skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Approximate factorization method for computation of nonhydrostatic mesoscale flows

Conference ·
OSTI ID:6120245

One approach which is widely used in the solution of the conservation equations governing nonhydrostatic meteorological flows employs an incompressible or anelastic formulation in conjunction with the primitive variables. It is well known, however, that the elliptic nature of such a formulation requires that the pressure be treated implicitly, i.e., it satisfies a Poisson equation. It is for this reason that we have recently explored alternative solution techniques and/or formulations to solve the three-dimensional problems of interest. We employ a modification of the compressible formulation called pseudo-compressibility. Just as for the compressible formulation, the pseudo-compressible approach results in a system of equations which are basically of hyperbolic type for the high Reynolds number flows of interest. To solve these equations we employ a very efficient method called the implicit approximate factorization technique. This technique is currently used in conjunction with finite difference schemes to solve three-dimensional incompressible flows. We are developing a similar model but with spatial discretizations based on the Galerkin finite element technique and tensor-product basis functions. It is anticipated that the additional complexity of this finite element approach will not significantly degrade the efficiency of our scheme if an implicit time integration technique is chosen. Moreover it is now computationally feasible to exploit the higher accuracy exhibited by the consistent mass matrix and higher order (e.g., quadratic) basis functions. 11 references.

Research Organization:
Lawrence Livermore National Lab., CA (USA)
DOE Contract Number:
W-7405-ENG-48
OSTI ID:
6120245
Report Number(s):
UCRL-91614; CONF-850682-1; ON: DE85007259; TRN: 85-007917
Resource Relation:
Conference: 7. AMS conference on numerical weather prediction, Montreal, Canada, 17 Jun 1985
Country of Publication:
United States
Language:
English