The 6th Int. Conf. on Hydroscience and Engineering (ICHE-2004), May 30-June 3, Brisbane, Australia 1

THE EFFECT OF THE SUBMERGENCE LEVEL ON THE RESISTANCE OF GROYNES

AN EXPERIMENTAL INVESTIGATION

Mohamed F. M. Yossef1

ABSTRACT

In this study, the resistance of groynes is investigated for different submergence levels. For this

reason, experiments have been conducted in a physical model with a scale of 1:40 for a schematised

river reach, which is based on the geometry of the Dutch River Waal. Four electromagnetic

flowmeters (EMF) were employed in order to obtain the horizontal velocity components. Transverse

velocity profiles for all test cases were measured. As the tests in this study were dedicated to

investigating the effect of submergence level on the flow in the groynes region, all test cases were

chosen to guarantee submerged flow conditions. From the analysis of the effect of submergence, it

was possible to devise a relation between the blockage by a groyne and the effective roughness in

the groynes region. Such a relation would allow an estimate to the effect the groynes on the effective

roughness of the river as well as an assessment to the effect of lowering the crest level of existing

groynes.

1. INTRODUCTION

Groynes are structures constructed in rivers at an angle to the flow direction in order to deflect the

flow away from critical zones. They serve to maintain a suitable channel for the purpose of flood

control, bank protection and to improve navigation conditions. On the one hand, the groynes

enhance the velocity in the main channel. On the other hand, the region between the groynes is

either a dead zone during emerged conditions, or a slow flow zone during submerged flow

conditions. This velocity difference leads to the formation of a mixing layer, through which,

exchange of mass and momentum between the groyne fields (slow stream) and the main channel

(fast stream) takes place.

The last few years have seen a number of experimental work to study the details of the flow

near groynes, e.g. Tominaga et al. (2001), Uijttewaal et al. (2001), Kurzke et al. (2002), and Yossef

& Uijttewaal (2003). Owing to the efforts of those researchers, the hydrodynamics of the flow near

groynes have been gradually clarified. However, an important question remains to answer for the

river manager when attempting to assess the conveyance capacity of a river with groynes. This

question is; what is the resistance of the groynes region during flood conditions? A closely related

question arises from the plans to lower the existing groynes along the Rhine branches in the

Netherlands (see Silva et al., 2001) is: what is the effect of lowering the crest level of the existing

groynes on the resistance of the groynes region?

1 PhD candidate, Faculty of Civil Engineering and Geosciences, Delft University of TechnologyP.O. Box 5048, NL-2600 GA Delft, The Netherlands ( M.F.Yossef@ct.tudelft.nl)

2

In order to answer these questions; a thorough understanding to the effect of submergence

level on the flow in the groynes region is required. Furthermore, to be able to investigate the effect

of lowering the groynes it is important to assess the effect of different submergence levels on the

resistance of groynes. Accordingly, all test cases presented in this study were chosen to guarantee

submerged flow conditions. The objective of this study was to devise a formulation to represent the

resistance of groynes to the flow that allows us to asses the effect of changing the crest level of

existing groynes on the total roughness in the groynes region.

2. LABORATORY EXPERIMENT

2.1. Setup

A physical scale model was built in the laboratory for Fluid Mechanics of Delft University of

Technology. The total model measured 5.0 m in width and 30 m in length. The model consisted of a

schematised river reach, which was based on the geometry of the Dutch River Waal. In the lateral

direction, the model represented half of the river width with a geometrical scale of 1:40. Groynes

were constructed on a shallower part. In the full test series four different types of groynes were

implemented, see Uijttewaal et al. (2002). Figure 1, shows part of the set-up, showing a schematised

top view of the fourth groyne field in which the measurements were performed. The model bed is

fixed and flat in the main channel; and it is sloping towards the bank in the area between the

groynes.

0.15 m

3rd groyne

field

5th groyne

field

4.5 m

3.0 m

2.0 m

v-direction

u-direction

total flume length = 30 m

0.25 m

1:30

1:3

1:20.08 m

0.75 m

section: A B C D

2.25 m

4th groyne

field

(a) (b)

Figure 1: Experimental set-up, for each experiment five identical groynes were implemented; (a) top

view with indication of the velocity measurements locations. (b) cross-sectional view.

In this study, we will consider only one of the tested groyne types, which represents the

standard groyne as it is typically found in the Dutch River Waal (type-A). It is a straight dike with a

horizontal crest perpendicular to the riverbank with slopes of 1:3 on all sides, see Figure 1b. For all

the tested cases; Froude number (Fr) was small enough to ensure subcritical flow condition (see

3

Table 1); and Reynolds number (Re) was guaranteed to be high enough to ensure a fully developed

turbulent flow in both the main channel region (Re ≅ 6·104) and the groyne fields region (Re ≅ 10

4).

In a separate test, the bottom roughness was evaluated for the main channel region and Nikuradse’s

roughness coefficient ks was estimated: ks ≅ 6.27.10-4

m.

Four devices of electromagnetic flowmeters (EMF) were used, measuring the velocity in the

horizontal plane, u- and v-components. In accordance with findings from earlier investigation

(Yossef & Uijttewaal, 2003), the velocity measurements were chosen to take place along section-C

(Figure 1a), to ensure that the upstream groyne is not affecting the velocity profile. Section-C was

located at a distance of 2.25 m from the upstream groyne and extended 2.6 m in the transverse

direction. Velocity was measured in 12 lateral points at mid-depth. At each point, the measurement

lasted for a period of 360 seconds with a sampling rate of 10 Hz.

2.2. Experimental conditions and procedures

All test cases presented in this study were chosen to guarantee submerged flow conditions. Four

different cases were considered; the hydraulic conditions of which are given in Table 1.

The procedures for running the experiment were as follows; the downstream tailgate was

adjusted to a certain level and the model was fed with water from the upstream side. From the

inflow control valves, the discharge was adjusted to highest discharge and then reduced to the lowest

on steps. Keeping the tailgate elevation constant and varying the discharge, we eventually vary both

the water surface slope and the flow depth to produce different hydraulic conditions to produce test

cases from 01 to 06 in Table 1. The hydraulic conditions could be represented best in the form of

Froude number (Fr = u/√gh).

After a complete set had been completed, the tailgate was set to a different elevation and the

same procedures are repeated to establish a new set. Before carrying out the measurements, the flow

was given enough time to adjust to the new discharge condition. In each test, the water level was

measured and the horizontal depth-averaged velocity was recorded using EMF at section-C as in

Figure 1.

Table 1: Hydraulic conditions for all test cases

SA SB SC SDTest

caseQ

(m3/s) Ht (cm) hg/Ht Fr Ht (cm) hg/Ht Fr Ht (cm) hg/Ht Fr Ht (cm) hg/Ht Fr

01 0.383 35.30 0.59 0.19 28.36 0.79 0.29 33.01 0.64 0.21 30.50 0.71 0.25

02 0.333 34.34 0.61 0.17 27.44 0.82 0.28 32.56 0.65 0.19 29.59 0.74 0.23

03 0.278 33.33 0.63 0.16 --- --- --- 31.59 0.68 0.18 28.58 0.78 0.21

04 0.220 32.15 0.66 0.14 --- --- --- 30.44 0.71 0.15 27.5 0.82 0.19

05 0.162 30.44 0.71 0.11 --- --- --- 29.04 0.76 0.12 26.09 0.88 0.15

06 0.102 29.03 0.76 0.08 --- --- --- 27.37 0.83 0.09 24.54 0.97 0.10

3. RESULTS AND ANALYSIS

3.1. Flow pattern

When the groynes are submerged the flow in the groyne fields region could be characterised as a

slow velocity region. The momentum transfer by the water flowing over the groynes is sufficient to

balance the momentum transfer through the mixing layer, that otherwise would have caused a

4

recirculating flow. It should be also noted that the flow over the groynes hinders the horizontal

recirculation, ultimately causing it to disappear when reaching a high enough submergence level; for

further details on the dynamics of the flow near groynes see Yossef & Uijttewaal (2003).

0.0

0.2

0.4

0.6

0.0 1.0 2.0 3.0 4.0 5.0

Y distance (m)

Vel

oci

ty (

m/s)

SA01 SA02 SA03 SA04 SA05 SA06

0.0

0.2

0.4

0.6

0.0 1.0 2.0 3.0 4.0 5.0

Y distance (m)

Vel

oci

ty (

m/s)

SB01 SB02

(a) (b)

0.0

0.2

0.4

0.6

0.0 1.0 2.0 3.0 4.0 5.0

Y distance (m)

Vel

oci

ty (

m/s)

SC01 SC02 SC03 SC04 SC05 SC06

0.0

0.2

0.4

0.6

0.0 1.0 2.0 3.0 4.0 5.0

Y distance (m)

Vel

oci

ty (

m/s)

SD01 SD02 SD03 SD04 SD05 SD06

(c) (d)

Figure 2: Transverse velocity profile for test cases SA, SB, SC, and SD – line without markers

indicates the bed level.

0

0.2

0.4

0.6

0.8

1

1.2

1 1.5 2 2.5 3 3.5 4 4.5 5

Y distance (m)

u/u

mc

SA01 SA02 SA03 SA04 SA05 SA06 SB01

SC01 SC02 SC03 SC04 SC05 SC06 SB02

SD01 SD02 SD03 SD04 SD05 SD06

channel side groyne fields side

Figure 3: Normalised velocity profiles; velocity is normalised by the main channel velocity (umc).

5

The transverse velocity profiles for the different submerged flow conditions were measured

and plotted in Figure 2, where we can observe that both the velocity in the main channel and the

groyne fields changes, whereas, the mixing layer width is nearly constant in all cases. The profiles

are then normalised by dividing the velocities by the velocity in the main channel, the resulting

normalised profiles are collected and given in Figure 3, where we can observe that, the mean

velocity in the groynes region (ugr) varies from 10 to 40% of the main channel velocity (umc). This

variation is thought to be due to the difference in resistance that a groyne poses to the flow in

accordance with the blockage ratio (hg/Ht) as given in Table 1.

To evaluate the resistance in the groyne fields region we need to estimate the resistance of the

groynes. There are two approaches to represent the resistance of the groynes. The first approach is to

consider the groyne as a submerged weir, and the second is to consider that the groyne acts as a

large obstacle i.e. a roughness element. In the following two sections, we discuss these two

approaches.

3.2. Groynes as a submerged weir

In this approach the discharge per unit width (q) over one groyne is represented in a weir form (see

Mosselman & Struiksma, 1992). The formula reads:

( ) 2

0

d t g t gm H h g h if H hq

else

⋅ − ⋅ ⋅ ∆ >=

(1)

where: md = weir discharge coefficient; Ht = total water depth; hg = groynes height; and ∆h =

pressure drop over a groyne (∆h = S·i), with S as the spacing between two groynes and i is the slope

in the main channel region

It is assumed here that the losses in the groynes region is solely due to the groynes, and the

hydraulic gradient is zero between two successive groynes. This assumption could not be verified

from the experiments as no water level drop was observed over the groynes. Furthermore, it is

assumed that the water level in the groynes region is equal to the main-channel water level

immediately downstream of the groyne.

After filling in Eq. 1, it was possible to deduce the discharge coefficient md for each test case.

The calculated md in this approach was found to be more than unity for all cases. This result

contradicts the fact that the maximum value of md should not exceed unity for any type of weir.

Apparently, the discharge in the groynes region is more than that passes over a weir with the same

hydraulic conditions of a groyne. This could be attributed to substantial transverse momentum

exchange from the fast stream in the main channel. Thus, the weir approach was dropped out and the

second approach was considered.

3.3. Groynes as a roughness element

In this approach, the groynes are modelled as a roughness element. Consequently, we could assume

that the resistance in the groynes region has two different sources; the first is the bed resistance, and

the second is the resistance due to the groynes. A formulation of the resistance of the groyne could

follow the form of a drag resistance, see for example Aya et al. (1997). Accordingly, the momentum

balance for a unit area in the groynes region away from the mixing layer could be written in the

following form:

6

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.5 0.6 0.7 0.8 0.9 1.0

Blockage ratio (h g/H t)

CD

SA

SB

SC

SD

Figure 4: Relation between drag coefficient of a groyne and submergence stage for all test cases.

�

2 2

2

1

2

g

gr D gr

base

hgghi u C u

C Stotal resistance

in the groynes region bed resistance groynes resistance

= +

����� �������

(2)

where: i is the local water surface slope; h is the local flow depth; Cbase is the base Chézy coefficient

(could be approximated to be similar to Chézy coefficient in the main channel); ugr is the velocity in

the groynes region away from the mixing layer; hg is the groyne height; S is the spacing between two

groynes; and CD is a representative drag coefficient for the groynes.

We must mention here that the drag coefficient CD in Eq. 2 is not a constant as a standard drag

coefficient should be. Yet, CD here is a characteristic drag coefficient that accounts for the resistance

of the groynes and it is a function of both the hydraulic conditions and the blockage that a groyne

poses to the flow.

0

10

20

30

40

50

0.0 0.2 0.4 0.6 0.8 1.0

Blockage ratio (h g/H t)

CD/F

r

2

SA

SB

SC

SD

3.717

2

2

76.422

0.9238

gD

r t

hC

F H

R

=

=

Figure 5: Relation between the blockage by a groyne and its drag coefficient CD normalised by Fr2.

Data points are defined in legend, solid line is a data-fit line for [CD/Fr2 = a

(hg/Ht)

b], with a = 76.4

and b = 3.7

7

During the test it was possible to measure both the velocity umc and flow depth h. With the

knowledge of the bed roughness height ks, it is then possible to deduce the local slope i in the main

channel for every test case. Assuming that the water surface slope in the groynes region is similar to

that in the main channel, all parameters in Eq. 2 can be filled and it is hence possible to deduce CD

for every test case. Figure 4 shows the calculated CD plotted against the blockage ratio (hg/Ht).

The calculated values for CD shown in Figure 4 include the effect of the different hydraulic

conditions for each test case. To eliminate that effect, the values of the estimated CD were divided by

Froude number squared (Fr2) (see Fenton 2003). The result of this operation is presented in Figure 5,

from which we can observe that; excluding the last point estimated from test case (SD06) which has

a blockage ratio near unity; the calculated (CD/Fr2) shows an obvious increasing trend with

increasing blockage ratio (hg/Ht).

Many functions could fit the data points in Figure 5, e.g. linear, exponential, or power.

Nevertheless, taking into consideration that the flow over a groyne could be as well represented in

the form of flow over a weir, the relation between the resistance of a groyne and the blockage ratio

would be better represented in the form of a power relation that takes the form;

2

b

gD

r t

hCa

F H

=

(3)

where: a and b are constants, and Fr is calculated for the main channel region. The result of the

fitting is shown in Figure 5 as a solid line, where the constants a and b were found to equal 76.4, and

3.72 consecutively

3.4. Effective roughness in the groynes region

As we mentioned before, it is of a great interest for the river manger/modeller to be able to

schematise the groynes in a 1-D model by, for example, using an effective roughness in the groynes

region. This effective roughness is representative to the additional resistance by groynes.

Utilising Eq. 2, we can further develop a definition for an effective Chézy roughness

coefficient (Ceffective) as a function of the groynes crest height. Here, Ceffective is a representation of the

total resistance in the groynes region. We first rewrite Eq. 2;

2 2 2

2 2

1

2

g

gr gr D gr

effective base

hg gu u C u

C C S

= +

(4)

hence, the definition for Ceffective takes the following form:

2

1

1 1( )

2

effectiveg

D

base

Ch

CC g S

=

+

(5)

with CD deduced from Eq. 3, and the values of a and b estimated from the laboratory data in

Figure 5, it is possible to calculate Ceffective for any given prototype conditions, see Figure 6.

Furthermore, it is now possible to assess the effect of lowering the crest level of the existing

groynes on the resistance of the groynes region. This could be achieved by evaluating the rate of

change of Ceffective with changes in (hg/Ht). For example, lowering the crest level of the existing

groynes along the River Waal by 1.0 m will reduce the blockage ratio during design floods from a

value of around 0.6 to a value of around 0.5. Such a measure would lead to increasing the effective

8

Chézy coefficient Ceffective by approximately 25%, for a Froude number around a value of 0.2 (see

Figure 6). This increase in Chézy coefficient in addition to the increase of the flow area could be

used to evaluate the reduction in design flood level.

4. SUMMERY AND CONCLUSIONS

Velocity measurements were carried out in a physical model of a river with groynes based on the

geometry of the Dutch River Waal. Identifying the resistance of the groynes to the flow was of main

interest. The effect of different blockage ratios was studied by changing the flow depth. Two

formulations to represent groynes were investigated. The first was to consider the groyne as a

submerged weir and the second was to consider that it acts as a large obstacle.

Cer

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

0.1

0.2

0.3

0.4

0.5

0.6

0.95

0.95

0.9

0.9

0.85

0.85

0.8

0.8

0.75

0.75

0.7

0.7

0.65

0.65

0.6

0.6

0.55

0.55

0.5

0.5

0.45

0.45

0.4

0.4

0.35

0.35

0.3

0.3

0.25

0.25

0.2

0.2 0.15 0.1

Fr

Ceffective

Cbase

Blockage ratio ( hg/Ht)

Figure 6: Contour plot for the effective Chézy coefficient in the groynes region as a function of the

blockage ratio and Froude Number; Ceffective is divided by Cbase, (Cbase = 50 m0.5

/s).

When using the weir formulation we arrived at a discharge coefficient md more than unity.

This could be attributed to a substantial transverse momentum exchange from the main channel (fast

stream) to the groynes region (slow stream).

The representation of the groynes as a large roughness element led to the definition of a drag

coefficient for groynes (CD) that is a function of both the blockage ration (hg/Ht) and Froude Number

squared (Fr2) see Eq. 3. Furthermore, we were able to define an effective Chézy coefficient that is a

function of the base Chézy coefficient, drag coefficient, groyne height and spacing between groynes

(see Eq. 5). Such a definition allows a possible representation of groynes in one-dimensional

mathematical models in addition to a quick assessment to certain river engineering measures.

ACKNOWLEDGEMENT

This research was financially supported by the Road and Hydraulic Engineering Division (DWW) of

the Ministry of Transport, Public Works and Water Management. The authors wish to thank M. van

der Wal for his support and M.H. Berg for conducting part of the experiments.

9

REFERENCES

Aya, S., Fujita, I., and Miyawaki, N. (1997). "2-D Models for flows in river with submerged groins"

In: Proc. 27th

IAHR Congress, San Francisco, CA. USA, 829-837.

Fenton, J. D. (2003). "The effects of obstacles on surface levels and boundary resistance in open

channels" in: XXX IAHR Congress, Thessaloniki, Greece, 9-16.

Kurzke, M., Weitbrecht, V., and Jirka, G. H. (2002). "Laboratory concentration measurements for

determination of mass exchange between groin fields and main stream" In: River Flow 2002 -

Proc. of the Int. Conf. on Fluvial Hydraulics, Louvain-la-Neuve, Belgium, 369-376.

Mosselman, E., and Struiksma, N. (1992). "The effect of lowering groynes" Q1462, WL | Delft

Hydraulics, Delft.

Silva, W., Klijn, F., and Dijkman, J. (2001). "Room for the Rhine branches in the Netherlands"

report 2001.031, Riza and WL | Delft Hydraulics.

Tominaga, A., Ijima, K., and Nakano, Y. (2001). "Flow structures around submerged spur dikes

with various relative height" In: Proc. 29th

IAHR Congress, Beijing, China, 421-427.

Uijttewaal, W. S. J., Lehmann, D., and van Mazijk, A. (2001). "Exchange process between a river

and its groyne fields - Model experiments" Journal of Hydraulic Engineering, ASCE, 127(11),

pp. 928-936.

Uijttewaal, W. S. J., Berg, M. H., and van der Wal, M. (2002). "Experiments on physical scale

models for submerged and non-submerged groynes of various types" In: River Flow 2002 -

Proc. of the Int. Conf. on Fluvial Hydraulics, Louvain-la-Neuve, Belgium, 377-383.

Yossef, M. F. M., and Uijttewaal, W. S. J. (2003). "On the dynamics of the flow near groynes in the

context of morphological modelling" in: XXX IAHR Congress, Thessaloniki, Greece, 361-368.