You are on page 1of 8

Chute Aerators.

II: Hydraulic Design


Michael Pfister1 and Willi H. Hager, F.ASCE2
Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

Abstract: Chute aerators are applied if cavitation damage on spillways is expected or observed. The aerator efficiency is usually
described with the ratio of the air discharge entrained through the air supply ducts and the water discharge, which does however not
account for the resulting air concentration distribution within the flow or for air detrainment. The present study investigates the streamwise
development of the air transport along the flow downstream of chute aerators. Based on an extensive test program in which six governing
parameters were systematically varied, the development of the average and the bottom air concentrations is provided up to the self-
aeration point. Based on this information, an optimization of aerators in terms of increased air entrainment and reduced detrainment rates
is possible, by assuming minimum required air concentrations. The main parameters influencing the air transport downstream of aerators
are the approach flow Froude number, the deflector angle and the chute bottom angle.
DOI: 10.1061/共ASCE兲HY.1943-7900.0000201
CE Database subject headings: Aeration; Cavitation; Entrainment; Spillways.
Author keywords: Air; Aerator; Cavitation; Chute; Deflector; Detrainment; Entrainment; Offset.

Introduction Göğüş 共2002兲, and Gaskin et al. 共2003兲. The chute aerator effi-
ciency was normally described with the global air entrainment
Major cavitation damage occurred on spillways between 1970 coefficient ␤ = QA / QW, i.e., the ratio of the air supply discharge
and 1980, e.g., at Keban Dam 共Turkey兲 or Karun Dam 共Iran兲. and the water discharge. However, these studies do not account
Chute aeration has been considered the most efficient technical for the streamwise air detrainment, as ␤ measures only the local
measure to counter cavitation damage, such that extensive experi- air entrainment at the aerator. The development of Cb was par-
ments have been conducted on chute aerators. Other means in- ticularly not considered in these studies. Chanson 共1989兲 mea-
cluding the treatment of chute concrete, the use of smooth sured air concentrations and discussed air entrainment and
shuttering formwork or the application of steel fibers were less detrainment regions along the chute downstream of aerators. He
also related the observed deaeration process at the reattachment
effective, in addition to increased cost 共Pan and Shao 1984兲.
zone to the associated increased bottom pressure. Kramer 共2004兲
Cavitation risk is usually characterized with the cavitation
further investigated the air detrainment characteristics in the far-
index K = 共h p + ha − hv兲 / 共V2 / 2g兲, in which h p = bottom pressure
field downstream of the jet impact zone including the natural
head, hv = vapor pressure head, ha = atmospheric pressure head,
self-aeration process. Information concerning the air transport in
V = chute flow velocity, and g = acceleration due to gravity. An
the vicinity of aerators is necessary to quantify the air transport
index of K ⬍ 0.2 is generally considered critical, requiring chute
between the aerator and the natural inception point including 共1兲
flow aeration 共Pugh and Rhone 1988; Falvey 1990兲. Following
the efficiency of an aerator and 共2兲 the distance to a second po-
Peterka 共1953兲 and Rasmussen 共1956兲, an average 共subscript a兲
tentially required aerator in terms of the overall aerated chute
air concentration Ca = QA / 共QA + QW兲 of approximately 0.01⬍ Ca reach downstream of an aerator.
⬍ 0.06 avoids cavitation damage, with QA = air supply discharge The present experimental study investigates the air distribution
and QW = water discharge. Although it is evident that the bottom and transport downstream of chute aerators, primarily as repre-
共subscript b兲 air concentration Cb is of prime interest in the con- sented by Cb and Ca, containing entrapped and entrained air as
text of cavitation prevention, the aforementioned criterion relates described by Wilhelms 共1997兲. Based on systematic measure-
to the average concentration. ments and on the analysis of 共1兲 the air-water flow surface; 共2兲
Bottom aerators were investigated by, among others, Ko- Ca共x / L兲; and 共3兲 Cb共x / L兲, general air transport zones downstream
schitzky 共1987兲, Rutschmann 共1988兲, Chanson 共1988兲, of chute aerators are described. Three flow zones are introduced,
Rutschmann and Hager 共1990兲, Skripalle 共1994兲, Kökpınar and namely 共1兲 the jet zone along 0 ⬍ x / L ⱕ 1; 共2兲 the reattachment
and spray zone along 1 ⬍ x / L ⬍ 3; and 共3兲 the far-field zone
1
Dr. sc. ETH, Laboratory of Hydraulics, Hydrology and Glaciology downstream of x / L = 3, with x = streamwise coordinate and L
共VAW兲, ETH Zurich, Zurich CH-8092, Switzerland 共corresponding au- = jet length. Furthermore, the general features of the main air
thor兲. E-mail: pfister@vaw.baug.ethz.ch transport parameters Ca and Cb were derived from experimenta-
2
Prof. Dr., Laboratory of Hydraulics, Hydrology and Glaciology tion. This paper highlights the results concerning Ca and Cb,
共VAW兲, ETH Zurich, Zurich CH-8092, Switzerland. E-mail: hager@vaw. while the general air transport zones and an equation for L are
baug.ethz.ch
introduced in a companion paper 共Pfister and Hager 2010兲.
Note. This manuscript was submitted on April 11, 2008; approved on
January 7, 2010; published online on May 14, 2010. Discussion period
open until November 1, 2010; separate discussions must be submitted for Experimental Setup
individual papers. This paper is part of the Journal of Hydraulic Engi-
neering, Vol. 136, No. 6, June 1, 2010. ©ASCE, ISSN 0733-9429/2010/ The experiments were conducted in a 0.3 m wide and 6 m long
6-360–367/$25.00. sectional chute model at the Laboratory of Hydraulics, Hydrol-

360 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010

J. Hydraul. Eng., 2010, 136(6): 360-367


Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Definition sketch with notation

ogy, and Glaciology 共VAW兲 of ETH Zurich. The chute bottom


angle ␸ was set to 12°, 30°, and 50° relative to the horizontal.
Flows of variable approach 共subcript o兲 flow depths ho and
Froude numbers Fo = Vo / 共gho兲0.5 were generated with a jet-box,
with Vo = approach flow velocity 共Fig. 1兲. It was possible to vary
Fo and ho independently by adapting the discharge, i.e., different
ho were tested with the same Fo, and different Fo with the same
h o.
The discharge was measured with an electromagnetic flow Fig. 2. Relations between ␤ and 共a兲 Fo; 共b兲 ␣; 共c兲 ␸; and 共d兲
meter 共Krohne, Germany兲. The flow depths upstream of the aera- 共s + t兲 / ho. Symbols represent test series according to Table 1.
tor were measured with a point gauge and the pressures in the air
cavity with an U-shaped manometer. The air entrained by the
aerators was supplied through a lateral duct and measured using a
thermoelectric anemometer 共Schiltknecht, Switzerland兲. The two- parameter is not considered significant, as it represents only the
dimensional 共2D兲 air concentration distribution of the flow was local air entrainment at the aerator without specifying the down-
measured using a fiber-optical probe 共RBI Instrumentation, stream air concentration development, it was measured to permit
France兲. comparisons with previous results. Herein ␤ is defined using QA
The following parameters influence the air transport process supplied from the air duct, thereby not accounting for the air
共Fig. 1兲: deflector height t, deflector angle ␣, offset height s, chute transport and entrainment along the upper jet trajectory. The char-
bottom angle ␸, approach flow velocity Vo, and depth ho, mea- acteristic parameters influencing ␤ were systematically varied in
sured upstream of the deflector. The jet length L was visually the hydraulic model 共Fig. 2兲. Test series were conducted varying
detected as the distance between x = 0 and the reattachment point one parameter and keeping the other parameters constant. As a
P of the lower jet trajectory. All experiments were conducted with consequence, the isolated effect of each single parameter on ␤
approximately zero subpressure head hs ⬇ 0 in the cavity below was found. Several test series resulted for each basic parameter,
the jet, i.e., the pressure head in the cavity corresponds to atmo- depending on the selected boundary conditions.
spheric pressure. The air-water flow surfaces were defined along The isolated effect of Fo on ␤ is shown in Fig. 2共a兲. Three
the isoconcentration lines of C = 0.90. In total, 93 tests were con- conclusions may be drawn: 共1兲 ␤ increases significantly with Fo;
ducted, including almost 60,000 air concentration measurement 共2兲 the extended curves of all test series intersect the abscissa at
points. Three types of aerators were systematically investigated: Fo ⱖ 4; and 共3兲 the lower set of curves including offsets without
共1兲 deflector without offset; 共2兲 offset without deflector; and 共3兲 deflectors is flatter, while those for aerators with deflectors are
combination of both. Five deflector geometries were tested, each more sensitive to changes of Fo. Furthermore, deflectors with
combined with offset heights of s = 0, 25, and 45 mm and partially relatively large ␣ induce a steeper curve than flat deflectors. The
also 100 mm. Furthermore, offsets of s = 25, 45, and 100 mm effect of ␣ on ␤ is given in Fig. 2共b兲: for 0 ° ⱕ ␣ ⱕ 11.3°, ␤
height without deflector were analyzed. Scale effects related to increases with ␣. Steep deflectors entrain thus more air than off-
the present investigation are discussed in the companion paper sets, for otherwise identical boundary conditions. Furthermore,
共Pfister and Hager 2010兲, their effect was argued to be small. tests with equal Fo are similar relative to ␤, while small Fo 共sym-
bol ⌬兲 generate smaller ␤. In contrast, no significant effect of ␸
on ␤ was found, as shown in Fig. 2共c兲. Consequently, ␤ is obvi-
Air Entrainment Coefficient ously independent of ␸. The effect of 共s + t兲 / ho on ␤ is shown in
Fig. 2共d兲. Again, no clear trend results and ␤ seems to be inde-
The relevant dimensionless parameters governing the aeration pendent of the dimensionless aerator geometry and flow depth
process may be derived from a dimensional analysis 共Koschitzky term. Each test series is denoted by a character in the legend of
1987; Chanson 1988; Rutschmann 1988; Wood 1991; Pfister Fig. 2 defining the related parameters, which were kept constant
2008兲 as Fo, ␣, ␸, and 共s + t兲 / ho, beside dimensionless terms for for the relative test series, according to Table 1.
the cavity subpressure, the viscosity and the surface tension. The It is known from literature that ␤ is primarily a function of Fo
latter three were not considered in the present data analysis. 共Rutschmann 1988; Chanson 1988; Rutschmann and Hager 1990;
The air entrainment coefficient ␤ is commonly used to de- Kökpınar and Göğüş 2002兲. The present investigation points also
scribe the air entrainment capacity of an aerator. Although this to the strong effect of ␣, provided ␣ ⬎ 0. The effect of the other

JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010 / 361

J. Hydraul. Eng., 2010, 136(6): 360-367


Table 1. Test Series from Fig. 2 with Parameters Held Constant for a
Specific Series
Fig. 2共a兲
Series ␸ 共degrees兲 ␣ 共degrees兲 共s + t兲 / ho 共⫺兲
A 50 5.7 0.3
B 50 0 0.9
C 50 0 0.5
D 50 5.7 0.8
Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

E 30 8.1 0.6
F 30 0 0.4
G 30 0 0.8 Fig. 3. Air entrainment coefficient ␤ measured 共a兲 on hydraulic mod-
H 30 8.1 0.9 els; 共b兲 validated on prototypes. Gray: limitation of Eq. 共1兲.
I 30 8.1 0.2
J 12 0 0.7
K 12 11.3 0.9 Ⰷ 0 causing pressure fluctuations with a reduced ␤ value. In Fig.
3共a兲 the measured ␤ values are plotted as a function of the term
L 12 11.3 0.6
⌽␤ = F2o关1 + Fo tan ␣兴 from Eq. 共1兲, intersecting the abscissa ␤ = 0
Fig. 2共b兲 at ⌽␤ ⬇ 36. As a consequence, a minimum Fo is required to gen-
Series ␸ 共degrees兲 Fo 共⫺兲 共s + t兲 / ho 共⫺兲 erate air entrainment. The minimum values are Fo ⬵ 6 for offsets,
A 50 7.5 0.2 Fo ⬵ 5 for aerators with flat deflectors 共␣ = 5.7°兲 and Fo ⬵ 4 for
B 50 7.5 0.9
aerators with steep deflectors 共␣ = 11.3°兲. Rutschmann 共1988兲
stated similar values. From Eq. 共1兲 the air entrainment depends
C 50 7.5 0.6
primarily on Fo and only slightly on the aerator geometry. Fur-
D 30 6.8 0.6
thermore, aerators with deflectors are more efficient in terms of
E 30 7.5 0.9
air entrainment than those with only offsets.
F 30 7.5 0.2 Fig. 3共a兲 includes also data of Koschitzky 共1987兲, Rutschmann
G 12 7.5 0.2 共1988兲, and Skripalle 共1994兲. Koschitzky accounted for the effect
H 12 7.5 0.9 of cavity subpressure. His tests with relatively small subpressures
Fig. 2共c兲 were adjusted between 0.6⬍ ␤ / ␤ M ⬍ 1 by Rutschmann and Hager
Series ␣ 共degrees兲 Fo 共⫺兲 共s + t兲 / ho 共1990兲, with ␤ M = maximum air entrainment coefficient for hs = 0.
Furthermore, only tests with a fully open air duct were consid-
A 5.7 7.5 0.2 ered. A coefficient of R2 = 0.91 resulted between the modified data
B 5.7 7.5 0.4 set of Koschitzky and Eq. 共1兲. The data set of Rutschmann was
C 5.7 7.5 0.1 limited to 共1兲 0 ⬍ hs / ho ⬍ 0.1; 共2兲 0 ⬍ ␤ ⬍ 0.8; and 共3兲 Fo ⬎ 4.2,
D 8.1 7.5 0.2 resulting in R2 = 0.94 between his data and Eq. 共1兲. The data set of
E 0 7.5 0.7 Skripalle describes offsets without deflectors on horizontal chutes.
F 11.3 7.5 0.9 Test with hs ⬇ 0 and with a smooth chute surface were considered,
Fig. 2共d兲 resulting in R2 = 0.89. These three additional data sets confirm the
accuracy of Eq. 共1兲 such that its limitations may be extended
Series ␣ 共degrees兲 Fo 共⫺兲 ␸ 共degrees兲
according to the boundary conditions of these studies to 5.8
A 5.7 7.4 50 ⱕ Fo ⱕ 16 . 1, 0 . 06 ⱕ 共s + t兲 / ho ⱕ 2.1, 0 ° ⱕ ␣ ⱕ 11.3°, and 0 ° ⱕ ␸
B 11.3 7.5 50 ⱕ 50°.
C 8.1 7.4 50 Furthermore, Wood 共1991兲 presented prototype measurements
D 0 7.7 50 of ␤. Five chute aerators were analyzed, namely Foz do Areia
E 0 9.5 30 Dam 共Brazil兲, Emborcação Dam 共Brazil兲, Amaluza Dam 共Ecua-
F 0 7.5 30 dor兲, Colbun Dam 共Chile兲, and Tarbela Dam 共Pakistan兲. Aerators
G 13.3 7.5 30 with subpressure coefficients smaller than 0.4, according to the
H 13.3 9.2 30 definition of Wood, were considered. These prototype data fit with
I 5.7 7.5 12
Eq. 共1兲 to R2 = 0.88 关Fig. 3共b兲兴.
J 13.3 7.5 12
K 13.3 9 12
Average Air Concentration

Downstream of the jet reattachment point P 共Fig. 1兲, the average


parameters ho, t, s, and ␸ on ␤ was found insignificant. It is
air concentration Ca, integrated over the flow depth defined be-
proposed that
tween z = 0 and the upper air-water flow surface zu at C = 0.90, is


␤ = 0.0028F2o关1 + Fo tan ␣兴 − 0.1, for 0 ⬍ ␤ ⬍ 0.80 共1兲 zu
1
which yielded a coefficient of determination R = 0.93. Eq. 共1兲
2 Ca = C共z兲dz 共2兲
zu 0
may be applied to aerators consisting of deflectors, offsets or
combinations, given that hs ⬇ 0. The air duct should consist of a In the context of cavitation prevention, the values of Ca共x / L兲 are
sufficiently large cross-section to avoid relevant subpressures hs of interest in the reattachment and the spray zone 共II兲 共1 ⬍ x / L

362 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010

J. Hydraul. Eng., 2010, 136(6): 360-367


Ca = Ca共3L兲 + 0.02 冉 冊
x
L
− 3 sin共␸ − 30°兲, for 3 ⱕ x/L ⱕ 9

共4兲

The model data of Ca共x / L ⱖ 3兲 and those computed with Eq. 共4兲
are compared in Fig. 4共b兲. Flows on “steep” chutes with ␸
ⱖ 30° have sufficient air transport capacity to maintain the aver-
age air concentration, while the flow deaerates otherwise. This
Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

average streamwise air detrainment is slightly smaller from Eq.


共4兲 than as given by Kramer 共2004兲, who investigated chute bot-
Fig. 4. Average air concentration Ca 共a兲 at x / L = 3 versus L / ho; 共b兲 tom angles between 0 ° ⱕ ␸ ⱕ 26°. An aerator designed for large
comparison between measured and computed values from Eq. 共4兲 values of Ca in the far-field zone 共III兲 should preferably produce a
along far-field zone 共III兲 for x / L ⬎ 3 long jet 共large L兲 on a steep chute allowing for a progressive jet
disintegration, combined with a high air concentration Ca共3L兲.

⬍ 3兲, as well as in the far-field zone 共III兲 共x / L ⱖ 3兲. The jet zone
共I兲 is irrelevant because no cavitation occurs as long as the flow is Bottom Air Concentration
separated from the chute bottom.
The bottom air concentration profile Cb downstream of point P is
Reattachment and Spray Zone defined experimentally using the values measured closest to the
model chute bottom, i.e., Cb = C共2 mmⱕ z ⱕ 3 mm兲. A data
The values of Ca strongly decrease in the reattachment region evaluation motivated a distinction between the streamwise devel-
共1 ⬍ x / L ⬍ 1.25兲 due to flow deflection, typically from 1.0Ca at P opment of Cb in zone 共II兲 for 共1兲 offset aerators without deflectors
to values of 共0.4 to 0.9兲Ca within the short distance of 0.25L. The 共s ⬎ 0 and ␣ = 0兲 and 共2兲 deflector aerators with or without offsets
relative detrainment rate is in particular large for steep and high 共s ⱖ 0 and ␣ ⬎ 0兲, while the far-field zone 共III兲 can be described
deflectors with large Fo. For such conditions, the corresponding ␤ independent of the aerator type.
is also large, such that the absolute downstream air transport is
still relatively large. A local maximum of Ca appears in the spray
region at approximately 1.5⬍ x / L ⬍ 2, generated by water par- Reattachment and Spray Zone
ticles ejected from the flow surface due to jet reattachment, Aerators consisting of offsets without a deflector are exclusively
thereby elevating the C = 0.90 isoconcentration line. As a conse- characterized by s, as t = ␣ = 0. Similar to the aforementioned in-
quence, the transport of entrapped air near the surface increases, vestigation of ␤, test series were conducted with a systematic
while the air concentration close to the chute bottom decreases. parameter variation to determine the isolated effect of the inves-
As described in the companion paper, local flow phenomena tigated variables on Cb. The analysis of all varied parameters
strongly influence Ca along the reattachment and spray zone 共II兲, provided Cb versus the streamwise coordinate 关共x / L兲 − 1兴.
complicating a quantification of Ca. A data analysis indicates that The effect of Fo was found as F−2 o , i.e., the values of
the value of Ca at x / L = 3 共subscript 3L兲 varies with the relative Cb关共x / L兲 − 1兴 decrease less for larger than for smaller Fo. The two
jet length L / ho as parameters ho and s were combined to form the relative drop
height ho / 共ho + s兲, and their effect was described with the expo-
L nent ⫺1.3. Large approach flow depths ho potentially generate
Ca共3L兲 = 0.008 + Cai, for 5 ⱕ L/ho ⱕ 40 共3兲
ho thick jets with an increased disintegration capacity resulting in
high values of Cb, while jets impinge with a relatively steep angle
The measured values of Ca共3L兲 are compared with Eq. 共3兲 in Fig. onto the chute for high offsets, resulting in a deaeration at the
4共a兲, resulting in R2 = 0.88. The data indicate an average incipient chute bottom 共Chanson 1988兲. The effect of the chute slope was
共subscript i兲 air concentration of Cai ⬵ 0.10 for L / ho = 0. Cai is retained as 共sin ␸兲0.4. Combining the individual effects of all pa-
equal or marginally smaller than Ca共x / L = 0兲 measured at jet take- rameters on Cb along 1 ⱕ x / L ⱕ 3 results in the function f O for
off, as Ca共x / L = 0兲 also includes the air entrapped along a deflec- offsets
tor.

Far-Field Zone
fO = 冉 冊 冉 冊
x
L
− 1 F−2
o
ho
s + ho
−1.3
共sin ␸兲0.4, for 0 ⱕ f O ⱕ 0.2

The analysis of Ca共x兲 along the far-field zone 共III兲 indicates that 共5兲
the values tend to decrease in the streamwise direction for ␸
= 12°, remain constant for ␸ = 30°, and slightly increase for ␸
= 50°. For given ␸, all respective curves similarly decrease or Fig. 5共a兲 shows Cb共f O兲 for all tests conducted with offsets, indi-
increase for x / L ⬎ 3, indicating an insignificant effect of the other cating good agreement. Tests with approach flow Weber numbers
varied parameters. The concentration curves were therefore de- Wo = Vo / 共␴ / ␷ho兲0.5 ⬍ 140 were excluded, as they underestimate
fined as a function of ␸ with respect to Ca共3L兲 at the beginning of Cb due to scale effects, with ␴ = surface tension and ␷ = viscosity
zone 共III兲 according to Eq. 共3兲. The values of Ca共x / L ⱖ 3兲 were of water. The trend of Cb may be approximated using the hyper-
found with R2 = 0.86 as bolic tangent function, i.e., Cb = 1 − tanh关4.8f O0.2兴, with R2 = 0.74, as

JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010 / 363

J. Hydraul. Eng., 2010, 136(6): 360-367


Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Bottom air concentration Cb in spray and reattachment zone


共II兲 for offsets: 共a兲 semilogarithmic streamwise development Cb共f O兲
of test data and Eq. 共6兲; 共b兲 comparison of measured and with Eq. 共7兲
computed values Cb共3L兲 at x / L = 3 Fig. 6. Exponents m and n versus sin ␸, with 共쎲兲 and 共䊏兲 mean
values from Table 1

冉 冉 冊 冉 冊
Cb = 1 − tanh 4.8
x
L
−1
0.2
F−0.4
o
ho
s + ho
−0.26
共sin ␸兲0.08 , 冊 offsets 共t = ␣ = 0兲, as summarized in Table 2. In the test series
for 1 ⱕ x/L ⱕ 3 共6兲 relative to Fo for example, all curves Cb关共x / L兲 − 1兴Fno collapse for
␸ = 50° with the exponent n = −2.5, for ␸ = 30° the exponent is n
Eq. 共6兲 starts at P 共f O = 0兲 with Cb = 1 decreasing then in the = −1.5, and n = −1.0 for ␸ = 12°. Accordingly, the exponent n is
streamwise direction. At the end of zone 共II兲 the values of Cb expressed as a function of ␸ as
become very small, unless self-aeration occurs 共Kramer 2004兲.
The length L, which is specified in the companion paper 共Pfister
n = − 1 − 共1.5 sin ␸兲3 共8兲
and Hager 2010兲, is influenced by the same governing parameters
as Cb in Eq. 共6兲, such that the effect of each parameter varied in As regards to the isolated effect of ␣, all curves Cb关共x / L兲 − 1兴
the model on Cb共x兲 must account for both, Eq. 共6兲 and L. A ⫻共tan ␣兲m collapse for ␸ = 50° with the exponent m = −1.0, the
sensitivity analysis indicates that Cb共x兲 mainly increases with Fo exponent is m = 0 for ␸ = 30°, and m = 0.5 for ␸ = 12°, such that
and ho, while the effects of s and ␸ are marginal. The parameters
Fo, ho and partially ␸ are usually given for a certain spillway by m = 0.5 − 共1.5 sin ␸兲3 共9兲
design and economic reasons. The efficiency of an offset aerator
can therefore only marginally be increased by s, such that it Both Eqs. 共8兲 and 共9兲 are shown in Fig. 6. The negative values of
should exclusively be used for large Fo ⬎ 6, as typically occur in n increase with ␸, indicating an amplified effect of Fo on the
bottom outlets or at chute ends. It can be concluded that offsets bottom air transport particularly for steep chutes, i.e., ␸ ⬎ 30°. In
operate generally less effectively in terms of bottom air entrain- contrast, the effect of m on 共tan ␣兲m is dual: The exponent is
ment than deflector aerators because the turbulence-generating positive for ␸ ⱕ 30° and negative for ␸ ⬎ 30°. The bottom air
deflector is absent 共Ervine et al. 1995兲, and L is relatively short. transport is reduced on “flat” chutes when applying steep deflec-
The concentration Cb共3L兲 at the end of zone 共II兲 is obtained by tors and increases on steep chutes with steep deflectors. This phe-
inserting x / L = 3 into Eq. 共6兲 as nomenon results from the jet impact angles at P: steep deflectors

冉 冉 冊 冊
on flat chutes generate a large impact angle of the aerated jet on
−0.26
ho the cute bottom, known to cause significant deaeration 共Chanson
Cb共3L兲 = 1 − tanh 5.5 F−0.4 共sin ␸兲0.08 , at x/L = 3
o
s + ho 1988; Ervine et al. 1995兲. In contrast, the effect of ␣ on the
impact angle is reduced on steep chutes while the disintegration
共7兲
process is amplified. The curve Cb关共x / L兲 − 1兴共tan ␣兲m is indepen-
The Cb共3L兲 values derived from Eq. 共7兲 and those measured are dent of ␣ for ␸ = 30° because then m = 0 and 共tan ␣兲0 = 1. How-
compared in Fig. 5共b兲, with R2 = 0.87. The test data were linearly ever, L depends on ␣ such that the bottom air concentration at the
interpolated between neighbor values if no measurement was absolute, not normalized, location x remains a function of ␣.
available at exactly x / L = 3. The two parameters ho and s were combined into ho / 共ho + s兲,
Aerator flows with deflectors and with offsets s ⱖ 0 are influ- i.e., the identical normalization as for offsets. The effect of this
enced by all varied parameters. For these, an effect of ␸ on both relative drop height is smaller than for offsets, resulting in a re-
Fo and ␣ was found. Hence, the effects of Fo and ␣ on duced exponent of ⫺0.2. Furthermore, no relevant effect of t on
Cb关共x / L兲 − 1兴 cannot be described with the same exponent as for Cb关共x / L兲 − 1兴 was found. As the effect of ␸ is included in the

Table 2. Exponents n and m Describing Effects of ␸ on Cb关共x / L兲 − 1兴Fno and on Cb关共x / L兲 − 1兴共tan ␣兲m. nv= No Value Available
Angle Parameter Exponent series 1 Exponent series 2 Exponent series 3 Mean
␸ = 50° Fo ⫺2.5 ⫺3 nv n = −2.5
tan ␣ ⫺1 ⫺1 ⫺1 m = −1.0
␸ = 30° Fo ⫺1.5 ⫺1.5 ⫺1.5 n = −1.5
tan ␣ 0 0 0 m = 0.0
␸ = 12° Fo ⫺1 ⫺1 nv n = −1.0
tan ␣ 0.5 0.5 0.5 m = +0.5

364 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010

J. Hydraul. Eng., 2010, 136(6): 360-367


Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. Bottom air concentration Cb in spray and reattachment zone


共II兲 for deflectors with or without offsets: 共a兲 semilogarithmic stream-
wise development Cb共f D兲 of test data and Eq. 共11兲; 共b兲 comparison of Fig. 8. Normalized trend of Cb / Cb共3L兲关⌽IIICb兴 in far-field zone 共III兲
measured and with Eq. 共12兲 computed values Cb共3L兲 at x / L = 3

⬎ 30° mainly ␣ influences Cb共x兲. In engineering practice, the


exponents n and m, the individual effects of the varied parameters spectrum of the typically applied geometrical parameters is 5 °
on Cb关共x / L兲 − 1兴 derived from all test series results in the function ⬍ ␣ ⬍ 12°, 0.1 m ⬍ t ⬍ 0.3 m, and s ⬍ 1 m. Note that s is defined
f D for deflectors along 1 ⱕ x / L ⱕ 3 as as chute bottom elevation difference not considering local aerator

冉 冊 冉 冊
grooves to optimize the transverse air flow in the jet cavity, and
−0.2
x ho that s and t are defined perpendicular to the chute bottom 共Fig. 1兲.
fD = − 1 Fno共tan ␣兲m , for 0 ⱕ f D ⱕ 0.3
L s + ho At the end of zone 共II兲 the values of Cb共3L兲 may be derived
from Eq. 共11兲 by substituting x / L = 3
共10兲
Fig. 7共a兲 shows the measured data Cb共f D兲 in a semilogarithmic
plot. All data essentially collapse, such that the resulting trend

Cb共3L兲 = 1 − tanh 5.7 F0.25n
o 共tan ␣兲0.25m 冉 冊 冊
ho
s + ho
−0.05
, at x/L = 3

follows the hyperbolic tangent function Cb = 1 − tanh关4.8f D0.25兴. Ac- 共12兲


cordingly, the bottom air concentration starts at P 共f D = 0兲 with
Cb = 1 and drastically decreases in the streamwise direction. The The measured values of Cb共3L兲 are compared with Eq. 共12兲 in Fig.
values tend to Cb → 0 at the end of zone 共II兲, if self-aeration is 7共b兲 with R2 = 0.81. There are few values Cb共3L兲 ⬎ 0.01 as tests
absent. The general trend of Cb for deflector aerators may there- with “large” values of Cb共3L兲 rarely reach the end of zone 共II兲 due
fore be expressed as to the limited model chute length.

冉 冉 冊
Cb = 1 − tanh 4.8
x
L
−1
0.25
Far-Field Zone

冉 冊 冊 −0.05
In the far-field zone 共III兲, i.e., between x / L = 3 and the self-
ho aeration point 共Kramer 2004兲, continuous deaeration occurs for
⫻F0.25n
o 共tan ␣兲 0.25m
, for 1 ⱕ x/L ⱕ 3
s + ho all tests with reduced rates as compared to zone 共II兲. Three as-
共11兲 pects resulted from the data analysis concerning Cb共x / L兲: 共1兲 a
semilogarithmic plot of the data indicates an exponential decrease
Eq. 共11兲 slightly differs from Eq. 共6兲, but the data analysis indi- of Cb; 共2兲 the detrainment gradients vary with Fo; and 共3兲 the
cated that two different normalizations were needed to describe effects of the other five investigated parameters are small. The
the distribution of Cb for aerators with deflectors or offsets. The concentration curves were normalized with their initial values
1 , 200 Cb data are compared with Eq. 共11兲 in Fig. 7共a兲 Cb共3L兲 from Eq. 共7兲 for offset aerators, or Eq. 共12兲 for deflectors
共R2 = 0.79兲. Again, tests with Wo ⬍ 140 were excluded for the and defined as a function of Fo. The following equation results
derivation of Eq. 共11兲. 共R2 = 0.91兲:

冉 冉 冊 冊
Considering the combined effects of all relevant parameters on
Cb共x兲 from Eq. 共11兲 and from L specified in the companion paper x
Cb = Cb共3L兲 exp − 8.5 − 3 F−1.5 , for 3 ⱕ x/L ⱕ 9
requires a distinction between flat chutes with approximately ␸ L o

ⱕ 30° and steep chutes with ␸ ⬎ 30°. An increase of the param-


共13兲
eters Fo, ␸, ␣, ho, t, and s causes an increase of the bottom air
concentration. Their effect is small on flat chutes, whereas on Eq. 共13兲 may be applied for aerators consisting of deflectors, off-
steep chutes large values of Fo and ␣ strongly increases the bot- sets, and combinations. Fig. 8 compares the measured values of
tom air transport. Cb / Cb共3L兲 as a function of ⌽IIICb = 关共x / L兲 − 3兴F−1.5
o with Eq. 共13兲.
The approach flow conditions are usually determined by the Few data with ⌽IIICb ⬎ 0.2 tend to values above those computed
spillway geometry and the discharge, such that they cannot be with Eq. 共13兲. These relate to tests with relatively small ho and L,
varied for unregulated chutes. Furthermore, the chute angle typi- and were located beyond x / L = 9. Therefore, they could be influ-
cally follows the topography and economical considerations. enced by incipient self-aeration increasing the air concentrations
These parameters may generally not be varied to optimize an slightly.
aerator. Hence, only the aerator geometry characterized by ␣, s, The Cb共x兲 values derived from Eq. 共13兲 and those provided by
and t may be selected to design an aerator. An optimization may Kramer 共2004兲 are almost identical, given that Kramer’s equation
be necessary for low Fo and should involve ␸: For ␸ ⱕ 30° a was applied: 共1兲 in the far-field zone 共III兲 defined in the compan-
combination of large ␣, s and t values is effective, while for ␸ ion paper and 共2兲 if the initial value Cb共3L兲 is derived from Eq. 共7兲

JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010 / 365

J. Hydraul. Eng., 2010, 136(6): 360-367


or Eq. 共12兲 of the present investigation instead of using Kramer’s Science Foundation 共SNF兲, Grant Nos. 200021-101548 and
estimation. Again, Fo is the governing parameter regarding the 200020-113448.
bottom air transport, beside a long L and a high initial value of
Cb共3L兲.
Notation
Conclusions
The following symbols are used in this paper:
This hydraulic model research accounts for the two-phase flow C ⫽ air concentration;
pattern and the air concentration distribution downstream of chute Ca ⫽ average air concentration;
Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

bottom aerators. The efforts were concentrated on a quantification Cai ⫽ incipient air concentration;
of the air entrainment at the aerator and on its detrainment further Ca共3L兲 ⫽ average air concentration at x / L = 3;
downstream. The entrainment process was primarily character- Cb ⫽ bottom air concentration;
ized by the commonly used air entrainment coefficient ␤. Two Cb共3L兲 ⫽ bottom air concentration at x / L = 3;
main parameters were considered for air detrainment: 共1兲 average
Fo ⫽ approach flow Froude number;
air concentration Ca and 共2兲 bottom air concentration Cb. The
f D ⫽ normalization function for deflector aerators;
relations of both air concentrations were described with the gov-
f O ⫽ normalization function for offset aerators;
erning parameters along the reattachment and spray zone 共II兲, and
g ⫽ acceleration due to gravity;
the far-field zone 共III兲. Equations are provided allowing to deter-
ha ⫽ atmospheric pressure head;
mine these two concentrations versus the streamwise coordinate
ho ⫽ approach flow depth;
and the boundary conditions, including the approach flow condi-
h p ⫽ bottom pressure head;
tions, the aerator geometry, and the chute bottom angle.
hs ⫽ cavity subpressure head;
The main parameters influencing the air transport downstream
hv ⫽ vapor pressure head;
of chute aerators are the approach flow Froude number Fo, the
K ⫽ cavitation index;
deflector angle ␣, and the chute bottom angle ␸. These govern the
L ⫽ jet length;
transport process, especially at relatively large values. However,
m ⫽ exponent of ␣;
Fo and ␸ are usually fixed for a spillway, such that primarily ␣
n ⫽ exponent of Fo;
can be selected for a sufficient air transport. Furthermore, the
QA ⫽ air discharge;
following conclusions were drawn concerning aerator design:
QW ⫽ water discharge;
• Aerators with deflectors generate higher air concentrations
R ⫽ coefficient of determination;
than do offsets;
s ⫽ offset height;
• Aerators mainly increase Ca, while Cb is only marginally af-
t ⫽ deflector height;
fected;
V ⫽ chute flow velocity;
• A large deaeration gradient is observed near point P, especially
Vo ⫽ approach flow velocity;
close to the chute bottom;
Wo ⫽ approach flow Weber number;
• Minimum Fo for efficient air entrainment by aerators are Fo
x ⫽ streamwise coordinate;
⬵ 6 for offsets, Fo ⬵ 5 for deflector aerators with ␣ ⬵ 6° and
z ⫽ coordinate perpendicular to chute bottom;
Fo ⬵ 4 for ␣ ⬵ 11°. Offsets are limited therefore to large Fo, as
zu ⫽ upper water surface at C = 0.90;
typically found in bottom outlets or at chute ends;
␣ ⫽ deflector angle;
• Steep deflectors are more efficient than those with small values
␤ ⫽ air entrainment coefficient;
of ␣, especially on steep chutes with ␸ ⬎ 30°; and
␤ M ⫽ maximum air entrainment coefficient for hs
• Aerators on steep chutes ␸ ⬎ 30° operate more efficient than
= 0;
those on flat chutes.
␳ ⫽ density of water;
Chute aerators may be designed based primarily on their bot-
␴ ⫽ surface tension of water;
tom and secondarily on their average air concentration curves, for
␷ ⫽ viscosity of water; and
which equations are proposed herein. The effects of the aerator
␸ ⫽ chute bottom angle.
geometry or the approach flow condition may be derived from the
resulting air concentrations. A general guideline for approach Subscripts
flows with “small” Fo is: on flat chutes with ␸ ⱕ 30°, a deflector
a ⫽ average;
aerator is efficient provided 10° ⱕ ␣ ⱕ 12°, 0.20 m ⱕ t ⱕ 0.30 m,
b ⫽ bottom;
and 0.50 m ⱕ s ⱕ 1.00 m, while these numbers are reduced for
i ⫽ incipient;
“steeper” chutes.
o ⫽ approach flow; and
The present results apply to 5.8ⱕ Fo ⱕ 10.4, 12° ⱕ ␸ ⱕ 50°,
3L ⫽ at x / L = 3.
0.1ⱕ 共t + s兲 / ho ⱕ 2.1, 0 ° ⱕ ␣ ⱕ 11.3°, and hs ⬇ 0, except for Eq.
共1兲 with a larger range. It is recommended that an optimized
aerator should have a sufficiently large air supply system to avoid
considerable energy losses in the air supply duct, which would References
reduce the aerator efficiency or even cause air duct choking.
Chanson, H. 共1988兲. “Study of air entrainment and aeration devices on
spillway model.” Ph.D. thesis, Univ. of Canterbury, Christchurch.
Acknowledgments Chanson, H. 共1989兲. “Study of air entrainment and aeration devices.” J.
Hydraul. Res., 27共3兲, 301–319.
The writers gratefully acknowledge the support of Prof. Dr. H.-E. Ervine, D. A., Falvey, H. T., and Kahn, A. R. 共1995兲. “Turbulent flow
Minor, ETH Zurich, and of Prof. Dr. A. Schleiss, EPFL Lausanne. structure and air uptake at aerators.” Hydropower and Dams, 2共4兲,
The first writer was financially supported by the Swiss National 89–96.

366 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010

J. Hydraul. Eng., 2010, 136(6): 360-367


Falvey, H.T. 共1990兲. “Cavitation in chutes and spillways.” Engineering draulics, Hydrology and Glaciology 共VAW兲, ETH, Zurich 共in Ger-
monograph, 42, USBR, Denver. man兲.
Gaskin, S., Aubel, T., and Holder, G. 共2003兲. “Air demand for a ramp- Pfister, M., and Hager, W. H. 共2010兲. “Chute aerators. I: Air transport
offset aerator as a function of spillway slope, ramp angle and Froude characteristics.” J. Hydraul. Eng., 136共6兲, 352–359.
number.” Proc., 30th IAHR Congress, D, Thessaloniki, Greece, 719– Pugh, C. A., and Rhone, T. J. 共1988兲. “Cavitation in Bureau of Reclama-
724. tion tunnel spillways.” Proc., Int. Symp. on Hydraulics for High
Kökpınar, M. A., and Göğüş, M. 共2002兲. “High-speed jet flows over Dams, Beijing, 645–652.
spillway aerators.” Can. J. Civ. Eng., 29共6兲, 885–898. Rasmussen, R. E. H. 共1956兲. “Some experiments on cavitation erosion in
Koschitzky, H.-P. 共1987兲. “Dimensionierungskonzept für Sohlbelüfter in water mixed with air.” Proc., Int. Symp. on Cavitation in Hydrody-
Schussrinnen zur Vermeidung von Kavitationsschäden 关Design con- namics, National Physical Laboratory, London, 1–25.
Downloaded from ascelibrary.org by UNICAMP - Universidade Estadual De Campinas on 12/01/20. Copyright ASCE. For personal use only; all rights reserved.

cept for chute aerators to avoid cavitation damage兴.” Mitteilung, 65, Rutschmann, P. 共1988兲. “Belüftungseinbauten in Schussrinnen 关Chute ad-
Institut für Wasserbau, TU Stuttgart 共in German兲. ditions for air entrainment兴.” Mitteilung, 97, D. Vischer, ed., Labora-
Kramer, K. 共2004兲. “Development of aerated chute flow.” Mitteilung, tory of Hydraulics, Hydrology and Glaciology 共VAW兲, ETH, Zurich
183, H.-E. Minor, ed., Laboratory of Hydraulics, Hydrology and Gla- 共in German兲.
ciology 共VAW兲 ETH, Zurich. Rutschmann, P., and Hager, W. H. 共1990兲. “Air entrainment by spillway
Pan, S., and Shao, Y. 共1984兲. “Scale effects in modeling air demand by a aerators.” J. Hydraul. Eng., 116共6兲, 765–782.
ramp slot.” Scale effects in modelling hydraulic structures, H. Kobus, Skripalle, J. 共1994兲. “Zwangsbelüftung von Hochgeschwindigkeits-
ed., 4共7兲, Technische Akademie, Esslingen, 1–5. strömungen an zurückspringenden Stufen im Wasserbau 关Forced aera-
Peterka, A. J. 共1953兲. “The effect of entrained air on cavitation pitting.” tion of high-speed flows at chute aerators兴.” Mitteilung, 124, Tech-
Proc., IAHR-ASCE Joint Conf. Int. Hydraulics Convention, Minne- nische Universität, Berlin 共in German兲.
apolis, 507–518. Wilhelms, S. C. 共1997兲. “Self-aerated spillway flow.” Ph.D. thesis, Univ.
Pfister, M. 共2008兲. “Schussrinnenbelüfter: Lufttransport ausgelöst durch of Minnesota, Minneapolis.
interne Abflussstruktur 关Chute aerators: Air transport due to internal Wood, I. R. 共1991兲. Air entrainment in free-surface flows. IAHR hydraulic
flow structure兴.” Mitteilung, 203, H.-E. Minor, ed., Laboratory of Hy- structures design manual 4, Balkema, Rotterdam, The Netherlands.

JOURNAL OF HYDRAULIC ENGINEERING © ASCE / JUNE 2010 / 367

J. Hydraul. Eng., 2010, 136(6): 360-367

You might also like