HYDRAULICS BRANCH {

,oe ;
OFFICIAL FILE COPY

FI.lrl»
AW, ;;, · fI
• "'
U ITED STATES
\(J ..

RETURN PRO�T . NT OF TH[ 1NTERIOR

WHEN BOAAO�D
BUR EAU OF RECLAMATION

H YDRAULIC MODEL STUDIES FOR SPIL LWAY

. CHANNEL AT ST EWART MOUNTAIN DAM-­
SALT RIVER PROJECT, A RIZONA
\

Hydraulic Laboratory Report No. Hyd.- 24 7

RESEARCH AND GEOLOGY DIVISION

BRANCH OF DESIGN AND CONSTRUCTION

DENVER, COLORADO

SEPTEMBER 13, 1948 · ...

 FOREWORD

The hydrau lic model studies of the spillway channel at Stewart

Mountain Dam were conducted by the personnel of the Bureau of

Reclamation, in the hydraulic laboratory on the campus of the

Colorado State College of Agriculture and Mechanic Arts, at Fort

C.ollins � Coloradoo These studies were begun in May 1935 and com­

pleted in March 1936, and were initiated under the direction of

E. W. Lane, but finished under J, E. Warnock.

These studies were made in the hydraulic laboratory, in co�junc­

tion with computations by Mr. Raymond A. Hill of Leeds, Hill, and

Jewett, Los Angeles, California, and Mr. D. C. McConaughy of the Dams

Division; however, such computations are not included in this report.

A report of these studies was begun in 1937 under the direction

of J. W. Ball, and a preliminary draft was written by R. R. Buirgy.

This report could not be completed at that time because of more urgent

work in the laboratory •

•
 CONTENTS

SUMMARY

1. Extension of the Stewart Mountain Dam Spillway

2. The Proposed Spillway Channel

J. Summary of the Tests

L1. Results and Conclusions

THE PRELIMINARY STUDIES

5. The 1:100 Model of the Original Design of Spillway Apron

6. Initial Studies on the 1:100 Model

7. Studies of Crest by 1:50 Model

8. Studies of the Revised 1:100 Model

9. Redesign of Channel of 1:100 Model

10. The Use of False Floor and Fillet To Imp rove Flow

DEVELOP1IBNT OF THE FINAL DESIGN

11. Design and Construction of the 1:50 Model

12. General Performance of Model and Discharge Capacity

13. Improvement of Flow Over Crest and Into Channel

14. Spillway Capacity
15. Comparison of Flow Through End Gates

• 16. Use of Wing Wall on Left End Pier

17. Effect of Training Wall in Channel
18. Effect of Wing Walls on Water Surface at Right Channel Wall

19. Revision to False Floor

20. Effect on Discharge of Excavating Bank Upstream From the Crest
 21. Measurements of Water Surface in Channel

22. Use of Circular Wing Wall at End Pier

23. Use of Fillet in Channel Downstream From The Crest

, 24. Use of Long Fillet Alongside Left Channel Wall

25. Modification of Sloping Bank Upstream From Left Pier

26. Gate Operating Schedule

27. Pressures in Bucket Do�mstream From Crest

28. Revision of Fillet in Channel and Modification of Right Wall

29. Streamlined Wing Wall Replaced, Pressures in Water Below Crest

30. Determination of Height of Channel Walls and Use of Overhanging Sea

Wall on Right Channel Wall

31. Water Surface Profiles Across Channel

32. Calibration of Model for Free Discharge Over Crest, 'and With All Gates

at S.ame Openings

33. Simplification Right Wing Wall of Spillway

34. Calibration of Flow in Right End Gate, Gate No. 1

35. ,Calibration of Flo,w in Center Gates

36. Calibration of Flow in Left End Gate
37. Velocity Distribution in The Channel
38. Splash Over Spillway Walls

39. Discharge Coefficients of Final Design

40. Discharge Coefficients With Spillway Channel Removed

41. Discharge Over Crest With Piers and Channel Removed

 UNITED STATES
DEPARTI�IBNT OF THE INTERIOR
BUREAU OF RECLAMATION

Branch of Design and Construction Laboratory Report No. 247

Reasearch and Geolo gy Division Hydraulic Laboratory
Denver, Colorado Compiled by: F. c. Lowe
September 13, 1948 Reviewed by J. E. Warnock
Subject: Hydraulic Model Studies for Spillway Channel at Stewart
Mountain Dam--Salt River Project, Arizona.

SUMMARY

1. Extension of the Stewart Mountain Dam Spillway

In 1936 a concrete discharge channel, 265 feet wide and 4SO feet

long, was built at Stewart Mountain Dam to replace the natural sidechannel

below the ogee crest of the spillway. (Figure 1.) This new channel was

curved to conform to the terrain and super-elevated to assure satisfactory

operation at any flow condition. The design was tested and improved by
hydraulic model studies as narrated in this report.

The Stewart Mountain Dam (Figure 2) is located about 30 miles east

of Phoenix, Arizona. Completed in 1930 by the Salt River VaJ:ley Water

Users Association, it was the third dam built in the Salt River Canyon
below the Roosevelt Dam, and is dovmstream from Horse Mesa and Mormon Flat
Dams. The Stewart Mountain Dam consists of a central arch between massive

concrete abutments with gravity sections on each side to close the gap

between the abutments and the sides of the canyon� The height from the
• stream bed to maximum water level is 119 feet, and a power head of 116 feet

is available for operating a 17,SOO-horsepower hydroelectric plant. In

addition to power production, the dam is used for irrigation s torage,

flood control, and to re-regulate releases from the dams above to fit

irrigation requirements downstream.

 Figure 1

A. The spillway channel built in 1936.

B. The original side-channel spillway.

SPil.LWAY ALTERATIOIIS AT STnlART MOUNT.AIR DAM

Figure 2
 The spillway, located on the left gravity section of the dam, is
I

a 267-foot ogee crest, at elevation 1506, placed between two training

walls, which extend upwards to the walkway on top of the dam. Eight
• piers 3 feet wide are placed on this crest to form nine openings to
accommodate 'nine radial gates 27 feet wide by 23 feet high. The

water surface in the reservoir may thus be .regulated between elevations

1506 and 1529, although during floods the water could rise to elevation
1535 before overtopping the dam. The topography downstream from the
crest forms a natural sidechannel (Figure 1.B), and 'When the dam was
built, further provision for returning the water to the riverbed was
considered unnecessary/ A small wing wall was built to protect the
east abutment of the main arch from scour by the spillway discharge.
At small discharges, there was a concentration of ·flow from this
natural sidechannel around the east abutment of the arch indicating
that large discharges would endanger the structure regardless of the
protective wing wall. Any undercutting at the abutment would be
especially serious, for the main arch of the dam was designed with
high unit stresses and the safety of the dam depends upon the stability
of the abutment. Moreover, a large discharge over the spillway
would wash soil and rocks from the banks of the sidechamel into the
riverbed, increasing the tailwater elevation at the powerhouse and
reducing the available power head. It would even be possible to wash
a bar across the river and flood the powerhouse.
2. The Proposed Spillway Channel
An improvement of the spillway channel was necessary to carry the
water along the hillside and into the,river at a point further downstream.

2
In the design of the new channel, the original sidechannel was filled,

and a new channel, 265 feet wide, was extended 450 feet to a lmoll
over which the water could flow (Figure 3) o It was contemplated that

erosion of this knoll would be negligible once the overburden w as

removed, for the main rock was of granite in fairly good conditiono
To reach this knoll, the.channel had to be turned through an arc of

about 32° 0 At the same time, the drop was from elevation 1506 at the
crest to approximately elevation 1475 at the end of the channel.

The channel was designed for a maximum flood of 140,000 second-feet,

and for smaller discharges when the flow was unbalanced, such as operation

with one gate only. The design was further complicated by the fact

that the channel turned through an arc, thus the floor had to be
superelevated to maintain a uniform·water depth to avoid the construction

of unreasonably high sidewallso A channel, superelevated to handle a

discharge of 140 ., 000 second-feet, may be unsatisfactory if a single gate

is opened, for the water may flow to the lower side of the super­
elevated curve, strike and overtop the lower sidewalL While it was
possible to prepare an analytical design of a spillway channel with the
degree of superelevation necessary for a maximum flood, the actual

success of the spillway operation was uncertain because of the above

conditions. A hydraulic model study V'Ja.S necessary to check and improve

the proposed spillway channel.

The Stewart Mountain Spillway v,ras completed in 1936, shortly after

the tests were completed. Since then, to the knowledge of the laboratory

personnel, no large discharges have passed through the spillway, the

3
 FIGURE 3

El. 1537. O··-, ... ···Motor generator house

'- 5
Normal w.s. /o
El.1529.00 ·.,

..original ground line at <f_

,.

,.� "I/

Ope_nin9 at . /
low point m pier····· '-..3'Min. into
2· 0
' SECTION A -A firm roe/<.
'

-Backfill Original ground line·

o. ,
' � X
J'- <:?6'
\ .-coo
/
· ·······--·--/32'·6''····· --· ·+· ········/32'.6"·········
1 .·· Backfill
-'---- �
.,.
-,_,-
\
c·-����······ �}��� __ ..... 265 ·• o ''. ......................

-------------------
···i
! ..Backfill
·'2::-� Spo1/ from ¾\&n�<<*-d:Zwn X
', <j' ,;:> s-'
-..I):' --- .... ___ �
excavation � --;;;\\\V,k<C:&« ::..._�'i-..
t x;So
Spoil from excavation .:'
.Si,':'¾
' ·.s- SE.CT/ON B - B SECTION C-C
XOo ''
'
·;-0
• .?-- '
+I
'
_..$/ope !J: I _ •.. -Original ground line
.. Present surface of
�-:,w,A\,1-&:"K"..::0,�
-----
-
--- --- - ,·Slope 1}: I Slope 2: 1, ./ sand and gravel
- ----"4

El. /40�;��-�.�--i�-------- ------- __ /

�\'f::#{(Tr.:--
�- - -------- - - ----------------- - 2ss '-o'! .. ---- -------------- --- -- __ ,...

SE.CT/ON O ·0 14-0'±············ .,.

SECTION E.-E.

CHANNEL ELEVATIONS
f<......... ·····-13216'! ....... -····<><
, ·'": f--4:......--Back fill
5 R··
LEFT RIGHT
STA.
TOP OF GRADE ELEVATION L, .
u�J�-"-*
TOP OF
f\J
ON�
€ /;.POINT AT WALL WALL
Ir- -f ::��·,
WALL AT WALL ij_POINT
0,50 14-98.82 1496.70 1493.30 1487,/2
1522,06 1485.29 1503.36
····-spoil from
1497.93 1495.25 14-9 /.24 1484.52
1519. 72 1479.66 /492.64
excavation
1,50 15/7.24-
1496.66 1493.51 1489.05 1482.05 1478./3 l48Z63
I ,OO

t'I 2, 00 1495.04- 1491.53 1486.74 1479.63

1514.50 1475.16 1484.73
SECT/ON F-F
i..:; 2 +50 1511.74 1493./0 1489.29 1484.30 1477. 26 1470.03 14-B/.82
3,oo 1508.70 1490.86 1486.80 1481.74 1474-.91 1455.33 1480.33
� 3 ,so /505.28 1488.20 1484.07 1479.05 1472.57 1463.93 1478.93
4,00 1501 .62 1485.22 1481.09 1476.24 1470.24- 1462.62 /477.62
4+50 /48/.86 1477.85 1473.28 /467.90 146/ .37
···----�---- 5+00 /478.2 1474.4 1470.3 1465. 6 1460.2
5,50 1474.J 1470.8 /467.3 14-63.5 1459.3
6+00 1469.8 146 7 ./ 1464.3 1461.4 1458.4
6+50 1465./ 1463.2 1461.3 1459.5 1457.7

I
I
I
I
I
I

.. El. !535.00
I
I

Normal w.s. ···E/.1530.00

THIS DRAWING SUPERSEDES DWG, 25•0•7/3
I El. 1529,00··. 0£PAATMENT OF THE INTERIOR
I
BUREAU OF RECLAMATION
SALT RIVER PRO-.JECT - ARIZONA
PLAN I
I
/ STEWART MOUNTAIN DAM
/
/ SPILLWAY
GENERAL PLAN
50 0 50 100 ELEV. G-G SEC. H - H
SCA LE OF FE ET
normal release of water being through the powerhouse o It was reported

that a small discharge or leakage through the gates flowing dovm

the right side of the channel has undercut the foundation to some
extent at the downstream end, and that the granite knoll dmvnstream

from the channel has decomposed at the surfaceo

J. Summary of the Tests

The tests were made with two models, one built on a scale of 1:100

and the other a scale of 1:50, and the testing program may be considered

as divided into two parts, the preliminary studies and the development
of the final design.

The preliminary studies included four tests. The first test

was on a 1:100 model of the original design with a curved superelevated

channel in which general flow characteristics were observed o A gate

operation procedure was studied." discharge capacity, water surface and

velocities in the channel measured, and the erosion on the knoll down­
stream observed. The second test was on a 1:50 model of the crest
without the channel downstream. In that test the crest itself was

studied by measurements of pressure and discharge. The third test was

on the 1: 100 model with channel downstream alined on a compoun:d curve

in which the superelevation was too steep. The fourth test was with

a similar channel� alined on a compound curve. In that test, the flow

in the channel was studied, the hydraulic losses measured, and a false

floor and fillet placed alongside the right, or inside, wall to reduce

the water depth and a fin which formed at the wall.

The final design was developed on the 1:SO model. The channel was

revised and alined on a spiral curve which became the final design.

4
Flow conditions were observed and the discharge capacity measured. The

flow over the crest and into the channel was improved, the t_ests being

_concerned with the effects of wing walls at the upstream end of the

piers, fillets in the pier offsets, a false floor· in the channel

alongside the right wall, and modification of the topography upstream

from the crest and determination of the height of the channel walls.

The development of the final design included (1) modification of the

sloping bank on the left side of the approach channel to the spillway and to

� streamlined-shaped wing w�ll at the head of the right ehd pier to

decrease contraction of flow at the end piers, thereby increasing the

dischapge capacity, (2) fillets in pier off sets to smooth the flow, (3)
a fillet on the channel floor alongside the right wall to decrease the

depth of flow at that point, and (4) an overhanging lip, or sea wall at the
top of this wall to divert splash at that wall back into the channel. · The
flow through the various gates was studied and a gate-operating schedule

established. When the final design was developed, discharge measurements

were made to establish rating curves·for various conditions of operation.
The final tests consisted in observation of splash over the channel
walls, and discharge measurements to note the effect of the channel
upon the capacity and-measurements of the velocity distribution in the
channel.

4. Results and Conclusions

While the original tests indicated the flow conditions which would
occur in the prototype _, it became apparent that the 1�100 model was too

small for the development ,of the final design. This was especially
noticeable in the velocity measurements in the channel which were

5
inconsistent and the losses larger than anticipated. From this 1:100

model it was concluded that erosion of the overburden on the knoll

downstream from the channel would have no serious consequences.

Subatmospheric pressure conditions would occur ·on the crest;

however, no changes were made since this structure was already

constructed.

When it was found that the superelevation of the channel was

too great, it was realized the basic problem was to design a curved
channel superelevated in such a manner as to keep the flow at a constant

depth across the section and that it was necessary to account for

frictional losses. Because of the friction factor, the model studies

had to be interpreted with caution. If any large d�fference in relative·

friction were to exist between the model and prototype, the superelevation

as indicated by the model would be inadequate. No slope or friction

corrections were made in the design of the model, for the very uncertain­

ties involved in friction corrections render them as unlikely to

represent the true prototype conditions as does the undisturbed model.
Moreover, the channel was comparatively short and the losses were not
large.
The spiral channel of the 1:50 model was used for the.final
design. A number of minor alterations, to improve the flow appearance
and the discharge capacity was necessary. The principal trouble was
the flow was too deep at the right channel wall requiring an undesirably

high wall. A high fin formed along this wall immediately downstream

from the crest. These conditions were alleviated largely by placing

6
a fillet on the channel floor along the wall, and constructing an

overhanging lip or sea wall at the top. Splash still occurred in the

rough flow downstream from crest to an estimated height 40 feet above

the right wall. It is uricertain as to what the actual splash condition

will be in the prototype structure.

I mprovements to increase the discharge capacity were based upon

reducing the contraction of flow at the piers to increase the effective

crest length. A streamlined wing wall or pier nose was placed at the
front of the right pier and tests on various types indicated this

was probably an optimum design of wing wall considering effectiveness

and construction -costs. From the �mprovements made, it appeared that the

maximum capacity was increased about two percent.

THE PRELIMINARY S'TUDIES

5. The 1:100 Model of the Original Design of Spillway Apron

A channel 271 feet wide and 429. 5 feet long, beginning at the

spillway crest, was contemplated in the original design of the spillway

apron for Stewart Mountain Dam. The centerline of this channel was
normal to the crest at Station 0+00 and extended in a straight line

to Station 0+17. 97. It then curved on a radius of 714.5 feet through

an arc of 33 ° to Station 4+29.50. This channel was superelevated
to give uniform depth of flow across the entire section with all gates

opened an equal amount and with the water surface in the reservoir at
elevation 1529·. This superelevated section was faired into the ogee
crest at approximately elevation 1492 at the top or outer edge of the
curved channel and at elevation 1475 at the lower edge. As the channel
dropped in elevation, the angle of superelevation increased to compensate

for increased velocities.

7
 A model of this original design, on a scale of 1�100, was built

in the Colorado Agricultural College Experiment Station Hydrau�ic

Laboratory, Fort Collins, Coloradoo This first model included a

• portion of the main arch, the spillway crest, and the adjacent topography

extending 325 feet upstream and 1,050 feet.downstream from the crest

(Figure 3ol) o The crest, gates, channel, and main arch of the dam were

built of sheet iron, the piers of redwood, and the topography shaped with,
sand. Aluminum paint on the structural parts protected them and

improved the photographic qualities. Portions of the sand topography

which represented solid rock was stabilized with cement, while that
p ortion representing overburden and soil was of loose sand to· show the

effects of erosiono The forebay,'representing a portion of the

· reservoir upstream from the crest, was sufficiently large to perrrd. t

water to approach the crest in a uniform mannero To further assure

this, a baffle was placed between the forebay and the source of supply
in the flume upstreamo The downstream section of the model was extended
to include sufficient length of channel to determine whether material
eroding from the knoll below the spillway would form a bar across the
riverbed.·, The model terrrd.nated :with an adjustable tailwater board.

Several features of this model were not similar to t he prototype

due to the lack of detailed information in the early stages of model
studies-. The·spillwaJ'." cr�st section upstream from the crest centerline

was 3 feet too wide, the piers extended 7-1/2 feet too far upstream and

the radius of the radial gates was 25 feet (prototype) instead of 20

feet. These differences, as shovm by dotted line on Figure J.lC were

8
 FIGURE 3. 1

.. Direction of fl ow
· entering water

Toilwoter control ·

Main arch
of dam ___
Toilwofer gage \

L_·_ · · -
_:;;-- - -=- -----

[.OJ ··Foce boy float goge A- P L A N

" '
E I J 5 3 5 --

:: Channe l !!

u
Surface slobilized

\ !!
Surface stabilized
with cement Crest ot excovotion
:: floor .. 11

1 6 gage .•·
� ron bents

Sand · Sand
I
- ---....- t\ i"ll" Angle iron b a s e -­
·· ·- 2 , 3 Slope

8 - S E C T I O N A-A

,Oa
I ·i75 !
!5 0 2s �1
Note . Pomt of pier
extends beyond
prototype pter.
Surface stabi lized Channel floor of 16 gage iron
with cement

PLAN

_ _ ___,
Top of p1 �r__E!-!?_3_5,

,,' .,.,"
Top of gate E!:!?.2_ �:·,
- - - - - - - - - - , -;.,

l,' :,,"
_San d . Redwood piers---- --­
..--1 11 Prototype -L--------­
iron base Model-······· -- ----
-­
Sheel i r o n lin i n g Crest Er. J 50J?:·:c . · "
..,,"111,?',
• JI ":f'-,

D - S E C T IO N B - B
r,20 Slope
2 0 goge iron

.
� 1 <1 1 5
C - CREST, P I E R A N D G A T E D E TA I L S

S A LT RIVER PROJ E C T A RIZONA

1 4tO \c-
0 -· ----c,'"'0 0cc
00c--- --!c
_ -----,,:;;- 0 --
00a_':0s;;0:;;- ----.a
, ,-;¢ ,os;;o,---,,�
,,;;; oo,ooo S T E W A R T MOUNTAIN D_AM
OtSCHARGE IN C.F.S
HYDRAULIC MODEL STUDIES O F SPILLWAY CHANNEL
TA I L WA T E R C U RVE 1 , 100 M O D E L OF O R I G I NAL D E S I G N
not significant in the preliminary studies because the flow in the

channel downstream would not be materially affectedo The height of the

channel walls was not to scale, but was extended to elevation 1529, as the
proper wall height was to be determined in the model studies .

_ Instrumentation of this model included a float gag� in the fore­

bay to measure the reservoir water surface, a point gage to measure

taiiwater, a movable point gage to observe the water surface on the

crest and in the superelevated channel, and a single leg pitot tube

to measure velocities. Discharge was measur,ed by a weir in the flume

upstream from the model . The movable point gage and pit�t tube were

mounted on a channel bar, which in turn was set on two horizontal,

pa,rallel bars fastened to the sides of the flume in which the model ·,

was placed.

6. Initial Studies on the 1 : 100 Model

The tests on the 1 : 100 model included (1 ) a study of gate operation

procedures with respect to flow conditions in the channel downstream,

(2 ) determination of spillway cap�city, (3 ) measurement of water surface

and velocity distribution in the channel, (4) general obseryations of flow

over the spillway crest, and (5) an estimate of the probable extent of

erosion in the river channel downstreamo In these tests, the water

surface of the reservoir was held at or above elevation 1529, the top
of the radial gates, for it was assumed that water would rise to this

elevation before the gates would be opened . The degree of superelevation

in the channel and this anticipated operating procedure was established

for flow at velocities acquired with the water surface at elevation 1529.

9
 Two methods of spillway gate operation were studied: (1 ) open all
gates the same amount, and (2 ) open individual gates completely, the

number depending upon the discharge. With all gates open an equal

amount, the flow was comparatively uniform in the channel , regardless

of dischar�e (Figure L.a) . The only undesirable condition observed was

the tendency to wash away the bank downstream f rom the right training wall .

By operating individual gates, the flow through the channel appeared

satisfactory with combinations of three, five, and seven gates open,

although it was preferable to use the gates on the right end of the

crest (Figure 4C1 ). When gates on the left end were opened , some flow
crossed the channel and piled against the right wall (Fieure 4D) .
With all gates wide open, the . spillway discharged ::J,.OO , QOO second­

feet when the water surface in the reservoir was at elevation 1529. 7 .

As the discharge increased , the water surface .c�ntinued to rise. At

a discharge of 139,200 second-feet ., the water reached the top of the

parapet on the reservoir, elevation 1535 . Thl,s was some 10, 000 second­

feet short of the design capacity of 150 , 000 second-feet. By operating

individual gates with the water surface at elevation 1529 , thJ'.'.ee gates

completely opened would discharge approximately 2 5 , 000 second-feet, five

gates 50, 000, and seven gates 75, 000 second-feet. All these measurements

were qualitative and no discharge coefficients were determined in this

test for the shape of the crest of the model was not similar to · that
of the prototype as previously mentioned ; moreover, it was found that

the 1 : 100 model was too small ·to obtain precise discharge measurements.
The tests included measurements of the water surface and velocity
to asqertain the applicability of the theoretical design of the

10
A. The model, all gates Jertially open . B. Erosion of river bed after discharge
Di scharge 75, 000 c .f . s . Reservoir at 139, 200 c . f . s .
at elevation 1529 , 0 ,

c. Five gate s at right si de open D. Five gate s at left s i de open .

Discba.rge · 50, ooo c . f . s . Reservoir
Discharge 50, 000 c . f. s . , Reservoir
at elevation 1529 . 0 . at elevation 1529 . 0 . (Note water
piling again st right wall . )

-I='"
HYDRAULIC MODEL STUD.JES OF ,STEWART M>Ul'f.I'AIN DAM SPn.n,AY
1 : 100 MODEL OF ORIGINAL DESIGN
 1/
superelevated channel� These measurements were made at two radial
sections of the channel (Figure 5-A) , oy placing the pi tot tube and

point gage at seven positions across each section (Figure 5 B) . Water

surface profiles for discharges of 50,000 , 100,000 , and 139, 200 second­

feet are shovm on Figure 5·Bo The depth of water is the minimum at

the inner edge of the curve, the right side, increasing gradually to

a maximum at the outer edgeo This condition was theoretically correct

because the velocities, a function of the energy head above the channel

floor were less at the higher elevations at the outer edge of the curve
and it follows that a greater depth is required at the outer edge to

maintain unif�rm discharge across the sectiono The velocity distribution

at the 14 positions is slrnvm in Figure 5C for discharges of 25 , 000 , 50 ,000 ,

75 ,000 , 100 ,000 j and 139, 200 second-feeto Vvhile some variation exists ,

the velocities are , roughly, independent of discharge. Theoretically ,

this is correct except for a discharge of 139, 200 second-feet when

the water surface is at elevation 1535 instead of 152 9 0 With an ideal ,

nonviscous fluid , and with the water surface in the reservoir at elevation
1529 , the velocity at any point may be expressed as 1/2 gh, where h is the
drop from elevation 1529. Actually, fricti onal losses over the crest

and down the channel reduced the effective head and velocity. By

estimating such losses , the design section obtained curves for discharge

1 / For a review of the design of a superelevated spillway channel ,

see Laboratory Report Hyd. 74 "Flow of Water in Superelevated Channels
at Velocities both Above and Below Critical , " by T . G . Owen. See also
"Design of Spiral Spillway Chutes, " by Raymond A.. Hi+l and D . C . Mcconaughy
in Civil Engineering , November 1945 , Volume 15 , No. 11 , Page 499 .

11
 �-- �
FIGURE 5

�
I .l

-
.0,50 ooo
-- ,..,
dU

-
150
0 1 39,200 . o,
'{ 1 -
--"-'I-- 1----
,_
_l I

I
1475
- c::::=:::

1450 EtiC (2 (3') 4l (5 <6'> /7

->�1\.,- 47.7 •->l 0,7 2 '�1 139.9'-, 186.3'·'1 226 .5'• 270.C ·� I<
- ·o AN E M ASUR ED F OM T S Pi°'NT 271.C-·
r r \ I t

�-
I
S E CT I O N "A"

,r,139,Jon
- -- -
�- -- --- _ ::.,,
.0!100 Joo o,5�,000
,

-
1500
- ->-

1475
_1-::::-;::

.,
c----
2 13, 4' 5"\ 1 (6) 7
t:'".�'ljSTANCE 54.4' 9'l5' 26r " ,;271(� .
1460 1 l1i9'- 2181't•
I 8.9•
MEA '°' FD,, .. r• I' Pl INT
A. LO C A T I O N OF S ECT I O N S A-A S E C T I O N "B"
AN D B-B I N C H A N N E L
B. WAT ER S U R FA C E S AT SECTIO N S A-A AND B-B

1595
1500
1500
1495 .- E levotion of f loor
o f chonnel. 1495
1490 POSITION 7

::1J lfi l 1 1
1490
16 55 50 16 50 55 60

: :IJ. I UMJ,N I : :IJ I 1 00 1 I

POSITION 6

: tt I Utl121\K I
POSITION 5

: lJ I I WA I :
:lJ I l� tJlj r�\ 1 1
POSIT I ON 4

POSITION 3

: : IJ. 1 1 Li!J 1 1
: :I.J I I IJJJA t_,j i POS I T I O N 2

: : u:1 11 I.W I
::t1 I I 1tU11 fj
16 20 25 30 35 40 45 50 55
VELOC I T I ES IN FEET PER SECOND

S ECTI O N A - A
60 65 70
POS I T I O N I
: ::1-+jl 1 1 1 Utl
16 20 25 30 35 40 45 50
V E L O C I T I E S IN F E E T PER S E C O N D
S ECTI O N B s B
55 60

C . R E L AT I O N O F V E L O C I T Y. T O D E P T H

NOTES S A LT R I V E R P R OJ ECT - A R I Z O N A
_ I. Fo r d i s c ho r ge s 2 5 , 0 0 0 , 50,000 , 75,000 on d 100,000 C. F s .
w o t e r sur foce i n reservoir ot elevation 1529 .00. For 1 3 9 , z O O
STE W A R T M O U NTA I N D A M
.-C. Fs. w o t e r s u r f a ce ot elevat ion 1 5 3 5.00 . H Y D R A U LI C MODEL ST,U D I E S OF S P I LLWAY C H A N N E L
2 . Woler su rfoce elevations obtained by point goge. Velocities
M E A S U R E M E N T S O F W A T E R S U R FA C E A N D
by s i n g l e leg pitot tube
V E LO C I T I E S I N O R I G I N A L I : 1 0 0 M O D E L
 ) .

of 100 , 000 second-feet as shown in Section A-A , Figure $C o The measured

velocities follow the same general pattern as the computed velocities,

but are less indicating greater losses in the model . An investigation

suggested that the losses i� the model were so large that · they were
questionable . In fact, it was believed that the model data should lie

close to the computed values. Although the values obtained from model

data for different discharges follow a logical pattern, they appear

widely spread . The only logical explanation was that the model

was so small that the frictional losses_ in it were comparatively large.

The flow over the crest contracted at the ends toward the center .?
rendering a portion of the spillway crest useless o If these entrance

conditions could be improved , the discharge might be increased f rom a

maximum of 1}9 » 200 second-feet to the desired 150 , 000 second-feet.

However ., any studies of this nature would have to be done on another

model with the crest and piers properly built.

One important factor to be shown by this model was the probable
erosion from the hillside to the riverbed downstream . In the proto­
type, the granite rock of the knoll , where the spillway discharges, was

covered with an overburden of soil and loose rocks o A conflict of

thought arose as to the manner of disposing of this overburden. If

i t were permitted to rernain .9 it would eventually be washed into the

riverbed. T he cost of removal woula not be increased .!) and this method
would be des:j..rable if a bar were not formed in the river· downstream

w hich would cause water to back into the, powerhouseo The model was run
at various discharges to ascertain the effect of erosion of the lmoll
upon the riverbedo The erosion patterns were similar ., becoming more

12
pronounced as the discharge increased D Figure 4B shows the erosion

downstream after a discharee of 139 .:, 200 second-feet. From this test,
it was concluded that a bar would not be formed which would seriously

retard the flow of the river below the powerhouse.

7. Studies of Crest by 1 : 50 Model

A spillway capacity of 150 3 000 second-feet was desired , but the

l ilOO model of the original design indicated that 139,200 second-feet

was the maximum obtainable. Since this model was too small to permit

precise measurements and the crest was not built correctly .:, the original

1 � 100 model was removed , and a l i 50 model of the spillway crest and
approach channel only was installed (Figure 6A ) . The piers and topo­

graphy upstream were included in this' model 9 but the regulating gates .,

the spillway channel ;, and . the topograph;y dovmstream were not included .
Such f'eatures were unnecessary, for the purpose of this model at this

stage of investigations was to study the properties of the crest itself,

the pressure distribution, the discharge coefficient ., and the extent

of contraction of flow at the end walls and around the piers. The

omission of the nin� radial gates was the equivalent of assuming that
tpey would always be ·operated in a wide-open position.

Piezometers were installed at the center of the crest to measure

pressures (Figure 6B) . The piers were removed to eliminate any effect
of contraction of flow. Pressure curves f or discharges of 10,200 .,

26, 100 , 49, 900 , 73,600, 100 ., 500 9 125, 500 _:, 155,000 , and 178,700 second­

feet are shown on Figures 6C and 6D . These curves represent the elevation

of the pressure on the crest compared with the elevation of the crest

itself , and indicate subatmospheric pressure at discharges greater thap

13
100 , 000 second-feet. Subatmospheric pressures on a crest of too great

1nagnitude are not desirable because air might be indra�m periodically

between the nappe of water and crest in such a manner as to cause

vibration, or .c avitation might occur with its attendant vibration and

pitting . The necessity for alteration of the crest to eliminate nega­

tive pressure was not indicated as they were not in excess of 15 feet
of water. Further investigation of pressures were not deemed necessary

although unfavorable negative pressures were anticipated in .the proto­

type not only ,rith the gates fully open, but also vlith the gates partially

open.

From the discharge coefficients obtained in this test, it was indi­

cated that the capacity of the spillway would be 150,000 second-feet

with the reservoir at elevation 153 5 with the piers removed to eliminate

the effect of contraction of flow. To study the crest further, calibra­

tions wer� made over a range of heads and discharges, for the follo"lJ\d.ng :

(1) without - piers, ( 2 ) with piers, and ( 3 ) with piers and with upstream
topography removed to change the approaci1 conditions . Without the piers,

the c apacit y of the spillway was 155 , 000 second-feet with the wq.ter sur­
face at elevation 1533. 6.

These figures have little significance unless the capacity is

expressed as a coefficient independent of th e crest length which may be
compared with a similar coefficient irith the piers in place.
For flow over a rectangular weir , similar to the Stewart Mountain
Dam Spillway, the coefficient of discharge, C , may be expressed as

C = where Q = discharge , L = length of the crest, and H = head

LHJ/2
over the crest o In an ideal situation, this coefficient would be a
constant, but in this case it increases with head (Figure 6E) o Several

factors cause this variation :: The crest shape, the contraction of the

water toward the center of the nappe away from the end walls or piers·,

the contraction of water away from the center piers , and the topography

upstream from the crestG At large discharges 9 the water was observed
to spring free of the end walls 9 both with and without the piers in

placeo This naturally reduced the effective cres� length o It follows

that a larger coefficient might be possible if this contraction were

elim inated o The effect of the piers was to further reduce the effective

crest length 3 and small coefficients were obtained with the piers in

place .ll as shown in Figure 6Eo The removal of topography upstream from

the crest did not materially affect the discharge capacityo This
suggested that a change of the upstream topography was not necessaryo

Bo Studies .of the Revised 1 : 100 Model

Upon completion of the crest studies, the 1 �·50 model was removed

and replaced by a second h:100 model of the eritire spillway and adjacent

topography." The crest was the same as used in the original 1 :-100 model .9
but the channel was revisedo It was alined on a compound curve extend­
ing to Station 4 + 50 .9 and set approximately five feet higher than the
channel in the original design (Figures 7A 9 B, and C ; . Instead of
extending the channel walls to the top of the gate 9 elE;lvation 1529, as

was done before, the right wall was on�y 14 feet above the chann�l floor
and the left or outside wall extended to elevation 1522. 28 . The spillway
crest, which was incorrectly built in the original 1 : 100 model, was not

changed , but new piers, properly proportioned 9 were used (Figure 7D) •

•
 �----�------- - ------------, 1 5 25
E X PLANATION
R E S E R V O I R EL. DISCHARGE SYMBOL
151 1.26 a = 10,200
------+---- -- - - -- --
- - -----1 1 520
1 5 1 5. 3 1 0=26, 1 0 0,

NOTE : Intermediote piers removed

1 5 1 9. 9 5 0=49,900
1523.67 0=73,600
1527.33
1 530.38
0=00,500
a =12s,s oo 1519.95··+�
,i'"""f" I I I ""
�
1515. 3 1· - -.
1533.58 0=155,000

I ���
1 536.04 0 =178,700 1 5 1 1.26·--

-�,
�� �
j 1510 �
w�
'- I :!

�-
NANUMETER·,,,

,._
't: 1505 ILi
�
_j A 7'
·.;,

/I 1 1 i.' \' �'- 's..� I 11 l''°o

,0 40 30 2. 0 io
P L A N PRESSURE IN FEET - PROTOT Y P E
D · PRESSU R E ON TOP OF C R E S T
C- P R E S S U R E O N UPSTREAM FAC E

V 'L'__
OF C R E S T

V
�J 0
I/1/p
-;;28

I/
26
1

I<,.
z
V
V,V
22

�r:
SECTION A·A §
PROTOTYPE SCALE IN FEET
1
1 I
1 �" I 11
g 20

MODEL SCALE IN INCHES

0 _...Va/V
L�
:: 1 8
A·D ETAILS OF MOD E L PIEZOMET E R LOCA T I O N S ,._
I
HOfUZ.DIST. FROM AXIS ELEVATION Dist BELOW CREST
V
� 16
PIEZ.

.,..,v:,V
No. P�OTO- FT. MODEL-IN. PROTO-FT. MODEL-INCHES z

,-!l_ES_T__E_L,1_5_ 0_�-99�
I - 2 .0 5 ' - 0. 4 9 2 � 1489.96 3.850"

, 2 - o. 7 2 7 " 1492.03 3. 3 5 3 '

/V
- 3.0 3' E 14
� 12

o Piers removed 1 crest leng th = 2 67.19'

- � . 2 4'
�-·20 Goge iron
+·--
, · ·-.,J_! .. .
3 - 1 . 2 5 7" 1494.24 2.02 2 · EX PLANATION

· � · - - -�
z

v v,
;oi: 4 - 7 . 45' - 1 . 7 8 6" 1496.45 2.292.

x Piers in ploce, c rest lengt h = 244.I '

� 10

"'"
·�:: .,,vV
/ ,,,
5 - 8 .00' - 1 . 9 2 0' 1497.84 1 . 95 8 "
r--

"
>

-��
6
7
- 8.0 0' - 1 . 9 2 0" 1499.66 1 . 52 2 ·
:: a Piers in place,but topography removed. r--

vv...
- 8 .0 0' - 1 . 9 2 0" 150 1 . 4 6 L.085"
0
8 - 7 .6 2' - 1 . 8 2 8" 1502.42 0.859.
w
9

��.l
- 6 . t 9' - 1 . 4 8 5" 1503.82 o.s23• X

-·-·--· ·--- --- - - 5. 652 ' . • . •••••••.••••

: 10 - 4. 37' - 1 , 0 4 9' 1504.93 0.257"
: ' II - 2.4 1 ' - 0. 5 79• 1505.7 1 0. 069 " ',V

: ;. ... ' ,r ·:. 2 �

12 0.00' 1506.00 0.000·
. .
0.0 00· 1
·,.:' : Note : Pr ototype cr e 13 t 2 . ! 5' t 0.5 1 6 " 1505.92 0. 0 l 9
n
2.,_,o 3.30 3.40 3.50 3.60 3.70 3.80 3.90 4.00
E = COEFFICIENT OF DISCHAR G E , ' c ' IN EQUAT ION o= C L H
'----------·· · equation ,,, " �:, 32' 1 14 t 4.23' t 1 . 0 1 5' 1505.68 0.077"

-"'- .,,
Mode l cre �� ;,;, determined . 15 t 7.0 9' t I. 7 0 2 • 1505.06 0.2 2 6 " E · D I SCHARGE C O E F F I C I E N T S
\
by poezom: e r !�cation s - · · .• _ 16 t 9 . 1 2' t Z. I 89" 1504...48 0.364"

"""'
17 + 13.05' + 3. 1 3 2" 1502.91 0.7 4 1 SALT R I V E R P R O J E CT - AR I Z O N A
···---16 Goge iron bents 18 t l6.75' t 4. 0 2 0" 1500.89 1.226
19 t23. 55' t 5 . 6 52" 1495. 95 2.4 1 2 ST E W A RT MOUN TAIN DAM
B- C R EST D E TA I L S 20 t32.45' t 7. 7 8 8 " 1487. t 6 4. 5 2 2
HYDRAU L I C MODEL STUD I E S OF SPI LLWAY
TEST 2 - CREST STUDIES ON 1°50 MODEL
 FIGURE 7

"

O t 50 1486.00 1472.00 1485. 1 5 1492 .83 149'?.39 1500.07 Constant

1 t oo 1484.69 1470.49 1483.09 1490.BB l495.77 149�.83 at
I t50 1483.27 1469.27 148 1 ,2 0 1489.93 1494.03 1497.44 1522.28
2 t 00 1482.28 1468.28 1479.43 1486.97 1492. 1 8 1495.80
2 t 50 1481.52 1467,52 1477.73 1485.02 1490.23 1494.0 1
3 t 0 0 1480.90 1466.90 1476 . 2 3 1 483.07 1488. l B 1492 .04
3 -t 50 1480.53 1466.53 1474.79 148 1 . 12 1486.03 1489.88

I I3 I6 I9 I!2
PROTOTYPE SCALE. IN FEET
0 25 50 75 100 4 t oo 140020 146s.20 1473.42 1419_ 1 6 1483.78 1401.sz
4 + 50 1480.00 1466.00 1472. 1 4 1 477.2 1 1481.45 1484.89
5 t 00
0
MODEL SCALE IN INCHES 1479.49 !465.49 1470.60 1474.96 1478.72 148 1 .9 7
5 t 50 1478.52 1464.52 1468.6 0 1472 .2 1 1475.43 1478.30
6 t 0 0 1476.B l 1462.8 1 1465.87 1468.67 1471.25 1473.63
..> 6 t 50 1474.77 1460.77 1462. 75 1464.63 1466.41 !468. 1 I

A-P L A N

.-·Top of left wa l l 14 feet

above floor Sand to represent
ave r burde n --- - ------ •. -� ·�
,. .., ,,-=-c:

Sand sta bilized with

cement to represent
bedroek ..•......... - -··· C- S E C T I O N B- B
B-SECT I O N A-A
PROTOTYPE SCALE IN FEET
20 40 60 80 100
I · ', , ', ,, ,,I
0

Q I 2 } 4 5 6 7 8 9 l0 11 12
MODEL SCALE IN INCHES

--Lei! wall Ei.1522.28

-,--
! i

!:-t' pl",;,.:'
Ei.1506· �
'-IC\I

l
'

1,
;

__J
I
I
/
; ELEVATI O N

I ' E- E X T E N SI O NS O R WING WALLS TO STREAM LI N E PIERS

I ;

l/
j'w
II
�

� / UNITED STATES
,
DEPARTM E N T O F T H E INTERIOR
BUREAU OF RECLAMATION
D - C R E ST, P I E R A N D GATE DETAILS AT it. SALT RIVER PROJECT ARIZONA

ST E W A R T M O U N TA I N DAM
H Y D R A U L I C M O D E L STU D I E S O F S P I LLWAY
T E S T 3 - R E V I S E D 1 • 1 00 M O D E L
This was necessary to further study the undesirable contraction of .

flow at the end pierso

Tests on this model consisted of r (1 ) observations of flow in

the channel at several discharges and various gate opening combin�tions,

(2 ) observation of the erosion of the knoll upon which the channel

discharged, ( 3 ) a comparison of the spillway capacity with that of the -

original l dOO model to show the effect of raising the channel floor, and

(4 ) installation of the several types of pier extensions to improve

the flow passing over the crest. The model was operated at discharges
of 2,5,000 , .50 , 000 , 7.5,000 , lOO,OOO, and 131, 800 second-feeto The water

surface was held at elevation 1.529 for the smaller discharGes and at

1.53.5 for the -ma.,"'Cimum di·scharge. The tests were made with all gates
equally open and with individual gates opened wide, the number depending

upon the discharge. With all gates open an equal amount, there was a
tendency for the water to shift to the right side of the channel overtopping
the 14-foot wall at discharges greater than 7.5, 000 second-feet (Figure 8A ) .

This indicated that the degree of superelevation was too great for the curva­

ture , and that the channel would not be satisfactoryo Another undesirable
condition was a high fin alongside the right wall irnmedi�tely downstream from

the crest (Figure 8B ) o Operation with individual gates wide open in groups

of three, five, and seven gave flow patterns in the channel similar to
those observed in the original design (Figures 4C and D) except for the

tendency of the water to flow against the right wall. The flow was

undesirable when the gates on the left end of the crest were opened.

Water crossing the channel piled against and overtopped the right wall.

16
A. The model , all gates open. Di scharge
100, 000 c . f . s . (Note water overtopping
right channel wall. )
B. Flow in channel below crest . Di scharge
100, 000 c . f . e . (Note fin at right channel
wall. )

C. Contraction o f flow at right wing wall. D. Condition o f flow with curved Wing Wall
Discharge 140, 000 second -feet. "C". Discharge 140, 000 second-feet .

:5!
HYDRAULIC IDDEL STUDIES OF STEW.ARI' M'.>UNTAill DAM SPILIMAY
1
CP
. TEST 3 - REVISED 1 :100 MODEL
 This overtopping of the channel walls could not be permitted in
the prototype beccl,use the earth fill alongside the wall would wash away

and the channel foundations undermined. In the model, loose sand,

representing overburden, was washed into the riverbed, not only from
the lmoll at the channel exit, but also alongside the channel wall almost

back to the crest.

The discharge capacity was 131, 800 second-feet with the water
surface in the reservoir at elevation 153.5 , while the desired maximum of

150 , 000 second-feet was with the reservoi r at elevation 1537. 4 . This was

less than that of the original model , whose capaci ty was 139,200 second-feet.

Tests to improve flov,r over the crest were necessary because water

tended to spring av-ray from the end piers , shortenin1:; the effective crest
length (Figure 8C ). This condition was especially bad at the right end
pier, in the deeper water near the main arch of the dam . The solution

appeared to be the use of some type of wing wall upstream from the end
, pier to suppress this contraction. To study the problem, five curved
wing walls or pier extensions were proposed (Figure 7 E ). All of these

designs appeared successful in that the contraction was eliminated to a

large degree. Of the five tested, wall 11 C 1 1 was considered most practical
(Figure 81) ) . However, the improvement obtained was not as much as was

desired , and a further test was made by extending a wall upstream from the
right end pier about 50 feet, set at an angle of about 15 ° with the

pier centerline . With this arrangement, the water surface was almost

level i n passing through the gates .

17
9. Redesign of Channel of l g lOO Model

· The channel used in the revised 1 : 100 model was unsatisfacto ry·­

mainly because the degre e of superelevation was so great that water piled

against and overtopped the 14-f oot right wall . The channel appeared to

have b e en designed for higher velo ci ti es than those actually occur.ring .

A new channel was built , on a com:Pound curve as before , but us ing radii

of 490 and 1 9 140 feet to the channel c enterline (Figure 9A ) . The ups tream

end of the channel floor was approximately at the s ame elevation as that

of the previ ous model .? but the downs tream end was approximately 4 o 5 feet

lowero In thi s model , the ·top of the channel walls was at elevation 1530 ,

but a line was drawn on the right wall at elevation 1490 to represent the

proposed top of the walL The sh ort cutoff wall b e tween _the main arch

and the cre s t was extended upstream until it was flush with the leading ,

edge of the piers .

The model was operated by opening all gates an equal amount , and by

opening individual gates with water surface in the re s ervoir held at

elevation 152 9 ·f or all discharges les s than 100 .9 000 s e cond-feeto With

all gates opened an equal amount, the- flow was spread uniformly across

the channel ., being somewhat deeper at the right side . Using individual

gates ., better flow co;iditions appeared when those on the right side were

opened . When , gates on the left side we re opened the fl ow crossed the

channel and piled against the right wall o An unde s irable fin appeared

alongside the ri ght wall (Figure 8B ) o

A s tudy was mad� to estimate the losses in the channel by plotting

1
a hydrauli c gradient comparing the results with losses found in the

original designo In the original design ,, the lo sses were calculated

18
 FIGURE 9

--� - _; _--: ....

l _:

' ''

\/,
', /

PROTOTYPE SCALE JN FEET

•·I " 'l""I

100 �o 100

J 2 IO � 6 � 12
MODEL SCALE IN INCHES

'"
P R O T O T Y P E E L EVATIONS OF C H A N N E L FLOOR
STA. ON RIGHT ,1, LEFT
.._ POINT SIDE
.._ SIDE POINT
0 t 50 147 1 . 5 5 1 4 8 5,02 1492 .83 !497.50 1 500.26
1 + 00 1469.99 1 4 8 2.67 1490.43 1495.40 1 497 57
1 + s q 1468.65 1480.40 1487 99 1493. 1 3 1 496.65

2 + 00 1467 3 7 1478. 1 8 1485.51 1490.74 1494.50

2 + 50 1466.29 1 4 7 6.02 1482 97 1488 1 8 1492.05

�-
3 t 00 1465.22 1473.92 1480.40 1485.44 1489.44
3 + so 1464. 1 0 147 1 .83 1477.77 14 82. 65 1 486.57

4 + 00 1463. 1 0 1 4 69.74 14 75.09 14 79. 67 1 483.46

4 t S O 1462.04 1467.69 1 472.37 1476.56 1480.1 4
A- P"L A N

ELl53 5·,

C - S EC T I O N 8 · 8

I" " ,, I
B - S E0 1 0 N A - A PROTOTYPE SCALE IN FEET
0 20 40 60 80 100
I' ,, ,, ,,
0 2 4 6 8 10 12
MODEL SCALE IN INCHES

ELEVATIONS O F CHANNEL FLOOR

"
M O D I F I E D BY CLAY FI LLETS
STA. ON RIGHT 'I• ,1, LEFT
SIDE POINT POINT SIDE
0 + 50 1474.20 1485.8_5 14 92.75 1497.50 1500.26

SECT ION - S T A . 2 -t- 50 I + 00 1472. 65 1484.25 1490.43 1495.40 1498-.57

E - F A L S E FLOOR AN D FILLET IN P L A C E I t 50 1482.04 1487.99 1493. 1 3 1496 65

I47 l .3 I
I N R IGHT H A L F OF CHA N N E L 2 + 00 1 470.05 1 479. 45 1 4 8 5 . 5 1 1490. 74 1494.50
2 t 50 1468.95 1478.25 1482.97 1488 18 1492 05
C O O R D I N ATES OF CLAY F I LL E T 3 +00 1467.88 1474.55 1480.40 1485.44 1 489.44
DIST. E L EV. S T A. DIST. ELEV. 3 t 50 1466.84 1472.95 1477.77 1482.65 1486.57
X' X'
S T A.
4 + 00 1465. 76 1_471 . 1 5 1475.55 1479.67 1 48 3. 46
Q -1- l Q 1.0 1504.7 0 + 50 0.0 1487.8
4 + 50 1464.70 1469.05 1472.37 1476.56 1480. ! 4

•o ,oo
30 1503.8 5.0 14830
1503.0 14808
0 • 20 1.0 1501 .8 20.0 1478.8
50 1498.9 27.0 1478.9
10.0 1496 9 '0+65 00 148 1 . 1
O • 30 0.0 1498.1 50 1 477.5
60 1 493 .2 10.0 1476. 2
90 1490.8 20.0 1477.0 SALT R I V E R P R O J E C T - AR I Z O N A
<5 0 1.:30 4 0+72 0.0 14800
20.0 1487. 8 5.0 14770
ST E W ART MOU N TAIN D A M

,o
O• 40 00 1492.6 I0.0 1475.6 H YDRAULIC MODEL STUDIES O F SPI LLWAY C H A N N E L
1490 l 15 0 1476.2 T E S T 4 : R E D E S I.G N OF C H A N N E L - \ : J O O M O D E L
D - C R EST, P I E R A N D G AT E DETA I LS AT <t 9.0 1485.0 0+86 00 1476.8
I5.0 1482 .8 60 1474.3
21.0 148 1 . 3 I0.0 1474.9
27.0 1480.6 20.0 1476.6
30.0 1480.9 30.0 1478.3
 from velocities determined by pitot-tube measurements o The l osses to a

station in the channel in this study were expressed as the difference

between the water surface elevation in the reservoir and the elevation

.. of the energy head above the station considered o Since the energy head

is the depth of water plus the velocity head , · v , the problem _resolves
2

2g
itself into a measurement of the depth of water in the channel , and

determination of velocity o

Stations 0+50 , 2 +50 , and 4+50 were selected and the gradient assumed

to lie along the centerline of the channel o The water surface was pro­

filed across the channel with a point gage f or discharges of 50,000 ,

100 ,000, and 150 ,000 second-feet (Figure lOA ) o The area was calculated

by integration of the depths across the section, and the average velocity
obtained by dividing the lmown discharge by the area o The gradients of

Figure lOB were based upon the velocity head of these average velocities.

While the results are approximate, the losses show a reasonable relation
with discharge, in that greater loss occurred in the channel when there

was less discharge. In contrast, the test on the original design by

pitot-tube measurements was qui"te inconsistent in this respect (Figure 5) •
The reservoir elevation was determined with the gates raised , at
discharges of 2 5 , 000 , 50 ,000, 75 ,000 , 100 3 000, 12 5,000 second-feet�
respectively, and a capacity of 139,000 second-feet at el evation 1535
was f ound by interpolation o This vras essentially the same discharge as

was f ound on the original design.

10. The Use of False Floor and Fillet to Improve Flow

In designing a superelevated spillway channel, the depth of flow at

the lower side or inside of the curve should be less than that at the

19
 F I G U R E 10

S T A T I O N 0 + 50

S T A T I O N 2 + 50

S TA T I O N 4 + 50
A - P R O FI L E OF W AT E R S U R FA C E IN C H ANNEL

""
RES. EL. 1536.5-,

.-·Energy gradients
RES. EL 15293·· 1530

R E S. EL. 1529.0''

�
1520

:::
,_w
0 1510

�
w
1500

�_,
1490

1480

1470

8 - H Y D R A U L I C G R A D I E N T S AT i

1 - Hydrouhc gradients= depth + ti2

NOTE

Where V= overoge velocity'Os

drschorge
determined
by water area under profile
2 - For pion of channel see Figure 9.
SALT RIVER P R O J ECT-A R I Z O N A

S TEWA R T M O U N TA I N DAM
H YD R A U L I C M OD E L ST U D I ES OF SPILLWAY
T E S T 4 - H Y D R A UL I C G R A D I E N T S
B A S E D ON AVE R AG E V E LO C I T I ES
 outer side. Since the depth of flow was greater in the l ower or right

side of the channel of the model (Figure lOA ) , it indicated that the

degree of superelevation was too great o The channel was modified by

.. constructing a false .floor on the right side (Figure 9E) o At the same
time, a clay fillet was built against the right wall to eliminate the

undesirab1e fin previously described (Figures BB, 9A 3 and 9E ) o The

improvements with these modifications were substantial because the depth

of flow was uniform across the channel, and the fillet materially reduced

the size of the fin o

This test concluded the series of studies wj,th the 1 : 100 model o The

development of the final design was made with a model to a 1 : 50 scale.

DEVELOPMENT OF THE FINAL DESIGN

11 . Design and Construction of the 1 : 50 M odel
The final design .studies of the Stewart Mountain Spillway were made ·

on a 1 ::50 model o This larger scale was necessary because the 1 : 100 model

was too s mall to study detailed improvements and to · obtain sufficiently

accurate discharge me�surements required for the final design. This

1 : 50 model was built similar to the former 1 : 100, and incl uded a portion
of topography upstream, the crest, piers , gates , and channel (Figure 1 1 ) .
The knoll a...r1d riverbed dmmstream from the chute vvere omitted. This

omi.ssion was not significal!lt in that flow over this knoll was· not ,to be
studied , and its presence or absence would not affect the flow in the

channel which was at supercritical velocities. Topography upstream from

the crest was made of sand stabilized with cement. The crest was the
same as used in Test 2 of the preliminary design � but with piers and gates

. added.

20
 Two c hannels were contemplated, first J the same design as the

1 : 100 model of T est 4 alined on a compound curve having radii of 490

and 1 , 140 feet 9 and second, a channel alined on a spiral curve w ith
• radii of 47 0, 560 3 690, and 905 feet (Figure lll) . The c hannel similar

to the l g l00 model� was not built because . of the necessity of proceeding

with s tudies o.f the f inal design. In view of the fact that the final
design was to be a channel alined on a spiral, the only value of tests on
this first channel would be to compare the flow- with that of the smaller

1 & 100 model. Even then, such a comparison would be of little value since

the roughness factor would not change as the same materials were used in

both models . In all J) 36 individual tests were made on this model of the
spiraled channel.

12 . General Performance of Model and Discharge Capacity

The model was studied in Tests 1, 2 , and 3 wit hout any refinements .,
with the c rest section as it existed in the field, and with the spiral

channel attached (Figure 11). Test 1 consisted of three runs at dis­

c harges 50, 000, 100, 000 J> and approximately l39 J) 000 second-feet with the

gates opened partially at the low-er discharge to hold the water surface

at elevation 1529, and opened wide at the larger discharges o The flow

appeared to spread uniformly across the c hannel , indicating that it was

correctly designedo There was some splash against the right wall down­
stream from the crest in a manner similar to that of the 1 & 100 model
(Figure 8B) . Fl ow over the crest was similar to that of the 1 : 100 model 1

Figure 8C , contracting at the ends and reducing the e ffective length of

the c rest o

21
 F I G U RE I I

l� /

.: - K
/ .! /

,,
/, ;'_ ___ _
I
,' O

,,
/ 2.

-�
I Cf

,
�
,
,
/

.' ,/ /
;
.s,
'•.,. t()
I

I ! //
I
•/

I /%�<
I
I /
I
' -� :::; I
I
/

,! J!U...l..lt::�_�_�_�_�_=_i�==i:�-

ii
,•', •',
_.L
'' '
, 1 _/ // i

�,
'
j ,/ ,,,,/ ,,: // ,.;sJ� //
· ·-Gote

• · -------------3' - 6!3:' ----------� j / ,,,, / ,,,//

,'' / ,// /'

,
:
:,
:

,
!/I/ ,,l /
//
MOO E L
, ,'
°" 'NCHES
,, '6
//

f-r-=l:::3-f-��-':_-.::· ·[ ___ E I. 153 5

:1 //,,,,, //
CALE

!>O ,50
!,/ /,
//
///
� ioo /,/' _,/ CHANNEL ELEVATIONS

'- '-
STA
PROTOTYPE SCALE J N FEET
GRADE ELEVATIONS
ON RIGHT ¾ POINT LEFT

,/ ,I/// ,,,//
WALL f P01NT WALL

. f--<-------------3'-i'i':- --------<,/
0+50 14750 1487.2 1493.3 1496 .3 1497.5
64 I + 00 1484.6 1491.2 1494.7 1496.5

c- --
1472.5
1 + 50 1470.7 1482.3 1489.0 1493.0 1495 2

// //
2 t 00 1469.2 1480. 1 1486 3 1491 .o 1493.5

'f
2 + 50 1 468. 1 1478.1 1484.7 1488.9 1491.7
3+ 00 1467.2 1 476.2 1482.5 1486 7 1489.6
1480.2 1484 ,3

," -il' /,/

" so 1466.5 1474.5 1487 3

l
1484. 7
---- --2'- i'i':------� _;:/ ,/ 4 + 00 1465.8 1472 7 1477.9 1481 .8
1 471 I

� ':jl" -----------------·j
1475.6 1479 2

-�
4 + 50 14653 1482. 0

iI
E l 1 5�

'//
1530 1530 �

,,���
5
I
Rig h t

j__�//
1495.2 ,_ :,\.
wa t r� · -·· t I
E 1 . 1496.as) �
Sta. \ + so
1410.1
TYPICAL C R O S S S E C T I O N OF C H U T E
A - PLAN
C - GATE A N D PIER

Left wall

0
B - SECTION A -A

S A LT R IVER PROJECT - A R I Z O N A

STE WA R T M O U N TA I N DAM
H Y D R A U L I C M O D E L S T U D I E S OF S P I LLWAY C H A N N EL
O R I G I N A L D E S I G N OF 1 • 50 MODEL
 In Test 2 9 discha rge measurements were made to compare t he spillway

capacity with that of the Test 4 on the l g l00 model, and to provide a

basis of c omparison for later tests after t he flow over th� c rest was
• improved. These measurements were made with all gates wide open J and the

discharges ranged from 12� soo second-feet to 136 , 000 second-feet, the

maximum c apacity with the water surface at elevation 1535. The most
significant aspect of this test was that the capacity of the 1 8 50 model

was less · than that of Test 4 on the 1 : 100 model, but the comparison was so

close when presented as head-discharge curves that the differences were

not significant (Figure 12A). A more desi rable comparison was to express
the capacities as a coefficient of discharge� as shOV'm by curves of

F igure 12Bo
Test 3 was made with the gates removed to eliminate the effect of
the s heet metal gate arms o This resulted in an inc rease in the spillway

capacity . to approximately 138, 000 second-feet, but this test was an

impossible condition for even with the gates c ompletely raised and out
of the water, the radial arms were still partially submerged in the stream .

The fact that the resistance of the_s e thin arms changed the capacity a

measurable amount indicated that minor changes could be . made · to improve

the fl ow over the c rest o

13 . Improvement of Flow Over Crest and Into Channel
The undesirable aspects of flOV<r over the c rest observed in the

l g l00 models were present in the l g 50 modelo There was a definite c on-�

traction at the end piers ., similar to Figure 8C 9 and a hig h fin at the
right wall dO\mstream from the · c rest (Figure 8B). Elimination of this

22
 F I G U R E 12

1534
�
�
28

��
1530 24

�
,p-

�
w

...
l---'.,.
w
O 1526 >- 20
0 0

...w /
er
0

...
Q.

er
,J"
Q.

�
� 1522 16

/1�
w

�
w
T E S T 2, 1 , 5 0 MODEL . . . . . . ·---

0
z
-.½
-
1518 12
� <[
w
>
r

�7
w
w �TH I R D l d OO LDEL
( Test 4 Port I )
1 51 4

1510
/;'
�//
1506 i/
0 10 20 30 40 50 60 70 BO 90 100 110 120 130
D I S C H A R G E IN T H O U S A N D SEC.-FT.

•
I ,, • /_
'I
/

I
/
T
.

i
1 534 28
I
T E S T 8 - --- - ·• ·
-1 I )/
b

'
(3 Left end gates only)
1530

.j
24
I I

v:
/

...
>-

...
w
T E S T 7-- ---
·y d - ·TEST 7, 1 = 50 MODEL

I ---1'
{ 3 Right � nd gates only) (End piers improved)
�{
... JI
1

-I
Q.

e 1526
0
>- 20
0 �;,,? T

...
er
0

�- �/
Q.

... 16 /)
TEST 10, 1 : 50 MODEL · ·····

/ ·
[ Folse floor in chonnel) /

�
Q.

/
lL. 1 52 2
w

�
w
T E S T 9 - - - - - - - - --

/�
__,,,,,,,.,v/
)
0
z

,,./
( 3 L e f t e n d gates with k:S T 2 , -T H I R D ldOO MODEL

�r
0

"'
wing-wall of r i g h t end p i e r ) 1 : 50 MODEL { Test 4 Port I )

L-,--<
1518 _

.;:: /
12
�
4

,.,<",,,,.
V ,,,,,,,-,
I/

---
,,.,.,.......,
>

__,�_,,. / /.
L-- � �
1514

-----
- ·-F I NAL D E S I G N

�10 _.,,,.v /

1506 0
3.00 3.10 3.20 3.40 3.50 3.60 3.70 •
COE FFI C I E N T OF D I S C H A R G E "c" C = L�.)/2

S A LT RIVER P R O J EC T - ARI ZONA

S T E WA R T M O U N TA I N DAM
H Y D R A U L I C M O D E L S T U D I E S O F S P I LLWAY C H A N N E L
D I S C H A R G E A N D C O E FF I C I E N T C U R V E S
c ontraction w ould increase the discharge capacityo If the high fin at
the right wall were eliminated 9 the height of the channel wall in that
region c ould be materially reducedo Tes ts 4, 5 .9 6 9 and 7 conside red
mea ns of improving the flow- ove r the c res t and into the channel o
In Test 4, 1-inch radius wing w alls ( 50 inches, prototype ) were
placed at each e nd pie r tangent to the 45 ° nose of the piers , as shown

in Figure 13A o Although satisfactory at small discharges , the w ing wall

on the right end pier was completely inadequate at large r discharges o

In Test 5 �· 3/4�inch radius· wing walls (37 o 5 inches .9 prototype ) were

placed at each end pie r tangent to the inside face of the end pie rs o

These wing walls were too small to improve the flow especially at the

right end pie r o It was concluded. that a large r wing wall was needed at
the right end pie r, but not at the left end pie r o

In Test 6 , a more general study was made by including four

alterations , o r improvements to the model � (1) a 2-1/4-inch radius wing

wall ( 9 o39 feet , prototype ) at the right end pie r (Figure 13A ) ; (2) a
clay fillet on the right end pier between the nose of the pier and the

offset dcwnstream (Figures 13A and B ) ; (3) a false floor about So feet
wide at the right wall extending from Station 0+10 to Station 2+50,
(Figure 13C ) ; and (4) a piece of sheet metal placed upstream from the
left end pier to represent a modification of the natural topography

upstream f rom the c rest to a slope of 1-1/2 8 1 (Figure 1.3D ) . These

alte rations considerably improved the flow over the crest and into the

channe.l . The large wi ng wall on the right end pie r.9 and modified topog­

raphy at the left end pier held the wate r surface at both e nds of the

23
crest nearly level with contraction of flow practically . eliminated. The

clay fillet on the right end pier streamlined the pier wall, tending to

hold the water_ against the wall and submerge the objectionable f in down­

stream f rom the crest.

To ascertain the probable height of the right wall , point gage

readings were taken to establish the water surface on this wall 9 using

a discharge of 132 9 000 second=feet (Figure 13E) .

The principal objection to this arrangement was the large circular

wing wall at the right end pier, which was extended below the c rest

approximately 40 feet ( prototype) to the ground surface. Therefore ,

Test 7 considered a more economical type of wing wall (Figures 13A and B) ,
which extended only a short distance below the crest. With this w ing

wall, the f low was nearly the same as bef ore, and the design was considered

tentatively satisf actory.

14 . Spillway Capacity

In Test 7 , six runs were made w ith the gates wide open w it h the dis­
charge varied · f rom 71,700 second-feet to 131 , 900 second-feet to obtain

a coeff icient c urve . The spillway capacity was materiall y increased such
that a f l ow of 140, 100 second-feet could be obtained w ith the water sur­
f ace in the reservoir at elevation 153.5 (Figure 12B ) • It was, therefore,

concluded that the proposed spillway alterations were worthwhile .

15. Corrparison of Flow Through End Gates

A w ing wall was placed at the right end pier, while the onzy alteration

at the lef t end pier was to grade the grormd surface upstream on a 1=1/2 � 1
slope. Since the a pproach conditions to the spillway were not symmetrical ,

24
it was desirable to know the comparative discharge next to the right and

left end piers. To do this, Test 7 inc luded five runs to measure dis­
charge with the three right gates openJ and Test 8 included five similar
runs to measure discharge with the three· left gates openo In Test 8 ,

the flow spread across the channel to such an extent that it was believed

that a fair test could be obtained only- if it were confined to the left

third of the channelo This was ac complished by placing a row of bricks

in the channel (Figure 13C) o

T he procedure of the tests was to hold the gates open a nd measure

discharge at various water surface elevations in the reservoir o Dis­

charge coefficients were then comput�d in the same. manner as with all
gates open, and are comparable because the coefficient of discharge is

theoretically independent of crest length o However, as s hown by curves

on Figure 12B , discharge coefficients for three gates only are less than

coefficients w ith all gates openo This was anticipated because with

three _ gates open there exists a contraction not only at the end pier, but
also at the fourth pier and this c ontraction at the fourth pier reduces
the effec tive c rest length.
More significant , however, was the fact that the coefficients for
the th ree left gates were materially less than those for the three right
gates o T his difference was c.aused by three factors g (1) the lack of a
wing wall at the le.ft end pier, (2) the ground surface upstream was close

to the c rest near the left end pier, but approximately 40 feet below the

c rest near the rig ht end pier, and (3 ) _in the superelevated channel down-�
1

stream ., the drop from the c rest to the channel was less at the left s ide ,
and any tendency for water i n the channel to submerge the crest would be
prevalent at that sideo While all three factors contributed to the

unsymmetrical flow over the spillway, it was not certain which was more

important, therefore , tests were made to establish this point o

16 0 Use of Wing Wall on Left End Pier

Without any other changes on the model , a wi ng wall having a radius

of 5 o l3 inches ( 21 035 feet, prototype) was placed on the left end pier to
establish the effect of this factor on di scharge (Figure 13D)o The runs

were made with t hree left gates open i n the same manner as i n Test' So

The coefficients indicated that the wi ng wall i ncreased the discharge

capacity a small amount (Figure 12B) , but not sufficient t o warrant the
constructio n i n the prototype o

l?o Effect of T rai ning Wall in Channel

In tests to measure flow through the th ree left e nd gates , a row of

bricks were used to represent a training wall to keep the flow in the

left third of the channel, since water passing over the crest tended to

s pread across the channeL To show the effect of this trai ning wall !) a

run was made in Test 8 in which the water surface below the centerline
of t he first gate was point gaged, w ith and wi thout the training wall o
N o significant differences could b e observed at least 100 feet downstrearno

The flow appear�d s imilar in Test 9, where it was furt her observed that

when the training wall was taken f rom the channel, there was no change in
the reservoir- water surfaceo This w ould indicate that the t raining wall

had no effect upon the di schargeo

26
l8 o Effect of Wing Walls on Water Surface at Right Channel Wall

The water surface along the right channel wall was point gaged in

Test 6 (Figure 13E) o In that test 9 there was a circular wing wall at
the right end pier, and no wing wall on the left end piero Later, the
wing wall at the right end pier was changed to a streamline shape, Test ? D

and a temporary wing wa ll was placed at the left end ·pier, Test 9 o In

Test 9, the water surface along the right channel wall was measured to

observe any effects caused by changing the wing walls . All gates wer�

open and the reservoir was held at elevation 1533 0 88 , virtually the same

as in Test 6 0 · The water surface was nearly identical to that in Test 6 ,

demonstrating that the different wing wall arrangement had no appreciable

effect upon the flow in the channel o

190 Revision t o False Floor

The false I,Loor used in Test 6 was removed and replaced by a new
false floor extending from Station 0+50 to 4+50� as shown by Figure 14A.

At the same time, the l-l/2 g l sloping bank on the left s ide of the spill�

way, upstream f rom the crest, represented· by a piece of sheet metal in

Test 6 through 9, was replaced by sand as permanent topography, for this

modification of the natural bank was desirable o
Test 10 included only discharge measurements with all gates wide
open to ascertain the capacity of ,this designo A comparison of the
coefficient curves, Figure 12B, shows that the capacity of this design
was less at low heads, but compared favorably with other designs at the

maximum head . In this test D the head was measured in a more refined

manner than formerly in that three gages at various points in the forebay

27
 ;g����tl� 9

�1
L���T /·Fillet
inside face of pier.

BOTH
END PIERS
RIGHT ENO
P I E R ONLY

TEST 5 TEST 6 TEST 7

A - P R O P O S E D W I N G W A LLS

. Nose o,t right

end pier.

I Tests a and 9 oppro i< i m o te

location of bricks in the'
1 channel to restrict the
flow--
i
"·-;; --Streom l i n ed wing wall
for Test 7.

B - S I D E V I E W OF W I N G WA LLS A N D
P I E R F I LL E T

C - P L A N S H O W I N G LOC A T I O N OF FA L S E
F L O O R -TEST 6 A N D B R I C K W A L L T E S T S 8 A N D 9

E L E V A T I O N S OF FALSE F LO O R
···· · · Slope r! r-- - - � ­

"- "-
STA. D I STANCE STA. D I S TA N C E STA. D I S TA N C E STA. DI STANCE

<t_ ct_
,�..,...,._..
·y .
ON FRO M WA L L E L EV. ON F R O M WA L L EL EV. ON FR O M WAL L E L EV. ON F R O M .WALL E L EV.

.· wing wall for Test 9

· · · · --Vedicol circ,loc PROTO M O D E L PROTO MODEL PROTO MODEL PROTO MODEL

_____ ________,,,
1 .00' 0.020' 1504 7 0+37.08 4.00' 0.080' 1486.65 1 + 00 3.25' 0.065' 1479.05 1 + 75 3.25' 0.065' 147480
3.00' 0.060' 1503 . 8 1 2 .50' 0.250' 1486.55 12.50' 0. 250' 1479.35 12.50' 0.250' 1475.B5
s.oo· a.too· 1503_ 4 22.50' 0.450' 1486. tO 22.so· o . 450' 1479.75 22.50' o .450' 1476.85
6.00' 0.120' 1503.0 32.50' o. 650' 1486.00 32. 50' 0.650' 1480.65 32.50' 0. 650' !477 85
PLA N

El. 1535···

,..�+�2=
0
10.00'
s-+-��
0
;
�7
:-
0
�:�

9.00'
12.00'
��
0.200' 1 496.9
��7
: +-:�:�--�:-+-_
.

0.180' 1494.6
0.240' 1493.5
_
+ 1
_ ,-+--!:�7:
0

,
61.75'
- +-�
��
:
-;
-
t.235'

: 148060

32.50' 0.650'
42.50' 0.850'
1485.95
�7>-'-"='+-'�"-+�
1479.70 t -t 50
,;=�
:
1481.50
1482.05
1483.05
'�
=:
54.50'

22.50'·
32.50'
,-�=: =�'":
1.090'

�
,.
� ,.
=9
o.450' 1477.95
0.650' 1478.95
=-r�
0
� -;=
:
55.00'

�
=
J . 100' 1478.15

s :_.._� �:�: ::::·:

12.50'
:
0.250· 1470.55

U PSTREAM E LEVATION 1 5.00' 0.300' 1493.2 52. 50' l . 050' 1484. 1 5 52.50' 1 . 050' 1480.45
61.00' 1 . 222' 1483.35 56.50' 1 . 1 30 ' 148070

D - T O P O G R A P H Y A T LEFT E N D OF
S P I L L W A Y - TEST 6

1 5 25 � - �--�--�--� - � - -� --�--� - � - �

S A LT R I VER P R OJECT - A R I Z O N A

S T E WA R T M O U N TA I N D AM
H Y D R A U L I C M O D E L S T U D I ES OF S P I LLWAY
OISTANCE ALONG RIGHT WALL FROM A XI S OF C R E S T
T E S T S 4 - 9 I M PRO V E M E N T S TO S P I LLWAY 1 • 50 MODEL
E - WATER S U R FA C E A LO N G RI GHT C H A N N E L W A LL
 were used to ompensate for any uncertainties of velocity effects.

Differences in the three gages were less than O o lO foot , prototype, and

this small difference was not significant being within the accuracy of
.. the gage itself o The location of the forebay gage was not critical as
long as it was placed at a reasonable distance upstream f rom the c resto

200 Effect on Discharge of Excavating Bank Upstream From the Crest

. It was suggested that the spillway capacity might be inc reased by

excavating the left bank upstream f rom the c rest down to elevation 1490

along the 1-1/2 � 1 slope 9 a9 shown .in Figure 14A o Several discharge runs
were made and c oefficients compared with Test lO o
1
The results indicated
an increase in discharge capacity of approximately O o ? of 1 percent , but

such a small gain was not sufficient to j ustify the cost of the proposed
excavationo The idea was abandoned and the embankment replaced as in
Test lOo

210 Measurements of Water Surface in Channel

With the water surface in the reservoir at elevation 1535 and with

the spillway discha rging 137 , 200 second-feet, the water surface in the
channel was measured along the channel walls and across the channel at
Stations 0+50, l+OO .., 1 +50 9 2 +00 9 2 +50 9 3 +00, 3+50, and 4+00 (Test 12) o

The water surface was . considered satisfactory along the left wall, but
too high along the right wall immediately downstream from the crest
(Figure 14B ) . Further changes were indicated to reduce the height of the

water along the right wall o

The measurements of the water surface across the channel are not
included as the model was material ly altered in lqter tests by removing

28
the false floor and installing i n its place a filleto It is sufficient
to note that t he water surface was reasonably smooth , and that the depth

was greater at t he outside of the curved channel (that is the left side)
as in the case of a theoretically c orrect design o
22 o Use of Circular Wing Wall at End Pier

In Test 6 9 a circular wi ng wall on the right end pier was used, which

was replaced by the smaller streamlined wing wall in Test 7 (Figures 13A

and B ) o In Test 9, it was demonstrated that the smaller streamlined wing

wall would be as efficient as the circular w ing wall, but to verify this
further, the ci rcular wal l was rebuilt and several runs were made to

measure discharge ., The discharge capacity was increased less than

l percent over that measured in Test l0 J> and as far as increasing the

discharge is c o ncerned j it was apparent that the c ircular wall had little

advantage over the streamlined type .,

T o compare this wing wall further, the water surface along the ri ght
channel wall was measured j but the results w ere not significantly different

from that in Test, 12 (Figure 14B ) o It was conclusively . shown that the
performance of the smaller streamlined wing wall .9 intraduced in Test 7 ,

was as efficient as the circular wall used i n Test 6 0

23 . Use of Fillet in Channel Downstream From the Crest

The primary objection to the design of Test 10 was the high water

surface along the right wall ., In forme r tests on the 1 � 100 model, the
flow was improved by placi ng a fillet on the channel floor alongside the
wal l, immediately downstream from the crest (Figure 9E) ., To demonstrate

again the advantages of such a fillet, one was placed i n the l i 50 model
(Test 13A) , extending from the crest to Station 0+80 ., patterned after the

29
fillet used in the 1 � 100 modelo In demonstration runs the height of

splash along the right wall was reducedo Measurements of the water

surface or coordinates of the fillet were not made, since it was planned

to imp�ove the shape of this fillet in later testso

24 0 Use of Long Fillet Alongside Left Channel Wall ·

After preliminary runs wi th the arrangement of Test lJA , the fillet

and false floor of Test 10 were replaced (Test 14) w ith a long fillet

alongside the right channel wal l (Figure 14A ) o This new fillet was formed
by shaping and reshaping until satisfactory flow conditions were obtainedo

Coordinates of t he fillet were. then determined by point gaging (see table

for Test 14 in Figure 14) o

With the gates c ompletely raised 9 and with the reservoir l evel at
elevation 1,53,5., 0 9 the water surface along the right channel wall was

point gagedo Comparison with the data of Test 12 (Figure 14B) shows that
the height of the water surface along this wa:11 was appreciably r�duced

indi cating a definite advantage in the use of the f1lleto

2.5. Modification of Sloping Bank Upstream From Left Pier

T he sl:oping bank upstream from the left end pier was revised (Test 14)
as shown in Figure 14A to improve the approach conditions and increase the
discharge c apacity as it was possible that t he embankment developed in
Test 10 was not large enougho Preliminary runs indicated that the flow
with this new embankment would be satisfactory, but no discharge measure·-
ments were made until other tests were c ompletedo •
26 .. Gate Operating Schedule

The need for a gate operating schedule was indicated.. Prior tests on

the l g ,50 model had been made with all gates opened equally, resulting in

30
 a flow down the spillway of uniform depth across the entire section3 the

design condition for this channel o In actual operation of the prototype

structure, it w ill seldom be necessary to require all gates to be opened,

, and f or most releases it may be more convenient to use individual gates

fully openedo A study (Test 14) was made by operating the individual
gates j then various combinations of two , three, and four gates o Through­

out this study, the water surface was held at the normal reservoir eleva=

tion of l.5'29 o 0 o It was found that when certain gates and certain· groups
of gates were opened the flow conditions i n the channel were not satis­
factoryo To facilitate the following discussion, the gates will be

numbered consecutivelyj) 1 to 9 3 from right to left, as means of their

identificationo
When Gate 1 or aey consec utive g roup of gates , beginning with Gate 1,

were opened the flow conditions were satisfactoryo

When G ate 1 was closed and Gate 2 or any consecutive group of gates J
beginning w it h Gate 2 9 were opened the�conditions were not satisfactory

because the flow piled against . the right wall to form a fjn as shown in

Figure l.5'C o

When Gates 1 and 2 were closed and Gate 3 or a�y consecutive groups

of gates y beginning with Gate 3 j) were opened the conditions were not

satisfactory because the flow piled against the right wall in the same
manne r as w ith Gate 2 openo
With Gate 4 open the water piled against the right wall, but not
sui'ficiently high to overtop ito

31
 ..
COORDINATES OF FALSE FLOOR TEST 10 COORDINATES O f fl L L E T T E S T 1 4
ST
STA. ON ' DI �
IEI; - FALSE TA ON �FRCN la. FALSE S TA.ON
IS .
STA.ON DIStFROM ELEV.OF STA.ON DIST FROM ELEV.OF STA ON DtStFROM ELEV.OF
I

\"ii,
RIGHMALL FILLET RIGHTWALL FILLET l 1GHTWAU. FILLET
{ ���� ct � ���
!FT.MOO ct 0 + 6.5 - o- 1505.23 - 0- l498 2 5 1 2 . 5 0 1478.45
o+ 1 . r 990 1503 95 0 1 4 5 32.50 1482.00 1479. 1 5 - o-
3.05 1 504.50 5.00 1 4 8 6.30 I t 20 23 00 147645
Q t 12.1 1 1. 5 0 1 50335 35 1 4 88.50 20.00 147625
5.0 0 1 5 0 3 .85 20 1492.30 5.00 1478.60

Note - Embankments for these tests - o- 150565 - o- 1 4 9 2. 7 5 -o- 1478.85

0 + I 5 1 � 1 5 0 1.95 Qt 50 32.00 1 4 8 1. 50 I t30 22,50 147605
eA tend down on o 1t I slope from
500 1 502.90 2 5. 00 14 8 1 . 2 5 20.00 1475. 85
eleva lion 1535
- o-
1500.00 5. 00 1483.75 1 0.00 147800
O t20 I 1 4 . 0 5 �98.85 50 1485.85 -o- 1 478.60

\� \I
2
150005 -o- 1488. 1 0 l t 40 2 00 1 475.60
I 20 1513.10 0+60 32.50 148 1.00 000 1477.50
1 5 1 5.40 25.00 M79.85 -o- 1478.45

0
1469.05 1�- -1°ci 149485 1 0 .00 1481.35 I t60 1474.75
2 5 21 45

Test 1 1 , exca vation

�00 1495 10 - o- 14 8 2. 1 5 0. 0 0 1476.55

8�
320

·�- �� ::;:::�
• 2.50 149960 O t 7 0 30.00 1 480.00 -o-
to EL. 1490.00 ·--:,_- ·· .975 1470.75 1509.25 22.50 1479.35 l t 80 23.45 1474 30

. -
.750 1 4 7020 -a- 1 5 1 065 10.00 ]480.50

----=a:-
.550 1469.50 3 1 4 5 - 1 4 89.95 -o- 1480.20
.-· .350 14saeo 10.00 1493.55 0 t80 27. 5 0 1478.90 1476.75
1490 /
"\
. 150 146845 .70 1497.50 25.00 1478.60 2 t oo 23.75 1 473.50
'/9 1477.05 I -o- 1467.65
6>0 I .65 1505.85 7 50 1 480. 1 5 ___!_7.50 1473.30 .
_ \
f -o- lsoiis 2.50 1 480.05 1 0.0 0 1474. 2 5

� ---
- --\-;·
Ot35 I 3 1 . 7 5 1 4 86.20 -o- 1480. 1 0 - o- 14 76.10

\_,,·

)
1 0.0 0 1490.60 0 1" 9 0 25.50 1477. 9 5 2+20 1 9. 0 5 1 4 72.00

,g .45 1 4 94.35 1 2.50 1 4 79.40 7 50 14 72.75

·- o -
------------- i
- o- 1480.15 ·· o - 1474.85
.35 1 501.75
'v

---
1 50285 "00 24 50 1477.40 21 5 0 9. 70 _1_469.60
1, o -t 4o l 3 1 . 1 5 1483.80 7 50 1479.45 -o- 1471.45
10 .00 1487.75 250 1479.35 2 + 70 4 20 1468.45

',, -.....
---- <,<,-,x,,_,
���
c
.45 1491.60 -o- 1479. 75 -o- 1469.15
. 25 140750 I t 10 23.6 5 1476.85 ' 2 t 8 5 - o- 1467.45

'•
. ....... I

',, ,
..

Folse floor Test',

"'·
"o
10 "'"'
�"
' COORDINATES Of F I L L E T TEST 1 6
(Replaces floor of Test 6) '-, .
ELEV.OF STA.ON DIST FROM ELEV.OF STA. ON

1473.85
STA.ON 1$tFfQA
�t<>!T...U. if,1 ST FROM ELEV OF
FILLET It. RlGHTWALL FILLET !G<TIW.L FILLET
It.

'"'
', 0+05 0 1505 8 7 O t 3 5 32.70 1486.30 2+00 22.00 1473.40

',\\
4.0 0 1505.52 0 + 5 0 0 1 485.24 2t25 0
OtlO 0 1505.50 16.00 1482.27 10.0 0 1471. 78
9.50 1504.22 34.00 1 4 8 2 . 30 2000 1472.30
1<1so -------­ Qtl5 0 1504.72 0 + 75 0 1 4 8 0 65 2 + 3 7 5 0 1 472.54
6.00 1502.86 1 6.00 1479.74 8.00 1470 73

I
A. PL A N S H OW I N G R E V I S I O N S ON M O D E L 12.90 150200 3 1.20 1480.17 16.00 1471.28

\
FOR T E S T S 1 0 - 1 6 0+20 0 1503.40 l t OO 0 1479.65 2+50 0 1471.30
8. 0 0 1 50Q22 16.0 0 1 478.30 8.0 0 1469.86
15.60 1 498,90 30.00 1478.83 14 00 1 4 70.58
O -t2 5 0 1 5 0 1.00 l t 50 0 1 478.50 2 + 62.5 0 1469. 7 7
Note- For test 15 pressure measurements
10 00 14 96.92 16.00 1 475.80 8.00 1469.2 8
on spillway bucket,see figure 16. I-
18.90 1494.90 2800 1476.34 2t75 0 1468.55

���-.---.---.--,---,---i------r-�--r--i�11-rTT-r1.1T7-r-7777 - -
0 1" 3 0 0 1498.00 1 175 0 1 4 7 7.46 6. 00 14sa60

�
_ 14 92.96
12 00 12.00 1475.20 2+87.5 0 1467.72
I
2 4.2 01490.02 14 74.50 1467. 8 7

;I I ·rst\ I I La�lierlurfJ I ! I I I I I
-�3.00
Ot35 0 1494.65 2+00 0 1475.84
16.00 1488.38 12.0 0 1473.70

Note - Water surface measurements with

reservoir at etevot1on 15350.
S A LT R I VER PROJECT - AR I Z O N A

S.T E W A R T M O U N TA I N DAM
H Y D R A U L I C MODEL S T U D I E S O F S P I LLWAY

,,
J--r�i-=t=t:t==t:=-�+-- �:.::- µ_J.:
TESTS 10 - 16 I M P RO V E M E NTS TO
\---I-
T H E C H A N N E L 1 : 50 MODEL

,�
_j G)

"':u
0 C
_
l
--:::f::::_
j__

' .J -

..,
lO 40 50 60 70 80 90 100 1 10 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280
OJ STANCE. FROM AXIS OF C R E S T
B. E L EVATIONS OF F I L L E TS AND AV ERAGE WAT E R S U R FA C E A LO N G RIGHT W A L L O F S P I L L W A Y
 /

With Gate 5 open the wate r spread a nd piled against both walls, but

not sufficient to overtop theme

With Gates 6 , 7 , 8, or 9 op� n the water pilep. against the left wall,

but not sufficie nt to overtop it o

'
It follows that consecutive combinations beginning with Gate 1 or

combinations of Gates 4, 5 , 6 , 7, 8, and 9 may be used satisfactorilyo

When alternate gates were discharging , high fins formed between them

directly downstream from the closed gate (Figt1re 15D). If alternate

combinations of gates were desired, it appeared that combinations such as

1 a nd 4, 1 and 5, 1 and 9 would be more suitable .

From the observations made in this test , it was recommended that the

gates be opened consecutively beginning with Gate 1, a nd under no circum­

stances should Gates 2 or 3 be opened unless Gate 1 was already open.

27 . Pressures in Bucket Downstream From Crest
The proposed spillway c hannel will be j oine� to the existing crest

section by a .30-foot radius bucket faired into the c rest to complete the
ogee from the c rest to the spillway channel . It was desirable to ascer­
tain the pressures in this bucket (Test 15) . Piezometers were installed

along the centerlines of the second and the fourth gates from the right
end of the spillway (Figure 16A) . Pressures were first measured with a ll
gates open wide a�d the water surface i n the reservoir a t elevation 1535 .o,
a nd then measured for various gate openings with the wa ter surface at
elevation 1529. 0 . I n all cases pressures were positive and apparently
�reater than the depth of water over .the pie zometero As indicated on

Figure 16A , the pressures were similar at both positions o

32
28 . Revision of Fillet in Channel and Modification of Right Wall
T he fillet ' along the right wall required simplification for

construction in t he prototype since t he fillet developed in Test 14 was

too hig h near the crest . The fillet was altered as given by the table

for Test 16 , in Figure 14 . It also - appeared desirable to change t he

alinement of t he rig ht wall immediately downstream from t he crest by

s hifting it inward to be flush wit h t he end pier and eliminate the step

or offset between the. wall and pier . T his alteration ' was accomplished in

t he model by a filler as s hown in Figure 14A.

To ascertain if t hese changes made any improvement , point gage

readings of t he water surface alongside the left wall were taken in t he

same manner as t hose of Test 12 and 14 . As shown by Figure 14C, some,

improvement was obtained in that t he heig ht of t he water surface was

decreased at several points .

29. Streamlined Wing Wall and Pressures i n Water Below Crest

In Test 13 t he streamlined wing wall on t he rig ht end pier, introduced

in Test 7 , was, replaced by a circular wing wall previously used in Test 6

(Figure 11 ) . At that time it was concluded t hat the different wing walls
had no appreciable effect upon flow in t he c hannel. As t he streamlined
wing wall was smaller and more economical to build 9 it was selected for

t he final design and was again placed on t he model. Test 17 was made to
measure the water surface alongside the rig ht c hannel wall to compare wit h

measurements obtained in Test 16 _j w hen t he circular wing wall was used .

T he results were nearly identical .

33
 Water coming down the spillway was guided into the channel QY the

bucket where pressures were increased by centrifugal force . In t he channel

itself , there was parallel flow and a normal hydrostatic variation of pres­

sure with depth . It was uncertain as to how far downstream the effect of

the curvilinear flow extended . To determine this , the variation of pressure

with depth was measured by static tubes at several sections downstream from

the bucket � and along the centerline of the second gate from the right end

pier . Two types of static tubes were used to check the results of one

against the other . As shown in Figure 16B , pressures at the end of the

bucket at Station 0+50 were nearly twice the normal hydrostatic variation

indi cating curvilinear flow in this section . The me asurements of pressure

with depth extending downstream 6 0 feet (prototype ) indicate that the

variation of pressure with depth becomes nearly hydrostatic approximately

JO feet downstream f rom the bucket .

JO. Determination of Height of Channel Walls and Use of overhanging Sea

Wall on Right Channel Wall

When improvements to the spillway and channel floor were completed,

steps were taken to determine t he final hei_ght of t he channel walls . The

'
proposed walls in the prototype were to exteµd 5 feet above the maximum
water s urface computed by the Design Section . The actual top of the walls

in the model was at elevation 1529. 0 (Figure 15A ). As revisions were made

to t_he c rest and channel floor 3 the height of the water surface was meas-
ured along the walls . The top of the proposed and revised walls were

sketched on this test wall as shown by the light lines of Figure 15A.

34
;. :;.:;:.,,.--

A. Fillet oh channel floor. B. Discharge with all gates open.

Reservoir at elevation 1535 . 0 .

D. Discharge with
gates No . l and 3
open. Reservoir at
elevation 1529 . 0.

C. Di scharge with gate No . 3 open.

Reservoir at elevation 1529 . 0.
�

HYDRAULIC }.[)DEL STUDIES OF STEW.ARI' M:>UNTAIN DAM SPILLWAY

FLOW IN l : 50 MJDEL WITH FILLET ON CHANNEL FLOOR
....
\Jl
 FIGU R E 16

I
I ---- -- - - e -- - I
151 B

Res. elev. 1535 - - - - - �

Gbtes opeh

Gales open 1
- ,'-'
/

Res elfv /535 - t · _ .,./ I

-
I
'

Res elev 1529 - ··· - -

Gates open

'
- '-----
1512
Gales open
Res elev 1529-- · -- - Res. elev. 1529- - - -
Gotes ro,;ed 18'
-· 7"--. 1
I
1

I --

i
, r--
-r--r=::::
Gates raised 15' j
I ,____

Res elev f529 - - - - - - -�

__

- - ----"\4�
Res elev 1529- · -
Gales raised 18' --
,,o • Gates raised 15' _ .:i
Res.elev 1529------·
I
Res efev 1529--
I I
- I
1500

I'- -
Gates raised 12·
;I'-,-__
Gales raised 12' ,: Gates raised 9 '
Res. elev.1529 --·· �------------ Res elev. 1529- - -- - -
,,__ ---------
I
z

Res. elev. 1529--·

Gates ro,sed 9' _,·'
�
'--..,_
r--------
r-----� .
0 1504 I
Res.elev. 1529 - ---- --
---------
r-----:___
..
z

'-
Ga)es ra,sed 6'
>

I
0

I
I
Gates raised 6 '
;:: 1502

'
� 1502

Res. elev. 1529-. .

-- '--------"-
>
I1 �
w

� �r-------- I�
w

V
� 1500

'------- Res elev. 1529-- - - -- .

-
-
Ga tes ra,s�d 3 '

I \
c-

I
I
'--------r---,_
Gates ro,sed 3i
Res. elev. 1 529 , ___ ;r-.,
r---,_
'r--------_�
•"
"" I'----,_ r--. l
--�
}

i -�
I
1 ""--

'�
�
I"-.
(��
SP/LL�A
�
----------- SPILL WAY-; · . .

�
. II
PIEZOMErE OPENINGS -'- - ·
I
I
�
I),._
--. '
�
. PIEZOMETER OPENING1 --'.�
I I I Pr--r--,,--- II
I
NOT£: Alternate p1e1ometers ore offset
20 22 24 26 28 30 32 34 36
DI STANCE F R O M C R E S T A L ONG 't. OF G A T E No. 4

[71--.__
;-,,,_
approx,matly i mch from gate , .
I '[ti----,_
"
I
28 '30 32 3,: 36 38
D I S TA N C E FROM C R E S T A L O N G �
40 42
OF GATE No. 2
,, Crest f -

A, P R E S S UR E D IS T R I B U T I O N I N
S P I L L WA Y BUCKET TEST 15

NOTE· 1 Reservoir elevat,on 15 35

r-, 2. Pressure i[Jd,cated wdh Sp,llwoy f - -,,
---- 1:p i --Sta. 0 150 .,/_,-Sta 1150
F:'sta 0100
pilot tube
'r---
�-,,r --.
1-- _
-- r----
;---, '1----.
• al gate Na 4--·· -----"---
,

:+=- -
��
150

r----1'--,,
:I-
f a/gate Na 2 ------
,,

'�\\, , '-,___ - "k

,
O,stance from

,r--,
- - ,,,______ ox,s ofcresf

r
,

-,
'l'y---,_'--;,,,
,
Sta 42.8 -;

'--,.__ ---!._
r--.

',, -,1,_ I', '--,___ ;----,. -!-,,__

P L A N O F S P I L L WA Y
, ·Sta 52.8

1"'-�: � '--r-,,, - ---r--- :,.:--..

,1,� � i---, , -Sta. 62.B
.. -· ·--
----------,
'-,-,

Sta 82.8- - - --·

��--, �, '
-
Sta. 72.8

r--- , -
,---
',, '-----,
',,

-� - '-- '--
Sta/02.8 --

',,
�I'--

00 10
.....� 13 14
""15 16 17 18 19 20I"'- 21 22 23 24 25 26 27 28 29
I',
30 31 32 33
PRESSURE JN FEET OF WA T E R
8,RELATION O F P R E S S U R E TO D E P T H I N C H A N N E L T E S T 17

SALT R I VER PRO J E C T - A R I Z O N A

S T E WA R T M O U N TA I N DA M
H Y D R A U L I C MODEL STU DIES OF SPIL L WA Y CHANNEL
PRESSURE DISTRIBUTION IN BUCKET AND CHANNEL
1,50 MODEL
 The first measurements were concerne d with the left wall (Test 18 ) .

The average and maximum water surface , for maximum discharge , was measure d

by a point gage and plotte d as shown in Figure 17D . This water surface

was quite uniform and close to the theoretical water surface compute d by

the Des ign Section. The proposed wall, 5 feet above the theoretical water
surface , was satisfactory and could be use d in the final design.

The height of the right wall could not be establishe d in this manner

(Test 19 ) . The propose d wall 1 5 feet above the theoretical water surface .1
was below t�e actual water surface between Stations 0+30 and 1 +25 , and

betwee n Stations 2 +00 and 2 +25 (Figure 17E ) . In previous tests , the height

of the water surface was re duce d as much as possible by using a fillet on

the channel floor alongsi de the right wall , but it was necessary to increase

the he ight of the wall over the desig�ed height . This was done between

Stations 0+37.5 and 2+50 as shown in Figure 17E . Even with this higher
wall, the water surface came nearly to the top, and the desirable freeboard

of 5 feet did not exist . Since it would not be economical to increase the
height of the wall, an overhanging lip , or sea wall , was place d along the

top between Stations 0+70 and 3+00 as shown in Figures 17B, 17C , 17E,

an d 18D. Flow conditions along this wall we re observe d for various com­

binations of gates discharging to o bserve the effectiveness �f the sea

wall (Figure 18D ). The study indicate d that the sea wall should be

exten de d upstream to Station 0+64 an d downstream to Station 3+50 to con­

trol all high points of the water surface along the wall.

31 , Water Surface Profiles Across Channel

With the sea wall installe d 1 the final design was .obtaine d 3 the water
surface a cross the channel was me asure d (Test 19 ) at various stations and

35
 FI GURE 17

-Wo l l of m odel extends

to elevotion 1530

0
0

f. ._,____,___

·u
---------------- 3�- ----------------------- �
Fillet --·, I
_.-Sto 4 t 50
Fl oor of chonne t - · ··

\·Ve r t i c a l f r l l e l

S E C T I O N LOOKING U P S T R E A M S H O W I N G
B S EAWALL D E T A I L A T S T A T I O N 2 + 50

I
A P L A N OF C H A N N E L

I
o , 140000 Sec. ft:· I I J - I
I

I
---a = 100000 Sec.W'" I I
1500 ' 500

-------�
t---.. 0 • 50000 Sec. ft.-·· I
Fioor of channel· _;-'
I-
r----

-
�I
Station Q t 50
;;;:---
I\\\\

- ----+--
l
I
�- F, llet !/'"
o, 140000 Sec f t:· ,,
I I ' 500
o , 10 0000 Sec. ft
I
"
I
' w'

-
--- · Seo wal l OSec.tt

�
Floor of c h a n n e l ··
Station I t 50

�\� -
1510
1
0,140000Sec. tt.L.

� - ·• ..--------=io,50000 5ec. 1t.I-- I

o, 100000 Sec. IL

�
---+-:::::::---
1490

bt �
./ , �
Floor of channel ·-··'
,.----Seo wa l l I '

I
<

_J
Station 2 + 50

� �
>
w

- ·-----Fillet'
L-------

---.--
J

0•'10000 Sec ft:·

I
O =IOOOOOSec ft.-.

I
I
0•50 c. t ·· , I '
I Floor of channel----·

,_________.---
, _
I Stot,on 3.f50

-----
-
0,140000 Sec. ft-..
0

o,100000Sec. rt- ·-. V

I
-"
0•500005ec. ft:·

r---
Floor of channel----·?
Station 4 + 5 0
I I
100 1 25 1 50 200
D I S T A NC E FROM RIGHT CHAN NEL WA L L

C WAT E R P R O F I L ES I N T H E C H U T E

�
S T A T I O N ON C H A N N E L {

"
1 + 50 2 +50 3t50
"I
7"'c··
I L II
O t 50

I Propes d wall! feet above

I -"
,I I
Left end pier
I I compu ed water surface - ·

IC:.:.;:'1_L
1 .
Average--·.,
I

I
I Crest., I I I:

I - -I - - - - - - -I - - - -
Meas:1 "ed water surface max1murn- - . P, I

I I -
/ Chornel floor -·

� � '20 ' 80
::.,

I I
440 460 500 520
D I S TANCE FROM C REST A L O N G L E F T C H A N N E L WA L L ·
0 W A T E R S U R FACE A L O N G L E F T C H A N N E L WA L L

STATION ON CHAN N E L

,I I I I I
i
3 + 50 4t50
'
[71 .I, �'
I
·. -ve/t,col 1!11et J
- - --F, let on oor ofc onnel
I
-J�
z :,- Propo sed w�(1 ·B I S A L:T R I VER PROJ ECT - A R I Z O N A
0 · - ottom of sea wal l

"\:�I .,�_ �,,;_ 1 I -r·

;:-,.__ .- - ·1 {Top coincides with top of proposed woll) STEWART MOUNTAIN DA M
� 1500
Measured wate r surface
._ Ave rage-...
[ M0A1mum·., HYDRAULIC MODEL STUDIES OF SPILLWAY CHA N N E L
j-.___ -+-=:::__ WAT ER SURFACE ALONG WALLS A N D
I
--i::._ I
.. IN C H U T E OF PROPOS E D DESIGN

'80 '
Channel fl o�r--
I I I
7 ---i --i-
I I I I
1 : 50 MOD E L

200 220 240 260 280 300 320 3 0 360

DISTANCE FROM CREST ALONG RIGHT C H A N N E L WA L L
E WAT E R S U R FA C E A L O N G RIGHT C H A N N E L W A L L
 B . Looking downstream,
showing sea wall and
fillets on piers and
channel floor.

A . View of model showing :position

of overhanging sea wall .

C . All gates p,.rtially open to D. Second gate from end open

discharge 85, 000 c . f. s , to showing water being turned
. ah.aw action of sea WBr-1 . into channel by sea wall .
�
.....
HYDRAULIC MODEL STUDIES OF STEWART IDUNTAIN DAM SPILLWAY 1
I-'
CP
USE OF SEA WALL ON LEFl' CHANNEL WALL, 1 : 50 MODEL
 for several discharges . In the first run, the radial gates were raised
sufficiently to maintain the reservoir at ·elevation 1529 . 0 with a dis­

charge of 50 .� 000 second-feet . Point gage readings were taken across the
� channel at Stations 0+50, 1+50, 2+50, 3 +50 3 and 4 +50 , to measure the depth

of water in the channel . The gates were completely raised and the water
surface held at elevation .1529 02 to give a discharge of 100, 000 second-feet,

and point gage readings were again taken across · the sections . The third

run was similar 9 with the gates raised, t he water surface at elevation 1535. 0

and the discharge 140, 000 second-feet . The results are shown on Figure 17C .

32. Calibration of Mo del for Free Discharge Over Crest 2 and With All Gates

at Same Openings

When an , acceptable design of the spillway �as obtained, the mo del was

calibrated (Test 20 ) to provide rating curves of discharge in the proto­

type (Figure 19 ) showi ng the relation of discharge to water surface

elevation . In p�evious tests, coefficient curves were used to compare the

capacity of t he different spillway designs .

A prototype rating curv.e for free discharge over the crest , showing

the relation between reservoir elevation and dis charge, w as computed from
a model coefficient curve (Figure 19E ) . Seventeen runs were made on the

mo del with the gates fully raised to measure discharge at various water
surfaces between elevation 1506 an d 1535 . The. coefficients were computing
using the relation : Q = CLH3 /2 , where Q = discharge, C = the coefficient ,

L = the crest length, and H = head in reservoir above the crest .

36
 GATE OPENING IN FEET
' '
GATE OPENING IN FEI:.T
' ' ' .''' I
GATE OPENING I N FEET

: ' ' i' i' i ' i l ")'1 1?Y1 t)iJ;;l_pv tl I 1 J'; ;""' I I/ ,,,v v
v
,'
'l 10 I I 12 13 14 15 16 17 1!

'
10 12 1 !

,I f I I, VI 1�
'
i I
,- I I II
I I I I I ���/l I I
/

I I I I ,1/AlAll /1/i '

!0
7
I- -

I I I I I,r;,l il!V ,0/

Ii
I
I
•- I , I I f /1:11 I I I I I I II/I II 1 1 /

f'/
;
I

I I I I Ii'/1 1 'II / I I I I II ·'

I

'
/(
� Freed,scticf'tje over the crest
' L; I
I I �I I /1 II /
0 /
/
1528 2

I I I / 'I; II V /
1528 2 for goh no l -..1fti gate no. 2

' ' I ---Free discharge

' '
�-. --Free d,scharge
-
in cperot,on

1/
i)'.I'/ /
o,,erthe cresf over'ttit crest
l
I I 1,1½
20 1526 2 1526t::J2

'I �
1526

'
-Freed1schorqe aver rhe crest-..1th
! ' I 'I 1 h
I -h -/7�
� gote no. l on!1 illoperof10n �

' '
t524

I I I I I I/ II- I /1 I/'I
1524 18 152<1 1 ,-
i �
' � ,,/ ,_
.' �1W:VVI I I II /
1
•--j

I II1 ljl�
1522 I
/
1522: 18 1522 1
� i

� I 1/1I'/,
1520 �
/
1520, I ,,_ 1520 ; I

I 1//.r1 V '
-
0
1/

. 11 /1 1 1/
- 0

'IIV
1518

' II '//.V
15!8 :r I

'
1518 l
�

' I'llV / II v . // I/1 1/

:r

1516 � 10 1516 � 1 1516 �

.' II ' '/ : I I/ '

' -;,
'I
', - . - 5 6 7 II 9 10 II l'l
:;
'- - - - 5 6 7 8 'l 10 II IZ ·-
:;
' - - . . .. .. ..
.. .. .. .. .. .. .. .. .. 5 6 7 8 9 10 n 12 . .,
DISCHARGE IN THOUSAND SECOND FEET DISC,..AAGE IN THOUSAND SECOND FEET DISC,-.AAGE IN THOUSAND SECOND FEET

DISCHARGE THROUGH GATES 2 TO 8 INCLUS I V E , DISCHARGE THROUGH GATES 2 TO 8 I N C L U SI V E ,

A DISCHARGE T H R O U G H G A T E NO. I B C
ONE ADJACENT GATE D I S C HARG I N G BOTH A D J A C E N T GATES DISCHARG I N G

. .
�--- �---
' ' y
GATE OPENING IN �EET

i . i i t/ 17? ?
12 3 1
V
' ' ' . ' ' '
GATE OPENING l N FEET

I I I/I I I I ;' ;'//

" " ' /1 ;s
' I I I IIi!///; V I/ V / V / /
; 7/ I I
0

II I ' I I
---
T

I I / I I I,r11ill ./ I I II II IVI I /I/ // / V /

/

-----
I
' I I lI 1'/ /1Iill ii / ' I / //I///

11 7 I-; j / / I I I I I I 7[7 V v1 /[/ ----

/

I I I I I/ I v/1 /V/ v/v �v--

'
/

.
'/
20

I . I 'Ir/ /1 II i/
1526
F�ee d,schof'ge ··- Freed15chor9e
/
�
over the crest over ttie crut

I I I II 1 /IV I II '/
1524 1

' '
� ·-
�

II /
1 522 <1 1 ./

I I '11
I/ /
I17,�6'.V
V/ /

I /1 I I !1
1520 ; I I /

' '
/ I I ;/ ,,,v
1518 I /

'
�

I II/I
<t- 1 516 � I

�
// - ,- 11 ,,
SALT RIVER PROJECT - ARIZONA

• I /
STEWA R T M O U N TA I N DAM
I/ HYDRAULIC MODEL STUDIES OF

:I
SPILLWAY CHANNEL
HEAD-DISCHARGE CURVES

';
'1
1<50 MODEL
I
0
o 1 2 l 4 5 8 9 10 II 12
DISCHARGE lN TliOUSAND SECOND FEET
I 14 15 16 !7 ,, " I 20 2 � 35 40 45 50 55 60 6� 70 75 80 85 'l0
DISCHARGE IN THOUSAND SECOND FEET
'l5 !00 105 110 US 120 l'l� 1]0 1]5 140 14 I�

DISCHARGE THROUGH GATE N O. 9 W J T H [ D I SC H A R G E T H ROUGH ALL GATES

Q
AlJJACENT GATE DISCHARGING
 Calibration measurements were made for gate openings as follows :

Gate
opening Number Range of water surface
feet runs made elevation in test

18 7 1530. 6 to 1535

15 9 1527 . 0 to 1535

12 10 1523 . 9 to 153 5

9 22 1518.5 to 1535

6 10 1515. 9 to 1535

13 1511.4 to 1535
1-1/2 11 1509.8 to 1535

In applying this data to prototype, a method of analysis was used whi.ch

subordinated the influence of in,dividual rea dings to the general trend in

a manner similar to the analysis with the gates completely raised . The

empirical expression Q = K (H-b ) n was used where Q = discharge, H = the

head, and b, K, and n = constants to be determined. The r elations of

Q and H were plotted on log-log coordinates . By a proper selection of

the value of the constant b , the relations of � and H-b plotted as a

straight line on log-log coordinates . Then f rom the expressions of data

as log Q and log H-b the values of K and n may be determined least squares ,
by connecting to prototype terms , an d by interpolation the curves of

Figure 19A, for various gate openings were obtained"

The curve for free flow over the crest was not joined with the curves

for discharge with the gate partially open since there is a discontinuity
• or unbalanced condition between the two types of flow . If the water surface

in the· reservoir is held constant , and the gates are closed from the wide
open position, free discharge over the crest will occur until the gates

37
 touch the flowing nappe . By impact , a head of water is create d against

the gate of sufficient force to cause the nap pe downstream from the gate

to cont�act re �ucing the effective area of the jet and re ducing the
•
discharge . A similar situation occurs when the gates are being raise d.

At the o pe ning whe re the nappe breaks from the gate to flow free , the.
discharge increases . No tests were made to study this condition beyond

general observat'ion bec�use of the difficulty of holding the water surface

in the fore bay of the mo del constant while a sudden change of discharge

occurre d .

33. Simplification of Right Wing Wall of Spillway-

As sliown in Figure 13B , the bottom of the wing wall at the right
side of the spillway w as forme d by a sloping fillet wit� a warped surfaca

below e le vation 1502 . 0 . As such a fillet would be difficult to form in

concrete, it was propose d that a flat bottom be use d if similar flow

conditions could be obtaine d by extending the w ing wall downward several

feet below the original position. A long wall was place d in the forebay

at the left to isolate this gate ( Figure 20A ) .

The discharge and head w ere me asure d with free discharge first , with

the fillet on the bottom, Figure 13B, (Test 21); second/ with a flat

bottom and the wing wall at elevation 1497 . 0 ( Test 22 ); and third, with
a flat bottom at elevation 1502 (Test 23 ) . There was n o appreciable

effect on the discharge capacity . Therefore J the pro totype design

include d a wing wall at the right end with a flat bottom at elevation 1502. 00
• (Figure 20A ) •

38
 Figure 20

A. Left end gate o:pen with wall at right

:pier to obtain ea.me a:p:proach conditi ons
as if adjacent gate were o:pened.

B. Building IBPer along r is}lt wall to show

extent of splash at maximum di scharge._

C. Building J;S:per along left wall to

show extent of splash.

HYDRAULIC MODEL STUDIES OF ffi'EWARI' MOUNTAIN DAM SPILDTAY

WING WALL ON PIER IN CALIBRATION TESTS
AND EXTENI' OF SPLASH OVER CHANNEL WALLS.
 34 . Calibration 'of Flow in Right End Gate, Gate 1

In addition to operating the spillway with all gates open the same

amount, it may also be operated by opening individual gates . As suggested

in Test 14 , a satisfactory schedule wo uld be to open the right gate,

Gate 1 , first , then the adj acent gates consecutively. The right gate was

calibrated with the adj acent gate closed, and then calibrated with it
open, because the contraction of the nappe at the left side will be

changed, a ffecting the discharge capacity. In Test 23, the gate was held

open and the free flow over the crest measured . Tbe long wing wall shown

in Figure 20A was placed at the left pier to suppress the flow in a manner

similar to the adj acent gate being open . Test 28 was similar, but with

the long wing wall removed to r epresent free flow through the r ight end

gate with the adj acent gates closed . The' procedure of the tests and

metho d for obtaining these curves was similar to that described' in Test 20.

In Test 29, the right end gate was cali brated for partial openings

with the water surface in the reservoir at various elevations . The curves

for partial openings and free discharge are shown in Figure 19A . Since

the long wing wall at the left pier used in Test 23 has been removed,

Test 29 represented the flow through the right end gate with the adjacent

gate closed . During this test , it was observed that vortices would form
at either, or both 9 �ides of the gate , usually when . the gate w as at a

wi de opening and when the head of water on it was not large . Some tests

were made by placing the long wall on the left pier to suppress these

• vortices . No material change could be observed and it was concluded that

the model was not sensitive enough to warrant a study of their effect .

39
 Mor�over , the gate was not calibrated for partial openings with the long

wa:p_ at the left pier i.n place which would represent the condition with

the adjacent gates also partially open . It was believed that such a
C,
calibration woul d not differ greatly from the curves shown in Figure 19A .

J5. C al ibration of Flow in Center Gates

•
If the recommended operating s chedule is followed, after the right end

gate, Gate 1, is open, the adj acent gates , Gates 2 to 8 , will be opened

consecutive� . Thus the flow through each of these gates must be £ali­
brated for two conditions : ( 1 ) with the adjacent gate to the right open,

( 2 ) with the gates on each side opened. If each gate were t hus calibrated ,

there would be a duplication of work , for it may be reasonably assumed

that the calibration of any one of the seven center gates will be repre­

sentative of the others . Gate 5 in the center of the c rest was selected

for the tests . In Tests JO and 31 .9 a single long wing wall similar to
that shown in Figure 20A, was placed at the right pier of Gate 5 to give
the same approach conditions that would o ccur if the adjacent gates to

the right were opened • . In Tests 24 and 25, two wing walls were used,
one at each si de of the gate , to give the same a pproach conditions that
would o ccur if adjacent gates on both sides were opene d . These tests were

conducted in the same manner as the calibration of the right end gate .

'rhe results of Tests JO and Jl are shown in Figure 19B and the r esults of

· Tests 24 and 25 in Figure 19C. The curves for a gate partially open -are
•
not v alid if the adj acent gates are not at the same opening a s the one

"' under consi deration .

40
 36.. Calibration of Flow in Left End Gate

In the suggeste d procedure for opening the gates indivudually,

Gate 9, at the left end, will be opene d last . Since the embankment at

. the side would cause an approach condition peculiar to this gate, it was

..
calibrate d separately. A long wing wall was place d at the right pier of

this gate to represent the approach condition that woul d occur if ap

adjacent gate to the right were opene d . T he proce dure for the calibration
was similar to that for the previous tests , and the results . are presente d

as shown in Figure 19D . The curves for specific gate openings are valid

only if the adjacent gates are opened the same amount as Gate 9.

37. Velocity Distribution in the Channel

It was desirable to measure the velocity distribution in this final

design for a comparison with theoretical velocities upon which the design

was base d._ · Velocities were me asure d on the original design of the

1:100 mo del , but the results were unsatisfactory because that model was

too small . In this Test 32, on the final d esign, the velocities were

me asure d at 15 positions indicate d in Figure 21B . The discharges were

50, ooq, 100, 000, and 139, 000 second-feet, with the water surface at

elevation 1529. 0 for the smaller discharges and elevation 1536 . 0 for the
maximum discharge . The pitot tube use d was a special static leg type

being small in diameter to give a minimum disturbance to the flow . Since

the dynamic pressure only was recorde d, the static pressure had to be
•
determine d by the depth of the water . The results o f these measurements

are shown on Figure 21A, and are compare d with the theoretical velocities .

The theoretical velocities were base d upon a d rop f rom elevation 1529 . 0

41
 to the point in question assuming no loss . Therefore, the theoretical

velo cities are · shown ru, increasing with depth although it might be argued

that they should be constant with respect to depth to account for the

static head . The velocities were higher ( Figure 21A ) at the right side

of the channel as anticipate d !! and somewhat less than the thebretical

•
velocities . It appears from these curves that the losses increase down

the channel for t he difference between the measure d and t heoretical

velocities at Station 4 +50 is greater than upstre am� but this data is too

irregular to draw any definite conclusion� .

38 . Splash Over Spillway Walls

During demonstration runs on the final desig�, it was observe d that

particles of water splashe d over the walls , eSJ)ecially at large discharges o

It was iq,ossible to pre dict what form this splash would take in the

prototype, whethe r it would be as S pray or as slugs of water. To show the

extent of this splash, a study (Test 33 ) was made with all gates wide open

and with the water surface in the reservoir at e levation 1535 0 0 to obtain

maximum discharge . A strip o f building paper was f astene d to each wall

and when a particle of water struck the paper, a stain resulte d which

coul d e asily be seen . The splash over the right wall (Figure 20B ) was

not s erious , although some spray struck the paper about 10 feet (prototy pe )
above the wall • . The splash along the left wall was more severe (Figure 20C ) .
The particles of water were concentrate d between Stations 9+30 and 2+30

with a maximum splash height of 40 feet (prototype ) above the wall at

., Station 1+00 . Such spray might be serious in the prototype only if the

spillway were ope rate d at maximum capacity for a long period of time .

Since this was unlikely , the condition was not considere d critical .

42
 39. Discharge Coefficients of Final Design

The discharge coefficients of the final design (Test 34 ) were

measured for a comparison with the previous tests to show the effect of

the various changes upon the capacity of the spillw�y , made after the

design described in Test 10 . Briefly, these changes were, a long fillet

in the channel downstream from the crest_ alongside the right wall , a

fillet on the right wall immediately downstream from the pier to make the

surface of the wall flush with the pier.� the installation of a sea wall,

mo dification of the sloping bank upstream from the left pier, and simpli­

fication of the bottom of the wing wall on the right end pier. These

changes increased the coefficient (Figure 12B ) slightly over that of

Test 10 at the lower discharges , but had no material .e ffect at the higher

discharges .

40 . Discharge Coefficients With Spillway Channel Removed

To show co nclusively the effect of the channel on the discharge

capacity of the spillway, Test 3 5 was made with the spillway channel

removed, and coefficients were measure d and compared with those of

Test 34. Vlhile no difference could be noted at the lower discharges, the
discharge capacity was increased about 2 percent with the maximum flow ,
with the water surface at el evation 153 5 . 0. Evidently , the flow over the
crest was partially submerged by the channel at the high discharges ,
proba bly at the. l.eft side .
" 41 . Discharge Over Crest Wit,h Piers and Channel Removed

To show the effect of the changes upstream from the crest, the

piers and channel were removed so the mo del was similar to Test 2 of

43
the preliminary studies . Figure 12B indicate s a slight re duction in the

coefficie nts over the final design. !here is no logical e xplanation for
· this condition as the chartges to the topography and conditions upstre am

s hould nave te nded to increase the capacity .

44
 " •
I
t.' ..,

1495 --
-
§ �'
'\
,os,r,o• ,

-
I
'
1 4 85

1 4so J f-- . . ' -'- I , j�

_;_I • . F,\� j'
1480 =====---r----------,--i---,.;;.,�;-;-�7
'- ' 1·
+- - - ',( _ _j -·i
'
J-
hese reod1ngs were token
i 4 1s -+-- 0J o point 15' from the

· -j\'
l left chOnnel woll_·· 1 ----
'"' I- -· ',t ··-· I , ,1
i--i- '"-==;=:::::,;::==;=���=�=�=�=�=::;:=::;:=:::==:::::::::::
·14832

,. ,'
' j -
6
114as
4 5
POSITION 2 1
,soo I - --'--- ---L-

- ,- --' -
'' 1 4 8 5 ,_______,_ __

' - -
-
-
-;
�
'"° �T"U I ±:=:e= :-i I I I ', I I I
I
' I I
I rLI
l
14�0
POSITION 3
!
T l _j_
-
POSITIO N 3

I� I I I
��,-!..._'' '
I- -'-- '-----

-- ""-' II II I'
I :,,_

'
. I '--
'""
'
IJ-,-
� 1475.3

I I l
__L_
I
---,: :..: --� 1 -
I I
-t=±-
jWoter surfoce POSITION 4

'
I

,, ..·· ·1 :f����:;e���'.
1500
I I
I I -Cl 9 0 1 �
T 1
I
losses.
{/�'s�ud�����e
. ' l�\'ft\
--r ,,
f---1--1-1-
I
'505
IIOO,ooosec ft-

of chonnel.
Ele11ot1on of floor
i
I __.:.-- /'
/',. ... . . r
;
- ,,;ago I I � i I ,,iao
b 1, �
�
:::� �::::::.;:-:;:::::::::!:::;':::;:::�:::::=
�

����:���:�----'- -+- ,- ----,---,---.----.------------------l

l I I
-. .:-sopoo sec.ft
' I T 1478.8
I

t,, '
1495
1505 POSITION 5 POSITION 5
!

'
"')

""
1 5 00

1491.5·:- ''-t,,,I I I I ' ;.·,

-

" " ,, "

T
�-

30 35 so 1480 �
00 45 50 60 45 60 35 40 so "
VELOC I T Y IN Ft P E R SEC. V E L O C I T Y IN FT. PER SiC. V E L O C I T Y I N FT. PER SEC.
STATION O t 50 STATION 2 t 50 STATION 4 t 50
"$ A RELAT I O N OF V E LO C I TY TO D E P T H

Note For water surface profiles

,·3 see figure 17

r-------___. --------------
. . ·Sta. 2 + 50

"'-'---l
---
3

---
----------
,•2 Sta 4 t so ...... :
S A LT R I VER PRO J E C T - A R I Z O N A
�G>
C
STEWART MOU N T A I N DAM ;c
rn
H Y D R A U L I C MODEL S T U D I E S OF S P I L LWAY C H A N N EL I\)
VELOCITY MEASUREMENTS 1,50 M O D E L