Begell House Inc.
Hybrid Methods in Engineering
HME
1099-2391
2
4
2000
FINITE ELEMENT ANALYSIS OF GAS-LIQUID TWO-PHASE FLOW ACROSS AN INLINE TUBE BUNDLE
20
Tomomi
Uchiyama
Nagoya University, Center for Information Media Studies
The gas—liquid two-phase flow across an inline tube bundle is calculated by a finite element method based on an incompressible two-fluid model. In a prior paper, the method was proposed by the author and successfully applied to calculate the two-phase flow across a staggered tube bundle. The present calculation yields unsteady flows with the volumetric fraction of the gas phase distributing quite unevenly. It clarifies that unsteady fluid forces act on the tubes and that the amplitude of the forces increases by an increment in the volumetric fraction of the gas phase upstream of the tube bundle.
PHASE FUNCTION ESTIMATION IN NATURAL WATERS USING DISCRETE ORDINATES METHOD AND MAXIMUM ENTROPY PRINCIPLE
16
E. S.
Chalhoub
Laboratorio Associado de Computacao e Matematica Aplicada (LAC), Instituto Nacional de Pesquisas Espaciais (INPE), Caixa Postal 515, CEP 12201-970—Sao Jose dos Campos SP, Brazil
Haroldo F. de Campos
Velho
Institute Nacional de Pesquisas Espaciais (INPE), Caixa Postal 515, 12201-970 São José dos Campos - SP, Brazil
Fernando Manuel
Ramos
Instituto Nacional de Pesquisas Espaciais (INPE), Laboratório Associado de Computação e Matemática Aplicada (LAC), Caixa Postal 515, 12201-970 - São José dos Campos, SP - Brazil
J. C. R.
Claeyssen
Instituto de Matematica, Universidade Federal do Rio Grande do Sul (UFRGS) Av. Ben to Goncalves, 9500, CEP 91509-900—Porto Alegre RS, Brazil
A technique for estimating the phase function in natural waters, based on the least square estimator, which is associated with a regularization function (maximum entropy principle), is presented. In the analyses, synthetic data (radiances and irradiances) corrupted with noise were used. These data were generated by the same analytical direct model employed in the inversion and implemented by PEESNC code. The inverse problem is iteratively solved by the quasi-Newtonian optimization algorithm. The recovery of the phase function is successfully performed for the selected sample problem.
INVESTIGATION OF INHALABLE AEROSOL DISPERSION AT CUBATÃO BY MEANS OF A MODELING SYSTEM FOR COMPLEX TERRAIN
20
Americo A. F. S.
Kerr
Instituto de Fisica, Universidade de Sao Paulo, Sao Paulo, SP, Brazil, CEP 05315/970; CP.66.318
Silvana A.
do Nascimento
Instituto de Fisica, Universidade de Sao Paulo, Sao Paulo, SP, Brazil, CEP 05315/970; CP.66.318
Domenico
Anfossi
Istituto di Cosmogeofisica del CNR, Corso Fiume4, 10133, Torino, Italy
S. Trini
Castelli
Istituto di Cosmogeofisica del CNR, Corso Fiume4, 10133, Torino, Italy
A modeling system, consisting of the mesoscale model RAMS (Regional Atmospheric Modeling System), the Lagrangian dispersion model SPRAY, and the interface code MIRS (Model for Interfacing RAMS and SPRAY), has been used to numerically investigate the dispersions of inhalable particulate matter emitted by fertilizer plants located at Cubatao (Brazil). RAMS produced the wind fields on the basis of large scale meteorological analysis; MIRS prepared the turbulence statistics field and transformed the RAMS output into SPRAY inputs; SPRAY also computed the concentration field. The results of these simulations were compared with the source apportionment performed by chemical mass balance on samples of aerosol collected during previous field campaigns.
ERROR ANALYSIS OF MIXED LUMPED-DIFFERENTIAL FORMULATIONS IN DIFFUSION PROBLEMS
28
Leonardo
Alves
Departamento de Engenharia Mecânica - TEM
Universidade Federal Fluminense - UFF
Leandro A.
Sphaier
Department of Mechanical Engineering – PGMEC, Universidade Federal Fluminense, Rua
Passo da Patria 156, bloco E, sala 216, Niteroi, RJ, 24210-240, Brazil
Renato M.
Cotta
Laboratory of Nano- and Microfluidics and Microsystems, LabMEMS,
Mechanical Engineering Department and Nanotechnology Engineering Dept.,
POLI & COPPE, Universidade Federal do Rio de Janeiro, Cidade Universitária,
Cx. Postal 68503, Rio de Janeiro, RJ, CEP 21945-970, Brazil; Interdisciplinary Nucleus for Social Development—NIDES/CT, UFRJ, Brazil;
Mechanical Engineering Department, University College London, UCL, United
Kingdom
Mixed lumped-differential formulations for diffusion problems are formally analyzed, with particular emphasis on heat transfer applications The main interest in these formulations resides in the task of modeling problems, prior to the choice of solution strategies, trying to reduce, as much as possible, and within prescribed accuracy requirements, the number of dimensions of any particular partial differential problem. This article illustrates appropriate integration strategies developed within a symbolic computation environment, which are employed to deduce mathematical formulations of comparable simplicity and improved accuracy in comparison with the classical well—established lumping procedures. Besides, approximate error expressions, based on the known boundary and/or initial conditions of the solutions, are readily obtained, for both steady and transient states formulations.
SOLUTION OF THE REYNOLDS-AVERAGED EQUATIONS FOR TURBULENT FLOW VIA INTEGRAL TRANSFORM AND ALGEBRAIC TURBULENCE MODEL
14
J. A.
Lima
Solar Energy Laboratory — LES, Universidade Federal da Paraiba-UFPB, 58059-900 -Joao Pessoa, PB, Brazil
C. A. C.
Santos
Universidade Federal da Paraiba, Laboratorio de Energia Solar (LES/DTM/CT/UFPB), Brazil
Luiz C. G.
Pimentel
LAMMA/Dpto. Meteorologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
The Generalized Integral Transform Technique (GITT) is used in the hybrid numerical-analytical solution of the Reynolds-averaged boundary layer equations for developing turbulent flow inside a parallel-plates channel. An algebraic turbulence model, a modified version of the local Van Driest effective viscosity model, is employed in modeling the turbulent diffusivity. The streamfunction-only formulation is solved by an eigenfunction expansion obtained from the homogeneous biharmonic problem associated with the original problem. Therefore, following the ideas in previous contributions to GITT, numerical results for different Reynolds numbers are obtained, both for illustrating the convergence characteristics of the integral transform approach and for critical comparisons with previously reported results obtained through different models and numerical schemes.
AN EFFICIENT NUMERICAL INTEGRATION ALGORITHM FOR INTEGRAL TRANSFORM SOLUTION OF NONLINEAR CONVECTION-DIFFUSION PROBLEMS
12
Humberto Araujo
Machado
Institute de Aeronautica e Espaco − IAE Pg. Mal. Eduardo Gomes, 50, Vila das Acacias 12228-904, Sao Jose dos Campos, SP, Brazil; Universidade do Estado do Rio de Janeiro − UERJ, Faculdade de Tecnologia − FAT, Rodovia Presidente Dutra km 298 - Polo Industrial, 27537-000, Resende, RJ, Brazil
The Generalized Integral Transform Technique (GITT) is employed to solve convection-diffusion problems in fluids with variable physical properties. In this case, some strong nonlinear terms are observed in the formulation where the analytical transform is not possible. With the simple algorithm presented, the coefficients can be obtained numerically and great flexibility is permitted in changing boundary conditions and source terms of the original equations. Results are shown for the compressible boundary layer equations in a parallel plate channel flow and for the classical lid-driven flow in a square cavity with all the properties (except density) as functions of the temperature.
A MULTIMODAL APPROACH TO NONLINEAR SLOSHING IN A CIRCULAR CYLINDRICAL TANK
22
I. P.
Gavrilyuk
Berufakademie Thuringen-Staatliche Studienakademie, 99817, Eisenach, Am Wartenberg 2, Germany
A. N.
Timokha
National Academy of Sciences of Ukraine, Institute of Mathematics, Tereschenkivska 3, 01601 Kiev, Ukraine
I. A.
Lukovsky
National Academy of Sciences of Ukraine, Institute of Mathematics, Tereschenkivska 3, 01601 Kiev, Ukraine
The fluid sloshing in a mobile container of circular cylindrical shape is examined in the context of potential theory. No roof impact, overturning waves are assumed. Combined numerical-analytical modal modeling is applied. The natural modes are used as the basis of Fourier approximation of free surface evolution. Primary natural modes are assumed to be dominant. The modal system couples nonlinearly the generalized time-varying Fourier coefficients (modal functions). The secondary modes supplement is accounted for by this system, which makes possible the calculation of realistic fluid response and the visualization of wave patterns. The calculation of fluid sloshing is reduced to a Cauchy problem. The Bubnov-Galerkin variational procedure gives approximate values of the initial conditions for simulation of steady solutions. The theory is validated by experimental data. Modal modeling guarantees time—efficient and robust simulations. Multidimensional structure gives more realistic surface wave patterns and may thereby improve the calculation of force (moment) response. The theory breaks down as the wave amplitude response exceeds 45% of the radius.
A FINITE ELEMENT TAYLOR-GALERKIN SCHEME FOR THREE-DIMENSIONAL NUMERICAL SIMULATION OF HIGH COMPRESSIBLE FLOWS WITH ANALYTICAL EVALUATION OF ELEMENT MATRICES
22
Horacio P.
Burbridge
Postgraduate Program in Civil Engineering, Universidade Federal do Rio Grande do Sul. Av. Osvaldo Aranha, 99-3°Andar. 90035-190 Porto Alegro-RS-Brazil
Armando M.
Awruch
Graduate Program in Civil Engineering (PPGEC),
Federal University of Rio Grande do Sul (UFRGS)
99 Osvaldo Aranha Ave. 3rd floor, 90035-190, Porto Alegre, RS, Brazil
An algorithm to simulate three-dimensional high compressible flows of viscous and nonviscous fluids is presented in this work. The time integration procedure was obtained from an expansion in Taylor series of the governing equations. Spatial discretization was carried out by using the finite element method based on the classical Bubnov-Galerkin technique. To obtain considerable improvements in CPU time and memory and take advantage of the fast vectorial processors in modern supercomputers, an analytical evaluation of element matrices was adopted to derive the corresponding expressions from the eight-node isoparametric brick element. Some practical examples are also presented to show the excellent computational performance and good agreement with results obtained previously by other authors.