OPTIONAL IORM 1 10 y JULY I073 EDITION /'j GSA FPMR (Al CFRI 101.11,6

r UNITED STATES GOVERNMENT

Memorandum Denver, Colorado

TO Open and Closed Conduit Systems Committee DATE: July 15, 1976

FROM : Chief, Hydraulics Branch

SUBJECT: Research into Uplift on Steep Chute Lateral Linings

In compliance with our agreement, Mr. Perry L. Johnson has undertaken and completed the subject research. The work was done in the Hydraulics Branch of the Division of General Research.

The purpose of the study was to develop sufficient data to allow the prediction of the uplift forces that result under concrete linings of steep chute laterals when flow passes over displacements in the linings. As figure 1 shows; a displacement in the lining can be in the direction of flow (we will refer to this as a horizontal offset), or into or away from the flow (referred to here as a vertical offset), or a combination of the two. Again as figure 1 shows, a portion of the flow passing over the displacement will be deflected downward into the opening. Then, depending on the geometry, part of the velocity head of this flow is lost to turbulence while the rest is converted to a stagnation or static pressure. This static pressure develops under the lining and exerts a force which attempts to lift the lining. If this pressure is strong

I, enough to overcome the lining weight, the weight of the water in the channel, and any structural strength of the lining, the lining could be lifted up and washed out. The objective of this research study was to relate uplift pressures to flow velocity, horizontal offset, and vertical offset. The findings would thus yield a means of evaluating the poten-tial uplift forces and, therefore, provide additional design data.

A two-dimensional hydraulic model (figure•2) was built to study the problem. The model was basically a 6-inch-wide, 8-foot-long flume. The last 3-foot section of the flume's floor was movable to allow estab-lishment of any desired displacement. The flume floor was made 2-1/2 inches thick near the displacement (figure 3) to represent a 2-1/2-inch-thick concrete lining. The opening at the displacement led to a watertight pocket under the flume floor. The uplift pressures observed were those that occurred within this pocket. Dynamic pressure data were taken which allowed evaluation of not only the average pressures but also instan-taneous maximum and minimum pressures. The pressure data were collected with an electronic pressure transducer and a strip chart recorder. Flow was supplied from the head box through a 6-inch by 8-inch slide gate to the flume. The head box could establish enough head to create flume flow velocities up to 16 feet per second, which were measured by a Pitot tube.

Buy U.S. Savings Bonds Regularly on the Payroll Savings Plan {DI0•I10

The model had several simplifications and limitations which affected the data and their interpretation. First, the model was two-dimen-sional and thus represented only channel centerline flow conditions for trapezoidal channels. A complete trapezoidal channel cross section was not studied. In addition, only floor displacements in a plane normal to the flow (.figure 4a) were studied. In reality, the displacement could occur in planes at an infinite number of angles' with respect to the flow (figure 4b) and the displacement would not only occur across the floor of the trapezoidal section but would also occur in the sidewalls. The implications are that the data obtained from the model do not describe the total pressure field that could occur under a trapezoidal lining, nor do they describe the effect of displacements other than those normal to the flow. What is described is the pressure developed at an element or slice of the lining. Until further information is obtained, these data could be integrated to obtain total uplift pressure. It should also be noted that although only displacements normal to the flow were studied, this condition should be the worst with respect to development of uplift pressures. In this respect, the results of the research should yield maximum possible pressures and; therefore, conservative lining designs.

Secondly, only breaks in the lining normal to the flow surface (figure 5b) were studied. As can be seen in figures 5a and c, breaks other than those normal to the flow surface have the potential to develop uplift pressures either greater than or less than those observed. If the break is at an angle less than 900 with the flow surface (figure 5a), the displacement will act as a scoop, deflecting more flow and more energy under the lining. Thus, the development of high uplift pressures could be expected. If on the other hand the break is greater than 900 (figure 5c), the opposite is true. Less flow would be intercepted and less flow energy would be converted to uplift pressure. In reality, many field breaks would be rough, con-taining breaks both greater than and less than 900. The significance of the angle of break might be a subject for future study.

The effect of the soil in which the lining is placed and the extent of the uplift pressure field under the lining were not evaluated. It is conceivable that uplift pressures could be relieved through the soil or along the soil lining interface. On the other hand, the soil might be sufficiently impermeable to prevent the uplift pressure from acting on large portions of the lining. In either case, the influence of the uplift pressure on the lining would be small. Of course, it is conceivable that conditions might result under which the full uplift pressure might develop under much of the lining. Tinder such conditions, the influence of the pressure would be significant. The effect of

2

soil on the magnitude and extent of the pressure field is, therefore, very important. In all cases, the soil would either have no effect on or would reduce the strength of the uplift pressure. Thus, use of the results of this study should yield a conservative design.

The test procedure consisted of setting a horizontal and a vertical displacement and then collecting uplift pressure data for a range of flow velocities. Average depth of flow over the displacement was measured and the uplift pressures were adjusted to eliminate the effects of static water depths. Horizontal offsets of 0.125, 0.25, 0.50, and 1.50 inches, each combined with vertical offsets of 0J 0.25, 0.75, and 1.50 inches, were studied. Some horizontal offset is required to allow an opening through which the water and its force can get beneath the lining. Data were taken for flow velocities ranging from 7.3 to 15.4 feet per second. It was understood that this range would adequately cover the possible velocities.

The strip chart pressure data were interpreted and are presented in two forms. Typical strip chart data are shown in figure 6. First, average or mean pressure values were determined from these charts. After these average values were adjusted to eliminate the effect of flow depth in the channel, they were expressed as a percentage of the average velocity head and plotted against vertical offset for various flow velocities and horizontal offsets. In a•similar manner, maximum pressure values which were greater than 95 percent of the instantaneous maximum peak pressures were obtained from the strip chart recordings. The 95 percent level was arbitrarily chosen to represent the maximum uplift pressures that can be expected. Again, these pressures were adjusted for flow depth and plotted against the vertical offset for various flow velocities and horizontal offsets. The uplift pressure was also expressed as a percentage of the velocity head.

The average uplift pressure data are shown in figures 7 through 10. The 95 percent maximum uplift pressure data are shown in figures 11 through 14. The data show similar patterns and tendencies for both sets of curves. In reviewing these curves, points that should be noted include:

1. Uplift pressures increase with reduced horizontal offsets. This situation appears to result because larger horizontal offsets allow the flow more room to circulate. When circulation cells develop in the break, a larger percentage of the flow's energy is lost to turbulence. Thus, the resulting uplift pressure is reduced. A horizontal offset of 0.125 inch was as small as could accurately be tested with the model. It is likely that horizontal offsets smaller than 0.125 inch will develop even higher percentages of the velocity head into uplift pressures.

3

2. Uplift pressures increase with an increase in vertical offset. For increases in vertical offset, the percentage of the velocity head converted to uplift pressure increases rapidly when vertical offsets are small. At larger vertical offsets, the percentage tends to become a constant. It appears that this is caused by the turbulent boundary layer which causes flow velocities, and therefore, velocity heads to be lower near the boundaries. As figure 15 shows, a typical turbulent flow velocity profile follows a logarithmic distribu-tion with the velocity going to zero at the boundary. The calculations in figure 15 show how the velocity profile curve was developed. As can be seen, velocities, and thus velocity heads, quickly increase in the flow near the boundary but tend to become constant in the flow farther away from the boundary. This pattern is similar to what was found in the curves relating uplift pressure to vertical offset. Small offsets affect only flow near the boundary which has reduced velocities and thus results in reduced uplift pressures. Larger vertical offsets affect flows with higher velocities and thus greater uplift pressures are developed. Since velocities tend to become constant at greater distances from the boundary, uplift pressures developed by larger offsets also tend to be constant. The maximum potential uplift that can be developed with a given vertical offset appears to be a function of horizontal offset and flow velocity.

3. Flow velocity over a displacement does affect the percentage of the velocity head converted to uplift pressure. The effect, however, is relatively small. Distinct patterns are not clear for the average uplift pressure data but for the 95 percent maxi-mum data, it appears that higher velocities develop a slightly smaller percentage of the velocity head into uplift pressure. It should be noted that because velocity heads are considerably larger at higher velocities, the 95 percent uplift pressure increases with velocity.

D. L. KING

Enclosures (15 figures)

Copy to: Regional Director, Boise, Idaho, Attention: Mr. Persson Regional Director, Sacramento, California, Attention: Mr. Brown Regional Director Boulder City, Nevada, Attention: Mr. Effertz Regional Director Salt Lake City, Utah, Attention: Mr. Murdock Regional Director, Amarillo, Texas, Attention: Mr. Nelson Regional Director, Billings, Montana, Attention: Mr. Worthington Regional Director, Denver, Colorado, Attention: Mr. Lidster 224 (Warden) 410 (Yocom) 1305 (Thomas) 1521 (Johns) 1532 (Zeigler)

Blind to persons on attached sheet

4

Blind to: 220 270 273 274

,--1530

5

VERTICAL OFFSET

-I

HORIZONTAL r OFFSET UPLIFT

PRESSURE REGION

FLOW PATTERN OVER DISPLACEMENT FIGURE I

y

m

HYDRAULIC MODEL

.figure 2

FLOW OVER DISPLACEMENT

figure 3

-<j 4a A

SECTION A-A

FIGURE 4 - PLANE OF LINING DISPLACEMENTS

FLOW FLOW FLOW ,

LESS THAN 90 0 SQUARE GREATER THAN 90 0

(a) (b) (c)

LINING DISPLACEMENTS FIGURE 5

0

meters of water C N

NX

=was

=MU_~ME EW ~ ~ AM=

_ON M

~.M.MM~

4 feet of water

UPLIFT PRESSURE

TYPICAL STRIP CHART DATA

figure 6

i

0 7.5 FT/S.

❑ 10.0 FT/S. FLOW 125 FT./S. VELOCITY

a 15.0 FT/S.

1.5

J n L J- D

J Q

z

HORIZONTAL OFFSET - 0.125 INCHES

10 20 40 60 80

AVERAGE UPLIFT PRESSURE (% Hv )

figure 7

Z

HORIZONTAL OFFSET = 0.25 INCHES

1"

• •

0 7.5 FT./S.

0 10.0 FT./S. FLOW

0 125 FT/S. VELOCITY

• 15.0 F T / S.

IU LU 40 60 80

AVERAGE UPLIFT PRESSURE (% Hv )

1.5

quo

ice

11YI\1L~J IY 11"tL \J t - t VC: 1 V,alV IIY VIIL~.7

o 7.5 FT./S.

O 10.0 FT./S. FLOW 0 12.5 FT./S. VELOCITY

o 15.0 FT./S

iu eu 40 60 80

AVERAGE UPLIFT PRESSURE (% Hv )

f figure 9

o 7.5 FT./S

❑ 10.0 FT/S.

A 12.5 FT./S.

0 15.0 FT/S.

FLOW VELOCITY

1.5

J Q

w

Z

rlUMILU(V I A L- vrrJGI - i.;Dv iiv~nca

[U 40 60 80

AVERAGE UPLIFT PRESSURE (% Hv )

HORIZUN IAL vr- r~)r. i

15

I.0

n L L

D

J Q

5

w

P 7.5 F T. /S.

0 10.0 FT/S. FLOW A 12.5 FT/S. VELOCITY

e 15.0 FT/S.

Iv c 1FV bV 0

95% MAXIMUM UPLIFT PRESSURE ( % OF Hv )

figure 11

HORIZONTAL OFF5ET ° 0.e5 INGHt5

L!

co w

U z 1.0

z

t~ w w Lj-

a T5 FT./S.

0 10.0 FT/a FLOW 4 12.5 FT/S. VELOCITY

Y 15.0 FT./S.

10 20, 40 60 80

95% MAXIMUM UPLIFT PRESI' URE (% OF Hv )

0 T.5 FT./S.

❑ 10.0 FT./S. FLOW A 12.5 FT./S. VELOCITY

O 15.0 FT./S.

1.5

U) w

U Z 1.0

Z

H W U) LL. LL O

HORIZONTAL, v1-1-5t I a U,~U INUMt5

10 20 40

95% MAXIMUM

60

UPLIFT PRESSURE (% OF Hv )

:•

fiF,ure 13

o 7.5 FT./S. I

❑ IQ0 FT/S FLOW 0 12.5 FT./S. VELOCITY

15.0 FT. /S.

F, . 1.5

U) W

U Z 1C

Z

h- W

LL L~ O

HORIZONTAL OFFSET a '.50 INCHES

iu Gu 40

95% MAXIMUM

IJ,

UPLIFT PRESSURE (% OF Hv)

:•

0.9

0.8

0.7

-- 0,6

4J W U-

Z

0.5 Z F- a w 0

W E- 0.4 Q 3

0.3

0.2

0.1

0 0 I 2 3 4 5 6

FLOW VELOCITY IN FEET/SECOND figure l~