Current World Environment Vol.

10(Special Issue 1), 288-293 (2015)

Compare and Evaluate Equations in

Velocity-Depth Distribution of Open Channels
Zandi Sara1 and Nemati Amir Reza2

Graduate student of Construction of Hydraulic Structures of North Tehran Branch,

1

Payame noor University, Tehran, Iran.

2
Young Researchers and Elite Club, Islamshahr Branch, Islamic Azad University, Islamshahr, Iran.

http://dx.doi.org/10.12944/CWE.10.Special-Issue1.37

(Received: November, 2014; Accepted: April, 2015)

Abstract

Awareness of depth distribution of velocity in open channels is very important in studies of

the flow of rivers and canals. Nowadays, many researchers use Logarithmic distribution equation for
measuring the velocity-depth distribution. This equation isn’t accurate estimate to flow velocity. The
research provides the main Statistical Indicators and develops Analysis of the full equations and
Logarithmic distribution equation. These equations contain power distribution, Logarithmic distribution,
modified Logarithmic distribution, wake distribution and share distribution. Each of the equations is
used Reliable experimental data’s like (Vanoni,1946), (Einstein and Chien,1955), for the accuracy
of each equation. The calculation illustrated that wake distribution is accurately estimate the speed
values both internal and external eras and recommended this equation. In the other hand, share
distribution was the low accuracy estimate and will create large errors Calculation.
.
Key words: Logarithmic velocity distribution, wake distribution, outflow in open channels.

INTRODUCTION experimental data’s too. The best relationship Will

be introduced by use the statistical analyses.
Experimental observations show that all
velocity profiles in a turbulent shear flow (like Flow The most famous relation is provided for the
in open channels and pressure) divided in to parts estimated velocity distribution, logarithm distribution.
(areas) on a straight line. The Disturbances internal Theodore van Karman used the logarithm distribution
area is affected with floor directly and external for the first time. The distribution is as follows:
area is affected by shear stress and floor indirectly.
u 1 y
The internal part divided three regions slow layer, = ln( ) + C1
u* κ y0
middle layer and overlapping layer. While the change ...(1)
in internal part to external area is gradual, the In relation. U Velocity at depth y,u/ Shear
overlapping layer is account the external part. Then rate, fixed van Karman k,(for clean water equal
the external part composes two areas overlapping ./41),y the depth is occur zero velocity.in the turn
layer and the Uplift layer (Circular movement’s part up maximum flow velocity, Logarithmic distribution
or wake).the figure 1 illustrated the Classification of is written as follows:
these areas. u y
u 1
= max + ln( ) for y ≤ d
u* u* κ d
Many studies performed the velocity ...(2)
distribution in numerically and experimentally In the relation, d is the height of maximum
and presented different relationships yet. First, speed Unix. The simple relation and also the low
the research introduces the validity relationship number of variables considered the advantages
and then, analyses the relationships and use the of distribution. Although the most of relation of
289 Sara & Reza, Curr. World Environ., Vol. 10(Special Issue 1), 288-293 (2015)

transmission of Sediment develop based on below relation is developed to estimate velocity

Logarithmic distribution. The logarithm law is valid for internal and external parts. This distribution is
to all of the internal area except near the channel famous the wake law And are as follows:
floor, but the external part would be deviation from u 1 y π y 
the experimental data’s. vernon , (Einstein and = ln( ) + 2Π sin 2 ( )
u* κ  y0 2 h 
Chien,1955), (Venoni and nomicos ,1960), (elata and ...(5)
aypen, 1961) showed the logarithm law is valid in In this relation, Ð is Coles coefficient or
total flow.in the paper, with consider the fixed amount wake coefficient estimated with uses experimental
of van karman is equal 0/4 in throughout the depth data’s that is based on (Guo and julien,2008)
of flow and analyzed the velocity distribution. opinions, because there is not valid relation to
calculate. Properties of distribution is applied only
The disadvantages of the distribution can the central part and it is not valid the flows occurs in
noted the large of errors in Speed estimation in the below layer by maximum speed. Maximum speed
external part ((2/0y/h>),in this part ,Logarithmic in the below layer occurs Channels by 5b/h< due of
distribution distracted the experimental data’s Strong shear stress near the wall and the occurrence
significantly .ever this deviation by the fixed van- of secondary flow. The case of Logarithmic law and
karman or Cannot modify add additional fixed wake law can’t be accurate estimation of speed in
to Logarithmic law .venoni claims that the Wide the near free layer in out region.
channels(5b/h>) and Central line speed channel
occurs surface water and Logarithmic law can N eve r t h e l e s s s e c o n d a r y f l ow a n d
use throughout Flow depth. But Normally, Narrow Occurrence the maximum speed in the below layer
channels (5b/h<) occurs in below the water surface water, the wake law cannot be estimate the speed
and Logarithmic distribution Will be applied only the during the three-dimensional in open channels. Add
internal area. wake function to modified Logarithmic law, proposed
the following equation (modified Logarithmic
This distribution is not valid in the close the distribution and wake):
wall .also in the close the edge of the boundary layer u 1 y πy y 
was distance from experimental data’s. Nevertheless, = ln( ) + 2Π sin 2 ( ) + α ln(1 − )
u* κ  y0 2h h 
this relation use all flat plate turbulent boundary layer,
because that in the distribution velocity shape related ...(6)
to Local shear stress, is very helpful. Guo and Julien modified the wake law that
known the modified Logarithmic distribution and
(Yang and Colleagues,2004) modified wake:
Logarithmic distribution by analyses nayver
stokes rinolda equation and assuming a parabolic u 1 yu  2Π p y 1 y 3
= ln( * ) + B  + sin 2 ( )− ( )
distribution for Eddy viscosity. They pay attention u *  k u  k 2d 3k d
to effect on channel in velocity flow. The modified
...(7)
Logarithmic distribution is as follows:
(Keulegan,1938) introduced parabolic
u 1 y y  distribution in his article and explained this distribution
= ln( ) + α ln(1 − )
u* κ  y0 h  with experimental data’s is suggested by bazin.
...(3)
Parabolic distribution as follows:
2
b u max − u  η 
α = 1.3 exp(− ) = C1 −
2h ...(4) u*  η dip 
 ...(8)
In this relation b is Channel width. A=0the
above equation would be Logarithmic distribution. In the relation, c is a Numeric constant and
determine to using experimental data’s, ç Depth
(Coles,1956) was analyzed the velocity of the seabed (y/h), çdip=yumax/h that yumax
wind distribution in Dimensional low and add wake is accrued Maximum speed depth in umax, this
function to velocity Logarithmic distribution, the distribution is applied throughout the Flow depth
 Sara & Reza, Curr. World Environ., Vol. 10(Special Issue 1), 288-293 (2015) 290

in narrow channels (5e”b/h) and the central part in Materials and methods
External area. (Vedula and Rao,1985) concluded
The point then is practical in parabolic distribution In the research uses the Reliable
for Containing the suspended sediment inis domain experimental data’s related to (Venoni,1946) and
ç=0/2 to ç=0/3 and also illustrated the c parameter (Einstein,1955), to survey accuracy of each equation
on This intersection. (Sarma et al.,2000) check out and determine the best equation. The experimental
the Limitations and the credit of parabolic law. They data’s is reliable data’s and most researchers use it to
claim Laboratory studies that the maximum in focus evaluate equations. These experiments performed
is while the maximum speed accurate in the water in a flume width85/0 and 0/307 meter.
layer, Ç=0/5 and the amount of maximum speed
decreases in the water below layer. Also, they show The main statistical indicators like the
that both of Logarithmic law and parabolic law are mean error (AE), least square root error (RMSE)
valid in Internal and external border areas and and the regression coefficient (2R) is applied for
Velocity at this point is unique. Determine the best equation, to calculate each of
these indicators is used the following relationships:

n
∑ (Pi − O i )
i =1
A
E =
n ...(9)

n
∑ (Pi − O i )2
i =1
RMSE =
n ...(10)

2
 n 
 ∑ t t
(O − O i )(Pi − P i )
i 
R2 =  t =1 
 n n 


∑
t =1
(O it − O i ) 2 ∑ (Pi t − P i ) 2 
t =1 
Fig. 1: Velocity distribution profiles in open
channels ...(11)

Fig. 2: Logarithmic distribution of the experimental data’s

291 Sara & Reza, Curr. World Environ., Vol. 10(Special Issue 1), 288-293 (2015)

Fig. 3: Modified logarithmic distribution of the experimental data’s

Fig. 4: Wake distribution of the experimental data’s

Table 1: The results of the various equations with using statistical parameters

Experimental Statistics Logarithmic Modified Wake Wake Parabolic

Data’s Parameters Distribution Logarithmic Distribution Modified Distribution
Distribution Logarithmic
Distribution

Vanoni (1946) AE 0.268 0.271 0.26 0.299 -

RMSE 0.353 0.355 0.349 0.382 -
R2 0.99 0.99 0.991 0.988 -
Wang.X & Qian.N AE 0.95 1.025 0.94 0.951 1.58
(1989) RMSE 1.03 1.12 1.016 1.02 2.33
R2 0.99 0.988 0.991 0.991 0.929
 Sara & Reza, Curr. World Environ., Vol. 10(Special Issue 1), 288-293 (2015) 292

In this relation pi is the calculated values, Oi Also, in the figure 5. The calculated values
is measured values, in the number of samples and Ô by modified logarithmic distribution –wake (equation
the average value measured. Positive and negative 6) Compared Laboratory values
values of AE Represents the Estimated at more than
and less than the actual value, in order. The value In the figure 6, is plotted the calculated
of this index is lower; indicate the reliability of the values of the velocity parabolic distribution (equation
equation. The amount of RMSE is always positive 8) in contrast, laboratory values. As can be seen
and approaching zero, Examined the relationship this distribution only can be estimate Speed values
between increased. The amount of RMSE states in the external area and Cleary, Distracted the
the estimated error in the entire curve. experimental data’s in the local area.

Results and discussion To determine the velocity distribution

equation indices was calculated score to each
In the figure 2, is plotted the logarithmic equation. In the table 1 presented The values of
velocity distribution (solid line) by experimental these parameters. It can be seen For all statistical
data’s. in this picture can be seen the changes in based on Examined, wake distribution Relative
local area is more than the external part. superiority Compared to other equations. In the
other hand, Parabolic distribution is less accurate in
In the figure 3. Modified logarithmic this equation. Of course, it must talk that Parabolic
distribution (the equation 1) compared with distribution is applied only the external area and
experimental data’s. Also, this distribution acts occurs the maximum speed in below water level.
similar the logarithmic distribution. Given the Wake distribution well estimated Speed values in
Modified logarithmic distribution is intended effect the total depth flow (local and external areas).
of channel on the velocity distribution and due the
figure can be concluded that the channel dimensions Conclusion
are not effect to the results. Comparison of the wake
distribution (equation 2) illustrate in figure 4 by In the research check out for Calculate
experimental data’s. The wake distribution estimates the velocity distribution in open channels by
a month velocity in this picture. Also, this distribution most developed equations. Such equations are
estimates a month velocity in the inner and external investigated logarithmic distribution, modified
areas. logarithmic distribution, wake distribution- modified
logarithmic distribution-wake and parabolic

Fig. 5: Modified logarithmic distribution- Wake of the experimental data’s

293 Sara & Reza, Curr. World Environ., Vol. 10(Special Issue 1), 288-293 (2015)

distribution. Calibration Each of these equations, is

used the reliable experimental data’s. Then with using
several statistical indicators, was select the best
equation among equations. The results of equation
illustrated that the wake distribution reasonably
accurate estimates Speed values both of the local
and external parts and Recommended using the
equation in calculation relate to Depth distribution
rate, on the other side, Parabolic distribution s less
accurate and the use it will create large of errors in
Calculation. Also, it has been said that the equation
can only to estimate amount velocity in the outer
region.
Fig. 6: Parabolic distribution of the
experimental data’s

References

1. Chow, V.T. ,”Open Channel Hydraulics”. Hydraulic Engineering, ASCE, 112(5), 356-
McGraw-Hill Book Company, New York. 375 (1986)
(1959) 9. Sarma, K.V.N., Prasad, B.V.R., and Sarma,
2. Coleman, N.L. ,”Effects of suspended A.K. ,”Detailed Study of Binary Law for Open
sediment on the open-channel velocity Channels”, Journal Hydraulic Engineering,
distribution, Water Resources Research, 22 126(3), 210-214 (2000)
(10), 1377-1384 (1986) 10. Schlichting, H. ,”Boundary Layer Theory”,
3. Coles, D. ,”The law of the wake in the turbulent McGraw-Hill, New York. (1979)
boundary layer”, J. Fluid Mechanics, 1,191- 11. Vanoni, V.A.,”Transportation of suspended
226 (1956) sediment by water”, Trans. ASCE, 111, 67-
4. Einstein, H.A., and Chien, N. ,”Effects of 133 (1946)
heavy sediment concentrationnear the bed 12. Vanoni, V.A., and Nomicos, G.N. ,”Resistance
on velocity and sediment distribution”, U. properties in sediment laden streams”, Trans.,
S. Army Corpsof Engineers, Missouri River ASCE, 125,1140-1175 (1960)
Division Rep. 8. (1955) 13. Vedula, S., and Rao, A.R. . “Bed Shear from
5. Elata, C., and Ippen, A.T. ,”The dynamics Velocity Profiles: A New Approach”, Journal
of open channel flow with suspensions of Hydraulic Engineering, 111(1), 131-143
neutrally buoyant particles”, Technical Report (1985)
No.45, Hydrodynamics Lab, MIT. (1961) 14. Wang, X, and Qian, N. ,”Turbulence
6. Guo, J, and Julien, P.Y. ,”Application of Characteristics of sediment-laden flow”, J.
the Modified Log-Wake Law in Open- Hydr. eng., 115(6), 781-799 (1989)
Channels”, University of Nebraska- Lincoln, 15. Wang, Z, Larsen, P. ,”Turbulent Structure of
Civil Engineering Faculty Publications. Water and Clay suspensions with Bed Load”.
(2008) J. Hydr. eng., 120(5), 577-601 (1992)
7. Keulegan, G.H ,”Laws of turbulent flow in open 16. Yang, S.Q., Tan, S.K., and Lim, S.Y. “Velocity
channels”, J. Res., Nat.Bureau of Standards, distribution and dip-phenomenon in smooth
21, 707-741(1938) uniform open channel flows”, J. Hydraulic
8. Parker, G., and Coleman, N.L. “Simple Engineering, 130(12), 1179-1186 (2004).
model of sediment-laden flows”, Journal of