Now more than ever, end users and designers require documentation on the erosion and sediment control (ESC) performance of rolled erosion control products (RECPs) in order to compare available materials and develop cost-effective designs. Many factors affect this performance, including (1) depth, mean velocity, and duration of flow; (2) cross-sectional geometry and longitudinal slope of the channel; (3) physical, mechanical, and hydraulic properties of the RECP; (4) anchor pattern and density; (5) soil type and erodibility; and (6) vegetative type and density (Gharabaghi et al., 1998).
Performance analysis via accelerated flow testing-crucial to successful RECP engineering, design, and installation-is an expensive undertaking. And because of the complex nature of fluid/liner/sediment interactions, laboratory and field studies to date have not been able to answer why and how some RECPs can protect a channel against erosion better than others or which physical, mechanical, or hydraulic properties of RECPs most influence their performance. End users often have had no choice but to rely solely on readily available or easily obtainable indices of physical, mechanical, and hydraulic properties of the RECPs to determine which products to use (Clopper, 1998).
The laboratory experiments and numerical simulation performed in the present study have provided a significant step toward (1) development and verification of semitheoretical and numerical models for flow above, through, and below RECPs; (2) identification and quantification of mechanisms by which RECPs control and reduce erosion and sediment transport; (3) identification of influential indices of physical, mechanical, and hydraulic properties of RECPs with respect to their erosion control performance; and (4) modification of the present design approach for RECPs in channel applications.
In the improved design approach suggested in this study, the concept of permissible shear-stress design for RECPs is modified based on a recognition of two main modes of failure of RECPs: soil-loss failure and liner-deformation failure. In the modified design approach, the two primary performance properties of RECPs-in accordance with these two modes of failure-are rate of soil loss and maximum shear capacity of the liner.
We are entering a new era of global awareness, concern, willingness, and active cooperation to conserve, improve, and sustain the planet’s natural resources and environment. Every year large storms dump heavy rains and cause flooding, landslides, and damage to roads, houses, and other structures. Excessive erosion and heavy sediment loads in waterways have enormous social and economic impacts. Concerns about erosion and sedimentation impacts on the quality and vitality of waterways have inspired innovative ideas, such as natural restoration of waterways, and have led to the introduction of new erosion control materials to the market. Use of live plant materials and vegetation for stabilizing waterways is gathering momentum as we are beginning to realize that nature has developed its own effective protection system against erosion. Grass is a low-cost and environmentally friendly means of protecting and stabilizing earth surfaces that are subject to intermittent flows of water (CIRIA, 1987). Vegetative lining provides a filtering medium for sediments and other pollutants and improves the quality of runoff before it enters natural or constructed wetlands. It permits infiltration and exfiltration, both of which are desirable in stormwater drainage channels. Vegetative lining also provides habitat opportunities for local flora and fauna and is aesthetically pleasing (Chen and Cotton, 1988).
The establishment of vegetative lining takes one to two growing seasons, depending on climate and weather conditions. Prior to the establishment of vegetation, there is a high risk that the seedbed might be washed away during initial storms. Since the mid-1920s, a variety of inexpensive, temporary RECPs have been introduced to the market. Temporary RECPs are composed of biologically and photochemically degradable materials that have been manufactured or fabricated into rolls and designed to temporarily reduce soil erosion and to enhance the establishment of vegetation. Vegetative lining has the disadvantage of being limited by the magnitude of erosive force that it can sustain. For waterways with highly erodible sandy soils and high flow velocities, vegetation alone might not be capable of stabilizing the soil. This means that a high-flow event can cause soil to dislodge from the surface, and the grass roots will eventually have nothing to grasp. Many long-term, nondegradable RECPs have been introduced to the market and not only furnish erosion protection prior to the establishment of vegetation, but also extend the erosion control limits of vegetative lining through root or stem reinforcement for the life of a project (Allen, 1997).
RECPs have been very effective in reducing the risk of failure and improving the erosion control performance of vegetative linings. However, RECPs generally are highly permeable, buoyant, and very flexible materials. Particularly prior to the establishment of vegetation, buoyancy forces, together with hydrodynamic shear-drag and uplift forces, can cause RECPs to float above the bed and separate from the soil surface, allowing a portion of flow to “pipe” between the bed and the liner and generate a wavy-geometry liner. Vertical oscillation of the liner, as a result of the liner’s dynamic interaction with the drag and lift forces of flow, can cause significant turbulence underneath the liner and cause erosion. The eroded sediments underneath the liner then have to pass through it to the higher-velocity flow above the liner in order to be transported downstream. The presence of a relatively high pressure zone on the front side of the wavy liner and low-pressure zones on the crest and on the lee side of the wavy liner can cause some fluid above the liner to pass through it on the front side and some equal amount of fluid from beneath the liner to exit through the liner to the higher-velocity flow above the liner (Gharabaghi et al., 1998).
Naudascher (1991) observed that in high-velocity turbulent flows, the great magnitude of the pressure fluctuations on the liner combined with its high permeability can result in a “blow vacuum” effect in which a significant amount of fluid passes through the liner in both directions. The exchange of fluid between the flow domains above and below the liner results in a net movement of sediments from under the liner to the higher-velocity flow above the liner. These mechanisms can play a significant role in the transport of eroded sediments through channel systems lined with RECPs. Thus, it is important for their design that there be an understanding of the mechanisms by which RECPs control and reduce erosion and sediment transport; hydraulics of fluid flow above, through, and below RECPs; hydrodynamics of shear and normal forces acting on the liner and on the sediment particles beneath it and influential indices of physical, mechanical, and hydraulic properties of RECPs with respect to their erosion control performance.
Evaluation of the performance of RECPs has usually been based on the following criteria (Chen et al., 1988; Northcutt, 1996; Clopper et al., 1998): (1) a soil-loss criterion-that is, the average channel deformation in the design flow event should be less than the standard permissible soil loss; (2) a liner-deformation criterion-that is, the deformation of the RECP and the development of ripples, sags, and tears from hydrodynamic drag-and-lift forces of flow should be less than the standard permissible value; (3) a vegetation growth criterion-that is, the vegetation density achieved over a single March-through-December growing season should be greater than the standard minimum vegetation density; and (4) a durability criterion-that is, photodegradation or biodegradation should not reduce erosion control capability and structural strength of the RECP below minimum standard values during the expected lifetime of the product. Although there are no standards for these RECP performance criteria as yet, the American Society for Testing and Materials (ASTM) committee D18.25 is working on developing such standards.
Sanders et al. (1990) and Israelsen et al. (1991) tested the performance of several different types of RECPs in nonvegetated conditions. The testing procedures in both studies consisted of several sequential 30-minute runs with increasing flow rates. After each run, the soil loss was measured.
The Texas Department of Transportation, through the Texas Transportation Institute, tested the performance of RECPs in vegetated conditions (Northcutt, 1996). Following installation, each product experienced a 90-day waiting period to promote the initial growth of vegetation to a minimum density of 70% prior to initiating a series of 20-minute runs with increasing shear stresses in subsequent runs.
The rates of soil loss (that is, soil loss per unit time) reported in the 1990, 1991, and 1996 papers noted above remained more or less constant during sequential runs and did not change with increasing shear stress from 98 to 384 Pascals. Indeed, if the sequential runs were continued and shear stresses were increased beyond the maximum shear capacity of the RECPs, they eventually would have failed and the rate of soil loss would have increased radically. In light of these observations, the ESC performance of RECPs can be characterized by (1) the rate of soil loss, which appears to be a constant for a given RECP, soil type, and vegetation cover as long as the shear stress does not exceed the bearing capacity of the RECP, and (2) the maximum shear capacity of the RECP, which is a function of the tensile strength of the RECP, the stapling pattern and density, and the vegetation cover and density. The most influential index properties of the RECPs with respect to their ESC performance-that is, the rate of soil loss and the maximum shear capacity-are permittivity according to ASTM 04491 and initial tensile modulus according to ASTM 04595, respectively.
The erosion control industry has adopted the concept of permissible shear stress for the design of RECPs (Chen and Cotton, 1988; Northcutt, 1996). Permissible shear stress for RECPs is defined as the maximum shear stress that a liner can withstand for at least a standard duration of time with an average channel deformation less than the standard permissible value and with negligible damage to the channel lining. The standard permissible soil loss is defined as the maximum soil loss that would not significantly affect the germination of the seedbed and the establishment of vegetation lining (Northcutt, 1996). According to CURIA (1987), the permissible velocity (or permissible shear stress) of RECPs is a strong function of the duration of the flow event and decreases significantly as the duration of the flow event increases. Depending on the climatic conditions of the site and the characteristics of the watershed draining into the channel, the duration of the design flow event might be greater or less than the adopted standard duration of flow for the performance tests (that is, 20 or 30 minutes).The design flow events for RECPs should include both short-duration, high-intensity flow events and long-duration, low-intensity flow events. It is more likely that the liner deformation criterion would fail during the short-duration, high-intensity flow event, while the soil-loss criterion would fail during the long-duration, low- (or moderate-) intensity flow event. In a satisfactory design, the duration of the design flow event multiplied by the rate of soil loss should not exceed the standard permissible soil loss, and the design shear stress on the liner should not exceed the maximum shear capacity of the RECP.
Dickinson et al. (1991) and Gharabaghi et al. (1998) observed that prior to the establishment of vegetation, buoyancy forces can cause RECPs to float above the bed and separate from the soil surface, allowing a portion of flow to “pipe” between the bed and the liner. Vertical oscillation of the liner, as a result of the dynamic interaction of the liner with the drag-and-lift forces of flow, can cause significant turbulence underneath the liner and cause erosion. The time-averaged, three-dimensional, wavy geometry of a flexible liner depended on (1) physical, mechanical, and hydraulic properties of the RECP; (2) the anchor pattern and density of the RECP; and (3) hydrodynamic drag-and-lift forces of flow on the liner. Because of the liner’s high permeability, significant discharge of turbid water and sediments could occur through the liner to the high-velocity flow above the liner and be transported downstream.
In the present study, an investigation examined the applicability of existing hydraulic models of the natural environment for turbulent shear flows over wavy, movable, and permeable boundaries for the prediction of flow resistance of RECPs in channel applications prior to the establishment of vegetation. Laboratory experiments were designed to study the flow-resistance relationship for RECPs in channel applications prior to the establishment of vegetation. Preliminary tests indicated that the flow resistance of the RECP decreased with an increase in flow velocity, depth of flow, and the anchor density of the liner. To investigate the significance of anchor density on flow resistance, two extreme cases were considered: a very high anchor density, which resulted in flat-bed flow conditions, and a very low anchor density, which resulted in wavy-bed flow conditions. A range of flow conditions was tested for each case by establishing steady, uniform flows along the flume and by measuring depth of flow, flow rate, slope, and turbulent velocity profiles at the centerline of the flume.
Laboratory experiments were performed at the Hydraulics Laboratory of the National Water Research Institute, CCIW, in Burlington, ON. The test facility included a 26-m-long, 1-m-wide, 0.75-m-deep variable-slope flume with Plexiglas sidewalls, hydraulic jacks to change the longitudinal slope of the flume within the range of 0-5%, a honeycomb flow tranquilizer at the entrance of the flume, and a slide gate at the downstream end. It also included a pump station comprising three pumps with capacities of 0.142, 0.284, and 0.426 m3/sec., respectively, which recirculated the flow through the system; a large volumetric tank for accurate measurement of flow rate in the system; a four-wheeled measurement cart mounted on steel rails over the flume; a high-precision Acoustic Doppler Velocity (ADV) meter that could measure all three flow velocity components at a sampling rate of 26 Hz in a remote sampling volume 0.050 m below the probe; point gages; a Conductance Water Surface Probe; and associated electronics.
A 1-m-wide strip of the selected RECP was attached along the flume bed with a stapling interval of 1 m in the longitudinal direction and 0.9 m in the transverse direction. At the entrance of the flume, right after the honeycomb flow tranquilizer, a 1- x 1-m plastic sheet was used to sandwich the front edge of the liner and attach it to the bed with a high stapling density, resting heavy lead bars on the sandwiched liner. Seven flow rates (from 0.05 to 0.35 m3/sec.) at seven depths (from 0.1 to 0.4 m) provided 49 test runs. To establish uniform flows in the flume, the longitudinal slope of the flume and the depth of flow at the downstream end were adjusted using a time-consuming trial-and-error procedure to obtain almost-equal (within the accuracy of 0.001 m) depths along the flume. The normal depth of flow and slope were recorded for each run. The steady flow rate in the flume was measured accurately using the large volumetric tank.
The three-dimensional geometry of the liner was measured at 0.25-m intervals in both longitudinal and transverse directions on a 1- x 1-m test section using a point gauge mounted on the measurement cart. Since the flexible lining was moving under hydrodynamic forces of flow, the maximum and minimum elevations of the liner were recorded at each measurement point during a two-minute observation period. Hydrodynamic forces of flow generated traveling waves on the flexible liner. The amplitude, period, and celerity of the traveling waves were monitored using a video camera. Velocity profiles were measured at the centerline of the flume, with approximately 20 measurement points for each profile, using the ADV meter, at a sampling rate of 25 Hz for a period of 2 mm at each measurement point.
The velocity profiles at the centerline of the flume for nine different runs for the flat-bed tests with mean flow velocities range from 0.1 to 1.1 m/sec. and depths of flow range from 0.3 to 0.4 m. Velocity profiles in the flat-bed tests departed from the log law in the outer layer of flow and obeyed the Coles’ (1956) log-wake law with a constant surface roughness parameter Ks and a constant zero plane displacement parameter d. The semitheoretical (based on the law-of-the-wall) flow-resistance relationship suggested by Simons and Senturk (1992, Equation 1 below) was applicable for the flat-bed tests with reasonable accuracy.
U/U*=1/k x 1n (r/Ks) + 6.25
where: U, U*, k, r, and Ks are mean velocity of flow, shear velocity, Von Karman constant, hydraulic radius, and equivalent surface roughness, respectively. The empirical constant of 6.25 in this relationship might slightly depend on the cross-sectional geometry of the channel and the depth of flow. The sidewall correction procedure of Vanoni and Brooks (1957) was implemented to reduce the effect of the cross-sectional geometry and depth of flow on the empirical constant of the flow-resistance relationship.The stapling interval was 1 m in the longitudinal direction and 0.9 m in the transverse direction, with a simple rectangular pattern. When the velocity of flow was very low, buoyancy forces caused the liner to float above the bed to an average elevation of about 0.1 m at the centerline of the flume. As the flow velocity was increased, the hydrodynamic shear and normal forces caused the liner to deform to a flatter and more skewed-toward-downstream geometry.
The velocity defect distributions were measured at the centerline of the flume and at the midpoint between stapling points for 12 different runs for the wavy-bed tests, with mean flow velocities ranging from 0.2 to 1.1 m/sec. and depths of flow ranging from 0.2 to 0.4 m. Unlike the flat-bed tests, in the wavy-bed tests, the equivalent surface roughness parameter Ks and the zero plane displacement parameter d were not constants but depended strongly on the three-dimensional geometry of the liner. The zero plane displacement d played a role in the calculation of the hydraulic radius r in the flow-resistance relationship and was very small (in the order of millimeters) for the flat-bed tests but quite large (approximately 40% of the mean elevation of the liner) for wavy-bed tests.
The drag force on the flat-bed runs was from surface roughness only, while the drag force on the wavy-bed runs was from surface and form roughness.
Computational fluid dynamics (CFD) is in part the art of replacing the governing partial differential equations of fluid flow with numbers and advancing these numbers in space and/or time to obtain a final numerical description of the complete flow field of interest (Anderson et al., 1992). CFD codes use “turbulence models” to represent the small-scale effects of turbulence. Most CFD codes make use of a two-equation turbulence model called the standard k-e model, which was formulated more than 20 years ago and has shown to have the potential to mimic nature closely. The role of CFD in engineering predictions has become so strong that today it can be viewed as a new “third dimension” to the other two classical methods of pure experiment and pure theory. Over the last 10 years, advances in computer technology have made it possible to thoroughly test increasingly complex turbulence hypotheses and apply them in practical calculations (Rodi, 1993).
In the present study, the CFD code ADINA was used to simulate flow over, through, and underneath an erosion control channel liner. The ADINA System has a unique capability of analyzing fluid-structure interaction problems and is based on the Finite Element Method (FEM). In FEM, the flow domain is divided into a number of simply shaped regions called finite elements. Porous media elements with a known (i.e., measured) permeability, thickness, and geometry were used to model the RECP. Appropriate boundary conditions were implemented for the inflow, outflow, free surface, channel bed, and interface between the liner and the fluid. The resulting nonlinear system of equations was solved to determine the velocity components and pressures at every node. The numerical simulation of flow showed that the presence of a relatively high pressure zone in the front side of the wavy liner and low-pressure zones on the crest and lee side of the wavy liner can result in the movement of the turbid water and sediments through the liner to the higher-velocity flow above the liner. This mechanism can play a significant role in the transport of sediments through channel systems lined with RECPs.
(1) The flow-resistance relationship for the RECPs in channel applications prior to the establishment of vegetation depends on not only the surface roughness of the liner but also the waviness of the liner (i.e., form drag). The flow-resistance relationship for flat-bed conditions obeys the Coles’ (1956) log-wake law with a constant surface roughness Ks and a constant zero plane displacement d. The flow resistance relationship for wavy-bed conditions obeys the log-wake law with an equivalent surface roughness Ks and a zero plane displacement d.
(2) Prior to the establishment of vegetation, buoyancy forces can cause RECPs to float above the bed and separate from the soil surface, allowing a portion of flow to “pipe” between the bed and the liner. Vertical oscillation of the liner, as a result of the dynamic interaction of the liner with the shear-drag and uplift forces of flow, can cause significant turbulence underneath the liner and cause erosion. The numerical simulation of flow showed that the presence of a relatively high pressure zone on the front side of the wavy liner and low-pressure zones on the crest and lee side of the wavy liner can result in the movement of the turbid water through the liner to the higher-velocity flow above the liner. In high-velocity turbulent flows, the great magnitude of the pressure fluctuations on the liner combined with the high permeability of the liner can result in a “blow-vacuum” effect in which a significant amount of fluid can pass through the liner in both directions. The exchange of fluid between the flow domains above and beneath the liner could result in the net movement of sediments from below the liner to the higher-velocity flow above the liner. This mechanism can play a significant rote in the transport of eroded sediments through channel systems lined with RECPs.
(3) The most influential index properties of RECPs with respect to their ESC performance are permittivity according to ASTM 04491 and initial tensile modulus according to ASTM 04595, respectively. Less permeable liners can prevent eroded sediments from entraining in the higher-velocity flow above the liner. Liners with higher tensile strength and flexural rigidity can have less deformation as a result of hydrodynamic shear-drag and uplift forces of flow and can remain in close contact with the soil.
(4) The rate of soil loss remained approximately constant for a given RECP, soil type, and vegetation cover as long as the shear stress on the liner did not exceed the maximum shear capacity of the liner. The maximum shear capacity of an RECP is defined as the threshold shear stress beyond which the rate of soil loss would increase rapidly, resulting in accelerated erosion. For example, the rate of soil loss during the last runs for Miramat 2400F and NAG P300 increased to a value almost as high as the rate of soil loss from the bare soil control. The maximum shear capacity of RECPs is a function of their tensile strength, the anchor density, and the vegetation cover. The application of RECPs would (1) significantly reduce the rate of soil loss compared to the unprotected, bare soil conditions; that is, prior to the establishment of vegetation, and (2) significantly increase the maximum shear capacity of the vegetation lining through root or stem reinforcement.
(5) The design flow events for RECPs should include both short-duration, high intensity flow events and long-duration, low-intensity flow events. It is more likely that the liner-deformation criterion will fail during short-duration, high-intensity flow events, while the soil-loss criterion will fail during long-duration, low- (or moderate-) intensity flow events. In a satisfactory design, the duration of the design flow event multiplied by the rate of soil loss should not exceed the standard permissible soil loss, and the design shear stress on the liner should not exceed the maximum shear capacity of the RECP.
