Bridge Scour in Nonuniform Sediment Mixtures and in Cohesive Materials: Synthesis Report

1. INTRODUCTION

This report presents the summary and synthesis of the various components of the experimental study entitled "Effects of Gradation and Cohesion on Bridge Scour" conducted at Colorado State University (CSU) from 1991 through 1996. This study encompassed experiments conducted under four major categories:

1. Effects of gradation and coarse material fraction on pier scour experiments.
2. Effects of gradation and coarse material fraction on abutment scour experiments.
3. Effects of cohesion on pier scour experiments.
4. Effects of cohesion on abutment scour experiments.

During this 5-year study, 5 different experimental flumes of varying sizes with 20 noncohesive and 10 cohesive sediment mixtures were utilized. In pier scour experiments, nine different pier sizes were tested; in abutment scour experiments, seven different abutment protrusion lengths were utilized. Depending on the purpose and scale of the experiments, varying degrees of data collection programs were implemented. The measurements utilized various levels of sophistication and collected detailed flow and sediment data in the vicinity of pier and abutment models. These experimental results were then utilized to quantify effects of coarse material fraction and cohesion on bridge scour. The details of the study were presented in a separate, six-volume set. These reports are:

In this synthesis report, the experimental facilities, experimental equipment and measurements, results, and analysis and conclusions for each of the components of the experimental study are presented individually in chapters 2 through 6. Each chapter tabulates the properties of the materials used, the approaching flow conditions, and the resulting scour measurements. The results from the various components of the study are also presented in terms of unified relationships and equations.

In chapter 2, results of experiments to study the effects of gradation and coarse bed material fraction on pier scour are presented. In the first phase of the study, effects of gradation and coarse bed material fraction were investigated experimentally in the 2.44-meter (m) wide by 60-m long tilting flume at the Engineering Research Center, CSU. Six different sand mixtures with the same median diameter, D₅₀, of 0.75 millimeter (mm), but with gradation coefficients ranging from 1.38 to 3.4 were used in the experiments. To study the coarse material fraction effects, mixtures with the same median diameter, D₅₀, and gradation coefficient, σ_g, but with varying D₉₀ and D₉₅ sizes were subjected to the same scour conditions. Experiments revealed that the coarse material fraction, rather than the gradation coefficient, is the controlling factor in pier scour. This set of experiments was limited to clear water scour around a circular pier with a diameter of 0.178 m. Flow intensities starting with incipient conditions were increased in the experiments until live-bed conditions were encountered. Extensive bed material samples from the scour hole and the approach were obtained and analyzed. Also, in chapter 2, the results of a smaller scale study for quantifying the effects of gradation on pier scour are presented using six sand mixtures with the same median sizes of 1.8 mm and 0.78 mm, but with different gradation coefficients. These smaller scale experiments, conducted in a 0.61‑m wide by 18‑m long experimental flume using circular piers with a diameter of 0.076 m, examined the effects of model scaling. These experiments, along with additional experiments conducted in the 5.1-m wide river mechanics flume, were used in defining the dependence of scour on pier diameter. It was found that local scour, D_s, varied with pier diameter, b, according to a simple power relationship: D_s ~ b^0.6. Finally, the study was extended to larger sediment sizes by conducting experiments using uniform and coarse fraction enriched gravel mixtures. As a result of this effort, a new equation was developed expressing pier scour in terms of the dimensionless excess velocity factor, flow depth, pier diameter, and a correction factor for the coarse fractions present in mixtures was derived. The new method was tested with available data from previous research. This equation shows that gradation effects are not constant through the entire range of flow conditions but vary with flow intensity.

In chapter 3, effects of size fraction on clear water abutment scour was studied using 16 different sediment mixtures ranging from very fine sand to gravel sizes. Experiments were conducted in 0.61-m, 2.44-m, and 5.18-m wide flumes to cover a range of geometrically similar flow conditions to analyze scale effects in physical modeling of abutment scour. In the experiments, for mixtures with median (D₅₀) sediment sizes of 0.1 mm, 0.55 mm, 0.78 mm, and 1.8 mm, the coarse size fractions present in mixtures corresponding to D₉₀ and D₉₅ sizes were varied while keeping the median size and sediment gradation constant. Results of abutment scour experiments were normalized using corresponding abutment scour values for uniform mixtures with the same median D₅₀ sediment sizes. Similar to pier scour analysis, a coarse fraction reduction factor for graded mixtures was derived for different gradation coefficients. It was found that, for very low flow intensities and for intensities approaching live-bed conditions, the coarse material fractions have little effect. However, for a wide range of intermediate dimensionless flow intensities, abutment scour is very much dependent on the coarse material fraction and can be as low as 15 percent of scour in uniform material with the same D₅₀ size. As a result of abutment scour experiments, two new equations were developed. The first equation was derived from a 0.1 mm uniform sand mixture and defines an envelope relationship. The second equation applies to mixtures with coarse fractions. A coarse size fraction compensation factor W_g is presented to account for the presence of varying amounts of coarse material in sediment mixtures under different dimensionless flow intensities. These new equations represent the experimental data very accurately but have not been tested with field data.

In chapter 4, effects of clay content on bridge scour are presented. For both pier and abutment scour experiments, clayey sand mixtures with varying amounts of clay were subjected to different approach flow conditions. The resulting scour values from these experiments were normalized with the corresponding scour experienced in pure sand used as the base material. In pier scour experiments, for Montmorillonitic clay mixtures, results showed that for sandy clays, increasing the clay content to up to 30 percent may reduce scour by up to 40 percent. Beyond a certain clay content (30 to 40 percent for the present mixture) parameters such as compaction, initial water content, degree of saturation, shear strength, etc., dominated the pier scour. Similarly, in abutment scour experiments, for Montmorillonitic clay mixtures, experimental results showed that for clayey sands, increasing the clay content up to 30 percent may reduce scour by up to 40 percent. For Kaolinitic clay mixtures the results are more dramatic; for the clear water scour range, scour reduction can be up to 80 percent of that observed in pure sands. Beyond a certain percentage of clay content (30 to 40 percent for the present mixture) parameters such as compaction, initial water content, degree of saturation, shear strength, etc., dominate the cohesive material scour. These experiments were conducted for the range of flows that represented clear-water scour conditions for the base material. However, it is expected that for velocities much in excess of those that could mobilize the base material, ultimate scour would be reached.

In chapter 5, effects of cohesion on pier scour are investigated experimentally using 1.22-m, 2.44-m, and 5.18-m wide test flumes at the Engineering Research Center, CSU. In these experiments, flows with corresponding Froude numbers ranging from 0.2 to 1.4 were used to investigate scour taking place in saturated and unsaturated Montmorillonitic sandy clay mixtures. Regression equations relating flow and selected cohesive soil parameters to pier scour were derived for a fixed circular pier geometry and a constant flow depth. These experiments showed that pier scour in cohesive material can be expressed in terms of cohesive soil parameters that can be obtained in the field. The equations derived from this study do not attempt to relate effects due to pier geometry, size, flow depth, gradation of sand in the mixtures, etc., and therefore are not intended for general application. These relationships are derived to explain the variability of pier scour with cohesive properties.

In chapter 6, effects of cohesion on abutment scour were investigated experimentally using 1.22-m and 2.44-m wide test flumes. In abutment scour experiments, Montmorillonitic and Kaolinitic sandy clay mixtures were used to examine clay mineralogy effects as well as effects due to compaction, initial water content, and soil shear strength. Froude numbers in these experiments varied between 0.2 and 0.6 to cover a wider range of hydraulic conditions. Regression equations relating flow and clay properties to abutment scour express scour in terms of sands scour and were therefore developed for the range of flows that represented clear water scour conditions for the base material.

The data resulting from the experimental study are presented in chapters 2 through 6 in tabulated, consistent units. The majority of these data were derived from the six-volume report mentioned above. The results of pier scour experiments in gravel and the results of pier scour experiments to study effects of pier width were conducted beyond the initial set of experiments and are included in this report in chapter 2.

Contents | Next