Professional Documents
Culture Documents
Concrete International
Fig. 2 - 1941 cavitation damage in the Arizona spillway Fig. 3 - 1983 cavitation damage in the Nevada spillway
tunnel at Hoover Dam. at Hoover Dam.
!
mary method of preventing cavita- It appears therefore, that the tion. Surface erosion due to sedi-
tion damage was to eliminate the most effective means of prevent- ment in the flow and not cavitation
cause by decreasing the allowable ing a recurrence of the 1941 inci- was the probable cause for the uni-
size of surface irregularities. dent would be to maintain the formity of the damage.
Experience and time have since tunnel lining as smooth as possi-
ble. As an additional precaution Areas out of tolerance upstream
shown that efforts should not be Mr. J. L. Savage, Chief Design of the elbow were repaired by
geared to constructing smooth flow Engineer, suggested that some patching, followed by grinding if
surfaces, but rather to improving means be devised to introduce air low, or by chipping, bushing, then
spillway designs to prevent cavita- into the spillway flow with the grinding if high. The first 200 ft
tion damage. Although there are expectation that the air first, (60.96 m) of the rough invert down-
many factors involved in choosing a would act as a cushion between stream of the elbow, was bushed,
spillway design, it is ironic that one the high-velocity water and the sand blasted, and stoned smooth
solution to preventing cavitation tunnel lining, and secondly, that with mortar. No repairs were made
problems may lie in designing large the same air would aid in reliev- beyond 200 f t (60.96 m) down-
uniform roughnesses in the form of ing any subatmospheric pressures stream of the elbow. A similar in-
which may occur along the sur-
steps on a spillway face. face of the tunnel invert. spection and repair of the invert
surface was then carried out in the
Nevada spillway.
Cavitation at Hoover Dam Although aeration would be A research study was conducted
Hoover Dam, completed in 1935, studied, the immediate repair of the in 1945 to investigate the design of
has a hydraulic height of 576 f t damage focused on reconstructing a aeration devices for the Hoover
(175.6 m) and 50 ft (15.24 m) di- smooth flow surface. The eroded spillway tunnels. Aerators were de-
ameter tunnels passing through each areas were repaired with concrete.' signed and tested at locations near
abutment, referred to as the Ari- After curing, t h e surface was the tunnel entrance and start of the
zona and Nevada spillways (Fig. 1). stoned, then ground smooth with a vertical elbow. Air concentration
The Arizona spillway was first small terrazzo machine. was measured along the elbow in-
operated for four months in 1941 at During the repair, the entire tun- vert to evaluate aerator perform-
an average discharge of 13,555 ft3/s nel invert was inspected to identify ance. The study concluded that the
(383.8 m3/s). Following operation, all misalignments and irregularities air induced into the flow did not re-
inspection revealed a hole, approxi- in the flow surface. Several addi- main along the invert in sufficient
mately 30 ft wide, 115 ft long and tional misalignments, small rock quantities to ensure cavitation pro-
45 f t deep (9.14 x 35 x 13.7 m), pockets, and calcite buildups were tection of the invert along its entire
eroded through the concrete lining identified between the tunnel en- length.
of the tunnel invert (Fig. 2). trance and the end of the vertical Although aeration of the flow
The inspection revealed a misa- elbow. The invert surface was continued to be studied, these and
lignment of the tunnel invert in the eroded downstream of the elbow, other experiences with cavitation
form of a hump a few feet up- leaving a nearly uniformly rough damage due to surface roughnesses
stream of the damaged area. This surface. This section of tunnel was formed the basis of concrete speci-
misalignment was generally ac- assumed t o have been damaged fications for high-velocity flows.
cepted as the source of the cavita- prior to spillway operation when the For nearly four decades after the
tion that initiated the damage. A tunnel section downstream of the repairs, there were no significant
subsequent Reclamation report2 elbow had also been used for diver- flows down either tunnel at Hoo-
stated: sion flows during dam construc- ver. During this time, the ability to
May 1991 59
'4
~tv'gincpltunnel surface 9m Ramp
Aerator or
downstream offset
Concrete International I
change direction. Structure toler-
ances (i.e., line, grade, etc.) control
the maximum allowable offset of
gradual roughnesses.
These surface tolerances have
been incorporated into Reclama-
tion's design process for hydraulic
structures. A structure's required
flow surface tolerance is deter-
mined by the minimum cavitation
index of the flow, shown in Table 2.
Experience has shown that cavita-
tion damage may be prevented
P Fig. 5 - Aerator operating at 20,000 ft3/s(566.3 m3/s)in the hydraulic model of when these guidelines are followed.
i; Hoover Dam. A minimum cavitation index of be-
tween 0.2 and 0.1 requires that spe-
cial aeration devices, such as air
or aerator. The .offset downstream Surface tolerance slots, be included in the design. Any
prevented water from filling the air specifications structure with a cavitation index of
groove during large free-flow dis- less than 0.1 must be redesigned to
Reclamation set up a team of engi- reduce its cavitation potential. Dif-
charges. It was determined that only neers to review the design and con-
one aerator was needed at Hoover ferent spillway concepts or realign-
crete tolerance specification process ment of the preliminary design
Dam. Further details of aerator de- for traditional smooth-surfaced hy-
sign are detailed in Reference 4. should be considered for any struc-
draulic structures. This team rec- ture with a flow sigma less than 0.2.
A prototype test of the effective- ommended that flow surface toler- The search for better spillway de-
ness of a similar air slot was con- ances (in the direction of flow) be signs for high-head dams often in-
ducted at Glen Canyon Dam just treated separately from surface fin- cludes stepped spillways.
prior t o installing the air slot at ishes and be limited to three levels,
Hoover. An area of eroded con- as presented in Table 1.
crete near the elbow of the Glen Surface roughnesses are treated Stepped spillways
Canyon tunnel spillway was not re- differently by the tolerance specifi- Stepped spillways have become a
paired prior t o the t e ~ t i n g The
.~ cations depending on whether they very popular method for such dam
successful outcome of these tests are abrupt or gradual in nature. The releases as those at Reclamation's
and other experiences with aerators abrupt roughness tolerance speci- Upper Stillwater Dam7because their
provided a strong basis for relying fies a surface offset dimension be- compatibility with roller-compacted
on aeration of the Hoover spillway low which any type of roughness concrete construction techniques
flow, instead of surface tolerance, (abrupt or gradual) is acceptable. produces low additional cost for the
as the main defense against cavita- Roughnesses which exceed the spillway. Traditional types of
tion damage. abrupt tolerance limits are evalu- smooth spillways with steep chutes
For Hoover, the decision was ated against the gradual roughness and expensive energy dissipators are
made t o relax surface tolerances tolerances. still being used with roller-com-
and leave unrepaired many surface The gradual roughness tolerance pacted concrete dams,8 but cavita-
roughnesses that exceeded previous is based on the maximum slope tion is still a concern with these
surface specifications. The lessons (height:length) of the roughness. An structures. Designers have been re-
learned through years of experience acceptable gradual roughness is de- luctant to use stepped spillways be-
and research with structures like fined as a roughness with a maxi- cause of the complex flow charac-
Hoover demonstrated the need to mum slope that is less than speci- teristics associated with them.
update Reclamation's concrete sur- fied by the gradual tolerance. The Several site-specific studies of
face specifications, standardize the more gradually a flow surface var- steeply sloping stepped spillways
design process for smooth spillways ies, the less likely it is to cause cav- have been c o n d ~ c t e d With
.~ each
with high-velocity flow, and to de- itation of the flow. Therefore, the study, the upper limits of unit dis-
velop innovative spillway designs maximum height or depth of a charge have been extended - q =
that are less susceptible to damage gradual roughness is not important, 154 ft3/s (14.3 m3/s) for Milltown
by cavitation. only how quickly the flow must Hill Damlo- and our knowledge of
I May 1991
the flow hydraulics have increased. ability t o predict the cavitation The pressure profiles indicate vi-
However, there is a lack of general characteristics of uniform steps. sual characteristics of the flow very
design guidelines for their use. well. The profiles are plotted over
Reclamation's initial experience Test facility the two steps representing the up-
with stepped spillways on high con- per, middle, and lower (toe) sta-
crete dams came largely from hy- The present laboratory facility con- tions for the appropriate overtop-
draulic model studies conducted for sists of two sloping flumes. One ping head (Fig. 8 and 9). Each pie-
Upper Stillwater Dam.' One objec- flume is designed for studying zometer tap location (No. 1 through
tive of the model studies was to de- 2H: 1V sloped stepped spillways 22) is indicated on the steps. The
termine if cavitation would be a typical of high embankment dams. pressure profiles indicate the jet im-
problem. Numerical analysis of a The second flume, sloping a t pact on the downstream end of the
smooth spillway of similar geome- 0.8H:lV, is used to study concrete step tread (high pressure) and an
try predicted a cavitation index of dam stepped spillways. The flume area of reduced pressure in the off-
0.2 would occur at high discharges. facilities are used to study the flow set below the pitch line of the steps.
Extensive tests were conducted to hydraulics of different step geome- An eddy forms in the offset area.
observe the flow field and measure tries and the effects of increasing For the step geometry, if cavitation
pressures and velocities. The re- depth of flow on the steps. forms, it is most likely to occur in
search found no indication that Initial investigations - Initial in- the fluid shear zone near the top of
cavitation would occur for the vestigations have concentrated on the eddy, well off the boundary.
stepped geometry. studying the hydraulics of step de- Also on the graph is a plot of the
Reclamation is currently con- signs on the 2H:lV sloping flume. approximate flow depth over the
ducting extensive research on the They have consisted of determining stepped spillway. The depth varied
hydraulic design of stepped spill- discharge capacity, flow depths, little down the stepped face and is
ways for high dams. The formation pressure profiles on the face of the shown as a single line regardless of
of cavitation due to the step geom- steps, and velocity profiles, and measurement station. Comparison
etry is still a concern. Although the documenting visual flow character- of this flow-depth measurement
complex flow that develops on a istics with increasing flow depth. with the pressure profiles clearly
very rough surface such as steps Data are recorded for scaled over- shows the different pressure zones
prevents easy analysis of cavitation topping heads (12: 1 geometric scale) associated with the jet impact and
potential, research has shown that of 5, 10, 15, 20, 25, and 30 ft (1.52, the return eddy. The eddy rotation
steps offer several positive aspects 3.04,4.56, 6.10,7.62,and9.14m). provides significant energy dissipa-
for preventing cavitation damage: The flow down the stepped face for tion as the flow passes down the
5 ft (1.52 m) of overtopping head is steps.
Research results" suggest that shown in an overall view in Fig. 6
a uniformly rough surface can have and close up in Fig. 7.
a lower cavitation potential than an Conclusions and
isolated roughness of the same ge- Pressure measurements - Pres-
ometry due t o reduced velocities recommendations
sures were measured on both the The following conclusions can be
and wake effects. vertical and horizontal faces of the
Large surface roughnesses pro- drawn from these investigations:
steps. Three measurement stations
mote self aeration of the flow. are located along the flume slope Constructing smooth concrete
Steps form large offsets away for gathering pressure profiles on surfaces on high-velocity spillways
from the flow direction. This inhib- the face of the steps. The upper sta- has not prevented cavitation dam-
its cavitation from residing on the tion (Steps 15 and 16) is located 30 age in many instances as natural
boundary. ft (9.14 m) below the crest, the mid- roughening of surfaces with age is
Step geometries can be de- dle station (Steps 47 and 48) is 94 ft difficult to monitor and control.
signed to prevent subatmospheric (28.65 m) down, and the lower sta- Concrete surface specifications
pressures on the surface. tion (Steps 79 and 80), near the toe, for many existing high-velocity
Reclamation currently has an ex- is 158 ft (48.16 m) down. At each spillways do not sufficiently reflect
tensive research program underway station two steps are instrumented, construction and maintenance limi-
to further study the characteristics each with 11 piezometer taps, for a tations.
and benefits of stepped spillways. total of 22 taps per station and 66 Research on stepped spillways
The objective of the research is to overall. The mean pressure from suggests that they may be viable al-
develop design criteria to determine each piezometer tap was recorded ternatives t o cavitation-prone
optimum step geometry based on with an IBM-compatible computer. smooth spillways.
hydraulic forces and energy dissi- The same computer was then used Based on these conclusions, the
pation. This research also will pro- to plot the resulting pressure pro- following recommendations can be
vide further data to improve the files over the steps at each station. made:
Concrete International
0.00 -1
0.00
... ,... ... , . .. , . , , ,
1.00 2.60 3.00 4.00 5.00
, , ,
6.00
, , , ,
7.00
, ,p . 9.60
8.00 . I 0.00
DISTANCE ALONG STEP (FT)
Fig. 6 - Overall view of overtopping embankment flume
facility. Fig. 8 - Pressure profiles of stepped spillway face at
three stations along slope: H = 5 ft (1.52 m); q = 33.96
ft3/s (3.1 5 m3/s).
.*
-
. * .
.*
-
'.,*
.
-
::
-
* * '
Roughness type
Abrupt-offset, in. -
1-0(25mm) -' .
'
Gradual<lope
-. . 1:4:
3*
.'-
L
"
-
':,: -Q-
% <
:e-r
,-.,
0-5(12mm); ':
\
,^
- '
,
r
: pg - -+-*<,
--<*-
.+
""i
v.
T3 , :7v .*
A .
0.25 (6mm) 1: 16 -:--?:
+-.--
May 1991
Surface tolerances alone should 5. Houston, K. L.; Quint, R. J.; and ACI member Kath-
not be relied upon to prevent cavi- Rhone, T. J., "Overview of Hoover Dam leen H. Frizell is
Tunnel Spillway Damage," Proceedings, In- an hydraulic engi-
tation damage t o a spillway struc- ternational Conference on Hydropower,
ture. Waterpower '85, V. 2, ASCE, Las Vegas, neer with the Hy-
Consideration must be given t o Sept. 25-27, 1985. draulic Structures
the cavitation potential o f a struc- 6. Frizell, K. W., "Spillway Tests at Glen Section, Hydrau-
Canyon Dam," U.S. Bureau of Reclama- lics Branch, Re-
ture early in the design process.
tion, Denver, July 1985. search and Labo-
Preliminary designs must be ana- 7. Houston, Kathleen L., "Hydraulic ratory Services Di-
lyzed using current spillway-flow Model Studies of Upper Spillway Dam vision, U.S. Bureau of Reclamation,
computer models which compute Stepped Spillway and Outlet Works," Bulle- Denver, Colorado. She has worked ex-
flow-cavitation indices as a func- tin No. REC-ERC-87-6, U.S. Bureau of tensively with developing the hydrau-
Reclamation, Denver, Oct. 1987.
tion o f stationing and discharge. 8. Houston, K. L., "Stepped Spillway De- lic design of stepped spillways, per-
Design alternatives, including sign with a RCC Dam," Proceedings, 1987 forming model studies of Upper Still-
stepped spillways, should be con- ASCE National Conference on Hydraulic water Dam, Milltown Hill Dam, and
sidered when the cavitation indices Engineering, Williamsburg, Virginia, Aug. McClure Dam, and performed the
are less than 0.2. 1987, pp. 122-127. model study of the aeration devices
9. Houston, Kathleen L., and Richardson, for the tunnel spillways at Hoover
Alan T., "Energy Dissipation Characteris- Dam.
References
- - - - tics of a S t e p ~ e dS ~ i l l w a yfor an RCC
1. Concrete Manual, 8th ed., U.S. Bureau Dam," ~nternationai~ ~ r n d o s i uon m Hy-
of Reclamation, Denver, 1981, pp. 359-364 draulics for High Dams, Beijing, China,
and 371-375. Nov. 1988. Brent W. Mefford
2. Bradley, J . N., "Study of Air Injection 10. "Final Model Study Results for Mill- is Head of the Hy-
into the Flow in the Boulder Dam Spillway town Hill Dam Spillway and Stilling Basin," draulic Structures
Tunnels-Boulder Canyon Project," HYD- U.S. Bureau of Reclamation, Denver, Aug. Section at the U.S.
22, 1990.
186, U.S. Bureau of Reclamation, Denver,
11. Falvey, Henry T., and Mefford, Brent Bureau of Recla-
Oct. 1945. mation's hydraulic
W., "Flow Surface Tolerances on the Invert
3. Walter, Donald S., "Rehabilitation and
Between Stations 20 + 20.08 and 46 + 61, Ne- laboratory in Den-
Modification of Spillway and Outlet Tunnels ver, Colorado. He
and River Channel Improvements at Hoover vada Spillway Hoover Dam, Arizona," U.S.
Bureau of Reclamation, Denver, Apr. 14, has a MS dearee in
Dam," MS thesis, Department of Civil En-
gineering, University of Colorado, 1957.
1986. civil engineGing from Colorado State
4. Falvey, Henry T., "Cavitation in University. He has been active in cav-
Chutes and Spillways," Engineering Mono- itation research and participated in re-
graph No. 42, U.S. Bureau of Reclamation, Received and reviewed under Institute publi- vising Reclamation's concrete surface
Denver, Apr. 1990. cation policies. specifications.
May 1991
Volume 13
Number 5
CONCRETE INTERNATIONAL
5 President's memo
10 Letters
69 Queries on Concrete
70 Standardization
71 Spanish translation of
synopses
72 Meetings
73 Calendar of
International Concrete
Conferences
75 News
78 Chapter news
80 On the move
Check your Concrete
Quotient
Products & practice
Calls for papers 13 GLASSGOLD, HOFF ELECTED; NEW DIRECTORS TAKE OFFICE
Bulletin Board ACI installs new officers at Boston, Mass. convention.
Advertisers' index 20 A YEAR OF ACHIEVEMENT
Bookshelf Annual membership report for 1990.
Membership -
May 1991 3