NIST logo

Publication Citation: Fire Dynamics Simulator Version 6: Complex Geometry, Embedded Meshes, and Quality Assessment

NIST Authors in Bold

Author(s): Randall J. McDermott; Glenn P. Forney; Kevin B. McGrattan; William E. Mell;
Title: Fire Dynamics Simulator Version 6: Complex Geometry, Embedded Meshes, and Quality Assessment
Published: June 17, 2010
Abstract: The Fire Dynamics Simulator (FDS) and Smokeview (SMV) are computational and visualization tools specifically designed for large-eddy simulations (LES) of low-speed, thermally driven flows \cite{fds-smv}. FDS Version 6 will offer improvements in the turbulence model (dynamic Smagorinsky \cite{Germano:1991,Moin:1991}), the wall model (Werner Wengle \cite{Werner:1991}), and the scalar transport scheme (Superbee \cite{Roe:1986}, CHARM\cite{Zhou:1995}). These features are already in the testing phase. In this work we discuss further developments related to complex geometry, embedded (or nested) mesh methods, and LES quality assessment. In FDS 6, we are developing two new approaches to handling complex geometry that take us beyond the limitations of the current stair-step method. For grid-resolved objects we are developing a complex geometry module with a second-order immersed boundary method capable of utilizing LES wall stress models \cite{McDermott:2009}. For subgrid objects, such as electrical cables or vegetation in wildland fires, we are developing a particle-based approach, similar to a multi-phase droplet formulation, where the gas phase subgrid sources of mass, momentum, and energy are governed by empirical laws for pyrolysis, drag, and heat transfer, respectively \cite{Mell:2009}. Since LES quality is principally tied to grid resolution, in FDS 6 we introduce an embedded (or nested) mesh strategy which utilizes a staggered grid framework and a fast direct Poisson solver on each mesh block. We are also developing three strategies for quality assessment. The first is to compute local metrics of resolution such as the local fraction of resolved kinetic energy \cite{Pope:2004}. A second strategy is to use local wavelet-based error estimators\cite{SchneiderVasilyev:2010}. Finally, with the latest release of Smokeview we have bundled a new executable called \emph{Smokediff}. With this tool the user can compare results at different grid resolutions or with dif
Proceedings: V European Conference on Computational Fluid Dynamics: ECCOMAS CFD 2010
Pages: 20 pp.
Location: Lisbon, -1
Dates: June 14-17, 2010
Keywords: Fire modeling; Large-eddy simulation; Immersed boundary method; Particle method, Nested grids; Quality assessment
Research Areas: Fire Modeling
PDF version: PDF Document Click here to retrieve PDF version of paper (3MB)