Algorithm For Numerical Solution Research Articles

An Ensemble Algorithm for Numerical Solutions to Deterministic and Random Parabolic PDEs

In this paper, we develop an ensemble-based time-stepping algorithm to efficiently find numerical solutions to a group of linear, second-order parabolic partial differential equations (PDEs). Particularly, the PDE models in the group could be subject to different diffusion coefficients, initial conditions, boundary conditions, and body forces. The proposed algorithm leads to a single discrete system for the group with multiple right-hand-side vectors by introducing an ensemble average of the diffusion coefficient functions and using a new semi-implicit time integration method. The system could be solved more efficiently than multiple linear systems with a single right-hand-side vector. We first apply the algorithm to deterministic parabolic PDEs and derive a rigorous error estimate that shows the scheme is first-order accurate in time and is optimally accurate in space. We then extend it to find stochastic solutions of parabolic PDEs with random coefficients and put forth an ensemble-based Monte Carlo met...

Algorithm for Numerical Solution of Diffraction Problem on the Joint of Two Open Three-Layer Waveguides

This paper describes the algorithm for the numerical solution of the diffraction problem of waveguide modes at the joint of two open planar waveguides. For planar structures under consideration, we can formulate a scalar diffraction problem, which is a boundary value problem for the Helmholtz equation with a variable coefficient in two-dimensional space. The eigenmode problem for an open three-layer waveguide is the Sturm-Liouville problem for a second-order operator with piecewise constant potential on the axis, where the potential is proportional to the refractive index. The described problem is singular and has a mixed spectrum and therefore the Galerkin method can not be used in this definition. One way to adapt the Galerkin method for the problem solution is to artificially limit the area, which is equivalent to placing the open waveguide in question in a hollow closed waveguide whose boundaries are remote from the real boundaries of the waveguide layer of the open waveguide. Thus, we obtain a diffraction problem on a finite interval and with a discrete spectrum, which can be solved by the projection method, as done in this paper.

A new perspective for quintic B-spline based Crank-Nicolson-differential quadrature method algorithm for numerical solutions of the nonlinear Schrödinger equation

In the present paper, a Crank-Nicolson-differential quadrature method (CN-DQM) based on utilizing quintic B-splines as a tool has been carried out to obtain the numerical solutions for the nonlinear Schrodinger (NLS) equation. For this purpose, first of all, the Schrodinger equation has been converted into coupled real value differential equations and then they have been discretized using both the forward difference formula and the Crank-Nicolson method. After that, Rubin and Graves linearization techniques have been utilized and the differential quadrature method has been applied to obtain an algebraic equation system. Next, in order to be able to test the efficiency of the newly applied method, the error norms, $L_{2}$ and $L_{\infty}$ , as well as the two lowest invariants, $I_{1}$ and $I_{2}$ , have been computed. Besides those, the relative changes in those invariants have been presented. Finally, the newly obtained numerical results have been compared with some of those available in the literature for similar parameters. This comparison clearly indicates that the currently utilized method, namely CN-DQM, is an effective and efficient numerical scheme and allows us to propose to solve a wide range of nonlinear equations.

Modelling Creep (Relaxation) of the Urinary Bladder

Abstract We first present the results of an experiment in which the passive properties of the urinary bladder were investigated using strips of rabbit bladder. Under the assumption that the urinary bladder had orthopaedic characteristics, the strips were taken in the longitudinal and in the circumferential directions. The material was subjected to uniaxial tension, and stress-stretch curves were generated for various rates of deformation. We found that the rates did not have a significantly effect on the passive response of the material. Additionally, the stress-stretch dependence during relaxation of the material when exposed to isometric conditions was determined experimentally. Next, we measured nonlinear stress-stretch dependence to determine the coefficients for this dependence in analytical form using a standard fitting procedure. The same approach was used to obtain the coefficients for the relaxation curves from the experimental data. Two constitutive laws, the nonlinear model for passive response and the creep model, were introduced within the shell finite element for geometrically and materially nonlinear analysis. We provide descriptions of the numerical procedures that were performed by considering the urinary bladder as a thin-walled shell structure subjected to pressure loading. The developed numerical algorithm for the incremental-iterative solution was implemented into the finite element program, PAK. The response of the urinary bladder was calculated for continuous filling, and the numerical and experimental results were compared through cystometrograms (pressure-volume relationships). We also present comparisons of the shapes and volumes of the urinary bladder obtained numerically and experimentally. Finally, the numerical results of the creep response, when placed under constant internal pressure, are provided for various stages of deformation.

Efficiency Analysis of the Parallel Implementation of the SIMPLE Algorithm on Multiprocessor Computers

This paper describes the details of the parallel implementation of the SIMPLE algorithm for numerical solution of the Navier–Stokes system of equations on arbitrary unstructured grids. The iteration schemes for the serial and parallel versions of the SIMPLE algorithm are implemented. In the description of the parallel implementation, special attention is paid to computational data exchange among processors under the condition of the grid model decomposition using fictitious cells. We discuss the specific features for the storage of distributed matrices and implementation of vector-matrix operations in parallel mode. It is shown that the proposed way of matrix storage reduces the number of interprocessor exchanges. A series of numerical experiments illustrates the effect of the multigrid SLAE solver tuning on the general efficiency of the algorithm; the tuning involves the types of the cycles used (V, W, and F), the number of iterations of a smoothing operator, and the number of cells for coarsening. Two ways (direct and indirect) of efficiency evaluation for parallelization of the numerical algorithm are demonstrated. The paper presents the results of solving some internal and external flow problems with the evaluation of parallelization efficiency by two algorithms. It is shown that the proposed parallel implementation enables efficient computations for the problems on a thousand processors. Based on the results obtained, some general recommendations are made for the optimal tuning of the multigrid solver, as well as for selecting the optimal number of cells per processor.

TESTING OF NUMERICAL SOLUTION OF THE PROBLEM OF DETERMINING SOURCES OF MAGNETOSTATIC FIELD IN MAGNETIZED MEDIUM

Purpose. Testing of numerical solution algorithm for integral equation for calculation of plane meridian magnetostatic field source distribution at interfaces of piecewise homogeneous magnetized medium by means of electrostatic analogy. Methodology. The piecewise homogeneous medium consists of three regions with different magnetic permeabilities: the shell of arbitrary meridian section, external unlimited medium outside the shell, and the medium inside the shell. For testing external homogeneous magnetic field effect on spherical shell is considered. The analytical solution of this problem on the basis of electrostatic analogy from the solution of the problem uniform electrostatic field effect on dielectric shell is obtained. We have compared results of numerical solution of integral equation with the data obtained by means of analytical solution at the variation of magnetic permeabilities of regions of medium. Results. Integral equation and the algorithm of its numerical solution for calculation of source field distribution at the boundaries of piecewise homogeneous medium is validated. Testing of integral equations correctness for calculation of fictitious magnetic charges distribution on axisymmetric boundaries of piecewise homogeneous magnetized medium and algorithms of their numerical solutions can be carried out by means of analytical solutions of problems of homogeneous electrostatic field effect analysis on piecewise homogeneous dielectric medium with central symmetry of boundaries – single-layer and multilayer spherical shells. In the case of spherical shell in wide range of values of the parameter λk, including close to ± 1, numerical solution of integral equation is stable, and relative error in calculating of fictitious magnetic charges surface density and magnetic field intensity inside the shell is from tenths of a percent up to several percent except for the cases of very small values of these quantities. Originality. The use analytical solutions for problems of calculation of external electrostatic field effect on piecewise homogeneous dielectric bodies for testing integral equations of magnetostatics and algorithms for their numerical solutions. Practical value. The described method of testing integral equations of magnetostatics and their numerical solutions can be used for calculation of magnetic fields of spacecraft control system electromagnets.

Numerical modeling of water hammer with fluid–structure interaction in a pipeline with viscoelastic supports

In the paper a numerical approach to the four equation model of water hammer (WH) with fluid–structure interaction (FSI) is presented. An algorithm for numerical solution of that model in time domain based on the method of characteristics (MOC) is proposed. Special attention is paid to modeling of the influence of viscoelastic pipe supports. The boundary condition at the support is formulated as a differential equation of junction motion which is solved numerically concurrently with the MOC compatibility equations. Numerical simulations were done with the use of an own computer program. The tests were conducted for a model of real pipeline built in the lab and fastened with specific, complex, supporting system. The numerical results were compared to the experimental records and after some modeling and calibration quite a good agreement was achieved. Basic behaviors in the pressure records were identified and existed discrepancies were discussed and explained. The other physical model tested and preliminary analyzed in the paper is a straight pipeline fixed to the floor only with viscoelastic supports. Such a design created the possibility of significant displacements of the pipeline on the supports and thus allowing for effective testing of the influence of supports stiffness and damping properties on the WH behaviors. The specific initial conditions for that case were formulated, solved and implemented in a computer code. It was found that proper selection of support parameters may produce significant reduction of WH pressure amplitudes, mainly due to effective energy absorption and dissipation at the supports.

Optimal harvesting strategy based on rearrangements of functions

We study the problem of optimal harvesting of a marine species in a bounded domain, with the aim of minimizing harm to the species, under the general assumption that the fishing boats have different capacities. This is a generalization of a result of Kurata and Shi, in which the boats were assumed to have the same maximum harvesting capacity. For this generalization, we need a completely different approach. As such, we use the theory of rearrangements of functions. We prove existence of solutions, and obtain an optimality condition which indicates that the more aggressive harvesting must be pushed toward the boundary of the domain. Furthermore, we prove that radial and Steiner symmetries of the domain are preserved by the solutions. We will also devise an algorithm for numerical solution of the problem, and present the results of some numerical experiments.

Mathematical modeling and numerical calculation of composite structures

The report is devoted to modeling the properties of composite materials. Two major approaches are considered: phenomenological and structural [1]. Within the framework of the first approach reinforced materials are modeled as homogeneous anisotropic medium with efficient physical and mechanical properties. In this case mechanical parameters of the material are determined basing on experimental data. In a structural approach, physical and mechanical parameters of the composite are expressed in terms of the parameters of its components and design of reinforcement that open up opportunities for improvement of the properties of composite structures. The mathematical relations describing the nonlinear elastic three-point bending of isotropic and reinforced beams with account of different strength and stiffness behavior in tension and compression are obtained. An algorithm for numerical solution of corresponding boundary-value problems is proposed and implemented. Results of numerical modeling have been compared to acquired data for polymer matrix and structural carbon fiber reinforced plastics.

Non-periodic acoustic and electromagnetic, scattering from periodic structures in 3D

Scattering of non-periodic waves from unbounded structures is difficult to treat, as one typically formulates the problem in an unbounded domain covering the unbounded periodic structure. The Floquet–Bloch transform reduces the latter problem to a family of decoupled periodic scattering problems. This reduction is in particular interesting from the point of view of numerical computations. We analyze a corresponding fully discrete numerical solution algorithm for three-dimensional scattering problems in acoustics and electromagnetics, proving in particular convergence rates under suitable assumptions on the geometry and the material coefficients. A crucial part of our analysis relies on the continuous dependence of the family of solutions to the quasiperiodic scattering problems on the quasiperiodicity. The latter part is actually more difficult to establish than for corresponding two-dimensional problems. We further provide a numerical example in 3D acoustics that illustrates feasibility of the proposed algorithm.

Geometrically linear continuum theory of dislocations revisited from a thermodynamical perspective

Continuum dislocation theory (CDT) allows the consideration of dislocation ensembles by introducing the dislocation density tensor. Though the kinematics of geometrically linear CDT are well established, the closure of governing field equations is not finished yet. The present study now brings together different principles for such a closure: It is shown how the field equations for the CDT can be obtained from potential energy minimization and from the phase field approach. These two energetic methods are integrated into a generic thermodynamic framework with twofold benefit: First, the rigorous thermodynamic treatment allows clarifying physical consequences of the energetic methods, among them the proof of thermodynamic consistency. Second, the framework provides a basis for consistent extensions of CDT. In this way, a new dynamic formulation of CDT is presented, which enables the analysis of the evolution of dislocation structures during plastic deformation. Moreover, a variety of possible dissipative phenomena is considered and the mechanical balance laws are deduced. For two special cases, the field equations are derived in the strong form and the stability of the solution is analyzed. Next, a flexible numerical solution algorithm is presented using the finite difference method. Solutions of various initial boundary value problems are presented for the case of plane deformations. Therefore, some of the dissipative phenomena are further investigated and two distinct sources of the Bauschinger effect are identified. Special attention is also given to different boundary conditions and their effect on the solution. For the case of uniaxial compression, the numerical results are confronted with experimental data. Thus, the simulations are validated and a new consistent interpretation of the experimental results is achieved.

On one boundary problem for nonlinear stationary controlled system

ABSTRACTThe object of study is nonlinear stationary controlled system of ordinary differential equations with constant disturbance in the right part. The problem of constructing the synthesising control function providing the transfer of this system from the initial state to the origin is considered. The sufficiently simple for numerical implementation algorithm of solution of the above-mentioned problem is obtained. It is shown that for local null controllability of the considered system, it is sufficient that the conditions of the Kalman's type were satisfied. In addition, the estimates restricting the choice of initial conditions and external disturbances under which the transfer is guaranteed are obtained. The main idea of the method of construction of the desired control function consists in reducing the original problem to stabilisation of a special kind linear non-stationary system and solving the Cauchy problem for an auxiliary system of ordinary differential equations closed by stabilising control. The simplicity of the realisation of this algorithm is determined by the construction of the auxiliary system and its stabilisation that could be obtained by analytical methods. The effectiveness of the method is illustrated by solving the problem of crane control and its numerical simulation.

Simulation of friction and stiction in multibody dynamics model problems

ABSTRACTA continuous model of Coulomb friction is used with a tangent space formulation of differential algebraic equations of motion for simulation of multibody dynamic model problems. Characteristics of the model problems studied are similar to those encountered in broad classes of multibody systems, without the associated geometric and analytical complexities. An implicit trapezoidal numerical solution algorithm is used to simulate dynamic response that includes the onset of stiction, its progression, and its termination, avoiding stiff behavior that is reported in the literature when index 3 formulations are used. Analytical criteria for stiction are derived for a three mass Coulomb friction model problem that defines the onset of and departure from stiction events with redundant equations of constraint. The tangent space formulation with implicit trapezoidal integration is applied to this analytical model to compute dynamic response, determine ranges of constraint forces that may occur during periods of stiction, and demonstrate that dynamic response is a discontinuous function of model parameters when stiction occurs. Accuracy of the continuous model of Coulomb friction is established, through comparison of results with those of the analytical model. Cartesian coordinate models of higher dimension are presented for three and four mass model problems that encounter a higher degree of redundancy in constraints during periods of stiction. Simulation of the Cartesian coordinate models, which have characteristics similar to more general multibody systems, yields accurate solutions, without any indication of stiffness in the tangent space equations of motion. Methods successfully demonstrated in model problems provide a foundation for simulation of spatial multibody dynamic systems with friction.

On Numerical Solution of Singularly Perturbed Optimal Control Problems

We study singularly perturbed optimal control problems in which optimal controls may take the form of rapidly oscillating functions. Such rapidly oscillating controls can be constructed on the basis of solutions of certain approximating averaged optimal control problems. We proposed an iterative algorithm for finding numerical solutions of the latter. The effectiveness of the algorithms is demonstrated with the help of numerical examples.

The study and numerical solution of some inverse problems in simulation of hydrophysical fields in water areas with ‘liquid’ boundaries

Abstract Some inverse problems related to mathematical modelling of hydrophysical fields in water areas (seas and oceans) under the presence of ‘liquid’ (open) boundaries are studied and solved numerically in the paper. Numerical solution algorithms for these problems are based on procedures of variational data assimilation.

Modelling of water droplets heat and mass transfer in the course of phase transitions. II: Peculiarities of the droplet radial coordinate and the time grid calibration

This paper continues optimization of numerical solution algorithm of iterative scheme grid for the droplet task, which was presented in the first article of this series. Assumptions were made by optimal assessable number of members which was already defined in numerical experiment in case of compound heat spread by conduction and radiation and an unsteady temperature field was described by infinite integral equation sum. For the convenience of numerical analysis, droplet thermal parameters PT were described by universal Fourier criteria Fo and by dimensionless radial coordinate η function PT(Fo,η). This function is given in form of infinite integral equation sum with each thermal parameter having a distinct initial member and individually defined subsidiary function. This function is given in form of infinite integral equation sum with each thermal parameter having a distinct initial member and individually defined subsidiary function. The droplet time and radial coordinate grading change influence for calculated function graphs PT(Fo,η) was evaluated by water droplets heat transfer and phase transformation numerical experiment. Summarizing by conduction and radiation heated water droplets thermal parameter variation patterns a methodology of forming an optimal grid for droplet task' task iterative solving, is provided.

A two-scale Stefan problem arising in a model for tree sap exudation

The study of tree sap exudation, in which a (leafless) tree generates elevated stem pressure in response to repeated daily freeze-thaw cycles, gives rise to an interesting multi-scale problem involving heat and multiphase liquid/gas transport. The pressure generation mechanism is a cellular-level process that is governed by differential equations for sap transport through porous cell membranes, phase change, heat transport, and generation of osmotic pressure. By assuming a periodic cellular structure based on an appropriate reference cell, we derive an homogenized heat equation governing the global temperature on the scale of the tree stem, with all the remaining physics relegated to equations defined on the reference cell. We derive a corresponding strong formulation of the limit problem and use it to design an efficient numerical solution algorithm. Numerical simulations are then performed to validate the results and draw conclusions regarding the phenomenon of sap exudation, which is of great importance in trees such as sugar maple and a few other related species. The particular form of our homogenized temperature equation is obtained using periodic homogenization techniques with two-scale convergence, which we investigate theoretically in the context of a simpler two-phase Stefan-type problem corresponding to a periodic array of melting cylindrical ice bars with a constant thermal diffusion coefficient. For this reduced model, we prove results on existence, uniqueness and convergence of the two-scale limit solution in the weak form, clearly identifying the missing pieces required to extend the proofs to the fully nonlinear sap exudation model. Numerical simulations of the reduced equations are then compared with results from the complete sap exudation model.

Statement and study of some inverse problems in modelling of hydrophysical fields for water areas with ‘liquid’ boundaries

Abstract In the present paper we formulate and study a class of inverse problems related to the mathematical simulation of hydrophysical fields in water areas (seas and oceans) in the presence of ‘liquid’ (‘open’) boundaries. Some algorithms for numerical solution of this class of problems are proposed. The base of these algorithms is formed by the variational data assimilation procedures.

A hybrid differential evolution and particle swarm optimization algorithm for numerical kinematics solution of remote maintenance manipulators

In the ITER or the future DEMO reactor systems, due to the neutron activation, the remote handling tasks such as inspection, repair and/or maintenance of in-vessel and ex-vessel components must be carried out using a wide variety of special tailored automatic manipulators. These manipulators can be designed as a pure serial articulated arm or a pure parallel mechanism, but for the sophisticated remote handling tasks requiring large workspace and high payload/weight ratio, it would be designed as a hybrid structure combined by these two mechanisms. The complex structure to obtain special remote operation makes the maintenance very time-consuming. To obtain the forward and/or the inverse kinematics of these manipulators, a global optimization method combined by differential evolution (DE) and modified particle swarm optimization (MPSO), here we call it as DEMPSO, is developed to save solution time. The general case studies focus on serial, parallel manipulators and the 10-DOF hybrid redundant serial-parallel robot designed for ITER vacuum vessel remote maintenance. The results would be extrapolated to solve the kinematic problems of the ITER or the future DEMO remote handling manipulators. The proposed optimization algorithm can also be used for the static and/or dynamic parameter identification in robot calibration systems.

A Hybrid Metaheuristic for the Efficient Solution of GARCH with Trend Models

GARCH with trend models represent an efficient tool for the analysis of different commodities via testing for a linear trend in the volatilities. However, to obtain the volatility of a given time series an instance from a particular class of scalar optimization problems (SOPs) has to be solved which still represents a challenge for existing solvers. We propose here a novel algorithm for the efficient numerical solution of such global optimization problems. The algorithm, DE–N, is a hybrid of Differential Evolution and the Newton method. The latter is widely used for the treatment of GARCH related models, but cannot be used as standalone algorithm in this case as the SOPs contain many local minima. The algorithm is tested and compared to some state-of-the-art methods on a benchmark suite consisting of 42 monthtly agricultural commodities series of the Mexican Consumer Price Index basket as well as on two series related to international prices. The results indicate that DE–N is highly competitive and that it is able to reliably solve SOPs derived from GARCH with trend models.