Front Cover
 Report documentation page
 Funding information
 Title Page
 Executive summary
 Appendix A: Sediment scour...
 Appendix B: Local structure-induced...
 Appendix C: Computational approaches...
 Appendix D: Structure-induced scour...
 Appendix E: Global sediment...

Group Title: Technical report – University of Florida. Coastal and Oceanographic Engineering Program ; 84
Title: Structure-induced sediment scour, local scour bounds & assessment of global scour
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00075470/00001
 Material Information
Title: Structure-induced sediment scour, local scour bounds & assessment of global scour final report
Series Title: UFLCOEL-TR
Physical Description: 1 v. (various pagings) : ill., charts ; 28 cm.
Language: English
Creator: Sheppard, D. M
Niedoroda, Alan W
University of Florida -- Coastal and Oceanographic Engineering Dept
Publisher: Dept. of Coastal and Oceanographic Engineering, University of Florida
Place of Publication: Gainesville Fla
Publication Date: 1992
Subject: Scour (Hydraulic engineering)   ( lcsh )
Scour at bridges   ( lcsh )
Erosion -- Florida   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references.
Statement of Responsibility: principal investigator: D.M. Sheppard ; co-principal investigator: A.W. Niedoroda.
General Note: "February, 1992.
General Note: "UF Project No. 451123412."
General Note: "UPN No. 90100412."
General Note: "Contract No. C-6852."
Funding: This publication is being made available as part of the report series written by the faculty, staff, and students of the Coastal and Oceanographic Program of the Department of Civil and Coastal Engineering.
 Record Information
Bibliographic ID: UF00075470
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida
Resource Identifier: oclc - 26521905

Table of Contents
    Front Cover
        Front Cover
    Report documentation page
        Unnumbered ( 2 )
    Funding information
        Unnumbered ( 3 )
    Title Page
        Title Page
    Executive summary
        Unnumbered ( 5 )
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
    Appendix A: Sediment scour bibliiography
        Appendix 1
        A 1
        A 2
        A 3
    Appendix B: Local structure-induced scour experiments
        A 4
        B 1
        B 2
        B 3
        B 4
        B 5
        B 6
        B 7
        B 8
        B 9
        B 10
        B 11
        B 12
        B 13
        B 14
        B 15
        B 16
        B 17
        B 18
        B 19
        B 20
        B 21
        B 22
        B 23
        B 24
        B 25
        B 26
        B 27
        B 28
        B 29
        B 30
        B 31
        B 32
        B 33
        B 34
        B 35
        B 36
        B 37
        B 38
        B 39
        B 40
        B 41
        B 42
        B 43
        B 44
        B 45
        B 46
        B 47
    Appendix C: Computational approaches to local structure-induced scour prediction
        C 1
        C 2
        C 3
        C 4
        C 5
    Appendix D: Structure-induced scour near slab-like structures due to steady currents
        D 1
        D 2
        D 3
        D 4
    Appendix E: Global sediment scour
        E 1
        E 2
        E 3
        E 4
        E 5
        E 6
        E 7
        E 8
        E 9
Full Text





D. M. Sheppard
A.W. Niedoroda

February 1992


U.S. Department of Commerce/NOAA
Department of Environmental Regulation

OM8 No. 0704-0188

ga1tl..'. Swl. l 4m.n rnt 0 A ne04d. 4 CO0.lnq 140 ..n q gn ,olM l.Oft Of M"aluOn. .40 CO0mfntt radnrQ thn o(ac1 m)ttt..te or *.. 01V M.wt of 1"I
oltn of mfo0ma .lson d 1ng 9uggetrslt l0o0tfq t"a m OrVo to W`neoqton .a4d0wne. Se(r0e. rMnO.rate or Informan 010oft aon f Reog 121I Jefferlon
OoofotM". 0toe 2l04. Arknqton. VA 22202-410. an to tw OtBLreo o f aqnget.q Pawor0 aeotactton PrOec (0704-188U). wenqton. OC 20501.

February 1992 Final Renort

Structure-Induced Sediment Scour Local Scour
Bounds and Assessment of Global Scour-6852
D.M. Sheppard
A.W. Niedoroda
*University of Florida
Gainesville, Florida 32611 PR 4910451123412
*Environmental Science & Engineering Inc. UPN No. 90100412
Gainesville, Florida
U.S. Dept. of Commerce/NOAA Dept. of Env. Reg. AGENCY REPORT NUMBER
OCRM Coastal Management
1825 Connecticut A., N.W. 2600 Blair Stone Rd.
Washington, D.C. 20235 Tallahassee, FL 32399


13. ABSTRACT (Maximum 200 words)

Live bed structure-induced sediment scour experiments were
conducted to help establish upper bounds on scour depths with regard
to depth mean velocity. The structure examined was a 4 inch diameter
vertical cylinder and the D50 for the sediment was 0.75 mm. Limited
experiments were also conducted on other shapes including multiple
vertical pile structures.

Structure-induced sediment scour / Live bed scour


Standard Form 298 (Rev. 2-89)
Ptrwesn by ANSI S 39.1

SN N 7540-01480.5500

Partial funds for this project were provided by the Department of Environmental Reg-
ulation, Office of Coastal Management using funds made available through the National
Oceanographic and Atmospheric Administration under the Coastal Zone Management Act
of 1972, as amended.

Final Report

Structure-Induced Sediment Scour
Local Scour Bounds & Assessment of Global Scour

Principal Investigator: D.M. Sheppard
Department of Coastal and Oceanographic Engineering
University of Florida

Co-Principal Investigator: A.W. Niedoroda
Environmental Science and Engineering Inc.

UF Project No. 451123412
UPN No. 90100412
Contract No. C 6852

Department of Coastal and Oceanographic Engineering
University of Florida
Gainesville, Florida 32611

February, 1992

Executive Summary

This report presents the results of an investigation into the depth limits of structure-
induced sediment scour with regard to depth mean water velocity. Live bed scour experi-
ments were performed on a vertical cylinder and the results compared with clear water scour
results of another investigator. The results clearly show a decreased equilibrium (structure-
induced) scour depth for velocities above the critical velocity (i.e. the depth mean velocity
needed to initiate sediment motion on the plain bed). However, at velocities above the crit-
ical value, sand waves (called dunes and antidunes) can and do form and migrate through
the scour hole. The total live bed scour depth (structure-induced plus sand dune amplitude)
can be more or less than the maximum clear water structure-induced scour depending on
the sediment size. Maximum structure-induced scour depths are believed to occur just prior
to transition from clear water to live bed conditions. The experiments were conducted in a
large tilting flume located in the Hydraulics Laboratory in the Engineering Research Center
at Colorado State University in Ft. Collins, Colorado. An instrument for in situ measure-
ments of scour was developed as part of this research. This instrument allowed the rate of
scour in addition to the scour depth to be measured.

Global (or dishpan) scour was also investigated from the standpoint of determining the
feasibility of studying this type of scour in a laboratory flume. Two multiple pile structures
were subjected to live bed scour conditions and the resulting scour measured. The feasibility
of such studies was verified. Local scour near a large rectangular structure (model swimming
pool) was also investigated.

The results of these tests along with the results of other investigators were incorporated
into the computer program, SCOUR, developed for the Florida Department of Natural Re-
sources under a previous contract.

A three dimensional flow solver (INS3D) was coupled with a sediment transport model
to determine the practicality of three dimensional structure-induced scour predictions with
present day algorithms and computers. It was concluded that additional basic research is
needed in both the numerical algorithms and in the turbulence models to be used for this
complex flow and sediment transport problem before this approach can be used for local
scour prediction.



When coastal structures are inundated by storm waters they often experience sediment
scour around their foundations. An understanding of these processes to the extent of being
able to estimate the depth and volume of sediment lost during an extreme storm event is of
utmost importance to the Department of Natural Resources Division of Beaches and
Shores. A study sponsored by the Florida Department of Natural Resources and University
of Florida completed in December 1990 addressed many aspects of the structure-induced
scour problem. A methodology for computing scour was developed in that study and the
results were reported in the final report to FDNR (Sheppard and Niedoroda 1990) and in
an international conference proceedings (Sheppard et al. 1990). In addition, a computer
program for computing local and global (dishpan) scour for a variety of structure shapes
was developed. In spite of the progress made in that study, there were important problems
remaining that needed to be addressed. Many of these problems were the subject of this
past year's work. At the heart of this effort was the question of what current velocity
should be used in the local scour calculations for coastal structures inundated by storm
surge. The water currents under extreme storm conditions are complex and difficult to
predict, yet the scour depth and volume are dependent on, among other things, the depth
average velocity. Analysis of the experimental results from this study and further analysis
of other investigators' research has produced information that should prove most useful to
FDNR's rather unique structure-induced scour prediction needs.
Another important topic treated in this study is that of scour near multiple pile structures.
For these structures, the scour is composed of two components: a local scour and a global
(or dishpan) scour. The local scour is similar to that experienced by a single element
structure while the global scour is a dishpan shaped scour hole that encompasses the
composite structure. This scour is of particular importance due to the large scour volumes
involved and the large number of multiple pile structures that exist in the coastal zone.
The remainder of the study dealt with 1) seeking better ways to compute local scour near
and under horizontal slab-like structures (such as would occur when a slab foundation is
undermined) subjected to currents and 2) testing the feasibility of modifying an existing
3-D computational fluid mechanics computer code to include a sediment transport model
for use in local scour computations. The level of effort on these topics was small in
comparison to those items described above.
The local sediment scour program developed in the earlier study has undergone some major
modifications based on the findings of this study. In addition to changes in the technical
content there have been enhancements to some of the help screens (addition of graphics).

All aspects of the study will be discussed in the body of the report but details are left for
the appendices.

Structure-Induced Scour Processes

Sediment erosion problems, in general, are complex and do not lend themselves well to
analytical treatment. This is particularly true when solid structures are involved since the
presence of the structure is an added source of turbulence, vortices and an enhancement to
the three dimensionality of the flow. As a result of this, most research in this field has been
experimental, and most of the scour prediction equations, empirical. Much of the work
done in this field has been directed at scour around bridge piers with a strong emphasis on
bridges over inland waterways and rivers. Under these circumstances the flow is constant in
direction and relatively constant in speed (as compared to tidally influenced waters). Thus,
with few exceptions, the water current responsible for causing local structure-induced scour
is of such duration that equilibrium scour depths will be achieved. This, along with the
fact that time dependency scour measurements have been extremely time consuming and
labor intensive in the past accounts for the lack of data reported in the literature. In
tidally influenced waters subject to rapidly changing hurricane storm surge, the rate of
scour becomes important. More work is needed in this area.
Researchers have made scour measurements for flow around vertical cylinders for a number
of years. When the depth average flow velocity exceeds the so called critical value,
transition from "clear-water" to "live-bed" conditions occurs, i.e. at this velocity the
median diameter sediment (away from the structure) starts to move. Once live-bed flow
conditions are reached sand waves (referred to as sand dunes in the literature) can be
generated. These waves can migrate in the direction of the flow or opposite to the flow
depending on the environmental conditions, water depth, flow velocity, sediment size, etc.
These dunes have been observed and discussed by several researchers conducting live-bed
local scour measurements (see e.g. Shen et al. 1966 and Jain and Fischer 1979). These
sand waves, which can have significant amplitudes, alter the flow near the bottom and in
the neighborhood of the cylinder. When the trough of the sand wave propagates past the
cylinder the observed scour depth is a combination of local structure-induced scour and
sand wave trough amplitude. Both processes are nonlinear so the total depth will most
likely not be the simple superposition of the two. If this nonlinear effect is similar to
surface wave interaction phenomenon, a simple superposition of the two depths will be
greater than the actual observed value. That is, the sum of the two depths would result in
a conservative or slightly over prediction of the scour depth. To the knowledge of the
authors, Jain and Fischer (1979) and Melville (1984) are the only investigators that have
separated the dune wave amplitude from the local scour depth prior to this study. Thus,
much of the scour values for live-bed conditions reported in the literature most likely are a
combination of dune amplitude and local scour.

It is the authors' contention that these two phenomena should not be combined, at least
from a scour prediction point of view. They are different processes governed by different
mechanisms and should be computed separately. Fortunately, there is a reasonable body of
data and knowledge on dune waves in the literature and predictive equations exist for both
their height and length as a function of the flow and sediment parameters. The question is
then, what happens to the local structure-induced scour above the transition from
clear-water to live-bed flow conditions? Since much of the data reported in the literature is
most likely "contaminated" by dune waves, only the data obtained by Jain and Fischer
(1979), Melville (1984) and in the tests conducted as part of this year's work are available
for analysis.
Based on the results of laboratory tests conducted by Chabert and Engeldinger in France
in 1956 and others, researchers concluded that the maximum structure-induced scour depth
occurs just prior to the transition from clear-water to live-bed conditions. Shen et al.
(1966) state that "the data of Chabert and Engeldinger show that for a given cylinder, as
velocity increases the scour depth reaches a peak value then drops off slightly and is
essentially independent of increasing velocity thereafter". Based on this "understanding"
most laboratory experiments were conducted for clear-water conditions. More recent
studies, however, indicate that this is not always the case. Melville (1984) conducted
live-bed tests for a range of cylinder and sediment diameters and found that for sediment
diameters greater than a 0.6 mm (i.e. non ripple forming sediments) the maximum total
local scour depth occurs at (or near) the transition from clear-water to live-bed conditions.
However, for sediment diameters less than a 0.6 mm (ripple forming sediments) the
maximum total local scour depth was found to occur at higher velocities. The total local
scour being the combination of "local structure-induced scour" and the apparent scour due
to the passage of a sand dune. In either case it appears that the portion of the scour due to
the structure reaches a maximum at the transition from clear-water to live-bed and
decreases as the velocity increases beyond the critical value. Dunes or sand waves have
been studied for a number of years and are reasonably well understood. Both experimental
data and empirical equations for predicting wave heights and lengths in terms of water
depth and flow and sediment conditions are reported in the literature (see e.g. van Rijn
1984, Raudkivi and Witte 1990, Haque and Mahmood 1986, Sumer and Bakioglu 1984).
The scour tests conducted as part of this study are described in Appendix B.
Dimensionless scour depth versus time for the three tests conducted are plotted in Figures
B25a, b and c along with the results of a test from Hannah (1978). Hannah's sediment was
almost identical in size and mass density to that in this study and his aspect ratios
(structure height to diameter ratio) were in the same range. The tests in this study were
all for live-bed conditions while Hannah's test was for clear-water flow. The live-bed test
data show variations of scour depth with time (see Figure B25a). These variations are real
and are attributed to small ripples migrating through the test area, turbulent burst
phenomenon causing avalanches in the scour hole, and variations in the flume discharge
(especially in Run C). Based on the limited available information on rates of scour, the

live-bed tests in this study should have reached equilibrium long before the clear-water test
of Hannah. Note that the structure-induced scour depths for all three live-bed tests are less
than that for Hannah's clear-water case. This suggests that the maximum local
structure-induced scour occurs near the transition from clear-water to live-bed flow
conditions for the sediment and flow conditions considered in these tests. This is consistent
with and adds strength to the findings of Melville (1984).
As a result of this study, it appears that there may be a better way to compute the local
scour depth and volume than has been done in the past. By separating the
structure-induced scour from that resulting from the formation and passage of sand dunes
several advantages are possible. These include possibly increasing the accuracy of the scour
prediction by expressing each of the two phenomenon in terms of only those parameters for
which they depend, e.g. the dune parameters are, for the most part, independent of the
structure parameters. Such an analysis would also allow for the computation of that
portion of the total scour depth and volume that is due only to the presence of the
structure. Since the flow velocities required to produce this condition (transition from
clear-water to live-bed condition) has a high probability of being achieved any time the
structure is submerged due to storm surge, the flow velocity needed for this computation
can be computed by the SCOUR PROGRAM relieving the program user of this task.
FDNR coastal engineers will, of course, also need estimates of the total local scour in order
to determine if the structure under consideration will fail during a particular storm. The
total local scour can be estimated by the SCOUR PROGRAM by superimposing the
structure induced component and that due to the passage of a sand wave.
This represents a new approach to analyzing the local scour problem and thus the concept
as well as the computer code needs to be tested. Two versions of the SCOUR PROGRAM
are presented with this report. Version 1.5 uses the approach taken in an earlier version
produced in the 1990 study. It includes enhancements in the number of data points used in
the four dimensional curve fit scheme for local scour as well as a complete rewrite of the
global (or dishpan) scour prediction algorithm. Version 2.0 uses the approach described
above and should be considered "developmental" at this point. It should be tested against
new field and laboratory data as it becomes available. In addition, the program's results
should be tested against its reasonableness based on experience and good engineering

Structure-Induced Scour Experiments

As stated in the introduction, one of the objectives of this year's work was to examine the
"maximum scour with respect to current velocity hypothesis". A series of live-bed scour
experiments were planed to provide information that would help in this assessment.
Several flumes were considered for these tests, but due to the high flow discharge
requirements the choice soon narrowed to a few with the best for the job being one in the
Hydraulics Laboratory in the Engineering Research Center at Colorado State University.

Live-bed local scour tests were performed on a vertical cylinder subjected to steady
currents. In situ scour (depth and volume) measurements were made using an instrument
developed in the Coastal Engineering Laboratory at the University of Florida.
Since the flume was large enough to accommodate several models at the same time the
authors took this opportunity to examine the scour around three other structural shapes of
interest to FDNR engineers. These shapes were that of a rectangular swimming pool, a
horizontal slab and (two different) multiple pile structures. Use of the flume was limited to
a three week period (for setup, calibration, tests, and cleanup) so the amount of data and
the parameters that could be changed was very limited but nevertheless the results proved
to be interesting and useful. The data obtained from these tests were used to check the
effectiveness of the SCOUR PROGRAM in estimating scour for these structural shapes.
These experiments not only produced valuable data but also provided further insight into
the local scour processes. This insight, eventually led to some of the conclusions discussed
above. All of the results have not been exploited to their fullest but the data has been
reduced and when time and funds are available this can be pursued. Details of the
experiments including apparatus and instrumentation used, procedures followed, and the
results are presented in Appendix B.

Global Scour

Another aspect of this year's work has to do with scour near multi-element or multiple pile
structures. This structure configuration results in two types of scour, one due to the
individual piles (local scour) and one due to the overall structure. The overall structure
scour is called global or dishpan scour. When the piles are in close proximity to each other,
a/D < 3.0, where a is the distance between pile centers and D the pile diameter, the
structure can be treated as a single, porous structure, Jones (1989). Dishpan scour does
not exist for this case. For 3 < a/D < 11 both dishpan and local are present. Once again
very little data has been reported in the literature on this subject. The report by Hannah
contains data for two, three, four and six pile structures for a limited range of conditions. A
procedure was developed in this study that allows the results to be used over a wider range
of environmental parameters. The data taken from Hannah and a description of how the
data was processed and implemented into the scour program is presented in Appendix E.

Scour Near Horizontal Structures

In the previous study a potential flow computer model was developed for use in estimating
sediment scour near a horizontal cylinder. One of this year's tasks was to expand this
model to include horizontal slabs in the presence of steady currents. Preliminary results on
this task showed that the limitations of the potential flow model were being exceeded for
this geometry and that the results were going to be of limited value. At this point it was

decided to determine if the two-dimensional canopy wind flow computer model being
developed for another FDNR project could be modified and used for this situation. The
initial results look promising but some additional work is needed before the output can be
parameterized and incorporated into the SCOUR PROGRAM. A brief discussion of the
potential flow and the two-dimensional second order closure turbulence models is presented
in Appendix D.

Three-Dimensional Flow and Scour Computations

The final item to be discussed is the study that was made to determine the feasibility of
adding a sediment transport (scour) model to an existing three-dimensional computational
fluid mechanics flow model for the purpose of computing local structure-induced sediment
scour. The results of this work are presented in Appendix C. A three-dimensional model
called INS3D was obtained from COSMIC, University of Georgia, Athens, Georgia, the
clearing house for NASA developed software. The model was already coded for steady flow
around a vertical cylinder which saved a substantial amount of time. A scour program was
developed and appended to this three-dimensional flow solver. As presently configured the
model solves the laminar Navier-Stokes equations, and systematically adjusts the bottom
until the bottom shear stress is uniform over the entire region. Even though the flow is
laminar and has a low Reynolds Number the scour hole pattern is surprisingly similar to
those observed in the laboratory and in the field. Certain scour producing mechanisms
such as the horseshoe vortex are produced by this model. Other important mechanisms
such as the turbulence produced by the bottom and the structure can not be modeled with
the present form of this model. Adding a meaningful turbulence model to this program will
require a significant investment of time and computer resources. The present code requires
approximately three hours of CPU time on the University of Florida IBM computer for a
typical run. Since this would have been cost prohibitive for the project, free computer time
was requested from Research Computing Initiative at the Northeast Regional Data Center
and approximately 15 hours per week for 32 weeks was granted. The authors are grateful
to The Northeast Regional Data Center for this support. A discussion of the INS3D and
scour model model along with some of the results are given in Appendix C.

Summary and Conclusions

The 1990 study resulted in a methodology and a computer program for estimating local
scour depths and volumes for a number of structure shapes and flow and sediment
conditions. The difficulty in applying this program came as a result of the lack of
knowledge regarding the flow velocity to use as an input to the program. It was
hypothesized that for a given structure and sediment there is a maximum local scour depth
with regard to depth mean velocity, i.e. a velocity beyond which the scour does not
increase. A series of high velocity (live-bed) experiments were designed to provide data for

assessing this hypothesis. The apparatus and necessary instrumentation were designed,
constructed, tested and transported to the Engineering Research Center at Colorado State
University in Fort Collins, Colorado where the tests were conducted. These tests not only
provided valuable data but perhaps more importantly, new insight into local scour
mechanisms which has led to a new approach to estimating local scour depths and
volumes. Since the approach is new it needs to be thoroughly tested before being
implemented, but it has the potential of providing maximum structure-induced local scour
depths as well as maximum total local scour depths for extreme storm conditions without
having to specify the current velocity. The structure-induced local scour depth is that
component of the scour due to the presence of the structure. The total local scour is the
sum of the structure-induced component and the component due to the propagation of
sand waves past the structure (when they exist).
The results of what appear to be some good experiments on global scour near a variety of
vertical pile combinations published in a thesis by Hannah (supervised by A.J. Sutherland)
at the University of Canterbury in Christchurch, New Zealand have been reanalyzed and
incorporated into the SCOUR PROGRAM. Due to the scarcity of data on this subject,
this represents a significant improvement in the accuracy in which the SCOUR
PROGRAM predicts global or dishpan scour.
The proposed approach for estimating sediment scour near horizontal slab-like structures
subjected to currents proved unsuccessful. The potential flow model was inadequate to
handle the complexity of the flow for this situation. Initial attempts at using a
two-dimensional, second order closure turbulence model appear to be promising. Some
additional work is needed before the output from this model can be parameterized and
included in the SCOUR PROGRAM. This work is continuing as time and funds allow.
The feasibility study regarding the use of an existing three dimensional flow solver coupled
with a simple scour model has resulted in the conclusion that while perhaps feasible much
work and expensive computer time is needed before practical results can be realized. The
research needed is of a fundamental nature and should be supported by such organizations
as the National Science Foundation. Only after certain advances are made in this field
should applied research funds, such as those provided for this work, be expended on this
Two improved versions of the SCOUR PROGRAM accompany this final report. Version
1.5 includes an improved algorithm for estimating local scour. Additional data has been
added and new coefficients for the S-N equation (developed in the 1990 study) have been
computed and incorporated into the program. The global scour portion of the program has
been rewritten as a result of the data in the Hannah thesis. Other enhancements include
the addition of graphics files in many of the help menus to make the program easier to use
and to minimize misinterpretation of the input and output parameters. Version 2.0 includes
a new and different approach to the computation of local scour and should be considered
developmental at this point. Both the approach taken and the code need to be thoroughly
tested prior to its use. If this approach proves to be valid, a number of important

advantages can be realized including the computation of that portion of the scour depth
and volume due to the presence of the structure, computation of the total scour depth and
volume for an extreme storm event without the need for a velocity input to the program.


There were a number of people that contributed to the success of this project. These

Mr. Maximo Ramos, Master of Science Student and Graduate Assistant.

Dr. Christopher Reed, Post Doctoral Fellow and computational fluid dynamicist.

Mr. Subarna Malakar, Associate in Engineering and Head of the Computations
Laboratory in the Coastal and Oceanographic Engineering Department.

Mr. Sidney Schofield, Assistant Director of the Coastal Engineering Laboratory and
electronics design engineer.

Ms. Erin F. Parker, Computer Systems Manager, Environmental Science
Engineering, Inc., Tampa office.

Ms. Laura Dickinson, secretary and word processing specialist, assisted with the
organization of the project and typed all reports and documents associated with the

The authors are most grateful to these people for their hard work and dedication to the
project. A special thanks also goes to Dr. Steven Abt, Professor and Director of the
Hydraulics Laboratory in the Engineering Research Center at Colorado State University,
Fort Collins, Colorado for allowing us use of the facilities in his laboratory. Dr. Alfred
Devereaux and his capable engineering staff provided valuable input throughout the
project. In addition to his technical input, Mr. Paden Woodruff also contributed by
serving as project monitor and administrator.

Appendix A

Sediment Scour Bibliography

Appendix A

Sediment Scour Bibliography

Arkhipov, G.A., (1984) "Consideration of Sediment Transport when Calculating Local
Scour." Hydrotechnical Construction (English trans. of Gidrotekhnicheskoe
Stroitel'Stvo), V.18, No.4, 149-153.

Baker, C.J., (1981) "New Design Equations for Scour Around Bridge Piers." J. of the
Hydraulics Division, ASCE, Technical Note, HY4, 507-511.

Baker, C.J., (1980) "Turbulent Horseshoe Vortex." J. of Wind Engineering and Industrial
Aerodynamics, V.6, No.1-2, 9-23.

Baker, C.J., (1980) "Theoretical'Approach to Prediction of Local Scour Around Bridge
Piers." J. of Hydraulic Research, V.18, No.l, 1-12.

Baker, C.J., (1978) "Vortex Flow Around the Bases of Obstacles." Ph.D. Dissertation,
Univ. of Cambridge, 216 p.

Batuca, D and B. Dargahi, (1986) "Some Experimental Results on Local Scour around
Cylindrical Piers for Open and Covered Flows." Proc. Third Intern. Symp. on River
Sedimentation, Estuarine and Coastal Sedimentation, The Univ. of Mississippi,
March 31-April 4, V.3, 1095-1104.

Breusers, H.N.C., G. Nicollet and H.W. Shen, (1977) "Local Scour Around Cylindrical
Piers." J. of Hydraulic Research, V.15, No.3, 211-252.

Chabert, J. and P. Engeldinger, (1956) "Etude des affouillements autour des piles de
points Lab., Nat. de Hydr. Chatou.

Hannah, C.R., (1978) "Scour at Pile Groups." Master of Engineering Thesis at the Univ.
of Canterbury, Christchurch, New Zealand, 92p.

Haque, M.I. and K. Mahmood, (1986) "Analytical Study on Steepness of Ripples and
Dunes." J. of Hydraulic Engineering, V.112, No.3, 220-236.

Hopwood, H.J., M.A. Peviani and H.M. Quinodoz, (1986) "Local Scour at Relief Bridges
with Soft Clay and Fine Sand Foundations." Proc. Third Intern. Symp. on River
Sedimentation, Estuarine and Coastal Sedimentation, The Univ. of Mississippi,
March 31-April 4, V.3, 1085-1094.

Jain, S.C., (1981) "Maximum Clear-Water Scour Around Circular Piers." American
Society of Civil Engineers, J. of the Hydraulics Division, V.107, No.5, 611-626.

Jain, S.C. and E.E. Fischer, (1980) "Scour Around Bridge Piers at High Flow Velocities."
J. of the Hydraulics Division, V.106, HYll, 1827-1842.
Jain, S.C. and E.E. Fischer, (1979) "Scour Around Circular Bridge Piers at High Froude
Numbers." Sponsored by the U.S. Dept. of Transportation, at the Iowa Institute of
Hydraulic Research, The Univ. of Iowa, Iowa City, Iowa, IIHR Report No. 220, 68 p.
Jones, J.S., (1989) "Laboratory Studies of the Effects of Footings and Pile Groups on
Bridge Pier Scour." Proc. of the Bridge Scour Symp., Oct. 17-19, 340-360.
Kwak, D., J.L.C. Chang, S.P. Shanks and S. Chakravarthy, (1986) "A Three-Dimensional
Incompressible Navier-Stokes Flow Solver Using Primitive Variables." American Inst.
Aeronautics and Astronautics J., V.24, No.3, 390-396.
Mao, Y., (1988) "Seabed Scour Under Pipelines." Proc. 7th Inter. Conf. on Offshore and
Arctic Engineering, Houston, Texas, February 7-12, 33-38.

Mao, Y., (1987) "The Flow Induced Pipe Vibration During its Sagging Process."
J. of Hydraulic Research, V.25, No.5, 565-582.

Mao, Y. and B.M. Sumer, (1986) "Experiments on Scour Below Pipelines Exposed to
Waves." Inst. Hydrodyn. and Hydraulic Engineering. Tech. Univ. Denmark, Prog.
Rep.64, 3-12.
Melville, B.W. and A.J. Sutherland, (1988) "Design Method for Local Scour at Bridge
Piers." J. of Hydraulic Engineering, V.114, No.10, 1210-1226.
Melville, B.W., (1984) "Live-Bed Scour at Bridge Piers." J. of Hydraulic Engineering,
V.110, No.9, 1234-1247.

Melville, B.W. and A.J. Raudkivi, (1977) "Flow Characteristics in Local Scour at Bridge
Piers." J. of Hydraulic Research, V.15, No.4, 373-380.
Raudkivi, A.J. and H.H. Witte, (1990) "Development of Bed Features." J. of Hydraulic
Engineering, V.116, No.9, 1063-1079.
Raudkivi, A.J., (1986) "Functional Trends of Scour at Bridge Piers." J. of Hydraulic
Engineering, Jan., V.112, No.1, 1-13.
Raudkivi, A.J. and R. Ettema, (1977) "Effect of Sediment Gradation on Clear Water
Scour." J. of the Hydraulics Division, Technical Note, V.103, No.HY10, 1209-1213.
Shen, H.W., V.R. Schneider and S.S. Karaki, (1969) "Local Scour Around Bridge Piers."
J. of the Hydraulics Division, V.95, HY6, 1919-1940.
Shen, H.W., V.R. Schneider and S.S. Karaki, (1966) "Mechanics of Local Scour."
Colorado State University. Civil Engineering Dept., Fort Collins, Colorado, Pub. No.
CER66HWS22, 56p.

Shen, H.W., V.R. Schneider and S.S. Karaki, (1966) "Mechanics of Local Scour, Data
Supplement." Colorado State University. Civil Engineering Dept., Pub. No.
CER66-67HWS27, 40p.

Sheppard, D.M. and A.W. Niedoroda, (1992) "Physical Effects of Vegetation on
Wind-Blown Sand in the Coastal Environments of Florida." Report, sponsored by
the Florida Dept. of Natural Resources, at the University of Florida, Gainesville,
Florida, 46p.

Sumer, B.M. and M. Bakioglu, (1984) "On the Formation of Ripples on an Erodible Bed."
J. of Fluid Mechanics, V.144, 177-190.

Tsujimoto, T., (1986) "Local Scour around a Bridge Pier and Migration of Dunes." Proc.
Third Intern. Symp. on River Sedimentation, Estuarine and Coastal Sedimentation,
The Univ. of Mississippi, March 31-April 4, V.3, 1105-1114.

van Rijn, L.C., (1984) "Sediment Transport, Part III: Bed Forms and Alluvial
Roughness." J. of Hydraulic Engineering, V.110, No.12, 1733-1754.

van Rijn, L.C., (1983) "Sediment Transport Phenomena: 1. Sediment Transportation in
Heavy Sediment Laden-Flow. 2. Prediction of Bed-Forms and Alluvial Roughness. 3.
Equivalent Roughness of Alluvial Bed." U.S. Dept. of Commerce, National Technical
Information Service, Document No. PB86-151800, 19p.

Yalin, M.S., (1963) "An Expression for Bed Load Transportation." J. of the Hydraulics
Division, V.89, No.HY3, 221-250.

Yanmaz, A.M. and H.D. Altinbilek, (1991) "Study of Time-Dependent Local Scour
around Bridge Piers." J. of Hydraulic Engineering, V.117, No.10, 1247-1268.

Appendix B

Local Structure-Induced Scour Experiments

Appendix B

Local Structure-Induced Scour Experiments


Some of the problems regarding structure-induced scour addressed in this year's work
required that laboratory experiments be conducted. Briefly, these problems are as follows:

1. The need for data that will test the hypothesis that, for a given structure-sediment
situation, there is a maximum structure-induced scour depth with regard to velocity.
2. The need for data to determine rate of scour .

3. The need for scour data for multiple pile structures and certain non-conventional
structures such as swimming pools and horizontal slabs.

The experimental program is described in this appendix. Included are: 1) the objectives of
the experiments, 2) choice of the flume used, 3) design of the experimental apparatus, 4)
design and calibration of the instrumentation used, 5) procedures followed in conducting
experiments and 6) experimental results.


The objectives of this experimental program were to design and conduct a sequence of
laboratory experiments on local structure-induced scour that would assist in testing the
maximum scour hypothesis and also provide some of the needed scour data indicated above.

Experimental Facility, Apparatus, Instrumentation and
Structural Shapes

Experimental Facility

Several flumes located in the Coastal Engineering Laboratory and the Hydraulics
Laboratory at the University of Florida, were considered for the local scour experiments,
but due to the requirements for high velocity flows and sediment recirculation, it was
decided that the flume in the Engineering Research Center at Colorado State University in
Fort Collins, Colorado would be more appropriate. Arrangements were made to lease the
flume for a period of two weeks in September, 1991. Tests were conducted in the two
hundred foot long, eight foot wide, four foot deep tilting flume, shown in Figures B1 and

This recirculating sediment flume has a maximum discharge of 100 ft3/sec and a maximum
of 500 horsepower is also available for pumping water to the sediment flume. Water for this
flume originated from either a 1 acre-ft sump located under the Hydraulics Laboratory
floor or from Horsetooth Reservoir which is adjacent to the laboratory. A photograph of
the Hydraulic Laboratory that was taken from the top of Horsetooth Reservoir is shown in
Figure B3.

Apparatus and Instrumentation

Since one of the objectives of these experiments was to measure rate of scour
instrumentation had to be developed for making insitu scour measurements (i.e.
measurements of the depth and volume of scour while it occurred).
Measurement of rate of scour has been a very labor intensive and time consuming process
in the past. Recent developments in underwater acoustics technology now allow techniques,
formerly used only for bottom profiling at prototype depths, to be used in the laboratory.
Even though complete systems are not yet commercially available, some of the key
components such as the "echo sounder" can be purchased.
Using one of these echo sounders, a narrow beamed depth measuring system was developed
at the Coastal Engineering Laboratory. The echo sounder sends an acoustic signal towards
the bottom of the flume. The amount of time it takes the signal to be reflected back to the
sounder measures the distance to the reflecting object. The amplitude of the pulse at any
point in time is a measure of the amount of reflecting particles in the water column. This
depth measuring system is shown in Figure B4 and it consists of a fiberglass dish mounted
on a rod such that the dish could be placed above the test cylinder used in the local scour
experiments. The echo sounder was attached inside the dish. Figures B5 and B6 are
photographs of the dish and cylinder inside the flume.
To allow the echo sounder to send acoustic signals to the bottom, a small window was cut
out of the dish and a plastic sheet was glued over the hole. Water was poured into the dish
and since the reflection off the plastic sheet was minimal, the acoustic signal could travel a
near homogenous path to the scoured bottom.
The dish was then lowered into the water such that the plastic sheet was beneath the
surface. Streamlining the bottom of the dish minimized the disturbance of the flow. A
detailed description of the echo sounder is included at the end of this appendix.

Structural Shapes

Three types of structures were used in the local scour experiments. These are a 4 inch
vertical cylinder, non-conventional structures such as a swimming pool or horizontal slab
and multiple pile structures. The only structural type for which time dependent data was
taken is the 4 inch vertical cylinder. Table B1 summarizes the tests performed on each
structural type while Figures B7 to B11 show the dimensions of each structure.

Table Bl:
Summary of Tests on Several Structure Types
Structure Runs Time Dependent Figure
Scour Data
4 inch Cylinder and Base A,B,C Yes B7
Swimming Pool Model B No B8
Horizontal Slab Model B No B9
Multiple Large Pile Structure B No B10
Multiple Small Pile Structure B No B11

Calibration of Instrumentation Used and Test Procedures

Calibration of Instrumentation

The vertical angle is the measurement of inclination of the echo sounder head as it swings
away from the cylinder and back, as it returns towards the cylinder. Figure B12 shows the
echo sounder depth measuring system. In order to measure this angle, a protractor was
mounted on the side of the sounder. To calibrate the vertical angle measurement, the
sounder head was moved to a specific angle and the output from the potentiometer was
recorded by the data acquisition system. Figure B14 shows the relationship between the
vertical angle and the output from the potentiometer and a best fit line is plotted through
the data points. The equation of this line is:

Angle (in degrees) = 0.05 x N 180.0. (1)

The horizontal angle is the measurement of the echo sounder system rotating around the
cylinder. Two protractors were glued to the top of the cylinder to measure the horizontal
angle. The calibration of the horizontal angle was similar to the vertical angle. The system
was rotated to different angles around the cylinder and the output from the potentiometer
was recorded by the data acquisition system. Figure B15 shows the relationship between
the horizontal angle and the output from the potentiometer. A best-fit line plotted through
the points yields the equation:

Angle (in degrees) = 0.190 x N 400.4. (2)

The acoustic signal from the echo sounder was calibrated by placing a flat metal plate
directly below the sounder and recording the output with the data acquisition system. The
metal plate was attached to two rods that were mounted on point gauges above the
cylinder as shown in Figure B13. By raising the metal plate a specific distance and
recording the output, a relation between the distance of the plate to the sounder head and

the output can be determined. This relation is shown in Figure B16 and the equation for
the best fit line is:
Distance (inches) = 0.0325N 66.7. (3)

A simple experiment was set-up to determine if the echo sounder could accurately record
changes in the bottom elevation of the flume. Four different bowls were placed on the
bottom as shown in Figure B13. The bowls ranged from a crystal salad bowl that was
9 inches in diameter and 3 1/2 inches deep to a plastic bowl, 6 inches in diameter and
1 1/2 inches deep.
By taking a scan of the entire bottom, the sensitivity of the sounder to changes in bottom
elevation was determined. Figure B17 is a contour plot of the bottom with the bowls. It
shows that the sounder is able to pick up slight changes in the bottom elevation. The noise
in the contour lines are caused by the interpolation routine in the SURFER plotting
package. Figure B18 is a section through the salad bowl. The general shape of the bowl
can be seen and the shape and dimensions of the hole are close to the measured dimensions
of the salad bowl.

Test Procedures

For Runs A and C, the only structure in the flume was the 4-inch vertical cylinder. Prior
to both runs, sand in the flume was both smoothed and compacted from 25 ft. upstream to
15 ft. downstream of the cylinder. The sand in this flume has a dso = 0.78 mm as shown in
Figure B19. A 3" x 3" piece of wood was used for this purpose while a plumb bob was used
to ensure that the bottom was level while the rest of the sand in the flume was raked to
approximate a flat bottom. The level of the sand was determined by reading a gauge on
the side of the flume.
The flume was filled slowly and a wooden device was placed in front of the cylinder to
prevent scour occurring during the beginning of the filling process. This wooden device and
the smoothed bottom is shown in Figure B20. When the water depth reached three to four
inches, the wooden device was removed and the flume continued to be filled slowly.
When the water depth was slightly greater than two feet, the process of adjusting the
sluice gate and pumps to achieve the desired flow rate and water depth began. As the flow
rate approached the desired rate, the echo sounder was used to determine the depth of the
original flat bottom. Further scans were conducted at approximately 20 minute intervals to
observe the development of the scour hole. Each scan consisted of a 360 degree rotation
around the vertical cylinder. The water temperature and depth were also recorded during
each scan. After the final scan was completed, the flow was slowly reduced and the sluice
gate was closed until there was no flow. This allowed the water in the flume to drain slowly
and minimize the disturbance of the scour hole. A point gauge or ruler was used to
determine the shape and depth of the scour hole after all the water had drained from the

The structures placed in the flume before Run B consisted of a vertical cylinder,
non-conventional and multiple pile structures. The same procedures used in Runs A and C
were used in the measurement of local scour around the vertical cylinder. The swimming
pool model was placed in the flume such that approximately two inches of the pool
remained above the bottom. Two small I beams were placed inside the pool to prevent it
from moving around during Run B. The sand immediately around the model was smoothed
and compacted and a plumb bob was used to level the sand. No scour depth measurements
of the swimming pool model were conducted during Run B. After the water had been
drained from the flume, a point gauge mounted on a moveable platform was used to
measure the scour depths along the front, back and sides of the pool.
A horizontal slab model was placed in the flume prior to Run B. The model was placed in
the flume such that the top of the horizontal slab was approximately 1 inch above the
bottom. The sand around the model was smoothed, compacted and levelled. No significant
scour developed around the horizontal slab so no bottom measurements were taken.
Two multiple pile structures were placed in the flume prior to Run B. The sand in the tank
had to be shovelled to ensure that the structures were resting on a flat bottom. A T-beam
was placed across the flume and the multiple pile structures were clamped to the beam to
add more stability. As with the previous two structural shapes, the sand around the
multiple pile structures was compacted and smoothed while a plumb bob was used to level
the bottom. No scour measurements were conducted during Run B. After the water had
been drained from the flume, a scale was used to determine the local scour at each pile and
the dishpan scour around each multiple pile structure.

Experimental Results and Data Reduction

Local Scour Around a Vertical Cylinder

Three separate runs were conducted on the vertical cylinder which are summarized in
Table B2.

Table B2:
Scour Test Flow Conditions
Run Discharge Depth Ave. Critical Number
Velocity Depth Ave. of
Velocity Scans
(ft3/sec) (ft/sec) (ft/sec)
A 33.3 2.1 1.62 43
B 28.8 1.8 1.62 22
C 78.4 4.6 1.62 16

All three runs were conducted under live-bed conditions, a situation which eventually led
to the development of sand waves that propagated down the flume. Runs A and C had to
be cut short due to the presence of sand waves, since these waves tend to make the local
scour depths fluctuate significantly. Figures B21 and B23 show the scoured bottom around
the vertical cylinder for Runs A and C. The sand waves that propagated through the flume
during Run C are shown in Figure B24. The amplitude of the sand waves are
approximately 10-12 inches. During Run A, sand waves were observed propagating down
the flume at approximately 6 hours.
Figure B25a is a plot of non-dimensional scour depths versus time for Run A compared to
Hannah's (1978) data. Hannah's experiments were conducted in clear water conditions and
therefore no sand waves formed in his experiments. As the sand waves propagate through
the location of the vertical cylinder, the non-dimensional scour depth fluctuates as a crest
of a sand wave passes through the scoured region making it difficult to measure the local
scour around a structure.
During the initial filling process, a scour hole developed around the vertical cylinder. After
the desired water depth in the flume was attained and the flow was being increased, some
sediment was deposited in the initial scour hole. This process is shown by the decrease in
the non-dimensional scour depth of Run A in Figure B25a. By adjusting for this initial
scour and deposition process, the rate of scour around the vertical cylinder can be
determined for each run. Figure B25b is a plot of adjusted non-dimensional scour depth
versus time for Runs A, B and C.
Run B had a depth mean velocity that was not much greater than U,. This prevented
sand waves from growing as large and moving as fast as those during Runs A and C. This
allowed the scour hole to develop without the influence of sand waves as shown in
Figure B22.

Local Scour Around Swimming Pool and Horizontal Slab Models

A model of a swimming pool measuring 35 1/2" by 17 7/8" was placed in the flume before
Run B. The model was made of hard plastic and was left open on top. Approximately two
inches remained above the bed after the model had been placed in the flume. The area
around the model was divided into several sections. The level of the bottom and the
volume of sand lost was computed for each section. Figure B26 is a definition sketch of the
different sections. The average velocity during Run B was 1.8 ft/sec and the run lasted
approximately 5 1/2 hours. After the tank was drained, measurements of the bed were
made by using a point gauge. Figures B27 to B30 show the elevation of the bed next to the
model. Using these bottom measurements, an estimate of the volume of sand lost around
the structure was made. Figures B31 and B32 show the deposition of sand along the back
face of the pool and the local scour along the sides and front of the pool.
A horizontal slab model measuring 24" x 12" was placed in the flume such that the top of
the slab was at the same level of the sand. After the tank was drained, it became apparent

Table B3:
Section Volume Lost
A- B 96.0
B C 94.9
C D 74.9
D A -140.6
A A 12.0
B-B 11.9
C C 11.2
D-D 5.7
Total Volume of Sand Lost = 166.1 in3

that the horizontal slab model did not cause significant local scour as shown if Figure B33
so no local scour measurements were taken.

Local and Dishpan Scour Around Multiple Pile Structures

Two multiple pile structures were placed in the flume before Run B. Each structure
consisted of nine piles. The smaller structure had pile diameters of 1/2 inch and pile
spacings of 2.5 inches, while the larger structures had pile diameters of 1 inch and pile
spacings of 3.5 inches.
Figure B34 shows the local and dishpan scour around the multiple large pile structure after
Run B. Figures B35 to B38 show traverses normal and parallel to the flow for both multiple
pile structures. The abrupt changes in bottom elevation indicate the depth of local scour at
the individual piles. The results of this multiple pile structure experiment indicates that
dishpan scour can be observed for a small diameter pile structure in uniform flow.

Data Reduction and Analysis

A data transformation program (DTRANS.FOR) was written to convert sediment scour
data obtained from the echo sounder and position potentiometers to bottom coordinates in
rectangular form (x, y and z). The raw data consisted of 1) voltage proportional to the
angular position of the echo sounder system, 2) voltage proportional to the angle of
inclination of the sonar head (measured from a vertical line with the positive angle
occurring when the beam swings outward) and 3) voltage proportional to the distance from
the sonar head to the bottom.
Figure B39 is a definition sketch of the measurements needed to convert the raw data to
x-y coordinates. The equations needed to convert the raw data and variable list are
included below:

d = Icos0,
r' = isinO,
r = r + r',
r = ri+lsin0,
x = r cos /,
y = -r sin /,
z = -d=-lcos0,
z' = D+z=D-d=D- Icos,

Variable Definition:
f = Angle of rotation of echo sounder,
ri = Radius of rotational path of echo sounder,
r' = Radial distance from echo sounder to bottom measurement
(positive away from cylinder),
D = Distance from echo sounder to original bottom,
d = Distance from echo sounder to scoured bottom,
0 = Angle of echo sounder with vertical axis
(positive away from cylinder),
z' = Difference between original and scoured bottom level
(negative if bottom is eroded).

Sonar Processing System for Acoustic Bed Profiling

The acoustic bed profiling system is based upon the use of a sonar echo sounder operating
at 5 MHz. The 5 MHz operating frequency was chosen to give reasonable accuracies at
short ranges, as well as to insure reflection from the bed. Other operating frequencies can
be used with this instrument for longer distances to the bed or bed penetration. A 10 psec
pulse is emitted at a 100 Hz rate. The returned pulse is an analog time/amplitude history
of the water column. The travel time of the signal measures the distance to a reflecting
object. The amplitude of the pulse at any point in time is a measure of the amount of
reflecting particles in the water column at each distance. This signal is supplied to the
detector of the range processor.
The processing system consists of a 2 MHz oscillator, elapsed time counter, clock
generator, control circuitry, detector and signal threshold circuitry and digital to analog
convertor. Figures B40 to B44 are schematics of several components of the processing
system. The 2 MHz oscillator provides all of the frequencies necessary for the other stages.
The 2 MHz signal is sent through the clock generator circuitry. The two megahertz signal
is divided by 20,000 to produce the 100 Hz sampling pulse for the control circuitry. The
signal from the clock generator is a 1 msec pulse with a 10 msec repetition rate.
The control circuitry generates the signals necessary to initiate and stop the measuring
sequence. The rising edge of the 100 Hz pulse from the clock generator zeros the counters

for the beginning of a new time measuring sequence. After the counters are cleared, a start
pulse is sent out to the echo sounder to send out the sonar pulse and to enable the
counters. The counters are then updated at a 2 MHz rate until a stop pulse is received.
The stop pulse disables the counters and stores the count in the 74HC373 latches. If the
counters overflow before a stop pulse is received, the overflow signal will stop the count and
latch zeros into the 74HC373's. To prevent the transmitted signal from being interpreted
as a stop pulse, a zero inhibit signal is also supplied to the control circuitry. The zero
inhibit is taken from the counter chain. The zero inhibit can be set to 2N clock cycles. N is
an integer between 0 and 12. With the Mesotech echo sounder, N is chosen to be 9 which
provides for a zero inhibit delay of 512 clock pulses. At a 2 MHz clock rate, this
corresponds to about 19 inches total distance. Since the distance to the reflecting object is
one half the total distance, the zero blanking will keep the unit from measuring any
distance closer than 7.4 inches from the echo sounder.
The detector and threshold circuitry receives the pulse from the echo sounder and provides
the stop pulse to the control circuitry. The signal from the echo sounder is a nominal 1 volt
peak to peak signal impressed on a 455 KHz carrier. The detector is an active half wave
rectifier with an amplification gain of 4. The output of the detector is filtered by a simple
RC low pass filter to remove the carrier frequency. The output of the filter is supplied to a
BNC connector on the front panel of the instrument to allow the user to see the returned
signal. This signal is also sent to the comparator circuitry. The comparator has an
adjustable threshold provided by potentiometer P1. When the returned sonar signal rises
above a level determined by the threshold adjustment, the comparator sends a stop signal
to the counters.
The digital to analog converter provides a 0-5 volt output from the counter chain. The
D/A converter has two adjustments to provide for calibration of the speed of sound and
total span length. Using a nominal 1475 m/sec for the speed of sound in water results in a
distance resolution of 0.0369 cm.

d = 1.475 105/(2 2.00 106) = 0.036875 (4)
The digital counters can count to 4096 resulting in a total distance measurement of
151.04 cms. The distance can be increased by lowering the clock frequency from 2 MHz.
This increase in distance will result in decrease of resolution. For example lowering the
frequency to 1 MHz will result in a total measuring distance of 302 cms, but the resolution
will be decreased to .073 cms.
Initial setup of the instrument is accomplished by setting the speed of sound potentiometer
to the center and adjusting the span potentiometer so that the output voltage is 2.000 volts
with a reflecting target 2 feet from the sounder. This will result in a calibration coefficient
of 1 volt per foot. Using a data acquisition system having a 12 bit A/D converter results in
a resolution of 0.015 inches. The speed of sound variation can be compensated for by
adjusting the speed of the sound potentiometer. This allows for a variation of the speed of
sound by 33%. This allows speed of sound compensation from 988 through 1988 m/sec.

Overall View of the Recirculating Sediment Flume in the
Engineering Research Center at Colorado State University.

View of the Flume from the Upstream End.

Figure B1.

Figure B2.


Figure B3. The Engineering Research Center at Colorado State University.
Taken from Horsetooth Reservoir.


Echo Sounder

Test Cylinder._

:::~:~ j'
: :~:~i:'. i.. ~ '
~ :~:
5'::~:: ::

Figure 84. Overall Set-up of Experimental Apparatus and Instrumentation.

Figure B5.

Experimental Test Set-up Showing Fiberglass Dish Inside
the Flume.

Overall Set-up of the Instrumentation and Apparatus.

Figure B6.


Base ,1 ,
Side View

Figure B7. 4-Inch Diameter Cylinder and Base Used in CSU Local Scour Experiments.

[<-1 77/8"-->


Front View

Figure B8. Swimming Pool Model Used in CSU Experiments.




Side View

ff--------3 5 i /2)~-------~



1/4" 111/2"
14------------------1 1/2"
<- 24" --

Side View

12 ~

Front View

Figure B9. Horizontal Slab Model Used in CSU Experiments.



'IA If

4.4 44


Side View






Figure B10. Multiple Large Pile Structure Used In CSU Dishpan Scour Experiments.




t 1/8"

Side View

Figure B11. Multiple Small Pile Structure Used in CSU Dishpan Scour Experiments.

Figure B12. Close-up of Echo Sounder Depth Measuring System.

Figure B13. Metal Plate and Bowls Used to Calibrate the Echo Sounder.



Figure B14.

Calibration Curve for the Angle of Inclination of the
Echo Sounder Head.



, 12.5
-.. 7.5









n 240.0
s3180.0 -

Z 120.0


,'\ r-


Angle (degrees) = 0.190*N 4

+-I . .


I I i 0 I


3200.0 3


S110.11 I I 0.0I
3600. 40. 0

Figure B15. Calibration Curve for the Angle of Rotation of the
Echo Sounder System.

00 I .0










Distance (in.) = 0.0325*N 66.7

2400.0 2450.0 2500.0 2550.0 2600.0 2650.0 2700.0 2750.0
Figure B16. Depth Calibration.Curve for the Echo Sounder.













-15.0 -10.0 -5.0
15.0 r 1 I --I-r---M-r-r-




>- -5.0


0.0 5.0

-15.0 1

Figure B17.

Contour Plot of Four Bowls Placed on the Bottom of
the Flume.


r 15.0

- 10.0







0.0 -
C -1.0


N -3.0

4.0 0 I I ,
-15.0 -10.0 -5.0 0.0 5.0 10.0 15.0
Y (inches)

Cross-Section of the Salad Bowl (A-A). (See Figure B17)

Figure B18.



0.01 0.1

1 10
DIAMETER (mm.) D50 = 0.78 mm.

Figure B19. Grain Size Distribution Curve of the Sand Used for
Scour Experiments.

Figure B20.

View from the Downstream End, of the Wooden Device Used
to Prevent Scour During the Initial Filling Process.



~~. 4 .Pb
I .,

Figure 821. Scour Hole and Sand Waves Around the Vertical Cylinder After
Run A.



Figure B22. Scour Hole Around the Vertical Cylinder After Run B.

::' ~
~ tP.r


d..a ~I 'j'

Figure B23. Scour Hole Around the Vertical Cylinder After Run C.

Sand Waves Upstream of the Vertical Cylinder After Run C.

Figure B24.

**** RUN A

1.5 -


0.5 -



Figure B25a.

Non-dimensional Local Scour Depth versus Time.
(Run A and Hannah's Data)


3.0 4.0 5.0
Time (hours)


1.0 2.0 3.0 4.0 5.0 6.C
Time (hours)
Figure B25b. Non-dimensional Local Scour Depth versus Time.






1.0 2.0 3.0 1.0 15.0 6.0
Time (hours)
Figure B25c. Smoothed Non-dimensional Local Scour Depth versus Time.



m 1.0



-.W N

Side of Flume
/ // / / //// // //// /// / / //// // / I//'/ /.


< 351/2"1

Side of Flume

NOTE: Not to Scale

Figure B26. Definiton Sketch for Volumetric Computations of Swimming Pool Model.






0 2.0

- 1.0
!< 0.0

0 -2.0





Figure B27. Traverse of Swimming Pool Model (East Side).




u 2.0

- 1.0 -
F- -
< 0.0



m -2.0 -

0.0 10.0

20.0 30.0
A-B (inches)

I 4 I I I 5
.40.0 50.0

Figure B28. Traverse of Swimming Pool Model (West Side).



~I I i


l i i l l l ,










0.0 10.0



Figure B29. Traverse of Swimming Pool Model (North Side Front).


















30.0 I I 1

Figure B30. Traverse of Swimming Pool Model (South Side Back).







I . I

Figure B31. Local Scour at Upstream End of Swimming Pool Model After
Run B.

Figure B32. Downstream View of Scour Around Swimming Pool Model
After Run B.

Figure 833. Horizontal Slab Model After Run 3.

Figure B34. Dishpan and Local Scour Near Multiple Pile Structure After
Run B.





10.0 20.0 30.0
Traverse Parallel to Flow (inches)

Figure B35. Traverse of Multiple Large Pile Structure
(Parallel to Flow).




Q "






Traverse Normal to Flow (inch<

Figure B36. Traverse of Multiple Large Pile Structure
(Normal to Flow).











Traverse Parallel to Flow (inches)

Figure B37. Traverse of Multiple Small Pile Structure
(Parallel to Flow).


. -1 .1



c -2.
i3 -











Traverse Normal to Flow inchese

Figure B38. Traverse of Multiple Small Pile Structure
(Normal to Flow).


. -



Rotational Path
of Echo Sounder







~ ,Botton
^^F ^, ----r'------|

Figure B39. Definition Sketch Showing Measurements Needed for Conversion
of Raw Data to x-y Coordinates.












+ 15 PIN 9

-15 T PIN 8


Figure B40. Analog Interface

.1 'L







Figure B41. Sonar Processor Function




490 10K






Control Circuitry

Figure B42.



44 1 5 9 1 9132- C START
74HC745 1 CC OD C c

74HC390 4 74HC390 14

74HC74 12
7 8 11

Elapsed Time Counter Pulse Generator

Figure B43.

S8R 9 1--1 1 I2 10K 10K SPEED OF SOUND
14-" A 4 MSB --- r
U4 13------ 356
12 4 6
11 3 2 7 2 10K
1CU17 2
1I 81 SPAN 5K2
9 356 OUT TO
14 18 19 1+15 10KU18
1l 12 10KU
U3 1 14 17 10.00V
11 13 12 i VREF
14 8 / 9 LSB

U2 4 I 6U16 30K 3.6K
11 3 REF 220K V POS (DIN PLUG PIN 3)
LATCH 1 5K 7500 2
U14 -W- I -,
U4, > 6- ,, .,,,,o, H POS (DIN PLUG PIN 7)


Figure B44. Digital to Analog Converter

Appendix C

Computational Approaches to Local Structure-Induced
Scour Prediction

.Appendix C

Computational Approaches to Local Structure-Induced
Scour Prediction

The prediction of structure-induced scour poses a difficult problem due to the intricate
physical interactions involved. The hydrodynamics alone produces complex phenomenon
including turbulence, flow separation, and three-dimensional boundary layers. Similarly,
the movement of sediment caused by the fluid motion involves complex particle-particle
and fluid-particle interactions. However, advances in the last decade in computational fluid
mechanics and in the understanding and modeling of sediment motion provides the
necessary basis for applying computational methods to the prediction of structure-induced
As a means of investigating the potential application of these methods, a prototype model
of structure-induced scour has been developed and implemented for scour around a vertical
cylinder. The model consists of two major parts 1) a flow solver and 2) a model of
sediment motion. The flow solver, called INS3D incompressiblee Navier-Stokes in Three
Dimensions) was developed at NASA Ames (see Kwak et al., 1986). It solves the
Navier-Stokes equations for the detailed velocity and pressure fields around an object using
a finite difference solution algorithm. It requires a grid network to represent the domain
and the boundaries which, for this case is the bottom and the vertical cylinder. The second
part of the model consists of a an algorithm to determine the sediment motion. For this
initial effort, it was assumed that the sediment moved only as bedload (no suspended load).
The two components of the computational model of structure-induced scour are linked
together in an algorithm that is analogous to their physical relationship. This process is
depicted in Figure Cl which shows a flow diagram of the model. The flow solver is used to
obtain the flow field around the cylinder. Once calculated, the velocity field near the
bottom can be used to determine the shear stress on the bottom. This shear stress is
spatially varying, due to the variation of the velocity near the cylinder. Figure C2 shows
the calculated velocity field around a vertical cylinder. The top views show the cross
section of the cylinder, the accelerated flow past the cylinder and the wake region for two
different heights above the bottom. Also shown is the flow in vertical planes parallel and
perpendicular to the free stream flow and through the center of the cylinder.
The resulting shear stress is then used to calculate the sediment motion. A bedload
formula analogous to Yalin's formula was used to calculate the sand transport at each grid
point on the bottom using the friction velocity at that grid point. The friction velocity is
based on the shear stress calculated from the velocity field (friction velocity, u. = Vr
where r shear stress and p mass density of water). The net rate of accumulation or
loss of sediment is then obtained at each point by (numerically) integrating the transport
rate around a control volume centered around each grid point. The bottom topography is

then updated based on the spatially varying accumulation/loss rate by multiplying the rate
by a small time increment.
With an updated bottom topography, a new grid network is generated based on the new
topography and used in INS3D to obtain an altered velocity field. This process is repeated
a number of times to mimic the development of the scour.
This process has been implemented for flow around a cylinder. It should be noted that the
INS3D code as obtained did not include a turbulence model. Incorporating a turbulence
model into the code was beyond the scope of work for this study. Thus, the sample
problem presented here represents laminar flow. The detailed structure of laminar flow is of
course much different than that for turbulent flow, however, the large scale flow structures
such as the wake and swirling flow still exist and are adequate to represent the (time
averaged) mean flow phenomenon necessary to produce scour. An additional consequence
of using laminar flow is that the bottom shear can be quite different than that for
turbulent flows. It was necessary to assume an artificially low value of the critical shear
stress in order to obtain sediment motion. Again, this does not greatly change the
qualitative aspects of the results.
The resulting scour from the model calculations are shown in Figure C3 for four different
stages in the development of the scour topography. The erosion of sediment in the
upstream and adjacent regions is quite evident as is the accretion in the region directly
downstream of the post. In general, the development of the topography is a reflection of
the spatial gradients of the bottom shear stress. However, in the last frame shown, the
ridge that is forming on each side of the post is becoming large and its influence on the
flow becoming significant.
The application of the prototype computational model has revealed a number of issues.
The state of computational fluid mechanics has progressed to the point where it is
reasonable to solve the fluid flow equations around complex geometries. However, large,
fast 'supercomputers' are necessary to obtain solutions in reasonable time frames. The
solution algorithm must also be robust, in order to deal with the complex geometric
boundaries that will be created by scouring. In addition, special grid generation techniques
will be needed to provide smooth well-behaved grids in the complex regions.
A turbulence model applicable to highly curved and separated flow conditions will also be
needed. This is a stringent requirement on a turbulence model and a thorough validation
phase for the flow solver with a turbulence model will be necessary before the sediment
motion problem can be addressed.
Inherent in the bedload transport relationship used here, is that the sediment responds to
the local shear stress and is unaffected by gradients in the stress field. This assumption will
require validation. Also, gravitational effects must be included in the transport model to
adequately represent transport into and out of realistic scour holes.

Computational Approach for Calculating
Structure Induced Scour

Determine flow and pressure field
around the structure

Obtain bottom stress from flow field

work Determine sediment transport due to
bottom stress

Update bottom topography based on
sediment transport rates

8-no converged?


Bottom topography contour plots

Figure C1 Flow chart for computational scour model.

Top View: z= 5.00

Flow Direction

Top View: z= 0.38

Flow Direction

xz grid and u.w vectors

___ __ __ __ ___ __ __ __ \\x r

___ ___ ___ __ ___ ~ \ \ I / -
_______________ ,\\~\ -

_________-~- \I\ I -- -

Flow Direction

yz grid and v.w vectors


Flow Direction(inward)

Figure C2 Velocity field for flow around a post.

Flow Direction --

After time step 2400

Flow Direction --

After time step= 2200


Flow Direction --

After time steB 2600


Flow Direction
Flow Direction --

Figure C3 Countour plots of bottom topography.



Appendix D

Structure-Induced Scour Near Slab-Like Structures
Due to Steady Currents

Appendix D

Structure-Induced Scour Near Slab-Like Structures
Due to Steady Currents

The original intent of this task was to modify the two-dimensional potential flow model
used to calculate the scour under a horizontal pipe and use it to calculate the scour under a
horizontal slab. The application of the potential flow model to the slab problem, however,
proved unfeasible due to the strong viscous effects associated with the slab problem.
Subsequently, attempts were made to modify a two-dimensional turbulence model,
originally developed for air flow through plant canopies for this situation. The initial
results look promising.
The original model for flow past a pipe calculated the inviscid irrotational flow field around
the pipe and used the flow near the bottom to estimate the bottom shear stress. This is
similar in concept to a potential flow/boundary layer calculation and is valid for the pipe
configuration since the boundary layer remains relatively thin compared to the gap
between the pipe and bottom. In the case of the slab, however, the boundary layer has
sufficient length over which to develop that it can extend throughout the gap. The
boundary layers for these two cases are depicted in Figure Dl. The assumption that the
flow field is irrotational is clearly inappropriate for steady flow around slab-like structures,
thus the potential flow model should not be used for these situations.
In order to better represent the flow characteristics, a two-dimensional turbulent flow
model developed to study flow through finite length vegetation canopies has been modified
for application to the slab problem. A detailed description of this model can be found in
Sheppard and Niedoroda (1992). Essentially, the model consists of the Reynolds averaged
continuity and momentum equations with a second order closure scheme to represent the
turbulence phenomena. The model equations are solved numerically on a domain which
extends upstream, downstream and above a vegetation canopy (or the slab-like structure in
this case). The presence of the canopy is represented in the equations by a momentum sink
term and by an adjustment to the turbulence length scale both of which are related to the
canopy's vegetation density. The specific location, length and height of the canopy can be
defined on the numerical grid representing the domain by assigning values for the density
function at the appropriate grid nodes. Applications of this model to canopy flow problems
can be found in the referenced report.
The canopy flow model can be applied to the slab problem without major code
modifications. Two changes to the input parameters are required: 1) the location of the
vegetation canopy is moved and scaled to correspond to the exact dimensions of the slab
and 2) a large, uniform vegetation density profile is used. The large density is required to
prevent flow through the slab (canopy) thus simulating the impervious surface of the slab.

The results from an initial attempt at this modelling approach are shown in Figure D2.
Two velocity profiles are shown, one for flow at the upstream end of the slab and one near
the downstream end of the slab. It is apparent that some of the flow must be bleeding
vertically through the slab since continuity is not satisfied for flow under the slab. This
bleeding can be prevented by increasing the density function further, however, a numerical
instability problem has been experienced when higher densities are used. This stability
problem must be investigated and corrected before the model can be applied to the
prediction of scour under slab-like structure configurations.

a. Steady Flow Over a Horizontal Cylinder.

b. Steady Flow Over a Horizontal Slab.

Figure D1 Flow Over Horizontal Structures.

Figure D2 Velocity Profiles Computed With Two-Dimensional Second Order Turbulence Model.

Appendix E

Global Sediment Scour

Appendix E

Global Sediment Scour

When multiple vertical structural elements such as building support piles exist in close
proximity to each other, the sediment scour in the vicinity of the structure is composed of
two components. These are called local scour and global or dishpan scour. The local
component is similar to that for individual structures. The global scour is a dishpan
shaped depression that encompasses the entire structure and is thought to be the result of
additional turbulence in the near flow field caused by the structural elements. Global scour
is important in many areas of engineering including those concerned with the sediment
budget of beach-dune systems, scour in the vicinity of bridge piers, scour around offshore
platforms, etc. Due to the complexity of the flow and sediment transport processes,
attempts at analytical and computational solutions have not been successful. For local
scour, this problem has been partially overcome by conducting numerous laboratory
experiments. Unfortunately, only a limited number of experiments have been performed for
global scour. Hannah (1978) conducted a series of experiments with two, three, four and
six vertical, circular cylinders with different spacings and flow orientations. He did not
distinguish between local and global in his analysis but presented enough information that
his data could be reanalyzed. The results of a reanalysis are given in Tables E1-E5. In an
attempt to generalize his results (and thus expand their usefulness), they were normalized
by the local scour depth around one of the structural elements. That is, for a given set of
flow and sediment conditions, the global scour depth was assumed to be proportional to
the local scour depth of one of the structural elements (alone in the same flow field). The
global scour depth is adjusted for the pile (center line to center line) spacing to diameter
ratio according to the relationships determined from Hannah's data. The constants of
proportionality were also obtained from Hannah's data.
Jones (1989) showed through a series of laboratory experiments that if the spacing to
diameter ratio (a/D in Figures E1-E4) is less than about 3, the multiple element structure
can be considered a single, porous structure for scour purposes. Hannah's results show that
if a/D is greater than about 11, global scour ceases to exist. Thus, according to these
limited results, global scour exists for 3 < a/D < 11.
Algorithms for computing global scour for this range of a/D were developed based on the
above discussion and implemented in both versions (1.5 and 2.0) of the SCOUR
PROGRAM that accompany this report.

Table El:
Equilibrium Scour Depths for Two Pile Structures
(In-Line Flow1) (Taken from Hannah 1977)

Local Global Aspect Ratio3
2 ont back de a
0.29 0.01 2.03 1
0.68 0.11 1.95 2
0.85 0.29 1.81 3
0.95 0.50 1.66 4
1.08 0.75 1.48 5
1.24 1.00 1.28 6
1.44 1.20 1.04 7
1.68 1.39 0.78 8
1.99 1.66 0.43 9
2.34 1.91 0.06 10

1 see Figure El, 0 Flow Direction.
2 de = 1.25d,(7hours).
3 see Figure El for definition.
depth mean velocity = 0.285 m/sec,
water depth = 140 mm,
median sand grain diameter = 0.75 mm,
pile diameter = 33 mm.


Table E2:
Equilibrium Scour Depths for Two Piles
(Normal Flow1) (Taken from Hannah 1977)

Local Global Aspect Ratio3
de2 de a
T average 77
0.00 4.53 1
0.26 2.56 2
0.53 2.10 3
0.84 1.76 4
1.10 1.48 5
1.28 1.24 6
1.50 1.00 7
1.73 0.74 8
1.91 0.50 9
2.15 0.25 10
2.38 0.00 11
1 see Figure El, 900 Flow Direction,
2 de = 1.25d, (7 hour) average of two piles,
3 see Figure El for definition.
= depth mean velocity = 0.285 m/sec,
= water depth = 140 mm,
= median sand grain diameter = 0.75 mm,
= pile diameter = 33 mm.


Table E3:
Equilibrium Scour Depths for Three Pile Structures
(In-Line Flow') (Taken from Hannah 1977)

Local Global Aspect Ratio3
Front ,middle -back de a
0.50 0.16 -0.15 2.03 1
0.64 0.21 -0.20 1.95 2
0.72 0.44 0.19 1.81 3
0.97 0.72 0.30 1.66 4
1.14 0.75 0.33 1.48 5
1.31 1.12 0.43 1.28 6
Ssee Figure E2, 0 Flow Direction,
2 de = 1.25d, (7hour),
3 see Figure E2 for definition.

Table E4:
Equilibrium Scour Depths for Four Pile Structures
(In-Line Flow') (Taken from Hannah 1977)

Local Global Aspect Ratio3
d2 front back de a
0.10 -0.60 3.28 1
0.49 0.11 2.26 2
0.44 0.19 1.96 3
0.46 0.29 1.71 4
0.67 0.48 1.48 5
1 see Figure E3, 00 Flow Direction.
2 de = 1.25d, (7hou).
3 see Figure E3 for definition.


depth mean velocity = 0.285 m/sec,
water depth = 140 mm,
median sand grain diameter = 0.75 mm,
pile diameter = 33 mm.

Table E5:
Equilibrium Scour Depths for Six Pile Structures
(In-Line Flow1) (Taken from Hannah 1977)

Local Global Aspect Ratio3
d front d middle back de a
0.85 -0.03 -1.03 3.28 1
1.27 0.89 0.24 2.26 2
1.27 1.04 0.54 1.96 3
1.13 1.02 0.62 1.71 4
1 see Figure E4, 0 Flow Direction,
2 de = 1.25d, (7hou),
3 see Figure E4 for definition.
U depth mean velocity = 0.285 m/sec,
h water depth = 140 mm,
D50o median sand grain diameter = 0.75 mm,
D E pile diameter = 33 mm.

Flow 90

--- Flow 00



Local Scour


Mean Water Level


' Bed Level
After Scouring

Two Vertical Piles, Circular Cross-Section.




Figure El.

Flow 900

0 0 O -- owO0



Three Vertical Piles, Circular Cross-Section.

Figure E2.

Flow 90






Dishpan Scour

Local Scour

Mean Water Level
--- Flow

Bed Level

Bed Level
After Scouring


Four Vertical Piles, Circular Cross-Section.


--- Flow 0


Figure E3.

Flow 900


TO 0 0


-- Flow 00


Local Scour


\ Bed Level
After Scouring

Six Vertical Piles, Circular Cross-Section.

Figure E4.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs