Hasil Pencarian  ::  Simpan CSV :: Kembali

Abstrak :
Persamaan Euler merupakan salah satu penyederhanaan persamaan Navier-Stokes dengan asumsi inviscid, adiabatik serta menghilangkan efek dari body force. Pada aliran kompresibel, persamaan Euler merupakan sistem pesamaan hiperbolik non-linear untuk hukum konservasi. Pada aliran kompresibel, munculnya fenomena diskontinuitas berupa gelombang kejut sering menimbulkan masalah dalam simulasi, terutama dalam hal akurasi. Pada skema Godunov, akurasi interpolasi untuk memperoleh fluks pada batas antar sel dapat ditingkatkan dengan penggunaan limiter. Salah satu limiter orde tinggi yang dapat digunakan dalam penyelesaian persamaan Euler adalah skema weighted essentially non-oscillatory (WENO). Masalah yang timbul dari penggunaan skema WENO sebagai limiter adalah beban komputasi yang sangat tinggi, terlebih jika sistem persamaan dan domain komputasi yang kompleks. Pengurangan beban komputasi dapat dilakukan dengan cara simplifikasi skema WENO itu sendiri atau dengan menggunakan skema hibrid dimana skema WENO akan digunakan pada kondisi tertentu. Pada penelitian ini dikembangkan skema hibrid orde tinggi yang mengadopsi WENO pada daerah diskontinu dengan deteksi diskontinuitas secara lokal. Metode cell-centered finite volume digunakan untuk diskretisasi ruang. Penyelesaian masalah Riemann pada batas sel digunakan skema Harten-Lax-van Leer contact (HLLC) dan Lax-Friedrichs, serta untuk integrasi waktu digunakan skema strong stability preserving Runge-Kutta orde ketiga untuk memberikan kestabilan yang baik pada skema numerik. Berdasarkan hasil yang diperoleh, skema hibrid yang dikembangkan cukup efektif digunakan dalam penyelesaian masalah aliran kompresibel. Pengurangan waktu komputasi yang signifikan dan akurasi yang baik menjadikan skema hibrid yang dikembangkan menjadi salah satu pilihan skema numerik orde tinggi yang baik untuk dapat diterapkan dalam simulasi aliran kompresibel. ...... Euler equation is a simplification of Navier-Stokes equation which assume the flows are inviscid, adiabatic, and eliminating the effects of body forces. In the compressible flow, the Euler equation is a non-linear hyperbolic conservation laws. The presence of the discontinuities phenomenon in the form of shock wave in the compressible flow often arise the problem in the simulation, mainly in the terms of accuracy. In the Godunovs scheme, the accuracy of interpolation to obtain flux at the intercell boundary can be improved by using a high order limiter. One of the high order limiter that can be used to solve the Euler equation is weighted essentially non-oscillatory (WENO) scheme. The problem that arises from the use of WENO scheme is high computational loads, moreover the system of equations or the domain are very complex. To reduce the computational cost, it can be done by simplify the WENO reconstruction or implement the hybrid scheme where the WENO scheme only applied in certain conditions. In this study, hybrid high order scheme are developed which adopt the WENO schem in the discontinuous region by detecting the local discontinuities. The cell-centered finite volume are used in the spatial discretization. Harten-Lax-van Leer contact (HLLC) and Lax-Friedrichs scheme are used to solve Riemann problem in the cell boundary, and third order strong stability preserving Runge-Kutta (SSP-RK) scheme is used for time integration to ensure the positivity and provide good stability in the numerical scheme. The results shows that the hybrid scheme developed in this work are effective for solving compressible flow problem. The significant reduction of the computational cost and the satisfactory accuracy results are make this hibrid scheme become one of the good choices of high order numerical scheme to be applied in the compressible flow simulation.

Abstrak :
This second edition, like the first, attempts to arrive as simply as possible at some central problems in the Navier-Stokes equations in the following areas: existence, uniqueness, and regularity of solutions in space dimensions two and three; large time behavior of solutions and attractors; and numerical analysis of the Navier-Stokes equations. Since publication of the first edition of these lectures in 1983, there has been extensive research in the area of inertial manifolds for Navier-Stokes equations. These developments are addressed in a new section devoted entirely to inertial manifolds. Inertial manifolds were first introduced under this name in 1985 and, since then, have been systematically studied for partial differential equations of the Navier-Stokes type. Inertial manifolds are a global version of central manifolds. When they exist they encompass the complete dynamics of a system, reducing the dynamics of an infinite system to that of a smooth, finite-dimensional one called the inertial system. Although the theory of inertial manifolds for Navier-Stokes equations is not complete at this time, there is already a very interesting and significant set of results which deserves to be known, in the hope that it will stimulate further research in this area. These results are reported in this edition. Part I presents the Navier-Stokes equations of viscous incompressible fluids and the main boundary-value problems usually associated with these equations. The case of the flow in a bounded domain with periodic or zero boundary conditions is studied and the functional setting of the equation as well as various results on existence, uniqueness, and regularity of time-dependent solutions are given. Part II studies the behavior of solutions of the Navier-Stokes equation when t approaches infinity and attempts to explain turbulence. Part III treats questions related to numerical approximation. In the Appendix, which is new to the second edition, concepts of inertial manifolds are described, definitions and some typical results are recalled, and the existence of inertial systems for two-dimensional Navier-Stokes equations is shown.