Page 153 - Air Pollution Control Engineering

03_chap_wang.qxd 05/05/2004 12:48 pm Page 132
132 José Renato Coury et al.

Table 12
Range of Body Diameters and Cyclone Reynolds Number Studied
Cyclone D (cm) Re
c cyc
Stairmand 1.9, 20.3, 25.2, 30.5 5,000–500,000
Lapple 3.8–14.0 2,000–64,000
Z 12–300 12,000–300,000
SRI II and III 3.1–3.7 6,000–32,000

Source: (Adapted from Lidén and Gudmundsson (49), with permission from Taylor and Francis;
http://www.tandf.co.uk/journals.)

Table 13
Dimensional Ratios for the Cyclones Studied

Cyclone D /D a/D b/D D /D S/D H/D h/D
e c c c in c c c c
Stairmand 0.50 0.50 0.20 0.50 4.0 1.5
Lapple 0.50 0.50 0.25 0.75 4.0 2.0
Z 0.25 0.45 0.11 ~0.82 ~3.4 ~0.86
SRI II 0.286 0.286 0.43 1.9 1.3
SRI III 0.27 0.27 0.35 1.6 0.45
Source: (Adapted from Lidén and Gudmundsson (49), with permission from Taylor and Francis;
http://www.tandf.co.uk/journals.)

Amultiple regression analysis was performed on ψ versus the annular Reynolds
50
number and several ratios of cyclone dimensions, and at the 5% level of significance,
the result was 2

ψ = exp( ) Re F 1  S   F (71)

F
ann
0
50
 D c 
The results of the multiple regression for the parameters F , F , and F are given in
0 1 2
Table 14, and are valid for the four cyclone types studied. Comparison between the
model and experiments were given for the Stairmand and SRI cyclones. From the respec-
tive curves, it was possible to extract the parameters a , a , and a [Eq. (68)], which are
1 2 3
listed in Table 15. Therefore, by calculating D from Eq. (71), the efficiency curve for
50
the cyclones can be derived from Eq. (67). The results for D = 2.5 µm are shown in
50
Figs. 15 and 16 for the Stairmand and SRI cyclones, respectively, and are compared to
the PM (WINS) curve. It can be noted that the SRI cyclone has a much better fit
2.5
than the Stairmand, which is not surprising, as the former was developed for sampling,
and the latter is a classical design for industrial use. The final fitting between model and
experiments, presented by Lidén and Gudmundsson, show a considerable scatter for the
Stairmand data and much less for the SRI data. In any case, these relations are useful
guidelines for cyclone scaling, but caution is advised before using the equations for
accurate cyclone design.