Escolar Documentos
Profissional Documentos
Cultura Documentos
Contents
1
CANMET Energy Technology Centre, Natural Resources Canada, Ottawa Canada
2
Energy Simulation Research Unit, University of Strathclyde, Glasgow U.K.
ABSTRACT INTRODUCTION
The standard k − ε model is The integration of CFD and BSim
commonly employed in the numerical Computational Fluid Dynamics
simulation of room air flows. Strictly (CFD) has been used to simulate air flow in
speaking this model is only valid for fully and around buildings for a quarter century.
turbulent flow, whereas weakly turbulent, In the case of interior air flow modelling—
relaminarized, or even stagnant flow can the subject of this paper—analysis is
exist in rooms (particularly in regions usually restricted to single rooms or spaces
remote from air-supply systems). within buildings due to high compute
Consequently, an important assumption is requirements. This presents a quandary, in
implicit when the standard k − ε model is that the boundary conditions of the problem
applied to a room: that the flow behaves as domain (wall-surface temperatures and air
though it were fully turbulent. This paper flows entering/leaving the room) cannot be
examines the implications of that determined a priori. The room does not
assumption. exist in isolation: wall temperatures and air
A room exposed to two-dimensional flows through openings are dynamic and
forced flow was simulated using the dependent on the external weather
standard k − ε model and it was found that excitations, states prevailing throughout the
the eddy viscosity (and thus the turbulent rest of the building, and the operation of
diffusion of heat and momentum) was over- plant equipment, these in turn depending on
predicted near the left wall, where flow was conditions within the room. Simplifying
almost stagnant. This excessive dose of assumptions are usually made, and the
eddy viscosity had a negligible affect on boundary conditions often treated as steady.
flow calculations, but had a significant Workers have begun to address this
influence on thermal predictions. Heat issue by integrating dynamic fabric models
transfer to the left wall was found to be and inter-surface radiation models into
highly sensitive to the value of the eddy CFD codes (Holmes et al 1990; Chen et al
viscosity in the surrounding air. 1995; Moser et al 1995). This allows room
Conclusions are drawn and air flow to be calculated by prescribing
recommendations made on the use of the boundary conditions external to the
standard k−ε model in whole building or in adjoining spaces, rather than
building/computational fluid dynamics within the room.
integrated simulators. Negrão (1995) extended this concept
by integrating a CFD code into the ESP-r
KEYWORDS (ESRU 1997) whole-building simulation
Turbulence modelling, computational fluid model (BSim), the two models interacting
dynamics, surface convection heat transfer. on a time-step basis, exchanging
information at their model boundaries. A
thermal and (optionally) a network air flow
representation of the whole building and
plant is established in the BSim program In contrast to these high-resolution
while a CFD model is created for a single techniques, turbulence models apply
room. BSim then establishes the boundary coarser grids and larger time steps and treat
conditions for the CFD model. Once the the random fluctuations with statistical
CFD solution converges, it passes the methods. The equations of motion are
thermal or air flow results (room air filtered with respect to time, so that rather
temperature, surface convection heat than modelling the details of the turbulent
transfer, air flows entering/leaving the motion, these methods account for the
room) to BSim, which uses the data to influence of turbulence on the mean motion.
calculate the surface temperatures, energy A plethora of turbulence models have
flows, and air flows throughout the been developed but one, the k − ε model,
building. This process is repeated each has enjoyed the greatest usage by far, not
time step. The reader is referred to Clarke only in the domain of buildings, but in most
et al (1995) for details on the ESP-r BSim- fields of study (aerodynamics, hydraulics,
CFD conflated simulator. combustion, etc). The preponderance of
With these conflated approaches, k − ε for modelling room air flows can be
CFD must calculate room heat transfer as seen by reviewing the literature (Whittle
well as room air flow. The quality of the 1986; Nielsen 1989; Jones and Whittle
heat transfer predictions not only affects the 1992; Chen and Jiang 1992; Lemaire et al
BSim results, but also influences the air 1993; Chen 1995).
flow predictions within the room, as the Nature of room air flow
boundary conditions for the CFD Despite many successful applications,
calculations are affected by the heat questions remain regarding the
transfer predictions at previous time steps. appropriateness of k − ε for room air flow
Turbulence modelling modelling. Strictly speaking, the model is
In essence, CFD involves the solution only valid for fully-developed turbulence
of a set of non-linear partial differential (notwithstanding near-wall regions for
equations using numerical techniques, the which adjustments are made to account for
equations expressing fundamental physical viscous effects). Consequently, the
laws—the conservation of mass, application of k − ε implies a very
momentum, and energy. Dealing with important assumption: that the flow is fully
turbulence (the presence of random turbulent or at least behaves like a fully
fluctuations which exists in most flows of turbulent flow.
practical interest) complicates matters But, in general, room air flows are not
considerably. Turbulent fluctuations fully turbulent. Baker et al (1994a)
enhance the transport of momentum, heat, characterize room air motion as typically
and pollutants, and must be considered in turbulent, although only weakly so. Baker
the formulation and solution of the et al (1994b) state that most room air flows
equations of motion. are at least locally turbulent, but flows
Techniques of various levels of away from HVAC supply systems and
complexity and computational intensity obstructions with edges tend to be subtly
have been developed to characterize this turbulent. According to Chen and Jiang
chaotic motion. Some approaches attempt (1992), room air flows may be laminar
to model the turbulent fluctuations, unsteady, locally artificially induced
necessitating very fine grids and time steps. turbulent, transitional, or fully turbulent.
Applications to room air flow modelling Measurements indicate that the flow in the
have already been made (Nielsen 1998; main body of ventilated rooms may be
Emmerich and McGrattan 1998) but transitional (Jones and Whittle 1992).
computational costs remain extremely high According to Chen and Jiang (1992),
and further refinement is necessary. few results on the subject are available but
it is doubtful that a k − ε model can represent the influence of the turbulent
successfully characterize a partially fluctuations on mean-flow quantities. To
turbulent flow. illustrate, examine the diffusion term in the
Research objectives and approach momentum-conservation equations:
The above discussion raises a number ∂ ∂ui ∂u j
of issues and questions. Given that room µ + − ρ u′j u′i (1)
air flows are rarely fully turbulent, can the
∂x j ∂x j ∂x i
the k − ε model lead to accurate predictions The term µ (∂ui /∂x j + ∂u j /∂x i ) involves
of air flow and heat transfer? Is k − ε the gradients of the time-mean velocities and
appropriate choice for the integrated the molecular viscosity, and represents the
modelling systems, in which CFD must diffusion of momentum through molecular
calculate heat transfer as well as air flow? motion. The term − ρ u′j u′i involves the
Can the thermal simulation of whole fluctuating velocities, and represents the
buildings be enhanced through the diffusion of momentum through turbulent
integration of CFD? motion.
The first step in addressing these Boussinesq proposed the eddy-
questions is to assess how k − ε reacts to viscosity concept a century ago, drawing an
room air flows which are not fully turbulent analogy between molecular and turbulent
and to assess the sensitivities of air flow diffusion. Like the viscous stresses in
and heat transfer predictions to laminar flow, the turbulent stresses are
uncertainties in turbulence characterization. assumed to be proportional the the mean-
This is accomplished by performing CFD velocity gradients:
and BSim-CFD simulations of a room
which has a region that can be µ t ∂ui ∂u j 2
−u′i u′j = + − k δ ij (2)
characterized as weakly turbulent. ρ ∂x j ∂x i 3
Modifications to the k − ε model are made where µ t is the eddy viscosity. The
to examine the significance of poor molecular viscosity ( µ ) is a property of the
turbulent characterization in this weakly fluid. In contrast µ t is a property of the
turbulent region. flow: it can differ significantly from one
The k − ε model is first described. flow to another and can vary throughout a
The numerical model is then outlined, the flow domain. Calculating the µ t
experiments described, results presented, distribution is the function of the turbulence
and finally conclusions drawn. model.
The fluctuating quantities are
k- ε TURBULENCE MODEL eliminated from the Reynolds-averaged
The equations of motion equations of motion with the eddy-viscosity
(conservation of mass, momentum, and concept, turbulent diffusion now being
energy) relate the instantaneous quantities completely characterized by gradients in
of pressure, velocity, and temperature. the mean quantities and by the eddy
These equations are filtered with respect to viscosity. The conservation of mass,
time, by replacing the instantaneous field momentum, and energy are expressed as,
quantities with the sum of a time-mean
quantity and a fluctuating quantity (eg. ∂ρ ∂
+ (ρ u j ) = 0 (3)
T = T + T′). This process results in the ∂t ∂x j
Reynolds-averaged form of the equations of
motion, which represent time-mean rather ∂ ∂ ∂P
( ρ ui ) + ( ρ u j ui ) = − + (4)
than instantaneous quantities. ∂t ∂x j ∂x i
The form of the equations of motion
is not altered by the Reynolds-averaging ∂ ∂ui ∂u j
( µ + µ t ) + − ρ β (T ∞ − T)g
process, but new terms are added to ∂x j ∂x j ∂x i
∂ ∂ ∂ k µ t ∂T q ′′′ 1994), this is a separate issue than the one
( ρ T) + ( ρ u j T) = + + (5)
∂t ∂x j ∂x j c p σ t ∂x j c p under examination in the current study.
Z (m)
simulations will indicate how sensitive air µt = 0 at left wall
300
relaminarization (as discussed by
Chikamoto et al 1992 and Nielsen 1998)
200
may address the eddy viscosity over-
prediction in the low-turbulence regions,
100 and should be investigated for use in BSim-
CFD simulators. Low-Reynolds-number
0
k − ε models (eg. Launder and Spalding
0 24 48 72 96 120 144 168
time (hours) 1974; Lam and Bremhorst 1981) may also
Figure 4: Reheat predicted with BSim- address the problem, but it is felt that these
CFD simulations methods remain too computationally
intense for use in BSim-CFD simulators at
CONCLUSIONS the present time.
The standard k − ε model tends to
over-predict the eddy viscosity—and thus REFERENCES
the turbulent diffusion of heat and Baker A.J., Williams P.T., and Kelso R.M.
momentum—in regions which are subtly (1994a), ‘Numerical Calculation of Room
turbulent. This has a negligible impact on Air Motion--Part 1: Math, Physics, and
mean flow calculations, due to the low CFD Modelling’, ASHRAE Transactions,
diffusion of momentum in these regions. 100 (1) 514-530.
However, this can have a significant Baker A.J., Williams P.T., and Kelso R.M.
influence on thermal predictions, due to the (1994b), ‘Development of a Robust Finite
importance of thermal diffusion in regions Element CFD Procedure for Predicting
of low flow. Substantial errors in the Indoor Room Air Motion’, Building and
prediction of surface convection can result. Environment, 29 261-273.
This influence was seen to be local,
Chen Q., Moser A., and Huber A. (1990),
indicating that the over-prediction of eddy
‘Prediction of Buoyant, Turbulent Flow by
viscosity in low-turbulence regions has a
a Low-Reynolds-Number k − ε Model’,
negligible influence on thermal predictions
ASHRAE Transactions, 96(1) 564-573.
in high-turbulence regions of the flow.
Chen Q., Peng X., and van Paassen A.H.C. of the ASME, 103 456-460.
(1995), ‘Prediction of Room Thermal Launder B.E. and Spalding D.B. (1974),
Response by CFD Technique with ‘The Numerical Computation of Turbulent
Conjugate Heat Transfer and Radiation Flows’, Computer Methods in Applied
Models’, ASHRAE Transactions, 3884 Mechanics and Engineering, 3 269-289.
50-60. Lemaire A.D., Chen Q., Ewert M.,
Chen Q. and Jiang Z. (1992), ‘Significant Heikkinen J., Inard C., Moser A., Nielsen
Questions in Predicting Room Air Motion’, P.V., and Whittle G. (1993), ‘Room Air and
ASHRAE Transactions, 98 (1) 929-939. Contaminant Flow, Evaluation of
Chen Q (1995), ‘Comparison of Different Computational Methods’, IEA Annex 20
k − ε Models for Indoor Air Flow Subtask 1 Summary Report.
Computations’, Numerical Heat Transfer, Liddament M (1991), ‘A Review of
B(28) 353-369. Building Air Flow Simulation’, AIVC
Chikamoto T., Murakami S., and Kato S. Technical Note 33.
(1992), ‘Numerical Simulation of Velocity Moser A., Schalin A., Off F., Yuan X.
and Temperature Fields within Atrium (1995), ‘Numerical Modelling of Heat
based on a Modified k − ε Model Transfer by Radiation and Convection in an
Incorporating Damping Effect due to Atrium with Thermal Inertia’, ASHRAE
Thermal Stratification’, Proc. Int. Symp. on Transactions, SD-95-14-4.
Room Air Convection and Ventilation Negrão C.O.R. (1995), Conflation of
Effectiveness, Tokyo. Computational Fluid Dynamics and
Clarke J.A., Dempster W.M. and Negrão C. Building Thermal Simulation, PhD Thesis,
(1995), ‘The Implementation of a University of Strathclyde, Glasgow UK.
Computational Fluid Dynamics Algorithm Nielsen P.V. (1989), ‘Airflow Simulation
within the ESP-r System’, Proc. Building Techniques--Progress and Trends’, Proc.
Simulation ’95, International Building Progress and Trends in Air Infiltration and
Performance Simulation Association, Ventilation Research, 10th AIVC Conf., 1
166-75. 203-223.
Emmerich S.J. and McGrattan K.B. (1998), Nielsen P.V. (1990), ‘Specification of a
‘Application of a Large Eddy Simulation Two-Dimensional Test Case’, IEA Annex
Model to Study Room Airflow’, ASHRAE 20 Research Item 1.45 Technical Report,
Transactions, SF-98-10-2. University of Aalborg, Aalborg Denmark.
ESRU (1997), The ESP-r System for Nielsen P.V. (1998), ‘The Selection of
Building Energy Simulations: User Guide Turbulence Models for Prediction of Room
Version 9 Series, ESRU Manual U96/1, Airflow’, ASHRAE Transactions,
University of Strathclyde, Glasgow UK. SF-98-10-1.
Holmes M.J., Lam J.K-W, Ruddick K.G., Whittle G.E. (1986), ‘Computation of Air
and Whittle G.E. (1990), ‘Computation of Movement and Convective Heat Transfer
Conduction, Convection, and Radiation in Within Buildings’, Int. J. Ambient Energy,
the Perimeter Zone of an Office Space’, 7 (3) 151-164.
Proc. ROOMVENT ’90, Oslo Norway.
Yuan X., Moser A., and Suter P. (1994),
Jones P.J. and Whittle G.E. (1992), ‘Wall Functions for Numerical Simulation
‘Computational Fluid Dynamics for of Turbulent Natural Convection’, Proc.
Building Air Flow Prediction--Current 10th International Heat Transfer Conf.,
Status and Capabilities’, Building and Brighton UK, 191-196.
Environment, 27 (3) 321-338.
Lam C.K.G. and Bremhorst K. (1981), ‘A
Modified Form of the k − ε Model for
Predicting Wall Turbulence’, Transactions