A segregated CFD approach to pipe network analysis

Abstract

SUMMARYThe most popular pipe network algorithms fall into three categories depending on whether node, loop or element solving equations are considered. Although node methods have some advantages over the other two methods, some authors have found the node methods to be more unreliable than the other two classes of methods. Node methods are also used in Computational Fluid Dynamics (CFD) to solve the Navier-Stokes equations. Since significant progress has been made in this field in the recent past it was felt that this should have some bearing on the development of more reliable node methods for pipe network problems. In this paper the well-known SIMPLE algorithm of Patankar and Spalding,' which is known in CFD as a segregated method, is extended to deal with pipe network problems. The method can deal with both compressible and incompressible flows. Special attention is given to the solution of the pressure correction equation, the stability of the algorithm, sensitivity to initial conditions and convergence parameters. It is shown that the present method is not very sensitive to initial conditions. The method is very reliable and it deals more effectively with compressible flows than the conventional Newton-Raphson method for incompressible flows.

Optimization of Air Distribution in a Preliminary Design Stage of an Aero-Engine Combustor

“…However, if the zone is too long, this would lead to an unnecessary increase in weight. Thus, the dilution zone volume is restricted by V dz,min ≤V dz ≤ V dz,max , leading to the following inequality constraints: (13) In order to ensure that the port diameters are within a certain range and that the mixing jets entering the flametube do not block the main axial flow, a circumferential blockage factor is introduced as…”

Fast, nonlinear network flow solvers for fluid and thermal transient analysis

“…Most of the existing network flow analysis methods deal mainly with specific flows such as flow in pipelines, incompressible flows, slightly compressible flows, or isothermal flows. They also use the pressure correction technique proposed in Patankar and Spalding (1972) for their computation (Greyvenstein and Laurie, 1994). In Majumdar (1999), a novel finite, volume-based network flow analysis procedure that is capable of analyzing unsteady compressible flows in complex networks involving both components and systems is described.…”

A flow network formulation for compressible and incompressible flow

“…The hybrid pipe network method (Osiadacz, 1987;Greyvenstein and Laurie, 1994;Greyvenstein, 2002) combines aspects of the nodal and loop-based techniques, and such that the advantages of these approaches are inherited. The nodal aspect of the formulation makes the network definition much easier and yields a sparser matrix, while the good convergence characteristics of the loop formulation are preserved (Osiadacz, 1988(Osiadacz, , 1987.…”

“…The nodal aspect of the formulation makes the network definition much easier and yields a sparser matrix, while the good convergence characteristics of the loop formulation are preserved (Osiadacz, 1988(Osiadacz, , 1987. Amongst the hybrid methods, is a technique pioneered by Greyvenstein and Laurie (1994), which makes use of the SIMPLE pressure correction technique developed by Patankar (1980). Their formulation is shown to effectively simulate incompressible flow (Greyvenstein and Laurie, 1994) as well as compressible HFF 18,2 flows at Mach numbers of up to 0.7 (Greyvenstein, 2002).…”

“…Amongst the hybrid methods, is a technique pioneered by Greyvenstein and Laurie (1994), which makes use of the SIMPLE pressure correction technique developed by Patankar (1980). Their formulation is shown to effectively simulate incompressible flow (Greyvenstein and Laurie, 1994) as well as compressible HFF 18,2 flows at Mach numbers of up to 0.7 (Greyvenstein, 2002). In the aforementioned work, separate governing equations are however solved for compressible gas and incompressible liquid flows.…”