Multiple Linear Regression Model Approach for Aerosol

Dispersion in Ventilated Spaces Using Computational Fluid

Dynamics and Dimensional Analysis
Shamia Hoque1; Bakhtier Farouk2; and Charles N. Haas3

Abstract: Aerosol dispersion in living spaces especially bioaerosols, due to accidents or deliberate acts, is of significant current interest.
Computational fluid dynamics 共CFD兲 provides an accurate and detailed platform to study the influence of different parameters on aerosol
distribution in indoor spaces. The simulations however are time consuming and site-specific. The work here introduces an approach
toward addressing this challenge. During emergencies, an accurate, quicker, and more general model is required to give rapid answers to
first responders. Significant parameters influencing aerosol behavior in an office room were identified and through dimensional analysis,
nine dimensionless groups were developed. Fractional factorial design was used to build sixteen scenarios to explore the design space.
These scenarios were then simulated using a comprehensive CFD model. Large eddy simulation with the Smagorinsky subgrid scale
model was applied to compute the airflow. Aerosols were modeled as a dispersed solid phase using the Lagrangian treatment. The
influence of the dimensionless groups on the temporal variation of the number of aerosols in the room and the spatial distribution of the
particles in the room was analyzed. The results showed that all the identified dimensionless groups were significant. Multiple linear
regression models were developed for the prediction of the number of aerosols in the room and their spatial distribution as a function of
the significant parameters influencing aerosol transport. The linear models accurately predicted the data on which they were based but did
not predict the results of the independent tests as well. The limited predictive ability of the model showed that the relationships between
the dimensionless groups are nonlinear and a higher level of experimental design will have to be applied to better explore the design
space.
DOI: 10.1061/共ASCE兲EE.1943-7870.0000201
CE Database subject headings: Air flow; Computational fluid dynamics technique; Design; Dimensional analysis; Regression
models.
Author keywords: Aerosols; Air flow; Computational fluid dynamics; Design; Dimensional analysis; Regression models.

Introduction would be crucial for decision makers planning to decontaminate

The anthrax mailing incidents subsequent to the September 11,
2001 attacks showed the ease with which aerosolized biological
agents could be released and dispersed in an indoor facility. Re-
sponse and decontamination costs after the 2001 attacks alone
exceeded 100 million dollars 共Webb 2003兲. For an efficient re-
sponse after any future bioterrorist attack we have to understand
how aerosols released in an indoor space would be dispersed or
how their number would change with time. The information
1 Tian and Tu 2007; Zhao et al. 2004兲. These studies provide in-
Ph.D. Candidate, Dept. of Civil, Architectural, and Environmental
Engineering, Drexel Univ., 3141 Chestnut St., Philadelphia, Pennsylvania
19104. E-mail: sh338@drexel.edu
2
J. Harland Billings Professor, Dept. of Mechanical Engineering and
Mechanics, Drexel Univ., 3141 Chestnut St., Philadelphia, Pennsylvania
19104. E-mail: bfarouk@coe.drexel.edu
3
L.D. Betz Professor and Department Head, Dept. of Civil, Architec-
tural, and Environmental Engineering, Drexel Univ., 3141 Chestnut St.,
Philadelphia, Pennsylvania 19104 共corresponding author兲. E-mail:
haascn@drexel.edu
Note. This manuscript was submitted on November 18, 2008; ap-
proved on November 20, 2009; published online on November 23, 2009.
Discussion period open until November 1, 2010; separate discussions
must be submitted for individual papers. This paper is part of the Journal
of Environmental Engineering, Vol. 136, No. 6, June 1, 2010. ©ASCE,
ISSN 0733-9372/2010/6-638–649/$25.00.
an infected site efficiently and with available resources.
Aerosol or particle transport and distribution are highly asso-
ciated with airflow motion and turbulence. With the advancement
of computer speed and algorithm sophistication, computational
fluid dynamics 共CFD兲 modeling has increasingly played an im-
portant role in understanding airflow dynamics and particle be-
havior in indoor spaces. Numerous studies have applied CFD to
understand the influence of different conditions on airflow pat-
terns and particle behavior 共Beghein et al. 2005; Chang et al.
2006; Gao and Niu 2007; Lu and Howarth 1996; Lu et al. 1996;
sight into the multiple factors influencing the dispersion of par-
ticles in a room. The results, of each study indicate how many
particles remain in the room 共deposited or suspended兲 over a cer-
tain time period. However, the results are site specific. During an
emergency it would require the development of a site-specific
CFD model, specific to the scenario, to answer questions about
the spatial and temporal variation of the aerosols in the room
which would be time consuming and not user friendly. Research-
ers in diverse applications have developed methods to simplify a
multiparametric problem and replace time consuming simulations
or complex analytical models with user friendly solutions 共Hand
et al. 1984; Musser and Bahnfleth 2001; Sáez and Rittman 1988;
Song et al. 2004兲. The concept is to apply the complex simulation
to multiple scenarios at different conditions to determine the pat-

638 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010


Fig. 1. Geometry of the representative system
tern of the influence of the independent variables on the result.
The process requires time and accuracy to ensure that the design
space is explored efficiently and the results can be applied for
predictive purposes. To determine which parameters influenced
the ultraviolet germicidal irradiation 共UVGI兲 air disinfection sys-
tems, Kowalski et al. 共2003兲 applied dimensional analysis to
group the parameters and used experimental design to decide on
the number of computer simulations to be conducted.
In the present work, comprehensive CFD simulations are ap-
plied to explore the multidimensional design space of the particle
dispersion system in an office room. The study develops a de-
tailed CFD model which is applied to simulate multiple scenarios
of the problem. The different conditions for each simulation were
decided by applying fractional factorial design principles. Dimen-
sional analysis was used to obtain dimensionless groups or factors
from the multiple parameters. The study identifies the important
dimensionless groups and determines the significant factors influ-
encing particle transport and removal. Finally a multiple linear
regression model is developed.
Methods
To develop the model, the following steps were performed:
1. Development of the CFD model;
2. Dimensional analysis;
3. Experimental design: Fractional factorial design; and
4. Simulation steps.
Geometry of the three-dimensional 共3D兲 room is shown in Fig.
1. Fig. 1 is a representative scenario of the problem where aero-
sols are released at a point location within a ventilated room. The
room is bare of any furnishings or partitions. Air is flowing in
through one end of the room and is exiting at the opposite end.
Infiltration and exfiltration of air is assumed negligible. To define
the system, thirteen dimensional variables were selected. “L” 共m兲,
“Hr” 共m兲, and “W” 共m兲 represent the length, height, and width of
the room. The vertical distance between the inlet and outlet of
area “A” 共m2兲 is denoted by “I” 共m兲. I can be either negative or
positive. A positive value of I indicates that the inlet is above the
outlet and a negative value indicates the reverse case. The inlet air
velocity is “U” 共m/s兲. The initial number of the particles released
in the room is “No” 共particles/ m3兲. The final number of particles
remaining in the room after time ␶共s兲 is denoted by N
共particles/ m3兲. The properties of air considered are density, “␳ f ”
共kg/ m3兲 and dynamic viscosity, “␮” 共N / m2 · s兲. The aerosols are
characterized by the diameter, “d p” 共m兲 and density, “␳ p”
共kg/ m3兲.
JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 639
Development of the CFD Model
The representative geometry of the CFD model is the 3D room of
Fig. 1. For simulating indoor airflow and aerosol transport, two
numerical models were used. First, a turbulent flow model was
used to simulate the indoor airflow patterns. Second, a Lagrangian
particle trajectory tracking model was applied to determine the
particle behavior in the airflow field. The numerical simulations
were conducted using CFD-ACE+ 共ESI, R. D. 2004兲 which al-
lowed transient particle calculations.
Air Flow Model
The 3D indoor turbulent flow was modeled here by large eddy
simulation 共LES兲. LES is a technique between direct numerical
simulation 共DNS兲 and Reynolds averaged Navier-Stokes 共RANS兲.
DNS is the most accurate of the three methods but it requires high
computing power 共high-speed and large capacity computer兲 for
airflow simulation in a building 共Beghein et al. 2005兲. RANS
modeling applies different turbulence models to calculate the
Reynolds stresses and close the system of Navier-Stokes equa-
tions. It generates the mean velocity field and not the instanta-
neous velocity field, which influences particle dispersion
significantly. LES generates instantaneous flow information re-
quired for particle simulation 共Beghein et al. 2005; Tian et al.
2007兲. In LES, small-scale turbulence is filtered out from the
Navier-Stokes equations, and a model is used to evaluate small
scales. The size of the filter is equal to the smallest grid size. The
filtered Navier-Stokes equations are solved for the large-scale mo-
tion, which is responsible for most of the momentum and energy
transport. The small scales are from subgrid-scale 共SGS兲 models,
which in turn influence the large eddies. For incompressible flow
of a Newtonian fluid the filtered Navier-Stokes equations are
⳵ ūi
=0 共1兲
⳵ xi
⳵ ūi ⳵ 1 ⳵ p̄ ⳵ ␶ij ⳵2ūi
+ 共ūiū j兲 = − − + ␯t 共2兲
⳵ t ⳵ xj ␳ ⳵ xi ⳵ x j ⳵ xj ⳵ xj
The effect of the small scales appears through the SGS stress
term, ␶ij 共Piomelli 1999兲. The small scales are more uniform and
isotropic than large-scales. The most commonly used SGS stress
model is the Smagorinsky model which is an eddy-viscosity
model of the form, ␶ij − 共1 / 3兲␶kk␦ij = 2␯tS̄ij. The SGS stresses are
related to the large-scale strain rate tensor by S̄ij = 共⳵ūi / ⳵x j
+ ⳵ū j / ⳵x j兲 / 2. The eddy viscosity, ␯t is obtained from the algebraic
model, Eq. 共3兲, where ⌬ is the grid size and Cs is the Smagorin-
sky constant
␯t = Cs⌬2兩S̄兩S̄ij 共3兲
兩S̄兩 = 共2S̄ijS̄ij兲1/2 共4兲
Near solid boundaries or in transitional flow this needs to be
adjusted and has been accomplished by ad hoc corrections such as
Van Driest damping 共Piomelli 1999兲. The Smagorinsky constant
共Cs兲 is 0.16. In the Smagorinsky model the eddy-viscosity coef-
ficient is maintained constant for the entire flow domain.
Particle Model
The particle trajectories were computed using the Lagrangian ap-
proach. The trajectory of the particle was determined by solving
the equations of motion. The equation comes from Newton’s sec-
ond law 共5兲, where, u p is the velocity of the particle of mass m.
Eq. 共6兲 determines the particle velocity at location xi. Appropriate
forces such as, drag 共FD兲 and gravitational forces 共FG兲 have been
included in the left hand side of Eq. 共5兲 hence, 兺F = FD + FG
du
兺 Fi = m dtp 共5兲
dxi
= up 共6兲
dt
The net gravitational force on a sphere is represented by Eq. 共7兲
where ␳ p = density of the particle
1
FG = ␲d3p共␳ p − ␳ f 兲g
6
共7兲
The drag force is represented by the Eq. 共8兲 where, CD = drag
coefficient, uជ = velocity of air flow, ␳ f = density of the fluid, and
Cc = Cunningham correction factor
FD = 共1/8Cc兲␲d2p␳ f CD兩uជp − uជ 兩共u
ជp − uជ 兲 共8兲
The Cunningham correction factor is a function of the Knudsen
number and is determined from Cc = 1 + 共2␭ / d p兲共1.257
+ 0.4e−1.1dp/2␭兲 where d p is the particle diameter. For particle mo-
tion with Reynolds’ number less than 0.1, CD = 24/ Re, when the
Reynolds’ number is from 1,000 to 2 ⫻ 105, CD ⯝ 0.44. In be-
tween, the correct value of CD is calculated from the following
expression: CD = 共24/ Re兲共1 + 0.16 Re2/3兲共Hinds 1999兲.
Numerical Scheme and Boundary Conditions
The control-volume method was employed to discretize the gov-
erning equations. A second order upwind scheme was applied for
the convective-diffusive terms and the Euler first order implicit
scheme was used for the temporal terms in the governing equa-
tions. The outlets were designated as pressure boundaries, with
Dirichlet conditions specified for pressure and Neumann condi-
tions 共i.e., zero normal gradients兲 specified for all other dependent
variables. The boundary conditions at the walls of the room rep-
resent no-slip conditions. The particles were initialized over a flat
surface slightly above the floor to avoid the influence of the
boundary layer. Particle trajectory calculations are terminated
when the particles reach the exit of the room. When a particle
strikes a surface it is assumed that it loses half its energy as it
bounces back. Following assumptions were applied:
1. Incompressible and isothermal conditions;
2. No heat and mass transfer between the air and the particles;
3. The coefficient of restitution was set to 0.5 to account for
some energy loss when the particle strikes the surface;
4. Spherical, solid and neutral particles;
5. Airborne particles has no influence on the surrounding air
flow field, i.e., one-way coupling;
6. Negligible particle-particle interactions; and
7. Negligible resuspension.
640 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010
Dimensional Analysis
Dimensional analysis is a technique for reducing the number and
complexity of experimental variables, which affect a given physi-
cal phenomenon, using a compacting technique. The Buckingham
⌸ theorem outlines a formal method of conducting dimensional
analysis 共Buckingham 1915兲.
The relationship between the input parameters or variables can
be written as
共9兲
f共N/No,L,W,Hr,I,U,A,␳ f ,␮,␳ p,d p,␶兲 = 0
where, N / No is the normalized number of aerosols in the room.
By applying Buckingham ⌸ theorem, out of the twelve variables,
nine dimensionless groups or ⌸ groups are developed. There is
generally no unique set of the ⌸ groups that describe a problem.
Hence, physical insight is used to determine a meaningful set of
dimensionless groups. Based on the developed ⌸ groups, Eq. 共9兲
becomes Eq. 共10兲 where the form of the function ␤ will have to
be determined through numerical investigations
冉 冊 冉
N
No
=␤
L L I WHr Q␳ f ␶ L ␳ p
, , , , , , ,
W Hr Hr A L␮ ␪ d p ␳ f
冊 共10兲
In Eq. 共10兲 the first four groups take into account the dimensions
of the room, the locations of the inlet and outlet and the area of
the inlet and outlet. The fifth group which is very similar to the
Reynolds number defines the flow characteristic in relation to the
room dimensions. Q 共m3 / s兲 is the volumetric flow rate of the air.
The sixth group is the ratio of the particle tracking time, ␶共s兲, to
the theoretical residence time of the room, ␪共s兲, ␪
= V共m3兲 / Q共m3 / s兲. It is the same as the mean residence time for a
CMFR 共completely mixed flow reactor兲 共Nazaroff and Alvarez-
Cohen 2002兲. Similar to the CMFR, air would reside for different
length of times at different locations in a room. Some of the air
would flow out at a time period shorter than the residence time of
the room while another portion of the air would reside for much
longer than the residence time. The second last group compares a
particle property to the length of the room while the last group is
the ratio of the density of the particles released to the density of
the fluid.
Experimental Design: Fractional Factorial Design
Based on Eq. 共10兲 the independent variables or factors are the
eight dimensionless groups: 共L / W兲, 共L / Hr兲, 共I / Hr兲, 共WHr / A兲,
共Q␳ f / L␮兲, 共␶ / ␪兲, 共L / d p兲, and 共␳ p / ␳ f 兲. The dependent variable is
the dimensionless group, 共N / No兲. To determine the significance of
these eight factors and to estimate the magnitude of their indi-
vidual effects as well as their interacting effects on the response
variable, a series of simulations have to be conducted. Factorial
designs are efficient exploratory tools applied to achieve these
objectives with as few simulations as possible. For the present
case a two-level full factorial design would require 28 = 256 simu-
lation runs or cases. Two levels are typically enough to determine
if a factor is significant or not 共Bingham 1997兲. Since a single
simulation would require at least 24–48 h to run on a single PC
the full factorial design is currently impractical. Instead a 16-case
fractional factorial design, 28−4 was developed 共Box et al. 1978;
Montgomery 2005兲. Fractional factorial designs take advantage of
the fact that higher order interactions often tend to become neg-
ligible. A common criterion based on which fractional factorial
designs are generated is “resolution” 共Ankenman 1999; Box et al.
 8−4
Table 1. Design Matrix for a 2IV Fractional Factorial Design
E F G H
Cases A B C D BCD ACD ABC ABD
1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1
2 1 ⫺1 ⫺1 ⫺1 ⫺1 1 1 1
3 ⫺1 1 ⫺1 ⫺1 1 1 1 1
4 1 1 ⫺1 ⫺1 1 ⫺1 ⫺1 ⫺1
5 ⫺1 ⫺1 1 ⫺1 1 1 1 ⫺1
6 1 ⫺1 1 ⫺1 1 ⫺1 ⫺1 1
7 ⫺1 1 1 ⫺1 ⫺1 ⫺1 ⫺1 1
8 1 1 1 ⫺1 ⫺1 1 1 ⫺1
9 ⫺1 ⫺1 ⫺1 1 1 ⫺1 ⫺1 1
10 1 ⫺1 ⫺1 1 1 1 1 ⫺1
11 ⫺1 1 ⫺1 1 ⫺1 1 1 ⫺1
12 1 1 ⫺1 1 ⫺1 ⫺1 ⫺1 1
13 ⫺1 ⫺1 1 1 ⫺1 1 1 1
14 1 ⫺1 1 1 ⫺1 ⫺1 ⫺1 ⫺1
15 ⫺1 1 1 1 1 ⫺1 ⫺1 ⫺1
16 1 1 1 1 1 1 1 1

1978兲. Resolution seeks to reduce confounding between lower minimum value of the flow rate Q was chosen slightly below the
order effects. For the Resolution IV design, no main effect is ASHRAE standard of 0.009 m3 / s person. The residence times, ␪
confounded with any two-factor interaction but two-factor inter- of the room was calculated from the air changes per hour 共ACH兲
actions are confounded with other two-factor interactions. The range of 1–10. This range accommodates the required ventilation
8−4
design chosen here is a Resolution IV design, 2IV . A higher res- rate for houses, offices, and hospitals. The inverse of the resi-
olution design could have been chosen but at the expense of more dence time of a room, ␪ gives the number of ACH. Minimum
simulation time 共64 cases兲.
particle tracking time, ␶ was equal to the minimum residence time
Table 1 illustrates the design matrix. Each of the dimensionless
group is labeled with a coded variable: A, B, C, D, E, F, G, and H. of the room while the maximum was chosen to be five times the
The design generators chosen are E = +BCD, F = +ACD, G = maximum residence time. Based on these values the correspond-
+ABC, and H = +ABD 共Montgomery 2005兲. For A, B, C, and D a ing dimensionless groups were evaluated and the highest and low-
low 共⫺1兲 and a high 共⫺1兲 level are assigned independently. For est values chosen. Table 3 shows the high and low values of each
E, F, G, and H the level is determined based on the design gen- group.
erators. Depending on the level of each coded variable the actual
minimum or maximum value of the dimensionless group is iden-
tified.
The maximum and minimum values of each dimensionless
group were determined by assigning a maximum and minimum Table 3. High and Low Values of the Dimensionless Groups
value for each parameter 共Table 2兲. The minimum values for each Dimensionless High value Low value
individual parameter were assigned by assuming an office room Factors group 共+兲 共⫺兲
for a single occupant while the maximum values were assumed
by considering a room capable of holding up to 30 people. The A
冉冊L
W
3.3 0.4

Table 2. Maximum and Minimum Levels for Each of the Parameters

B
冉冊I
Hr
0.6 ⫺0.6

L m
Maximum
10
Minimum
3
C
冉 冊
WHr
A
280 15

Hr
W
m
m
4
7
3
3
D
冉冊
␶
␪
5 1

I
A
m
m2
2.5
0.8
⫺2.5
0.1
E +BCD
冉 冊
Q␳ f
L␮
5033 50

冉 冊
␶ s 18,000 360 F +ACD 3.3 0.8
L
Q m3 / s 0.24 0.008
Hr
␪
冉冊
s 3,600 360
G +ABC L 1 ⫻ 107 6 ⫻ 105
␳f kg/ m3 1.1614
␮ kg/m s 1.85⫻ 10−5 dp
␳p
dp
kg/ m3
m
1,000
5 ⫻ 10−6
800
1 ⫻ 10−6
H +ABD
冉冊␳p
␳f
861 689

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 641

Fig. 2. Evolution of the instantaneous velocity 共in the x direction兲 at
the center point of the room; Case 10
Simulation of Air Flow and Aerosol Transport in a
Room
Each of the 16 cases was repeated once to obtain duplicate re-
sults, i.e., in total 32 cases were simulated. For each case, first the
dimensions of the room, the inlet air velocity, and physical prop-
erties of air and particles were determined from the levels of each
dimensionless group or factor as given in Table 1. The length of
the room was initially assumed to be a certain value and the
corresponding dimensions calculated from the high and low val-
ues shown in Table 3. The ACH was then calculated and the
length adjusted until the ACH was within the range of 1–10. Next
the air flow is developed for one residence time. After one resi-
dence time the airflow achieves a pseudo steady state condition.
Fig. 2 shows the velocity profile in the x direction over time for a
sample case 共Case 10兲 with a residence time of 405 s.
After developing the airflow, 100,000 particles were released
and tracked for the required amount of time based on the level of
D, 共␶ / ␪兲. Particles were initialized over an area of 0.01 m2. The
duplicate case was simulated by randomizing the positions of the
particles over the same area when released from the floor of the
room. Randomizing the particle locations took care of the fact
that if particles were released from the same region again they
would not occupy the same exact location. For all the simulation
cases, nonuniform grids were generated. The smallest mesh size
was 0.01 m. The time step used in the calculations was 0.05 s. It
took 2 months to complete the 32 cases, requiring on an average,
48 h for each case. At the end of each case the number of particles
remaining in the room and the spatial distribution of the particles
in the room was determined.
Results and Discussions
Validation of the CFD model
To validate the CFD model it was applied to simulate scenarios
from the literature and the results obtained were validated against
experimental measurements of indoor air velocities 共Posner et al.
2003兲 and mass concentrations of particles 共Lu et al. 1996兲. The
same procedure was applied by other writers to validate their
models 共Chang et al. 2006兲. Although the room geometries are
different from the scenarios being studied the objective was to
gain confidence on the CFD model by validating its prediction
against available experimental data.
Posner et al. 共2003兲 conducted experiments in a scaled down
model room of dimensions 0.91 m共x兲 ⫻ 0.46 m共y兲 ⫻ 0.31 m共z兲.
642 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010
Fig. 3. Comparison between LES predictions and data from Posner
et al. 共2003兲 of the vertical mean velocity at the midpartition height
The room was partitioned into two zones with a partition located
at the center of height 0.15 m. The square inlet and outlet area
was 0.01 m2. The velocity of the uniform inflow was 0.235 m/s.
Based on the inlet dimensions the Reynolds’ number was 1,600.
The Reynolds’ number in the room could range between laminar
and turbulent depending on the length and velocity scales. This
scenario was adopted and used with the current numerical model.
For the numerical results the velocity data was averaged over the
last 60 s. The vertical mean velocity along the midpartition height
obtained from the numerical simulations and experimental data
has been compared in Fig. 3. The plot shows that the numerical
and measured results are in very good agreement.
The Lagrangian particle model was verified against experi-
mental data obtained by Lu et al. 共1996兲. The experiment was
conducted in a two zone room with a connecting door. The di-
mensions were 5 m共x兲 ⫻ 2.4 m共y兲 ⫻ 3 m共z兲. The inlet and outlet
dimensions were 0.15共x兲 ⫻ 0.5共y兲 ⫻ 1.0共z兲 m3. The dimensions of
the door were 0.95 m共y兲 ⫻ 0.7 m共z兲. To simulate the scenario,
airflow was first established in the room for one residence time
and then particle trajectory calculation was started assuming the
particles were uniformly mixed in the first zone. Fig. 4 shows the
change of the concentration of the particles over time in Zone 1 as
particles exit through the connecting door into Zone 2. Numerical
results agreed very well with the experimental data as shown in
the plot.
Fig. 4. Comparison of the simulated particle mass concentration with
measured data from Lu et al. 共1996兲 for Zone 1
Table 4. Design and Results of the 16 Duplicate Cases
Fraction of particles Mean of the SD of the
E F G H remaining in the room remaining particles remaining particles
Cases A B C D BCD ACD ABC ABD N1 / N0 N2 / N0 ␮1 ␮2 ␴1 ␴2
1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 ⫺1 0.49 0.71 156.52 120.47 972.94 784.86
2 1 ⫺1 ⫺1 ⫺1 ⫺1 1 1 1 0.00 0.00 0 0 0 0
3 ⫺1 1 ⫺1 ⫺1 1 1 1 1 0.00 0.00 0 0 0 0
4 1 1 ⫺1 ⫺1 1 ⫺1 ⫺1 ⫺1 0.88 0.91 0.12 0.12 6.15 6.30
5 ⫺1 ⫺1 1 ⫺1 1 1 1 ⫺1 0.50 0.50 0.072 0.07 0.27 0.27
6 1 ⫺1 1 ⫺1 1 ⫺1 ⫺1 1 0.02 0.01 0.003 0.003 0.053 0.058
7 ⫺1 1 1 ⫺1 ⫺1 ⫺1 ⫺1 1 0.03 0.02 2.12 2.43 28.60 28.61
8 1 1 1 ⫺1 ⫺1 1 1 ⫺1 0.88 0.92 236.47 222.74 487.92 352.47
9 ⫺1 ⫺1 ⫺1 1 1 ⫺1 ⫺1 1 0.08 0.08 0.033 0.033 0.54 0.53
10 1 ⫺1 ⫺1 1 1 1 1 ⫺1 0.02 0.02 0.0064 0.0062 0.081 0.101
11 ⫺1 1 ⫺1 1 ⫺1 1 1 ⫺1 0.02 0.03 4.78 3.13 26.62 14.99
12 1 1 ⫺1 1 ⫺1 ⫺1 ⫺1 1 0.00 0.00 0 0 0 0
13 ⫺1 ⫺1 1 1 ⫺1 1 1 1 0.06 0.07 15.45 13.43 36.61 31.5
14 1 ⫺1 1 1 ⫺1 ⫺1 ⫺1 ⫺1 1.00 1.00 46.75 46.85 832.62 832.62
15 ⫺1 1 1 1 1 ⫺1 ⫺1 ⫺1 0.01 0.00 0 0.0027 0 0.052
16 1 1 1 1 1 1 1 1 0.01 0.01 0.0013 0.0011 0.036 0.034

Computational Results
The data from the 16 duplicate cases from the fractional factorial
design were analyzed to determine the effect of the factors on the
temporal variation of the number of particles in the room and on
the spatial variability of the particles within the room. Table 4
lists the levels of the 16 duplicate cases, the normalized number
of particles remaining in the room, i.e., 共N / No兲 and the mean and
standard deviation 共SD兲, ␴ of the number of particles at different
locations of the room at the end of each case. By randomizing the
particle positions during release, the results of duplicate cases
showed some variability.
Analyzing the results; for five cases, half or more of the par-
ticles remain in the room. In some cases the particles remained
clustered at different locations in the room while for others they
were more evenly dispersed. No particular dimensionless group
or parameter appeared to be the main reason for the particle be-
havior. Figs. 5共a and b兲 show the velocity vector plots for two
cases 共3 and 10兲 averaged over 60 s. The plots are representative
of the airflow patterns observed in all the 16 duplicate cases. A
central circulation zone forms at the core of the room and smaller
circulation zones near the corners. The size of the circulation
zones and the velocity magnitude at the different locations varied
with the geometry and boundary conditions of each case. The
inlet velocities are 0.06 and 0.33 m/s and the dimensions are
6 m共L兲 ⫻ 8 m共H兲 ⫻ 14 m共W兲 and 9 m共L兲 ⫻ 12 m共H兲
⫻ 2.7 m共W兲 for Cases 3 and 10, respectively. The Reynolds’
number based on the inlet dimensions are 11,052 and 30,822,
respectively. Fig. 5共c兲 shows the root mean square of ūi on the
vertical axis located at the midpoint of the central plane for Cases
3 and 10. The velocity magnitude for both cases is highest at the
inlet region. The turbulent kinetic energy, k 共m2 / s2兲 profiles for
the same cases are shown in Fig. 5共d兲. Near the vicinity of the
inlet the kinetic energy is the highest. Fig. 6共a兲 shows the location
of the particles for Case 10 after 240 s. The particles have entered
the central circulation zone of the room and some are exiting.
After 2,025 s, the particle tracking time, nearly 80% of the par-
ticles had exited from the room. The remaining particles were
mainly dispersed near the top of the room 关Fig. 6共b兲兴.
To determine the main and interacting effects of the factors on
the particles, the results were analyzed via ANOVA and regres-
sion methods using the statistical toolbox of MATLAB 共Math-
Works 2007兲.
Temporal Variation of the Aerosols Released in the
Room
To determine the significant factors, an ANOVA table was gener-
ated. The residuals were plotted against the predicted value to
assess the adequacy of the model 共Devore 2004兲. The points on
the residual plot 关Fig. 7共a兲兴 are evenly distributed on both sides of
the horizontal line at zero. Some of the data points are clustered
because most of the data does not vary over a wide range. The
normalized number of particles in the room was log-transformed
to aid presentation of the data 关Fig. 7共b兲兴. Table 5 summarizes the
results of the analysis. All the main effects and the two-factor
interactions are listed in the first column of the table. The con-
secutive columns show the sum of square errors, the degrees of
freedom, mean square error and the F-test statistic. The last col-
umn shows the p values. This is the probability that the observed
effect is due to random error. Any p value below 0.05 was noted
to be significant, i.e., the observed effect is greater than would be
expected by chance.
Looking at the p values the factors A共L / W兲, D共␶ / ␪兲,
E共Q␳ f / L␮兲 and G共L / d p兲 were identified as not significant. How-
ever, the p values for the two way interactions indicate that except
for the two way interactions of 关AD兴 and 关AG兴 all other interac-
tions are significant. This analysis seems to lead to the conclusion
that the groups L / d p and ␶ / ␪ was not significant. These dimen-
sionless groups are further discussed below.
The magnitude of difference between the scales of the length
of the room in meters and the particle diameter in micrometers is
106. As a result the change of L / d p may not significantly influence
the number of particles remaining in the room and could be ig-
nored from future calculations. The other dimensionless group
which was classified as not significant is ␶ / ␪. Identifying this
group as not significant indicates that as more time passes, the

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 643

Fig. 5. 共a兲 Velocity vector plot at the end of one residence time for Case 3, maximum vector length represents 0.08 m/s 共XY plane located at the
mid section of the room兲; 共b兲 velocity vector plot at the end of one residence time for Case 10, maximum vector length represents 0.50 m/s 共XY
plane located at the mid section of the room兲; 共c兲 RMS of the instantaneous u velocity component along the normalized central vertical axis on
the mid XY plane; and 共d兲 turbulent kinetic energy profiles along the normalized central vertical axis on mid-XY plane

moment at which the particles were released in the room becomes

less important to the future state of the particles. Hence, the num-
ber of particles that remained in the room was determined within
the residence time of the room and the length of the particle
tracking time did not significantly change the final number. Figs.
8 and 9 are plots of number of particles remaining in the room
over time for Cases 5 and 16. Most of the particles exit the room
after 200 s for Case 5 共␪ = 875 s兲 and for Case 16, the particle exit
rate declines rapidly after 400 s 共␪ = 405 s兲. For both cases the
maximum number of particles exits within the residence time in

Fig. 6. 共a兲 Particle positions after 240 s for Case 10 showing that the Fig. 7. 共a兲 Plot of standardized residuals against predicted values
particles have moved into the central circulation zone of the room 共dependent variable: 关N / No兴兲 determined via ANOVA; 共b兲 plot of
共front view of the room兲; 共b兲 particle locations at the end of the standardized residuals against predicted values 共dependent
particle tracking time, 2,025 s for Case 10 共front view of the room兲 variable: log关共N + 1兲 / 共No + 1兲兴兲 determined via ANOVA

644 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010

Table 5. ANOVA 兵Dependent Variable: log关共N + 1兲 / 共No + 1兲兴其
Source Sum square df Mean square F Significance
A 0.399 1 0.399 0.247 0.626
B 34.713 1 34.713 21.551 0.000
C 58.307 1 58.307 36.194 0.000
D 3.577 1 3.577 2.220 0.156
E 0.802 1 0.802 0.498 0.491
F 26.965 1 26.965 16.739 0.001
G 6.763 1 6.763 4.199 0.057
H 165.406 1 165.406 102.676 0.000
AB 62.824 1 62.824 39.003 0.000
AC 24.684 1 24.684 15.322 0.001
AD 0.982 1 0.982 0.609 0.446
AE 68.477 1 68.477 42.510 0.000
AF 9.030 1 9.030 5.606 0.031
AG 0.184 1 0.184 0.114 0.740
AH 48.233 1 48.233 29.945 0.000
Error 25.778 16 1.611
Total 537.125 31
Note: R squared= 0.952 共adjusted R squared= 0.907兲.

accordance with the above conclusion. However, a different trend
is noticed for Case 11 共Fig. 10兲. The residence time is 875 s but
none of the particles exit the room until after 1,000 s. The par-
ticles also exit in two steps with the second step happening after
3,200 s. Analysis of the results of this case showed that the par-
ticles had become entrapped in the vortex in the center of the
Fig. 8. Number of particles remaining in the room as a function of
time 共Case 5兲
room with some portion of the particles first exiting as they be-
came entrained into the main stream. Because of the presence of
a case which is not in agreement to the conclusion reached from
ANOVA, further investigation is required before screening out
␶ / ␪ from future simulations.
The multiple linear regression model was developed through
forward step-wise regression. Based on the fractional factorial
design developed, there are seven main effects and eight two way
interaction effects. Each interacting effect is also a summation of
the effect of other two way interaction effects. The same conclu-
sions as ANOVA were reached through this method as well, but
the main effect of G 共L / d p兲 was also identified as significant. Eq.
共11兲 is the regression model in terms of the coded variables. The
R2 for the regression model is 0.94. Based on the magnitude of
the coefficients, H 共␳ p / ␳ f 兲 appears to be relatively the most sig-
nificant group among the main effects followed by C共WH / A兲,
B共I / H兲, and F共L / H兲. G共L / d p兲 is the least important. The magni-
tude of the coefficients of the interacting effects is similar to the
coefficient of G. Among the interacting effects the most important
is 关AE+ BF+ DG+ CH兴 and the least is 关AF+ CD+ BE+ GH兴. Fig.
11 is a plot of the standardized residual against the predicted
values from the regression model. The standardized residual is
obtained by dividing the residual by the estimated SD of the
residual 共Devore 2004兲. The points do not follow any trend show-

Fig. 9. Number of particles remaining in the room as a function of Fig. 10. Number of particles remaining in the room as a function of
time 共Case 16兲 time 共Case 11兲

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 645

Fig. 11. Plot of standardized residuals against predicted values 共de-
pendent variable: log关共N + 1兲 / 共No + 1兲兴兲 for the regression model, Eq.
共11兲

ing that the residuals are not dependent on the predicted values.
Two points 关coordinates 共⫺7.8, ⫺3.8兲 and 共⫺7.8, 3.2兲兴, however
appear to be outlying values falling beyond the range of ⫺2 to
+2. The other residuals are within two SDs of their expected
value of zero Fig. 12. 共a兲 Positions of 49,795 particles in the room at the end of
Case 5; 共b兲 histogram showing the frequency of the number of par-
log关共N + 1兲/共No + 1兲兴 = − 4.37 − 1.042关B兴 + 1.349关C兴 + 0.918关F兴 ticles in each domain 共0.1⫻ 0.1⫻ 0.1 m3兲 at the end of Case 5, the
− 0.460关G兴 − 2.274关H兴 + 0.350关AB + EF y-axis has been log-transformed for a clearer presentation

+ CG + DH兴 + 0.220关AC + DF + BG
+ EH兴 + 0.366关AE + BF + DG + CH兴
− 0.133关AF + CD + BE + GH兴
− 0.307关AH + BD + CE + FG兴
Forward step-wise regression was again applied to determine
the coefficients of the linear regression model and the same con-
clusions were reached regarding the significance of the factors.
共11兲 Eq. 共12兲 is the developed regression model

Spatial Variability of the Aerosol Distribution in the

Room
To determine the influence of the factors on the spatial distribu-
tion of the particles in the room, each room was broken down into
small domains of size 0.1⫻ 0.1⫻ 0.1 m3. The mean and SD of
the number of particles at different locations was then calculated
by counting the number of particles in each subdomain. This pro-
cess was repeated for all the cases. The results have been shown
in Table 4. If the SD was different from the mean by a large
margin it indicates that there was more clustering of the particles
in the room. Less difference between the mean and SD values
indicated more uniform dispersion of the particles in the room.
Histograms were used to graphically display the frequency of the
number of particles in each volume of the room. Figs. 12 and 13
shows the snapshots of the particle positions at the end of Cases 5
and 16 as well as the corresponding histograms for the particle
distribution. For example, 818 particles remain in the room at the
end of the simulation 关Fig. 13共a兲兴. In Fig. 13共b兲 it can be seen that
most of the subdomains are empty while ⬃800 domains 共fre-
quency兲 contain one particle each, which correspond to the num-
ber of particles remaining in the room at the end of the
simulation.
ANOVA was used again to distinguish between the different
factors and their influence on the SD. Table 6 summarizes the
analysis of the data with the dependent variable being the log-
transformed SD 共␴兲. The p values indicate that the main effects of Fig. 13. 共a兲 Positions of 818 particles in the room at the end of Case
all the factors except factor F 共L / Hr兲 and all the two way inter- 16; 共b兲 histogram showing the frequency of the number of particles in
actions are significant. Fig. 14 is the residual plot of the predicted each domain 共0.1⫻ 0.1⫻ 0.1 m3兲 at the end of Case 16, the y-axis
values against the standardized residuals. has been log-transformed for a clearer presentation

646 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010

Table 6. ANOVA 共Dependent Variable: log关␴ + 1兴兲
Source Sum square df Mean square F Significance
A 0.764 1 0.764 53.94 0.000
B 1.354 1 1.354 95.52 0.000
C 7.347 1 7.347 518.53 0.000
D 2.688 1 2.688 189.72 0.000
E 88.427 1 88.427 6240.53 0.000
F 0.047 1 0.047 3.32 0.087
G 5.208 1 5.208 367.57 0.000
H 38.369 1 38.369 2707.83 0.000
AB 3.975 1 3.975 280.52 0.000
AC 24.122 1 24.122 1702.36 0.000
AD 0.609 1 0.609 42.99 0.000
AE 3.635 1 3.635 256.55 0.000
AF 4.126 1 4.126 291.55 0.000
AG 0.193 1 0.193 13.61 0.002
AH 17.834 1 17.834 1258.58 0.000
Error 0.227 16 0.0142
Total 198.927 31
Note: R squared= 0.999 共adjusted R squared= 0.998兲

log关␴ + 1兴 = 2.018 − 0.155关A兴 − 0.206关B兴 + 0.479关C兴 − 0.290关D兴
− 1.662关E兴 − 0.403关G兴 − 1.10关H兴 + 0.088关AB + EF
+ CG + DH兴 + 0.217关AC + DF + BG + EH兴
+ 0.035关AD + CF + EG + BH兴 + 0.084关AE + BF
+ DG + CH兴 + 0.090关AF + CD + BE + GH兴
+ 0.019关AG + BC + DE + FH兴 − 0.187关AH + BD
+ CE + FG兴 共12兲
The R2 for the regression model is 0.99. In Eq. 共12兲 based on the
coefficient; E共Q␳ / L␮兲 is the most significant followed by
H共␳ p / ␳ f 兲, C共WH / A兲, G共L / d p兲, D共␶ / ␪兲, B共I / H兲, and then
A共L / W兲. Among the interacting effects 关AC+ DF+ BG+ EH兴 is
relatively significant. Fig. 15 is the residual plot for the regression
model. The points also do not follow any trend and appear to be
distributed evenly about the horizontal line with most points
Validation Studies Applying the Multiple Linear
Regression Models
Thirty two separate and independent CFD simulations were con-
ducted, with the magnitude of the input variables kept within the
range of the maximum and minimum levels of each dimension-
less group, to obtain a data set for validation of the linear models.
The data obtained from the simulations were compared to the
results predicted from the developed models. First the regression
model, Eq. 共11兲 was applied to calculate the lognormalized num-
ber of particles remaining in the room for the thirty two indepen-
dent cases. The average and SD of 共simulated− predicted兲 results
was 0.927 and 4.00 while the standard error, from Eq. 共13兲 共De-
vore 2004兲, was 4.99. The standard error was 1.23 when the
model was applied to predict the data from which the model was
developed, i.e., the fractional factorial data set
Standard error

冑
being within two SDs of the expected value, zero. One significant
outlier was located at coordinates 共3, ⫺3.8兲.
=
兺 共simulated results − predicted results兲2
number of runs − 共number of parameters + 1兲
共13兲

Fig. 14. Plot of standardized residuals against predicted values 共de- Fig. 15. Plot of standardized residuals against predicted values 共de-
pendent variable: log关␴ + 1兴兲 determined via ANOVA pendent variable: log关␴ + 1兴兲 from the regression model, Eq. 共12兲

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 647

 8 Conclusions
7

Simulated log[σ+1]
6 The Buckingham pi theorem applied here helped to identify eight
5 dimensionless groups which are important in describing the dis-
persion of aerosols in a ventilated space. The application of clas-
4
sical experimental design methods allowed the development of an
3
efficient design plan to explore the relationship and influence of
2 these groups. Fractional factorial design approach was used to
1 identify the parameter values for detailed numerical simulations.
0 A comprehensive and validated CFD model was applied to con-
0 1 2 3 4 5 6 7 8
(a) Predicted log[σ+1] duct the simulations as planned out.
From the simulations two dimensionless groups 共the ratio of
8 the length of the room to the particle diameter and the ratio of
7 particle tracking time to residence time兲 were identified that did
not significantly influence the temporal change of the particle
Simulated log[σ+1]

6
5
4
3 as well since all interactions proved to be significant. Hence, all
2 the identified dimensionless groups will be considered for future
1 analysis. The use of two-level design which is based on assigning
0
0 1 2 3 4 5
(b) Predicted log[σ+1]
Fig. 16. 共a兲 Plot of predicted log关␴ + 1兴 关applying Eq. 共12兲兴 against
the simulation results from the 32 CFD cases generated through fac-
torial design used to develop the linear model; 共b兲 plot of predicted
log关␴ + 1兴 关applying Eq. 共12兲兴 against the simulation results from the
32 CFD cases not used to develop the linear model 共validation data
set兲
Based on the standard error it is evident that the model behaves
much better in predicting the data set it was developed from but
not as well for another data set. Similarly Eq. 共12兲 was also ap-
plied to predict log关␴兴 for the same 32 independent simulations.
The average and SD of 共simulated− predicted兲 was ⫺1.94 and
1.504, respectively. The standard error for the independent data
set was 3.35 while for the fractional factorial design data set it
was 0.127. In this case as well, the linear regression model, Eq.
共12兲, is predicting the data from which it was developed well but
not doing as well for a separate and independent data set.
Figs. 16共a and b兲 are plots of the predicted results applying Eq.
共12兲 against the simulated results from the fractional factorial
design data set and the separate validation data set. This provides
a visual assessment of model effectiveness in making predictions.
If the model behaved well most of the points would be close to
the 45° line. Fig. 16共a兲 shows that the linear model predicts the
data set from which it was developed very well. In Fig. 16共b兲
there is more scatter about the 45° line.
A multizone airflow simulation program, CONTAM 共Walton
and Dols 2005兲 was used to simulate nine cases. The CONTAM
predictions were compared to the CFD results and to the predic-
tions from the regression model, Eq. 共11兲. Information about the
specific location of the particles cannot be obtained from CON-
TAM, hence comparison with Eq. 共12兲 was not possible. The sum
of square of errors 共SSEs兲 between simulated results from CFD
and CONTAM predictions was 298. While the SSE between the
simulated results from CFD and predictions from Eq. 共11兲 was
196 indicating that the linear model showed marked improve-
ment.
number in the room. However they influenced the spatial distri-
bution of the particles within the room at the end of the simula-
tion. The dimensionless groups cannot be treated independently
maximum and minimum values to each factor implies the as-
6 7 8 sumption of linearity in the factor effects.
Forward step-wise regression method was applied to develop
the linear regression models for predicting the normalized number
of particles remaining in a room and for determining their spatial
dispersion. Thirty two more independent scenarios were simu-
lated by choosing points within the design space and the simu-
lated data compared with the predicted data from the developed
linear regression models. The results showed that the linear model
predicted the data used in the fractional factorial design very well
but was not as accurate at predicting an independent data set.
A step by step method has been described in this work using
CFD toward the development of a robust model for predicting the
number of aerosols and their distribution pattern in a room. The
method outlined here shows how to apply the detailed results of a
time consuming but accurate CFD model in a systematic manner
for the development of more user friendly solutions. Dimensional
analysis provided a convenient way of reducing a multidimen-
sional problem into controllable proportions. The work can be
expanded by considering more variables or incorporating more
complexities or phenomena into the model.
Acknowledgments
The work described in this paper was supported by the L.D. Betz
endowment.
Notation
The following symbols are used in this paper:
A 共m2兲 ⫽ area of the inlet and outlet;
Cc ⫽ Cunningham correction factor;
CD ⫽ drag coefficient;
d p 共m兲 ⫽ particle diameter;
FD 共N兲 ⫽ drag force;
FG 共N兲 ⫽ gravitational force;
Hr 共m兲 ⫽ height of the room;
I 共m兲 ⫽ vertical distance between the inlet and outlet;
L 共m兲 ⫽ length of the room;
m 共kg兲 ⫽ particle mass;

648 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010

 No 共particles/ m3兲 110共2兲, 440–456.
⫽ initial number of organisms Hinds, W. C. 共1999兲. Aerosol technology, properties, behavior and mea-
per unit volume; surement of airborne particles, 2nd Ed., Wiley, New York.
N 共particles/ m3兲 Kowalski, W., Bahnfleth, W. P., and Rosenberger, J. L. 共2003兲. “Dimen-
⫽ final number of organisms sional analysis of UVGI air disinfection systems.” HVAC&R Res.,
per unit volume; 9共3兲, 347–362.
Lu, W., and Howarth, T. A. 共1996兲. “Numerical analysis of indoor aerosol
Q 共m3 / s兲 ⫽ volumetric flow rate;
particle deposition and distribution in two-zone ventilation system.”
U 共m/s兲 ⫽ inlet air velocity;
Build. Environ., 31共1兲, 41–50.
u p 共m/s兲 ⫽ particle velocity; Lu, W., Howarth, T. A., Adam, N., and Riffat, S. B. 共1996兲. “Modelling
V 共m3兲 ⫽ volume of the room; and measurement of airflow and aerosol particle distribution in a ven-
W 共m兲 ⫽ width of the room; tilated two-zone chamber.” Build. Environ., 31共5兲, 417–423.
xi 共m兲 ⫽ particle location; MathWorks. 共2007兲. MATLAB, version 7.5.0, Mathworks Inc., Natick,
␪ 共s兲 ⫽ residence time of the room; Mass.
␮ 共N / m2 · s兲 ⫽ fluid viscosity; Montgomery, D. C. 共2005兲. Design and analysis of experiments, 6th Ed.,
␳ f 共kg/ m3兲 ⫽ fluid density; Wiley, Hoboken, N.J.
␳ p 共kg/ m3兲 ⫽ particle density; Musser, A., and Bahnfleth, W. P. 共2001兲. “Parametric study of charging
␴ ⫽ SD; and inlet diffuser performance in stratified chilled water storage tanks with
␶ 共s兲 ⫽ time particles are tracked. radial diffusers: Part 2—Dimensional analysis, parametric simulations
and simplified model development.” HVAC&R Res., 7共1兲, 51–65.
Nazaroff, W. W., and Alvarez-Cohen, L. 共2002兲. Environmental engineer-
ing science, Wiley, New York.
References
Piomelli, U. 共1999兲. “Large-eddy simulation: Achievements and chal-
lenges.” Prog. Aerosp. Sci., 35, 335–362.
Ankenman, B. E. 共1999兲. “Design of experiments with two and four level Posner, J. D., Buchanan, C. R., and Dunn-Rankin, D. 共2003兲. “Measure-
factors.” J. Quality Technol., 31共4兲, 363–375. ment and prediction of indoor air flow in a model room.” Energy
Beghein, C., Jiang, Y., and Chen, Q. Y. 共2005兲. “Using large eddy simu- Build., 35, 515–526.
lation to study particle motions in a room.” Indoor Air, 15, 281–290. Sáez, P. B., and Rittman, B. E. 共1988兲. “Improved pseudoanalytical solu-
Bingham, T. C. 共1997兲. “An approach to developing multi-level fractional tion for steady-state biofilm kinetics.” Biotechnol. Bioeng., 32, 379–
factorial design.” J. Quality Technol., 29, 370–380. 385.
Box, G. E. P., Hunter, W. G., and Hunter, J. S. 共1978兲. Statistics for Song, J., Bahnfleth, W., and Cimbala, J. 共2004兲. “Parametric study of
experimenters: An introduction to design, data analysis, and model single-pipe diffusers in stratified chilled water storage tanks.”
building, 1st Ed., Wiley, New York. HVAC&R Res., 10共3兲, 345–365.
Buckingham, E. 共1915兲. “Model experiments and the forms of empirical Tian, Z. F., and Tu, J. Y. 共2007兲. “CFD studies of indoor airflow and
equations.” Trans. ASME, 37, 263–296. contaminant particle transportation.” Particul. Sci. Technol., 25, 555–
Chang, T.-J., Hseih, Y.-F., and Kao, H.-M. 共2006兲. “Numerical investiga- 570.
tion of airflow pattern and particulate matter transport in naturally Tian, Z. F., Yu, J. Y., Yeoh, G. H., and Yuen, R. K. K. 共2007兲. “Numerical
ventilated multi-room buildings.” Indoor Air, 16, 136–152. studies of indoor airflow and particle dispersion by large eddy simu-
Devore, L. J., 共2004兲. Probability and statistics for engineering and the lation.” Build. Environ., 42, 3483–3492.
sciences, Thomson Brooks/Cole, Calif. Walton, G. N., and Dols, W. S. 共2005兲. “CONTAM 2.4 user guide and
ESI, R. D. 共2004兲. CFD-ACE+, version 2007.2.25, ESI R&D Inc., Hunts- program documentation.” NISTIR 7251, NIST, Gaithersburg, Md.
ville, Ala. Webb, G. F. 共2003兲. “A silent bomb: The risk of anthrax as a weapon of
Gao, N. P., and Niu, J. L. 共2007兲. “Modeling particle dispersion and mass destruction.” Proc. Natl. Acad. Sci. U.S.A., 100共8兲, 4355.
deposition in indoor environments.” Atmos. Environ., 41, 3862–3876. Zhao, B., Zhang, Y., Li, X., Yang, X., and Hunag, D. 共2004兲. “Compari-
Hand, D. W., Crittenden, J. C., and Thacker, W. E. 共1984兲. “Simplified son of indoor aerosol particle concentration and deposition in different
models for design of fixed-bed adsorption systems.” J. Environ. Eng., ventilated rooms by numerical method.” Build. Environ., 39, 1–8.

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 649

Copyright of Journal of Environmental Engineering is the property of American Society of Civil Engineers and
its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.