The influence and interaction of sediment grain size distributions and channel flow hydraulics over the formation of armor layers in gravel bed rivers has yet to be fully defined. Imbrications and microclusters develop as part of the armor layer and contribute to the overall increased stability of armored beds. Recent research on microclusters has focused on how the turbulent flow field is affected by their presence, which has not been fully elucidated. This research will connect the processes of bed armoring, microcluster formation, and bed stability and provide the quantification necessary to manage flow releases so that the downstream channel surface does not erode and potentially harm the downstream ecosystem. Flume experiments will create armored beds testing 4 sediment mixtures against 4 flow rates. Microcluster density and bed topography will be measured from laser profile scans of the bed surface. Acoustic Doppler Velocimeters (ADVs) will record point velocities (u,v,w) and fluctuations (u?,v?,w?) over the armored surface and around microclusters, from which turbulence descriptors, such as Turbulent Kinetic Energy (TKE), are calculated. Flow data will be analyzed using the Double Averaged Navier-Stokes (DANS) approach and quadrant analysis, which together with the detailed bed topography will provide insight into the momentum transfers occurring and the coherent flow structures present in the flow field at specific locations over the armored bed. Using ADV measurements, it will be possible to differentiate the boundary stresses as due to microcluster, topographic surface, viscous stress, or stress from the interaction of multiple clusters. At the conclusion of the proposed research, the effect of bed substrate grain size distributions and flow rate during armor layer formation will be quantified and linked to the creation of structure in the armor layer topography and the stability of the channel bed. An understanding of the controlling variables over armor layer formation and bed stability will be gained. This research will provide a means of predicting the flow rate that will break the downstream armored bed layer as a function of the substrate in the channel and the existing bed microtopography, both of which can be measured in the field.

Project Report

A gravel bed river surface is often defined by whether or not it has an armored surface, and clusters have long been recognized as developing within the surface armor. In this research project, detailed laboratory measurements were made of the turbulent flows around clusters naturally formed during the armoring process in a gravel bed channel. Through a series of flume runs using three different armoring flows rates, we created armored beds and clusters using gravel sediment to which sand was added to create grain size distributions of 1%, 9%, 24%, and 38% sand. Flows were measured around 134 clusters and also around unclustered areas of the armored beds. The impact of small scale roughness associated with the armoring process was limited to the inner flow region, increasing turbulent flows only near the bed. The cluster form extended above the local, mean bed surface and into the flow field farther than the topography created by individual clasts on an uneven armored surface. Where a cluster was present, the impact of flow hydraulics extended throughout the flow field as turbulent flows near the bed transferred momentum and energy to the outer flow area. Flow separated as it was forced over the cluster crest, creating an increase in the turbulent flow properties and formation of a flow recirculation cell where the flow re-attached downstream of the cluster. Although a similar recirculation cell formed at each cluster, the strength and orientation of these cells varied with bed condition. As the sand content of the channel bed increased, the bed transitioned from a framework to a matrix bed subsurface structure. This change correlated to an increase in the energy and momentum in the recirculation cell, but a reduction in cluster influence across the flow profile. In a low sand content bed, essentially a framework bed, the cluster increased local turbulent flows throughout the vertical flow profile while the magnitude of the impact on the flow profile increased with flow rate. Framework beds provided the greatest increase in bed surface stability upon armoring as indicated by both the time required to mobilize the bed surface and the dimensionless shear stress ratio. Clusters formed on sediment beds transitional from framework to matrix were distinguished by their reduced yet evenly distributed influence on local flows. Turbulent flows around clusters formed from a bed with sand content high enough to be considered a sand matrix developed recirculation cells that were of high magnitude but oriented low in the flow profile, remaining predominantly below the cluster crest. These clusters had a large impact only on turbulent flows in the near bed region. The lowest apparent flux of momentum and energy to the outer flow region occurred over clusters from the sand matrix beds which were also the least resistant to mobilization. The spatial arrangement of clusters on the different armored channel beds was also investigated for its role in increasing or decreasing local turbulence. Clusters developed as: isolated cluster, coupled clusters- a pair of clusters immediately adjacent; grouped clusters - three clusters in close proximity. The net effect of increasing cluster density was an increase in the magnitude and variability of the turbulent flow field around coupled clusters, followed by an overall dampening with an increase to grouped clusters. A larger amount of energy transfer to the outer flow profile occurred in the flow field over the coupled clusters, while momentum transfer was greater through the flow field over the entire grouped clusters. The results have direct implications for river restoration and management. For example, in managed systems where the release of water is controlled, the potential impact of the release on downstream bed stability is often a concern. These results show that a range in armored, clustered bed stability can be expected based on the overall sand content of the sediment bed. Where the channel downstream of a dam includes cluster bedforms important for aquatic habitat, a larger flow release may be planned if the subsurface is a matrix bed with a large sand content. Rock clusters are popular feature in river restoration design. However, the impact of created clusters on channel hydraulics will depend on the bulk sand content of the channel bed. Under conditions of a framework bed, the inclusion of individual cluster forms will generate highly turbulent flows, which may be desirable for aquatic habitat, but these clusters will also be unstable. If an upstream supply of cluster sized rocks is not present or a threshold condition is desired, clusters would not present a long term design option. It is equally possible to envision a river restoration scenario where high magnitude turbulent flows are desired near the bed surface, low in the flow profile. If the channel bed has a matrix character, the inclusion of cluster forms would be a beneficial to meet restoration goals.

National Science Foundation (NSF)
Division of Earth Sciences (EAR)
Application #
Program Officer
Thomas Torgersen
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Virginia
United States
Zip Code