Eugenio Prokhorov

Method to Cope with Zero Flows in Newton Solvers for Water Distribution Systems

AbstractThis paper presents and discusses a simple method to deal with zero flows in Newton solvers for water distribution systems, in particular in a previously published global gradient algorithm. The method consists in replacing the Hazen-Williams head loss–flow relationship for flows below a certain threshold with a linear relationship, which coincides with Hazen-Williams’ at zero, but has a nonzero derivative at that point, thus avoiding computational troubles associated with zero flows. Manual and automatic random testing of example networks shows that the proposed method compares well with both the algorithm adopted in water distribution modeling software and the recently proposed regularization method in convergence rate while outperforming them in computational veracity and applicability range.

Pseudotransient Continuation-Based Steady State Solver: Extension to Zero Flow Rates

This paper presents and discusses an extension of the pseudotransient continuation-based steady state solver for hydraulic networks proposed previously to the case of zero flow rates. The original solver, which reduces the solution of the governing nonlinear algebraic equations to the numerical integration of an initial-value problem, has problems in situations in which the head derivative of the flow rate tends to infinity, as is the case with zero flow rates. The extension is on the basis of the use of a model headloss-flow relationship that coincides with the true one at zero and has a finite head derivative at that point. This modified steady state solver is free from some convergence problems that occur in Newton-Raphson method-based solvers when analyzing a pipe network with control devices. The paper includes the results of the numerical analysis of test networks, which demonstrate the convergence of the modified steady state solver for cases in which existing steady state solvers have troubles.

HYDRAULIC SIMULATOR TESTING: METHODS,' TOOLS, AND RESULTS

This paper presents methods for automatic testing of static solvers of hydraulic simulators. The proposed methods are implemented in programs of evolutionary testing of solvers for convergence and the correctness of the converged solution. The programs search for network parameter combinations that result in divergence or incorrect solutions using genetic algorithms: divergences and incorrect but converged results are searched for by maximizing the number of iterations required to obtain a converged solution and the residual of the governing equations, respectively. The use of the methods is illustrated and the efficiency of the programs is shown by the example of the EPANET static solver. Examples of networks that present problems both for EPANET and for some commercial simulators are given. When making quantitative assessments and comparing the original version of the solver with its improved version, use was made of the random testing programs developed. A list of some unsolved problems in the simulation of water distribution systems is presented, and a solution to some of them is suggested.

Capacitance–voltage characteristics of selectively doped AlxGa1−xAs/GaAs heterostructures containing deep traps

The low- and high-frequency capacitance–voltage characteristics of a selectively doped AlxGa1−xAs/GaAs heterostructure containing deep traps in the AlxGa1−xAs layer are calculated. It is shown that the low-frequency capacitance–voltage characteristic features a rising portion when the deep traps are localized at the heterointerface, this rising portion being due to the onset of emptying of the deep traps. The calculated results agree with the experimental data.

Experimental Analysis of Hydraulic Solver Convergence with Genetic Algorithms

A procedure for the experimental convergence evaluation of a hydraulic-network solver is proposed, based on using genetic algorithms to search for network parameter values that maximize the number of iterations of the hydraulic-network solver under test. The efficiency of the method is demonstrated by the example of convergence evaluation for the EPANET hydraulic simulator. Examples of a pipe network and of combinations of parameter values for which the static solver of the simulator fails to converge in a reasonable number of iterations are given. The features of the EPANET 2.00.12 solver responsible for loss of convergence are discussed. New criteria for the automatic start of solution damping aimed at improving the convergence of the solver are proposed. The better convergence of the EPANET solver modified in accordance with these criteria is confirmed by the random and the proposed search-based testing method.

Nonmonotony of the extrinsic photoconductivity of n-type GaAs thin-film structures under backgating

Abstract The extrinsic photoconductivity of an n-type GaAs thin-film structure under backgating has been studied. It is shown that this photoconductivity is a nonmonotonic function of negative substrate voltage, namely, the magnitude of the photoconductivity shows a maximum. This is due to a sharp increase in the magnitude of the photoconductivity caused by an illumination-induced change in the backgating threshold voltage and to its subsequent decrease caused by the accumulation of excess carriers near the film–substrate interface. The calculated results are in a qualitative agreement with the experimental data.

Effect of backgating on the field distribution in planar thin-film GaAs structures

Abstract The distribution of the electric field in planar film–substrate GaAs structures under backgating is considered. It is shown that backgating can make the film exhibit a long-length region of a low-gradient electric field exceeding the threshold of N-type negative differential mobility, the magnitude of negative differential mobility in this region being high enough. At values of the film doping density and film thickness typical of GaAs transferred-electron devices, this region can be as long as several tens of micrometers. The underlying physical mechanism is discussed.

Effect of the impact ionization of deep traps on the field distribution in planar thin-film GaAs structures

The distribution of the electric field in planar GaAs structures made up of a thin film and semi-insulating compensated substrate is considered allowing for the impact ionization of deep traps in the substrate near the film–substrate interface. It is shown that there exists a critical film thickness below which the impact ionization of deep traps can make the film exhibit a long-length region of a uniform electric field exceeding the threshold of N-type negative differential mobility without recourse to special doping profiles.

Sidegating mechanism as a function of the sidegate-to-channel spacing

Abstract It is shown that the type of nonlinearity of the substrate current–voltage characteristic beyond the sidegating threshold changes from super to sublinearity as the sidegate-to-channel spacing increases. Based on this, it is concluded that the sidegating mechanism is a function of the sidegate-to-channel spacing. This conclusion is confirmed using a simple one-dimensional computational model.