spillway tes hidraulic

8/13/2019 Spillway Tes Hidraulic

1/71


2/71

United States Department of the InteriorDOUGLAS McKAY, SecretaryB ureau of R eclam ationW . A. DEXHEIM ER, Com missioner

L. N. McCLELLAN, Assistant Commissioner and Chief Engineer.

Engineering M onographNo. 16

SPILL WAY TESTS CONFIRl\fMODEL-PROTOTYPE CONFORMANCE

ByA. J. Peterka

Engineer, Hydraulic Laboratory BranchEngineering Laboratories Division

. Technical Information BranchDenver Federal CenterDenver, Colorado

.


3/71

.

ENGINEERING MONOGRAPHS are publishedin limited editions for the technical staff of theBureau of Reclamation and interested technicalcircles in government and private agencies.Their purpose is to record developments, inno-vations, and progress in the engineering andscientific techniques and practices that are em-ployed in the planning, design, construction, andoperation of Reclamation structures and equip-ment. Copies may be obtained from the Bureauof Reclamation, Denver Federal Center, Denver,Colorado, and Washington, D . C .

.


4/71

CONTENTS

INTRODUCTION. . . . . . . .. .. .. .. .. .. .. . ..HEART BUTTE DAM STUDIES. . . . . . . . . . .. . . . . . . . ..

Description of Project. . . . . . . . . . . . . . . .. .. .. .. . . .Model Tests . .. . . . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. ..1. Spillwayand pier tests. . . . . . . . . . . . . . . . . . . . . .2. Deflector and v ertical bend tests. . . . . . . . .. .. .. ..3. Outlet works and tunnel junction tests . .. .. .. .. .. .. .. .. .. .. ..4. Stilling basin tests. . . . . . . . . . . . . . . . . .. .. .. ..5. Spillway air tests .. .. .. .. .. . .. .. .. .. .. .. .. .. .. .. .. ..

The 1950Spring Flood. . . . . . . . . . . . . . . . . . .. .. .. ..M odel-Prototype Com parison Tests. .. .. .. .. .. .. .. .. .. .. .. ..1. Spillwaycapacity. . . . . . . . . . . . . . . . . . . . . . . . .

2. Spillw ay perform ance--Free and subm erged discharges .. .. .. ..3. Spillwayair demand. . . . . . . . . . . . . . . . . . . . . . .4. S tilling basin perform ance . .. .. .. .. .. .. .. .. .. .. .. .. ..5. Erosion downstream from stilling basin .. .. .. .. .. .. .. .. .. .. .. .. .. ..6. Tail-water elevations . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..7. River erosion. . . . . . . . . . . . . . . . . . . . . . . . . .8. Inspection of structure following 1950 a nd 1951 floods. . . . . . . .SHADEHILL DAM STUDIES .. .. .. .. .. . .. . .. .. .. .. .. ..

Description of the ProjectSummary of Model Tests

.. .. .. .. .. .. .. .. .. .. .. .. .. ..

.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..1. The m odel .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..2. Spillway air tests .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..3. Stillingbasin tests. . . . . . . . . . . . . . . . . . . . .

1952 Spring Flood. . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..ModelPrototypeComparisonTests. . . . . . . . . . . . . . . . . . .1. Spillwaycapacity. . . . . . . . . . . . . .. .. .. .. .. .. .. .. ..

2. Spillway performance--submerged and free discharges3. Vortex action and effect of ice. .

.. .. .. .... .. .. .. .. .. .. .. .. .. .. .. .. ..

4. V ibration. . . . . . . . . . . . . . . . . . . . . . . . . . . .i

Page111235668

11121214192124272929303035353535354141414549


5/71

5. Debris. . . . . . . . . . . . . . . . . . . . . . . . . .

5. Tunnel air demand .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..7. Spillway head loss .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . ~ ..8. Stilling basin perform ance .. .. .. .. .. .. . .. .. .. .. .. .. .. .. .. ..9. Tail-water elevations. . . . . . . . . . . .. .. .. .. ~

10. Erosion below the stilling basin . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. efuspection of the Prototype after the Flood .. .. .. .. .. .. .. .. .ACKNOWLEDGMENTS. . . . . . . . . . . . .. .. .. .. .. .. ..

ii

5050515252586262


6/71

Number12

3

4

5

67

8

9

1011

1213

14

15

16

171819202122

2324

LIST OF FIGURESDescription

Index and vicinity maps .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..Plan and sections of spillway and outletworks at Heart ButteDam. . . . . . . . . . . . . . . . . .Laboratory model of spillway and outletworks for Heart Butte Dam. . . . . . . . . . . . . . . . .Heart Butte spillway model tests at adischarge of 3,750 cf s . . . . . . . . . . . . . . . . . . .Entrance details of spillway and outletworks at Heart ButteDam. . . . . . . . . . . . . . .Flow through vertical bend in Heart Butte spillway model .. .. .. ..Tests of junction of spillway and outlettunnels for Heart Butte Dam. . . . . . . . . . . . . . . . .Stilling basin details for spillway andoutlet works at H eart Butte Dam. . . . . . . . . . . . . . .Testing Heart Butte stilling basin modelat a discharge of 5, 600 cfs . . . . . . . . . . .. .. .. ..Hydrographs of April 1950 flood on Heart River. . .. .. .. ..Reservoir elevations and hydraulic dataof 1950 flood a t Heart Butte D am. . . . . . . . . . . . . . .Spillway discharge capacity curve for Heart Butte Dam .. .. .. .. ..Discharge comparison of Heart Buttemodelandprototype. . . . . . . . . . . . . . . . . . . .Performance of spillway and outlet worksat Heart Butte Dam. . . . . . . . . . . . . . . . . . . . .Reservoir at el. 2073.8, crest submerged9. 3 ft., discharge 3 ,250 cfs. . . . . . . . . . . . . . . . .Operation of Heart Butte spillway withp iers su bm erg ed . . . . . . . . . . .. .. .. .. .. .. .. .. ..Reservoir at el. 2074.0, discharge 3,260 cfs .. .. .. .. .. ..Reservoir at el. 2071, discharge 3,090 cfs. . . . . . . .Reservoir at el. 2068.6, discharge 1,600 cfs .. .. .. .. .. .. .. .. ..Reservoir at el. 2070.0, discharge 2,650 cfs,36 inches of ice on the w ater. . . . . . . . . . . . .. .. .. ..Spillway air vent system at Heart Butte Dam .. .. .. ..A ir demand curves for model and prototypefor varying discharges. . . . . . . . . . . . . . . .Outflow 3, 250 cis, stilling basin operating efficientlyPerformance of the stilling basin at adischarge of 3,600 cfs . . . . . . . . . . . . . . . . . . .

ii i

Page2

33

4

56

7

8

910

1112

12

13

14

15161617

1819

2021

22


7/71

Number')~~0

26

? ',

28

29

30

31

32

33

34353637

3839404J

42

4344

45

46

474848

DescriptionPerformance of the stilling basin ata discharge of 3, 600 cfs. . . . . . . . . . . . . . . . . . .Profiles of comparative discharges ofmodel and prototype spillway . . . . . . . . . . . . .Profiles of comparative discharges ofmodel a nd prototype spillway. . . . . . . . . . . . . . . .Comparative cross sections of riverchannel before and a fter 1950 flood. . . . . . . . . . . . . .Detail cross sections of river channelbefore and after 1950 flood. . . . . . . . . . . . . . . . .Erosion of the excavated channel dow nstreamfrom the s tilling basin. . . . . . . . . . . . . . . . . . .Comparison of measured and computedtail water elevations. . . . . . . . . . . . . . . . . . . .Flow conditions at junction of excavatedand natural channels. . . . . . . . . . . . . . . . . . . .Loss of bank material through erosionduring 1950flood. . . . . . . . . . . . . . . . . . . . .River channel erosion, compare with figure 23 . . . . . .Cast of eroded area in. vertical bend. . . . . . . .Cast of another eroded area . . . .. . . . .General plan and sections of spillwayat ShadehillDam. . . . . . . . . . . . . . . . . . . . . .Details of spillway intake for Shadehill Dam . . . . . . . . . .Details of spillway stilling basin at Shadehill Dam . . . . . . . .ShadehillDamspillwaymodel. . . . . . . . . . . . . .Test of Shadehill spillway model ata discharge of 5, 000 cfs. . . . . . . . . . . . . . . . . . .Reservoir elevations and hydraulic dataof the 1952 flood at Shadehill Dam. . . . . . . .Hydrographs of the 1952 flood at Shadehill DamDischarge capacity curve of the servicespillway at ShadehillDam . . . . . . . . . . . . . . . . . .Comparison of the discharges of the modeland prototype spillways at Shadehill Dam. . . . . . . . . . .Transition from submerged to free flow atShader..ill Dam spillway. . . . . . . . . . . . . . . .Vortex of submerged spillway flow at Shadehill Dam.Effect of ice on vortex action of the Shadehill spillway. . .Vortex action was continuous above el. 2284.9in Shadehillspillway. . . . . . . . . . . . . . . . . . . .

iv

Page2324

24

25

25

26

27

2829303132

33343637

38

3940

42

43

444546

47


8/71

Number505152

53

54

55

56

57

58

59

60

6162

DescriDtion.Vortex action at reservoir e1. 2283.6 was sporadic. . . . . . .No vortex visible at reservoir e1. 2277.8 and below . . . .Air demand and spillway discharge curvesfor ShadehillDam. . . . . . . . . . . . . . . . . . . . .

Air flow in tunnel was insufficient todeflect a trickle of water. . . . . . . . . . . . . . . . . .Overall head loss through Shadehillservice spillway structure. . . . . . . . . . . . . . . . .Comparison of water surface profilesof model and prototype spillways. . . . . . . . . . . . . . .W ater surface profiles of Shadehillspillway, 1, 300 to 3, 950 cfs ................Shadehill spillway stilling basindischargL '1g4, 400 cfs. . . . . . . . . . . . . . . . . . . .Shadehill spillway stilling basindischarging 4, 600 cis. . . . .. . & . .. .. . . .. . . . .Computed and actual tail water curvesfor service spillway at Shadehill. . . . . . . . . . . . . . .Flood erosion and repairs to spillwayoutlet channel following 1952 flood. . . . . . .. .. .. .. .Channel erosion caused by 1952 flood at Shadehill. . . . . . . .Erosion of excavated channel banks by1952 flood at Shadehill. . . . . . . . . . . . . . . . . . .

v

Page4848

4950

51

53

54

55

56

57

5960

61


9/71


10/71

IN TR OD U C TlO NThere is a general need for dab whichcan be used to compare the perform.ance of

models and prototypes and extend the rangeof usefulness of hydraulic models as all aidin design. Ordinarily, prototype data aredifficult to obtain and usually the model dataare not in the prototype range of heads ordischarges, which makes direct comparisondifficult. A lso, several years may elapsebetween model tests and the time the proto-type is subjected to large flows. The HeartButte and Shadehill Spillways, however, op-erated during the first flood season followingtheir completions and almost immediatelyafter the hydraulic model tests were made.W ith the test data on the models still freshit was possible to obtain prototype data onshort notice that could be compared withm odel tests.

This monograph compares the perform-ances of both the Heart Butte and ShadehillDam. morning glory spillway models with theperformances of the prototype structures.The results of these comparisons add furtherproof to the premise that prototype perform-ance can be predicted with accuracy fromm odel tests.

Brief discussions are given of the nec-essary hydraulic model tests conducted on thescale models to aid in the design of the struc-tures and to obtain data useful in operatingthe prototype structures. Following a de-scription of the 1950 flood on the Heart Riverand the 1952 flood on the Grand River, whichproduced discharges of 68 percent and 88 per-cent, respectively, of the maximum antici-pated flows, the perform.ances of the proto-type structures are described.

Direct model-prototype comparisons aremade of spillway performance and dischargefor free and submerged conditions; spillwayair demand; stilling basin performance, in-cluding erosion downstream from the basin;and tail-water elevations in the excavatedchannel. Photographs and charts a.re usedto illustrate the agreement found betweenmodel and prototype performance.

Certain aspects of the prototype per-formances which are beyond the scope ofmodel tests are also discussed. This in-cludes the effect of ice completely covering

the morning-glory during submerged dis-charge, the erosion of the downstTeam rive::--banks, and the effectiveness of the riprapused on the excavated cham 'el banks, Theresults of inspections of the spillway tun-nels and structures following the floods arealso given.HEART BUTTE DAM STUDIESDescriution of Pro,lec .

Heart Butte Dam is on the Ileart River60 miles west of Bismarck, North Dakota,and is part of the Heart H ive::- Unit of theIVIissOIlri River Basill Pr'ojed (figure 1). 'T 'hedam is of compacted eartl1 fill w ith arock rip-rap cover, rises 135 feet above str-ear:: bcd,and is 1, 860 feet long. It Serves both Lrriga-tion a..'1dflood-control purposes, A t m axim urnwater-surfaceeievatlOn the reservoir wiEcontain 392,500 acre-feet of water collectedfrom a drainage area of 1,810 square mile's.

Figure 2 s ho ws plan and sections of thenood-control spillway and of the outlet works,which, as the transparent model in figure 3clearly shows, is an integral pcu:..tof the spill-way structure. The spillway is a morning-glory type, 32 feet 6 inciles in outs ide diame-ter. It discharges into an 11-foot diametervertical shaft and a 900 diverging elbow thatlead to a nearly horizontal tunnel 14 feet indiameter and a.bout 800 feet long. Aroundthe vertical shaft of the spillwa;;- is the en-trance to the outlet works, wbch dischargesinto a 5-foot 3-inch dian1eter tunnel locateddirectly above the spillway tunnel and fromthere into the larger tunnel from above. Fromthe junction point the spillway tunnel carriesboth discharges into the hydraulic-jump still-in g b asin .

The morning-glory spillway is unusualin that tis designed to operate throughoutthe range of free discharge, the transitionrange between free and submerged discharge,and up to submergences a.s as 53.7 feetof water above the ,rest. ' The spillwa.y crestis equipped with six spaced piersplaced radially in plan, bu t does nothave con-trol gates of any kind. The outlet works dis-charge i6 controlled at the lower end of the

*To the author's knowledge, Heart Butte andShadehill spillways are unique L'1 tr.is respect.

1


11/71

outlet works tunnel by a 4- by 5-foot high-pressure slide gate.

The capacity of the spillway is 5,450second-feet at maximum reservoir elevation2118.2. The maxim urn vertical fall fromheadwater to stilling basin floor is about 130feet. The capacity of the outlet works is 650

i I \NORTH , D A KO TA : I ~ 8ISMARCK: . J Forgo I (HEARTiMONT i BU TTE D AM > \1 -I E -- - 9I / ,SHADEHIi])\ '1--- 7 - DA M \' TA: / SOUTH DAKO ,i I PI E RRE ' Huron ,, '0:WYO I Rapid. .: C ity

;i- N D E X MAP

niIu ItI III) NORTH I I ) ~ I:: ) B e l f i e l d DICKERSON

SOUTH0I 10ISCALE OF MILES

MAPICINITY

second-feet at spillway crest level elevation2064.50.M odel Tests

The discharge structures were tested ona 1:21. 5 scale model. The morning-glory,the outlet works intake structure, and thesurrounding topography were constructed

FIGURE l--Map of portions of North Dakota an dSouth Dakota that shows the locations ofboth Heart Butte and Shadehill Dams.

DAKOTAU.S,fO

N.J r

DAKOTA20 30 4-0 50I I I

2


12/71

4x5HPSiideo a t e o p e n in a , ,~ /:---;+''-~, 3 D' ( /.,- \ '~. l a - ' - ' - _ ~ E) -, ? 2 - 2 ,,' (H , \ ,-5R,(y~1') -) t.~,_I.i : J J ;. L-ju.-t:2-I flu \i(( ',,- I ( )- < -- - 0 \ -< - -\ ' tructures was larger than necessary. Con-

FIGURE 3--The essential parts of the laboratory model of the spillway andoutlet works at Heart Butte Dam were molded of transparent plastic. Theintake for the outlet works surrounds the vertical spillway shaft and thedischarge is controlled by the slide gate. Model scale was 1 :21. 5.

~


13/71

A--A violent vortex formed Lv; the splilw -ay at heads exceedir,g t:-~esubm ergence point. L'1etail of the vortex extended dO;.;Y lwardinto the horizontal tu... L 1el.

B-Six piers placed ra.dially en the crest reduced the vortex tonegligible size.FIGURE 4-Tests of the ,r.G rning-glory spill-.:ay model fo r Heart B-.rtte Dam at adischarge of' 3,:750 ef's.

4


14/71

sequently, the vertical shaft diameter wasreduced from 14 to 11 feet. This requiredthe reshaping of the spillway profile to fitthe vertical shaft, and the installation of a900 vertical transition bend at the bottom ofthe shaft. The discharge capacity was thenfound to be approximately correct, accordingto the irrigation and flood control require-ments.

Vortices which formed in the model whenthe spillway was submerged, figure 4A, werethoroughly investigated both experim entallyand mathematically. Since these same vor-tices could form to scale in the prototype,attempts were made to eliminate them. Var-ious arrangements of piers, dividing walls,and floating and fixed rafts were tested. Asa result, six spillway crest piers were rec-ommended for use on the prototype, see figure4B. It was found. unnecessary to extend th epiers as high as the maximum headwaterp.levation, a distance of 54 feet. Since vor-tex adion diminished rapidly when the headon the crest approached 14 feet, it was nec-essary to extend the piers only to this height.

2. Deflector and vertical bend tests.--Tests to determine the most satisfactory typeof vertical bend showed that a ~$1r~~lQQ,~joining the ll-foot-diameter shan with the14-foot-diameter horizontal tunnel had a dis-tinct advantage because it hE ,Q :Y JciE gg:::E ~tEs~~.~.P~tWE E .n. ..Uw..w.~t .gr .s i. lJ ,' 7f~~~~t11;~ ~.~-n el c rown f or :.YE n til? :t ng hE Y~ rt J~ af be p'gfrom th , atInosPl1ere at the tunn~l outlet.

However, even with this arrangement,difficulty was encountered in preventing thehorizontal tunnel from filling unexpectedlywhen the spillway and outlet works were bothoperating. Flow passing through the benddid not break cleanly from the crown of thebend. The flow had a tendency to follow thecroVl.'Il t hroughout the bend, causing a changein the location of the flow control. When thecontrol moved downstream the head on thesystem increased, causing an incr ~ase indischarge which filled the tunnel. This, inturn, caused a: s till greater head with a cor-respondingly greater discharge and resultedin negative pressures of considera:hle mag-nitudes on the spillway face. Once the tun-nel had filled, it was impossible to obtainopen channel flow again unless the head on

the spillway was reduced to a point consid-erably below where it had filled. To correctthis condition a small deflector was placedat the base of the vertical shaft on the down-stream, or crown, side of the shaft, seefigure 5. The deflector accomplished threethings: (I) it provided a positive control atthe base of the vertical shaft and preventedthe tunnel from filling, (2) it had a stabilizingeffect on smaller flows and provided a flatwater surface on all flows passing into thevertical bend, and (3) it provided a clear pas-sage for air to circulate as far upstream asthe base of the deflector. The thickness ofthe deflector at the base was varied in themodel to determine the size necessary toexactly meet the discharge requirements atcertain heads since precise tests had shownthat the ll-fpot-diameter vertical shaft wasslightly too large. Spillway flow in the tun-nel was found to be satisfactory after thestructure had been modified as described.Figure 6 shows the flow entering, passingthrough, and leaving the vertical bend withthe deflector in place. Note the smooth andflat water surface on the flow entering thetunnel.

4( A.;=S'~t / 3 2 C 6 ~ C _ ,.,.

,. '...

.1:::>

. . A. . '~ , '---'.if -, '/ - ,. ,. - ~ \ . \_ ;,~n/~CJ, -r-----.\.,/ /.

- -7 c--- .~~( ' ,'~~~;,- ...-PLAN

, ,-EL207850r :;:r'F- ;--', --In. .,' I I .,; I -, 2n 45 0I : :', I /'::1. vD .~ . '-.0i ~:~O 'RAi r , .' . ' \' ... ' ' . IE. . . ' ' .'ve'n+---; ~-,.- jL~i2~J I:. . ' ;j ~ :_~- '. :. '[: , ,


15/71

FIGURE. 6--The deflector at the base of the vertical shaft produced smoothflow and a flat w~ter surface in this test of the vertical bern at 3,750cfs. Al though the air flow appeared continuous to the eye, high-speedphotographs showed that air entered the flow in bursts. -c:::'' :;--, '(...~ ~ --:

-, ~--TI11ilii;

-- ~ T- t5'-91'

-~: --::-:---Z=..'-=-.; -'-.----.----.---'--'. -.~':::'~':::>-- ~D '''C ' C ' -,.

::52''r--=:-,v~+o~ :';..::~;--.:=~~~ n~ '='---=--3.l~--~ - . ~ -, -; 4-; -

L- StC_i3+24-.50

'-,., +6 ,,~

I

~3 :.~:: c:.+r.i~+ q -'.

-',y- ~,'El.2C'3.00'i-S : .

UW ~: ,~?:C~;~-~.-~.

'.

,.',..

~

.,'

'.

'~'.'5

~

O;~'

.

- E ' ~ : ~ ~ o i C +.50;,'~:\

,t 'f~'i9 5C I

- - . . -~-_.~ ~-__:~:~9::-~ I'_'J.~~1i

g9i.C14 - . -41-001 -SECTiON

----....-OETA:L.. A

. 0 i 2 ' 4 9 5 0 ~ , : : : : : : > : : : ~ ~ , c',::-:,,-

IiII

--, 3VQ:)e~-

P LA, NDETAiL C

- . 150 '- ( '_~.i

.FIGURE 8--Stilli.. 1g bs.sir.:. de-ta.ils for the spillw-ay and ou.tlet work.s at Hef:w ardhroughthe flow for high discharges but produced thedesired effect of distributing tile flow from 14feet wide at the tunnel portal to an ultimate42.5 feet wide in a horizontal distance of 75fepl:

Chute blocks and baffle piers were usedto increase the fine grain turbulence in thebasin and thereby reduce the required lengthof the stilling basin. . The shape of the bafflepiers, dividing wall noses. and trajectorycU Ves weremodif ie d to p ro v:ir le a tmo sphe ric

pressures or above on critical areas. sincetests on prelim inary designs had indicatedthat pressures as low as 18 feet of water be-low atmospheric pressure occurred down-stream from sharp corners. The recom-mended stilling basin is shown in figure 8.The perform ance of the developed stillingbasin. was evaluated from erosion tests madeon a. m ovable bed located downstream fromthe model basin and from wave height obser-va.tions m ade in the excavated ta.ilrace chan-nel. Erosion tests were made using a well-graded sand (100 percent passed. a No.4 sieveand 3 percent passed a No. 50 sieve). Thesetests showed that erosion tendencies wereless severe on the channel bottom than on 'thesloping sides. W ave action originating in thehydraulic jum p com bined with a slight surging

action caused rapid decay of the banks. Everyeffort was made to keep the wa.ves and surgesto a minimum, but it was deemed necessaryto riprap the banks of the prototype. Figure9 shows the perform ance of the recom mendedbasin.5. Spillway air tests. nWhen the morn-ing-glory spillw ay w as designed, it was an-

8


18/71

A-The reco:m :nended bas:L. '1, shown in operation, was redu.ce6. in dirre~sionsto the mL 'l1mum consistent with acceptabJ..e perfor ltlll'lce.

~-A l though S O m . E ~ erOS2G D 0 ccured, the opera tion.,1E.S consideredccept-able, since riprap protection w 'as to be u.sed ir l t he p ro to ty -- pe . Themodel produced the effect shown in a half-nour s operation.

FIGTJrtE 9--Tests of the model stillL'1.g ba.si::. for HeartButte Lfuil at 2. dischargeof 5,600 cfs and at tail -..later elevation 2012.Q


19/71

- {f'MGJI. infiow 30,500 CTS.'l-H Ui I I i I i~~ I .25 ~-r-: . I -t~---:--r---'-+ i --t--~-lr-T--+--+---t--12 0 ~ i : i : J U - - l ~ ~ ~j (/ i j Ii I l / \ j , i ii~ : ~ _ _ L Y _ + - - - ' - + r - ' \ i i -+--- - ,I;::: I , . . I ~--+---~ . ---:::< i5 :---+---1- . , I I---i t j - .I~ . I- ~lLi 1 \ \ \ \

: ..: I~ j- 1-i3crw i Jjj-l i I i - L I I I i '--~'---, -Max. Outflow 3760 cfs.: . I , II 1'=tv~ -~--+ L ~j5frr~-r~:r~~T~~.I-- L I - ~ r ~ i - 1

APRiL. 14 15 ,6 \7 ,8 19 20 2( 22 23 24 25 26 27 28 29 30

ticipated that air introduced into the spillwaydischarge at a point just below the spillwayCY'st might help to cushion the impact of thef'l.ow passing around the vertical bend. It wasL TD .portant that ur..necessary im pact and vibra-tions caused by the flowing water be elim-inated' because the entire structure was tobe constructed on sand. I'urtherIl1ore, ifforany reason cavitation should occur inornear the vertical bend, thepresence of theentrained air might reduce the tendency todac--nage the concrete tunnel lining. Labora-tory tests have shown that even very smallau~tities of air introduced into the flow willdelay the appearance of cavitation damage. '

Model tests on the many devices pro-posedto increase the entrained air in the flow

*A. J. Peterka, The Effect of Entrained Airon C avitation P itting, P roceedings, M innesotaInterp..ational H ydraulics C onvention, 1953,1 .J niv ers ity o r M in . r es ota , ML '1 Ile ap olis , M in n.~-~.,~-~ ,

showed that only a relatively small amountof air entered the flow regardless of how theair-entraining devices were arranged. How-ever, it was known that air flow in small hy-draulic modeL' is uncertain and that a greaterpercentage of air can be expected to enter asimHar prototype structure. The amount ofincrease to be expected in the protot:Y'pe isnot known and could not be computed sincethe fa.ctors governing the entrainment of aira~e not known. After tests on many differentmodel arrangements, it was finally decidedto construct the prototype airvents shown i...'1figure 5 and to provide measuring facilitiesin the prototype structure so that air quantitydeterminations could be made. Figure (;shows the vertical bend discharging 3,750cis with air introduced by the air deflectorsentrained i...'1the flow. To the unaided eYEthe air flow appeared continuous but in the]

/15. ODD-second exposure photograph the airis seen to enter in gusts. This was moreclearly illustrated L l the extremely slow mo-tion pictures (3,000 frames per second) madeof this condition.

FIGURE 10--Eydrcg:raphs of the fle-cc. OL 1:1:6 Heart River at H ea.rt 3'L1t-r.e DaIrlill Ap:ril 195C.

10


20/71

iI~I ;:I~ v -CZ;i g . r0 _fOB:I4~/lQXJ:L S 5 O fiO' ~IlL--~ nTITl'.-.1

1

u ir--:. j ' '('_., It Lv, ,~1i , ', '. ,

\) V l'~)C:~ 2 U 6 ([, tl. ~' 1 011n ..:', I [ : i ?O~,2 [j '~ I1 . I .. ~v,~. ,'d~, i. ~ ,)J4SI:' i}J ~~?_~~ 20 3

I L-iE1 - : : g'1I

303,200

2 27;500 ~~Li...;

165,500 LL i- - , A S:?OQ is Ii 6,000 - it-77,2CC G


21/71

Model-Prototype Comparison TestsIt w as recognized that m odel-prototype

comparison data pertaining to the spillwaydischarge and the air demand would be par-ticularly valuable and that comparisons ofthe erosion in the excavated channel, waveheights below the stilling basin, and profilesbelow the stilling basin would also be of in-terest. In the course of recording these data,other comparisons were made which includedobservations on vortex formation above thespillway and a comparison of the computedand actual tail-water curves in the excavatedchannel and in the river. Water-surface pro-files in the stilling basin and data on the rip-rap protection were also obtained.

1. Spillway capad :y. --During the 1950run-off, readings were taken each morningand afternoon on the headwater gage locatedin the gate operating house. These are shownplotted in figure 11. Using the discharge-capacity curve obtained from the model testson the morning-glory spillway, figure 12, anoutflow hydrograph was prepared, see figure13. On April 17, 19, 25, and May 1 the UnitedStates Geological Survey made stream gagemeasurements in the river downstream fromthe stilling basin to determine the dischargeof the spillway. During these measurements,the irrigation outlet works was closed. The

H+ri-1 III: I

10i C u r v e ~~~a~~edrom ::m I

I,,:

I'

tests on a 1:21.5 s c a l e : ~ ~, ' II'

hydraulic m odel. , ,~ 'I: 21001 ' '~111

12 I I II'I

i ~ 91 II, m,' ,T -~ Ii i~'I i i L--1~ I ,I,~ 8oL-L-l_

,I

i I :: q---,~,

iI' I ~ I I , ii' i ' i~Ii I

61i

3 1J

2DiSCH ARGE - '~HOU SAND S O F CFS.

F IG UR E 12-S pill-w -ay d is:::h e.rg e cap -acitycurve for Heart Butte Dam.

38 pt' t 'No V a r l o Io nI 4,6%2 i I3 1.74 27.4

32,' '

, I;' ,'1\Ii. i

I3oi-H-W i,

h-~ ' , -i , i,~W -l,1 ~,

,

II i,

iI b i I I .' I '\ I . ,i'I'; 2c U . I i, i i --1- ', ~--1L L~ULtJ ~1 ; .' I..IL) I I ' . . i: :tt1f+t+l. .I ~ : : H ; ; i ;, ,I~ i8 +t-ii-i- ,I I' i i

.


22/71

A--On April 16, 1950, the reservoir b..a.disen to the spillway crest,elev~tion 2064.5. The upstream face of the dam is at the right.

B--Ice that had formed from leaks around the outlet works control gatewas removed before the spillway operated.FIGURE 14--Performance of spillway and outlet works at Heart Butte D&'Il.

13


23/71

graph of figure 13. In general. however. theagreem ent between m odel and prototype dis-charges is co nsidered excellent. particu larlyat the higher discharges, and it i s beli evedthat the rating curve obtained from the modelw ill adequately serve to determ ine dischargevalues through the prototype m orning-gloryspillway.

2. Spillway performance--Free aIldsubmerged discharges. --During the modeltests it was noticed that, for certain arrange-mentsof the structure. the transition fromfree to submerged flow and vice versa wasaccom panied by violent surging in the verticalshaft. In some cases the unstable flow condi-hon existed over several feet of change inreservoir elevation. A mushroom-shapedcolumn of water rose and fell in the shaft.cau sin g ex cessiv e sp lash in g a nd tu rb ulen ce.In addition to giving poor hydraulic condi-tions it was feared that the prototype struc-hIre w ou ld be sub jected to objection able forcesand vibration. Consequently, the structurerecommended for field construction was de-veloped by tests to provide a minimum tran-sition range, 1. e., less than 0.2 foot. Therating curve determined by model tests, seefigure 12, indicates the definite change from

one type of flow to the other. It was for thisreason that the prototype spillway was closelyobserved when the headwater reached thetransition range.On April 16, 1950. the reservoir had

risen to the spillway crest elevation 2064.5.Ice covered most of the reservoir area, butthere was some open water close to the spill-way, see figure 14A. Before flow startedover the spillway, the tunnel was inspectedand ice which had formed around the outletworks gate was removed, see figure 14B. ByApril 17, 1950, the reservoir had risen suf-ficiently to submerge the spillway and pro-vide Ii head of 9.3 feet on the crest, corre-sponding to a discharge of 3, 250 cfs, seefigure 15. Sometime during the night thereservoir elevation had passed through thecritical region where the flow changes fromfree to submerged. Some ice had been dis-charged through the spillway, but it h ad c au se dno apparent difficulty. On April 18, the pierswere completely covered and the reservoirwas covered with ice, see figure 16A, whichappeared to be about 12 inches thick. A smallamount of trash had collected over the spill-way and slight movement of the trash was theonly evidence that the spillway existed. The

FIGURE 15--0n April 17 the reservoir had risen to elevation 2073.8, submerg-ing the spiliway crest to a depth of 9.3 feet at a discharge of 3,250 cfs.

14


24/71

A On April 18 the reservoir elevation w as 20 80.2, the head on thecrest 15.7 feet, piers were submerged 1.7 feet, and the discp~rgewas ,3,650 cfs. Arrow indicates the spillw ay loc atio n.

B BW April 21 the reservoir had receded slightly, but water stillstood a foot above the tops of the piers. Spilhm.y at arrow.FIGURE 16--0peration of Heart Butte spillway with piers suh'Ilerged.

1


25/71

FIG UR E l7--R eservoir elevation ha d been reduced to 2074.0 April 26.a.ny disturbance in the morni.TJ.g glory. Discr..arge 'a s 3,260 cfs.Ice hid

F IG UR E 18 --S pillw ay discharg e ''as 3,090 cfs April 28, and reservoir elevationwas 2071, about 0.7 foot above the point at which flow changes from submergedto free. The musr.ro 'omll surged no more tr..an a few i.'1che s.

16


26/71

.II

FIGURE 19--W ith the reservoir elevation at 2068.6 on April 30, the spillwaywas discharging freely at a rate of 1,600 cfs.reservoir continued to rise throughout A pril19, but on April 20 it started to recede. OnApril 21 the reservoir was still a foot or soabove the piers. see figure 16B. The ice wasbreaking up fast and the wind was shifting itaround the spillway area. Regardless ofwhether ice or water was over the spillwayentrance. the operation was satisfactory.with no evidence of serious vortex actiOn.

On April 26. with the reservoir at ele-vation 2074 and the piers again visible. op-eration was also satisfactorv. see figure 17.The reservoir was down to elevation2071 on April 28, or about 0.7 foot above thepoint w here the flow changes from subm ergedto free discharge. see figures ~8 and 12. Theph oto grap h. ind ic ate s the m ild c on ditio n in sid e

the spillway. There was no pulsation orrising and falling of the m ushroom .On April 29. the reservoir had fallen toO. 8 foot below the critical subm ergence pointand although the mushroom was lower inthe shaft. it was still stable with no risingand falling evident. A gain.the flow had passedthrough the critical range during the nightwhen photographs and observations were Un-

practicable. Indications are that the proto-type submerged at the headwater elevationshown by the break on the curve of figure 12and that the change occurred abruptly as in-dicated on the model curve.The spillway was discharging freely onA pril 30. (1.600 cfs) w ith reservoir elevation2068.6. see figure 19. No'spray emergedfrom the glory hole at this or lower heads ashas occurred on some other glory hole spill-ways.Throughout the flow range. no vibrationwas noticeable in the structure. Several ex-cursionsw ere m ade dow n into theoutlet w orksgate access well while the spillway was op-erating. Efforts were made to detect vibra-tion in the structure by feeling the various

parts of the structure. but no vibration couldbe detected. Also. from a location on the topof the dam or from the reservoir ba.nks. nonoise could be detected. The outlet worksgate was opened and closed on April 21, 1950.No noise or vibration was evident during thisoperation.One year later the spillway again wentinto operation, reaching a m ~Yim um reser-

17


27/71

FIGURE 2D--On March 27, 1951, the reservoir stood at elevation 2070.0, about0 .2 f oo t below the sulEJergencf.> point. Ice thickness oi' 36 L lches in Feb-ruary had caused no operatL '1.g dii~ficclti6S~ Discharge wo-s 2,650 cfs.

IB


28/71

voir elevation of 2075, or about 7 feet lessthan occurred in 1950. On March 27, 1951,the reservoir was at elevation 2070, see fig-ure 20, about 0.2 foot below the submergencepoint. Again the operation was satisfactorywith no visible difficulty despite t.'1e fact thaton February 15, 1951, the ice in the reser-voir was 36 inches thIck.

3. Spillway air demand. --M easurementswere mad e in the model to determine thequantity of air being entrained by the spill-way discharge as it passed over the air-en-training deflectors located on the spillwayface just below the spillway crest, see figure5. A ir-flow m easurem ents in the model weremade using a 3/B-incb-diameter sharp-edgedorifice connected to a differential water ma-nometer. All air entering the model passedthrough the orifice before entering the ventingsystem. Since the differential was extremelysmallfor the air quantity flowing inthe model,a specially constructed gage was used whichmultiplied the actual differential so that moreconsistent rea,dings could be obtained through-out a series of tests. The gage was calibrated

to provide reasonably accurate air measure-ments, but COlIsistency was considered moreimportant than absolute accuracy.

At the time of prototype construction,pipe was extremely difficult to obtain on shortnotice. Since the model tests continuedthroughout most of the construction period,only a small amount of pipe and special fit-tings could be provided for measuring sta-tions in the prototype. Thus, the data ob-tained from the prototype are not sufficientto determine pressures in various parts ofthe venting system, but do indicate the quan-tHy of air flow in the prototype for varioussp illw ay disch arg es.

The air quantity flowing in the prototypevents was determined by measuring the airvelocity with an anemometer hand-held in the18-inch-diameter air vent pipe. Air velocitydeterminations were made in one of the ver-tical pipes contained in the wall of the gateoperating housE and in one of the horizontalpipes just upstream from the point where itemerges into the tunnel junction section. see

~.

...

..

.

'

,.

' .. .. ~~--- .~~ ~,

:.~..'-'-' SlOE' Sect,on A-I\.~ - ~---D:E..=-._,. P LAN' o p e n t o a tm o sp he reT o p o f D q m a t t o p o f g a te h o u s e

A ir v e io ci : y r n e o su re n 'p n T Sm o d e ih ru t/:Jn 51 1 p i p eo p e n i n g s -

1',>'---:~__,~-J2-6 P i p e s

1 1 1 1 ' lco~:' ;; ,:'''l< / ' ',. . '-,,~O~ --'--

iRPI .I: < 1 ',11,,- 0t~ I~ :i'F==H i:

SEe. A-A

-'---'---.---SECTION

G J ;a s h a f T

A -r ' v 'e J o c i ty m e a ~ u r e m e n t sm c d e th r u th i sc a la p ip e . .Dcr ssu rc m 'eas .,rp mp n+s I' ,r.; : ,:. 'WI, ~,' . ,, - I .m o d e in th iS r p i p e

I

I~~::ii' - ~ 8' -. .~~J 8 SECTION 8-8'. - V e n t s oper. t o T U n n c j; u n C + : Q ~ : :)2CTI(r,S ee C ';g ure ;.FIGLlRE 21--Spi1.1way air ver:~ s ys tem a t H ea rt B ut te Dam.

19


29/71

a :wI-< :[3 :u ..0

:: : - 0 , f t -, I ,I~ ~i ~ l i tI ~ -02H-iI C l. I I

0

0~ - ji:iiiii P R E S S U R E IN VENT PiPES V S : D IS C H A R G E~. I.. I . ~ - 3 ~ tS': : I I< :[ 3 0 0 , ,. ' I I 'c I I I I i, ' I I t , . ,r J ) I O A - 5 0 ' . . i . [ , . ,~ I I O w - 2 0 . Y o ' i .: . I . 2 0 0 1 i 1 II ' -r---'-:0 I : i : i : z I ii ' i i I i I i. < : [

~ '. .

, - - ; - - - ' - r -- I ~ ) : : i i : i ~ i: ,, I1 00 - - - 0 '- ' I i .I ' '~ ' 5 5 0 J - ~ - i i ' I< :: t O ,w ' JO , 'I . i II Q A i ,;--0-=7.7%

OI Y INI.

ii -

if-I

~5 0 0 0

M O D E L A I R D E M A N D2 0 0 0 3 0 0 0

DiSCHARGE -WATER -CFS-Ow010 1000

:-;; .:.:-.FIGURE 22-Air d em an d c urv es f or m od el a nd p ro to ty pe f or v ary in g d isc ha rg es.

figure 21. Concurrent ..lth the air-velocitym easurem ents, pressure m easurem ents w eremade on the other horizontal air vent using aU-tube contaiIUng water for an indicator. Thepressure-measuring station is also shown infigure 21.Air flow in the prototype was not sm ooth,as evidenced by the so1.md of the air flow andfrom the difficulty experienced in holding theanemometer steady. There was chance forconsiderable error in anyone anemometerm easurem ent and so several determ inationswere made for each flow in both the verticaland horizontal vent pipes. Readings weretaken until the observer was satisfied that atrue average had been obtained; a consistent

set of readings over a period of 10 or moreminutes was obtained. The anemometer re-corded lineal feet of flow w hich. w hen dividedby the elapsed time, gave the air velocity infeet per second. Each observation lastedabout 2 minutes so that the average velocityof air flow was that occurring for a testingtime of 6 to 12 minutes. Pressures meas-ured in the U-tube also indicated that the airflow was not steady. Differentials varied

from plus to m inus, .but an average pres-sure reading was easier to obtain than wasthe air velocity.The results of the air quantity and pres-sure determinations are plotted on figure 22.The percentage of air entrained in the spill-w ay discharge, for both m odel and prototype,show ed a decrease as thedischarge increased.In this respect the model predicted the per-formance of the prototype. The prototype,however, entrained roughly four times asmuch air as was predicted by the model. Inthis respect. also. the prototype performedas anticipated except that accurate predictionsof how much more air the prototype wouldentrain could not be made from the model

'tests. W here the model showed air entrain-ment of 5.5 percent of the water dischargefor 1,000 second feet of spillw ay discharge,the prototype showed 20.5 percent. For 3, 600second feet the model showed 1. 9 percent andthe prototype 7.7 percent.The points from which the curves of fig-ure 22 were drawn are also shown in the fig-

ure. The prototype air demand curve was not

20


30/71

drawn through the points for 2, 500 secondfeet because the pressure values, which wereconsidered more reliable, indicated that thecurve should be drawn below the velocitypoints. Moreover, the shape of the curvewas then similar to the model curve whichwas based on very consistent data. To fur-ther prove the validity of the shape and valuesof the prototype air demand curve, computa-tions of air flow were made using the meas-ureel pressm-es, assuming that both vent pipescarried equal quantities of air and using theusual losses for bends, friction, inlet, etc.The computed values were found to be in fairagreement with those shown in figure 22.

4. Stilling basin performance. --Theperform ance of the stilling basin w as satis-factory in every respect and, furtherm ore,it performed according to the predictionsmade from the.model tests. A general viewof the basin and surrounding area is shownin figure 23.

Water leaving the tunnel appeared to befully aerated and at the approximate depthindicated in the model studies, see figure24A. The entire basin contained extremelyturbulent water, see figure 24B, and was

long enough to obtain the full jum p height be-fore the flow entered the excavated channel,see figures 25A and 25B. At times consid-erable amounts of spray shot into the air, ata point where the outflow from the tunnelplunged beneath the tail water but most of thespray fell back into the basin. The smal1amount of spray which fell adjacent to thebasin caused no difficulty. Much of the timethe flow entered the basin smoothly as shownin figure 25A. Flow leaving the basin had arelatively quiet w ater surface w ith few m eas-m-eable waves, figure 25B. There were longperiod swells, however, with a maximumheight of 12 to 18 inches which were causedby pulsation in the stilling basin. The dis-turbances below the stilling basin were sim i-1ar to those noted during the model tests.

W ater-surface profiles were m easured inthe prototype for discharges of 3, 700, 3, 300,2,350, and 1,050 cfs. These are shown :infigures 26 and 27, along with the profile ob-tained during the model tests for 5,600 cfs.Although no exact comparisons can be made,the prototype profiles seem to be in goodagreement with the model profile. If differ-ences do exist, they are probably due to thegreater air entrainment in the prototype.

F IG UR E 2 3- -0 ut fl ow at 3,250 cfs, about 58 percent of capacity.basin effectively dissipates the energy of the discharge.

The stilli 1g

21


31/71

A--Flow G118 basin wH,S .\f f:1...i, s.(;':~'ated an d 'We ll distributedacross the b&si:: ';,' ~L.'1tL

B--Profile of the hj liraulict he b as in 1.Jas &lGotL and along the wall. F lo i'[ le av in gFIGlJRE 2 4- Pe rf or ma nc e o f t he s ti ll hg basin at a discharge of 3,600 cfs.

22


32/71

A--Dividingwalls, submerg-ed here, wereeffective inspread ing flowt o e nt ir ebasin w:Wth.

B--Flow leaving the basin r..adwells and boils, but no chopPY;''aves.FIG URE 25--Perform aIlce of the still:L.'1g basin at Heart Butte Dam in dissipatingenergy o f d isc ha rg e at a flow of .3,600 cf:J,


33/71

EI.2017 ,Model- S600 c fs .------- ,II1:'0 .>'.9:

--=

A ril 2 -3700 cfs.~,I/-

TM ay I-IO SOcfs.-EI.1987.50 .. ' .'\). . ,:::~_.' : .0..

i jI 75 ' 0 , , It ~ - - ,\-Sta.12 t 49.5 Sta.13 t 99.5-/FIGu'F.E 26--Profiles of comparative discharges of model and prototype spillway.

:..- '...: . :C>.': . 'FC'.I,IIII '1'I II ,I II I, II -, I II I I II IIf ' - - - - - - - - - - {5 -0 - - - - - - -- - - u>K-- - - - - - - - - - -75 -0 -- - - - -- - n --~'~-Sta.f2t49.5 Sta.13+99.5~/

1his w ould m ake the prototype profiles slightlyhigher than those in the model for the samedischarge.5. Erosion downstream from stillingbasin. --Erosion tests in the m odel indicatedtiiat the channel banks just downstream fromthe stilling basin w ould be subjected to greatererosion forces than the channel bottom andthat rock riprap would be necessary in theprototype to prevent bank dam age. The chan-nel bottom was shown by the model tests tobe relatively free from erosion tendenciesand no dam aging erosion was expected there.As a precaution, however, because of thefine-textured friable m aterial com posing thechannel, rock riprap was used in the proto-

E I . 2 0 1 7 ,,-Mod el-5 60 0 cfs.---

type channel bottom . No riprap was used inthe model tests.

Before the run-off in the spring of 1950,cross sections had been taken in the proto-type channel on May 31, 1949. Following the1950 run-off, cross sections were again takenon June 15, 1950. Cross sections for bothdates at a station located just downstreamfrom the eIrl sill and at a station 50 feet down-stream from the sill are shown in figure 28.These typical sections show the maximumerosion depth to be less than 12 inches. Closeto the end sill there is no significant erosion.Using all the cross sections taken, see fig-ures 28 and 29, calculations indicated thatless than 20 cubic yards of material was

, ~ A p r i I 26 - 3300 cfs.

A p r i I 29 - 2350 c f s ~ ' - = t - =IIIIEI.1987.5 ~ . . . . -. . -/'. .:c):.:

FIGURE 27--Profiles of comparative discharges of model and prototype spillway.

24


34/71

~2000f-c: t>w~ 1995

20'I

II St~. 14 + 50la '

----

2000

,,iI,II

St~. 13+ 99.5At end s i l ii

z0f-~ 1995w-IW

1990

O r i g i n a IA f te r F lo o d

I

I

I I'I ,1 'II -- - OriginalI I/hl:/. A fter F l o o d


35/71

A--SweJ..ls eroded fine rns.t61'i::l f:r'om tx}hin:j the coarsel.eft bank of the excav5~ted cr..a.nnel below the stillircg baEd..D .. t.X16

~~'

B--Erosion e...LS (\ ~C C1Z'I'e(: a..L Cr::[ t,ne X 'lght, D sn.L neaT .the j.:.:rlcc.br:. of thenatural river Cnei.::Ule..L s.f.\d the 8X2.s.vE.tec. cbunn?} helo ,.;- th~~~ b&SL .L.

F IG tlHE 3G --Er osw: :: . C -.> . th e e xc a7 Ets ;d c tJ Rn .r 1.8 ..:- c ..c\.tn s 'G :r ea ID . i ro ll. th'2 basin.

26


36/71

eroded and removed from the ch:umel bott0mduring the entire run-off.The channel banks, despite their riprap

cover, were eroded to a grenter degree. Theriprap, however, had been placed in a thinlayer, was not weE graded as to size, andin places the earth banks could be seen be-tween the individual rocks. Swells wereobserved to rise over local areas and pene-trate very easily into the large voids. VVllenthe water receded, some of the earth wasremoved from behind the riprap. This wasevidenced by the darker, earth-colored waterwhich could be seen adjacent to the riprap.After several days of spillway operation, theriprap had slumped and the earth banks hadcaved as shown in figure 30. In spite of theapparent dam age tothebanks the riprap stillcontinued to provide a good mea.sure of pro-tection against further cutting.

The bank damage was caused primarilyby swells originating from surges in the hy-draulic jump, and not bv waves of the ordir..arvvariety, since these ~ere only OG-:': 5000 6000 1: :r -: ; C ~ , .\ =< CE .' SEC. - r:-:-F IG1JRE3 1--C ompa .riso n o f m ea .s ure da nd c ompute d tail 'W-aterelevations.

200 feet downstream from the axis of the damin the Heart River.During the prototype operation it wasreadily apparent that the tail-water elevationin the river was considerably lower than tha.tin the excavated channel. Water entering theriver from the channel had a steep surface

slope and a much higher velocity than antici-pated, see figure 32A. Observations, how-ever. were not sufficient to establish whetherthe tail water in the channel was too high orthat in the river too low . Levels were runto determine the tail-water elevation at fourseparated points for five different discharges.The location of these points, together withthe tail-water elevation and discharge, isshown in. figure 31. These values are plottedbelow the tail-water curve used in the modeltests. These data show the computed tail-water curve to be 2.3 feet higher at 1,000s ec on d :fe et and. 4.1 feet higher at 3,600 sec-ond feet than the actual measured points in theHeart River. Tail-water elevations meas-ured in the excavated channel more near Ivcoincided with the computed curve, but;t3,600 second feet the elevation at Point Cfigure 31. taken in a quiet area adjacent t~the wing wall at the end of the apron, was 1foot below the computed curve. Elevationsaotained from water-surface profiles takenalong the basin center line agreed with thecomputed curve, but only because they in-

, .


37/71

A--The 4-foot difference in elevation between the flow in \,~.le excavatedand natural char.w. 1.elsat 3,700 CIS tlcceleratcd the d':: 8~.:tlL: 'g(;t

E--Tb.e higb-velocit~J discharge into the natural c.canne.l se t up a. back:eddy that. damaged the river bank upstream from the point of' impact.FIGlmE 32--F low conditions at the j~~ction of the excavated and natural channels.

28


38/71

FIGURE 33--loss of ban.i{:material through erosion during the 1950 flood is readilyapparent. The eddy sho'W l1 in figure 32-3 caused this damage.cluded the boil height at the erld of the apron,which was slightly higher than the adjacenttail w ater.

The model stilling basin was tested. todetermine the permissible reduction in tailwater before the jump would be swept off theapron. In the model it was possible to lowerthe tail water only 3 to 4 feet before the jumpwas swept out for the maximum discharge of5,600 second feet. S ince the tail-water ele-vation in the Heart River is 4.1 feet lower,at a discharge of 3,600 second feet, than thecomputed tail wq.ter, it is imperative that aclose watch be kept on the excavated. channelto prevent damage which might lower the tailwater to the level of the Heart River. If thisshould happen the jum p w ill undoubtedly sw eepout: and the apron will operate as a flip bucket.Since the structure is not designed for thistype of operation, damage could result.

7. River erosion. --The difference inwater levels between the excavated channeland the river was the cause of the high-ve-locity flow entering the Heart River. Waterleaving the stilling basin was of relativelylow velocity and would not have caused illeffects as it entered the river, see figure25B. The 4-foot difference in elevation, how-ever, caused an increase in velocity which

pro ved. to be sufficient to cau se considerabledamage to the unprotected riverbank down-stream , see figure 32B. Some of the damagewas caused by the direct effects of the cur-rent flowing diagonally across the river andcutting into the far or left bank. A greatshare of the damage, however, was caused.by a large induced eddy in the river, seefigure 32B. This eddy caused. an upstreamcurrent along the left bank which removedlarge volumes of material from areas con-siderably upstream from the point wherethe main flow impinged on the bank. Althoughthe damage was considerable in extent, ithad no ill effects on the structure or its op-eration. Riprap placed in the eroded areawill prove of value, however, since the dam-age would become greater with each succes-sive run-off and the end result would be dif-ficult to predict. The bank damage is illus-trated in figure 33. A comparison of figures23 and 34 shows the extent of the bank dam-age which occurred between the start of therun-off and May 5, 1950.

8. Inspection of structure following1950 and 1951 floods. --An inspection of the;spillway conduit was made following the 1950flood and again following the 1951 flood. Cer-tain findings are of interest and are discussedherein.29


39/71

The conduit inspection, in both instances,revealed that the concrete was in excellentcondition with the exception of four smalleroded areas located in the 900 bend. Fol-lowing the 1950 flood, plaster casts weremade of the two most prom inent areas. Theseare shown about full size in figures 35 and35. The largest area is about the size of amanls haIld a..'1d by actual measurement hasa maximum depth of erosion of 3; 4 inch. Thesmaller area shows a maximum depth of1/2 inch. The surfaces shown in., figures35 and 36 were molded in sponge rubber a-gainst the plaster casts made in the fieldand are therefore an exact replica of thetunnel surface followL'1.g the 1950 flood.

These areas are located near the invertand near the bottom of the 900 bend. Con-struction timbers or ice falling into the shaftcould have caused the surface damage shown.Persons who have viewed th.e rubber castshave been of the unanimous opinion that thedamage was not started by cavitation.

Following the 1951 flood, these areaswere again noted. There did not appear tobe any marked change in these areas as aresult of the 1951 spring floods. No repairswere believ~d necessary.

Inspection of the riprap downstreamfrom the stilling basin, following the 1950flood, indicated that repairs would be ad-visable. The slumped riprap in the channelimmediately downstream from the stillingbasin structure was repaired in May andJune 1950. Gravel backfill was placed onthe slopes to bring them to grade and rockreplaced. over the graveL The eroded river-bank just downstream from the end of theriprapped channel was sloped and coveredwith gravel and rock.SHADEHILL DAM STUDIESDescription of the Project

Shadehill Dam is a part of the MissouriBasin Project and is located near Lemmon.South Dakota, on the Grand River a tributaryof the Missouri. Figure 1 is ~map of th~general area. The dam was constructed toprovide irrigation and flood control benefitsand is sim ilar in many ways to Heart ButteDarn.

Shadehill Darn is a compacted earth-fillstructure, rising 125 feet above the streambed to elevation 2318. At normal water sur-face elevation the reservoir contains about

FIGURE 34-Extent of dow-nstream erosion can be seen by comparing this view afterthe flood had receded withfigure 23, at the height of the flood flow.'1 0


40/71

FIGURE 35--Eroded areas in the 90 bend subsequent to the 1950 flood were fourin number. Tnis, the deepe st, is shown by a cast made in the bend from theactual depression. Erosion was 3/4 inch deep. Tn e illustration is full size.

31


41/71

FIGu73 36--Another of the eroded areas is also shown full size. Jepth ofe ro si on h er e 'a s1/2 ir.:ch. The 1951 flood did not extend the size ordepth of the eroded areas.

130,000 acre -feetof water and at m axim umwater surface elevation it contains 480, 000acre-feet. The drainage area is 3,070 squarem iles. The service -spillway is a morn:ing-glory type which discharges into a horizontaltunnel under the dam as shown in figure 37.The emergency spillway on the left abutmentis a saddle 1, 500 feet wide excavated. to ele-vation 2297, which is 21 feet below the topof the dam. .An expendable earth dike, 5 feethigh, was constructed across the saddle toprevent mmecessary small flows through theemergency spillway. It was expected. that alarge flow -would overtop the dike and washit out.

An outlet worKs for release of irrigationwater is located. near the left riverbank. Sincethe outlet works has no physical connectionwith the spillway and was not in operationduring t..>te prototyPe tests, no further men-tion of it will he made in this report.

As at Heart Butte, the morning-gloryspillway is designed to operate throughoutthe range of free discharge, the transitionrange between free and submerged discharge,and for submerged flows as great as 40 feetof water above the crest. Likewise the crestis uncontrolled but has six equally spacedpiers placed radially in plan to reduce vor-tex action (figure 38). The spillway has anoutside diameter of 33 feet, 8 lnches. Itscrest elevation is 2272. 00. Its capacity is5,000 second-feet at normal reservoir ele-vation 2297.0, the elevation of the emergencyspillway crest. At maximum reservoir ele-vation the spillway capacity is about 5, 700second -feet.

The spillway crest profile merges intoa short vertical shaft with a small deflectorat its base, then into a 90 vertical bend. Thebend leads into a horizontal tunnel 13 feet 6inches in diameter and 520 feet long, which


42/71

~::' (:~=::ooo,, ~~ ~~ up ,' C re st E I 1 30 2 ,0 -- - - -- I ~ , ; ; ~ = - : ~ ~ - = ~ ~ = : = = = - = - ~ - : = ~ = - ~ ~ ~ = ~ ~ ~ : - - ~ , , ~ ~ - : r~ A X Is o f d om ( Cr es t 1. 23 1B0) 2 ~ - ; } f[, / / ', i t , ~ ~ I f . Outlet w o r k sA'

,

,

S

,

f,,,.; ' ,,' ~,

',

'ir,

1'ar n ---9 ~ i '.,-- p ..i S e n f ' : ~~ sp J IWJY

~I:f 'i '

:, 1,

'10.,- lJ. l A . if \ ~:,:~:>'-


43/71

/-~/,c~\\ i//~ '\\' \~ ~ ~/l p . - j~''1--~II le, / / ';'._- 'j ,./'- ~~If / / 1./ I (~')f( .' / / // ;/1 /1 / / ./ ~ -t k'~--, ~' / / v' ii_.w ': :oo~-_ , ~~-- r\ -\~ \ \~~'\ '\\ \\;.t='~\ llm\'J~< : ,II.J.~~~

0.--'. 8

:__H ',..

;:~ ; 22 :.- .tC1-- - - ~ ? S C . CO =~\ . - -i . ~-_.._---

,~I,.

,., .I...:.. .:1 '- - - >. :

\'--'-' .

,~~~ \~j~1,< ) '\, . ',. c:ic ;\7


44/71

terminates in a hydraulic-jump-ty-pe stillingbasin. An excavated channel about 1,400feet long conducts dischc lI'f es from . the sti .l-ing basin to the Grand River, Figures 37and 38 show the general feature,; of the spiJ.l-way.

The stillin;:; basin contaIns C::t row eachof chute blocks and baffle piers and an endsill (figure 39). Spreader wal , were notused in the basin. The maxim 1ZC, vertiealfall from headwater to stilling ;)35ix\ floor isabout 130 feet, which is the sa..'7le as at HeartButte.

The spillway approach. spillway, tUIll . el,stilling basin, and a portion of the dcwr:streamchannel were tested and dcveloped ,1Sing a1:20.73 scale model. see figure 40.Su.-rnmary of M odel Tests

1. The model. --The Sha.dehiHmodel tests were made soon after th.e H cartButte tests and since the .struct1,;res weresim ilar in many respectL, TIear: Butte.'nodel was m odified for the Sha.derJll tes s.The Shadehill studies were of the samegeneral nature as the Heart Butte studies,and generally sim ilar recom rnendertion.s were

made. The several important differencesthat do exist between the structures are dis-cussed below.

2. Spillway air tests. --The air-entrain-ing device used on the Heart Butte spillwaywas not used on the Shadehill spillway. Onthe Shadehill spillway, however, an la-inchair vent was placed near the base of the de-flector in the vertical shaft (figure 38), tosupply air to the upstream end of the hori-zontal tunnel. The vent was located near thecrown of the tunnel and supplied air to thespace between the surface of the wa.ter andthe tunnel crown; t..'1us no air was intr{.x: .ucedinto the flowing water. The purpose of thevent was to reduce the possib: Uty of lo',' pres-sures occurring in the upstream end of thetunnel in the event that sufficient c.ir did netflow upstream from the hmne1. portal. Thisvent was not constructed in L te model sinceit was not considered ahsojuteJ~c necessaryfor satisfactory operation. TLe decision toinstall the vent in the prototYPwas rna.deafter the model tests bad br.-en completed a:.1J:iit was installed as a prec~.utiDnar:y measure.

3. Stilling basin tests. --Tests on thestilling basin were sunilar to those describedfor the Heart Butte basin except that it wasfo1lIld Lh..at spreader w alls were not essentialto good distribution in the basin. The lowerangle of divergence in the basin was probablytJ:1e facter that made the walls urmecessary.J\tlodel tests wer,,, made both with and withoutthe waUs in place and the tests showed thatsatisfactory distribution of the flow acrossthe basin. w idth was obtained without the di-vidi~,; walls. Figure 41 shows the model inoperation.

During the final tests to develop the mosteconomical stilling basin, a question wasra(seQ. as to whether the basin might be short-cned an. additior ,, ll14. feet. Tests were madeon the shorter basinand its use in the proto-type was considered. However, at that timeit was decided that the longer basin wouldprovide hetter protection. Therefore, mostof tbe final data were taken on this basin(figur


45/71

.~.,,~~ , ~ ~~__ ,~__R ' --_R_' . - '~ '~- '---,

Sto 6+02 ,1 Sto 6+34'''''''i---=:::::i===-~-- .- -+-----===r=====-~ ,---r= j~Y :: 'Origin: :' 2062', ;--2 9 '

22'. O~ X' X'/(I.75Q}2 + 10'52512 ' I

,- ,+--~eeD e t a i l C

, -


46/71

i.I

:

i

-,i

\.

I

~

\I

{,':\

I\J

.IH0

-n

:Id~c

i'

2>

:

i

-i_#

,

IF-~8D

;1-

,cmG

i

i

L,1

Ii

iiI:

,

i:1

.

l_

'1I

.~~

-~>

t

z.,

iI

iJ

-

-.. C

'


48/71

_R~~ ~'-~~-'~ '---' '-'- '-'~ '-- '-~-'- '_. 2315

M ax . W .S E I 2 31 20 0- -,l-'-'-~--'~---=-~-~-'': ' '-'-'~---:--~,

~,~..,

. , :rg en (:y S ~i'I., ...; ~.,. 'dIe E12?9

':;',~c.','

I,

I

Ic, i

r~'~,l\-I \eme

E xp en da bl ed ik e, cr cs t [ 12 30 2- -, .~~--

,/'/-~~~/

...~~

---l~j [Cl'r1

~ L - = J = ~ =~-- '- - .-----.----

,-bJIJ} -';':>9 z-'N

70.- ;.:85U..J

-. . -_C

,

j.

~':::=I

-'

.

i

.

~IT.., 1

~:-::::::1

.

i-C=-~-1--

1

--1- T ut UI-u-r . ..0000- - i ' .I l_.. -- -- --1- ~--- '90,000: f -E I 2298 MAX W 5 1952 F L O O D i,', 66% 'of f l o o d s to ;a ge t o m ax . W . o i,' 84'1. o ff lo ad s to ra ge to d i ke c r es t.S p i l l w a y ji~r:horgp.: H8~/r,of a nt ic ip at ed m ox, i . I . -- - . ..- . -. ..

'

.

--[ '

.

. , : ~ ~

23tO

? 3 0 5 . .- .. .. -

230

??':,1;~,:,~~Oj~

.J);J.Ja.:11.7 ~')O

105,000

30 I. t - --, LJ10-MAY-R'}, ~)()O

7 0, 0 00

55,000

FIG ill.E i,2-.-Reservoir elevations and hydraulic dats of the 19'52 flood at ShadehUl D a m .


49/71

J.

.

.

11-~---+

.,...l- n

34

32

~'O ._. .~...~,.,

28

26

24 ~

tI>-0

'0LL .


50/71

was 5,020 second-feet on April 10 (figure43). Extensive data were taken on the grad-ually decreasing discharge until May 7, whenthe flow was 800 second-feet. Spillway flowcontinued for several more weeks, however,before ceasing altogether. 'Ll'1e spillway was'thus in operation at appreciable dischargesfor more than a month, which provided agood opportunity to obtain data on the per-form.ance of the structure and ample tlineto evaluate the effectiveness of the stillingbasin and riprap protectiOn in the excavatedchannel.

The greatest spillway discharge duringthe flood represented 88 percent of maxL'YlillIlspillway capacity, while the highest reser-voir elevation during the flood indicated that66 percent of reservoir storage had beenutilized. For these conditi.ons the spillw aycrest was submerged 25. H feet and the topsof the piers 15.9 feet. The maximum. head-water to tail-water fall was 9.:t 9 ieet and theenergy entering the stilling hasin amountedto about 40, 000 horsepower.Model-Prototype Comparison Tests

Data on the performance of the Shadehillspillway were taken to compare model andprototype performance a.t1d to evaluate theperformance of the entir'e prototype struc-ture. Model-prototype comparison data weretaken on spillway discharge, vortex actionfor submerged flow, head losses through thestructure, water surface profiles in the still-ing basin. and erosion below the stilling basin.Other data, taken to evaluate the performanceof the structure, included wave heights, w.atersurface profiles in the excavated channel,comparisons of the predicted and actual tail-water elevations, and the effectiveness of theriprap in preventing bank erosion.

1. Spillway capacity. --Headwater ele-vations were taken daily at 9 a. m., usingthe headwater gage located in the outlet worksgate operating hou.'se. Heservoir elevationsthroughout the flood period are plotted infigure 42. An outflowi:ydrograph (figure 45)was prepared, using the 1'Lsen/Oir elevationsand the discharge-.~apacity Clirve obtainedfrom the Inodel. The 13cnargecurve is shown in 44 On the datesshown by the clrcle;;; in ;~ure 45 the UnitedStates Geological Survey made current metermeasurements in the river downstream from

the excavated charmel to get an independentcheck on spillway discharges. The circledpoiIlts show the river discharge obtainedfrom the current meter traverses; the fig-ures adjacent to the circles indicate the dil-ferences in percent betv,'een measured dis-charges and those predicted by the modeltests.

The agreement between predicted andmeasured discblrges is considered excdlent.The average variation from the model cali-bration curve is only 2. 7 percent for the 12measurements made by the United StatesGeological Survey. In eight instances, thestream gaging measuremenis were less thanthe discharge predIcted by the model. The2verage variation was 2.2 percent. In thefeur inf,tanr:,,~ wherE ' thE model predictedless water L~an was actually measured, theaverage variation waS 3.6 percent. Sinceboth plus a.nd minus variations v/ere srnallta l d since the 12 Inea:S'i:~','Tnents c:overed thedischarge uniforrn.,y from GOO to 5,000 sec-ond-feet during both the :risin,; and fallingEtages, it can be (concluded that the modelwas capable in every respect. of accuratelypredicting prototype d5.scharges.

2 .. S pillw ay p erfo rm an ce --su bmerg edan d free d ischa.r: es. --As discussed for theHeart Butte ti';'.t'~ the model was designedto p::ovide a minimillIl transition range, 0.2foot, between free and submerged flow, orvice versa. The prototyne tests indicatedthat this raJJge was not exceeded by an appre-ciahle anlount; and~ from evidence available,th e tr an si ti on in flo,,,ccurred with ou t in ci-dent. As at leaI' Butte, however, the flowtransitions occurredwHho ut e yewitn ess es.The rate of rise was so rapid during the fillingcycle that field personne: cacid not be alertedin time to make the obse:::-vatiol1. During thedraining cyele the flow ch~mged from sub-r:nerged to free du:ring the night.

At 2: 3 0 a. In. > on May:::', however. theU. S. G . S. gaging station recorder, locatedin the downstre2.J. T] river chan.'1E '1, started tochange slope, indicating a reduction in dis-chargE not attr;,}Jutahle to the uTIiformly de-creasing hpuri. W hcn the slope change started,the head water W B .3~ lt e1 evatio n 22 76 .. 9 6.8 w hichis be c1evatior :1 which l11.odd indicatedthat free discharge wou.ld begm (figure 44).Figure 46A is )l photograph of the spill-

41


51/71

2315

2310NOTE: Curve obtained fromte sts on a I: 20.732 sea Iehydraulic model.

2305

I- 2 3 0 04J4Ju. .]

I::~z0;.:22954J..JWa: :

I ;I II

}- -~~2290~ I-~ Lill I1-i

: Ii .: ~

a: : -,2285

t . I, A - Subm erged flow -,II I'L i

2275

2272 lSoillway crest EL 2272.00

I I i ~23 4 ~DiSCHARGE IN THOUSANDS OF C. F.S. 6

FIGu~ 44--Discharge capacity curve of the service spill~~y at Shadehill Dam.

42


52/71

5452

50484644424038

Cf )L.i...36UL.i...340Cf )03 2I.Ua:03 0Z:::>:r ; 2 8Z- 26w0 4.20/00 : :24:r;u2 2Cf )02 0

181614:21086

I4 4

+---

I ,-t--+ii I I~ I

II

I. IIIII

iNOTE - Solid curve based on i i Irototype headwater goge

:. -: . --I :. -rea ding s and hyd ra uiicm odel ,i idischarge capacity curve.;; Circles are discharges i +- 1m easured in river dow nstre am i-r-, ' ~.'.rom dam. I '. i. i I I I I i. I I ,i I' ,i : I J-t--It. : ' I1 I j,i i

'. II~ i i i14

--- i\A4Y-

FIGURE 45-Comparison of the discr..arges of the l '...odel a.nd prototype spillwaysat Shadehill Dam.

43


53/71

A--Submerged fIow at heaa'iater elevation 2277.2, 8.oo:.:t 0. 3 f oo t a-oovefree discharge POL'1t. About 7 :00 P. nl., jvi,ay1, 1952.

B--Discharge changed to free flow during the night. A t headwater elev-ation 2276.7, it is about 0.2 foot below free discharge point.FIGDhE 46--~Tansition from submerged to free flow at Shadehill Dam.

44


54/71

FIG UR E 47--T 'ne vortex w as continuous and did not shift laterally ' more than a fe,;feet. There was no unusual noise. Tree trunk was 18 to 20 inches in diaJneter.way taken just before dark on May 1. Theheadwater elevation was 2277. 2 or about (). 3foot above the free discharge point. The flowwas smooth with no apparent pulsation, eventhough the mushroom had started to recedein the shaft. Early the next morning anotherphotograph, figure 46B, was taken with head-water at elevation 2276.7. or about 0.2 footbelow the free discharge point. The flow wasfree with no column of water visible in theshaft. The three elevations discussed aboveare plotted on figure 44 as Points A , B, andC. and indicate that the prototype perform -ance was exactly as predicted by the model.

3. Vortex action and effect of ice.--Observations of the vortex formation duringsubmerged conditions were of interest be-cause they provided an evaluation of the effec-tiveness of the spillway piers in reducingvortex action. Model tests had shown thatpiers 14 feet high would provide the optL'l1Uffireduction in vortex action. Taller piers had

little additional effect in reducing vortexaction while shorter piers were less effectivein reducing vortex action. The appearanceof the model in operation with and withoutspillway piers was practically identical tothat illustrated in figure 4 for Heart ButteDam.

The vortex observed at Shadehill Damwas somewhat larger (estimated to be twicethe diameter) than expected for piers 14 feethigh. For structural reasons, however. theprototype piers had been reduced to 10 feet.Undoubtedly this was the reason for the largervortex observed. The vortex is shown infigure 47. Its size may be estimated by com-paring it w ith the 20-i.11.ch diameter tree trunklodged against the spillway. It is believedthat the vortex would have been several timeslarger without piers.

On April 17, reservoir elevation 2293.2,th e reservoir was covered with partially

45


55/71

A --Spillway area. covered with decaybg sheet ice about 4 L '1ches thick.:t> ornL '1gglory is to right of tilted oil drum .

B--Sheet ice was broken up IL 1d discharged through vortex, keepbg areaarou11d spillway free of ice.FIGURE 48--Effect of ice on the vortex action of the Shadehill spillway.

46


56/71

decayed sheet ice about 4: to 6 inches thick(figure 48A). In the vicinity of the spillwaythe ice appeared to be in continual verticalmotion. Ice partic~es seemed to be risingand falling and to be causing a tinkling soundwhich could be heard plainly on top of the dam.It is believed that the partially decayed icew as b eing subjected to compressive forcesby the convergi,'lg flow, causing the ice toshatter and splinter. Other action over thespillway was so mild, however. that it wasdifficult to find the spillway location fromthe appearance of the reservoir surface. OnApril 18 the reservoir was 0.2 root lowerand a fairly large area near the spillway wasfree of ice. A vortex was evident at all times,see figure 48B. The vortex was stable anddid not change its location laterally morethan a few feet during the entire day. Themodel vortex was less stable in this respect,being in lateral motion ahnost continuously.This lateral movement of the vortex in the

model may have been. caused by the greateramount of turbulence in the flow approachingthe model spillway.

Occasionally an ice sheet drifted overthe prototype vortex opening, causing thevortex action to stop momentarily. But aftera few seconds the ice sheet would disinte-grate and fairly large pieces of ice would becarried down by the vortex. The vortex re-sumed its normal action when the entire sheethad been broken up and discharged throughth e s pillw ay .

The strength of the vortex was demon-strated by the ease with wbch the large sheetsof ice were broken and discharged throughthe tunnel. T11is strength was also manifestedwhenit was found that several of the 50-gallondrums used as floats to hold the safety boomin place around the spillway had torn loosefrom their moorings and had been passed

FIGURE 49--At headwater elevations above this point, elevation 2284.9, vortexaction was continuous. The vortex is partially hidden by a pier strut.

47


57/71

FIGURE 50--At headwater elevation 2283.6, vortex action was sporadic, withsurface waves a few inches high suficient to stop the action.

through the vortex and discharged throughthe tunnel. Two of the drums had caught inthe branches of two partially submerge'd treesin the excavated channel downstream fromthe stilling basin. O ther drums could not be.accounted for and it was presumed that thesetoo had passed through the s~illway and ondownstream.

Despite its size and demonstrated force.the vortex caused no apparent ill effects on

the structure or its operation. An audiblegurgling sound accompanied the vortex ac-tion but this sound was easily drowned byordinary conversation. wind noises, or othernormal sounds. The force of the vortex didnot extend appreciably beyond the visiblelim its of the swirling water. although a smallboat passing close to the vortex would prob-.ably be endangered. Evidently if the spill-way piers had been shorter or elim inatedaltor;;ether. the vortex and consequently its

FIG UR E 51-N o vortex was observed when the water surface in the shaft was belowreservoir level. E levation here was 2 277 .8, a bo ut 1 fo ot a bo ve s uh~e rg en ce .

48


58/71

140 6Curve 1-,,,

IL l>12 0 0::5;:)uN -UJ UJ>~a:::;:) a: :u co t,100 :I:A 2310u 0u> (/)-u.: 0U a::: If)i UJ f= wf- >a co t I...J 80 33 2300 ;LL . LI.. Ua::: 0 Cu rve 2 ~zco t 0 a3: f-a co t60 ~2 I 2290~a::: I ...J- Wco t I a:::-0>40 I 2280 ~(/)I wI a:::1:::.

effects would have been larger. face. No vortex action could be detectedbelow elevation 2279. Figures 49, 50, and51 show spillway performance and vortexaction at three different headw ater elevations.

.

When ice was not covering the spillwayarea, vortex action occurred continuouslyfrom reservoir elevation 2298 to about ele-vation 2285. Between elevations 2285 and2 2 79 vortex action was sporadic and wasaffected by the turbulence in the spillwayopening and by waves on the reservoir sur-

4. Vibration. - -Throughout the flowrange no vibration could be felt in the struc-ture. While the opportunity to get close tothe spillway tunnel was less than at Heart

16 0 7

20 0 2 3 4- 52270S PILL WA Y D ISC HA RG E - THOUSANDS C.F.S. -(CURVES 1,2,83)FIGURE 52--Air demand and spillway discharge curves for & ':1adehill Dam.

49


59/71

.~

FIGD?..E 53--Trickle of water droppi.'1g straight down into flow at tunnel portalshows that air flow through the tunnel was not sctficient to deflect thetrickle. Straw dropped into the flow showed only slight ai r movement.

Butte, vibration was not perceptible in theparts that were accessible. Despite the pound-ing or the water in the stilling basin, no vi-bration of the stilling basin walls could befelt.

5. Debris. --The reservoir was remark-ably free of floating debris. The dead treevisible in the photographs of spillway opera-tion was the only large piece in evidencethroughout the flood. This tree became lodgedagainst the outside periphery of the spillwayand piers but had no noticeable effect on spill-way performance. On the morning of May 1,1952, the falling reservoir exposed a largeportion of the tree, and it apparently toppledover and was passed through the spillway.Observers watching the tree trunk float down-strea..m after passing through the tunnel esti-mated its over-all length as about 30 feet.

6. Tunnel air demand. --The quantity ofair entering the 18-inch-diameter vent locatedat the base of the deflector in the verticalshaft, figure 38, was measureCl using an ane-

mometer held in the 18-inch-diameter intakepipe located at ground level just upstreamfrom the stilling basin (figures 37 and 57A).The lineal feet of air flow was measured forsix different discharges using an anemometerwell braced in the pipe. Each measurementwas sufficiently long to establish an averagevalue~ The quantity of flow was then com-puted and plotted against water discharge(Curve 2 of figure 52). The maximum airquantity measured was 141 second-feet andoccurred for a water discharge of 2, 400 sec.ond-feet. The corresponding velocity in theair pipe was about 80 feet per second. Sinceair vents w ill work satisfactorily with veloc-ities up to about 300 feet per second, thisair demand is regarded as moderate. Thespillway rating curve is included a::; Curve 3in figure 52 to show the relationship b~twe~nair and water discharges. Curve 1 showsair flow as a percent of the water flow . To-gether, Curves 1 and 2 show that the quantityof air increases w ith increasing dischargeuntil the water flow reaches about 2,400 sec-ond-feet, then decreases rapidly until after

5 0


60/71

80

.70

~0[


61/71

the 31 T'CY 'C:z;:Lt h C~~ld lo~s ;:nuasure:c In theprutot)rpe '. ~rhest:; figli:r:e~~ indicate that themodel,. D l.L,LL: tD 3. scale 01' 1:23 5traD.Sr>arc:rtprO'totyl.iC as an den .f :I g :: ;/ j .0 SSf :S~

~5. . SU .1 lj .J lf } b as i. n p er I'o rr na nc e - -the hydra'u.lic model tests.t a basil' , was d(2-

VeIopl ..'peeted during modeltesting tl-.at tail-water elevations at the basin

..

52


62/71

oJ ... .

EXPLANATIONAverage prototype profile for 4000 c.f.s.Mode I profi Ie for 5000 c.f.s.Prototype profile (Apri I 28, 1952J for 4050 c.f.s.

2240

._m__., '--'-- . ,~- ~-- m.- -..-...Model '.',Prototype

\ \--- -

. I . . . . .-:r :->.4 n.. Portol E1.2216.39---J---t-.-+--


63/71

EXPLANATIONApril 30,1952- 3950 cf.s.May 3,1952 - 2500 cf.s.May 5,1952- 1300 c.f.s.2240

r,-; --'.'Pod(JI EI. ;'216391 .1LL2200z

,p. z0f- 21 DISTANCE FROM PORTAL IN FEETW_Il,J

2Z00

DISTANCE FROM PORTAL IN FEET

~.,-~~..~...~~,,~~ ~ n' ~'.~ '~'--, ~~.-,~_a ~~ ~.. .............-

FIGURE 56--Hater surface prof:iJ.es of Shadehill spElvm.y, 1,300 to ],950 efs.

., ... .


64/71

..

f

A--Stilli. 'lg basin, excavated ChanJlel, and the Grand Rive:: on April. 24.A ir i. 'ltake structure in the foreground.

&--Stilli. 'lg basin performance showi. 'lg turbulence at the toe of thehydraulic jump and fi. 'ls along walls of the transition. Spray fromthe jump caused wet areas adjacent to the walls.

FIGURE 5?--S:hadehill spillway stillL lg basin dischargi. 'lg 4,1.00 cfs.55


65/71

.

A--Eoil &.t end of still:L.YJ.g basi..'1 extended into excavated Channel.,making exact determine. tioD of tailv.taer elevation d ifficul t.

B--Erosionof cr.nnelbanks isolated trees &nd w~~ened the channel.Far barl..'k: was eroded by eddy cutting behind. riprap Con drl1. 'n .s caught.b: trees were part cf safety boom tr2t had passed through spillway.

?IGL11E 58--Sb.adehill spillwa'.l stilling basiIl discharg:L.~g 4,600 cfs.

S~


66/71

would be higher than the values shown on thecomputed curve. Because these tail-waterelevations were lower, rather than higherthan those anticipated, the operation of thes t illing basin was adversely affected to agreater extent than the figures above indi-cate.

This low tail-water caused the boil at the

end of the basin to be higher, the surging inthe basin to be greater, the velocity of theflow leaving the basin to be higher, and thewave action to be greater than would other-wise have occurred. Also, the velocitiesthroughout the excavated channel were higherb eCRuse of the steep water surface slopebetween the end of the basin and the riverchannel (figure 58B). Had the true tail-water

.2206

2204

Jo

:z0....> 2202lL J...JlL JQ:lL J....~ 2200J

Ii.. . ....... O i, I

.... ,0 '

2198

21960 2 3Q IN THOUSANDS OF C.F.S. 4 5

0 Prototype river T.W . /430' downstream from basinr::. Computed tailwater curve used for design andhydraulic model testsFIGURE 59--Computed and actual tail water curves for service spillway at Shadehill.

57


67/71

elevaUo115 been lm own during thE; m odel tests,the entire basin wouLd have bCE;l1 set 3 feetlower to eliIni late the excessive difference.It is fortUI1i:ite that the excavated channel

was not sufficiently wide and deep to aJlowthe tail-water at the stilling basin to approachthe elevation of the river surface, see figures55 and 56. Had this occurred the jump wouldhave been near the sweep out point and theenergy dissipating ability of the jump wouldhave been greatly reduced. Serious damageto the structure could thereby have occurred.

10. Erosion below the stilling basin. --Despite the low tail-water conditions therewas no significant bottom erosion as a re-s u lt of the flood. Apparently, the bottomvelocities after passing through the jumpwere directed to the surface by the end sill,producing a boil on the surface rather thanerosion on the channel bottom. The chan-nel bottom was protected for 100 feet down-stream from the end of the basin, however,by dumped riprap 3 feet thick. The riprapwas fairly well graded in size irom fist-sizepieces up to 24 inches in rnax:L'num dimen-Slon.

Figure 60 shows the contours of the chan-nel bottom after the flood. Although exactelevations of the bottom before the flood oc-curred are not known, it is certain that nosignificant erosion occurred. This was aspredicted by the model tests.

The :model erosion tests were made usinga movable sand bed consisting of well gradedsand (0.5 b 2.0 mm mean diameter), withno rim 'ap protection on the bank.. ' or channelbotto~, -and. also with a movable bed of sta-bilized sand. 'The stabilized sand was com-posed of a cured m .ixture ofsand and cem enthavill,S; a resistance to erosion equivalent tothe estim.ated res.istance of theprototypechan.'1el material. Both model tests showedbottom erosion abom 1 foot below apron ele-vation at the apron corners for a discharge of5,000 CIS second-feet (figure 41B). The testsal o indicated t1:Jat s evere b3.l:..Kerosion w ouldoccur un.lf~Ss riprap protection wereprovided. Bank erosion occu.rred in the pro-totype.

'TIle :flow after the boil at the endof the pro1 .:Jtype apron accelerated rapidlybecause of the differencE' in elevation between

the top of the hoil and the channel water sur-race, It was this velocity combined with thewaves generated at the end of the stillingbasL'1 which produced the erosion of the chan-nel banks.

(

The right bmk of the channel consistedof;.:, low riprapped slope more or less parallelto the basin training \valls, as shown in fig-ures 60 and 61. The two trees located about6 and 8 feet from thebank lbe in figure 61,were on dry lan.d before the outflow started.Tn e left baw\: was a higher and steeper slopeabout 15 feet righ., and was riprapped for 100f ee t d owns tre am . Near the end of the riprapthe left bank started a sweeping curve to theright as shown in figure 81.

t

During operation the riprap on the rightbank was SOOii submerged and the waves be-gan to erode the soft earth. The trees wereunderm ined aIld a new channel cut to the rightof the trees (figures 58 and 62B). The sur-face velocity a..'ld waves attacked the left bankjust downstream from the end of the riprap,causing rapid decay of the soft material forseveral hundred feet dov.-n.stream. A sizablearea was eroded at the end of the riprap,causing an eddy to form in the area. Thisaction became progressively more violentand the eroded area enlarged illltil the riprapwas attacked from behind. The effective-ness of some of the riprap was lost, sinceit slumped into a low pile, see figure 62A.It w as estim ated that t...1-teeft bank line down-stream from the riprap had receded as muchas 80 feet as a result of the combined veloc-ity, wave, and eddy action. The waves thatwere partly responsible for the left bankda.'I12.ge w ere sim ilar to those shoV\.'D.ll figure58 .

T he h an k d aI n. a .:; e in . its el f was consideredinconsequential 1.>1hat the Jam: in the imm edi-ateareawas e xp en da ble , However, the bankd am ag e ca u..'S ed c on cern becauseof the possi-bility of the tail-water being f1lrlher loweredand further aggravatL 1g operation of the still-ing basin. The extent of erosion of the banksis evident b figure 60. To repair the bankdama.ge, a ba nd of riprap 50 feet long wasplaced across the char..nel bottom and up thechannel banks. rne right bank, figure 62B,was replaced to an elevation h1gher than theorigin;::. and riprapped to prevent wideningof the channel. Riprap was also placed alongthe left bank beyond the proposed 50 -f

spillway tes hidraulic

Documents