Building and Environment 82 (2014) 180e189

Contents lists available at ScienceDirect

Building and Environment

journal homepage: www.elsevier.com/locate/buildenv

The effect of particle resuspension during walking activity to PM10

mass and number concentrations in an indoor microenvironment
Norbert Serfozo a, Soa Eirini Chatoutsidou a, b, *, Mihalis Lazaridis a
a
Department of Environmental Engineering, Technical University of Crete, Polytechneioupolis, 73100 Chania, Greece
b
Norwegian Institute for Air Research (NILU), Instituttveien 18, 2007 Kjeller, Norway

a r t i c l e i n f o a b s t r a c t

Article history: Mass and number concentration of particulate matter (PM10) was measured during walking experiments
Received 6 June 2014 inside a laboratory. In total 20 walking experiments were conducted in a period of 3 months. Particle
Received in revised form number concentration was measured in the size range of 0.02e1 mm, while, mass concentration was
29 July 2014
measured in the range 0.1e10 mm. The PM10 mass increased during the activity, by average 84%, whereas,
Accepted 14 August 2014
Available online 23 August 2014
the walking activity had no impact on particle number concentration. Human induced resuspension rates
were examined under different dust loadings. The different dust loadings used (25, 15, 5, 1 g/m2) in order
to evaluate the impact of surface loading on the indoor PM10 mass concentration and on the resus-
Keywords:
Indoor PM
pension rate. Walking style was the same in all experiments. Moreover, the experiments involved two
Indoor particle emissions different walking patterns inside the laboratory (rectangular and line). The impact of different speed was
Resuspension rate also examined. The average resuspension rate was calculated to be equal to 10
2e10
3 hr
1. No impact
Surface loading on resuspension rate was observed for different walking patterns or walking speed. On the other hand,
the measured mass concentration inside the room was increased when using higher dust loading on the
oor, although the estimated resuspension rate found independent on the initially deposited surface dust
loading.
2014 Elsevier Ltd. All rights reserved.

1. Introduction Characterization of indoor particulate matter in respect to

The indoor air quality has very important implications to human
health since people spend most of their time indoors (80e90%) [1].
Human exposure to ambient particulate matter (PM) can cause
serious health problems on the respiratory and cardiovascular
system or even mortality [2e4]. Many studies have focused on
determining the impact of particle exposure to human health
[5e8]. The results suggest that PM health effects depend both on
exposure time and exposure concentration [3]. Particles generated
by indoor sources found to have the dominant role on human
exposure [8], however, signicant correlation found between the
outdoor originated particulate matter indoors [5]. The particle size
is also highlighted as a major issue in risk assessment of human
exposure [9]. Hence, the determination of indoor sources charac-
teristics and origin of indoor particles are necessary in order to
investigate the impact of indoor PM to everyday life.
* Corresponding author. Norwegian Institute for Air Research (NILU), Insti-
tuttveien 18, 2007 Kjeller, Norway. Tel.: 47 6389 8167.
E-mail address: sic@nilu.no (S.E. Chatoutsidou).
chemical composition depends on the precursor material or source.
Personal exposure and health effects is highly associated with
chemical composition of atmospheric particles [10,11]. Elemental
carbon, heavy metals and organic compounds are usual compo-
nents of the suspended particulate matter [12e14]. Moreover, dust
particles can be easily transported, thus, particulate matter char-
acteristics and chemical composition depends signicantly on the
origin region [15,13,16]. Hence, the knowledge of the origin of
particles and the associated chemical composition is a crucial issue
both on in terms of indoor air quality and human exposure.
Airborne particles in the indoor environment can be derived
either from outdoors or from indoor sources [17e19]. The indoor
concentration of particulate matter is arising from ambient parti-
cles that penetrate indoors [20,21] and by emission of particles
originating from indoor activities. Common activities such as
printing, painting, cooking, burning wood, burning candles and
smoking can generate new particles in the indoor environment
[22e29]. However, particles in the indoor air are not only the result
of newly formed particles, but arising also from the resuspension of
the already existing particles, such as settled dust on indoor sur-
faces. Any kind of human activity in indoor microenvironments can

http://dx.doi.org/10.1016/j.buildenv.2014.08.017
0360-1323/ 2014 Elsevier Ltd. All rights reserved.
 N. Serfozo et al. / Building and Environment 82 (2014) 180e189 181

result in particle resuspension [30e32]. The most common activ-
ities that resuspend particles indoors are vacuuming and walking
[33e38].
Particle resuspension from indoor surfaces is a crucial issue that
affects indoor particle dynamics. Many studies investigated particle
resuspension experimentally or using theoretical models [32e50].
The experiments were conducted in controlled conditions using
chamber or wind tunnels [34,36,37,41e43], whereas, several real
environments were used to study the impact of resuspension ac-
tivities to indoor particle concentration [32,33,35e37]. Hence,
particle resuspension was investigated throughout two aspects: the
inuence of human activities and when a uid ow acts on the
deposit particles. Considering the rst case, it was found that par-
ticle resuspension depends on many factors such as the amount of
particle size, number of present people, the type of the oor, the
walking style, the type of the shoes, bottom roughness, walking
velocity and foot size [34,36,37,44]. Two types of mechanical forces
are proposed, foot tapping on the oor and displacement due to
foot penetration [38,44,45,49] Moreover, indoor conditions were
examined (inltration system, presence of fan, relative humidity) in
regard to their impact on resuspension, but still uncertainties exist
on how they are affecting the resuspension rate of indoor particles
[34,37]. However, insight to resuspension in relation to particle
dynamics and interactions between external and intermolecular
forces proposed by several models [46e49]. Reeks and Hall [46]
considered a kinetic approach of particle resuspension where the
bound particle oscillates about a pivot on the surface under the
inuence of the drag and lift force of the uid ow. Lazaridis et al.
[47] adopted the same kinetic approach but modeling the parti-
cleesurface interaction using the Lennard-Jones intermolecular
potential. Guingo and Minier [48] and Goldasateh et al. [49] used
Monte Carlo computations to simulate surface roughness and dis-
tribution of the adhesion forces. The proposed models share com-
mon features such as the stochastic description of particle
resuspension, the strong dependence on surface roughness and the
importance of interaction between the deposit particles and the
surface. It was found that high surface roughness corresponds to
easier detachment [52] due to a signicant reduction in adhesion
forces [47,53]. Moreover, particle size plays an important role by
enhancing resuspension, since the removal forces (drag and lift)
depend strongly on it [54]. It was also proposed [46] that resus-
pension depends on time with high resuspension rates at a short-
term regime followed by a long-term regime with reduced rates.
The time dependence observed also experimentally [35,42].
The aim of this work was to evaluate the particle resuspension
rate induced by walking inside an indoor microenvironment using
continuous measurements of indoor particle mass/number con-
centration. The main objectives were to determine the impact of
different dust loadings on the oor, the indoor particle concentra-
tion and the resuspension rate and to investigate the impact of
different walking patterns, style and walking speed to the resus-
pension rate.
The resuspension measurements were performed in a paper
pool (4.3 m  1.4 m and 1.1 m high) built in the laboratory. During
the resuspension measurements, the instruments were placed at
the farther end of the paper pool with inlets 1.38 0.03 m above
the ground. The selected height of the instruments (1.38 m) cor-
responds to the breathing zone of the person performing the
walking activity. The instruments used for the measurements were
Dust Trak, Dust Trak II, P-Trak, SMPS C, NanoScan SMPS, OPS, IAQ
and Tiny Tag data loggers (see Fig. 1). A detailed description of the
instrumental setup is presented in section 2.1.2. The emission
period coincided with the walking activity (rectangle and line paths
with different dust loadings) and lasted for 20 min. The duration of
the background measurement was 60 min, which considered suf-
cient enough for the evaluation of the background period. No
signicant temporal variations of the background concentration
recorded in all experiments.
During the experiments only one person was present in the
laboratory, with the window and the door closed all times. To
determine the reproductivity, all 8 variations (combination of
rectangle and line paths with different dust loadings e 25, 15, 5
and 1 g/m2) of resuspension experiments were performed twice.
Additional measurements were conducted with increased and
decreased walking speed in order to investigate the effect of
walking speed on resuspension rate. Two experiments were per-
formed with 20% higher walking speed, whereas, two experiments
performed with 20% lower walking speed. Thus, four additional
experiments with different walking speed. The experiments were
conducted following a line path and only in the case 5 g/m2 and
1 g/m2 of dust loading. The increased or decreased speed was
obtained based on the average walking speed on each set of
measurements (5 g/m2 and 1 g/m2 in the present case). The ma-
terials used to build the paper pool were a simple plotter paper
(0.94 m width, r 80 g/m2) and a masking paper tape. The
footwear used during the walking were a classic long stripe cotton

2. Materials and methods

2.1. Experimental setup

2.1.1. Laboratory setup

The measurements were carried out in the period between June
to September 2013 in a laboratory at the Technical University of
Crete. The area of the laboratory is 18.5 m2 of rectangle shape and
its volume is 53.7 m3 with one door and one window placed
opposite of the door. The effective volume was estimated to be
77.3% of the total volume, thus 41.5 m3. The effective surface area
was equal to 6 m2 (4.3 m  1.4 m). Fig. 1. Scheme of the laboratory and the instrumentation around the paper pool.
182 N. Serfozo et al. / Building and Environment 82 (2014) 180e189
socks, 3 pairs (75% cotton, 25% polyester) which have been washed
after every single use. A simple paper respirator mask was used
during the walking as well.
2.1.2. Instrumentation
Several instruments were used in order to measure indoor
particle concentration. More precisely a Dust Trak (Aerosol
Monitor, Model 8520, TSI) and a Dust Trak II (Aerosol Monitor,
Model 8532, TSI) instruments were used for mass concentration of
PM10. A P-Trak (Ultrane Particle Counter, Model 8525, TSI) in-
strument was used for the determination of the number concen-
tration characteristics. The Dust Trak and Dust Trak II instruments
are based on a 90 light scattering and have a particle size range
0.1e10 mm at a ow rate 1.7 lpm with a concentration range
1e100,000 mg/m3 for Dust Trak and 3.0 lpm with a concentration
range 1e150,000 mg/m3 for Dust Trak II, respectively. The log in-
terval was setup to 1 min. The P-Trak uses high-purity isopropyl
alcohol to grow microscopic particles for easier detection and
counting in the optical chamber. Its concentration range is be-
tween 0 e 5  105 particles/cm3 and it is able to measure particle
concentration in the size range 0.02e1 mm at a sample air ow
rate of 0.1 lpm.
In addition, resuspension measurements were performed by a
SMPS C (CPC Model 5.403 and L-DMA e Vienna type, GRIMM),
NanoScan SMPS (Model 3910, TSI) and OPS (Model 3330, TSI) for the
determination of the number size distributions. The SMPS C system
measures particle number concentration up to 107 particles/cm3 in
size range 10e1100 nm at sample air ow rate 0.3 lpm and a sheath air
ow rate 3.0 lpm taking sample every 7 min and 34 s. The principle of
operation is based in the use of bipolar charge of the sample air in the
DMA where the negatively charged particles drift down the positively
charged inner electrode to the slit and are drawn to the CPC where the
particles pass through a butanol vapor in the saturator chamber
(35  C), cool down in the condensation unit (10  C) to grow the par-
ticles for easier counting and afterwards they are detected and coun-
ted by 90 light scattering.
Unlike the previous, the portable NanoScan 3910 is based on an
unipolar charging of the particles where the particles are posi-
tively charged in a mixing chamber and sent to radial DMA
(RDMA) for size classication and then are counted in
isopropanol-based CPC. It measures particle concentrations from
102 to 106 particles/cm3 in the size range of approximately
10e420 nm at a scan time of 60 s. The OPS instrument is based on
120 light scatter and lter sampling and it is able to count mass
concentrations 0.001e275,000 mg/m3 and particle concentrations
up to 3000 particles/cm3 in the size range 0.3e10 mm. It operates
with up to 16 user-adjustable channels at a sample air ow rate
and sheath air ow rate 1.0 lpm.
In order to measure the CO2 levels indoors/outdoors of the
laboratory for the determination of the air exchange rate, the IAQ
instrument (Indoor Air Quality Meter, Model 8762, TSI) was used
which measures the CO2 level in parts per million (ppm). The log
interval was setup to 1 min.
The air temperature and relative humidity during the resus-
pension measurements were recorded by four Tiny Tag data log-
gers placed in the corners of the laboratory. The average air
temperature in the laboratory during the measurements period
was recorded to be 29.62 0.95  C and the relative humidity
45.03 3.83%.
Nonetheless, the PM10 mass concentration data obtained by
Dust Trak were corrected based on work of Chalvatzaki et al. [55]
where comparative measurements of PM10 concentrations by
the Beta attenuation monitor (FH 62 SEQ) and the Dust Trak
were performed at an urban background site close to the labo-
ratory for a period of 12 months. The Dust Trak instrument
measurements were then corrected based on these comparative
measurements. The correction equation can be expressed as:
h . i h . i
Beta conc: mg m3 0:80  Dust Trak conc: mg m3 10:4
2.1.3. Measurements protocol
At the beginning, two testing measurements were conducted
using 25 g/m2 dust on the oor of the paper pool, one in a rectangle
and another in a line path for a walking period of 30 min. The other
experiments were conducted using a walking period of 20 min.
Each resuspension experiment (20 experiments in total) lasted 5 h
and 30 min and were completed in 3 stages:
Stage 1: Empty laboratory: 0e60 min (background measurement)
Stage 2: Walking period: 60e80 min (resuspension e emission
period)
Stage 3: Post-walking period: 80e330 min (empty laboratory e
particles removal)
The 20 min of the emission period were used to determine the
resuspension rate, since it was the only period of activity inside the
laboratory room. However, only the last 10 min of background period
and the rst 210 min of the removal period used to evaluate the
background and the removal period respectively. Hence, 4 h in total for
each experiment. The background period was evaluated only at the
last 10 min, since, no signicant temporal uctuations of the indoor
concentration observed during the pre-activity period. The removal
period was evaluated only for the rst 210 min because the indoor
concentration maintained similar values after the chosen period.
During stages 1 and 3 no one was present in the room and all the win
dows and doors were closed. During the stage 2 only one person was
present in the room with the door and window closed. Two walking
paths were used for the experiments. Path A in which the person was
walking in a line path 4.3 m long in the middle of the paper pool and
path B in which the person was walking in a rectangle shaped path of
8.6 m long around the edge of the paper pool in a clockwise direction.
2.1.4. Dust analysis
The dust used in all the experiments was collected with a scoop
into a 10 L container outside the laboratory window from the side-
walk on 05/06/2013. It is the same dust that could possible get inside
the indoor environment during air exchange while opening the
window or by penetration. The collected dust was then sieved on a
laboratory sieve with an aperture of 53 mm and stored in an air-tight
box at a room temperature in the laboratory. With the same sieve the
dust was spread manually on the oor of the paper pool with
extreme care in order to maintain a uniform layer. The dust was
spread at 0.15 0.05 m height above the ground at the most 4e5 h
before the measurements. Prior to new measurement, the paper
with a dust loading at the bottom of the paper pool was overlayed
twice from the top with a new paper (during the experiments with
25 and 15 g/m2 of dust loading the bottom was overlayed four times
with a new paper in order not to affect the next measurements) and
the walls were cleaned with a piece of semi-wet cloth.
Heavy metals analysis (Na, Mg, Al, Si, K, Ca, Fe, Ti, V, Cr, Mn, Ni, Cu,
Zn, As, Se, Cd, Ba, Hg, Pb) was performed to determine the metal
content of the dust used by Inductively coupled mass spectrometry
(ICP-MS 7500cx coupled with Autosampler Series 3000, Agilent
Technologies). The collision (He) and reaction modes of the instru-
ment were used during the analysis procedure in order to eliminate
possible isotopes interferences (e.g. interference of 40Ar12C in the 52Cr
analysis). The results were calibrated using an external calibration
curve (at least 5 levels of concentration, with a correlation coefcient
r 0.99 for each element) and the corresponding Merck standards.
 N. Serfozo et al. / Building and Environment 82 (2014) 180e189 183

Although, the results did not reveal any signicant levels of major During the resuspension experiments, the total surface area of
crustal elements (Al, Si, Fe, Ca, K, Mg, Ti) which were all below 0.62% the laboratory was used. In this case, Ar can be substituted with A,
with exception of Calcium where the analysis showed a high con- and a simplied version of the system of equations (4) and (5)
centration equal to 14.94%. On the other hand, trace elements (Zn, Pb, yields:
Cu, Ni, Cd, Cr) and other element (Na, V, Mn, As, Se, Ba, Hg) concen-
trations were not higher than 0.01%, except Sodium (0.81%). Miner- dCin  
V PaVCout rAL
a k VCin (6)
alogy, particle size distribution and chemical composition of the soil dt
samples were studied jointly with the nutrient content of soil solution
from study areas of Crete including four types of parent rocks with dL
A
rAL kVCin (7)
different geological origin, from acidic to ultramac [56]. Regarding dt
the current study, no interrelation between the chemical composition The equation (6) describes the change in mass concentration
of used dust and the Cretan soil dust was found, resulting in conclusion indoors and equation (7) is describing the change in particle mass
that the used dust is not of natural soil origin. More likely, it is concentration on the oor. The resuspension rate r is estimated
attributed to road dust because soil particles as well as anthropogenic by equation (6) using a forward difference approximation, while,
metals like Pb and Cr are included in the chemical prole [57,16]. equation (7) can be solved analytically. Thus, the resuspension
Furthermore, the high percentage of Ca in the used dust implies the rate and the particle surface loading for each time step are given
inuence of construction activities in the area. Sulphates related to by:
this factor are likely to originate from sulfate salts such as CaSO4 and

Al2(SO4)3 which are typical cement components [16].   V Cin t dt
Cin t    
r t dt k a Cin t
ALt dt

2.2. Model description  

aPCout t (8)
2.2.1. Resuspension rate model
A resuspension model was used for the analysis of the resus-
  kVCin t    
pension experiments and the calculation of the resuspension rate. L t 1
e
rt L 0 e
rt (9)
The resuspension of an indoor aerosol is usually dened by the rA
resuspension rate r (min
1), which is the fraction of particles The requirements for solving the above system are the initial
removed from the surface per unit of time [58]: values of mass loading L(0) and the change in concentration C(t)
with time. Given the initial mass loading L(0) and the change of
R aerosol concentration, r can be estimated for time step one using
r (1)
L equation (8). Since r is known, equation (9) is then used to estimate
the surface loading at the rst time step, L(1). The same procedure
where R is the resuspension ux (mg/m2min) and L (mg/m2) is the
is followed for every time step.
particle surface concentration. Considering a room of well-mixed
Moreover, the particle concentration in the indoor air can be
air volume, the particle concentration can be described by a dy-
predicted from equation (6) [35]:
namic mass balance model [19]:
  A      
dVCin Cin t dt r L t dt Cin t 1
a kdt aPCout t dt
PaVCout S
aVCin
kVCin (2) V
dt
(10)
where Cin is the particle concentration inside the laboratory (mg/ The rst term on the right side of the equation (10) is the
m3), Cout is the particle concentration outside the laboratory (mg/ resuspension contribution to indoor air concentration, the second
m3), V is the effective volume of the laboratory (m3), P is the term refers to the reduction of indoor concentration by deposition
penetration efciency, a is the air exchange rate (min
1), k is the and air exchange rate and nally the third term is the penetration of
deposition rate (min
1), S is the emission of the particles indoors outdoors origin particles indoors.
(mg/min) and t is the time (min
1). In particular, in respect to par-
ticle emissions by resuspension e S, the emission rate can be
written as: 2.2.2. Inltration rate
The particle indoor dynamic model for a well-mixed air volume
S rAr L (3) is given by equation (6). The inltration rate of outdoor origin
particles can be estimated using equation (6) for a period without
where r is the resuspension rate (min
1), Ar is the oor surface area any activity inside the laboratory. Thus, equation (6) is written as:
used for resuspension (m2) and L is the oor loading (mg/m2). Thus,
equation (2) can be rewritten as follows: dCin  
V PaVCout
a k VCin (11)
dt
dCin  
V PaVCout rAr L
a k VCin (4) The analytical solution of the above equation is:
dt

Introducing a mass balance model on the surface, the change in   aPCout   aPCout
akt
Cin t Cin 0
e (12)
aerosol mass concentration on the oor can be written as [34]: ak ak

dL The right side of equation (12) consists of two terms. The rst
A
rAr L kVCin (5) term is the steady state concentration of the particles inside the
dt
laboratory which can be dened as the fraction of particles that
where A is the total surface area of the room (m2). Thus, the change penetrates from outdoors and remain suspended indoors. The
in particle concentration indoors is described by a set of two second term represents the total losses of the indoor aerosol due to
equations, namely equations (4) and (5). deposition and air exchange from indoors to outdoors.
184 N. Serfozo et al. / Building and Environment 82 (2014) 180e189

The inltration rate was calculated based on equation (12) after concentration was 2037 particles/cm3, whereas, average concen-
the end of the activity, where, the particle concentration is expo- tration during the activity was 2110 particles/cm3. The same char-
nentially decreasing with time. Due to lack of outdoor data, the acteristic found in all experiments. Thus, the three periods
inltration was estimated as the fraction of particles that remain (background, walking period, post-walking period) were strongly
suspended inside the laboratory (the steady state particle concen- correlated only with the indoor particle mass concentration.
tration or inltration concentration Cinf): Particle resuspension from indoor surfaces depends on particle
size [42,59]. Larger particles detach more easily from the surface
aPCout [54,59]. Previous studies verify that small particles (<1 mm) are not
Cinf (13)
ak easily resuspended or not resuspended at all [30,43,47]. Moreover,
studies on resuspension by human walking concluded that resus-
pension rate increases with particle size [34,35,37].
2.2.3. Air exchange rate
Our results indicate that in terms of mass concentration the
The air exchange rate was estimated based on the exponential
walking activity inuences strongly the PM10 indoor particle con-
decay of CO2 inside the laboratory. The concentration of CO2 after
centration. On the other hand, in terms of number concentration
the emission period is described by the following equation:
the indoor particles are not affected by the activity in the laboratory
dCCO2 room inside the measured particle size range. Considering the
aCCO2 (14) different size range used for the mass and number concentration,
dt
the present results imply that the resuspension of indoor particles
where a is air exchange rate (hr
1). Thus, CO2 concentration at any takes place at particle sizes higher than 1 mm. It is likely that par-
time inside the laboratory is given by: ticles at size that correspond to the measured number concentra-
tion (0.02e1 mm) is not resuspend at all. Since, the measured size
   
CCO2 t CCO2 0 e
at (15) range of the mass concentration corresponds to higher particle
sizes (0.1e10 mm), it is believed that the resuspended particulate
The average air exchange of the laboratory calculated to be matter lies in the size range of 1e10 mm.
0.16 hr
1 during the experiments. Table 1 summarizes the average PM10 concentration for back-
ground and activity periods during all experiments. Indoor PM10
3. Results and discussion mass concentration was increased by average 84% during the ac-
tivity period. This is indicating a considerable change in indoor
3.1. Indoor particle concentration mass concentration inside the laboratory. Indoor mass concentra-
tion is increased by one or two orders of magnitude in a period of
Indoor particle number and mass concentration was measured few minutes (20 min). On the other hand, particle number con-
simultaneously during the walking experiment. The number con- centration maintained values close to the background level in all
centration was measured in the size range of 0.02e1 mm, whereas,
the measured mass concentration corresponded to the size range of
0.1e10 mm. Table 1
Fig. 2 presents the indoor mass and number concentration Comparison between average background concentration and average walking
concentration for different dust loadings (R-rectangular, L-line path). The average
versus time during the experiment on 04/07/2013. The mass con- emission rate for each experiment is also shown.
centration is inuenced from the beginning of the activity with an
increase in the indoor concentration which is followed by a Experiment-walking PM10 average PM10 average % Increase Average
prole (L-line path, background walking period emission rate,
decreasing period after the end of the walking activity. The back- R-rectangular path) concentration, concentration, mg/min
ground average indoor mass concentration was 73 mg/m3, while, mg/m3 mg/m3
during the walking activity the average mass concentration was
Dust loading, 25 g/m2
3320 mg/m3 indicating an increase of 98%. On the other hand, the 04/07-R 73 3320 98 17.3
number concentration maintained similar concentration levels 10/07-L 58 3205 98 14.6
before and during the walking activity. Average background 26/07-R 39 3839 99 17.6
30/07-L 32 3463 99 19.4
Average 51 3457 99 17.2
Dust loading, 15 g/m2
21/06-R 47 1433 97 6.5
26/06-L 36 1579 98 7.1
08/08-R 42 1690 98 7.8
10/08-L 35 1716 98 9.2
Average 40 1605 98 7.7
Dust loading, 5 g/m2
02/07-R 38 462 92 2.0
28/06-L 42 573 93 2.4
12/08-R 41 556 93 3.5
13/08-L 49 642 92 2.9
10/09-L 34 680 95 2.8
11/09-L 39 573 93 4.1
Average 41 581 93 3.0
Dust loading, 1 g/m2
12/07-R 47 215 78 0.9
24/07-L 27 325 92 2.2
15/08-R 52 259 80 1.2
14/08-L 50 240 79 1.4
12/09-L 35 381 91 2.6
13/09-L 24 156 85 1.0
Fig. 2. Mass and number concentration versus time during walking in rectangle shape
Average 39 263 84 1.6
inside the paper pool with dust loading 25 g/m2 (R2 on 04/07/2013).
 N. Serfozo et al. / Building and Environment 82 (2014) 180e189
experiments as indicated in Fig. 2. Note that the % increase of PM10
mass concentration was below 90% only in 4 experiments (12/07,
15/08, 14/08, 13/09) highlighting the strong inuence of the activity
to the indoor particle mass concentration.
No dependence of indoor PM10 concentration found in respect to
the two different walking proles followed during the activity.
Walking in a rectangular manner or in line inside the laboratory had
no effect on indoor aerosol concentration and therefore on the
resuspension rate. Qian and Ferro [34] examined the effect of different
walking paces and weight of the person while walking inside a
chamber and found that the main contribution in resuspension rate is
walking style. In the present study walking style, time of walking, dust
type, footwear and walking person were the same in all experiments
and the only variable during the activity was the walking pattern
which had no effect on the indoor particle concentration.
3.2. Indoor PM10 mass concentration at different dust loadings
Fig. 3 presents the PM10 indoor mass concentration versus time
for different dust loadings. One experiment from each dust loading
(thus 4 in total) was chosen in order to compare the indoor mass
concentration and examine the effect of the dust loading to the
measured PM10 mass concentration. All chosen experiments cor-
responded to walking in a line path in order to isolate the impact of
using different dust loadings on the oor.
Higher surface loading is leading to higher indoor mass con-
centration. Increased concentration on the oor corresponds to
higher initial dust loading, which, can be potentially removed from
the surface. In fact, calculating the emission rate S from equation (3)
for each surface loading reveals that the emission rate increases
with increasing the initial dust loading on the oor (Table 1). The
average emission rates for 25, 15, 5 and 1 gr/m2 dust loading were
17.2, 7.7, 3.0 and 1.6 mg/min respectively. Gomes et al. [60] also
observed increased mass of resuspended particles at higher dust
loadings and Tian et al. [43] reported that higher dust emission
rates were provoked by higher surface dust loadings. Moreover,
Qian et al. [61] reported that the emission rates of particle mass due
to walking were in the range 10
2e102 mg/min, which is in
agreement with our ndings.
3.3. Estimation of inltration rate
In the absence of indoor sources, the indoor particle concen-
tration is the net result of particle penetration from outdoors,
Fig. 3. Mass concentration versus time for different dust loadings. Line path was used
in all experiments (10/07, 26/06, 28/06, 14/08).
185
particle deposition on indoor surfaces and air exchange rate of the
study room [62e64]. Thus, the total indoor particles losses include
deposition by Brownian diffusion and gravitational settling [18]
and the indoor/outdoor air exchange rate [19].
In the present study the post-walking period was evaluated
using the equation (11), which reects the indoor particle con-
centration inside the laboratory room in the case of no present
indoor source. Applying equation (12) to the experimental data
after the end of the activity and using also equation (13), the
inltration rate and total losses inside the laboratory can be esti-
mated. Fig. 4 presents a comparison between the measured PM10
mass concentration and the modeled one after the end of the
walking period. The same pattern found in all experiments. Addi-
tionally, in Table 2 are listed the inltration concentration Cinf and
total losses as they were calculated by equation (12) for each
experiment.
The correlation between the modeled indoor concentration for
the experiment R5 (Fig. 4) and the observed one is high with
R2 0.98. This is indicating that equation (12) is well-describing the
measured indoor particle concentration starting with the initial
value of 255 mg/m3 and ending at the steady state mass concen-
tration of 46 mg/m3 inside the laboratory.
High correlation between the observed and the modeled indoor
concentration found in all experiments with R2 always above 0.9
(0.91e0.99). The Cinf is independent from the initial surface dust
loading. According to equation (13) the steady state indoor con-
centration is a function of penetration of outdoor particles and total
losses inside the laboratory. Moreover, no correlation found be-
tween total losses (a k) and the initial surface loading. On average
the total losses inside the laboratory were 2.85 0.50 hr
1. The
total losses rates found in the present study are consistent with
other studies dealing with indoor particle dynamics [63,65e67].
3.4. Dust loading
The most important factor for determining the resuspension
rate is the surface dust loading for every time interval. The resus-
pension rate and surface loading are irreversibly proportional
(equations (8) and (9)) and cannot be derived independently. An
increase of indoor mass concentration is increasing the resus-
pension rate and reducing the dust loading on the oor. The L(t) can
be estimated only by knowing the resuspension rate at that specic
Fig. 4. Observed mass concentration and modeled one after the end of the activity.
Walking in rectangle shape inside the paper pool with dust loading 25 g/m2 (R5 on 12/
07/2013).
186 N. Serfozo et al. / Building and Environment 82 (2014) 180e189

Table 2 same behavior was observed in all experiments. The average vari-
Results of the modeled indoor concentrations after the end of the activity period. ance of the relative uctuations of the surface concentration was
Experiment Dust Initial Steady state Total losses, R2 found 0.013%, conrming that a very small fraction of the deposited
loading, concentration, concentration, a k, hr
1 dust on the oor is actually resuspended. These ndings are in
g/m2 C(0), mg/m3 Cinf, mg/m3 agreement with Tian et al. [43] where the authors also suggested
04/07 25 4582 95 2.60 0.97 that only a small fraction of particles is resuspended. Nevertheless,
10/07 25 4026 63 2.35 0.97 indoor PM10 concentration increases drastically with higher dust
21/06 15 2024 48 2.39 0.99
loading during the walking period resulting in higher mass con-
26/06 15 3932 48 2.91 0.97
02/07 5 658 39 2.25 0.98 centration inside the laboratory. A reasonable explanation is that
28/06 5 757 59 2.36 0.94 the higher the dust loading on the oor is, the higher is the po-
12/07 1 255 46 1.91 0.98 tential fraction of the particles that can detach from the surface and
24/07 1 286 25 2.90 0.95
consequently resuspend.
26/07 25 4226 48 3.46 0.98
30/07 25 3920 44 3.04 0.97
Particle detachment is closely related with resuspension and it
08/08 15 1997 38 3.24 0.98 depends both on the adhesive and removal forces [44,47,51,54,72].
10/08 15 1884 43 3.00 0.98 Moreover, resuspension occurs easily for larger particles
12/08 5 621 36 2.78 0.97 [33,34,36,37,43,59]. In human induced particle resuspension, par-
13/08 5 736 46 3.27 0.94
ticle detachment is caused by mechanical forces, such as foot tap-
15/08 1 274 62 3.00 0.99
14/08 1 272 48 2.91 0.95 ping [38] or displacement due to foot penetration [45]. Moreover,
10/09 5 802 31 2.65 0.98 Goldasteh et al. [49] propose that the main mechanism for particle
11/09 5 747 33 4.20 0.91 resuspension is the high speed air ow generated at the gap of the
12/09 1 444 27 2.69 0.98 shoe sole during the upward and downward motion of the foot.
13/09 1 186 26 3.13 0.97
Average 46 2.85 0.98
Considering that the walking style was the same for all experi-
ments, particle detachment happened under the same conditions.
Removal forces that are connected with these conditions were
likely in the same order of magnitude and responsible for the
moment. Giving the initial value L(0), a repeated numerical pro-
fraction of particles leaving the surface.
cedure is followed by estimating at each time step the resuspension
rate and then the instant dust loading on the oor.
Several variations of dust loading were taking place while the 3.5. Resuspension rate
experiment was performed. A typical example of surface concen-
tration variations versus time during the activity period is shown in The resuspension rate was calculated using the coupled set of
Fig. 5. The initial dust loading on the oor was estimated at 24.57 g/ equations (8) and (9). Table 3 summarizes the resuspension rate for
m2 and this value is compared with the calculated dust loading all the experiments.
derived from equation (9). The data points under the line e which On average the resuspension rate for all the experiments was
shows the initial surface concentration-indicate an increase in in- 6.61$10
3 hr
1. In general the range of the resuspension rate is
door mass concentration, whereas, the data points above the line between 10
2e10
3 hr
1. These results are in agreements with
indicate a decrease in indoor mass concentration (various temporal other studies for resuspension by walking in a chamber or in real
uctuations took place in most of the experiments). Numerically
the rst case is related with a positive number in the calculation of Table 3
Resuspension rate for different dust loadings.
resuspension rate, while, the second case is related with a negative
number. Moreover, the mass concentration on the oor shows Experiment Resuspension rate, r, hr
1
small variations. In Fig. 5 the modeled dust loading uctuations Dust loading, 25 g/m 2

varied between 24.5 and 24.6 g/m2. Even though the indoor PM10 04/07 7.15$10
3
concentration increases up to two orders of magnitude, the surface 10/07 6.64$10
3
26/07 7.31$10
3
dust loading L(t) retains values close to the initial dust loading. The
30/07 8.60$10
3
Average 7.43103
Dust loading, 15 g/m2
21/06 4.46$10
3
26/06 5.05$10
3
08/08 5.28$10
3
10/08 6.19$10
3
Average 5.25103
Dust loading, 5 g/m2
02/07 4.05$10
3
28/06 4.63$10
3
12/08 6.65$10
3
13/08 6.48$10
3
10/09a 5.56$10
3
11/09a 8.38$10
3
Average 5.96103
Dust loading, 1 g/m2
12/07 7.81$10
3
24/07 2.06$10
2
15/08 1.10$10
2
14/08 1.35$10
2
12/09a 2.39$10
2
13/09a 9.60$10
3
Average 44.44102
Fig. 5. Initial dust loading on the oor and modeled dust loading versus time during
the activity period, Lo 24.57 g/m2 (R2 on 04/07/2013). a
Different walking speed.
 N. Serfozo et al. / Building and Environment 82 (2014) 180e189 187

Table 4 3.5.2. Resuspension rate time dependence

Average walking speed during walking period for different dust loadings.
Dust loading, Average walking Average resuspension
g/m2 speed, m/s rate, hr
1
25 0.85 7.43$10
3
15 0.83 5.25$10
3
5 0.84 5.96$10
3
1 0.84 1.44$10
2
environment which found values in the range 10
2e10
6 hr
1
[34,35,37,43].
Table 3 also suggests that resuspension rate is independent from
the initial dust loading on the oor. The different dust loadings are
affecting the indoor PM10 concentration but not the resuspension
rate. Factors such as walking speed, weight of person, number of
persons inside the room, shoes, type of oor, style of walking
contribute different on the resuspension rate
[34,36e38,43,45,59,61]. In our analysis the different walking pat-
terns had no effect on indoor concentration or on resuspension
rate. Even though different walking speeds were examined in 4
experiments (10/09, 12/09 higher walking speed, 11/09, 13/09 lower
walking speed) no correlation found between the walking speed
and the resuspension rate. Probably, the effect of walking speed on
particle resuspension should have an effect for a wider range of
speeds. The average walking speed and resuspension rate for every
dust loading is presented in Table 4.
3.5.1. Inuence of environmental factors
The air temperature and relative humidity had no effect on
resuspension rate for the range of the environmental conditions
occurred in the laboratory. The average temperature inside the
laboratory for all the experiments was 30 1  C and the average
relative humidity was 45 4%. The relative humidity was found to
affect the adhesion force between the particle and the substrate
through capillary forces [36,54,61,68e70]. However, the inuence
of relative humidity on particle detachment becomes signicant on
wider ranges of 20%e90% RH [69,70]. Qian and Ferro [34] also
found no effect on resuspension rate for RH range 26.4%e51%,
while, Tian et al. [43] found a dependence on resuspension for two
levels (40% and 70%) of relative humidity.
The resuspension rate of indoor particles was estimated for each
time step of every experiment using the set of equations (8) and (9).
In order to nd the time dependence of resuspension rate, the
walking period was further divided. This was due to the fact that in
several experiments uctuation of indoor mass concentration was
observed, thus, the activity period divided into two stages. In stage
1, which is the burst of particle emission, the rst 4 min of the
activity were chosen due to higher rates observed at this period,
and in stage 2 the remaining time of the activity was chosen. Fig. 6
presents the resuspension rate for the two stages for all the
experiments.
Fig. 6 demonstrates a strong dependence of resuspension rate
versus time. The rst minutes of the event are characterized by a
higher resuspension rate than the remaining time. The resus-
pension rate at stage 1 is higher by 68% in comparison to stage 2.
Strongly adhered particles are more difcult to detach from the
surface. Thus, in the rst minutes of the walking activity resus-
pension occurs for particles which are less adhered on the surface.
As a result, the resuspension rate is higher on the rst minutes of
the activity and more particles detach from the surface. Similar
results were found both in cases of human induced particle
resuspension [35] or in wind tunnel studies [42,71,72]. Turbulent
air ow generated by human walking, adhesion forces and type of
deposit (monolayer or multilayer) are responsible for the uctua-
tion on resuspension rate through exposure time [59].
3.6. Indoor model performance
An evaluation of the performance of the indoor model is pre-
sented in this section. Using the equation (10) the indoor PM10
concentration can be predicted for each time step. The resus-
pension rate r and surface loading L, are already known from
equations (8) and (9). Thus, the indoor concentration of aerosol is
derived by setting the values for every time interval. The results are
presented in Fig. 7 for different surface dust loadings.
The indoor model is based on a coupled of differential equations
(6) and (7). These equations refer to a mass balance on the air and
on the oor, respectively. The physical processes that the model
takes into account is penetration of particles from outdoors, ex-
change rate of the indoor air, deposition of particles onto surfaces

Fig. 6. Comparison of resuspension rate for stage 1 and 2 for all the experiments.
188 N. Serfozo et al. / Building and Environment 82 (2014) 180e189
Fig. 7. Mass concentration of indoor PM10 during walking experiment and comparison
with model for dust loadings 25, 15, 5 and 1 g/m2. Data from Fig. 3.
and emission by resuspension. Coagulation and condensation are
assumed to have no major effect on indoor particle dynamics. The
coagulation process inuence directly the particle number distri-
bution and has minor effect in total mass concentration of particles
[18]. Moreover, it is important only in high number concentrations
and therefore, can be neglected when total number concentration
is lower than 104 cm
3 [65]. The number concentration of indoor
aerosol in all experiments was under that limit and independent of
any walking activity, supporting the assumption of negligible
coagulation during walking experiments inside the laboratory.
The deposition found to be dominant in the calculations of the
total losses rate inside the laboratory. Although, the air exchange
rate was one order of magnitude lower than the deposition rate,
nevertheless, it was signicant enough not to neglect it. The net
result from inltration of outdoor particles, air exchange of indoor
particles and deposition on the surfaces of the laboratory leaded to
a steady state concentration, Cinf, and as dened by equation (13)
represents the fraction of particles penetrating from outdoors and
remain suspended indoors. The Cinf, is directly correlated with the
background PM10 concentration inside the laboratory before the
activity period. The indoor PM10 concentration decreases after the
activity period tending to reach the background concentration.
Indoor concentration reached the background concentration
approximately 3 h after the end of the experiment (Fig. 7).
4. Conclusions
The current study focused on the estimation of resuspension
rate by human walking inside an experimental laboratory. Different
dust loadings used to cover the oor while the walking style was
the same for all the experiments.
The resuspension rate for PM10 induced by human walking
estimated in the range of 10
2e10
3 hr
1. Different dust loadings
on the oor contribute directly to the indoor mass concentration of
particles but have no effect on the resuspension rate. Although,
higher dust loading resulted in higher emission rates and is asso-
ciated with higher potential of dust particles than can detach from
the surface, the fraction of particles leaving the surface is related to
the magnitude of the removal forces. Further investigation on this
is needed in order to illustrate the nature of removal forces induced
by human walking. Moreover, it is believed that the resuspension
rate depends mainly on walking speed rather on walking pattern or
dust loading. Wider walking speed range should be examined in
order to investigate the impact of the removal forces.
The indoor mass balance model used in this study well-
predicted the indoor particle concentration in both activity and
post-activity period. However, resuspension of indoor particles is a
complicated physical process dealing with the adhesion forces of
particles on the oor and the removal forces acting upon them, thus
a simultaneous investigation on particle dynamics both on the oor
and the ambient air is necessary.
Acknowledgments
This work was supported by the European Union 7th framework
program HEXACOMM FP7/2007-2013 under grant agreement N
315760.
References
[1] Robinson J, Nelson WC. National human activity pattern survey data base. NC:
United States Environmental Protection Agency, Research Triangle Park; 1995.
[2] Morawska L, Salthammer T. Indoor environment, airborne particles and
settled dust. Willey-VCH GmbH & Co. KGaA; 2003.
[3] Pope III CA, Dockery WD. Health effects of ne particulate air pollution: lines
that connect. J Air Waste Manag 2006;56:709e42.
[4] Anderson HR. Air pollution and mortality: a history. Atmos Environ 2009;43:
142e52.
[5] Janssen NAH, de Hartog JJ, Hoek G, Brunekreef B, Lanki T, Timonen KL, et al.
Personal exposure to ne pariculate matter in eldery subjects: relation be-
tween personal, indoor, and outdoor concentrations. J Air Waste Manag Assoc
2000;50:1133e43.
[6] Samet J, Krewski D. Health effects associated with exposure to ambient air
pollution. J Toxicol Environ Health A 2007;70:227e42.
der S, Schlink U, Borte M, Diez U, et al. Respiratory
[7] Franck U, Herbarth O, Ro
effects of indoor particles in young children are size dependent. Sci Total
Environ 2011;409:1621e31.
[8] Morawska L, Asfari A, Bae GN, Buonanno G, Chao CYH, H anninen O, et al.
Review article, indoor aerosols: from personal exposure to risk assessment.
Indoor Air 2013;23:462e87.
[9] Cao AG, Yu G, Chen YS, Cao QM, Fiedler H, Deng SB, et al. Particle size: a
missing factor in risk assessment of human exposure to toxic chemicals in
settled indoor dust. Environ Int 2012;49:24e30.
[10] Poschl U. Atmospheric aerosols: composition, transformation, climate and
health effects. Angew Chem Int Ed 2005;44:7520e40.
[11] Sarigiannis DA, Karakitsios SP, Gotti A, Liakos IL, Katsoyiannis A. Exposure to
major volatile organic compounds and carbonyls in European indoor envi-
ronments and asssociated health risk. Environ Int 2011;37:743e65.
[12] Srivastava A, Jain VK. A study to characterize the suspended particulate matter
in an indoor environment in Delhi, India. Build Environ 2007;42:2046e52.
[13] Kopanakis I, Eleftheriadis K, Mihalopoulos N, Simantris-Lydakis N, Katsivela E,
Pentari D, et al. Physico-chemical characteristics of particulate matter in the
Eastern Mediterranean. Atmos Res 2012;106:93e107.
[14] Bzdek BR, Pennington R, Johnston MV. Single particle chemical analysis of
ambient ultrane aerosol: a review. J Aerosol Sci 2012;52:109e20.
[15] Koak M, Theodosi C, Zarmpas P, Se guret MJM, Herut B, Kallos G, et al. Inuence of
mineral dust transport on the chemical composition and physical properties of
the Eastern Mediterranena aerosol. Atmos Environ 2012;57:266e77.
[16] Karanasiou AA, Siskos PA, Eleftheriadis K. Assessment of source apportion-
ment by positive matrix factorization analysis on ne and coarse urban
aerosol size fractions. Atmos Environ 2009;43:3385e95.
[17] Abt E, Suh HH, Catalano P, Koutrakis P. Relative contribution of outdoor indoor
particle sources to indoor concentrations. Environ Sci Technol 2000;34:
3579e87.
[18] Nazaroff WW. Indoor particle dynamics. Indoor Air 2004;14:175e83.
[19] Chen C, Zhao B. Review of relationship between indoor and outdoor particles:
I/O ratio, inltration factor and penetration factor. Atmos Environ 2011;45:
275e88.
[20] Qian J, Tavakoli B, Goldasteh I, Ahmandi G, Ferro AR. Building removal of
particulate pollutant plume during outdoor resuspension event. Build Environ
2014;75:161e9.
[21] Tian L, Zhang G, Jinghua T, Zhou J, Zhang Q. Mathematical model of particle
penetration through smooth/rough building envelop leakages. Build Environ
2009;44:1144e9.
[22] Long CM, Suh HH, Koutrakis P. Characterization of indoor particle sources
using continuous mass and size monitors. J Air Waste Manag 2000;50:
1236e50.
cek J, Dohanyosova P, Zdmal V, H
[23] Hussein T, Glytsos T, Ondra ameri K, et al.
Particle size characterization and emission rates during indoor activities in a
house. Atmos Environ 2006;40:4285e307.
[24] Ott WR, Siegmann HC. Using multiple continuous ne particle monitors to
characterize tobacco, incense, candle, cooking, wood burning, and vehicular
sources in indoor, outdoor, and in transit settings. Atmos Environ 2006;40:
821e43.
 N. Serfozo et al. / Building and Environment 82 (2014) 180e189
[25] Pagels J, Wierzbicka A, Nilsson E, Isaxon C, Dahl A, Gudmundsson A, et al.
Chemical composition and mass emission factors of candle smoke particles.
J Aerosol Sci 2009;40:193e208.
[26] Slezakova K, Pereira MC, Alvim-Ferraz MC. Inuence of tobacco smoke on the
elemental composition of indoor particles of different sizes. Atmos Environ
2009;43:486e93.
[27] Glytsos T, Ondracek J, D
zumbova L, Kopanakis I, Lazaridis M. Characterization
of particulate matter concentrations during controlled indoor activities.
Atmos Environ 2010;44:1539e49.
[28] Torkmahalleh MA, Goldasteh I, Zhao Y, Udochu NM, Rossner A, Hopke PK,
et al. PM2.5 and ultrane particles emitted during heating of commercial
cooking oils. Indoor Air 2012;22:483e91.
[29] Zhang H, Wang S, Hao J, Wan L, Jiang J, Zhang M, et al. Chemical and size
characterization of particles emitted from the burning of coal and wood in
rural households in Guizhou, China. Atmos Environ 2012;51:94e9.
[30] Thatcher TL, Layton DW. Deposition, resuspension, and penetration of parti-
cles within a residence. Atmos Environ 1995;29:1487e97.
[31] Spilak MP, Boor BE, Novoselac A, Corsi RL. Impact of bedding arrangements,
pillows, and blankets on particle resuspension in the sleep microenvironment.
Build Environ 2014;81:60e8.
[32] Ferro AR, Kopperud RJ, Hildemann L. Source strengths for indoor human ac-
tivities that resuspend particulate matter. Environ Sci Technol 2004;38:
1759e64.
[33] Corsi RL, Siegel JA, Chiang C. Particle resuspension during the use of vacuum
cleaners on residential carpet. J Occup Environ Hyg 2008;5:232e8.
[34] Qian J, Ferro R. Resuspension of dust particles in a chamber and associated
environmental factors. Aerosol Sci Technol 2008;42:566e78.
[35] Qian J, Ferro R, Fowler KR. Estimating the resuspension rate and residence
time of indoor particles. J Air Waste Manag 2008;58:502e16.
[36] Rosati JA, Thornburg J, Rodes C. Resuspension of particulate matter from
carpet due to human activity. Aerosol Sci Technol 2008;42:472e82.
[37] Shaughnessy R, Vu H. Particle loadings and resuspension related to oor
coverings in chamber and in occupied school environments. Atmos Environ
2012;55:515e24.
[38] Kubota Y, Higuchi H. Aerodynamic particle resuspension due to human foot
and model foot motions. Aerosol Sci Technol 2013;47:208e17.
[39] Kim Y, Gidwani A, Wyslouzil BE, Sohn CW. Source term models for ne par-
ticle resuspension from indoor surfaces. Build Environ 2010;45:1854e65.
[40] Goldasteh I, Ahmadi G, Ferro AF. Monte Carlo simulation of micron size
spherical particle removal and resuspension from substrate under uid ows.
J Aerosol Sci 2013;66:62e71.
[41] Goldasteh I, Ahmadi G, Ferro AF. Wind tunnel and numerical simulation of
dust particle resuspension from indoor surfaces in turbulent ows. J Adhes Sci
Technol 2013;27:1563e79.
[42] Wang S, Zhao B, Zhou B, Tan Z. An experimental study on short-time particle
resuspension from inner surfaces of straight ventilation ducts. Build Environ
2012;53:119e27.
[43] Tian Y, Sul K, Qian J, Mondal S, Ferro AR. A comparative study of walking-
induced dust resuspension using a consistent test mechanism. Indoor Air
2014. http://dx.doi.org/10.1111/ina.12107.
[44] Zhang X, Ahmadi G, Qian J, Ferro AR. Particle detachment, resuspension and
transport due to human waling in indoor environments. J Adhes Sci Technol
2008;22:591e621.
[45] Oberoi RC, Choi J, Edwards JR, Rosati JA, Thornburg J, Rodes CE. Human-
induced particle re-suspension in a room. Aerosol Sci Technol 2010;45:
216e29.
[46] Reeks MD, Hall D. Kinetic models for particle resuspension in turbulent ows:
theory and measurement. J Aerosol Sci 2001;32:1e31.
[47] Lazaridis M, Drossinos Y, Georgopoulos PG. Turbulent resuspension of small
nondeformable particles. J Colloid Interface Sci 1998;204:24e32.
[48] Guingo M, Minier J-P. A new model for the simulation of particle resuspension
by turbulent ows based on a stochastic description of wall roughness and
adhesion forces. J Aerosol Sci 2008;39:957e73.
189
[49] Goldasteh I, Tian Y, Ahmadi G, Ferro AR. Human induced ow eld and
resultant particle resuspension and transport during gait cycle. Build Environ
2014;77:101e9.
[50] Barth T, Preu J, Mller G, Hampel U. Single particle resuspension experi-
ments in turbulent channel ows. J Aerosol Sci 2014;71:40e51.
[51] Ibrahim AH, Dunn PF, Qazi MF. Experiments and validation of a model for
microparticle detachment from a surface by turbulent air ow. J Aerosol Sci
2008;39:645e56.
[52] Soltani M, Ahmadi G. Particle detachment from rough surfaces in turbulent
ows. J Adhes 1995;51:105e23.
[53] Ziskind G, Fichman M, Gutnger C. Resuspension of particulates from surfaces
to turbulent ows e review and analysis. J Aerosol Sci 1995;26:613e44.
[54] Hinds WC. Aerosol technology. New York: John Wiley & Sons; 1999.
[55] Chalvatzaki E, Kopanakis I, Kontaksakis M, Glytsos T, Kalogerakis N,
Lazaridis M. Measurements of particulate matter concentrations at a landll
site (Crete, Greece). Waste Manag 2010;30:2058e64.
[56] Moraetis D, Perdikatsis V, Pentari D, Manoutsoglou E, Apostolaki C, Lydakis -
Simantiris N. A study on the correlation of the properties of parent rock and
soils of different geological origin. In: 2nd International Conference on: Ad-
vances in Mineral Resources Management and environmental Geo-
technology, Hania, Greece; 2006.
[57] Watson JG, Chow JC, Lu Z, Fujita EM, Lowenthal DH, Lawson DR, et al.
Chemical mass balance source apportionment of PM10 during the Southern
California air quality study. Aerosol Sci Technol 1994;21:1e36.
[58] Nicholson KW. A review of particle resuspension. Atmos Environ 1988;22:
2639e51.
[59] Boor BE, Siegel JA, Novoselac A. Monolayer and multilayer particle deposits on
hard surfaces: literature review and implications for particle resuspension in
the indoor environment. Aerosol Sci Technol 2013;47:831e47.
[60] Gomes C, Freihaut J, Bahneth W. Resuspension of allergen-containing par-
ticles under mechanical and aerodynamic disturbances from human walking.
Atmos Environ 2007;41:5257e70.
[61] Qian J, Peccia J, Ferro AR. Review: walking-induced particle resuspension in
indoor environments. Atmos Environ 2014;89:464e81.
[62] Rim D, Persily A, Emmerich S, Dols WS, Wallace L. Multi-zone modelling os
size-resolved outdoor ultrane particle entry into a test house. Atmos Environ
2013;69:219e30.
[63] Thatcher TL, Lunden MM, Revzan KL, Sextro RG, Brown NJ. A concentration
rebound method for measuring particle penetration and deposition in the
indoor environment. Aerosol Sci Technol 2003;37:847e64.
[64] Schneider T, Kildes J, Breum NO. A two compartment model for determining the
contribution of sources, surface deposition and resuspension to air and surface
dust concentration levels in occupied rooms. Build Environ 1999;34:583e95.
[65] Hussein T, Hruska A, Doha nyosov a P, D
zumbova  L, Hemerka J, Kulmala M,
et al. Deposition rates on smooth surfaces and coagulation of aerosol particles
inside a test chamber. Atmos Environ 2009;43:905e14.
[66] Lai ACK. Particle deposition indoors: a review. Indoor Air 2002;12:211e4.
[67] He C, Morawska L, Gilbert D. Particle deposition rates in residential houses.
Atmos Environ 2005;39:3891e9.
[68] Corn M, Stein F. Re-entrainment of particles from a plane surface. Am Ind Hyg
Assoc J 1965;26:325e36.
[69] Biggs S, Cain RG, Dagastine RR, Page NW. Direct measurements of the adhe-
sion between a glass particle and a glass surface in a humid atmosphere.
J Adhes Sci Technol 2002;16:869e85.
[70] Bateman AP, Belessein H, Martin ST. Impactor apparatus for the study of
particle rebound: relative humidity and capillary forces. Aerosol Sci Technol
2014;48:42e52.
[71] Wu YL, Davidson CI, Russel AG. Controlled wind tunnel experiments for
particle bounceoff and resuspension. Aerosol Sci Technol 1992;17:245e62.
[72] Ibrahim AH, Dunn PF, Brach RM. Microparticle detachment from surfaces
exposed to turbulent air ow: controlled experiments sand modeling.
J Aerosol Sci 2003;34:765e82.