors operating in a closed uranium–plutonium fuel cycle. The estimation of probable LCOE values for various thermal and fast reactor technologies shows that the spread in possible LCOE values for different types of fast reactors turns out to be comparable and largely overlaps with the spread in LCOE values for thermal reactors. This suggests that, based on the LCOE estimations, taking into account the uncertainty in their values caused by the spreads in fuel cycle unit service cost and costs for the construction, operation and maintenance of nuclear power plants, it is impossible to make an unequivocal conclusion about the greatest attraction of any fast reactor concept, relying only on the LCOE analysis, and it is also incorrect to make statements about the lower economic performance of fast reactors as compared to thermal reactors: the corresponding statements should be considered solely as probabilistic ones.

   The number of documents processed during the organization's operational processes is dramatically growing from year to year. As a result, there is a strong demand for automatic processing of such information. Unfortunately, an essential part of such documents is unstructured information. Unstructured document formats are very variable and strongly different from document to document. In this case, automation, through the rule based coding and parsing templates is quite complex and ineffective for further support. It could be analyzed manually, but it is time-consuming and resource intensive. This paper presents the general architecture of an information extraction system based on modern approaches to Natural Language Processing (NLP) and machine learning. The article presents the statistical results of the experiments carried out to solve a number of practical problems of processing unstructured documents. The presented solution allows to process a great amount of unstructured information without coding and preparing parse templates.