University of Southern California Dissertations and Theses

Wave-Induced Scour Around Natural And Artificial Objects

(USC Thesis Other)

Wave-Induced Scour Around Natural And Artificial Objects

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 70- 11,382 PALMER, Harold Dean, 1934- WAVE-INDUCED SCOUR AROUND NATURAL AND ARTIFICIAL OBJECTS. University of Southern California, Ph.D., 1970 Geology University Microfilms, A X E R O X Company, Ann Arbor, Michigan WAVE-INDUCED SCOUR AROUND NATURAL AND ARTIFICIAL OBJECTS by Harold Dean Palmer A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Geological Sciences) January 1970 UNIVERSITY OF SOUTHERN CALIFORNIA TH E GRADUATE SCHOOL U NIVERSITY PARK LOS ANGELES, C ALIFO RN IA S 0 0 0 7 This dissertation, written by under the direction of his.... Dissertation Com mittee, and approved by all its members, has been presented to and accepted by The Gradu ate School, in partial fulfillment of require ments of the degree of ,HARQLD..DEM..EAmEEL. D O C T O R O F P H IL O S O P H Y Dean Date Januarx.. 19. 7. 0. DISSERTATION COMMITTEE Urman TABLE OP CONTENTS Chapter Page INTRODUCTION.................................. 1 Geheral statement .......................... 1 Definition of terms ........................ 2 Previous work.............................. 4 Pield methods.............................. 7 Limitations................................ 16 Laboratory methods ........................ 17 Acknowledgments ............................ 20 DYNAMICS OP SCOUR.............................. 22 Introduction .............................. 22 The problem of threshold velocity ......... 22 Sediment and fluid Interaction under scouring conditions ................... 34 Secondary flows.......... 34 The primary vortex................... 45 Other vortices and pressure gradient effects.............................. 57 Dyed sand study........................ 65 Summary of scour phenomena ........... 70 DATA REDUCTION AND ANALYSIS................... 73 Films and photographs..................... 73 ii Chapter Page Water motion at the sea floor............... 74- Fluid velocities and sediment parameters . . 76 Sediment transport rates and scour development.............................. 85 Summary of scour analysis ................... 91 INTERPRETATION OF D A T A .......................... 93 Lateral extent of scour ..................... 93 Permanent stations .......... . . . 103 Platform studies ............................ 108 Comparison of scour under unidirectional and oscillatory f l o w ................... 110 Interpretation of time-lapse films ......... 118 GEOLOGICAL SIGNIFICANCE .......................... 134 Nearshore environment........................ 134- Submarine erosion related to s c o u r ......... 14-7 Paleoenvironmental analyses ................. 147 Scour at d e p t h .......... 132 CONCLUSIONS...................................... 161 REFERENCES...................................... 166 iii ILLUSTRATIONS Plate Page 1. Time-lapse motion picture system ......... 9 2. Operating positions for time-lapse camera . 11 3. USC surgemeter and record............ 14 4. Primary vortices in front of obstructions . 41 3. Formation of turbulent "gusts" ........... 43 6. Scour depressions around tall pilings . . . 46 7. Sand slumps in the primary vortex ripple . 50 8. Accelerations of flow adjacent to obstructions .............................. 59 9. Sequence of scour and turbulent flow around cylinder 10 cm diameter ........... 61 10. Formation of horseshoe vortex ............. 66 11. Close-up sequence of wake effects behind cylinder 10 cm diameter.............. 68 12. Unidirectional scour around rock ......... 114 13. Progressive scour around rock on sea floor 136 14. Scour around boulder outcrop on sea floor . 138 15. Scour channel adjacent to bedrock outcrop . 145 16. Submarine erosion by scour ................ 148 17. Submarine erosion by scour ......... 150 18. Relationship of objects on deep sea floor to sediments and microtopograp'hy... 155 iv Figure Page 1. Index map of diving stations............. 18 2. Combined threshold velocity curves from various sources ......... » • ....... 24 3. Variations in flow regime between uni directional and oscillatory f l o w ......... 28 4. Three flow regimes established around a cylindrical obstruction ................... 36 5. Schematic diagram of flow regime in scour pit.................................. 39 6. Growth of primary vortex and scour pit . . 53 7. Velocity and weight of sand moved for establishing equation approximating V-^o • • 77 8. Relationship of sand weight to sediment number N s .................................. 80 9. Relationship of effective Ns to weight of sand m o v e d ............................ 83 10. Relationship of effective Ns to strong surge frequency............................ 86 11. Approximation of scour pit geometry .... 89 12. Geometry of test objects employed........ 94 13. Relationship of cylinder diameter to scour pit development...................... 96 14. Relationship of object diameter to ratio of pit diameter/object diameter ........... 99 15. Curves of scour volume related to object height..................................... 101 16. Scour development at permanent stations . . 105 17. Combined data from sea floor studies and tidal flat s c o u r .......................... Ill v Figure Page 18. Relationship of object size to pit width and velocity.............................. 119 19. Growth curves for cylindrical objects . . . 121 20. Growth curves for various irregular objects ..... 123 21. Comparison of two measures of sediment transport rates used in this study .... 129 22. Comparison of effective Ns and percentage volume for various objects . ........... 131 23. Proposed scour in the vicinity of boulder outcrops.................................. 142 Table I. Calculated and observed surge velocities . 32 II. Tabulation of sediment and water motions for selected studies ..................... 125 vi 1 INTRODUCTION General Statement One of the major endeavors in marine geology is the study of erosion, transportation and deposition of natural sediments on the sea floor. With the advent of self-contained underwater breathing apparatus (SCUBA) and refinements in underwater cameras,a number of marine geologists have approached the problem of near-shore sedi ment motion through direct observation and sampling on the sea floor. In many cases, their observations show little correlation either with model studies performed under rigorously controlled laboratory conditions or with theoretical solutions based upon wave mechanics. Careful measurements of wave-induced water motion on the sea floor have departed significantly from theoretical values (Zenkovich, 1967) and problems inherent in scaling sedi ment parameters and hydrodynamic factors in model studies impose limitations to flume experiments (Powell, 1955)* This study was undertaken to determine, under natural conditions, the behavior of unconsolidated marine sediments in the vicinity of objects which obstruct the flow of wave-induced surge. Such disturbed flows produce distinctive depressions in the sediment surface which can be used to determine current directions and the velocity of the flow producing them. The primary objective was the determination of factors affecting scour configuration around a wide variety of objects in various orientations to the flow direction. In addition, sediment transport rate measurements provided the basis for attempts at cor relation between various factors such as object size, grain size, surge velocity and other measured or calculated para meters . Discussions and correspondence with domestic and foreign sedimentologists indicated that marine geologists are not currently conducting sea floor studies of scour. It is hoped that the data presented here may provide in sight into the nature of sediment transport influenced by obstructions to flow. Definition of Terms The excavation of material adjacent to natural and artificial objects has been termed scour. Many writers employ this term to describe the generation and modifica tion of ripples, the depressions produced by wind in snow and desert sands, the lowering of beach and stream pro files, and, in general, the wave and current-induced dis placement of subaqueous sediment. This study emphasizes the erosion of near-shore marine sediments in response to perturbations in wave-induced flow created by obstacles. The orbital paths of water particles under a train of gravity waves describe increasingly flattened ellipses with increasing depth until, at the sea floor, the paths become horizontal linear motions termed oscillatory flow. This to-and-fro motion of water is termed surge, and the velocity of such surge has been shown to be related uo wave period, wave height and water depth. Prom these relationships, it is clear that for a given wave spectrum surge velocities increase toward shore to the breaker line. Depending on sediment texture, there is a critical depth at which, under given wave conditions, the increasing surge velocity will attain the "threshold velocity" of the sedi ment. At this point, individual grains will be moved and, with increasing velocity, the flow attains the critical erosion velocity (Kuenen, 1967), forcibly displacing and transporting sediment grains from their previous location. Under these higher velocity conditions, sediments are transported in several modes. Near the bottom, surge creates a dense flow of sediment-charged water. Observa tions made during this study indicate that this layer, which has a measured maximum thickness of 6 cm, transports grains at a velocity equal to the surge. Transport is, therefore, by turbulent suspension (Leliavsky, 1957). The writer could find no reference describing a term for this layer, and to facilitate discussion, this feature will be termed the "suspension carpet." This is the suspended equivalent of the "traction carpet" (bed load) discussed by Allen (1965a). The traction carpet, characterized by sediment motion on the bottom at less than the flow velocity, is much less efficient in transporting sediment than the suspension carpet. Under conditions of turbulent flow, forces tending to lift grains predominate over drag forces (Manohar, 1955) and transport of particles is achieved by suspension of the sediment. Turbulence which develops during this flow also raises sediment above the level of the suspension carpet. This material may not settle to the bed before reversal of flow occurs, and some sediment grains may be deposited quite near their original position prior to surge-induced motion. Previous Work A bibliography of references, describing scour phenomena, has been compiled, but its length precludes incorporation in this paper. Many are more appropriate to hydraulic engineering as they treat remedial measures for scour-induced damage around obstacles such as bridge piers whereas others are reports on laboratory studies which may not be relevant to sea floor studies. It seems desirable to separate laboratory and theoretical studies from actual field investigations so that findings based upon these distinctly separate approaches may be placed in perspective. Therefore, readers who may consult the list of references for sources helpful to their interests will benefit from noting the abbreviations listed after each citation. Such notation is explained in the heading under "References." Because of its immediate application to design considerations for foundation members placed underwater, the study of scour effects is of practical interest to engineers. Various configurations of flumes and wave tanks utilizing unidirectional flow have been employed in model studies of scour around submerged obstacles. Tison (1961), Moore and Masch (1962) and Shen, Schneider and Karaki (1966) evaluated the effects of various cross- section geometries to scour development. A review of sub aqueous vortex behavior and its relation to scour has been prepared by Richardson (1968). Wind-generated scour in snow was described by Allen (1965b). Peabody (194-7) dis cussed current-induced depressions preserved in the Triassic Moenkopi Formation, and a similar paper describ ing contemporary current marks in wadi floors is des cribed by Karcz (1968). Allen (1968) included a discus sion of scour in his thorough review of fluid motion and sediment transport associated with current ripples. Although this study is concerned with scour in sand-sized materials, current marks in finer materials have been evaluated by several workers. A general classi fication of scour forms was presented by Dzulynski and Sanders (1962) and the use of such features in determining direction of flow in turbidites is described by Sanders (1965). Model studies designed to simulate the oscillatory wave-induced flow on the sea floor are less common than studies conducted with unidirectional flow. To the writer's knowledge, only two closed flumes were in opera tion in 1965; one at the Technical University of Denmark (Copenhagen) and one at the Georgia Institute of Tech nology in Atlanta. At the latter facility, Carstens and his colleagues published a number of reports treating incipient motion phenomena, ripple formation, and the generation of scour around obstacles (Carstens and others, 1963, 1965, 1966, 1967a, and 1967b). Several measure ments programs conducted by the United States Navy (R. P. Dill, Personal Communication, 1968) employed diving geologists in studies of sea floor scour around objects but the results are classified. To the writer's knowledge, the only other field study of scour performed by diving geologists is that by Bradford (1966) who conducted sedi ment tracer and photographic investigations of scour around objects having regular geometric cross-sections. Scoured depressions in the vicinity of deep sea outcrops are indicative of prevailing current conditions on the sea floor. In most cases, such pits are similar to those developed under a unidirectional flow regime. Photographic evidence of deep-water scour presented by Heezen, Tharp and Ewing (1959)» Laughton (1963), and Pratt (1963, 1968) provides graphic evidence of deep sea current velocities of a magnitude sufficient to erode, or to at least preclude the deposition of, globigerlna ooze and pelagic sediment. In a recent publication devoted to deep- sea photography, Owen and Emery, Hollister and Heezen and Laughton (in Hersey, 1967) exhibit photographs which dis play scour probably generated by turbidity ourrents or relatively high-velocity deep currents'to depths as great as 4220 m. However, the maximum depth of scour noted in available deep-sea photographs is that shown by Laughton (1963) at a depth of 56I8 m. Continuing photographic studies of the deep sea fl*.or will undoubtedly provide additional evidence that erosive scour around obstructions to flow takes place in all depths in the sea. Field Methods Field studies were performed with the aid of SCUBA which permitted direct observation, measurement and photography of scour development around natural and artificial objects. Supporting studies conducted from a small submersible and field investigations at offshore platforms and in a tidal inlet provided additional data for evaluation of scour effects. Still photography at various sites was conducted with conventional underwater cameras but a special motion picture system was fabricated to provide the time-lapse capability required by this study. This camera system, which is shown in Plate 1, consists of a Minolta K-7 Super 8mm motion picture camera coupled with a variable increment timer which activates the shutter release. A water-proof pressure housing and tripod collar were fabricated to provide three modes of operation for vertical and oblique filming. The camera is shown in various operating positions in Plate 2. Time settings of 2.75 and 5.1 seconds per frame permitted filmed studies of 160- and 310-minute durations, respectively, for a single subject. Study objects were replaced after a significant scour pit had developed around them, and these scour dimensions were compared against identical control ob jects placed nearby. Thus, the scour observed in 60 to 90 minutes of time lapse could be compared to scour around identical controls in place for the duration of the day's survey; usually 4 to 5 hours. On a real-time basis, both natural scale and close-up motion picture photography provided the technique necessary to study water velocities near obstacles, grain behavior, vortex magnitude and dura tion, and other short-term effects. Slow motion and frame-by-frame projections of these and the time-lapse films provided the major portion of the data reported here. For long-term studies, the writer made periodic PLATE 1 Time-lapse motion picture camera system and housing fabricated for scour studies described in this report. Lower sketch shows components. Camera has been modified since this photograph to permit adjustment of setting on the sea floor. 9 FRONT PORT PLATE 2 Operating positions for the time-lapse camera employed in sediment transport studies. a. Diver adjusting camera in vertical position shown in (b). b. Camera mounted in tripod position 1.6 m above sea floor. Fixed tripod legs hold camera shroud and housing in vertical position at constant elevation above the sea floor. c. Camera and housing in oblique position on monopod base. This configuration was used for close-up photography in time-lapse and real-time photography. d. Modified tripod configuration for distant oblique photography. Legs are those used in vertical position, with modification made on the sea floor by diver. 11 ■S^ OPERATING POSITIONS FOR CAMERA visits to a pier which had been marked with a reference level. In addition, four stations were established at depths of 7, 14, 21, and 28 m off the windward coast of Santa Catalina Island, and these stations were also checked periodically for progressive scour development. Data required for computing actual and theoretical factors (wave period and height, water depth, current velocity and direction) were obtained during the study after suitable devices had been constructed. Wave period and height were determined at each station by means of a stop-watch and wave staff and these measurements provided the theoretical wave length. Surge velocities and directions at the sea floor were measured by a device which recorded the angular displacement of small plates aligned normal to the surge direction (PI. 3). This "surgemeter," designed by Mr. H. J. Summers and the writer, was cali brated against a sensitive strain gauge device maintained by the Naval Underwater Warfare Center, San Diego. This meter, in turn, was calibrated in a towing tank under carefully controlled flow conditions. The observed velocities reported here were found to agree within a 95 percent confidence level in relation to calibration tests involving both meters which were operated simultaneously on the sea floor off Mission Bay, California. PLATE 3 a. Diver placing surge orientation device which records a "surge rose" similar to a wind rose in meteorology. Device standing at right is the USC surgemeter which records surge velocity over a 13-15 minute period. Lower cylinder within the housing atop the supporting rod is the recording drum, upper cylinder contains drive motor. Vertical plates, or "wings", maintain an orientation into the surge flow. Two paddles marked "1" and "2" are deflection plates which measure surge velocity and duration. b. Combined records from calibration runs of USC surgemeter with the Naval Underwater Research and Development Center surgemeter. Center trace with arcuate record is from USC meter, two bordering records are from Rustrak re corder mounted in the NUWC meter. 14 iTABT SURGEMETER RECORDS Limitations 16 The conclusions of this report are based primarily upon time-lapse and real-time motion picture photography of scour phenomena. Successful underwater photography is largely dependent upon a fortuitous combination of wave and surf conditions, tide, proximity to shore and light ing. These factors, combined with unfavorable sea condi tions which precluded safe operations, caused cancellation of about one-half of the scheduled diving days. In spite of these limitations, 137 productive dives on 49 days at sea provided useful data. Appropriate sites for scour analyses were, by the nature of the problem, in shallow water where wave- induced surge was of sufficient magnitude to displace sea floor sediments. In many coastal areas, the sediments are either too fine or too poorly sorted to permit the rapid settling of suspended grains necessary for good photo graphy. A minimum vertical visibility of 2 m was essential for the employment of the time-lapse camera and this con dition is consistently met at only a few local sites. Near-shore turbidity caused by ebbing tides which trans port suspended materials seaward was frequently an addi tional restriction to photography. On days when high surge velocity was encountered, work was attempted further offshore in deeper water. In some cases, this maneuver was unproductive owing to reduced illumination at greater 17 depths. Overcast and fog were additional factors contri buting to poor lighting conditions. In various combinations, these variables often made successful photography impossible. However, two areas in which adverse conditions were usually at a minimum were exploited: the region in the vicinity of Point Pertain on the Palos Verdes Peninsula and the windward side of Santa Oatalina Island. These sites are shown in Figure 1. Time-lapse studies of scour around objects were made for periods ranging from 20 minutes to 2 hours depend ing upon the rate of scour development. No underwater lights were employed so long-term effects were not re corded. However, measurements on Identical control objects located adjacent to the subject observed during the same day for periods up to 6 hours show that the time-lapse invariably captured at least 90 percent of scour develop ment in any given period of photography. Laboratory Methods Grain size distributions of samples collected dur ing this study were determined from analyses employing the automatic settling tube recently Installed in the U.S.C. Sediment Laboratory. The use and calibration of this de vice are described by Cook (in press) and Felix (in press). Grain densities were determined by conventional pycnometer techniques. FIGURE 1 Index map for diving stations occupied during scour study. 18 L O S INDEX MAP DIVING STATIONS ANGELES • no" s c a I e (km) © — frequent stations O — single stations □ — permanent stations X - deep station'vith submarine o SANTA" CATALINA ISLAND SAN CLEMENTE I S L A N D . SAN DIEGO VO 20 Acknowledgments Investigations of the sea floor conducted with SCUBA diving techniques require the support of many indivi duals. Mr. David 0. Cook acted as the indispensable div ing partner for these studies and his assistance is grate fully acknowledged. The Ahoyoha III, supporting vessel for all diving studies, was piloted by Mr. Richard Bergeron whose skill and patience made these observations possible. Discussions with Drs. R. P. Dill and E. L. Hamilton of the U. S. Naval Underwater Research & Development Center, San Diego, were enlightening, and Dr. E. C. LaPonde and Mr. E. C. Buffington made possible the observations from the U. S. Navy Electronics Laboratory oceanographic tower and the calibration of necessary instrumentation. Studies conducted from a small submersible vehicle and dives on the outer shelf in Santa Monica Bay were made possible by Dr. James W. Vernon of General Oceanographies, Inc. The time-lapse camera used in these studies could not have been assembled without the aid of Mr. H. J. Summers, and his interest and assistance in the entire project has been of great value. Dr. D. S. Gorsline, Chairman of the writer's committee, gave freely of his time in reviewing the progress of this study. The preliminary manuscript was critically reviewed by the writer's Guidance Committee consisting of Drs. Donn S. Gorsline, Richard Stone, and Richard Tibby. Correspondence, discussions, reports and data from the following individuals are gratefully acknowledged: Dr. M. R. Carstens, Georgia Institute of Technology; Dr. Norman Brooks, California Institute of Technology; Dr. P. D. Richardsen, Brown University; Dr. L. Draper, National Institute of Oceanography, Great Britain; and Dr. F. E. Masch, University of Texas. The writer is indebted to the scores of graduate students, businessmen, academicians, and other personnel who assisted him during this study. “ “Tire"writer was granted leave of absence from his employer, Dames & Moore, to com plete the dissertation research and academic requirements leading to the doctoral degree. Their generous support made this study possible. Support for vessel operations was made available through National Science Foundation Grant &B 8206, and the fabrication and calibration cost of the USC surgemeter were included under the grant from the Office of Naval Research Contract number N00014-67-A- 0269-0002. 22 DYNAMICS OF SCOUR Introduction Because scour processes represent an accelerated mode of erosion, there exist many similarities between scour and net sediment transport at the sea floor. How ever, the magnitude of the object whose presence produces scour is usually much greater than common sea floor bed forms, and thus the magnitude of velocity and pressure changer; is accordingly higher than those over a rippled bed. The evaluation of critical factors bearing upon the understanding of scour is followed by analyses of the field data obtained during this study. The Problem of Thresho'1 ' Velocity In attempting to determine the flow conditions under which noncohesive sediment experiences initial displace ment, geologists have sought relationships between velocity and grain size which could be applied to sediment transport studies. The velocity of the fluid must in crease as it passes around objects and, thus, in the vicinity of obstructions, the threshold velocity for a given sediment size is reached sooner, and transport con tinues longer, than in neighboring regions of unobstructed flow. This factor influences the rate and duration of 23 sediment scour and is, therefore, essential in calculating sediment transport functions under scouring conditions. 4 * Most marine geologists are aware of the Hjulstrom w diagram (Hjulstrom, 1939) and the more recent work of Sund- borg (1956). Other studies conducted under laboratory and field conditions provide a broad spectrum of velocities. The range of values is bewildering, as demonstrated by inspection of Figure 2, which graphically summarizes some of the published references on threshold velocity for given grain sizes. A discussion of fluid mechanics and sediment behavior is provided by Inman (194-9) and, more recently, by Allen (1968). Hjulstrom's values have been criticized on the basis of his employment of the settling velocity curve for obtaining the minimum velocity required for transport (Kuenen, 1967; Briggs and Middleton, 1965). Employment of any threshold criteria in sediment transport functions must also recognize that the velocity required to initiate movement is greater than that required to maintain motion. Thus, the settling velocity adequately approximates the termination of motion, but is less than the velocity re quired for initial displacement. It has been proposed that the critical factor in the movement of grains is the magni tude of bottom shear, or the shear velocity (Briggs and Middleton, 1965). Under controlled conditions, for a given sediment and flow condition, it is possible to calculate FIGURE 2 Combined "threshold velocity" contrasting fluid velocity with sediment median grain size. See REFERENCES for publications. (S) Sundborg (1956) (H) Hjulstrom (1939) (M) Menard (1950) (b) Bagnold (1946)# (e) Eagleson and others (1958)# (v) Volkov (1960)# (1) Larras (1957)* (g) Gugnyaev (1959)* (*) Experimental value for laboratory sands (Carstens and Neilson, 1967, see p. 86) * References marked with asterisk (*) were ob tained from Table 7, p. 96, in Zenkovich, 1967. 24 25 100 5 0 o 0) 10 s E o >- h- O O _J L x J > 5.0 0,5 0 . 1 GRAIN DIAMETER (mm) the critical tractive force (critical shear stress per unit area of the bed) at which the sediment is placed in motion. Modification of the Shields diagram (Shields, 1936) by Vanoni (1964) permits calculation of critical shear stress providing a number of variables are known. However, Brooks (1958) suggested that the initiation of sediment motion is not a unique function of shear stress and Bagnold (1963, p. 525-526) observed that the drag induced in bottom flow by a rippled bed Invalidates many of the empirical relationships derived from experiments conducted over artificially smoothed beds. He also noted a significant source of error arising from the fact that the laboratory analyses assume a homogeneous fluid acting on the sediment surface. Sediment suspended by wave surge creates a heterogeneous fluid (a suspension carpet) Immediately above the bed, and thus the shear stress act ing on the sediment surface bears no relation to measured velocity gradients in a clear fluid. It is not surprising that coefficients and empirical relationships derived under controlled laboratory conditions do not agree with field measurements. An additional factor bearing upon the application of published critical shear stress and threshold velocity equations results from the generation of these data from experiments conducted under unidirectional flow. Some investigators (Einstein, 1948; Kalkanis, 1964) feel that 27 some of the basic principles of stream flow hydrodynamics are valid concepts in the study of wave-induced oscillatory flow. Others cited previously do not agree. A field experiment reported here tends to confirm Einstein's and Kalkanis' interpretation, at least for cylinders up to about 14 cm in diameter. In flumes or natural streams and rivers, the velocity gradient in the water column is relatively con stant for a given short period of time. Consider the magnitude of a reasonable change in the horizontal velocity gradient at the bed which may occur within 1 minute under established unidirectional flow (Pig. 3). This may be on the order of several centimeters per second due to fluctuations in the flow regime. Contrast this with the reciprocating horizontal velocity gradient under progressive waves which decreases exponentially to zero at a depth equal to the wave length. At shoaler depths, the flow in the water column reverses with a frequency twice that of the wave period. Under such con ditions, the velocity and acceleration at the sea floor vary with the period, depth and individual wave height. This means that on the sea floor at shallow and inter mediate depths under a strong surge, flow passes quickly from essentially still water conditions through brief periods of laminar flow to fully turbulent flow in a few seconds and then decelerates to approximately still water FIGURE 3 Variations in flow regime between unidirectional flow and wave-induced oscillatory flow. A. The velocity gradient under unidirectional (stream) flow for an assumped depth of 10 m may be postulated as shown. A laminar sub layer probably exists under most stream flow, and the average fluctuation in the velocity gradient during 1 minute of flow may be ap proximated by the range of arrow (a), or perhaps 10 cm/sec. B. The motion of water particles under progres sive (Airy) waves may be approximated by the enlarged portion of (B) at a depth equivalent to the stream flow shown in (A). Water depth in this case is assumed to be about 30 m with a wave length of similar dimensions. Note that turbulence under fully developed surge, shown only in one direction (here left to right), is achieved to the sediment surface (see text for discussion). The fluctuation in the velocity gradient ranges between 0 and an unspecified maximum, as shown schematically by the length of arrow (b). 28 VELOCITY GRADIENT: UNIDIRECTIONAL (STREAM) FLOW IOm —^TURBULENCE S3=n-SUBLAMINAR LAYER mm 10 20 30 40 50 O VELOCITY (cm/sec) VELOCITY GRADIENT: OSCILLATORY (WAVE) FLOW , Wftve type shallow IOm intermediate mm I-- 30 40 50 VELOCITY (cm/sec) conditions in one-half the wave period. According to Carstens and Martin (1963) the laminar boundary layer is negligible or absent during turbulent scouring events and thus it appears that sediment transport functions may be formulated quantitatively without reference to fluid viscosity. Natural bed forms such as ripples further complicate the motion of water over the sea floor and, for practical purposes, the net effect of these irregularities produces fully turbulent flow to the sediment surface. Under these conditions, incorporation of the Reynolds number (the ratio of inertial forces to viscous forces), which is often a consideration in laboratory model studies, is not required for scour studies reported here. It is interesting to note that the onset of exten sive turbulence has been observed to occur during the period of deceleration (Martin, unic.; Inman, 1963), a phenomenon supported by the writer's motion pictures. With longer periods of observation, velocity com ponents related to bottom slope, surf beat, longshore cur rents, tidal currents, and possibly internal wave orbital motions may influence the velocity gradient. Owing to the complex interactions of the numerous variables affecting flow regime and their nonperiodic nature, it seems im possible to arrive at a singular value for net sediment transport. For this reason, estimates of wave-induced sediment transport based merely upon the grain size of the 31 samples and the theoretical bottom velocities calculated from wave observations and depth frequently bear no similarity to observed velocities or measured sediment transport. Perhaps one discrepancy may be related to the asymmetry of waves in shoaling water, as described by Zenkovich (1967). Asymmetry in the shoaling wave profile creates inequalities in the surge velocity gradients with the result that on-shore velocities generally exceed the return flow velocity. Net sediment transport thus becomes a function of the magnitude of this inequality. According to Cook (personal communication, 1969), preliminary analyses of surge magnitude and differential sediment transport (by weight) support Zenkovlch's statements re garding the effect of asymmetric wave profiles. Examination of Table I reveals the range of values from calculations of predicted orbital velocity which are contrasted with observed velocities at the bottom during this study. It will be noted that roughly one-half of the observed values are less than those predicted by most theories. A similar discrepancy in predicted versus measured wave attenuation with depth was presented by Draper (1957). Although his measurements indicate that for deep-water waves (in general, a depth exceeding one- half the wave length) classical hydrodynamic theory is valid, intermediate and shallow water wave attenuation TABLE I Comparison of surge velocities.obtained with cali brated USC surgemeter and those predicted by wave theory proposed by various authors cited in column headings (see REFERENCES). "V-^q observed" column lists average of actual observed upper 10 percent of all surges recorded during a typical 13 minute run of the USC meter. "V10 (curve)" lists values derived from linear curve shown in Figure 7. 32 TABLE I OBSERVED, CALCULATED AND THEORETICAL VALUES, MAXIMA OF AVERAGE SURGE VELOCITIES (cm./sec) Date Average Observed USC Surgemeter Inman; Hunt U = *** H---- m T sinh kh Inman u - ‘V m T Lamb u = — ----- m 2Ccoshkh V10 Observed v10 (° UI .184 (wt s + 24 5 Nov 28.3 21.5 28.0 19.0 41.1 40.4 13 Nov 16.0 23.8 33.1 25.5 27.7 30.4 14 Nov 16.5 23.0 34.2 21.5 30.0 28.4 19 Nov 22.6 16.1 27.9 14.1 34.1 40.9 3 Dec 14.4 15.1 23.4 14.4 38.8 37.6 5 Dec 26.5 15.5 24.4 14.1 31.1 31.1 10 Dec 17.5 19.8 35.0 17.0 29.8 28.7 11 Dec 18.5 14.5 13.0 16.4 29.8 26.2 12 Dec 14.6 18.3 27.8 16.9 26.3 24.3 17 Dec 15.2 15.8 20.4 15.5 30.0 25.1 18 Dec 20.4 14.5 22.0 13.3 34.7 35.4 9 Jan 18.5 15.5 14.8 11.9 33.3 39.5 34 measured by a pressure transducer was less than the theoretical value by as much as 20 percent. Zenkovich (1967) discussed the problem of significant departures from theoretical bottom velocities measured in laboratory studies (too low) and field measurements by divers (too high). The mass of conflicting data highlights the problem of threshold velocity but does little to solve the problem of determining a common relationship which is applicable to both laboratory and field measurements. A discussion of threshold velocities assumed for this study is presented under Data Reduction and Analysis. Sediment and Fluid Interaction Under Scouring Conditions Secondary Flows Sediment particles in the vicinity of obstructions to flow are subject to displacement by vorticity and turbulence created by the obstacle. These "secondary flows" initiate scour when the drag and lift forces acting upon the sediment grains exceed the critical shear stress of the bed material. In order to study the process of scour around natural objects of irregular shape, it was necessary first to evaluate the agents and processes pro moting scour around a regular, simple shape. For purposes of discussion, this object, a right cylinder placed verti cally upon the sea floor will serve as the example. 35 It should, however, be noted that with the exception of long objects of low profile, the physics of scour are similar for all obstructions. Only the rate and magnitude of scour will vary with changes in cross-sectional con figuration of the obstacle. For the following discussion, the writer has drawn freely from papers by Masch and Moore (I960), Moore and Masch (1962), Tison (1961), and Richardson (1968). Figures 4a , 4b and 40 display the three-dimensional flow conditions which prevail around a cylindrical ob struction during a single surge. With the initiation of flow under the advancing wave front, stream lines repre senting the trajectories of water particles must separate to pass around the obstacle. Inasmuch as the velocity gradient increases upward from the sea floor, instantan eous current velocities passing the upper portions are higher than at separation near the sediment surface. At the front of the obstacle, a region of relatively stagnant water accumulates between the diverging stream lines and along the face of the obstruction, forming a vertical zone of relatively quiet water, or "stagnation line," down the front of the cylinder. More fluid accumulates in the upper portions of the obstruction than at the base where the velocity is lower. In the extreme case where the obstacle protrudes above the surface of the water, this accumulation of fluid will raise the water level and form FIGURE 4 The three flow regimes established around a cylindrical obstruction to flow. The ultimate flow pat tern under fully developed turbulent flow is the sum of these three regimes. See text for full description. A. Formation of the primary and "horseshoe" vortices by downward deflection at the up stream face of the obstruction and helical flow at the sides and rear of the object. B. The accumulation of vortex filaments, as proposed by Richardson (1968). Approaching filaments are accumulating around the up stream face of the obstruction and, in the process, are stretched and thinned. This promotes an increase in angular velocity for successively smaller diameter filaments with an accompanying increase in erosion of noncohesive sediments adjacent to the ob struction. C. The downward deflection of accelerating streamlines as proposed by Tison (1961) and the accompanying development of a wake plume and reversed local flow in the lee of the obstruction (Masch and Moore, i960). 36 37 a "bow wave" along the leading edge of the cylinder. For submerged obstructions, the magnitude of this effect is reduced since a portion of the fluid passes freely over the top of the object. However, the same effect is present to some degree in all objects which protrude above the local surface roughness features such as ripples. Under these circumstances, the pressure, or piezometric head, at the front of the obstacle increases with increasing dis tance above the sea floor. In oscillating flow, this situation quickly establishes a pressure gradient along the vertical axis of the cylinder which induces accelerat ing flow down the face of the obstruction. The result is a fluid jet which strikes the sea floor with a velocity equal to that of a fluid flowing past the uppermost portion of the cylinder (Figs. 4b and 5). Deflection of the jet is concentric to, and away from,the front of the ob struction. This outward deflection is met by oncoming fluid and the result is the nearly instantaneous genera tion of strong turbulence, or the "primary vortex" (Shen, Schneider, and Karaki, 1966) at the base of the cylinder facing the flow. This effect is shown on Plate 4. The strength of this jet oscillates in response to turbulence and vortex accumulations with the result that the pulsat ing primary vortex is maintained by strong "gusts" within the descending flow (PI. 5). The effect of gusts and turbulent jets on the sur- FIGURE 5 Schematic flow regime for surge from left to right. Length of arrows Is related (non-dimensionally here) to velocity of fluid. Increased arrow length represents increased velocity. Note the development of two vortices in the scour pit facing flow: the major, or "primary," vortex develops at the base of the scour depression, whereas a weaker "separation" vortex is established due to the presence of the cavity. This vortex is analogous to those formed in the lee of ripples. The wake plume and rising velocity gradient behind the obstruction is due to the Bernoulli effect which lowers the pressure gradient above the scour pit. 39 ONCOMING FLOW WAKE PLUME SEPARATION.’ VORTEX PRIMARY VORTEX SCHEMATIC OF VELO CITY GRADIENTS AROUND A CYLINDRICAL OBSTRUCTION o PLATE 4 a. Fully developed primary vortex along the leading edge of a 10 cm-diameter cylinder. Surge is from right to left. Note rotation of vortex (in sketch) developed by oncoming flow and frontal portion of the "horseshoe vortex" (HV) rising from the lower center of the photo toward the middle left. Rotation of primary vortex is counterclockwise as viewed. b. Fully developed primary vortex at the front of the small I-beam (flange 11 cm wide). Clear fluid immediately adjacent to the front of the object represents the descend ing Jet (sketch at right). Oncoming fluid in surge (right to left) causes counter clockwise rotation in photo. Lesser turbulence associated with the separation vortex (S in sketch) appears to left of pit lip (PL). 41 O f y \ < PLATE 5 Formation of "gusts" associated with the rapid descent of the fluid Jet down the face of a cylindrical obstruction (diameter 25 cm). Time increments are ob tained from rate of film advance through camera. a. Initial impact of vertical Jet with sediment surface. Note sediment on adjacent sea floor is in motion. Surge is from right to left. Dark stripe on cylinder is 5 X 10 cm. b. Gust of primary vortex Jet 0.05 seconds after (a). Note rising front of gust and fine radial "tendrils" of sediment. c. A second gust, 0.45 seconds after (b); "tendrils" have reformed, probably in response to minor obstructions within sediment. Photos from single frames close-up motion picture film. 45 face of a granular sediment has been discussed by Suther land (1967). His experiments with vertical jets produced a toroidal vortex which resembles the primary vortex developed in front of obstacles observed in this study. Sediment transport under his controlled conditions occurred in two stages; grains initially rolled on the surface but accelerations quickly raised the particles above the sediment surface in leaping trajectories which removed them from the point where the jet encountered the bed. This behavior is analogous to sediment behavior observed in the writer's films. It is interesting to note that organisms such as sea urchins, barnacles, scallops and mussels which are attached to, or settle on, pilings and rocks create a rough surface which is effective in disrupting the descending fluid. The establishment of a permanent biologic colony on objects greatly diminishes the effect of the jet on the sediment surface. Examples of such growth are shown on Plate 6. The Primary Vortex The relatively high velocities associated with a primary vortex generate a distinctive ripple on the inner surface of the scour pit (PI. 7). Analyses of films taken during this study revealed that the location of this ripple migrates outward from the obstruction to a position PLATE 6 a and b. Broad scour depression around tall pilings (diameter 30 cm) below Marine- land Pier off Palos Verdes, California. This, and subsequent photos of Plate 9* were taken on an extremely "quiet" day when no appreciable surge was noted. Ripple patterns around the cylindrical pilings reveal directions of flow such that streamlines form orthognals to the ripple crests. The "refraction" of ripples thus reveals the limit of in fluence of these obstructions to flow. Dark crescentic shapes are sand dollars. Prevailing surge is from top to bottom in both (a) and (b). 46 A PLATE 6 c and d. Scour depressions around Marineland Pier pilings. Yardstick was leveled prior to placement. Note heavy organic growth and attached starfish. Smoothed area within scour pit reveals effective ness of infilling and smoothing by organisms and sedimentary processes. Angle of steepest portion of pit is 17°• Later visit during storm conditions revealed pit slopes of 34° with as sociated deepening of pits adjacent to pilings. 48 49 ,*v? -£ ■ < PLATE 7 Slumping of sediment from an oversteepened primary vortex ripple is feeding sediment into the scour pit (arrow). Small "fingers" of sand are individual slump lobes. Cylinder is 10 cm in diameter. An unstable primary vortex ripple within a scour pit developed around a small I-beam (flange width, 11 cm). Lip of scour pit appears in upper right corner of photo. Failure of the ripple in (b) and extension of slump "fingers" into scour pit. This photo also reveals a slump scar above the fingers which results from caving of the initial ripple. Generation of multiple and concentric ripples and scars produces a series of concentric ripples within an individual scour pit. 52 midway between the obstacle and the lip of the scour de pression. This situation continues until a critical lip distance is attained, beyond which the primary vortex ripple remains static as the pit continues to widen. The maximum distance of this ripple from the surface of the obstruction is crudely related to the diameter of the obstacle (Fig. 6). Although the location of the primary vortex remains static, it continues to affect the growth of scour pit diameter by maintaining the slope of the inner pit at the angle of repose for a given sediment. Measurements of maximum slope angles in developing scour pits indicate that for the range of sand-size materials under consideration the angle of repose is 35° • Excavation of the toe of the pit slopes by the rotating primary vortex perpetuates an unstable slope which periodically slumps into the base of the pit (PI. 7). Sub sequent surges remove most or all of this slumped material and, together with turbulent suspended transport, the pit continues to deepen and widen in steps related to the frequency of strong surges. As this step-wise enlargement progresses, a series of concentric ripples may develop on the slope of the scour pit. These are miniature slump scars which migrate outward to the lip of the scour pit as successive slump events occur. Published photographs of scour pits in snow developed around obstacles exposed to unidirectional air FIGURE 6 A. Growth cycle of a scour pit to terminal scour conditions. Within the initial pit, the primary vortex migrates away from the ob struction at a rate such that its distance from the obstacle (X) is consistently one- half the distance to the pit lip (L). B. The terminal value of (X) for cylindrical obstacles is related to the cylinder diameter, as shown by the curve in Figure 6B. The extreme scatter of points is due to the varied cross-sectional configurations. How ever, a reduction of (X) values with in creasingly streamlined cross-sections is apparent in the lens and block orientations with sharp edges facing the flow. Large I-beam shapes also exhibit low (X) values due to the short radii of curvature of streamlines around the sharp corners con fronting flow. 53 54 < FLOW 0 35, GROWTH OF P IT TO TERM INAL SCOUR CONDITION A E o CL LlI h- i i i 2 < I — o i L l “3 CD O -20 55 / o j o dI q a a / / / / 05000W 0 / X VA LUES (cm) O B J E C T S 0 -C Y L IN D E R S a ROCKS 0 - I - BEAMS BLOCKS 0 - LENS AVERAGE FOR CYLINDERS g flow (Allen, 1967), shallow subaqueous flow (Karcz, 1968) and deep sea photographs show a major ripple, assumed to be the primary vortex ripple, at a distance of approxi mately one-half the scour pit lip distance. This suggests that, under unidirectional flow conditions, the lateral extent of a scour pit is limited to twice the distance of that achieved by the primary vortex. Thus, the continued enlargement of the scour pit lip beyond twice the primary vortex distance may be a unique feature of scour under oscillatory conditions. In addition to the primary vortex, a weaker vortex with similar rotation develops on the upstream side of an obstacle. This vortex is related to separation of flow at the lip of the scour depression (Pigs. 4a and 5) and is analogous to the separation vortex developed behind ripples. When the scour pit is relatively small, this vortex reinforces the primary vortex but with increasing pit width, it migrates up the slope and becomes a separate disturbance, contributing to slope instability. Although weak in comparison with the primary vortex, turbulence associated with the separation vortex is still effective in lifting sediment from the pit slope. Once suspended, the sediment is subject to entrainment in oncoming flow and may be swept from the scour depression. Consideration of the velocity and turbulence generated in the formation of the primary vortex indicates that the bulk of sediment transport is achieved by this process. Not only does the vortex erode materials from the immediate base of the obstacle, but the severe turbulence raises much of the sediment to a height where oncoming flow can entrain it and sweep it past the sides of the obstacle, into the turbulent wake zone, and well beyond the scour depression. Other Vortices and Pressure Gradient Effects Richardson (1968) described the generation and maintenance of scour induced by vortices which develop near the sediment surface, primarily in response to fluid drag effect. Vorticity is also augmented by ripples on the bed. If the axes of rotation of these vortices are connected by a line, this line and the diameter of the vortex form a "vortex filament" (Pig. 4b). Plow over the sea floor contains trains of such filaments which ac cumulate in front of an obstruction to flow as the stream lines separate to pass around the obstacle. In this situation, the filaments wrap around the obstacle and, in the process, are stretched. Inasmuch as they cannot be cut, even by a sharp leading edge, the stretching leads to a reduction in filament diameter and consequently an increase in the rate of rotation within the filament. The resulting increase in vorticity close to the bed initiates 58 scour by lifting grains from the sediment surface. Slip page of the filaments around the sides of the obstacle en trains suspended sediment and, as in the case of the primary vortex, grains are quickly carried beyond the edge of the existing scour depression. Turbulent transport near the bed is also accelerated by a diving motion of streamlines as they separate to pass the obstruction (Tison, 1961). As these streamlines pass the obstacle, they begin to rise (Pig. 4a) and lift sus pended sediment above the rim of the scour pit. Results from flume studies indicate that acceleration of stream lines adjacent to obstacles increase the velocity to twice that of the unobstructed flow (Moore and Masch, 1962; Carstens, 1965). This promotes sediment displace ment adjacent to the obstacle before the unobstructed flow exceeds the threshold velocity of local bed materials (PI. 8), and serves to maintain movement after the velocity falls below the critical value required to initiate move ment. Inasmuch as the fluid Jet down the front of the cylinder does not reach the sea floor instantaneously, transport of sediment in a plane roughly parallel to the sea floor will commence before that initiated by the establishment of the primary vortex (PI. 9b). At the rear of the cylinder, the pressure gradient is the reverse of that established at the front of the obstacle. The crowding of streamlines is carried beyond PLATE 8 a. At the commencement of surge, the velocity of the fluid immediately adjacent to an ob struction is accelerated by separation of streamlines. Note that sediment is in violent motion at the base of the cylinder (18 cm diameter) but adjacent sea floor bed- forms have not yet been set in motion. This object has just been placed into the sediment, and no significant scour pit has developed. Note turbulence beginning at the "rear" of the obstruction. Surge is from right to left. b. A weak surge, incapable of displacing sediment from ripples, has exceeded the critical shear stress within the scour pit. A weak plume has developed in the wake region at the left side of the cylinder, and a weak primary vortex has created incipient slumping at the "front" of the object within the scour pit. Asymmetry of ripple crests and turbulence in the vicinity of the obstruction reflect the surge from left to right. 59 60 * U . t ^ . » * r * r gggg ' i ~ V f PLATE 9 a. Scour depression around 10 cm diameter cylinder. Sediment median diameter = 0.17 mm, Santa Catalina Island 10 December 1968. Note primary vortex ripple at left (seaward) side of object. Coarser lag deposit left by previous surges is visible at base of far side of cylinder. b. Initial surge from left to right begins to establish a vertical plume in the wake region "behind" the cylinder. Bedload out of pit commences as a sheet of entrained sand responds to acceleration adjacent to the base of the obstruction. Note that sand is in transport within the pit but not on the ad jacent sea floor, although asymmetric ripple crests indicate incipient motion prevails over the entire sea floor. An effective primary vortex has not been established. 61 62 PLATE 9 c. Suspension carpet has begun to form, and primary vortex and horseshoe vortex in their initial form have developed. Entire sea floor is in motion, and plume in wake is grow ing. Time = 1.3 seconds after (b). d. Fully developed classical scour around a cylindrical obstruction 2.0 seconds after (b) Suspension carpet still displays ef fects of bed roughness (ripples) but tails of horseshoe vortex are lifting sand above carpet in turbulent suspension. 63 64 the obstruction with the result that a wake of turbulent but relatively static water forms behind the obstacle. Because the velocity gradient increases upward, the pres sure gradient established by this imbalance causes fluid to rise behind the cylinder. This behavior is determined by the Bernoulli Principle which states that the fluid pressure reaches a minimum where the velocity is highest. The presence of this gradient is revealed by both a strong turbid vortex, or "sediment-laden tail," which rises be hind the obstacle and by the sediment on the surface be hind the cylinder which is sucked Inward to the base of the obstruction. Although a similar separation of flow lines, which reduces velocity and increases pressure, also occurs at the front of the obstacle, the Bernoulli effect cannot be established along the surface of the obstacle facing the flow because accumulated fluid and vortices in the stagnation zone overcome this effect. The cumulative effect of these turbulent flows is shown in Plates 9, 10 and 11, and Figure 4a, 4b and 4c. Dyed Sand Study Qualitative studies employing dyed sand as a tracer were conducted to determine deposltional patterns for sedi ments eroded from scour pits. These observations confirm Bradford's (1966) findings that sediment displaced from the scour pit is deposited essentially in two lobes in front of PLATE 10 Incipient "horseshoe vortex" forming "behind" cylinder 5 cm diameter. Primary vortex is visible along the leading edge of the object at the left (PV in sketch). Sheet of sedi ment is being ejected from the pit in the lower right of the photo. Surge from left to right; photo from single frame close-up motion picture film. Cylinder in (a) O.57 seconds later. Primary vortex is fully developed, as is the "horse shoe vortex." Rotation of the near element of the horseshoe vortex is shown in sketch at right. Turbulent plume at edge of flat plate, flow right to left. Low pressure area behind flat plate (flange of small I-beam 11 cm wide) is identical to Bernoulli effect behind cylinder in (a) above, and (b), Plate 11. Note multiple ripples associated with primary vortex at far side of flat plate. CPV PLATE 11 a. Initiation of surge around cylinder 10 cm in diameter. Sediment is in rapid transport adjacent to cylinder wall at base of scour pit and is starting to rise in the lee wake. Surge is from left to right. Photographs are from single frames of close-up motion picture film. b. Cylinder in (a) 0.22 seconds later. A sheet of sand is being expelled from the pit at' the bottom, and a distinct plume is rising along the "back" of the cylinder. Sediment at middle right is being drawn into the low pressure region at the base of the scour pit. c. Cylinder in (a) 0.61 seconds later. Strong plume has developed behind the cylinder and the two elements of the "horseshoe vortex" have begun to form (broad flat arrows). 68 70 and behind the obstruction in the directions of oscillatory flow. These displaced sediments assume the configuration of local bed forms, and do not form the noticeable mound commonly found downstream from obstructions to unidirec tional flow. Summary of Scour Phenomena The primary vortex, diverging streamlines and vortex filaments combine to form a U-shaped vortex often termed the "horseshoe vortex," which is wrapped around the obstruction (PI. 9d). The sediment plume between the two elements of the horseshoe vortex is efficient in transport ing sediment beyond the limits of the scour depression and, if turbulence is severe, sediment may be raised to a height where it remains in suspension after the flow has reversed. In this case, it may fall back into the pit or be carried beyond it in the opposite direction. In combination, the turbulence, and increased flow velocities, collectively termed "secondary flows," create a flow regime which will quickly erode a depression around an obstacle. Where sorting of the natural materials is poor, the coarsest grains may not be swept from the pit base and a lag, or "armor" deposit develops. This armor ing effect produces a net decrease in sediment transport rates as demonstrated by Hooke (1968) in flume studies. The reduction of"scour accompanying an increase in grain 71 size has been recognized by engineers as one technique in the prevention of scour around bridge and pier pilings. With the exception of long horizontal cylinders (a special situation not discussed in this report), the scour depressions around any object exposed to oscillatory flow ultimately assume a generally circular outline. This condition holds for clusters of objects as well as indivi dual obstructions. Studies of cylinder clusters revealed that, after the initial individual pits coalesce, the pit rapidly assumes a circular pattern. The diameter of such a pit is equivalent to that which would develop around a cylindrical object having a diameter equal to the com bined width of the cluster elements. This "neighbor effect" often creates large depressions around clusters of cobbles and boulders. The sloping sides of scour pits are initially in clined at the angle of repose during active scour, and progressive enlargement of the pit requires the removal of a progressively larger volume of sediment. The ulti mate pit volume is achieved when the ratio of sediment transport into and out of the depression approaches unity. Due to various fluctuations in the flow regime, a static pit is never really attained. During quiescent periods, biologic reworking of sediment (PI. 6d), slumping and similar processes of in-filling,smooth the depression and the slopes of the pit are reduced to roughly half the angle of repose. This process was observed at the Marlneland Pier under storm conditions when scour slopes originally- measured at 17-18° (PI. 6) were observed to have deepened and attained slopes of 34-35°• This alternating cut and fill effect in the near-shore region is probably a seasonal phenomenon related to annual beach cycles and frequency of storms. 73 DATA REDUCTION AND ANALYSIS Films and Photographs Objects placed Into or on the sea floor were painted with a grid which provided the scale factor required for measurements taken from motion pictures and still photo graphs. The latter provided data for overall scour pit geometry while the time lapse motion pictures, when pro jected frame by frame at l/2 full scale, permitted the plotting of scour growth rate curves based upon the time interval of single-frame exposures. To avoid problems of parallax, a yardstick was inserted in the field of view at bed level for several frames. The image was projected on a sheet of paper and the successive positions of the scour pit lip marked with a line when detectable outward displace ment of this lip was observed. The calculation of volume approximations from such curves appears in a later section. A close-up lens was employed to analyze sediment behavior inside the pit during scour events. These films, when projected as single frames and in slow motion, pro vided insight into sediment and fluid interactions during brief and violent displacement of materials adjacent to the obstructions (Pis. 7b, 7c; 10, 11). Water Motions at the Sea Floor 74 Theoretical calculations of wave-induced surge on the sea floor include wave length as a dependent variable. This factor is difficult to measure from a vessel and, as noted by Inman (in Shepard, 1963, p. 58) a useful approxi mation for the wave length at any water depth is provided by Eckart (1952): where: L = wave length at any depth 2 Lj = deep water wave length = 1.561 m where T = wave period in seconds tanh = hyperbolic tangent kd = wave number = 2'7r/L( i h = water depth. Once the wave length for a given site is calculated, it may be substituted in various equations yielding the theoretical maximum orbital velocity at the sea floor (umax)» The 'two most frequently used equations for Umax are provided by Inman (in Shepard, 1963, p. 61): tanh k^h (1) U . H (2) max t sinh kh where H = wave height 75 and Lamb (194-5): Umax = gH (3) max 20 cosh kh where: 0 = phase velocity of wave train g = acceleration of gravity. Solution of these equations is facilitated by Tables of sinh kh and cosh kh for d/L listed in Wiegel (1964). Cal culations based on Lamb's equation (3) are shown in Table I. Using equation (2) above, Hunt (1961) prepared a set of graphs which permit the estimation of wave-induced surge velocities when wave period, height and water depth are known. Records of bottom velocities obtained with the USC surgemeter were divided into velocity decades of 10 cm/sec increments. Prom these plots, it was possible to derive average bottom velocity for comparison with theoretical average velocities derived from equations (2), (3) and Hunt's graphs. Comparison of values are presented in Table I. This table reveals the magnitude of dis crepancies between observed bottom velocities measured at about 30 cm above the bottom and theoretical velocities obtained from calculations based on wave and depth measurements. Similarly, calculation of theoretical orbital diameters by use of the relationship: Orbital Diameter = ___2___ (4) sinh kh yields values on the order of 0.5 m (Table I). Observa- 4, i tions of kelp fragments, sand clouds and other objects in motion above the sea floor commonly revealed excursions of 2-3 m. Strong surges usually exceeded this distance. At present, the cause of these discrepancies is not known. Fluid Velocities and Sediment Parameters Inspection of the observed average bottom velocities noted in Table I reveals that in most cases they lie below reasonable threshold velocities for sand-slze materials (Fig. 2). Inasmuch as sand was observed in motion at all sites, a velocity in excess of the average was required for calculating sediment transport rates. Study of films and field observations indicate that sand was never in motion for more than 10 percent of the time. In an attempt to relate this maximum rate to recorded bottom velocities, the upper 10 percent of bottom velocities was calculated from the surge spectra for each station. Comparison of this velocity, termed V-^q , with the weight of sand moved per hour, determined by Cook (personal communication, 1968), produced a scattered but consistent relationship approximated by the line in Figure 7* The equation of this line provided an empirical relationship: V10 = 0.184 (weight sand) + 24 (5) which was used to estimate V-^q for those studies conducted prior to the construction of the USC surgemeter. In order to compare inertial (flow) forces tending FIGURE 7 Correlation of the dry weight of sand (gms/hr) with the upper 10 percent of the observed velocity of bottom surge <V10 in cm/sec). A best fit curve yields the equation shown in the figure, and this relationship was applied to sediment transport values, in weight of sand per hour, obtained prior to the construction of the USC surgemeter. 77 60 50 E o o 40 >* >- O O 30 _l UJ > 20 0 20 40 60 80 100 120 DRY WEIGHT SAND (gms/hr) 00 79 to displace sediment grains with gravitational forces tend ing to hold them in place, it was useful to employ a sediment Froude number, or sediment number (Oarstens and Martin, 1963). This relationship is: Ns = _ V _ (6) \f[ s—l) gd where: Ms = dimensionless sediment number V = velocity, here V1Q in cm/sec S = ratio of specific weight sediment to specific weight of fluid g = acceleration of gravity d = median grain size. Analyses of obstructions to flow in laboratory flumes have shown that accelerations of the fluid diverted around an obstacle raise its velocity to twice the "ambient field" velocity, here the V-j_g« Consequently, in computing values of Ns, the velocity component V10 was doubled. Contrasting the Ns (for 2V^q ) with the submerged weight of sand derived from Cook's measurements yields two curves (Pig. 8) which indicate general agreement between sand weight, grain size and the Ns values. The net effect of Ns is more apparent if the relationship (Ns2 - Nsc2) = "effective Ns" (7) is used in place of Ns. Henceforth, this factor is termed FIGURE 8 Relationship of submerged weight of sand to the "sediment number," (eq. (6) in text, p. 79). The minimum value of Ns shown as 3.9 on the abscissa is derived from critical shear stress experiments and represents"thres hold velocity" (see text, p. 76). 80 (for 2 V . 0 • 1 I ** n o 3 Q • CO 0 ) Q o 3 CO Q. O CO 3 WEIGHT SAND (submerged, gms) o 100 the effective Us. Use is the value of Us for a velocity at critical shear stress (threshold velocity) conditions, and (7) thus provides the value of the effective Us above critical shear stress conditions. Both quantities are squared to provide proportionality between the ratio of inertial and gravitational forces (Carstens and Ueilson, 1967). For the purpose of this study, a constant value of 3.9 for Use was used. This value was derived from labora tory studies employing oscillatory flow and natural sands (Carstens, 1965). Por grain size of 0.19 mm, and specific gravity of 2.67, the velocity of Use is 21.8 cm/sec. This value is shown as a point on Figure 2, and from its posi tion on this graph, it appears to provide a suitable ap proximation for values of Use. It will be noted in Figure 8 that the curve approaches the origin at an Us value (using 2 V1Q) of approximately 4, a factor which supports the choice of 3.9 for Use. Plotting the effective Us against weight of sand moved per hour yields a curve (Pig. 9) similar to Figure 8, but one which includes the balance of forces acting upon the sediment. Although an increase in the magnitude of the ef fective Us is primarily due to higher surge velocities, an increase in the frequency of strong surges is also ef fective in raising its value. Although water depth is a factor, the number of strong surges per minute is primarily a function of wave height and period, the former fluctuat- FIGURE 9 Relationship of the "effective Ns" to the net weight of sand transport in grams per hour as submerged weight. See text, p. 82:and eq.. (7). 83 NET W EIG HT SAND (submerged, gms/hr) o o o o o o 85 ing with variations in the combined wave trains forming local swell conditions. If the frequency of strong surge, here defined as surge velocities exceeding 20 cm/sec (an approximation of the velocity for Use), is contrasted with the effective Ns (Pig. 10), the frequency effect is ob vious. Although the scatter of points is pronounced, the general trend of increasing effective Ns with increasing surge frequency demonstrates the relationship between the two parameters. Sediment Transport Rates and Scour Development In his study of contour maps of scour pits developed around a cylindrical obstruction to oscillatory flow, Carstens (1965) suggested that the configuration of the initial depression resembles the inverted frustrum of a right circular cone having a base diameter equal to the obstacle diameter and a slope equal to the angle of re pose for a given material. Field measurements and study of photographs and films obtained in this study support this approximation, but a condition closer to reality may be achieved by introducing a factor to account for the flat floor of the pit adjacent to the object. This flat portion results from the severe turbulence associated with the primary vortex. Applying the data from Figure 6, the calculation of the frustrum volume was modified to include this reduction in volume not incorporated in Carsten's FIGURE 10 Relationship of "effective Ns" to the strong surge frequency. See text, p. 82-85. 86 87 -12 -10 o0 “0 500 100 50 sc Strong surge frequency units = surges/min > 2 0cm /sec 88 calculations. The volumetric and linear relationships in the equation for volume at a given increment of growth are shown in Figure 11, equation (8). The absolute volume of the frustrum was based on an angle of repose at 35° which was measured in both still and motion picture films as the maximum angle at which the granular sediments would stand before slumping occurred. Employing a maximum value of the scour pit lip dis tance from the object (value of L in Fig. 6), it is pos sible to calculate the volume of the scour pit knowing only the L and Xmax values. A computer program was pre pared to print out the frustrum volume for increasing values of L in increments of 0.5 cm. By varying the ob stacle diameter, a set of volumetric tables was computed which provided the incremental volume changes necessary for calculating sediment transport rates under scouring conditions. Comparison of the tables with curves of the growth rate of L obtained from single frame analysis of the time-lapse films permitted conversion of these data to curves of volumetric change. Analyses of the volume curves provided two measures of sediment transport rates based upon: Sediment transport rate Qs = dV dt (9) in cnrVmin and: FIGURE 11 Approximation of scour pit geometry through use of volumetric relationship for the frustrum of an inverted right circular cone incorporating basal area dictated by progressive growth of X values to X_Q„. Computer program I I I q a provided volumes in increments of 0.5 cm L distance, here shown as (R-j_ - r). 89 90 F R U S T R U M Volume = TT -{£• ( R2 + R, R2 ♦ R 2 ) r* T H J . CYLINDER Volume = TTr2 H 35 Volume S C O U R PIT Volume = TT-g-(R2 + R,Rt+ R2)-1 T r2 H (Eq.8) L = 0 —> L m aX in 0.5 cm increments X = = L max R, = r + L R ,, = r + x H = ton 3 5 ° ( L - x) or 0 . 7 0 0 2 l ( L - x ) 91 Volume growth rate Qv = d# dt (10) in percent/min. Determination of Qv eliminates the scale factor inherent in Qs calculations and provides insight into volumetric rates for all obstructions regardless of size and cross- section configuration. Separate equations were written to provide computer solutions for the scour volume developed around the square object and horizontal cylinder 1 m in length. It should be recognized that the volumes of scour pits computed for given obstructions are absolute volumes and, as such, neglect porosity. An adjustment may be introduced by reducing a given volume by: True volume of sediment = 0.57 (calculated volume). (11) The value 0.57 is obtained from the average porosity of shelf sands cited as 43 percent by Hamilton (personal com munication, 1968). Summary of Scour Analysis By means of the equations presented in this section and the measurements of scour development based upon visual observation and time-lapse photography, it is pos sible to make quantitative approximations of the rate and magnitude of scour development. The inclusion of refined geometrical relationships resulting from measurements of primary vortex phenomena in scour analyses provides a realistic estimate to scour ing processes. The interpretation of these processes is presented in the following section. 93 INTERPRETATION OP DATA Lateral Extent of Scour The primary objective of this study was the deter mination of the effects of various object geometries and orientations on scour development. Shapes and orienta tions are shown in Figure 12 with dimensions rounded to the nearest cm. Studies conducted prior to construction of the time-lapse camera (in operation by October 1, 1968) were simple measurement programs around various objects deployed at a given site. The maximum scour noted at the end of each study period was plotted against the object's diameter or width, with the result that a general curve of scour dimensions was obtained (Fig. 13)• The scatter is due to differences in local conditions for a given date. The cylindrical piles having a 30 cm diameter are those sup porting the Marineland Pier and their position on the graph reflects long-term adjustment to local conditions. If the test objects of approximately the same diameter (25 and 33 cm) had been left in place at the study sites for a significant period of time (perhaps several weeks) they, too, would have higher ordinate values in Figure 13. In addition, the pilings may be considered to have infinite height, a factor which will be discussed in a later section. FIGURE 12 Geometry of test objects employed during this study. A broad double arrow indicates orientations used in deployment of objects under various test conditions. 94 LENS ROCKS BLOCKS O CLUSTERS Q Q o o 10 20 30 OVAL 40 50 SCALE (cm) VO VJ1 FIGURE 13 Relationship of cylinder diameters to scour pit lip distance (L). A linear relationship between variables is evident up to diameters of 18 cm. Beyond this point, curvature (dashed) reflects time factor which did not permit full scour to develop around larger objects. Points in upper right portion of graph display scour pits measured around pilings supporting the Marineland Pier at Palos Verdes. 96 DISTANCE TO PIT LIP(L, cm) o o o o oo “ o o O •• • CO oo r o pier pilings / N VO In order to reduce scale effects induced by the various sizes of objects, a second plot of scour extent was prepared which contrasts the ratio of pit diameter to object diameter for various test objects (Pig. 14). Here the scour rate becomes significant because the duration of the study period did not permit full-scale scour to occur around the larger objects. If the object diameter is large compared to the length of excursion (orbital path length) of water particles in the given surge, there may be insufficient time for full development of primary vortex and wake plume turbulence. This is reflected in the lower ordinate values (Pig. 14) for larger diameter objects. Again, the duration of studies precluded the full develop ment of scour, as shown by contrasting long-term values for objects left in place for up to 51 hours with the 30 cm pilings at Marineland Pier. Time studies of scour development were conducted on several occasions prior to construction of the time-lapse camera. One such study was designed to observe differences in scour rates around similar obstructions with varying height. The results, shown in Figure 15, reveal a pro nounced acceleration in rates of scour development with increasing cylinder height. However, it will be noted that at the end of the first 60-minute period, the lower objects had begun to approach the volume of material scoured from the taller cylinders. Examination of the FIGURE 14 Relationship between object diameter and ratio of pit diameter to object diameter. Line reflecting average Indicates decreasing values of the ratio with increasing object diameter. As in Figure 13, this reflects a time factor which was too brief to permit terminal scour development around larger obstructions. 99 -8 2 < - 7 O I- O UJ C D P -6 - 5 - 4 . < Q - 3 H Q_ -2 O I- < cr -i 20 25 30 35 FR E QUENCY - many O - few >*✓“ one — average OBJECT DIAM. (cm) A — maximum pits developed around test objects in long-term studies (>24hrs) i —1 o o FIGURE 15 Curves of scour volume in relation to object height for uniform diameter cylinders. Note that taller cylinders promote rapid scour shortly after placement, but that shorter cylinders approach values for taller cylinders with increasing time. Examination of scoured pits 24 hours after placement revealed that all objects, regardless of height, had generated scour depressions of equal extent and volume. 101 500 DATA 5.1 cm cylinders Md=O.I7lmm 400 ^ 0= 28.7cm/sec eff.Ns= 110.8 o o UJ 2 ID _1 O > 300 200 ••'---. -r.-I-.-. •: • - - 100 FLOW Qv (%/min) Height (cm) Qs (cc/min) 2 5 .4 I 9 .0 2 .7 6 .3 7 4 /2 0 4 3 /2 0 7.6 5.5 TIME(min) 5 io i _________« t i i i i i i 103 three remaining cylinders 24 hours later revealed uniform pits of 23 cm in diameter had formed around all three ob jects. This suggests that given sufficient time low ob jects are scoured as efficiently as tall objects. Consideration of the action of the primary vortex led to experiments with cylinders placed at an inclined attitude 45° to the bed. After 2 hours of scour under heavy surge conditions, there was little difference be tween the extent of scour developed around inclined cylinders and the vertical controls. It must be assumed that inclination of an obstruction to flow does not materially affect the generation of secondary flows produc ing scour. Permanent Stations In an attempt to study long-term scour effects around the test objects, a series of four stations were established on November 12, 1968 off the windward side of Santa Catalina Island. Cylinders having diameters of 5> 10, 18, and 25 cm, and concrete blocks measuring 38 by 30 by 19 cm were placed at depths of 7> 14, 21 and 28 m. Re peated visits to the sites provided Insight Into prolonged scour at various depths. The location of the 24 m station was lost when the buoy and attached block were displaced by human agents after November 14, and the 7 m station was likewise lost shortly thereafter. Data from all observa 104 tions are shown in Figure 16. The loss of some objects from undisturbed stations is attributed to fouling by dis lodged kelp and/or organisms and complete undermining by scour of some objects. Several features of the curves on Figure 16 are apparent: 1. The pits surrounding the 18 and 25 cm cylinders show similar fluctuations at 21 m (Day 30), sug gesting that infilling and smoothing of the pits by combined benthic organisms and weak sediment transport operated at equivalent rates. Inasmuch as the depth of scour around the 25 cm cylinder continued to increase during this period, the smoothing effects were apparently not effective immediately adjacent to the cylinder. 2. The curves for the smaller cylinders at the same station do not indicate that infilling occurred during the same period. This may be a reflection of the scale factors between small and large objects previously discussed. 3. Blocks, which also served as anchors for the buoy lines, were initially buried to one-half their height (8 cm). Within 48 hours, scour had effectively undermined these blocks at the 7, 14 and 21 m stations and they were found FIGURE 16 a. Combined curves for test cylinders placed at stations off Santa Catalina Island. Note the distorted time scale on abscissa. b. Scour pit development with time for concrete blocks. Block at 40 m station was completely buried 30 days after placement. 105 106 40 - 3 0 O' •20 18cm cyl- 21 m 18 cm cyl- 14m 25 cm cyl- 21 m 20 -0 io- 14m 21 m 28m 10.2 cm CYL. -20 14 m 28m 5.1 cm CYL. 65 30 TIME (days) 30 -20 14 m 28m BLOCKS 30 8 65 T I M E ( days) B. lying on their sides in large depressions. Very little drag could be provided by the small cork buoys and neutrally buoyant polypropalene line so it is assumed that the blocks were not pulled over by fouling of the lines. That the blocks were in the exact location of Initial burial argues against human intervention. 4. After 30 days, the block at the 14 m station was found to be completely buried. Probing indicated the minimum depth of burial was 5 cm. At the same station, as at all stations during the entire study, examination of reference stakes adjacent to the objects revealed negligible fluctuations of the sea floor level. Differences of 2-3 cm were attributed to the random distribution of ripples. Thus, the burial of the blocks was not due to local de position which raised the level of the sea floor, but to excavation and ultimate burial of the objects by scour processes. Platform Studies To investigate scour effects around large objects, studies were conducted at two offshore platforms, Union Oil Company's Platform Eva off Huntington Beach (depth 21 m, supporting legs about 2 m diameter, sediment Md = 0.112 mm) 109 and USNUWC's Oceanographic Platform off Mission Bay (depth 21 m, supporting legs about 1 m diameter, sediment Md = 0.52 mm). At Platform Eva, the supporting piles are sheathed with a heavy growth of mussels to a depth of 18 m, but the supports are relatively clear for the remaining 5 m. The sea floor under the platform, consisting of a thick carpet of mussel shells, starfish and crabs, rises several meters above local bed level. The fact that no scour effects were observed at any supports may be attri buted in part to the blanket of shell debris. Although the supports were clear of attached mussels to an elevation about 3 m above the sea floor, their absence is attributed to such ecological factors as light and turbidity rather than to mechanical abrasion. An identical vertical dis tribution of mussels was reported for a similar platform off Summerland, California (Carlisle, Turner and Ebert, 1964). The study at the USNUWC platform spanned 2 days and included deployment of most test objects and 12 hours of time-lapse filming of sea floor microrelief. Low swell and coarse sediment precluded full development of scour and only minor pits were observed after 27 hours of ex posure. Discussions with Navy divers who conduct experi ments from the tower on a regular basis revealed that scour never has been observed around the tower legs, even under heavy swell conditions. They attribute this to the coarse- 110 ness of the local sediment which consists of the common relict "red sands" frequently encountered at similar depths off southern California. Comparison of Scour Under Unidirectional and Oscillatory Flow The controversial question of the applicability of unidirectional flow effects to scour on the sea floor prompted a study designed to compare scour developed under both conditions. A survey was made of tidal flat sedi mentation on the Gulf of California at Percebu, Baja California, Mexico. Test objects used in sea floor studies were transported to the site and deployed at flood tide. At the study site, maximum water velocities at the surface were clocked at 70 cm/sec. Water clarity was such that measurements of scour pit dimensions could be made during the period of flow was well as at full exposure at low tide. The results appear in Figure 17, where it can be seen that the slope of the best fit curve provides a pit/ object ratio of approximately 2. This ratio appears to hold regardless of object size. This supports Brooks' statement (personal communication, 1969) of general con sistency of similar measurements under laboratory condi tions. In the enlarged portion of the graph it can be noted that at Percebu, pit dimensions during flow were greater than the same measurements obtained after cessation FIGURE 17 Combined data from sea floor scour studies and tidal flat investigation of scour under unidirectional flow. Note that sea floor scour generates a wider pit than does stream flow. Inset reveals relationships be tween rocks and test objects under both conditions. See text for discussion. Ill ' 112 "150 60— 50 -100 4 0 9 T 30 -50 I Q - 30 20 40 50 60 OBJECT DIAM.(cm) RANGE OF ► PITS ON SEA FLOOR n TEST OBJECTS DURING FLOW o TEST OBJECTS AFTER FLOW • ROCKS AFTER FLOW A ROCKS SEA FLOOR X TEST OBJECTS SEA FLOOR AVERAGE' 113 of flow. This factor is attributed to waning turbulent flow and a reduction in sediment transport rather than in filling by diminishing bed load or slumping. If these factors were effective, one would expect a gradational deposit of finer sediment within the pit and a smoothing of the pit lip and primary vortex ripple. Neither condi tion was observed. The finest example of unidirectional scour available to the writer is shown in Plate ll(x). This pit reveals all the classic features previously described, including a secondary pit behind the obstruction which represents the point of "reattachment" where the turbulent fluid raised in the horseshoe vortex again encounters the bed. In order to compare scour measurements made under oscillatory flow with those at Percebu, the range and average for the same objects were reproduced from Figure 13. For diameters less than 14 cm, the dimensions of the average pits developed under oscillatory flow are remark ably close to those measured under unidirectional flow at Percebu. The relationship apparently does not hold for larger diameters. This tends to support the contention that scour under oscillatory flow is similar to that under unidirectional flow, at least for the small cylinders. However, they are similar only to the degree that their extent and approximate values are related. The sediment PLATE 12 The scour pit developed around the obstruction in this plate displays all the classic features of uni directional scour. Pit width is approximately twice the diameter of the rock. Orthogonals to the ripple pattern in the wake zone reveal the distribution of streamlines in the flow behind the rock. Higher velocities and turbulence have created these ripples whose height and wave length greatly exceed those visible outside the pit. Also note the depression immediately in front of the observer's feet. This pit represents the point where flow within the horseshoe vortex impinges upon the bed after passing the obstruction. Laboratory experiments show that at this point the fluid strikes the bed as a jet which is capable of producing a secondary pit. Plow is from upper left to lower right. Picture taken at Cornwall, England, by D. J. Leeds. 114 1 115 . . V I *■ « v * ‘ » / * A * V f ’ t * PM < . " ' H V 1 £'*****'*{ p * * > - M m m . W W m m * * * * S i * .............. „ . . r . ...V .“i . : ' ^ ' “ ;: . r • f ‘ V V^, ' ■ ' * • ; ' • a's*ta ; --rxv. .....• . . '. ^ r r t & ' , • J , v • . • ,. . : • -.• " " • /• • \ • , . - .■ - » V ' - . - e - ;:r.I"'*? V * 1 , • • , - i , * * y- , j , ■ • ;* ir ’ /> • ■ * 'y ■ . • , ' <»:.■••. ■ , . . , > • ! • j ' . - . r v . - ; . * - V * ; - , . i . ,* yi*. ^ c * S - W f.-V I v I iff * r , /jr ^ * *• r # f ' K> 1 & * i ' f f a \ 116 transport rates are different, and the configuration of the unidirectional pit with its tail and vortex ridges is quite asymmetric compared to the essentially round pit around cylinders exposed to oscillatory flow. As cited later, large objects of low profile (boulders) on the sea floor may display pits of pronounced asymmetry which super ficially resemble unidirectional scour. In laboratory studies, Carstens (1965) noted that if the ratio of the depth of flow to object diameter is less than 2, free-surface effects (the free surface is the upper surface of the fluid) such as the "bow-wave effect" may cause significant acceleration of scour processes. At Percebu, the fluid depth ranged to zero so this condition was met for all obstacles. This factor probably accounts for the upward displacement of points representing measurements under conditions of fluid flow. The rapidly increasing differences between values for rocks and the larger cylinders cannot be attributed to "free-surface effects" acting upon these objects on the sea floor as the depth/diameter ratio for even the largest cylinder (33 cm) was never less than 20. The dif ferences can only be attributed to significant variations in the two flow regimes, as discussed earlier (Pig. 17). The few measurements of rock scour on the sea floor tend to depart from the unidirectional curve at about the same point as the test objects, suggesting that this effect 117 applies to natural objects as well. Their position on the plot below the averages for test objects probably re flects the low height of the rocks relative to the artifi cial (and taller) objects. Intuitively, it might be suspected that the extent of scour would be a function of grain size and velocity. However, a plot of terminal scour pit distance versus median grain size merely supported the observation that scour extent is primarily a function of object size. Con sideration of the velocities and associated accelerations occurring in scour events at depths included in this study leads to the conclusion that they are of sufficient magni tude to place essentially all bed materials in suspension. Continued slumping of the pit walls constantly feeds new material into the region subject to extreme vorticity and the frequency of strong surges assumes a dominant role in the removal of displaced materials (Fig. 10). Laboratory studies under unidirectional flow (Shen and others, 1966) indicate that scour depths are independent of grain size up to a median diameter of about 0.6 mm. However, at greater depths where bottom velocities are marginal in relation to critical shear stress, coarse sediment may in fluence the extent of scour development. This situation may prevail at sites where coarse relict sands are en countered such as the USNUWC Oceanographic Platform. Velocity and scour pit width relationships are shown in Figure 18. Although scattered, points for given ob jects tend to fall into vertical classes with the object shape and size increasing away from the ordinate. It ap pears that there is no relation between increased velocity and changes in pit width. This effect is, no doubt, due to the consistent relationship between object shape and size and the location and magnitude of the deflection of streamlines. Interpretation of Time-Lapse Films Time-lapse photography provided the continuity of observation required for calculations of sediment trans port and volumetric growth rates. Several examples of net volume curves are provided in Figures 19 and 20. These observations, plus the results from all time-lapse studies, are presented in Table II. Although gross relationships are evident, attempts to determine consistent relation ships between specific variables were unsuccessful. The absence of correlation is attributed to the nature of the study which had, as its primary objective, evaluation of scour around a wide variety of objects and their various orientation to flow. To this end, Table II provides a summary of such measurements. Although the scatter of points observed in plotting various combinations of variables was discouraging, several relationships were ap parent. It will be noted that some curves in Figures 19 FIGURE 18 Comparison of pit width (L) to V10 component of surge reveals that scour pit dimensions are independent of the velocity, and that object diameter or cross-sectional area is the controlling factor in scour processes. 119 120 60 50 _ 4 0 XX 3 0 000 a 1 oo oo X 0 20. 40 50 20 PIT WIDTH L1 , ' in cm) o 5.1 cm cylinder • 7.6 cm cylinder a 10.2 cm cylinder A 14 cm cylinder X 17.6 cm cylinder 0 oval & 25 cm cylinder 1 email T-beom I Ia rg e X-beam ^ 20.4 cm cylin der 33 cm cylinder FIGURE 19 Growth curves for scour pits developed around cylindrical objects. Sediment transport rates (Qs) were obtained from single frame projection of time-lapse motion picture film. The low values for the 7.6 cm cylinder are due to relatively weak surge the day of measurement, and the low values of the 14 cm cylinder are caused by the low profile (height) of this object. 121 VOLUME (cc) X 10 10.2 cm 14 cm 7.6 ci 20 4 0 6 0 8 0 100 T I M E (min) 122 FIGURE 20 Growth curves for irregular objects. The small I-beam obstruction is shown in two positions. Surge direction is shown by the arrow at the bottom. Note the right-hand ordinate values differ from those on the left. 123 1* R— Rock X - Small T-beam — Large I-b e a m O-FIQW [ > 4 0 50 TIME (min) 60 70 8 0 24 20 IO O X a <u X) ■ H a> cn 16 -2 i o o 12 8 0 90 111 2 =3 _1 O > * ro -P-- TABLE II Tabulation of water and sediment properties for selected studies. Pinal column on right displays object orientation to surge flow and lists object dimension fac ing flow. Parameters are cited in text. 125 TABLE II WATER. PROPERTIES SEDIMENT PROPERTIES Object (Flow) < ------> Diam. or Width Period T(sec) Wave Height (m) Water Depth (m) Md (mm) Sed. trans. rate Qs (cc/min) Volume growth rate Qv (7./rain) Subm. wt. sand (gms) .6 .40 7.6 .152 18.7/5.5 1.87/0.55 O 7.6 6. .40 7.6 .152 47.0/6.1 2.5/0.46 o 10.1 12. .46 6.1 .148 134/46 2.5/0.9 72.7 o 10.1 12. .46 6.1 .148 1200/86.5 35/3 72.7 o 14.0 11.2 .46 6.1 .150 33.8/15.8 9.6/5 17.9 o 12.7 11.2 .46 6.1 .150 200/2500(7) 12.5/167(2) 17.9 □ 14.0 10.5 .28 7.6 .123 160/29 3.6/0.9 38.2 H 10.1 10.5 .28 7.6 .123 117/42.5 3.6/1.25 38.2 I 10.1 10.5 .28 7.6 .123 123/183 1.8/2.3 38.2 10.1 9.0 .37 9.1 .265 103 4.5 56.3 o 10.1 6.6 .30 6.1 .150 73.4/50 1.6/1.4 22.4 H 10.1 6.9 .30 4.6 .171 380 1.7 15.6 H 23 5.8 .52 10.6 .255 158/126 2.5/2.5 7.27 o o o o 10.1 5.8 .52 10.6 .255 400 14 7.27 0 10.1 5.8 .52 10.6 .255 460 25 7.27 5.1 5.6 .43 7.6 .470 46 2.3 3i 5 5 20.3 6.0 .28 5.5 .185 36/18.7 7.5/3.1 37.8 o o o o 5.1 6.0 .28 5.5 .185 110/20.4 11.1/2.7 37.8 O O 5.1 6.0 .28 5.5 .185 32.3/440 2.7/25 37.8 O 21.6 126 and 20 revealed points of inflection indicating a decrease in sediment transport rates. This generally occurs between percentage growth ranges of 55 to 75 percent. These in flections correspond to similar inflections in the pit lip growth curves which increase in a generally progressive manner with increasing object diameter. A reduction in sediment transport rates is to be expected under a given level of surge velocity due to the fact that the absolute volume to be removed for an incremental advance of the pit lip increases with increasing pit width. However, this should impose a smooth linear decrease in sediment trans port rates rather than the observed break in slope. It is suspected that this break is related to some critical pit width (under the given conditions) beyond which the bed load becomes significant, and perhaps dominant, in scour processes. Prior to reaching this point, the pit may be sufficiently small to permit all of the suspended sediment to leave the pit. As the pit grows in extent, an increas ing amount of suspended sediment falls back to the bed inside the depression. Experiments with fluid jets directed against a sediment surface revealed that the transition period between suspended (initial) and bed load (later) transport rates was very narrow (Laursen, I960). On the sea floor, these relationships are not as clear, primarily due to wide fluctuations in the velocity gradient. However, this explanation appears reasonable for most ob 128 served changes in sediment transport rates. The anomalous values of higher Qs beyond the initial inflection points may be related to object shape (blunt or flat) but again the causes are not clear. It is not, however, due to the merging of the four corner pits for blocks and I-beams. This event invariably occurs much earlier than the ob served break in the volume rate curves. Figure 21 contrasts the two measures of sediment transport rates used in this study. Although these varia bles are related, by comparing them, the effect of shape and size emerges as a dominant factor in scouring rates. An increase in object size causes an increase in scour development, as shown by higher Qs values for the larger objects. Furthermore, the rate of pit growth (Qv in percent/min) decreases with increasing object size. In combination, these effects are apparent in the field where small objects attain terminal scour conditions rapidly whereas large objects may fail to generate significant depressions during the same period. The general relation ship between the effective Ns and both net weight of sand moved (Fig. 9) and strong surge frequency (Fig. 10) sug gests that similar agreement should be found by comparing effecting Ns with Qs. Inasmuch as many objects were em ployed in this study, it is difficult to establish any thing but gross trends such as the two shown in Figure 22 for the 10 cm cylinder and small I-beam. The scatter in FIGURE 21 Comparison of the two sediment transport rates used in this study reveals the reason for rapid scour around small objects. Note that the region of high Qv values is occupied by smaller objects, while the higher Qs values are generally restricted to larger shapes. Two points at the top of the figure fall off the limits of the plot and should be projected to higher ordinate values. 129 (cc /m in) 200 100 - * ? T 380 430 a a a , ^ 0 0 ^ D “ D o o V 03 * A , Cj mm - \ / a 0 ^ 0 * 0 a ° X 3 a * V ° *> V o o — a “ • ° o a • a o ° / < 0 o 0 .1 0.5 I 10 o 5.1 cm cylinder • 7.6 cm cylinder 0 10.2 cm cylinder ft small 1 - beam 0 lens A 14 cm cylinder X rocks ■ 14 cm blocks 130 FIGURE 22 The relationship between effective Ns and maximum Qs shows several trends. Although the scatter is pro nounced, the trends shown by the two objects most frequently studied (10 cm cylinder and small I-beam) suggest a relationship exists such that increasing ef fective Ns yields an increased Qs. See text for dis cussion. 131 IOQVI^s '^sc I 500 1000 OC / / 1/ / / / / / TL/ I / / / oC 132 X o to _S£_ in. J_ i » i l_L 15 50 100 MAX. Qs (cc/min) OBJECTS (dimensions in cm.) 500 5.1 cylinder ^-5.1 cyl.-ripple crest oC- 5.1 cyl. - cluster • - 7.6 cylinder n - 10.2 cylinder PC-10.2 cylinder-cluster 1 - 1 4 blocks X - rocks X - 10.7 small I-b e a m - 2 2 large I-b e a m 133 the 5 cm cylinders must be due to some combination of factors not readily apparent. Although Figure 22 permits little more than speculation, it does suggest that, as might be expected, an increasing effective Ns is accompani ed by a rise in the Qs. Attempts to compare the weight of sand moved per hour, as measured by Cook (in progress), with the Qs were abandoned. Although it is possible to equate the two by density and porosity approximations, Cook's figures are primarily for bed load caught in a small sediment trap set in the sea floor. The Qs value is a function of the area of a pit, and because this is constantly increasing, it is difficult to compare it with the static area of the sedi ment trap. In addition, an unknown percent of the Qs is due to suspended transport, and even though it may be relatively small, it precludes comparison between the two factors. 134 GEOLOGICAL SIGNIFICANCE Nearshore Environment Any process which places sea floor sediment in motion becomes an effective component of the sum of forces working to erode, transport and deposit submarine sedi ments. Inasmuch as the process of scour has been ob served to greatly accelerate rates of sediment transport, the presence of obstructions to the ambient flow will greatly accelerate sea floor erosion. Thus, if the ob structions are reasonably close together, the scour pro cess becomes effective in: 1. Accelerating the rate of erosion in the vicinity of obstructions, 2. Creating better sorting by inhibiting deposi tion of hydraulically finer sediments, and 3. Accelerating the rate of deposition in neighbor ing areas which are exposed to lower ambient surge velocities. These effects may have seasonal significance in those nearshore areas consisting of coarse (cobble and boulder) substrate deposits or outcrops. During fall and Winter, local beaches are eroded by higher waves as sociated with storms. This material is deposited offshore beyond the breaker zone. During the summer, the sand is 135 returned to the beach, primarily by longer period swells which are more effective in transporting sediment than the shorter but higher winter waves. In the offshore region, the effect of obstructions producing scour is to retard deposition in the fall and accelerate erosion in the Spring. An example of accelerated erosion due to scour is shown on Plate 15. A rock buried with only its uppermost corner protruding through the surface of a fine sand (Md 0.136 mm) simulates conditions which prevail as a rocky substrate is exhumed. Within a period of only 30 minutes, one-half of the rock was exposed by scour. If similar obstacles lay in close proximity, the entire area of the sea floor would have been lowered by coalescing depressions. Time-lapse photographs of these rock obstacles provided sediment transport rates shown graphically in Figure 20. During a study in January, 1969, deep asymmetrical scour pits were observed around boulder outcrops off Point Vicente (PI. 14). This asymmetry does not resemble the generally concentric scour pits commonly observed around either the test objects or tall pilings. In plan view, its shape is more closely approximated by pits formed by unidirectional flow. It may be that the primary onshore flow passes over the low profile and generates a strong vortex on the shoreward side of the obstruction, while the seaward flow generates a vertical jet on the more exposed shoreward face of the rock (Fig. 23). Acceleration of PLATE 13 Time study of progressive scour around test cobble, Santa Catalina Island, December 11, 1968. Pertinent data are: Md sediment = 0.136 mm Cobble dlam. = 20 cm Strong surge freq.. = 5.2 surges/min V]_q = 4-8.1 cm/sec Ns for 2V10 =21.0 Effective Ng = 425.8 See text for discussion. 136 I minute 10 minutes 30 minutes PLATE 14 a. Boulder outcrop on sea floor off Point Vicente, Palos Verdes, California, at a depth of 7 m. Deep scour has developed between two exposed boulders by jet action of surge passing be tween the two obstructions. Ripples at the front of the larger boulder indicate that flow parallel to the lower face of the boulder passes into the jet region. Surge passes from right to left. b. View of boulders in (a), looking seaward. Surge moves parallel to plane of view. Note rim of scour pit at bottom. Yardstick pro vides scale. Photographs taken January 9> 1969. 138 PLATE 14 c. Boulder outcrops at a depth of 7 m off Point Vicente, Palos Verdes, California. Jet effect between boulders has eroded deep channels and maintains scour depressions in foreground. Yardstick provides scale. Photographs taken January 9» 1969. 140 FIGURE 23 Proposed scheme of scour In the vicinity of a boulder outcrop shown in Plate 13a and 13b. a. Scour with surge directed toward shore establishes eddies behind the boulder and promotes a Jet between obstructions. b. Flow in offshore direction develops a signi ficant vortex in front of the more exposed shoreward face of the boulder. 142 143 SHORE SURGE SURGE B 144 currents passing among adjacent rocks also serves to maintain high velocities on the pit floor. Such a gap may establish a pressure "leak*' which channels accelerating flow away from the low face of the obstruction. This relationship requires further investigation. It appears that the low profile and large size of the object are the significant factors in promoting ef fective scour on the shoreward side of low obstructions. This effect was observed at a number of nearshore stations during this study. With increasing height, an effective primary vortex may be established on the seaward side and the pit would tend to assume the more circular and concen tric configuration. If the obstruction is a major feature of relief, such as an islet, submerged ridge or breakwater, its margins may show scoured depressions regardless of its orientation. This occurs when the obstacle either reaches the surface or, if submerged, is large enough to create wave refraction and thus alter the direction of bottom surge. A scoured channel of this nature is shown on Plate 15, where high wave energy and refraction over a bed rock ridge maintain a scoured channel flanking the ridge. PLATE 15 Scoured channel adjacent to bedrock outcrop near Plymouth, Massachusetts. Heavy swell from “northeasters" is refracted around outcrop and effectively scours marginal depression along contact between medium sand sea floor sediment and resistant outcrop. Note mound of sand in lower right of bathymetric chart indicative of deposition outside of strong surge area. Datum of chart is MLW, echo sounder record uncorrected for tide. Record and chart prepared February 14, 1967» during local storm season. 145 A" 146 BSDHCCK SCOUR SAND ‘ aM fcdii u * ■ t i i i i i i j j -120 1 & & -1 8 0 - \/a \3 £ / A \ ? o : i : /aiya i 4 t y ' ■ / a ! ■ ' -125- -1S5- BLUinvORtjfU MARINE, fl.Y. E5-20 : ■ ! DEPTH IN FEET 35 R O C K Y VER Y IRREGULAR 3o 25 co n to u rs in f e e t 100 100 200 300 meters 14? Submarine Erosion Related to Scour The process of vigorous scour in the nearshore region is effective in promoting submarine erosion. Large outcrops and boulders are frequently undercut at their sides where the scour velocity is greatest. Excellent ex amples of this condition are shown on Plates 16 and 17. The abrasive action of local sediment as it sweeps past the outcrop does not permit a protective sheath of sessile organisms to become established in the region of maximum abrasion. Paleoenvironmental Analyses Analyses of scour geometry should provide a useful tool in paleoenvironmental studies. Pettijohn and Potter (1964) display several examples of "current crescents" (unidirectional scour pits) and other scour features pre served in rocks as old as Precambrian (Pahrig, 1961). A well-illustrated classification of scour features developed under unidirectional flow, complete with outcrops and hand sample examples of preserved scour marks, is provided by Dzulinski and Saunders (1962). Although the published references are sparse, the fact that ripple marks and similar small sedimentary structures are frequently found in outcrops suggests that scour depressions also may be preserved. PLATE 16 a and b. Megaripples and undercut schist boulders, Santa Catalina Island, at water depth of 6 m. Sediment in both photos has median diameter ranging from coarse sand to cobbles. Photographs courtesy of T. H. Loop. 148 PLATE 17 a. Sea floor erosion by scour at Santa Catalina Island, depth 6 m. Note incised rock plat form at base of outcrop. Rock behind left range pole protrudes beyond softer material (schist) due to greater resistance to abrasion. Gradations on range pole = 10 cm. b. Megaripples and undercut schist, Santa Catalina Island, depth 6 m. Absence of at tached growth in undercut reveals height of effective abrasive scour, here about 25 cm. Both photographs by T. H. Loop. 150 152 If the scour pit configuration can be established, it should be possible to determine the nature of flow (oscillatory or unidirectional) producing the scour. If a trace of the primary vortex is preserved, and the pit lip is greater than twice its distance from the object, scour was probably by wave-induced surge. Asymmetric scour pits developed around large objects on the sea floor may be confused with unidirectional flow conditions, but if smaller objects in the same outcrop exposure had circular pits, wave-induced scour may be proposed. In this case, the major axis of the large pit should serve as a direc tional distinguishing criterion, unidirectional scour pits display a raised mound in the lee of the obstruction, a feature which has not been observed on the sea floor. Scour at Depth According to wave theory, scour by oscillatory flow at depths greater than the local maximum wave length should not occur. An Investigation of portions of the mainland shelf at depths greater than the limits of practical SCUBA diving was conducted with the aid of General Oceanographies' submersible "Nekton." The site selected for this dive was an area of rock outcrops lying at a depth of 63 m on the shelf within Santa Monica Bay (Pig. 1). The dives were made in clear weather on December 23, 1968, A days after a period of intense storms and high seas. In the 3 hours of sea floor observation at the site, the writer failed to observe any depressions attributable to wave-induced or current scour around any of hundreds of exposed rocks. The sea floor was composed of fine silty sands which should have displayed scour ef fects if wave-induced surge had occurred. However, numerous depressions resembling scour pits were observed around many rocks, but in every case, these definitely could be attributed to biological excavations. Numerous crabs and small fish were observed within these depressions and it is apparent that the rocks are exploited as havens by such animals. This is especially significant because, in many cases, the most extensive pits were immediately ad jacent to natural havens provided by overhanging ridges in the rock. A vertical photograph at this site, such as that often obtained by deep-sea cameras, would thus re veal an excavated pit surrounding a rock which would appear to be a scoured depression when, in reality, it was the result of the work of organisms. This suggests that care should be exercised in analyzing deep-sea photos for scour effects. Wave-induced scour effects have been noted on the summits of seamounts which lie in relatively shallow waters. Walter C. Sands (personal communication, 1968) stated that television and photographs reveal effective scour on the summit of Oobb Bank at a depth of 40 m during 154 the winter storm season. Similar scour effects have been observed on summit platforms of seamounts in the Atlantic (Pratt, 1963, 1968). Examination of hundreds of photo graphs obtained by the research submerslbles Trieste, Soucoupe, and Deepstar in deep dives off southern Cali fornia and Mexico show only a few examples of scour around natural objects and debris (PI. 18). However, deep dives with the Soucoupe in San Lucas Canyon (off the southern tip of Baja California, Mexico) revealed definite asym metric scour pits in the axis of the canyon at a depth of 290 m (Shepard and Dill, 1966). Similarly, many deep sea photographs (3jn Hersey, 1967) reveal asymmetric scour pits around rocks. Such pits are useful in interpreting deep current directions and relative velocities. LaPond (1961) examined current velocities at the sea floor developed by internal waves. He suggested that, given sufficient constriction by a thermocline lying 6 m above the sea floor, water velocities of 3 cm/sec can be developed at the bed. In comparing this velocity to curves for sediment motion (HJulstrom, 1939)» he erroneous ly concluded that internal waves generating a velocity of 3 cm/sec will move sand 0.4 mm in diameter. This value was derived from the curve separating transportation from deposition, not the threshold velocity curve. The magni tude of velocity components generated by internal waves merits scrutiny, as it may be that velocity components PLATE 18 a. Partially buried beer can in 290 m water, San Diego Trough. Note lack of scour and sedi ment atop can. Numerous pit and tracks sug gest partial burial by bioturbation. Sea floor in this area had high concentration of spiny urchins (upper left). (Official U. S. Navy photograph by E. 0. Buffington.) b. Bathroom sink on floor of San Diego Trough, depth 290 m. Two fish suggest this object acts as a haven to swimming organisms. See text for description of sedimentation and apparent scour. (Official U. S. Navy photo graph by R. P. Dill.) 155 157 generated by Internal wave motions are ineffectual in either eroding the sea floor or developing scour depres sions around obstructions to such flow. The observation that partially buried concrete blocks at the test stations off Santa Catalina Island were undermined and re-buried in a relatively brief period (30 days) leads to the conclusion that estimates of accretion based on the partial burial of objects may be misleading. If periodic turbidity currents, density currents or other agents producing relatively high velocity on the sea floor develop scour depressions around unsupported objects, then biologic reworking of the bottom may soon smooth and fill the pit, partially burying the obstruction. Estimates of rates of deposition based upon artificial datable ob jects, such as a beer can (Dill, Personal Communication, 1969) should be qualified by evaluating the scour potential at the site. In the case of Dill's observations, un questionable scour effects were noted farther down the axis of the canyon, but at the site of the container no ripples or scour were noted. A similar situation was observed by Buffington (Personal Communication, 1969) in a dive aboard the Soucoupe off San Diego. A beer can (PI. 18) at a depth of 290 m was observed to be buried in a fashion Identical to that of Dill's. Bioturbation of the fine bottom materials is obvious in this photograph, and traces of sediment may be seen atop the container. This appears 158 to be a case of partial burial by biologic activity rather than .accretion. Interpretation of scour around larger, irregular objects may be more difficult. Plate 18b depicts a bath room sink in the same area as Plate 18a. This object was photographed by R. P. Dill during a dive in Deepstar. Sig nificant points to observe in this case are: 1. The presence of two fish immediately adjacent to the object. One appears close to an over hang, a situation identical to that observed in similar "rock havens" on the Santa Monica shelf. 2. Attached organisms on the upper surface of the sink which suggests both relative stability (no overturning) and a relatively long period of elapsed time since "deposition" of the sink. 3. The left-to-right onlapping wedge of mixed sediment atop the Inclined flat portion of the sink. 4. A deposit of fine materials within the bowl proper. 5. The essentially round configuration of the sur rounding depression, suggesting scour occurred by oscillatory flow. Regardless of the round pit, in combination, these factors suggest current-induced scour with a prevailing current acting from right to left. The fine materials 159 within the bowl imply that currents of a magnitude suf ficient to lift fine sediments to this level were active at the site. It is doubtful that bioturbation alone could achieve this effect, but the pit obviously has been modified by biologic activity. It may be that part of the pit in Plate 18b repre sents an impact feature. After falling to the sea floor, the combined action of acceleration of weak currents and biologic activity could preclude deposition within the pit. These effects, plus unidirectional current flow, may have maintained the depression. The sink is an unsupported object similar to the blocks at the fixed stations at Santa Catalina Island. If scour was an effective process at this depth, one would expect that the sink would be partially buried rather than lying essentially at the sur face . Scour depressions up to 1.5 m in depth have been observed around outcrops on the floor of Santa Barbara Channel (J. W. Vernon, personal communication, 1968). Numerous pits and high current velocities (up to 0.5 m/sec) at a depth of 220 m supports the contention of active but localized scour at considerable depths. A similar, but concentric scour pit, at a depth of 900 m, was described by Dietz and Dill (1969). If surge generated by internal waves is discounted, it is difficult to conceive of an agent producing oscil latory flow at depths below a value equivalent to the maximum wave length at a given site. The presence of con centric scour pits at greater depths must be due to some other cause. The writer concludes that scour pits at depth are initiated by unidirectional current flow but are preserved by a combination of biological grazing and ex cavation, and the deflection of weak but locally competent currents which inhibit deposition in the region of ac celerated flow. If scour pits are preserved in sedimentary rocks, the use of pit configuration to establish flow regime must be used in conjunction with other evidence (grain size, microfossils) to estimate water depth. If shallow water conditions are indicated, the concentric pit will indicate oscillatory flow. 161 CONCLUSIONS With regard to the limitations imposed on this study by the physical conditions of the environment, the accuracy of the devices and techniques used to record data, and the scatter imposed by the use of a variety of objects, the following conclusions are proposed: 1. The height and shape of the obstacle has a sig nificant effect on sediment transport rates early in the history of scour. Eventually the object diameter becomes the dominant factor. Initially, objects with a blunt or flat surface exposed to flow will scour more rapidly than round objects. 2. The upper 10 percent of surge velocities measured at the sea floor ranged from 24 to 61 cm/sec for a 95 percent confidence limit. It appears that, within^this range, scour extent (pit diameter) is independent of velocity. 3. For the observed range of median grain sizes from 0.120 to 0.630 mm, scour extent is independent of grain size. However, at depths greater than those occupied during this study, grain size may influence scour rate. 4. For the test objects considered, the locus of high vorticlty associated with the primary vortex migrates outward at a rate equal to one- half that of the pit lip. Upon reaching a critical distance, which increases with in creasing object diameter, the primary vortex for given flow conditions remains static while the pit continues to expand. This point has no visible effect on volume growth rate curves. Individual elements in clusters are initially scoured as individual obstructions. After pits coalesce, scour around the cluster approaches that for a single object having a diameter equal to the combined diameter and separation distance of the elements. Scour extent from the outermost elements of the cluster approxi mates that for a single element. Inclining test cylinders at a 45° angle to the sea floor, both parallel and normal to the surge orthogonals, produce scour depressions of a size equal to that developed around identical vertical cylinders. This suggests that effec tive scour occurs without regard to the atti tude of the obstruction. Comparison of the effective sediment number (effective Ns) with the weight of sand moved as bed load demonstrates that the effective Ns provides a useful dimensionless parameter for sediment transport studies. Its relationship to sediment transport rates (Qs) for scouring conditions is not well-defined, but an increase in effective Ns promotes increasing values of Qs. 8. As the cross-sectional area of the obstruction increases, the absolute rate of sediment trans port (Qs) increases, but the percentage growth rate (Qv) decreases. This relationship re flects the increasing net volume of sediment which must be scoured in order to achieve a given increase in scour extent for larger ob jects. Therefore, for a given effective Ns, the overall increase in the dimension of scour pits generally progresses more slowly for larger objects than for smaller objects. 9. The maximum extent of scour around rocks and tall test objects is uniform for both oscil latory and unidirectional flow up to a critical diameter of about 14 cm. For larger diameters, the differences in scour extent become pro gressively greater, with oscillatory flow pro ducing a larger pit than unidirectional flow for objects of similar dimensions. 10. Long-term studies of vertical cylinders with identical diameters, but placed at different 164 depths, indicate that the effectiveness of scour decreases with increasing depth. 11. Large, unsupported objects may develop scour pits of sufficient depth to lower them below the level of the sea floor. Infilling by natural processes during periods of low surge velocity ultimately buries these objects with no net accretion of the sea floor. 12. Strong surge frequency is a critical factor in scour processes. An increase in the strong surge frequency raises the effective Ng which, in turn, increases sediment transport rates. 13. Scour processes accelerate erosion and retard deposition on the sea floor. 14. Large, low objects generally display asymmetric scour pits. Major topographic features such as ridges and islets may redirect flow with the result that scour forms marginal depressions. 15. Scour geometry and distribution are useful in paleoenvironmental studies as indicators of current and flow regime and shoreline orienta tion. 16. Scour at depths below that equal to the maximum wave length at a given site is generated by unidirectional flow. Modification of the initially asymmetric pit by biologic agents and 165 flow acceleration may produce a concentric scour pit, resembling that developed under oscillatory flow. 17. Depressions similar in appearance to scour pits may be developed by biologic excavation. 18, Currents developed on the sea floor in re sponse to the passage of interval waves are in effectual in eroding or scouring the sea floor. REFERENCES To assist readers in locating references relevant to their interests, a letter code has been added at the j end of appropriate citations. A letter or letters follow- j ing such references indicates the general topic or ap- j proach used by other investigators. As used here: L = Laboratory study (including theoretical treatments) j F = Field study i t 0 = Oscillatory flow j U = Unidirectional flow | Thus, a citation of (F,U) implies a field study j under unidirectional flow conditions. | REFERENCES ■Allen, J. R. L., 1965a, A review of the origin and characteristics of Recent alluvial sediments: Sedi- mentology, v. 5, p. 91-191. (F,U) _______ , 1965b, Scour marks in snow: Jour. Sed. Petrology, v. 35, p. 331-338. (F,U) :_______, 1968, Current ripples; their relation to patterns and sediment motion: North Holland, Amsterdam, 433 p. | (F,U) jBagnold, R. A., 1946, Motion of waves in shallow water. Interaction between waves and sand bottoms: Roy. Soc. London, Proc., Ser. A., v. 187, p. 1-15. (L,0) ;_______, 1963, Beach and nearshore processes, Part 1, Mechanics of marine sedimentation; in The Sea: Ideas and Observations, v. 3, Interscience Publishers, New York, p. 507-528. (L,0) Bradford, W. T., 1966, Surge induced scour: unpub. Master's thesis in Geology, University of So. Calif., 71 P. (F,0) Briggs, L. I. and Middleton, G. V., 1965, Hydromechanical principles of sediment structure formation; in Primary Sedimentary Structures and their Hydrodynamic Interpretation; Soc. Econ. Paleontologists and Mineralogists Special Pub. 12, 265 p., p. 5-16. (L,U) Brooks, N. H., 1958, Mechanics of streams with movable beds of fine sand: Am. Soc. Civil Engineers, Trans., v. 123, 526-659. (L,U) Carlisle, J. G., Jr., Turner, C. H., and Ebert, E. E., 1964, Artificial habitat in the marine environment: Fish Bulletin 124, Calif. Dept. Fish and Game, 93 p. Carstens, M. R., 1965, Part 1 - The effect of dunes upon localized scour: Final Rpt., ProJ. A-770, Engineering Experiment Station, Georgia Inst. Technol., 19 p., App. A, B, C (mimeo). (ii,U) _______ , 1966, Similarity laws for localized scour: Jour, of Hydraulics Div., Amer. Soc. Civil Engineers, v. 92, p. 13-36. (L,U) 167 168 Carstens, M. R., and Martin, C. S., 1963, Scour: in Tech. Rpt. 3, Pour Topics Pertinent to Sediment Transport and Scour; Pro;]. A-628, Engineering Experiment Station, Georgia Inst. Technol., 18 p. (mimeo.). (L,U) Carstens, M. R. and Neilson, P. M., 1967a, Evolution of a duned bed under oscillatory flow: Jour. Geoph. Res., v. 72, p. 3053-3059. (L,0^ Carstens, M. R., Neilson, P. M., and Altinbilek, H. D., 1967b, An analytical and experimental study of bed ripples under water waves: Pinal Rpt., Pro;]. A-798, Engineering Experiment Station, Georgia Inst. Technol., 63 p., App. A, B (mimeo.). (L,0) Cook, D. 0., in press, Calibration of the University of Southern California automatically recording settling tube: Jour. Sed. Petrology, v. 39. Dietz, R. S., and Dill, R. P., 1969, Down into the sea in ships: Sea Frontiers, v. 15, p. 2-9. Draper, L., 1957* Attenuation of sea waves with depth: Houille Blanche, v. 12, p. 926-931. (F,0) Dzulynski, S. and Sanders, J. E., 1962, Current marks on firm mud bottoms: Connecticut Acad. Sci., Trans., v. 42, p. 57-96. (F,0,U) Eagleson, P. S., Dean, R. G. and Peralta, L. A., 1958* The mechanics of the motion of discrete spherical bottom sediment particles due to shoaling waves: Mass. Inst. Technol., Hydrodynamics Lab., Tech. Rpt. no. 26. (L,0) Eckart, C., 1952, The propagation waves from deep to shallow water: in Gravity Waves, Nat. Bur. Standards, Circ. 521, p. 165^173. Einstein, H. A., 1948, Movement of beach sands by water waves: Am. Geoph. Union, Trans., v. 29, p. 653-655.(1,0) Fahrlg, W. P., 1961, The geology of the Athabaska forma tion: Geol. Survey of Canada, Bull. 68, 41 p. Felix, D., in press, Design of an automatic settling tube for analysis of sands: Jour. Sed. Petrology, v. 39. Gugnyaev, Ya. E., 1959* The modelling of sea drift: (in Russian), Trudy Okeanogr. Korn. Akad. Nauk. SSSR, no. 4. 169 Heezen, B. C., Tharp, M. and Ewing, M., 1959, The floors of the Oceans: I: North Atlantic, Geol. Soc. America Spec. Paper 65, 122 p. Hersey, J. B., 1967, Deep Sea Photography: Johns Hopkins Oceanographic Studies no. 3, 1967, 310 p. # • HJulstrom, P., 1939, Transportation of detritus by moving water: in Recent Marine Sediments, P. D. Trask, editor, 3m. Assoc. Petroleum Geologist, Tulsa, p. 5-31. (L,U) Hollister, C. D. and Heezen, B. C., 1967, The floor of the Bellinghausen Sea: in Deep Sea Photography, J. B. Hersey, editor, Johns Hopkins Oceanographic Studies no. 3, P. 177-189. Hooke, R. L., 1968, Laboratory study of the influence of granules on flow over a sand bed: Geol. Soc. America Bull., v. 79, P. 495-500. (L,U) Hunt, L. M., 1961, Wave generated oscillatory currents along the bottom of the eulittoral and sublittoral zones: Mine Advisory Committee, Nat. Acad. Sciences, Nat. Res. Council, 23 p. (mimeo.). (L,0) Inman, D. L., 1949, Sorting of sediments in the light of fluid mechanics: Jour. Sed. Petrology, v. 19, p. 51-70. (L,0) _______ , 1963, Chapter 3, Ocean waves and associated currents, and Chapter 5, Sediments: Physical pro perties and mechanics of sedimentation: in Shepard, P. P., Submarine Geology, Harper & Row, New York, p. 49-81, 101-147. (L,0) Kalkanls, G., 1964, Transportation of bed material due to wave action: U. S. Army Coastal Engineering Res. Center, Tech. Memo. 2, 38 p. (L,0) Karcz, I., 1968, Pluviatile obstacle marks from the wadis of the Negev (southern Israel): Jour. Sed. Petrology, v. 38, p. 1000-1012. (p,U) Kuenen, Ph. H., 1967, Emplacement of flysch-type sand beds: Sedimentology, v. 9, p. 203-244. LaPond, E. C., 1961, Internal wave motion and its geological significance: in Mahadevan Volume, 6 May, 1961, p. 61-77. 170 Lamb, H., 1945, Hydrodynamics: 6th Edition, Dover, New York, 738 p. Larras, J., 1957* Plages et Cotes de Sable; Paris. Laughton, A. S., 1963* Microtopography: in The Sea: Ideas and Observations, v. 3* Interscience Publishers, New York, p. 437-472. _______ , 1967* Underwater photography of the Carlsberg Ridge: in Deep Sea Photography, J. B. Hersey, editor, Johns Hopkins Geographic Studies no. 3* p. 191-206. Laursen, E. M., i960, Scour at bridge crossings: Amer. Soc. Civil Engineers, Jour. Hydraulics Div., v. 86, p. 39-54. (F,U) Leliavsky, S., 1966, An introduction to fluvial hy draulics: Dover Publishing, Inc., New York, 257 p. Loop, T. H., 1969, Physical oceanography and sediment characteristics of Little and Shark Harbors, Santa Catalina Island, California: unpub. Master's thesis, in Geology, University of So. Calif., 144 p. Manohar, M., 1955, Mechanics of bottom sediment movement due to wave action: U. S. Army Corps of Engineers, Beach Erosion Board Tech. Memo. 75, 121 p. (L,0) Martin, C. S., unk., The effect of uplift on incipient motion of bed particles; mimeo. rpt., 11 p. (L,U) Masch, P. D., Jr., and Moore, W. L., I960, Drag forces in velocity gradient flow: Jour. Hydraulics Div., Amer. Soc. Civil Engineers, v. 86, p. 1-11. (L,U) Menard, H. W., Jr., 1950, Sediment movement in relation to current velocity: Jour. Sed. Petrology, v. 20, p. 148-160. (L,U) Moore, W, L. and Masch, P. D., Jr., 1962, Experiments on the scour resistance of cohesive sediments: Jour. Geoph. Res., v. 67, p. 1437-1446. (L,U) Owen, D. M., Emery, K. 0., and Hoadley, L. D., 1967, Effects of tidal currents on the sea floor shown by underwater time-lapse photography: in Deep Sea Photography, J. B. Hersey, editor, Johns Hopkins Oceanographic Studies no. 3, p. 159-166. 171 Peabody, P. E., 1947, Current crescents in the Triassic Moenkopi formation: Jour. Sed. Petrology, v. 17 > P. 73-76. Pettijohn, P. J. and Potter, P. E., 1964, Atlas and glossary of primary sedimentary structures: Springer- Verlag, New York, 370 p. Powell, R. W., 1955» The use and misuse of hydraulic models: The Port Engineer, July, 1955> p. 21-26. Pratt, R. M., 1963> Bottom currents on the Blake Plateau: Deep-Sea Res., v. 10, p. 245-249. _______ , 1968, Atlantic continental shelf and slope of the United States; Physiography and sediments of the deep-sea basin: U. S. Geol. Survey Prof. Paper 529-B, 44 p. Richardson, P. D., 1968, The generation of scour marks near obstacles: Jour. Sed. Petrology, v. 38, p. 965-970. (L,U) Sanders, J. E., 1965» Primary sedimentary structures formed by turbidity currents and related resedimenta tion mechanisms: in Primary Sedimentary Structures and their Hydrodynamic Interpretation: Soc. Econ. Paleontologists and Mineralogists, Special Pub. 12, 265 p., p. 192-219. Shen, H. W., Schneider, V. R. and Karaki, S.. S., 1966, Mechanics of local scour: Engineering Res. Center, Colo. State Univ., Rpt. no. CER66HWS22, 55 p. (L,U) Shepard, P. P., 1963, Submarine Geology: Harper & Row, New York, 557 p. Shepard, P. P. and Dill, R. P., 1966, Submarine Canyons and Other Sea Valleys: Rand-McNally, Chicago, 397 p. Shields, A., 1936, Applications of similarity principles and turbulence research to bed load movement (in German): Mitteilungen der Preussischen Versuchanstalt fur Wasserbau und Schiffbau, Berlin, Heft 26, 26 p. Translation by W. P. Ott and J. C. van Uchelen, U. S. Dept.- Agriculture, Soil Conservation Service Coop. Lab., Calif. Inst. Tech. (L,U) O 1 1 Sundborg, A., 1956, The River Klaraven, a study of fluvial processes: Geograf. Ann., v. 38, 127-316. (P,U0 172 Sutherland, A., 1967, Proposed mechanism for sediment en trapment by turbulent flows: Jour. Geoph. Res., v. 72, p. 6183-6194. Tison, L. J., 1961, Local scour in rivers: Jour. Geoph. Res., v. 66, p. 4227-4232. (L,U) Vanoni, V. A., 1964, Measurements of critical shear stress for entraining fine sediments in a boundary layer: Calif. Inst. Tech., W. M. Keck Lab. of Hydraulics and Water Resources, Rept. KH-R-7* 47 p. (L,U) Volkov, P. A., i960, The velocity below which waves do not erode a loose bottom. Rechn. Transp., no. 6. Wiegel, R. L., 1964, Oceanographical Engineering: Prentice-Hall, Englewood Cliffs, New Jersey, 532 p. Zenkovich, V. P., 1967* Processes of Coastal Development: Interscience Publishers, New York, 738 p.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Oceanography And Late-Quaternary Planktonic Foraminifera, Southwestern Indian Ocean

PDF

Sedimentology And Pleistocene History Of Lake Tahoe, California - Nevada

PDF

Distribution And Transport Of Suspended Matter, Santa Barbara Channel, California

PDF

The Aleutian-Kamchatka Trench Convergence: An Investigation Of Lithospheric Plate Interaction In The Light Of Modern Geotectonic Theory

PDF

Synoptic Surf Zone Sedimentation Patterns

PDF

Sand Transport By Shoaling Waves

PDF

Late-Neogene Paleomagnetic And Planktonic Zonation, Southeast Indian Ocean - Tasman Basin

PDF

Continental Slopes Of The World

PDF

Recent And Upper Pleistocene Sediments Of The Southwestern Portion Of Losangeles County, California

PDF

Petrology And Depositional History Of Late-Precambrian - Cambrian Quartzites In The Eastern Mojave Desert, Southeastern California

PDF

Early Diagenesis In Southern California Continental Borderland Sediments

PDF

Marine Geology Of The Andaman Basin, Northeastern Indian Ocean

PDF

The Marine Geology Of San Miguel Gap Off Point Conception, California

PDF

Shelf Sediment Transport System

PDF

Marine Geology Of The Santa Catalina Basin Area, California

PDF

Marine Geology Of The Baja California Continental Borderland, Mexico

PDF

Preservation Of Fossil Fish In The Miocene Monterey Formation Of Southern California

PDF

Quartz Grain-Shape Variation Within An Individual Pluton: Granite Mountain, San Diego County, California

PDF

The characterization of Huntington Beach and Newport Beach through Fourier grain-shape, grain-size, and longshore current analyses

PDF

Cyclostratigraphy and chronology of the Albian stage (Piobbico core, Italy)

Asset Metadata

Creator Palmer, Harold Dean (author)

Core Title Wave-Induced Scour Around Natural And Artificial Objects

Degree Doctor of Philosophy

Degree Program Geological Sciences

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Geology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Gorsline, Donn S. (committee chair), Stone, Richard O. (committee member), Tibby, Richard B. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-394792

Unique identifier UC11361850

Identifier 7011382.pdf (filename),usctheses-c18-394792 (legacy record id)

Legacy Identifier 7011382.pdf

Dmrecord 394792

Document Type Dissertation

Rights Palmer, Harold Dean

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA