An easy-to-use turbulent inlet boundary condition for large-eddy simulation of radial diffusers

When applying time-accurate computational fluid dynamics, large-eddy simulation (CFD LES) to room ventilation, a common problem is that the supply jets do not model properly if there is no realistic turbulence in the inflow boundary. In this paper we introduce a novel, easy-to-use method of generati...

Full description

Saved in:
Bibliographic Details
Published inBuilding simulation Vol. 9; no. 5; pp. 569 - 581
Main Authors Saarinen, Pekka, Koskela, Hannu, Siikonen, Timo
Format Journal Article
LanguageEnglish
Published Berlin/Heidelberg Springer Berlin Heidelberg 01.10.2016
Springer Nature B.V
Subjects
Atmospheric Protection/Air Quality Control/Air Pollution
Boundary conditions
Building Construction and Design
Computational fluid dynamics
Computer simulation
Diffusers
Engineering
Engineering Thermodynamics
Flow velocity
Fluid flow
Heat and Mass Transfer
Inflow
Large eddy simulation
Mathematical models
Monitoring/Environmental Analysis
Research Article
Simulation
Turbulence
Velocity distribution
Vortices
CFD
radial diffuser
inlet boundary condition
turbulent inflow
large-eddy simulation
radial jet
Online AccessGet full text

Cover

More Information
Summary:When applying time-accurate computational fluid dynamics, large-eddy simulation (CFD LES) to room ventilation, a common problem is that the supply jets do not model properly if there is no realistic turbulence in the inflow boundary. In this paper we introduce a novel, easy-to-use method of generating a turbulent inlet boundary condition for simulating a radial diffuser. The basic idea is to apply a time-independent velocity distribution with realistic spatial statistics, and to add temporal fluctuations by rotation. The method is simple to use, requires only one extra parameter, and does not necessitate large storage files. It needs a momentary velocity distribution on the inflow surface with the desired volume flow rate obtained from a time-accurate internal simulation of the diffuser alone. In a room-scale simulation, this velocity distribution is put in rotating movement in order to mimic a turbulent inlet boundary. The method is verified by comparing the outcome with published measurements and with a simulation containing the interior of the diffuser.
ISSN:1996-3599
1996-8744
DOI:10.1007/s12273-016-0286-3