Revised Clear-Water and Live-Bed Contraction Scour Analysis (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3-1   Laboratory Testing Setup and Test Plan 3.1 Overview As summarized in Chapter 2, the laboratory and numerical applications focused on priority aspects of contraction scour presently not adequately understood or formulated. The laboratory tests were designed to eliminate the deficiencies of earlier investigations that were discovered during NCHRP Project 24-34 (see Section 2.2.2) and during the literature and data evaluation phase of this study. NCHRP Project 24-34, for example, found that several earlier contraction scour studies suffered from flaws in the experimental design that degrade the utility of the datasets resulting from those studies. The literature and data review indicated that much of the prior data mix short-contraction and long-contraction depths of scour, and in some cases, it is not clear which scour depth was documented. The numerical experiments were designed to augment the laboratory modeling, focusing particularly on how contraction geometry affects the flow field through the contraction con- nection between the approach channel and the narrowed channel downstream. The laboratory experiments were conducted using the 8-ft-wide, 200-ft-long sediment recirculating flume at Colorado State University (CSU). Various geometric configurations of the contracted section were examined for a fine sand sediment size under a range of flow rates for each configuration. Non-erodible (rigid) bed tests were also conducted. 3.2 Test Setup This section describes the test setup, test program, and the array of instruments used and measurements taken during the laboratory testing program. Figure 3-1 is a schematic diagram of the relevant geometric information and hydraulic conditions available with which to investi- gate contraction scour. The hydraulics of flow through the test setup was checked by means of the 1D numerical model HEC-RAS. The test setup used the 8-ft-wide flume at CSU as shown in Figure 3-2. The setup included a 25-ft-long approach channel that contracts to an approximately 85-ft-long narrowed channel. The contraction angle (indicated as a tapered wall set at angle α) in Figure 3-1 was held at 45o for all tests. The setup in Figure 3-1 indicates a short-contraction reach and a long-contraction reach, followed by a reach with the channel returned to full width, similar to the layout shown schematically in Figure 2-1(a). This setup conveniently encompasses both short- and long- contraction scour conditions and the relationship between the two conditions. Moreover, this setup enabled the transition between the two conditions to be altered as necessary. An initial series of test runs was completed at a Severe contraction ratio B2/B1 of 25%. A second series of runs was made at a Moderate contraction ratio of 50%, and a final series of tests was completed C H A P T E R 3

3-2 Revised Clear-Water and Live-Bed Contraction Scour Analysis Figure 3-1. Plan view of the test layout in CSU’s 8-ft-wide, 4-ft-deep, 200-ft-long tilting flume. (a) (b) 8.0ft Figure 3-2. Views of the 8-ft-wide, 200-ft-long flume: (a) an overall view; and (b) a view looking downstream.

Laboratory Testing Setup and Test Plan 3-3   at a Mild contraction ratio of 75% for both clear-water and live-bed conditions. Rigid-bed (non-erodible) tests were also conducted at a Severe contraction ratio using a plywood bed instead of a sand bed. Flow was recirculated through the flume for all test runs. Downstream of the contracted section, a sediment trap captured the sand, and sediment-free water was returned to the recirculation sump. For the live-bed tests, sediment was supplied to the upstream approach reach using a hopper equipped with a screw auger and movable distribution board. The flow capacity of the flume’s main pumps was ample for the range of flow rates needed for the tests. 3.3 Data Measurements and Instrumentation The laboratory tests included the following measurements: (a) Rating curves for flow depth, y1, before contraction scour occurs, were prepared for the two contraction configurations; B2/B1 = 0.25 and 0.75. The approach was to create a non- erodible bed with the plywood base of the flume. After the rating curves were established, the bed was covered with the 0.26-mm fine sand that constituted the bed material for the clear-water and live-bed tests. (b) Water surface profiles during selected tests for each configuration were measured to accurately determine the evolving bed and flow conditions upstream, through, and down- stream of the contracted section before and during scour. In particular, measurements were made of the flow depth, y1, during tests at frequent intervals from the movable data collection carriage. These measurements were made using SeaTek sonar transducers, which yield records of the time-variation of the evolving water surface. The resulting data were essential for evaluating the current scour depth equations in HEC-18, which compute scour as the difference between the flow depths on the initial bed and scoured bed. (c) Point velocity measurements during selected tests were taken using 3D Acoustic Doppler Velocimetry during most tests as scour progressed. These velocity data provided the relationship between velocity and scour depth as scour developed. (d) The main features of the contracting flow field were delineated using large-scale particle velocimetry (LSPIV), which was used through the region of contraction scour for selected tests. LSPIV provides quantitative digital visual records of flow contraction and large-scale turbulence features, indicating how they change as scour progresses (Ettema et al. 2010). The Fudaa software in use at CSU was utilized. Figure 3-3 shows an example of a water surface flow field delineated using LSPIV. (e) Temporal variation of bed elevations through the short and long portions of the contraction scour region were measured during the course of each test using an array of small SeaTek sonar transducers. (f) Equilibrium bathymetry was measured. Once an equilibrium scour depth was achieved, the equilibrium scour condition was LiDAR-scanned using CSU’s TOPCON TLS System LiDAR scanner. This provided a high-resolution digital map of the entire bed surface, not just a profile along the centerline. (g) Flow-field visualization using several visualization methods (notably, dye and digital photography) was used, as necessary, to provide diagnostic and documentary insight into the processes and extent of contraction scour. Data reports on all tests conducted were prepared, including a brief description of each test and associated electronic files of measured data from the various instrumentation arrays and post-processed information, such as time-varying data of bathymetry, water surface profiles,

3-4 Revised Clear-Water and Live-Bed Contraction Scour Analysis and velocity distributions. Data files include tabular and graphical information, as appropriate, and the following: • Metadata associated with each test (e.g., test number, date, discharge, contraction geometry, etc.) • Spatial reference for all instrument measurements • Temporal reference for all instrument measurements 3.4 Laboratory Test Plan During Phase II of the project, the following laboratory flume experiments were conducted. The experiments addressed the main knowledge gaps identified in Section 2.2. The program of experiments focused on the following conditions of contraction scour: (a) Clear-water contraction scour (b) Live-bed contraction scour (c) Rigid-boundary hydraulics of contracted reaches (d) Comparison of short-contraction and long-contraction scour depths and additional com- parison with normal flow depths in the approach and exit channels The sediment in the flume was a uniform fine sand; the particle size distribution is shown in Figure 3-4. The median particle size d50 is 0.26 mm and the geometric standard deviation σg = (d84/d16)1/2 is approximately (0.39/0.15)1/2 = 1.6. The specific gravity of the sand was 2.68. Figure 3-5 illustrates the water surface and bed profiles through a contracted reach as scour develops. The contraction ratios B2/B1 were 0.25, 0.50, and 0.75. The contracted sections were con- structed with a transition entry angle of 45°. Clear-water, live-bed, and rigid-bed contraction scour test conditions are summarized in Tables 3-1 through 3-3, respectively. The naming convention for the tests provides information on the sediment transport condition, contraction ratio, and approach velocity ratio. For example, Test CW_0.25-0.75 indicates a Clear-Water test at a contraction ratio B2/B1 = 0.25, and an approach velocity ratio Vn1/Vc = 0.75. Similarly, Test LB_0.50-2.0 indicates a Live-Bed test at a contraction ratio of 0.50 and an approach velocity ratio of 2.0. Figure 3-3. An example view showing CSU’s use of LSPIV to reveal the flow field (water surface) around and over a bendway weir locally constricting the flow. The dashed line indicates the axis of a bendway weir.

Laboratory Testing Setup and Test Plan 3-5   4 8 14 25 35 45 60 80 120 200 0 10 20 30 40 50 60 70 80 90 100 0.010.1110 Pe rc en t F in er b y w ei gh t - % Grain Size in millimeters GRAIN SIZE DISTRIBUTION Target gradation CSU delivery - Sample 1 CSU delivery - Sample 2 Sieve Size GRAVEL Fine SAND Coarse Medium Fine SILT or CLAY Figure 3-4. Sediment particle size distribution for the CSU contraction scour tests. Figure 3-5. A generalized plan view and profiles of water surface and bed through a contracted section as scour develops.

3-6 Revised Clear-Water and Live-Bed Contraction Scour Analysis Test Number Approach width B1 (ft) Contracted width B2 (ft) B2/B1 Discharge Q (ft3/s) Y2- tailgate (ft) Vn1/Vc Duration (hours) CW_0.25-0.55 8.0 2.0 0.25 2.24 0.58 0.55 18 CW_0.25-0.65 8.0 2.0 0.25 2.65 0.58 0.65 18 CW_0.25-0.75 8.0 2.0 0.25 3.05 0.58 0.75 30.5 CW_0.25-0.80 8.0 2.0 0.25 3.26 0.58 0.80 18 CW_0.50-0.55 8.0 4.0 0.50 2.26 0.58 0.55 18 CW_0.50-0.65 8.0 4.0 0.50 2.67 0.58 0.65 18 CW_0.50-0.75 8.0 4.0 0.50 3.09 0.58 0.75 18 CW_0.75-0.75 8.0 6.0 0.75 3.09 0.58 0.75 18 CW_0.75-0.85 8.0 6.0 0.75 3.50 0.58 0.85 18 CW_0.75-0.95 8.0 6.0 0.75 3.91 0.58 0.95 18 Table 3-1. Clear-water contraction scour tests. Test Number Approach width B1 (ft) Contracted width B2 (ft) B2/B1 Discharge Q (ft3/s) Y2- tailgate (ft) Vn1/Vc Duration (hours) LB_0.50-1.2 8.0 4.0 0.50 4.89 0.58 1.2 7.0 LB_0.50-1.4 8.0 4.0 0.50 5.70 0.58 1.4 7.0 LB_0.50-1.65 8.0 4.0 0.50 6.72 0.58 1.65 7.0 LB_0.50-2.0 8.0 4.0 0.50 8.14 0.58 2.0 7.0 LB_0.75-1.2 8.0 6.0 0.75 4.89 0.58 1.2 7.0 LB_0.75-1.4 8.0 6.0 0.75 5.70 0.58 1.4 7.0 LB_0.75-1.65 8.0 6.0 0.75 6.72 0.58 1.65 7.0 LB_0.75-2.0 8.0 6.0 0.75 8.14 0.58 2.0 7.0 LB_0.75-2.5 8.0 6.0 0.75 10.18 0.58 2.5 7.0 Table 3-2. Live-bed contraction scour tests. Test Number Approach width B1 (ft) Contracted width B2 (ft) B2/B1 Discharge Q (ft3/s) Y2- tailgate (ft) Vn1/Vc Duration (hours) RB_0.25-0.55 8.0 2.0 0.25 2.24 0.58 0.55 2 RB_0.25-0.65 8.0 2.0 0.25 2.65 0.58 0.65 2 RB_0.25-0.75 8.0 2.0 0.25 3.05 0.58 0.75 2 RB_0.25-0.80 8.0 2.0 0.25 3.26 0.58 0.80 2 Table 3-3. Rigid-bed contraction tests.

Laboratory Testing Setup and Test Plan 3-7   Figure 3-6 illustrates the flow regions associated with short- and long-contraction effects in a rigid-boundary system. In this figure, region “a” is uniform flow (UF); “b” is gradually varied flow (GVF), and “c” is rapidly varied flow (RVF). Important features of the test program were the emphasis on the following analytic consid- erations and outcomes: (a) Hydraulics of contracted flows (b) The temporal evolution of contraction scour along the short- and long-contraction reaches (c) The acquisition of detailed data and observations that facilitated additional numerical simulation (d) Information sufficient to conduct limited statistics on bedform development (e) The provision of sufficient information to advance the current methods for estimating contraction scour depth (f) Linkage to abutment scour, which relates closely to short-contraction scour No prior study had conducted as comprehensive a set of laboratory experiments as those conducted for this project. In addition, the experiments involved larger physical dimensions than used heretofore. 3.5 1D Hydraulic Model Calibration Procedure Comprehensive calibration for all clear-water, live-bed, and rigid-bed tests was performed using HEC-RAS. The rigid-bed runs were used to determine Manning’s n value for plywood, with a resultant average of 0.013. This value was then used for the wall roughness for the sand-bed tests under both clear-water and live-bed conditions. For each test, the final bed profile (after scour) was input to HEC-RAS. The Manning’s n value was varied to determine the best-fit value by minimizing the objective function: h h (3.1)i observed i predicted 2 1 n ∑( )χ = − Figure 3-6. Water surface and bed profiles through a non-erodible contracted section.

3-8 Revised Clear-Water and Live-Bed Contraction Scour Analysis where hi observed are the observed water surface elevations at the 13 measurement stations, and hi predicted are the water surface elevations at the same stations, as predicted by HEC-RAS for a particular Manning’s n. With the optimal n-value for plywood determined from the rigid-bed tests, the same approach was then used to determine the optimal n-value for the observed final conditions with a sand bed, including the ripple-dune bedforms. Figure 3-7 provides an example of the optimization process for CSU Tests RB_0.25-0.75 and CW_0.25-0.75. The optimal Manning’s n values for the plywood walls were determined from the rigid-bed tests. The optimal Manning’s n values for the sand bed (including bedforms) were then used with HEC-RAS to determine the initial (pre-scour) and final (post-scour) hydraulic conditions for the clear-water test. The results from the final HEC-RAS calibration for Test CW_0.25-0.75 are provided as an example in Figure 3-8. Note that the water surface elevations for initial conditions (shown as the fine black dashed line in Figure 3-8) and final conditions (solid blue line) could not have 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 O bj ec tiv e fu nc tio n = su m o f s qu ar ed d ev ia tio ns fr om o bs er ve d va lu es Manning's n Optimal Manning's n for plywood = 0.013 Optimal Manning's n for plywood walls and ripple-dune sand bed = 0.028 Figure 3-7. Optimization functions for determining best Manning’s n values, Tests RB_0.25-0.75 and CW_0.25-0.75. -1.0 -0.5 0.0 0.5 1.0 0 10 20 30 40 50 60 70 80 90 100 El ev ati on (ft ) Station (ft) Test CW_0.25-0.75 Plywood (HEC-RAS) Plywood (observed) Initial WS (HEC-RAS) Final WS (HEC-RAS) Final WS (observed) Initial bed Final bed Initial bed Final bed Figure 3-8. HEC-RAS calibration results for rigid-bed and clear-water tests, CW_0.25-0.75.

Laboratory Testing Setup and Test Plan 3-9   been estimated properly without the results of the rigid-bed (plywood) tests. Table 3-4 provides a summary of the HEC-RAS calibrations for all tests. Note that the bed shear and velocity values from the calibrated HEC-RAS models reflect a lumped parameter Manning’s n, which is a combination of plywood wall roughness, bed- particle (grain) roughness, and form drag caused by the ripple-dune bedforms. Test Number Optimal Manning's n Intermediate Contraction Hydraulic Conditions (sta. 40-80) Initial (pre-scour) Final (post-scour) Plywood Plywood walls, sand bed Average shear stress (lb/ft2) Average velocity (ft/s) Average shear stress (lb/ft2) Average velocity (ft/s) CLEAR-WATER TESTS CW_0.25-0.55 0.013 0.028 0.067 1.64 0.051 1.45 CW_0.25-0.65 0.013 0.023 0.062 1.91 0.039 1.57 CW_0.25-0.75 0.013 0.028 0.100 2.02 0.046 1.46 CW_0.25-0.80 0.013 0.023 0.078 2.18 0.037 1.59 CW_0.50-0.55 0.013 0.034 0.032 0.92 0.026 0.84 CW_0.50-0.65 0.013 0.043 0.064 1.03 0.056 0.97 CW_0.50-0.75 0.013 0.037 0.063 1.18 0.052 1.09 CW_0.75-0.75 0.013 0.037 0.034 0.82 0.027 0.79 CW_0.75-0.85 0.013 0.037 0.050 0.90 0.047 0.87 CW_0.75-0.95 0.013 0.038 0.049 1.00 0.048 0.95 LIVE-BED TESTS LB_0.50-1.2 0.013 0.030 0.098 1.83 0.066 1.55 LB_0.50-1.4 0.013 0.019 0.051 2.10 0.029 1.64 LB_0.50-1.65 0.013 0.027 0.108 2.19 0.056 1.62 LB_0.50-2.0 0.013 0.020 0.082 2.59 0.029 1.68 LB_0.75-1.2 0.013 0.037 0.035 1.28 0.031 1.20 LB_0.75-1.4 0.013 0.037 0.040 1.47 0.038 1.45 LB_0.75-1.65 0.013 0.037 0.058 1.65 0.044 1.46 LB_0.75-2.0 0.013 0.037 0.063 1.94 0.045 1.68 LB_0.75-2.5 0.013 0.037 0.098 2.18 0.060 1.76 Table 3-4. Optimal Manning’s n value, velocity, and shear stress from 1D HEC-RAS calibrations.