RESEARCH TRIANGLE INSTITUTE
WIND TUNNEL TEST REPORT NO. 29A
TEST OF THE RUPPRECHT AND PATASHNICK TEOM PM10 SAMPLER INLET
AT 2 AND 24 KM/H
Prepared by:
D. W. VanOsdell
Research Triangle Institute
P. 0. Box 12194
Research Triangle Park, NC 27709
May 1991
EPA Contract No. 68-02-4550
RTI Project No. 432U-4699-101
Project Officer
Kenneth A. Rehme
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
POST OFFICE BOX 12194 RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709-2194
-------
ABSTRACT
Wind tunnel tests of the Rupprecht and Patashnick (R&P) 10-pm inlet for the TEOM Series 1400 PM-
10 monitor have been conducted at 2 and 24 km/h. The purpose of the test was to compare the R&P inlet
to the Sierra-Andersen (SA) 246b Dichotomous Sampler inlet. The test program was conducted in the EPA
Aerosol Test Facility (ATF). The procedures used were those specified in 40 CFR Part 53 except that a
reduced number of test particle sizes were used. All tests utilized liquid challenge particles, and tests were
conducted at either 2 or 24 km/h.
Based on these limited tests, the R&P inlet appears to be functionally identical to the SA 246b
Dichotomous Sampler Inlet. Using the standard PM10 data analysis procedure, the R&P inlet cut-point was
estimated to be 9.8 pm at 2 km/h and 9.6 \im at 24 km/h (compared to 9.8 and 10.0 pm, respectively, for
the SA 246b.)
Concurrently with the R&P Inlet test, the Saturation Monitor (SM) and Combustion Engineering
Portable Indoor Particulate Sampler (PIPS) were tested. The results for these inlets are given in VanOsdell
(1991).
-------
CONTENTS
Section Page
ABSTRACT II
1.0 INTRODUCTION 1
2.0 CONCLUSIONS » 2
3.0 EXPERIMENTAL PROCEDURES 3
3.1 WIND TUNNEL ARRANGEMENT 3
3.2 AEROSOL GENERATION 5
3.3 SAMPLER POSITION AND OPERATION .., 6
3.4 INLET TESTS 8
3.5 ANALYSIS OF MASS COLLECTED ON FILTER SAMPLES 8
3.6 DATAANALYSIS 8
4.0 RESULTS AND DISCUSSION 10
4.1 EFFECTIVENESS RESULTS 10
5.0 REFERENCES 14
APPENDIX 15
-------
FIGURES
Pane
Figure 1. EPA Aerosol Test Facility and Wind Tunnel 4
Figure2. Arrangement of Samplers in the ATF Wind Tunnel 7
Figures. Comparison of SA 246b and R&P Inlets 11
Figure4. R&P I0-|jm Inlet Performance at 2 and 24 km/h 13
TABLES
Table 1. Wind Tunnel Set-Up for 2 and 24 km/h 5
Table 2. Summary of Multiple! Corrected R&P Test Results 10
Table A-l. Effectiveness Data 16
-------
TEST OF THE RUPPRECHT AND PATASHNICK TEOM PM10 SAMPLER INLET
AT 2 AND 24 KM/H
1 .0 INTRODUCTION
This report documents a test whose primary purpose was to evaluate the R&P 10-|jm inlet for the
R&P TEOM Series 1400 Continuous PM-10 Monitor. The R&P inlet is essentially identical to the widely-used
Sierra-Andersen (SA) 246b 10-pm inlet, which has been commercially available for a number of years.
VanOsdell and Chen (1996) previously reported the results of a wind tunnel test of the SA 246b at the EPA
ATF. Because the R&P inlet is a copy of the SA 246b, it was expected to be a fully satisfactory PM,, inlet.
During this test, the R&P inlet was tested in the EPA ATF following an abbreviated PM10 test protocol that
included 6 particle sizes at 2 and 24 km/h. By itself, this test was not sufficient to show that the R&P inlet
met the criteria of 40 CFR Part 53 for reference and equivalent PM10 samplers. K was sufficient to find any
significant differences in performance between the SA 246b and the R&P inlet.
Also tested during this test program were the Saturation Monitor (SM) and the Portable Indoor
Particulate Sampler (PIPS) sampler heads. The SM, constructed of plastic pipe, was designed to be an
inexpensive outdoor PM,, monitor. The PIPS was designed for indoor monitoring of particles in rooms, Two
SMs and two PIPSs were run during each test of the R&P Inlet. Results for these inlets can be found in
VanOsdell (1991).
-------
2.0 CONCLUSIONS
Based on this test of the R&P 10-|jm size-selective inlet, the following conclusions are drawn:
1. The wind tunnel effectiveness performance of the R&P inlet is substantially the same as that
of the SA 246b inlet at 2 and 24 km/h.
2. Given the physical similarity of the two inlets, the R&P inlet appears to meet the
requirements of 40 CFR Part 53 for a PM10 inlet. Because 2 and 24 km/h are the extremes
of the measurement range, it can be reasonably inferred that the R&P inlet would also
perform satisfactorily at 8 km/h.
-------
3.0 EXPERIMENTAL PROCEDURES
The test procedures used in the EPA Aerosol Test Facility were the same as those used and reported
previously (VanOsdell, Chen, and Newsome, 1988). Individual tests met the requirements of 40 CFR Part
53. Because the test program was designed primarily to compare the R&P inlet to the SA 246b only 2 wind
speeds and about half the number of particle sizes called for in 40 CFR Part 53 were tested during the
present work. A brief overview of the test procedures is given below, and details may be found in the report
by VanOsdell, Chen, and Newsome (1988).
3.1 WIND TUNNEL ARRANGEMENT
Figure 1 gives an overview of the EPA Aerosol Test Facility and the wind tunnel. Flow in the wind
tunnel was counterclockwise. There are few flow obstructions, and a number of access doors are provided
to allow all sections of the wind tunnel to be cleaned. The test aerosol was generated on top of the wind
tunnel where indicated, and injected through a distributor into the 1.83 m square cross-section region below.
The sampler test area is also indicated in Figure 1. At the test area the wind tunnel cross-section is 1.52 m
wide by 1.22 m high. The blower downstream of the sampler test area is capable of driving the wind tunnel
at speeds up to 50 km/h (1550 m3/min).
Some wind tunnel arrangement details not shown on Figure 1 were required to achieve acceptable
particle and velocity uniformity at the 2 wind speeds. A plywood baffle was placed about 1 m upstream of the
1.83 m square cross-section particle injection zone to promote mixing. The baffle was 1.22 m square and
mounted in the center of the wind tunnel transverse to the air flow. A counter-flow fan, 0.4 m in diameter and
centered in the cross-section, was operated about 1 m downstream from the injection zone to provide
additional mixing.
At 24 km/h, the large blower in Figure 1 powered the wind tunnel, and the filter/chiller was not turned
on except to clean the wind tunnel air for 30 min before beginning each day's testing. The large blower could
not be slowed enough to power the wind tunnel at 2 km/h. To operate at 2 km/h, the damper indicated on
Figure 1 was closed and the filter/chiller fan used to power the wind tunnel. To prevent flow channeling along
the wall of the wind tunnel during the 2 km/h tests, a center-hole baffle was placed 2 m downstream of the
sampler test area (and about 1 m upstream of the filter/chiller inlet. This baffle blocked the wind tunnel except
-------
Fluorometer/ >•
Microscope
Figure 1. EPA Aerosol Test Facility and Wind Tunnel
-------
for the 30-cm square hole in its center, and provided a symmetric flow profile at 2 km/h.
The velocity uniformity and turbulence intensity have been previously measured, with the wind tunnel
in the same physical configuration. The results are given in Table 1. The values at the center of the wind
tunnel were checked before beginning this test and found to be within the ranges in Table 1. These flow
parameters are within acceptable limits for PM10 testing.
Table 1. Wind Tunnel Set-Up for 2 and 24 km/h
Mean
Wind
Speed
2 km/h
24 km/h
Baffle
Arrangement
1.22 m2 centered
1.22 m2 centered
Mixing
Fan
On
On
Velocity
Uniformity
± 5%
±4%
Turbulence
Intensity
in Test Zone
3-4%
4-5%
Note: Velocity uniformity was calculated as the deviation from the mean within the test zone. Velocity was
measured with a hot-film probe.
3.2 AEROSOL GENERATION
The test was conducted with monodisperse test aerosols generated using a vibrating orifice aerosol
generator (VOAG). The aerosol material, oleic acid, was tagged with uranine, a fluorescent dye, and the oleic
acid and uranine were both dissolved in an ethanol carrier. The concentration of nonvolatiles (oleic acid and
uranine) in the ethanol varied as required to obtain the desired particle size after the ethanol evaporated.
Typical VOAG operation utilized a 20 \im orifice, 0.165 mL/min feed rate, and a frequency of about 70 kHz.
Particle size was calculated from the VOAG and particle solution parameters, and verified microscopically
using Nye-Bar treated glass slides and a flattening coefficient determined by Olan-Figueroa et al. (1982). The
liquid particles generated for the test had nominal diameters of 5, 7,9,10,12, and 25 \im.
The test aerosol was blown down into the wind tunnel through a dispersion manifold, and dispersed
-------
across the wind tunnel cross-section within the 10 m between the injection site and test zone. The uniformity
of particle dispersion and particle challenge concentration were evaluated during each test using an array of
five isokinetic samplers placed within the test zone and operated simultaneously with the samplers being
tested, The results of a day's tests were not useable if the particle mass collected by each individual
isokinetic sampler that day was not within +/-10 percent of the mean particle mass from the 5 isokinetic
samplers, No tests during the present test program had to be rejected. The isokinetic samplers are
described more fully below.
3.3 SAMPLER POSITION AND OPERATION
Figure 2 shows the arrangement of the samplers in the wind tunnel in a view along the direction of
wind flow. The inlet of each sampler was positioned in the same axial plane of the wind tunnel (the same
distance from the particle injection point.) That is, the upstream edges of the R&P inlet, the SMs, and the
PIPSs were all in the same plane as the upstream ends of the isokinetic sampler nozzles. (The isokinetic
samplers are indicated as 11 through 15. Dimensions above the wind tunnel floor are referenced to the center
of the PIPS inlet hole, the bottom of the SM wind cap, and the center of the isokinetic samplers.
The isokinetic samplers were 47 mm fitter holders fitted with sharp-edged conical nozzles, and were
operated isokinetically. The suction pipe at the back of each sampler was clamped to a support frame to hold
the sampler in position with the nozzle inlet about 25 cm upstream of the support frame. At 2 km/h, the
nozzles' inlets were 2.94 cm in diameter and the samplers were operated at 22.6 L/min. At 24 km/h, 1.22
cm diameter nozzles operated at 46.8 L/min were used. The flow rate through each sampler was controlled
with a manual valve that was preset to the required flow rate. During a test, the total flow through each
sampler was measured with a dry gas meter. The house vacuum manifold was used to draw the sample
through the isokinetic samplers,
The R&P inlet was attached to a 3.2 cm OD aluminum riser tube and supported at the center of the wind
tunnel as shown in Figure 2 A 47 mm fitter holder was mounted at the bottom of the tube, and a Gelman
A/E glass fiber filter collected the aerosol that penetrated the inlet. The flow rate through the R&P inlet was
controlled manually with a valve that was adjusted to the required 16.7 L/min prior to the test. During a test,
the total flow was measured using a dry gas meter. Suction was provided by the house vacuum manifold.
The other samplers were also positioned as shown in Figure 2. They were held in place using 3-
-------
152cm
SM1
PIPS2
19cm
14
--IOJ
122cm
R&P
PIPS1
11©
M 7 cm+16 err
SM2
013
ftf
«j«16cm-4»-17cnH
10cm
16.7cm
34.3cm
Figure 2. Arrangement of Samplers in the ATF Wind Tunnel
7
-------
fingered laboratory clamps that were themselves clamped to the support frame. With flow rates of 5 L/min
for the saturation monitors and 10 L/min for the PIPSs, their sampling did not affect flow near the R&P inlet.
3.4 INLET TESTS
Three sequential tests of the inlets were conducted on the same day using the same test aerosol for
most particle sizes. The R&P inlet, two SMs, two PIPSs, and 5 isokinetic filter samplers were operated
simultaneously during each of the tests. The duration of each test was set to ensure that-the aerosol mass
captured on the sampler filters was sufficient to provide a reliable measurement. Most runs lasted 1 hour,
but the 5 and 25 \im particle runs at 24 km/h were 3 hours long.
The sampling effectiveness of the R&P inlet was computed as the ratio of the measured mass
concentration to the mean of the mass concentration measured by the five isokinetic samplers shown in
Figure 2.
3.5 ANALYSIS OF MASS COLLECTED ON FILTER SAMPLES
Following the EPA Aerosol Test Facility standard procedures, the fitters from the samplers were
placed in clean 2 oz. bottles to which 20 ml of 0.01 N NaOH were added. The uranine was extracted from
the filters into the NaOH solution by soaking overnight following 20 minutes of ultrasonic mixing. The mass
of test aerosol collected on the filters was determined fluorometrically using standard ATF procedures. The
nozzles of the isokinetic samplers were washed and the uranine found in the wash was added to the uranine
collected on the filter to obtain the total challenge aerosol mass. The inlet sections of the R&P inlet, SMs,
and PIPSs were not washed.
3.6 DATA ANALYSIS
The raw effectiveness data from the samplers was analyzed using the PM,, data analysis procedure
normally used at the ATF. The three effectiveness values for each test were averaged to obtain a value at
each test particle size. These effectiveness values were then input to the PM,, data analysis computer
program (VanOsdell, Chen, and Newsome, 1988). For each sampler and wind speed, the effectiveness data
8
-------
were adjusted to account for the presence of muttiplets of the primary challenge particle. A robust-spline
curve (in log-normal space) was then fit to the muKiplet-corrected data The PM10, data analysis procedure
outlined in 40 CFR Part 53 requires that the effectiveness-particle size data be fit with a smooth curve and
that the ends of the curve be smoothly extrapolated to 100 percent at 1 urn and 0 percent at 50 \un, and this
requirement has been implemented mathematically in the data analysis program. (Because the curve fit is
generated in log-normal space, values above 100 percent are suppressed.) The program usually fits
effectiveness data well, especially in the region of the cut-point, and it provides an impartial estimate of an
inlet's performance parameters. The robust spline curve-fit process does not impose any preconceived
functional form on the data The DM, expected mass collection for the PM,, ambient particle size distribution
(40 CFR Subpart D, Part 53,Table D-3), and expected mass ratio were all computed based on the robust-
spline curve.
-------
4.0 RESULTS AND DISCUSSION
4.1 EFFECTIVENESS RESULTS
Figure 3 presents a direct comparison between the current test of the R&P inlet and a previous test
of the SA 246b inlet (VanOsdell and Chen, 1990). (To facilitate comparisons with the results presented by
VanOsdell and Chen (1990), the effectiveness values in Figure 3 have not been multiplet-corrected.) At both
2 and 24 km/h, the R&P and SA 246b effectiveness values agree extremely well, and show little wind speed
dependence.
A summary of the test program results (after correction for multiplets) is presented in Table 2. The
Table 2. Summary of Multiple! Corrected R&P Test Results
2 km/h Dso Mm
2 km/h Expected Mass, pg/m3
2 km/h Mass Ratio to Ideal PM,,
Sampler
24 km/h D50 fjm
24 km/h Expected Mass, pg/m3
24 km/h Mass Ratio to Ideal PM10,
Sampler
R&P
Inlet
9.82
148.0
1.028
9.58
147.8
1.027
Note: All values computed using standard PM10 Data Reduction Program. All effectiveness values
were corrected for multiplets.
10
-------
Effectiveness (Uncorrected), %
10
Particle Size, um
1ZU
<4 f\/\
100
f\f\
90
on
oU
~ff\
70
&r\
60
Cf\
OU
/tf\
40
QO
ou
nf\
£\J
1 A
10
n
.! 4 M M
I <> I <> M
f; :
; :
! M MA
: : : : : ^y
I \ I' III
' , '. '. ' ' ' J
I '!!!=:
L ^ ; ; ....;........
f I ! M ! !
I i i i i i i i
S
I
i
j
D 2km/hSA246b i
A 2km/hR&P : i
09A, km/h QA 94fih : :
V 24km/hR&P ; i
..xx i... . I i i M ii
<§> i i j | ; ; i ;
v : : :.::,:::
i /s. ! I i M M
I I : ^^ : : : : ! : :
i l"v i i i i i i i
100
Figure 3. Comparison of SA 246b and R&P Inlets
11
-------
Expected Mass and Mass Ratio to Ideal Sampler are values used to compare PM10, samplers. The ideal
sampler effectiveness performance curve and the ambient particle mass distribution are given in 40 CFR Part
53. The expected mass is obtained by multiplying the mass in each size fraction of the size distribution by
the sampler's effectiveness at that size and summing over the size distribution. The ratio is self-explanatory.
The complete data sets for each wind speed are given in the Appendix
Figure 4 shows the data and curve-fits for the R&P inlet at 2 and 24 km/h. The data are seen to be
well-behaved, and the D50, expected mass, and mass ratio values given in Table 2 provide good
representations of the R&P sampler's behavior. Within the limits of this data set, the R&P lO-pm Inlet
appears to easily meet the wind tunnel sampling requirements of 40 CFR Part 53. While the 8 km/h data
were not gathered, the 2 and 24 km/h data span the limits of interest and are the most likely velocities for
a sampler to fail the test procedure.
12
-------
120
110
100
90
80
70
60
50
40
30
20
10
Effectiveness (Multiple! Corr.), %
1
.;....;..-:...;*»
D 2 km/h Data
2 km/h Fit
A 24 km/h Data
24 km/h Fit
10
Particle Size, um
100
Figure 4. R&P 10-nm Inlet Performance at 2 and 24 km/h
13
-------
5.0 REFERENCES
Lundgren, D. A and Paulus, H. J. (1975) The Mass Distribution of Large Atmospheric Particles, JAPCA.
25:12, pp. 1227-1231.
Olan-Figueroa, E., McFarland, A R, and Ortiz, C. A (1982). 'Flattening Coefficients for OOP and Oleic Acid
Droplets Deposited on Treated Glass Slides.' AIHA Journal. Vol. 43, pp. 395-399.
VanOsdell, D. W. (1991). Wind Tunnel Test Report No. 29. Test of the Rupprecht and Patashnick PM10
Sampler inlet, the Saturation Monitor Inlet, and the Portable Indoor Particulate Sampler Inlet at 2 and
24 km/h. US Environmental Protection Agency, AREAL, Research Triangle Park, N.C., K A Rehme,
Project Officer.
VanOsdell, D. W. and Chen, F.-L (1990). Wind Tunnel Test Report No. 28. Test of the Sierra-Andersen
246b Dichotomous Sampler Inlet at 2, 8, and 24 km/h. US Environmental Protection Agency,
AREAL, Research Triangle Park, NC., K, A Rehme, Project Officer.
VanOsdell, D. W., Chen, F.-L, Newsome, J. R. (1988) The PM,, Sampler Evaluation Program: Annual
Report August 1987 to July 1988. US Environmental Protection Agency, AREAL Research Triangle
Park, NC., K A Rehme, Project Officer.
14
-------
APPENDIX
15
-------
Table A-1. Effectiveness Data
Date
2kmft
3/12/91
3/12/91
3/12/91
3/13/91
3/13/91
3/13/91
3/18/91
3/18/91
3/18/91
3/19/91
3/19/91
3/19/91
3/21/91
3/21/91
3/21/91
3/26/91
3/26/91
3/26/91
3/27/91
3/27/91
3/27/91
3/28/91
3/28/91
3/28/91
Aero.
Oil
6*0
Te«»
10.03
10.03
10.03
5.04
5.04
5.04
24.71
24.71
24.71
5.01
5.01
5.01
7.04
7.04
7.04
1209
1209
1209
9.05
8.05
9.05
7.04
7.04
7.04
Oeo.
Std
Dev.
1.025
1.025
1.025
1.028
1.028
1.028
1.015
1.015
1.015
1.035
1.035
1.035
1.022
1.022
.1.022
1.025
1.025
1.025
1.016
1.016
1.016
1.023
1.023
1.023
%Doub
6.3
6.3
6.3
7.1
7.1
7.1
23
23
23
23
23
23
6.6
6.6
6.6
5.1
5.1
5.1
8.6
8.6
8.6
3.4
3.4
3.4
%Trip.
0.0
0.0
0.0
1.3
1.3
1.3
ao
ao
0.0
0.0
0.0
0.0
1.7
1.7
1.7
3.4
3.4
3.4
0.0
0.0
0.0
0.0
0.0
0.0
R&P
Inlet
428
44.8
44.2
97.8
826
85.7
1.7
1.0
ao
97.8
91.6
95.3
71.3
71.5
64.5
30.3
30.3
321
58.0
57.4
54.7
75.1
78.5
75.9
Part
Cone.
254
256
261
27
25
26
234
1.08
214
26
25
25
1.47
1.37
1.31
4.11
4.20
4.36
208
203
210
1.49
203
1.52
COV
FWwMeai.
W
20
21
1.8
5.3
6.0
4.0
6.4
5.0
5.8
4.1
25
26
22
.9
.7
1.0
ao
1.6
.8
20
.7
5.5
3.1
1.6
-------
Otte
Am.
DU.
ftim)
24kmATe»t»
4/15/91
4/16/91
4/16/91
4/16/81
4/17/91
4/17/91
4/17/91
4/18/91
4/18/91
4/16/91
4/24/91
4/29/91
4/29/91
4/29/91
5.00
9.00
9.00
9.00
11.83
11.83
11.83
7.00
7.00
7.00
24.99
10.00
10.00
10.00
Geo.
Std
Dev.
1.035
1.018
1.018
1.018
1.015
1.015
|_ 1.015
1.024
1.024
1.024
1.007
1.028
1.028
1.028
«0oufa
2.3
3.6
3.6
3.6
4,8
4.8
4.8
Z7
2,7
27
6.6
6.5
6.5
6.5
ttTrip.
ao
0.9
0.9
0.9
1.9
1.9
1.9
a7
0.7
0.7
Z2
1.9
1.9
1.9
R»P
InM
BL(K)
86.5
54.8
57.4
61.0
31.9
30.8
31.3
77.2
65.7
67.6
4.6
44.5
41.0
45.6
Pert.
Cone.
(W/m1
20
.31
.31
29
.46
.48
.48
.41
.40
.41
.63
.47
.46
.50
cov
• Rake MM*.
W
Z8
5.0
2.5
4.0
5.5
1.4
5.2
5.7
3.7
3.1
7.8
4.8
5.0
5.0
•COV » Coefficient of Variation (Std. Dot/Mean) for concentration measurements fay the 5 rake umplen
------- |