CLEAR WATER SCOUR AT CYLINDRICAL PIERS IN CLAY SAND MIXTURES

Thesis submitted to Indian Institute of Technology, Kharagpur in partial fulfillment of the requirements for the award of the degree of

Master of Technology
in

Hydraulic and Water Resources Engineering

Submitted by,

Mr. Langhi Manojkumar Namdeo (07CE6108)

Under the guidance of

Prof. Subhasish Dey Chair Professor, IIT Kharagpur

DEPARTMENT OF CIVIL ENGINEERING INDIAN INSTITUTE OF TECHNOLOGY

KHARAGPUR- 721302, INDIA

2009

Dedicated to,

My lovely Parents and Friends

Departmrnt of Civil Engineering, Indian Institute of Technology. Kharagpur-721302 Certificate This is to certify that the thesis entitled Clear water scour at cylindrical piers in clay sand mixtures is a bonafeid work carried out by Mr. Manojkumar N. Langhi under my supervision and guidance for the partial fulfillmet of the requirements for Postgraduate degree of Master of Technology in Hydraulic and water Resources Engineering during the academic session 2007-2009 in the Civil Engineering Department, Indian Institute of Technology, kharagpur, India.

Prof. Subhasish Dey Department of Civil Engineering Indian Institue of Technology Kharagpur India

Acknowledgement
I would like to express my heartfelt gratitude to my project supervisor Prof. Subasish Dey, Chair Professor, Department of Civil Engineering, Indian Institute of Technology, Kharagpur, for his invaluable guidance, constant encouragement, talented and versed advice and helpful suggestions. I am grateful to Prof. L. S. Ramachandra, Head of the Department, Civil Engineering and also thankful to all the faculty of Civil engineering department. I am very much thankful to Mrs. S. Talukdar madam, Head of Laboratory, Indian Institute of Technology, Kharagpur, for providing necessary facilities during the research work. I would like to thank Mr. S. Sarkar, Mr. R. Das, Mr. R. Acharya and Mr. D. Deb for their co-operation and encouragement during the research work. I am also thankful to Amol, Anirudha, Avinash, Irfan, Nilesh, Parag, Pinaki, Ramesh, and Santosh for their cooperation during my project work. I extend my sincere thanks to all, officer, laboratory staff and my friends, who were very co-operative and always eager to help me. I owe a great deal of love, to my parents, my sister, brothers, sister in law and a friend Sanghu, for their blessing and consistent moral support during my study. Finally, I bow before the Almighty who has enable me to complete the project work successfully.

IIT, Kharagpur Date: . 12. 2009 (Manojkumar N. Langhi)

CONTENTS Chapter Description List of Tables List of Figures List of Symbols Abstract I 1.1. 1.2. II 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.6.1 2.6.2 2.6.3 III INTRODUCTION General Objectives of present investigation LITERATURE REVIEW General Scour and its classification Mechanism of local scour Scour in non-cohesive and cohesive soils Parameters influencing scour depth at piers Influence of parameters on scour depth Approaching flow velocity Approaching flow depth Time - variation of scour EXPERIMENTAL PROCEDURE 3.1. 3.2. 3.2.1 3.2.2 3.2.3 3.3. 3.3.1 3.3.2 3.4. 3.4.1 3.4.2 3.4.3 3.4.4 General Experimental setup Flume Water supply system Instrument carriage Scheme of Experiments Non-cohesive sediments The pier model Method of measurements Discharge Bed and water levels Scour depth Velocity and flow field SETUP AND 11 11 11 11 11 12 12 12 12 12 12 13 13 13 Page No. i ii v vi 1 1 3 4 4 4 4 6 7 8 8 9 10

3.5. 3.5.1 3.5.2 3.5.2 IV 4.1. 4.2. 4.2.1 4.2.2 4.3. 4.3.1 4.3.2 4.3.3 4.4 4.4.1 4.4.2 4.4.3 V VI

Experimental Procedure Non-cohesive sediments After 5 % mixing After 10 % and 20 % mixing RESULTS AND DISCUSSION General Time variation of scour depth Scour for non-cohesive soil Scour for mixture of clay and non-cohesive sand 7.5 cm pier model Time-Velocity variation Turbulent Intensity Reynolds stresses 3.8 cm pier model Time-Velocity variation Turbulent Intensity Reynolds stresses SUMMARY AND CONCLUSIONS REFERENCES

15 15 16 16 17 17 17 17 18 28 28 31 33 35 35 38 40 43 45

LIST OF TABLE Table Title Page No.

4.1

Experimental data of obtaining maximum scour depth for different percentage of clay for different pier model 19

LIST OF FIGURES Figure 2.1 2.2 Title Flow pattern around bridge piers Time-variation of clear water and live bed scour after Chabert and Engeldinger (1956) 3.1 4.1 4.2 4.3 Schematic diagram of the experimental set-up Time-depth variation for 1 cm pier model Time-depth variation for 2 cm pier model Time-depth variation for 3.8 cm pier model (before mixing of clay in non-cohesive sediment) 4.4 Time-depth variation for 7.5 cm pier model (before mixing of clay in non-cohesive sediment) 4.5 Time-depth variation for 3.8 cm pier model (after mixing 5 % of clay in non-cohesive sediment) 4.6 Time-depth variation for 3.8 cm pier model (after mixing 10 % of clay in non-cohesive sediment) 4.7 Time-depth variation for 3.8 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 4.8 Time-depth variation for 7.5 cm pier model (after mixing 5 % of clay in non-cohesive sediment) 4.9 Time-depth variation for 7.5 cm pier model (after mixing 10 % of clay in non-cohesive sediment) 4.10 Time-depth variation for 7.5 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 4.11 Photograph of the scour hole for 3.8 cm pier model (before mixing of clay in non-cohesive sediment) 4.12 Photograph of the scour hole for 3.8 cm pier model (after 5 % mixing of clay in non-cohesive sediment) 4.13 Photograph of the scour hole for 3.8 cm pier model (after 10 % mixing of clay in non-cohesive sediment) 4.14 Photograph of the scour hole for 3.8 cm pier model (after 20 % mixing of clay in non-cohesive sediment) 26 25 25 24 24 23 23 22 22 21 21 20 10 14 19 20 Page 6

ii

4.15

Photograph of the scour hole for 7.5 cm pier model (before mixing of clay in non-cohesive sediment) 26

4.16

Photograph of the scour hole for 7.5 cm pier model (after 5 % mixing of clay in non-cohesive sediment) 27

4.17

Photograph of the scour hole for 7.5 cm pier model (after 10 % mixing of clay in non-cohesive sediment) 27

4.18

Photograph of the scour hole for 7.5 cm pier model (after 20 % mixing of clay in non-cohesive sediment) 28

4.19

Time-velocity variation for 7.5 cm pier model (before mixing of clay in non-cohesive sediment) 29

4.20

Time-velocity variation for 7.5 cm pier model (after mixing 5 % of clay in non-cohesive sediment) 30

4.21

Time-velocity variation for 7.5 cm pier model (after mixing 10 % of clay in non-cohesive sediment) 30

4.22

Time-velocity variation for 7.5 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 31 Vertical distribution of u+ and w+ at vertical section (before mixing of clay in non-cohesive sediment) 32 Vertical distribution of u+ and w+ at vertical section (after 5 % mixing of clay in non-cohesive sediment) 32
+ +

4.23

4.24

4.25

Vertical distribution of u and w at vertical section (after 10 % mixing of clay in non-cohesive sediment) 33 Vertical distribution of u+ and w+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 33 Vertical distribution of uw+ at vertical section (before mixing of clay in non-cohesive sediment) 34 Vertical distribution of uw+ at vertical section (after 5 % mixing of clay in non-cohesive sediment) 34 Vertical distribution of uw+ at vertical section (after 10 % mixing of clay in non-cohesive sediment) 35 Vertical distribution of uw+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 35

4.26

4.27

4.28

4.29

4.30

iii

4.31

Time-velocity variation for 3.8 cm pier model (before mixing of clay in non-cohesive sediment) 36

4.32

Time-velocity variation for 3.8 cm pier model (after mixing 5 % of clay in non-cohesive sediment) 37

4.33

Time-velocity variation for 3.8 cm pier model (after mixing 10 % of clay in non-cohesive sediment) 37

4.34

Time-velocity variation for 3.8 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 38
+ +

4.35

Vertical distribution of u and w at vertical section (before mixing of clay in non-cohesive sediment) 39 Vertical distribution of u+ and w+ at vertical section (after 5 % mixing of clay in non-cohesive sediment) 39 Vertical distribution of u+ and w+ at vertical section (after 10 % mixing of clay in non-cohesive sediment) 40
+ +

4.36

4.37

4.38

Vertical distribution of u and w at vertical section (after 20 % mixing of clay in non-cohesive sediment) 40 Vertical distribution of uw+ at vertical section (before mixing of clay in non-cohesive sediment) 41 Vertical distribution of uw+ at vertical section (after 5 % mixing of clay in non-cohesive sediment) 41
+

4.39

4.40

4.41

Vertical distribution of uw at vertical section (after 10 % mixing of clay in non-cohesive sediment) 42 Vertical distribution of uw+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 42

4.42

iv

LIST OF SYMBOLS Particular d50a H l U Ua Uc Uca u* u*c

u

Description Median diameter of sediment particles Approaching flow depth (L) Transverse length of abutments (L) Average approaching flow velocity (LT -1) 0.8Uca (LT-1) Critical velocity for sediment particles (LT-1) Critical velocity for armor particle size d50a (LT-1) Shear velocity of approaching flow (LT -1) Critical shear velocity of bed sediment (LT -1) Fluctuating component of streamwise velocity (LT -1) Fluctuating component of vertical velocity (LT -1) Normalized streamwise turbulent intensity component (M0L0T0) Normalized vertical turbulent intensity component (M0L0T0) Normalized vertical depth (L0) Normalized Reynolds stresses (M0L0T0)

u+ w+ y+ uw+

ABSTRACT Scour holes created by three-dimensional flow of water around bridge piers are a major cause of failure of bridge pier foundations. An evaluation of the effects of scouring around bridge piers forms necessary step in bridge design. The problem of scouring at cylindrical pier model on non-cohesive sand and on a bed containig different percentage of clay in non-cohesive sand was investigated experimentally. All the experiments were performed in a 12 m long, 0.6 m wide and 0.71 m deep horizontal flume. Non-cohesive sand of diameter 0.15 mm, different percentage of clay such as 5, 10 and 20 % and different pier models were used in the experimental runs. The time-averaged velocity components, turbulent intensity components, vertical depth components and Reynold stresses within the scoured bed were taken by the Acoustic Doppler Velocimeter (ADV) at the upstream side of two different cylindrical pier models. Four pier size of diameter 7.5 cm, 3.8 cm, 2 cm and 1 cm were considered for depth measurement in initial set of experimental runs for noncohesive sand. In such bed condition velocity measurements were performed only for 7.5 cm and 3.8 cm pier model. For further sets of experimental runs, thoroughly mixed clay content of 5 %, 10 % and 20 % in non cohesive sand were used for depth and velocity measurements in the vicinity of 7.5 cm and 3.8 cm pier model. An experimental result have shown that the time required to attain maximum constant scour depth in non-cohesive sand is less and therefore, low maximum constant scour depth was obtained due to increment of clay content in non-cohesive sand. The volume of scour hole at the upstream of the pier model was decreased with increased in clay content and the flow velocity in the scour hole of non-cohesive sand with higher clay content was also got reduced. Due to flow separation, pronounced bulges were observed in the vertical distribution of normalized streamwise turbulent intensity component and Reynolds stresses, while spike was observed near the bed for turbulent intensity components because of the shuddering effect of the primary vortex.

Keywords: Pier models; three-dimensional flow; scour.

vi

CHAPTER I

INTRODUCTION
1.1 General An alluvial river bed is subjected to continuous change. Flowing water erodes, transports and deposits sediment in the river, altering its bed elevation and adjusting its boundaries. Changes in bed elevation may be due to natural causes or by the activities of man which lead to changes of river bed or river geometry. Scour around bridge piers is just one example of the many different effects resulting from human interference with the river. Scour holes created by flow of water past bridge piers are a major cause of failure of bridge pier foundations. Failure of bridges due to such scour at their foundation is a common occurrence and each year a colossal amount is spent to repair, reconstruct or replace bridges whose foundations have been under-cut by the scouring action of stream flow. In the year 1947 the considerable bridge losses in the State of Iowa were in large measure responsible for the determination of the Iowa State Highway Commission to sponsor an intensive study of the problem with the goal of evolving means for predicting probable scour depths (Laursen et al. 1956). As of 1995 it was estimated that approximately 84 percent of the 575,000 bridges in the National Bridge Inventory are built over waterways (Richardson et al. 1995). Of these bridges, approximately 121,000 are considered to be scour susceptible and of those 121,000, approximately 13,000 are considered to be scour critical (Jones 1993). A study completed by the Transportation Research Board in 1984 estimates that an average of 150 bridges in the United States fail each year due to sediment transport and local scouring of piers or abutments (Davis 1984). Between the years 1985 and 1987, a total of 90 bridges were destroyed in New York, Pennsylvania, Virginia and West Virginia due to either pier or abutment failure. In 1994 the state of Georgia experienced over 500 bridge failures due to scour caused by Hurricane Alberto (Jones 2002). It is apparent that failures of bridges have brought significant life and financial losses. To ensure public safety and minimize the losses of bridge failures, more extensive studies on scour at bridge crossings are necessary. In particular, comprehensive studies deciphering the mechanisms themselves which initiate scour should be at the forefront of any current or future research. Until these initiating 1

mechanisms are well understood, the potential for scour around bridge support structures could prove to be a major concern for bridge design engineers. Due to the overall complexity of field conditions there is no generally accepted principle for the prediction of scour around bridge piers and abutments have evolved from field experience alone. The flow of individual streams exhibits a multiple variation, and great inequality exists among different rivers. The alignment, cross section, discharge, and slope of a stream must all be correlated with the scour phenomenon, and this in turn must be correlated with the characteristics of the bed material ranging from clays and fine silts to gravels and boulders. Finally, the effect of the shape of the obstruction itself - the pier or abutment - must be assessed. Since several of these factors are likely to vary with time to some degree, and since the scour phenomenon as well is inherently unsteady, sorting out the influence of each of the various factors is virtually impossible from field evidence alone. An analytical approach is equally difficult. If an obstruction, such as a pier, is placed in a stream, the flow pattern in the vicinity of that obstruction will be modified. Because the capacity for the transport of sediment is a function of the flow, the transport-capacity pattern will also be modified. In any area where, as a result of the modified pattern, the capacity for transport out of the area is greater than the rate at which material is supplied to the area, scour will occur. Conversely, where the transport capacity is less than the rate of supply, deposition will occur. The resultant changes in the stream bed will further modify the flow pattern - and the capacity pattern - until equilibrium between capacity and supply is again achieved at every point on the stream bed. An analytic solution would have to combine a prediction of the flow pattern and a description of the local transport capacity of the flow. Although an approximation of the flow pattern might be attempted, a comparable solution for the capacity is not yet possible. The experimental approach has been tried in the past with limited success, usually because the goal was restricted to a particular installation or to some special phase of the general problem. The earliest report on a laboratory study which has done by Engels at Dresden, Germany, in 1894. In that report it described that study reference is made to an earlier one in France in 1873 by Durand-Claye. Neither these early experiments nor subsequent studies done in later period by various investigators in various countries have been sufficiently general to obtain the desired result - a means of predicting scour in the field. However, considerable investigations on pier scour have been carried out further and a reliable design method is now available 2

(Melville and Sutherland 1988). These all the investigations pertain to scour around piers founded in cohesionless sediment. Study on the problem of local scour around bridge piers in cohesive sediments is still in its intial stage. Unlike in the case of noncohesive sediments, the flow condition at which cohesive material gets eroded is difficult to predict as it depends upon a variety of factors such as the type and percentage of clay content present, stage of compaction or consolidation etc. Further, only limited study has been carried out on the temporal variation of scour depth around bridge piers founded in cohesive sediments. 1.2 Objectives of present investigation The aim of the present investigation is to study experimentally the flow field, influence of different parameters on equilibrium scour depth, time variation of scour depth at cylindrical piers under clear water scour condition. The main objectives of the study are as follows: Investigation of the three-dimensional turbulent flow fields in the vicinity of two different cylindrical pier models placed on non-cohesive sand and on a bed containing different percentage of clay in non-cohesive sand. Determination of time-variation of scour depth for various bed conditions around different cylindrical pier models. Determination of time-velocity variation for various bed conditions around two different cylindrical pier models.

CHAPTER II

LITERATURE REVIEW
2.1 General In this Chapter, a comprehensive review of the investigations on local scour at bridge pier is presented. Scour and its classification, scouring mechanism, parameters affecting scour depth and time-variation of scour are discussed in successive section. 2.2 Scour and its classification Scour is a natural phenomenon of lowering the level of riverbeds by the erosive action of flowing stream. Scour is classified into two types, general scour and local scour. General scour in the river occurs due to change in the characteristics of river while local scour develops near the structure due to modification of the flow field as a result of obstruction to the flow by the structures. On the basis of time taken for scour development, general scour can be categorized as short-term scour and longterm scour. Short-term general scour develops during a single or several closely spaced floods. It may occur due to convergence of flow, a shift in the channel thalweg or braids within the channel, and bed-form migration. On the other hand, the longterm general scour is the general aggradation or degradation of streambed elevation due to natural (e.g. channel straightening, volcanic activities, and climate change) and human causes (e.g. channel alterations, streambed mining, dam/reservoir construction, and land-use changes). Local scour is classified as clear-water scour and live-bed scour. Clear-water scour occurs when the sediment is removed from the scour hole but not supplied by the approaching stream. In contrast, the live-bed scour occurs when the scour hole is continuously fed with the sediment by the approaching stream. 2.3 Mechanism of local scour The boundary layer in the flow past a bridge element undergoes a threedimensional separation. The dominant feature of the flow about a pier is the system of vortices which develops. The most important of these are the horseshoe vortex and the wake-vortex system. Laursen and Toch (1956) described the formation of horseshoe vortex. At the nose of the pier the approach flow velocity goes to zero. Since the flow velocity decreases from a maximum at the free surface to zero at the bed, the stagnation pressure decreases with distance from the water surface and this pressure 4

difference drives the flow. Therefore, separation occurs at the upstream face of pier and shear layer rolls up along the obstruction to form a vortex system in front of the element which is swept downstream by the river flow. Viewed from the top, this vortex system has the characteristic shape of a horseshoe and thus called a horseshoe vortex. The horseshoe vortex results from a concentration by the pier of vorticity already present in the approaching flow. However, the wake-vortex system is generated by the pier itself (figure 2.1). The formation of the horseshoe vortex and the associated downflow around the bridge element results in increased shear stress and hence a local increase in sediment transport capacity of the flow. This leads to the development of a deep hole (scour hole) around the bridge element, which in turn, changes the flow pattern causing a reduction in shear stress by the flow thus reducing its sediment transport capacity. The temporal variation of scour and the maximum depth of scour at bridge elements therefore mainly depend on the characteristics of flow, pier and river-bed material. The formation of the horseshoe vortex and the associated downflow cause scour at different elements of a bridge such as pier, abutment and spur dike. The mechanism of scour around bridge piers has been studied by Melville (1975), Kothyari et al. (1992a & b), Dey (1995), Dey et al. (1995), Dey (1999), Horst (2004) whereas, studies on the mechanism of scour around abutments and spur dikes have been studied by Kothyari et al. (2001), Barbhuiya (2003), Dey et al. (2004 & 2005), Barbhuiya et al. (2004a & b).

Pier

Surface Roller

Wake Vortex

Downflow

Flow Scour hole

Sediment bed

Horseshoe vortex

Fig.2.1 Flow pattern around bridge piers

2.4 Scour in non-cohesive and cohesive soils Non-cohesive soil consists of the bed material ranging from very fine to very coarse. When bridge pier is constructed in such a strata and the discharge is sufficiently large, the scour development would progress. For non-cohesive sediment, the submerged density of the soil and gravity forces provides the main resistance to erosion. During scour development, the coarser particles would accumulate in the scour hole and partly inhibit further development of the scour. Ultimately the accumulated coarser material would stop further scour and the scour depth obtained would be much smaller than that in uniform material. The mechanism of cohesive material scour is fundamentally different from scouring of alluvial non-cohesive materials. The process involves not only the balancing of flow induced shear stresses and the shear strength of soils to withstand scour, but also the chemical and physical bonding of individual particles and the properties of the eroding fluid. Hence scour in cohesive materials is more complex and less understood than the scour in non-cohesive sandy material. It is believed that 6

scour in cohesive soils occurs when the fluid shear is sufficient to overcome the tensile strength of the bed material and the submerged unit weight of the soil. Very little work has been carried out on the basic mechanism involved on the scouring of cohesive soils. One reason could be complexity of the problem; the physico-chemical aspects and the resistance to scour in cohesive soils, particularly, governed by wide variations in the sediment properties. Many investigators have studied the scour phenomenon in coarse-grained soils while scouring in cohesive material was studied by Partheniades (1965), Kamphuis and Hall (1983), Briaud et al. (1999), Rambabu et al. (2003). From previous research Rambabu et al. (2003) concluded that the rate of erosion in cohesive soil is dependent on many parameters such as induced shear stress, moisture content and density of the soil type, shear strength of the soil, type of clay and its adsorbed complex, temperature etc. Whereas, according to Molinas et al. (1998a) cohesive materials, once eroded, remain in suspension. As a result, the phenomenon identified as clearwater local scour in non-cohesive materials always prevails. Along with eroding fluid properties, the scour process in cohesive soils is strongly affected by the amount of cohesive material present in the soil mixture as well as the types of mineral clay, initial water content, soil shear strength, and compaction of the clay. Hence by the knowledge gained in the past in cohesive material scour Molinas et al. used two different types of clay mixtures and studied local scour around abutments and analyzed the effects of compaction, initial water content, soil shear strength, and the approach flow conditions on abutment scour. Molinas et al. (1998b) studied pier scour in montmorillonite clay soils and along with analyzing the effects of compaction, soil shear strength, and the approach flow conditions on pier scour in unsaturated cohesive soils and influence of initial water content of saturated clay on pier scour they developed scour prediction equations in unsaturated and saturated cohesive soils to quantify the scour which may occur around circular piers. 2.5 Parameters influencing scour depth at piers Scour at piers is influenced by various parameters (Breusers et al. 1977), which are grouped as follows: Parameters relating to the pier: Size, shape, spacing, number and orientation with respect to the approaching flow direction.

 Parameters relating to the bed sediment: Median size, particle size distribution, mass density, angle of repose, cohesiveness. Parameters relating to the approaching flow condition: Approaching flow velocity, approaching flow depth, shear velocity and roughness. Parameters relating to the fluid: Mass density, viscosity, gravitational acceleration and temperature (may not be important in scour problems). Parameters relating to the time: Time of scouring for an evolving scour hole. Parameters relating to the unsteadiness: Passage of flood wave in rivers and waves in marine environment. 2.6 Influence of parameters on scour depth 2.6.1 Approaching flow velocity The depth of the local scour hole is closely related to the undisturbed approach flow velocity. The idea about the effect of approach flow velocity on local scour depth under live-bed conditions have changed over the years. Early researchers related the relative scour depth (normalized by the flow depth) to the Froude number. Most of the conclusions drawn that for a given flow depth, the scour depth increase indefinitely, either at an increasing or a decreasing rate, with increasing velocities. The numerous equations relating normalized scour depth and Froude number are summarized by Melville (1975). Kandasamy (1989) showed that the scour depth increases with increase in flow depth due to incorporation of the flow Froude number. It is generally recognized that the shear velocity u* is an important parameter not only in distinguishing clear water scour from the live bed scour but also in representing the erosive power of the flowing stream for a given sediment size. Clear water scour occurs for the approaching flow velocity up to the critical velocity U c for bed sediments, that is U / U c 1 ; while live bed scour occurs when U / U c 1 .For nonuniform sediments, Melville and Sutherland (1988) defined an armor velocity U a , which marks the transition from clear to live bed conditions for a sedimenttransporting flow and is equivalent to U c for uniform sediments. Thus, for nonuniform sediments, live bed conditions prevail when U / U a 1. However, if U / U a 1 , armoring of the bed occurs as scour proceeds and clear water conditions exist. Dongol (1994) conducted an extensive series of experiments to study the effect of approaching flow velocity on scour depth at vertical-wall, wing-wall and spill-through abutments 8

under live bed conditions in uniform and nonuniform sediments. His results are complimentary to the studies of Chiew (1984) and Baker (1986) for live bed scour at bridge piers in uniform and nonuniform sediments, respectively. Chabert and Engeldinger (1956) stated that as the approach flow velocity exceeds the critical velocity for sediment entrainment, the scour depth decreases to about 10 % less than the maximum scour depth at the critical velocity and thereafter, an increase in the velocity has no effect on the local scour depth. However, it was recognized that under clear water conditions, the maximum scour depth occurs when

U U c . This scour depth is called the threshold peak. For U / U c 1 , that is under
live bed conditions, scour depth initially decreases with increase in approaching flow velocity reaching a minimum value and then increases again toward a second maximum. The second maximum occurs at about the transitional flatbed stage of sediment transport on the channel bed and is termed the live bed peak. 2.6.2 Approaching flow depth According to Laursen (1952), the approaching flow depth H is an important factor to determine scour depth. Experimental results of Kandasamy (1989) indicate that for a constant value of the shear velocity ratio u* / u*c , the maximum scour depth increases with the increase in approaching flow depth. It was also observed that the maximum scour depth increases at a decreasing rate with increase in approaching flow depth. According to Kandasamy (1989), for shallow flow depths, the scour depth increases proportionally with H, but is independent of l. On the other hand, for intermediate flow depths, the scour depth depends on both H and l. However, Melville (1992) distinguished short and long abutments. He concluded that for short abutments ( l / H 1 ), the scour depth is independent of flow depth; and for long abutments ( l / H 25 ), the scour depth is dependent on flow depth. However, most abutments are neither long nor short, as a result of which the scour depth is influenced by both H and l. There is a consensus that the maximum scour depth increases at a decreasing rate with increase in approaching flow depth and there exists a limiting depth corresponding to which the maximum scour depth is independent of the flow depth.

2.6.3 Time - variation of scour Figure 2.2 shows the schematic diagram describing the time-variation of scour depth at cylindrical pier after Chabert and Engeldinger (1956). Time to reach equilibrium scour depth varies widely, ranging from a day to a fortnight. Anderson (1963) stated By virtue of the logarithmic character of the development of the scour region with time, a practical equilibrium is reached after a relatively short time, after which the increase in the depth and extent of scour becomes virtually imperceptible. Rouse (1965), however, stated that scour is an ever-increasing phenomenon and there is no real equilibrium scour depth. Some of the researchers thought that the variation of scour depth with time is logarithmic but, few researchers proposed an exponential time-variation of scour; while Bresuers (1967) and Cunha (1975) gave a power law distribution. (see Barbhuiya 2003). General consensus is that the equilibrium scour depth at pier is attained asymptotically.

Fig. 2.2 Time-variation of clear water and live bed scour after Chabert and Engeldinger (1956)

10

CHAPTER III

EXPERIMENTAL SETUP AND PROCEDURE

3.1 General Experiments were carried out in Hydraulic and Water resources Engineering Laboratory of the Indian Institute of Technology, Kharagpur, India. The details of the experimental setup, scheme of experiments, experimental procedures and method of measurements are given in this Chapter. 3.2 Experimental Setup 3.2.1 Flume Experiments were performed in a horizontal, re-circulating flume with a rectangular cross-section 12 m in length, 0.6 m in width and 0.71 m deep. At the test section, the side walls of the flume were made of transparent glasses. At the inlet section of the flume concrete stilling basin was provided through which water enters into the flume. The stilling basin consisted of one perforated baffle wall and two vertical steel screens covering the full cross section for damping the flow turbulence and waves. An adjustable tailgate was installed at the downstream end of the flume to control the flow depth. The location of test section was made in such a way that the flow became fully developed before it reaches the test section. The sediment recess consisted of rectangular box made up of perspex sheet 12 mm in thickness with a dimension as 0.85 m length, 0.60 m width and inner depth of 0.165 m. The test section was located 3.5 m from the flume entrance. On the upstream side of sediment recess, the false floor of height 0.177 m were constructed above the original bed level of the flume in such a way that it allows water to pass uniformly over the test section without causing its turbulent characteristics over the sediment particles, while on the downstream side, the false floor of same height and 0.8 m in length was constructed. Five small holes were provided at the bottom of the downstream wall of the sediment recess to drain out the water from the sediment bed. Provision was also made to trap the washed-out sediment particles at the downstream side of floor by constructing barrier wall near tailgate with same height of the false floor as shown in fig 3.1. 3.2.2 Water Supply System The flume was connected to the water supply system comprised of a constant head reservoir about a height of 4 m above the ground level, an inlet tank, a large 11

underground reservoir and the pumps. The water was pumped to the constant head reservoir and supplied to the inlet tank through the valve fitted at their junction. A calibrated V-notch weir was fitted at the inlet tank through which water entered into the flume via stilling basin. 3.2.3 Instrument carriage An instrument carriage comprised of a main unit, which travelled on the rails in the longitudinal direction and the auxiliary unit, which carried the instruments, such as point gage, ADV probe etc. travelled in the transverse direction. 3.3 Scheme of Experiments 3.3.1 Non-cohesive Sediments The Indian Standard sieves were used for the preparation of sediment samples. The data of the sieve analysis were plotted to draw particle size distribution curves. From the curve the mean diameter of sample was selected to be 0.15 mm. This was used as base material to which clay was added in different proportions for further experimental runs. 3.3.2 The pier model The experiments were performed using four different types of perspex sheet pipes with diameters, 7.5 cm, 3.8 cm, 2 cm and 1 cm to symbolize a small scale model of a bridge pier. All the four type of piers were used for the experiments in noncohesive sediment. However, for the experiments in the mixture of sand-clay only two models (7.5 cm and 3.8 cm) were used. 3.4 Method of measurements 3.4.1 Discharge The discharge into the flume was regulated by a valve fitted at the junction of constant head reservoir and inlet tank and decided using calibrated V-notch fitted at the inlet tank. The V-notch was calibrated and calibration equation was used for the measurement of discharge Q, given as a function of head of water H above the sill level of the V-notch as

Q 0.9174H 2.159

3.1

The water level in the inlet tank was measured using the vernier point gage with an accuracy of 0.1 mm.

12

3.4.2 Bed and Water Levels Instrument carriage carrying a vernier point gage with an accuracy of 0.1 mm was used for the measurement of bed level and water surface level above the bed. The water surface level above the bed was adjusted by using tailgate at the downstream side of the flume. 3.4.3 Scour Depth The maximum scour depth near the pier for all the experimental runs were measured using a vernier point gage with an accuracy of 0.1 mm. 3.4.4 Velocity and Flow Field The Vectrino Velocimeter was used for the measurement of instantaneous three-dimensional component of velocity. The Vectrino velocimeter operated on a pulse-to-pulse coherent Doppler shift to provide instantaneous three-dimensional velocity components at a rate of 50 Hz. The acoustic sensor comprised with transmitting transducer and receiving transducers. The receiving transducers were mounted on short arms around the transmitting transducer at 120 0 azimuth intervals. The transmitting transducers emitted acoustic beams with a frequency of 10MHz. The beams travelling through the water arrived at the measuring point which is 5 cm below the transducer, where they were reflected by the ambient particles within the flow being received by the receiving transducers. The processing module performed the digital signal processing required to measure the Doppler shift. A real-time display of the data in graphical and tabular forms was provided by the data acquisition software. There was no requirement of seeding of the flow during experiments, as the signal-noise ratio (SNR) was in the range of 12 to 16. Because of the interference due to echoes from the flume bed, the receiving signal might be disturbed near the bed, which may result in inaccurate velocity measurement. The measurement by the Vectrino probe was not possible in the zone located 5 cm below the free surface. A special carriage structure was made to facilitate the movement of Vectrino Velocimeter along different radial line with respect to pier centreline. Sampling rate and sampling volume adopted for present experiment was 100 Hz and 2.5 cm3 respectively.

13

Perforated baffles Outlet Tailgate Sediment recess Sediment trap Pier model False floor Screens From inlet tank

0.177 m

0.5 m

0.85 m

3.5 m

(Dimension not in scale) Fig. 3.1 Schematic diagram of the experimental set-up

14

3.5 Experimental Procedure The experiments were carried out in four different parts. In the first part of the experiments non-cohesive sediment was used. 3.5.1 Non-cohesive sediments Initially, non-cohesive sediments of mean diameter 0.15 mm were used to fill the sediment recess around the pier model of 7.5 cm. The preparatory work for the experimental runs involved the following subtasks 1. Prior to the commencement of the experimental runs, pier model was placed in the middle of sediment recess. The bed was properly levelled using planner and final bed level was checked using a point gage. 2. On the upstream and the downstream side of the pier in the sediment recess, armor layer of desired thickness and 0.15 m wide was placed. 3. At the time of actual runs, in order to avoid the undesirable scour, which otherwise would happen by the action of sheet flow with inadequate flow depth, the flume was first slowly filled with the water by a pipe at a low rate at the downstream side. Once the water level of desirable height was reached, the experimental runs were started by adjusting the inflow rate and maintaining the required flow depth within a flume by a downstream gate. 4. The runs were taken for a maximum period of 2 hours, to ensure that the maximum scour depth was obtained. 5. To avoid the partial filling of scour hole by the sediments while draining out the water from the flume, the water was first drained out by opening a valve in the upstream end of the flume and adjusting the tailgate so that a minimum flow velocity occurred at the sediment recess. Finally, water was drained out very slowly by opening the holes at the bottom of downstream walls of the sediment recess, sediment trap and downstream gate. 6. For further runs in non-cohesive sediments 3.8 cm, 2 cm and 1 cm pier model was placed successively in the sediment recess and procedure from 2 to 5 was followed. 7. In all the experiments, two runs for 7.5 cm and 3.8 cm pier model were performed. Initial run carried out for depth measurement and second run conducted for velocity measurement. Both depth and velocity were measured

15

at the upstream of pier. For 2 cm and 1 cm pier model only depth measurement were taken at the upstream of the pier. 3.5.2 After 5 % mixing In the second and further part of the experiments two pier models (7.5 cm and 3.8 cm) were used. The non-cohesive sand was dried completely and clay soil of 5 % by weight was mixed in it thoroughly. Procedure from 1 to 5 was followed by using 5 % clay and sand mixture to measure the depth and velocity at the upstream side of the pier. At each run care was taken to maintain 5 % of clay soil in non-cohesive sediments. 3.5.3 After 10 % and 20 % mixing In the case of 10 % and 20 % of mixing, the non-cohesive sediment was dried completely and 10 % and 20 % of clay soil by weight was added thoroughly in noncohesive sediments, respectively. For both cases procedure from 1 to 5 above, were followed for measurement of depth and velocity at the upstream side of the pier. The care was also taken to maintain the required percentage of clay soil in respective run.

16

CHAPTER IV

RESULTS AND DISCUSSION

4.1 General The experiments were conducted in four different sets. In each set two runs for two pier models were performed. First run carried out for depth measurement and according to such measurement and initial bed condition velocity measurements took place in the second run. Velocity measurements were taken during the formation of scour hole. Both the depth and velocity were measured for a period of 2 hours to ensure that the maximum constant scour depth could be reached. Non-cohesive sediment of mean diameter 0.15 mm was used in initial set of experimental runs. Clay soil of percentage 5, 10 and 20 was thoroughly mixed in non-cohesive sediment for further sets of experimental runs. Data collected in each run was used to plot Time-Depth variation, Time-Velocity variation, Turbulent Intensity and Reynolds stresses discussed further in this chapter. 4.2 Time variation of scour depth 4.2.1 Scour for non-cohesive soil Depth measurements were taken as an initial measurement in each set of experimental runs to study the behaviour of scour at the upstream side of the pier model. For first set, non-cohesive sediment of mean diameter 0.15 mm was used for four different types of pier model, such as 1 cm, 2 cm, 3.8 cm and 7.5 cm. Time variation of scour depth for these pier models is shown in figures 4.1 - 4.4. These entire scour profiles showed that during the initial periods of scouring the pick-up rate was very high (for about 20 minutes) but it decreased and gradually become asymptotic to the time axis in the final periods. This was because of the horseshoe vortex. The particles at the base of the cylinder are removed due to fluid-induced forces under the combined effect of bed shear stress, turbulent agitation, and oscillation of the horseshoe vortex (Dey, 1996). At the initial periods of scouring, due to small dimensions of the scour hole the size of the horseshoe vortex was small. Consequently, the high bed shear stress developed beneath the vortex which caused rapid dislodgement of the sediment particles. Hence, there was rapid increment of profile in short period of time. As scour hole increased with time, the size of the horseshoe vortex also increased and therefore its strength decreased. Thus, the 17

bed shear stress induced by the vortex gradually decreased which resulted into the process of sediment pick-up to proceed at decreasing rate. Therefore, the profile changed its trend (after about 20 minutes) and gradually increased till the maximum constant scour depth could be reached. 4.2.2 Scour for mixture of clay and non-cohesive sediment The main cause of scour occur in clayey soil is due to different types of forces act between soil particles which resist the dislodgement of particles. These are Van der Waals forces, electric surface and other bonding mechanisms such as hydrogen bond, and chemical cementation between particles. Hence scour in clayey materials is more complex and less understood than the scour in non-cohesive sandy material (Garde et al., 1998). Therefore for next three sets of experimental runs different percentage of clay in non-cohesive sediment was used. Clay content of 5 %, 10 % and 20 % were mixed thoroughly in non-cohesive sediment and respective run carried out for depth measurement. The plot of depth variation with time for all these runs is shown in figures 4.3 4.10. Time-variation of scour depth for all these experimental run showed that initially sediment pick-up rate was very high but it decreased and gradually become asymptotic to the time axis in final periods as in case of non-cohesive sediment. The major difference between a non-cohesive and a cohesive sediment scour is that the erodibility for a fully consolidated, cohesive clay material is much less than that of sand (Hsu, FWRRC Annual Technical Report 2006). Thus, increment of clay content in noncohesive sediment caused less scour depth compared to full non-cohesive sediment. As shown in figures 4.8 4.10, although the initial rate of scouring was high for all runs the maximum depth attained for non-cohesive sediment was maximum as compared to the other runs which carried out by mixing different clay content given in the table 4.1 as well as the volume of scour hole around the pier model decreased with increased in clay content, shown in figures 4.11 4.18. Even run (not given in the table) were conducted for velocity measurement. For all the experimental runs discharge, depth over the bed and time for scouring process were kept constant. The same result was observed for both 7.5 cm and 3.8 cm pier model.

18

Table 4.1 Experimental data of obtaining maximum scour depth for different percentage of clay for different pier model Clay content in nonRun cohesive sediment (%) 1 3 5 7 9 11 13 15 0 0 5 5 10 10 20 20 Pier model diameter (cm) 7.5 3.8 7.5 3.8 7.5 3.8 7.5 3.8 Maximum scour depth attained after 2 hours (cm) 15.5 8.9 14.8 8.7 14.5 7.8 13.65 7.5

5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 0 20 40 60 80 100 120 140

Depth (cm)

Time (min)
Fig. 4.1 Time-depth variation for 1 cm pier model

19

5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 0 20 40 60 80 100 120 140

Depth (cm)

Time (min)
Fig. 4.2 Time-depth variation for 2 cm pier model

8 7 6

Depth (cm)

5 4 3 2 1 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.3 Time-depth variation for 3.8 cm pier model (before mixing of clay in non-cohesive sediment)

20

16 14 12

Depth (cm)

10 8 6 4 2 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.4 Time-depth variation for 7.5 cm pier model (before mixing of clay in non-cohesive sediment)

8 7 6

Depth (cm)

5 4 3 2 1 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.5 Time-depth variation for 3.8 cm pier model (after mixing 5 % of clay in non-cohesive sediment)

21

8 7 6

Depth (cm)

5 4 3 2 1 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.6 Time-depth variation for 3.8 cm pier model (after mixing 10 % of clay in non-cohesive sediment)

8 7 6

Depth (cm)

5 4 3 2 1 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.7 Time-depth variation for 3.8 cm pier model (after mixing 20 % of clay in non-cohesive sediment)

22

16 14 12

Depth (cm)

10 8 6 4 2 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.8 Time-depth variation for 7.5 cm pier model (after mixing 5 % of clay in non-cohesive sediment)

16 14 12

Depth (cm)

10 8 6 4 2 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.9 Time-depth variation for 7.5 cm pier model (after mixing 10 % of clay in non-cohesive sediment)

23

16 14 12

Depth (cm)

10 8 6 4 2 0 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.10 Time-depth variation for 7.5 cm pier model (after mixing 20 % of clay in non-cohesive sediment)

Fig. 4.11 Photograph of the scour hole for 3.8 cm pier model (before mixing of clay in non-cohesive sediment)

24

Fig. 4.12 Photograph of the scour hole for 3.8 cm pier model (after 5 % mixing of clay in non-cohesive sediment)

Fig. 4.13 Photograph of the scour hole for 3.8 cm pier model (after 10 % mixing of clay in non-cohesive sediment) 25

Fig. 4.14 Photograph of the scour hole for 3.8 cm pier model (after 20 % mixing of clay in non-cohesive sediment)

Fig. 4.15 Photograph of the scour hole for 7.5 cm pier model (before mixing of clay in non-cohesive sediment) 26

Fig. 4.16 Photograph of the scour hole for 7.5 cm pier model (after 5 % mixing of clay in non-cohesive sediment)

Fig. 4.17 Photograph of the scour hole for 7.5 cm pier model (after 10 % mixing of clay in non-cohesive sediment) 27

Fig. 4.18 Photograph of the scour hole for 7.5 cm pier model (after 20 % mixing of clay in non-cohesive sediment) 4.3 7.5 cm pier model 4.3.1 Time-Velocity variation Velocity measurements for all experimental runs were taken as secondary measurement during the formation of scour hole for both the pier models. It was observed earlier that, the obstruction of the flowing stream by a bridge pier caused a threedimensional separation of flow, as it travels by the side of the pier, forming a vortex flow field around the pier which moved downstream (Dey et al., 1995; Dey, 1995). Thus, streamwise velocity changed its direction near the bed. In the present experimental runs, the vertical distributions of time averaged streamwise velocity component for this pier model is plotted. Almost similar pattern was observed for all the profile depicted in the figures 4.19 4.22. Due to flow separation negative streamwise velocity was observed during the formation of scour hole upstream of the pier as well as almost constant negative velocity was observed during the formation of scour hole. Higher negative velocity was observed for such type of pier model.

28

In successive experimental runs, clay content was increased in non-cohesive sediment and velocity measurements were taken at the upstream side of the pier model. It was observed that, the volume of scour hole upstream of the pier decreased with the increment of clay content. Consequently, size of horseshoe vortex became small. Therefore, the flow velocity in the scour hole of sediment with higher clay content was lower than the lower or without clay content in non-cohesive sediment. Hence, the time averaged streamwise velocity profile exhibited higher constant negative velocity for a run performed after mixing 20 % of clay content in non-cohesive sediment than the run conducted after mixing 10 %, 5 % and 0 % of clay content in non-cohesive sediment in succession.

0.3 0.25 0.2

Velocity (m/s)

0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.19 Time-velocity variation for 7.5 cm pier model (before mixing of clay in non-cohesive sediment)

29

0.3 0.25 0.2

Velocity (m/s)

0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.20 Time-velocity variation for 7.5 cm pier model (after mixing 5 % of clay in non-cohesive sediment)

0.3 0.25 0.2

Velocity (m/s)

0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.21 Time-velocity variation for 7.5 cm pier model (after mixing 10 % of clay in non-cohesive sediment)

30

0.3 0.25 0.2

Velocity (m/s)

0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.22 Time-velocity variation for 7.5 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 4.3.2 Turbulent Intensity The vertical distribution of normalized streamwise turbulent intensity component u+ [= u 2

0.5

/ U ], where u is the fluctuation of U] at the upstream of pier model is

illustrated in the figures 4.23 4.26. It was observed that the streamwise turbulent intensity varies little in the zone for y+ < -0.15 having distribution more or less linear. On the other hand, in the zone for y+ > -0.15, where the reversal of flow occurred, normalized streamwise turbulent intensity component increases towards the scoured bed, but it reduced in the vicinity of the bed. A most significant feature of the distribution is the pronounced bulges immediately above y+ = -0.16 line, due to flow separation inside the scour hole. Near the scoured bed at the upstream side of pier model the spike was observed. These are due to the shuddering effect of the horseshoe vortex. The experimental runs conducted using different percentage of clay in non-cohesive sediment showed similar pattern of turbulent intensity component. The vertical distribution of normalized vertical turbulent intensity component w+ in the scour hole at the upstream side of the pier model is shown in the same figures 4.23 4.26. The distribution pattern of w+ is almost similar to that of u+. However, it is apparent that w+ do not show spike near the scour bed, as there is no shuddering effect of horseshoe vortex in the vertical direction. For all runs performed in different experimental conditions, same distribution pattern of w+ was observed. 31

0 0 -0.05 -0.1 -0.15

0.5

u+, w+

1.5

2.5

-0.2 -0.25 -0.3 -0.35 -0.4

u+ w+

Fig. 4.23 Vertical distribution of u+ and w+ at vertical section (before mixing of clay in non-cohesive sediment)

0 0 -0.05 -0.1 -0.15

0.5

u+, w+

1.5

2.5

-0.2 -0.25 -0.3 -0.35 -0.4

u+ w+

Fig. 4.24 Vertical distribution of u+ and w+ at vertical section (after 5 % mixing of clay in non-cohesive sediment)

32

0 0 -0.05 -0.1 -0.15

0.5

u+, w+

1.5

2.5

y+ -0.2
-0.25 -0.3 -0.35 -0.4

u+ w+

Fig. 4.25 Vertical distribution of u+ and w+ at vertical section (after 10 % mixing of clay in non-cohesive sediment)

0 0 -0.05 -0.1 -0.15

0.5

u+, w+

1.5

2.5

y+ -0.2
-0.25 -0.3 -0.35 -0.4

u+ w+

Fig. 4.26 Vertical distribution of u+ and w+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 4.3.3 Reynolds stresses Figures 4.27 4.30 represents the vertical distributions of normalized Reynolds stresses uw+ (= u w / u*2 ) at upstream section of pier model for different experimental condition. Reynolds stresses shows the distinguishable swell immediately below y+ = 0 line, inside the scour hole, as a result of flow separation. However, in the vicinity of the 33

scoured bed it reduces drastically. Also normalized Reynolds stresses changed its sign due to reversal of flow near the scoured bed. Similar pattern of observation was also made for all conditions of experimental runs.

uw
-20 0 -0.05 -0.1 -0.15 0 20 40

60

80

100

y+ -0.2
-0.25 -0.3 -0.35 -0.4

Fig. 4.27 Vertical distribution of uw+ at vertical section (before mixing of clay in non-cohesive sediment)

uw
-20 0 -0.05 -0.1 -0.15 0 20 40

60

80

100

y+ -0.2
-0.25 -0.3 -0.35 -0.4

Fig. 4.28 Vertical distribution of uw+ at vertical section (after 5 % mixing of clay in non-cohesive sediment)

34

uw+
-20 0 -0.05 -0.1 -0.15 0 20 40 60 80 100

y+ -0.2
-0.25 -0.3 -0.35 -0.4

Fig. 4.29 Vertical distribution of uw+ at vertical section (after 10 % mixing of clay in non-cohesive sediment)

uw+
-20 0 -0.05 -0.1 -0.15 0 20 40 60 80 100

y+ -0.2
-0.25 -0.3 -0.35 -0.4

Fig. 4.30 Vertical distribution of uw+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 4.4 3.8 cm pier model 4.4.1 Time-Velocity variation The vertical distribution of time averaged streamwise velocity component is shown in figures 4.31 4.34. Almost similar distribution pattern was observed for all the profile. The phenomenon of three-dimensional separation of flow caused by the bridge 35

pier which obstructs the flowing stream as it travels by the side of the pier formed a vortex flow field around the pier which moved downstream. Thus, streamwise velocity changed its direction near the bed. Due to flow separation negative streamwise velocity observed during the formation of scour hole upstream of the pier is shown in figures. Almost constant negative velocity was observed during the formation of scour hole. This is more apparent in the profile plotted for such pier model than the profile plotted for 7.5 cm pier model. For further experimental runs clay content was increased in non-cohesive sediment and velocity measurements were taken at the upstream side of the pier model. In this case also the volume of scour hole around the pier decreased with the increment of clay content. Consequently, size of horseshoe vortex became small. Therefore, the flow velocity in the scour hole of sediment with higher clay content was lower than the lower or without clay content in non-cohesive sediment. Thus, because of the flow, which was more contained in the scour hole the time averaged streamwise velocity profile illustrated in figure exhibited higher constant negative velocity for a run performed after mixing 20 % of clay content in non-cohesive sediment than the run conducted after mixing 10 %, 5 % and 0 % of clay content in non-cohesive sediment in succession.

0.3 0.25

Velocity (m/s)

0.2 0.15 0.1 0.05 0 -0.05 -0.1 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.31 Time-velocity variation for 3.8 cm pier model (before mixing of clay in non-cohesive sediment)

36

0.3 0.25

Velocity (m/s)

0.2 0.15 0.1 0.05 0

-0.05 -0.1 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.32 Time-velocity variation for 3.8 cm pier model (after mixing 5 % of clay in non-cohesive sediment)

0.35 0.3 0.25

Velocity (m/s)

0.2 0.15 0.1 0.05 0

-0.05 -0.1 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.33 Time-velocity variation for 3.8 cm pier model (after mixing 10 % of clay in non-cohesive sediment)

37

0.3 0.25

Velocity (m/s)

0.2 0.15 0.1 0.05 0

-0.05 -0.1 0 20 40 60 80 100 120 140

Time (min)
Fig. 4.34 Time-velocity variation for 3.8 cm pier model (after mixing 20 % of clay in non-cohesive sediment) 4.4.2 Turbulent Intensity The vertical distribution of normalized streamwise turbulent intensity component u+ [= u 2

0.5

/ U ], where u is the fluctuation of U] at the upstream side of the pier

model depicted in figures 4.35 4.38. Due to the reversal of flow occur for the zone y+ > -0.04, normalized streamwise turbulent intensity component u+ increases towards the scoured bed and reduced in the vicinity of the bed. A distinguishable feature of pronounced bulges was observed above y+ > -0.04 line, due to separation of flow inside the scour hole. Because of the shuddering effect of the horseshoe vortices less distinct spike as compared to 7.5 cm pier model was observed near the scour bed at the upstream side of the 3.8 cm pier model. Same vertical distribution of normalized streamwise turbulent intensity component u+ and pronounced bulges were observed for runs conducted using different percentage of clay in non-cohesive sediment. The vertical distribution of normalized vertical turbulent intensity component w+ in the scour hole at the upstream side of the pier model is also illustrated in the same figures 4.35 4.38. As in this case, there was no shuddering effect of horseshoe vortex in the vertical direction no spike was observed near the scour bed for w+ but this pier model also depicted almost similar distribution pattern of w+ to that u+. Every run conducted in different experimental condition containing different percentage of clay content in noncohesive sediment for such pier model shows similar result. 38

0 0

0.5

u+, w+

1.5

-0.05

-0.1

u+ w+

-0.15

-0.2

Fig. 4.35 Vertical distribution of u+ and w+ at vertical section (before mixing of clay in non-cohesive sediment)

0 0

0.5

u+, w+

1.5

-0.05

y+ -0.1
-0.15

u+ w+

-0.2

Fig. 4.36 Vertical distribution of u+ and w+ at vertical section (after 5 % mixing of clay in non-cohesive sediment)

39

0 0

0.5

u+, w+

1.5

-0.05

y+ -0.1
-0.15

u+ w+

-0.2

Fig. 4.37 Vertical distribution of u+ and w+ at vertical section (after 10 % mixing of clay in non-cohesive sediment)

0 0

0.5

u+, w+

1.5

-0.05

-0.1

u+ w+

-0.15

-0.2

Fig. 4.38 Vertical distribution of u+ and w+ at vertical section (after 20 % mixing of clay in non-cohesive sediment) 4.4.3 Reynolds stresses Figures 4.39 4.42 exhibits the vertical distributions of normalized Reynolds stresses at the upstream section of the pier. In this case also, inside the scour hole, distinguishable bulges was observed immediately below y+ = 0 line, as a result of flow separation. In the vicinity of the scoured bed Reynolds stresses changed its sign due to 40

reversal nature of flow and became small near the scoured bed. Similar distribution pattern was observed for all experimental runs performed in different experimental conditions.

uw
-20 0 0 20 40

60

80

100

-0.05

y+ -0.1
-0.15

-0.2

Fig. 4.39 Vertical distribution of uw+ at vertical section (before mixing of clay in non-cohesive sediment)

uw
-20 0 0 20 40

60

80

100

-0.05

y+ -0.1
-0.15

-0.2

Fig. 4.40 Vertical distribution of uw+ at vertical section (after 5 % mixing of clay in non-cohesive sediment)

41

uw
-20 0 0 20 40

60

80

100

-0.05

y+ -0.1
-0.15

-0.2

Fig. 4.41 Vertical distribution of uw+ at vertical section (after 10 % mixing of clay in non-cohesive sediment)

uw
-20 0 0 20 40

60

80

100

-0.05

y+ -0.1
-0.15

-0.2

Fig. 4.42 Vertical distribution of uw+ at vertical section (after 20 % mixing of clay in non-cohesive sediment)

42

CHAPTER V

SUMMARY AND CONCLUSIONS

The experiments were conducted in four different sets using different percentage of clay content in non-cohesive sand such as 20 %, 10 %, 5 % and without clay (0 %). Depth and velocity measurements were taken at the upstream side of the pier models by using Acoustic Doppler Velocimeter (ADV) in the laboratory flume. Measurements were taken in the vicinity of the pier model for a maximum period of two hours. Four pier size of diameter 7.5 cm, 3.8 cm, 2 cm and 1 cm were considered for depth measurement in initial set of experimental runs containing non-cohesive sand of mean diameter 0.15 mm. In such bed condition, velocity measurements were performed for 7.5 cm and 3.8 cm pier model. For further sets of experimental runs, thoroughly mixed clay content of 5 %, 10 % and 20 % in non- cohesive sand were used for depth and velocity measurements in the vicinity of 7.5 cm and 3.8 cm pier model. Data captured in each run was used for the development of relationship between time and depth and time and velocity as well as turbulent intensity in streamwise and vertical direction and Reynolds stresses were plotted, which leads to the following conclusions: 1. The time scale required to attain the maximum constant depth is important parameter. The time required to attain maximum constant depth in noncohesive sand is less compared to the bed containing even small clay content in non-cohesive sand. Therefore, low maximum constant depth was obtained due to increment of clay content in non-cohesive sand compared to full noncohesive sand bed condition. 2. The state of equilibrium provides the most important step toward simplifying the erosion problem from an engineering point of view because the maximum equilibrium scour can be estimated as the most conservative design. Thus, predicting the maximum constant scour is the most fundamental step to study a scour problem. 3. The volume of scour hole at the upstream of the pier model was decreased with increase in clay content. As the flow is more contained in the scour hole, the flow velocity in the scour hole of non-cohesive sand with higher clay content was lower than the lower or without clay content. 43

4. The vertical distribution of normalized streamwise turbulent intensity component and Reynolds stresses for both type of pier model (7.5 cm and 3.8 cm) showed distinguishable features of the pronounced bulges and were reduced in the vicinity of the scoured bed. 5. The distribution pattern of vertical velocity component w+ is almost similar to that of streamwise velocity component u+ except the spike near the scour bed, as there was no shuddering effect of primary vortex in the vertical direction. 6. The data captured for different conditions of experimental runs would be useful for the development of mathematical models of flow field in a scour hole at bridge pier. Thus, the accurate estimation of scour depth would be possible in different bed conditions using flow field model. 7. Using similitude modelling (geometric/ dynamic/ kinematic) obtained results can be extended for real world situation.

44

CHAPTER VI

REFERENCES
Anderson, A. G. (1963). Discussion of Sediment Transportation Mechanics: Erosion of Sediment, by the Task Force on Preparation of Sedimentation Manual. J. Hydraul. Div., ASCE, 89(6), 237-248. Ansari, S. A., Kothyari, U. C., and Ranga Raju, K. G. (2002). Influence of cohesion on scour around bridge piers. J. Hydraul. Res., 40(6), 717-729. Baker, R. E. (1986). Local scour at bridge piers in nonuniform sediment. Rep. No. 402, School of Engineering, University of Auckland, Auckland, New Zealand. Barbhuiya A. K. (2003). Clear water scour at abutments. PhD thesis, Indian Institute of Technology, Kharagpur, India. Barbhuiya A. K., and Dey, S. (2004a). Measurement of turbulent flow field at a vertical semicircular cylinder attached to the sidewall of a rectangular channel. J. Flow meas. Instru., 15, 87-96. Barbhuiya A. K., and Dey, S. (2004b). Turbulent flow measurement by the ADV in the vicinityof a rectangular cross-section cylinder placed at a channel sidewall. J. Flow meas. Instru., 15, 221-237. Breusers, H. N. C., Nicollet, G., and Shen, H. W. (1977). Local scour around cylindrical piers. J. Hydraul. Res., 15(3), 211-252. Briaud, J. L., Ting, C. K., Chen, H. C., Rao, G., Perugu, S., and Wei, G. (1999). SRICOS: Prediction of scour rate in cohesive soils at bridge piers. J. Geotech. Env. Eng., 125 (4), 237246. Chiew, Y. M. (1984). Local scour at bridge piers. PhD thesis, University of Auckland, Auckland, New Zealand. Chabert, J., and Engeldinger, P. (1956). Etude des affouillements autour des piles de ponts. Serie A. Laboratoire National dHydraulique. Chatou, France (in French). Davis, S. R. (1984). Case Histories of Scour Problems at Bridges. Transportation Research Record 950, Second Bridge Engineering Conference (2), Transportation Research Board, Washington D. C. 45

Dey, S. (1995). Three-dimensional vortex flow field around a circular cylinder in a quasi-equilibrium scour hole. Sadhana, Proc., Indian Acad. Sci., 20, 771-785. Dey, S. (1996). "Sediment pick-up for evolving scour near circular cylinders." Appl. Math. Modelling, 20(7), 534-539. Dey, S. (1999). Time variation of scour in the vicinity of circular piers. Proc., Inst. Civ. Eng., J. Wat. Maritime Eng., 136, 67-75. Dey, S., and Barbhuiya A. K. (2004). Clear water scour at abutments. J. Wat. Management, Proc., Inst. Civ. Eng., 157, 77-97. Dey, S., and Barbhuiya A. K. (2005). Turbulent flow field in a scour hole at a semicircular abutment. Can. J. Civ. Eng., 32, 213232. Dey, S., Bose, S. K., and Sastry, G. L. N. (1995). Clear water scour at circular piers: a model. J. Hydraul. Eng., ASCE, 121(12), 869-876. Dongol, D. M. S. (1994). Local scour at bridge abutments. Rep. No. 544, School of Engineering, University of Auckland, Auckland, New Zealand. Garde, R. J., and Kothyari, U. C. (1998). Scour around bridge piers. PINSA, 64 A (4), 569-580. Horst, M. W. (2004). Time variation of shear stress measurements at the base of rectangular piers. PhD thesis, The Pennsylvania State University, Pennsylvania. Hsu, T. J. (2006). Measurement of erosion around hyrdaulic structures. Annual Technical Report, Florida Water Resources Research Center, Civ. Coast. Eng., University of Florida. Jones, S. (1993). Preliminary Studies on Pressure Flow Scour. Proc. 1993 National Conference, J. Hydraul. Eng., ASCE, 916-921. Jones, S. (2002). Indirect Methods of Measuring Shear Stress in the Bottom of a Scour Hole. Conference Proceedings, First International Conference on Scour of Foundations, 140-155. Kamphuis, J. W., and Hall, K. R. (1983). Cohesive material erosion by unidirectional current. J. Hydraul. Eng., 109 (1), 5061. Kandasamy, J. K. (1989). Abutment scour. Rep. No. 458, School of Engineering, University of Auckland, Auckland, New Zealand. 46

Kho, K. T., Valentine, E., and Glendinning, S. Unpublished paper. An experimental study of local scour around circular bridge piers in cohesive soils. Kothyari, U. C., Garde, R. J., and Ranga Raju, K. G. (1992a). Temporal variation of scour around circular bridge piers. J. Hydraul. Eng., ASCE, 118(8), 1091-1106. Kothyari, U. C., Garde, R. J., and Ranga Raju, K. G. (1992b). Live-bed scour around cylindrical bridge piers. J. Hydraul. Res., IAHR, 30(5), 701-715. Kothyari, U. C., and Ranga Raju, K. G. (2001). Scour around spur-dikes and bridge abutments. J. Hydraul. Res., IAHR, 39(4), 367-374. Laursen, E. M. (1952). Observations on the nature of scour. Proc., 5 Hydraul. Conf., State Univ. of Iowa, Iowa city, Iowa, 179-197. Laursen, E. M., and Toch, A. (1956). Scour around bridge piers and abutments. Bull. No. 4, Iowa Highways Research Board, Ames, Iowa. Melville, B. W. (1975). Local scour at bridge sites. Rep. No. 117, School of Engineering, University of Auckland, Auckland, New Zealand. Melville, B. W. (1992). Local scour at bridge abutments. J. Hydraul. Eng., ASCE, 118(4), 615-631. Melville, B. W., and Sutherland, A. J. (1988). Design method for local scour at bridge piers. J. Hydraul. Eng., ASCE, 114(10), 1210-1226. Molinas, A., Hosny, M., and Jones, S. (1998). Pier scour in montmorillonite clay soils. J. Wat. Res. Eng., (1), 292-297. Molinas, A., Reiad, N. G. Y., and Jones, S. (1998). Effect of cohesion on abutment scour. J. Wat. Res. Eng., (1), 252-257. Partheniades, E. (1965). Erosion and Deposition of Cohesive Soils. J. Hydraul. Div., ASCE, Proc. Paper 4204, Jan., 105-138. Rambabu, M., Narasimha Rao S., and Sundar V. (2003). Current-induced scour around a vertical pile in cohesive soil. J. Ocean Eng., 30, 893-920. Richardson, E. V., Harrison, L. J., and Davis, S. R. (1995). Evaluating Scour at Bridges. HEC-18, Report FHWA-IP-90-017, Washington D. C. Rouse, H. (1965). Engineering hydraulics: sediment transportation. John Wiley and sons, New York, NY. 47