Relationship Between Erodibility and Properties of Soils (2019)

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.

20 This chapter describes some of the most common tests developed over the past century to quantify soil erodibility. The drawbacks and advantages of these testing methods are evaluated and identified. Generally, erosion testing is divided into the following two types: 1. Laboratory erosion testing and 2. Field erosion testing. The erosion tests presented in this chapter are divided into two sections: laboratory tests and field tests. Some of the erosion tests have applications in both the laboratory and the field; however, they are discussed only once, in either Section 2.1 or Section 2.2. 2.1 Laboratory Erosion Testing 2.1.1 Erosion Function Apparatus In the early 1990s, the idea of the erosion function apparatus (EFA) was first developed and established by Briaud at Texas A&M University (TAMU) (Briaud et al. 1999, 2001a, 2001b). Today, the EFA is being manufactured by Humboldt, Inc., and is used widely by many engineer- ing organizations. This test was originally developed to evaluate the erodibility of a wide range of both cohesive and noncohesive soils, including gravel, sand, clay, and silt. Figure 4 shows a schematic diagram and image of the EFA device. Soil samples are taken by using ASTM standard Shelby tubes with an outside diameter of 76.2 mm (ASTM D1587). With soft rocks, a rock core sample can be extracted and placed into a Shelby tube for rock erosion testing. A pump is used to drive water, the eroding fluid, into a 1.2-m-long rectangular cross section of 101.6 × 50.8 mm, as shown in Figure 4. The water flow can be adjusted with a valve, and the average water velocity is measured by a flow meter in line with the flow. One end of the Shelby tube is placed on the bottom of a circular plate that is connected to a piston. The piston is designed to push the sample up into the flow when necessary. The sample is pushed upward until it becomes flush with the bottom of the rectangular cross section pipe. The test procedure is as follows (Briaud et al. 2001b): 1. Place one end of the Shelby tube on the circular plate piston and push it up until it becomes flush with the bottom surface of the rectangular cross-section pipe. 2. Fill the rectangular pipe with water and wait for 1 hour. 3. Initiate the flow with a small flow velocity, typically 0.2 m/s. 4. Start recording time. Hold the sample surface flush with the bottom of the rectangular pipe during the induced flow velocity. The test operator needs to make sure that the soil surface C H A P T E R 2 Existing Erosion Tests

Existing Erosion Tests 21 is kept flush at all times by pushing the soil with the piston as it is eroded by the water and maintaining a level interface. Continue this procedure until 50 mm of the soil is eroded or 30 min have passed. Read the protrusion height by observing the change in the height of the bottom of the piston. 5. Repeat Steps 3 and 4 for a new and higher flow velocity (i.e., 0.2 m/s, 0.6 m/s, 1 m/s, 1.5 m/s, 2 m/s, 3 m/s, 4.5 m/s, and 6 m/s). The scour rate versus flow velocity is plotted. Figure 2 (Chapter 1) shows an example of EFA test results as both erosion rate versus velocity and erosion rate versus shear stress. The shear stress on the eroded surface of the soil is calculated by using the Moody chart (Moody 1944). 1 8 (7)2f vt = r where t = shear stress (Pa), r = density of water (1,000 kg/m3), v = flow velocity (m/s), and f = friction factor obtained from Moody chart. Advantages 1. Minimize the sample disturbance effect, as it takes the sample directly from the field using Shelby tubes. 2. Can be used for natural samples as well as man-made samples. 3. Gives the critical velocity and the critical shear stress. Can give the erosion function directly. 4. EFA test results are directly used as input to the TAMU-SCOUR method for bridge scour depth predictions [see Hydraulic Engineering Circular No. 18 (Arneson et al. 2012, Chapter 6)]. (a) Cross Section of Pipe 76.2 mm Pressure Port 50.8 mm 1016 mm Pressure Port 101.6 mm τ 50.8 mm Water Flow Soil PISTON PUSHING AT THE RATE = Z (b) Figure 4. EFA: (a) schematic diagram and (b) image (courtesy of TAMU).

22 Relationship Between Erodibility and Properties of Soils 5. The EFA can test the erodibility of the soil at any depth as long as a sample can be recovered. 6. While the EFA test is a surface erosion test, it can be used to evaluate internal erosion as well, because the EFA erosion function represents the erodibility of the soil at the element level. 7. The EFA can be used to test from very soft to very hard soils. This test has very broad applications. Drawbacks 1. Shear stress is indirectly measured from the average velocity according to the Moody chart, so the measurement might not be accurate. Also, the average flow velocity is used in the calculations instead of the actual velocity profile. 2. In cases in which field samples are required, obtaining samples is difficult and costly. The test needs to be done on the sample before the sample is affected by long periods of storage. 3. Particles larger than 40 mm cannot be tested with confidence, as the diameter of the sampling tube is 75 mm. 4. The EFA is a fairly expensive device (around $50,000 to purchase). Several other organizations have utilized the concepts and principles of the EFA and developed similar devices, such as the Sediment Erosion Rate Flume (SERF) and the ex situ scour testing device (ESTD), which are also presented in this chapter. 2.1.2 Sediment Erosion Rate Flume The SERF was developed by Sheppard and his colleagues at the University of Florida to measure the erodibility of cohesive and noncohesive sediments (Trammel 2004). Figure 5 shows a schematic diagram and photograph of the SERF. The SERF has a 9-ft-long rectangular channel with dimensions of 5.08 × 20.32 cm elevated at 5.5 ft that is fed by two 500-gallons-per-minute parallel pumps from a large 1,100-gallon water tank. The flow channel is designed to have a 1-ft straightener in the beginning to reduce the turbulence of the water discharge. The specimen cross section is placed in the center of the rectangular channel at the base of the flume. Flow is driven through a 3-ft-long channel and right after reaches the specimen’s cross section. It then proceeds another 4 ft of the channel and is directed to the reservoir tank. The reason that two pumps are used is to account for harder (b) SERF Apparatus Control Room (a) Flume Wall Thickness = 0.5' Support Flume Pumps inflow Stepper Sample Figure 5. SERF apparatus at the University of Florida (a) schematic diagram and (b) photograph (Trammel 2004).

Existing Erosion Tests 23 soil samples, for which both pumps can be running. Also, erosion of the Shelby tube size sample is continuously monitored by the control computer with an video camera attached next to the test section. An array of sonic SeaTek transponders is attached at the top of the flume right above the test section; these transponders give the mean elevation of the sample surface, which is used to prompt the computer to advance the piston and keep the sample surface flush with the bottom of the flume. Basically, the SERF is controlled and monitored automatically by computer software. The summation of upward movements recorded and steps by the motor for a specific flow velocity (shear stress) divided by that particular time period reflects the erosion rate. SERF includes pressure ports at 2 ft upstream and downstream from the center of the test section. The pressure drop in the flume is calculated with the following equation: area 2 2 (8) p w h L ( ) ( ) t = D × + × where t = hydraulic shear stress (Pa), Dp = recorded pressure drop (Pa), area = cross-section area of the rectangular channel, L = distance between pressure ports (4 ft), and 2w + 2h = hydraulic radius in the channel. In addition to similar advantages mentioned for the EFA, the SERF is independent of the operator and runs with an automated system. Advantages 1. Can be used for natural and man-made samples. 2. Can give the erosion function directly. 3. Runs with an automated system. Drawbacks 1. The SERF is no longer being used; the device requires a bulky setup. 2. The automation of the process requires the use of very expensive instruments that also require significant expertise when they break down or need to be adjusted. 3. The samples have to be prepared in the cylinder of the test device. This limits the use of the SERF to disturbed or man-made samples. 2.1.3 Ex Situ Scour Testing Device The ESTD was developed by Kerenyi and his colleagues at the Federal Highway Administration (FHWA) Turner–Fairbank Highway Research Center (Shan et al. 2015). The purpose was to simulate the velocity profile for open channel conditions. Figure 6 shows a schematic diagram along with a photograph of this test device. A cylindrical soil specimen with a diameter of 63.5 mm and height of 15 mm is placed on the top of a direct force gauge. A 580-mm rectangular channel with dimensions of 12 × 2 cm connects the inlet tank to the outlet tank. As with the EFA, a flow meter is attached to the device to measure the flow velocity in the channel. A moving belt, as shown in Figure 6, is used to entrain the water and reproduce the expected log-law velocity profile in the field. The rough- ness of the channel is controlled by attaching a range of wide grit of sandpapers to the bottom of the channel surface. Instead of calculating the hydraulic shear stress at the interface of the eroding fluid and the soil indirectly from velocity or pressure drop, a direct force gauge is used

24 Relationship Between Erodibility and Properties of Soils to instantaneously capture both normal forces and shear stress induced on the soil surface. Samples for the ESTD are prepared in the lab, typically with a Pugger-Mixer, which prevents the existence of air bubbles in the specimen. Samples are left in water to slake before the ESTD test is performed. The advantages and drawbacks of the ESTD are listed below. Advantages 1. The ESTD is automated. 2. The ESTD is designed to reproduce an open channel flow condition. 3. The existence of sensors that measure the vertical force and shear stress directly is very helpful. 4. The effect of turbulence can be studied more precisely by using the results of the vertical force on the interface. Drawbacks 1. Setting up the ESTD is time consuming (up to 2 days). 2. The ESTD cannot reflect actual field conditions, as the soil specimens are all handmade in the lab; that is, an intact sample from the field cannot be tested directly in the ESTD. 2.1.4 Sediment Erosion at Depth Flume The High Shear Stress flume, or SEDflume (sediment erosion at depth flume), was originally developed at the University of Santa Barbara for the purpose of measuring sediment erosion at high shear stress and with shallow depth (McNeil et al. 1996). It has been used by researchers for coastal applications and by the U.S. Army Corps of Engineers. The primary application of this device is studying the sediment transport and suspension rate during high-stress floods. Figure 7 shows a schematic diagram and photograph of the SEDflume. The channel of the flume is 120 cm long, and the rectangular cross section is 10 × 2 cm, as shown in Figure 7. The SEDflume also exists as a portable lab that can be moved to a site. At the end of the straight channel is a 15-cm-long section where the top of the soil sample is exposed. The rectangular soil sample is 1 m long and has cross-section dimensions of 15 cm long × 10 cm wide. Test samples are either collected directly from the site or recreated in the lab by using a sediment slurry. In deeper waters, divers might be needed to place the coring tube (a) (b) Inlet Tank Outlet Tank 1 2 Clay specimen Direct force gauge Figure 6. ESTD: (a) schematic diagram and (b) photograph (Shan et al. 2015).

Existing Erosion Tests 25 and extract a sample from the site. In soft soils, the sampler can be pushed into the soil, whereas in stiffer soils, a Vibracoring head can be used. The test sample is then placed on a piston with a hydraulic jack, which is used to manually adjust the height of the sample. As with the EFA, the operator needs to keep the sample flush with the flume surface. The flow also is adjusted by a valve, and a flow meter is attached to the flume to measure the flow rate of the eroding fluid. Water that is powered by a pump flows into the flume, and testing proceeds in the following steps (McNeil et al. 1996): 1. Place one end of the rectangular coring tube on a piston located at the bottom of the test section. 2. Use the piston to extrude the sample upward until it becomes flush with the bottom of the flume surface. 3. Make sure that as the soil surface erodes, a level interface is maintained between the sample and the bottom of flume surface. 4. Record the amount of eroded sediments by recording the upward movement of piston. 5. Repeat Steps 2 to 4 for higher flow velocities and, thus, higher shear stress values. As is done for the EFA, the shear stress is calculated with Equation 7. The advantages and drawbacks of the SEDflume are as follows: Advantages 1. The field version helps minimize the sample disturbance effect. Also, with the field version, the existing water in the field can be used in the test. 2. The SEDflume directly measures the erosion rate versus shear stress curve from the field and determines the critical shear stress Drawbacks 1. The shear stress calculation is based on the mean flow velocity and the use of the Moody chart instead of on direct measurements. (a) (b) Figure 7. SEDflume: (a) schematic diagram (McNeil et al. 1996) and (b) photograph (U.S. Army Corps of Engineers).

26 Relationship Between Erodibility and Properties of Soils 2. The lab apparatus is very bulky and costly (more than $100,000). Preparation of samples is time consuming, as is the test setup in the field. 3. Only disturbed or reconstructed samples can be tested. 4. There are several limitations in sample collection, especially for offshore conditions. Roberts et al. (2003) developed a device similar to the SEDflume called the Adjustable Shear Stress Erosion and Transport (ASSET) Flume. The ASSET Flume was designed to be larger than the SEDflume to overcome a problem common to all flume tests: the effects of the channel wall on the flow. The other difference between the ASSET Flume and the SEDflume is that in the ASSET Flume, the eroded sediments are collected and then dried to obtain the bed load and suspended fractions in the flume. 2.1.5 Some Other Flume Tests In addition to the EFA and similar tests, several researchers have performed flume tests to study the erodibility potential of different soils. Gibbs (1962) conducted some large-scale flume tests in the laboratory on intact soil samples [mostly clay of low plasticity (CL) and silt (ML)] from canal banks. The purpose of his tests was to investigate the influence of different plasticity properties and in situ density of the soils on the erosion resistance of these soils. Gibbs’ findings are summarized in the next chapter of this report. A few years later, Lyle and Smerdon (1965) of TAMU constructed a laboratory flume test to study the effect of soil properties (especially compaction) on the erosion resistance of soils. They conducted erosion tests on seven Texas soils in a 22-m-long hydraulic flume with a 76- × 40-cm cross-sectional area. The slope of the flume was 0.2%. The 5.5-m-long sample soil was placed in the center cross section of the flume. The velocity of the water flow was measured by using six pitot tubes installed at six different points of the flow. The depth of flow was also measured accurately by using 17 piezometers along the flow channel. The shear stress induced on the eroded surface was calculated as the product of the hydraulic gradient and water unit weight. Lyle and Smerdon then defined the critical shear stress as the shear stress that initiates the erosion of the soil. The samples that were replicated by using the suspended samplers during each increase in the flow were tested in the lab, and the plasticity index and void ratio (compaction) were recorded. The critical shear stress was then plotted against the compaction and plasticity properties of the soil. The findings of Lyle and Smerdon (1965) are discussed further in the next chapter of the report. Kandiah and Arulanandan (1974) also used flume tests to study the erodibility of Yolo clay loam. The main purpose of their research was to compare the erodibility results obtained from the flume test and the rotating cylinder test. Also, the effect of compaction and water content on the critical shear stress was investigated. Some of Kandiah and Arulanandan’s findings are discussed in the next chapter of this report. Arulanandan and Perry (1983) studied the erodibility potential of dam core materials used for better representing the common dam filter design method that was being practiced during that time. To quantify the critical shear stress imparted on the cracks within the core of the dam, Arulanandan and Perry used both flume tests and rotating cylinder tests. Research on evaluating soil properties continued, and different researchers used different approaches to find the erodibility parameters. Shaikh et al. (1988a) performed flume tests to evaluate the influence of clay material and compaction of soil on the erosion resistance of soils. To do so, they constructed a 250–cm-long rectangular channel with a 15.5- × 11-cm cross section (see Figure 8). The slope of the flume was adjustable. As shown in Figure 8, three samples that were 15.2 cm long with a 10.5- × 2.25-cm cross section could be tested at the

Existing Erosion Tests 27 same time with the same slope. The flow depth could be adjusted between 80 cm and 210 cm. The flow velocity was also measured with a pitot tube, as shown in Figure 8. Chow (1959) calculated the shear stress induced on the surface of the sample by using the smooth channel flow equation: * 5.5 5.75 log * (9) V V V y v = + ×    where V* = shear velocity w t r     , V = velocity of flow at a depth y in the turbulent zone, rw = water density, and v = viscosity of water. Shaikh et al. (1988a) defined the erosion rate as the rate of weight removal in a given time. Then they could plot the erosion rate (N/m2/min) versus hydraulic shear stress (Pa). Six years later, Ghebreiyessus et al. (1994) developed a new enclosed flume test to study soil resistance to erosion as well as the influence of geotechnical parameters in soil erodibility. For this purpose, a 250-cm-long rectangular flume with a 20.3- × 2.5-cm cross section was constructed (Figure 9). The flume dimensions were selected to generate a steady flow condition according to Chow’s (1959) equations. The test samples were cylindrical, with a 10.2-cm inner diameter and were mounted on a mechanical piston to maintain a level interface with the bottom surface of the flume. Erosion Figure 8. Schematic diagram of flume test used by Shaikh et al. (1988b).

28 Relationship Between Erodibility and Properties of Soils rates were calculated as the rate of dried mass removal in a given time. The shear stress on the soil sample was predicted as (10) h R L w Lt = γ × × where t = hydraulic shear stress (Pa) on the soil surface; hL = head loss (m) measured by using two standpipes, one at each side of the sample; R = hydraulic radius (m) of the flume; L = length (m) of flume; and γw = unit weight (N/m3) of water. Several other attempts were also made to develop a flume erosion-testing apparatus. Some examples are the attempts made at Georgia Institute of Technology by Navarro (2004), Hobson (2008), and Wang (2013) to modify the EFA method. 2.1.6 Jet Erosion Test The jet erosion test (JET), which can be credited to Hanson (1990a, 1990b), was developed at the U.S. Department of Agriculture, Agricultural Research Service. Hanson first developed this testing device for the purpose of measuring soil erodibility in situ. The JET was standardized as ASTM D5852 in 1995 and includes both in situ and lab versions. It included a nozzle with a diameter of 13 mm, which was held 22 cm away from the center of soil surface. Figure 10 shows a schematic diagram and photograph of the in situ version of the JET apparatus. A pin profiler is used after each jet sequence to read the change in the depth of the hole made in the soil by the jet (Hanson 1990a, 1990b). The JET has been modified since its inception. Hanson and Hunt (2007) developed a new laboratory version of the JET apparatus. The circular jet submergence tank in this version has a diameter of 305 mm and a height of 305 mm. Scour readings are made by using a point gauge Figure 9. Schematic diagram of constructed enclosed flume (Ghebreiyessus et al. 1994).

Existing Erosion Tests 29 that is aligned with the orifice and measures the scour in the center of the specimen. The soil specimen is compacted in a 4-in. standard compaction mold that is centered in a submergence tank and placed below the jet nozzle. The distance between the nozzle and the soil surface in the standard compaction mold is 33 mm. The JET is currently used by some departments of transportations and engineering firms. The step-by-step procedure of a JET done in the laboratory is as follows (Hanson and Hunt 2007): 1. Compact the sample in the 4-in. standard compaction mold and trim the top surface. 2. Center the specimen in the submergence tank right below the jet orifice. Fill the tank with water. (b) (c) (a) Figure 10. JET: (a) schematic diagram of submerged JET apparatus for field testing (ASTM D5852-95) and photographs of (b) lab version of JET (Hanson and Hunt 2007) and (c) in situ version of JET (Hanson and Cook 2004).

30 Relationship Between Erodibility and Properties of Soils 3. Adjust the pressure head at the jet orifice to be 775 mm. 4. Direct the water jet at a given velocity perpendicularly to the soil surface and record the depth of the hole made by the jet as a function of time (not more than 2 hours), while holding the jet in a stationary position. The last version of the JET is a miniature of the original JET apparatus, called the mini-JET. It was first used in the field by Simon et al. (2010) at 35 sites in Oregon (Al-Madhhachi et al. 2013). Compared with the previous versions, it is easily portable and can be used both in the field and the lab on the 4-in. standard compaction mold sample. The submergence tank in the mini-JET has a dimension of 101.6 mm, and height of 70 mm. The adjustable mini-JET nozzle is 3.18 mm in diameter, and the head pressure at the nozzle is typically 450 to 610 mm. Figure 11 shows the stress distribution at the soil surface proposed by Hanson and Cook (2004). The erosion rate is calculated as the slope of the curve linking the depth of the hole to the time of jetting. The shear stress associated with the jetting process is calculated as a function of the maximum stress due to the jet velocity at the nozzle using the following equation: (11)02 2 C U J J f p i t = × r × ×    where Cf = coefficient of friction (typically 0.00416), U0 = velocity of the jet at the origin ( 2gh), r = fluid density, Ji = initial jet orifice height from soil surface, and Jp = potential core length (6.3 × nozzle diameter). The critical shear stress, tc, is defined as the stress that exists when the hole is deep enough that the jet is no longer adequate to cause additional downward erosion (Hanson and Cook 2004). Figure 11. Stress distribution at soil surface in JET (Hanson and Cook 2004).

Existing Erosion Tests 31 To describe the relationship between the JET erosion rate and the jet velocity or calculated shear stress (erosion function), Hanson and colleagues used a linear relationship and called the slope of the line the erosion coefficient, KD (Hanson 1991, 1992; Hanson and Cook 2004):  (12)z K D c( )= t − t Based on many JETs performed over time, Hanson classified the erodibility of soils accord- ing to their KD value as shown in Figure 12. Advantages 1. The JET can be run both in the field and in the lab. 2. The latest version of the JET is simple, quick, and inexpensive (around $15,000 to purchase) compared with other types of erosion test. 3. The JET can be performed on any surface—vertical, horizontal, and inclined (Hanson et al. 2002). 4. The JET is good as an index erodibility test. Drawbacks 1. Particles larger than 30 mm cannot be tested with confidence because of the small size of the sample. 2. Coarse-grained soils (i.e., noncohesive sand and gravel) tend to fall back into the open hole during the jet erosion process, thereby making the readings dubious. 3. The scale of the test application is very small. 4. The JET is typically used for man-made samples. Natural samples are more difficult to test. 5. The flow within the eroded hole and at the soil boundary is complex and difficult to analyze. sc, Pa k d cm 3 / N -s Figure 12. JET: Hanson’s classification according to erosion coefficient (Hanson and Simon 2001; Chedid et al. 2018).

32 Relationship Between Erodibility and Properties of Soils 6. The JET gives only three erodibility parameters [tc, Et, and the erosion category (EC)] out of the five possible parameters. 7. The elements of erosion are inferred rather than measured directly. 8. There are multiple interpretation techniques for predicting the critical shear stress, and these techniques give significantly different results. Prior to Hansen (1990a, 1990b), a few scholars had conducted some studies on erodibility of soils by shooting a jet into the surface of the soil. Sections 2.1.7 and 2.1.8 summarize some of that work. 2.1.7 Jet Apparatus to Measure the Tractive Resistance of Cohesive Channel Beds Dunn (1959) used a jet test to calculate the critical shear stress of cohesive channel beds. This test contained a vertical submerged impinging jet perpendicular to the soil surface. Dunn observed that the location of the maximum shear stress was the same for different pressure heads at the nozzle. To measure the induced shear stress, Dunn used a device that included a steel plate that was almost fully covered with soil particles, except for a 1-in.2 area at the location of maximum shear stress that was not covered with any soil. Dunn also measured the vane shear strength of the tested samples. Using this approach, he was able to observe the change in maximum shear stress with change in vane shear strength for each soil sample. Dunn found that the vane shear strength was proportional to the maximum shear stress at the start of the erosion process. Dunn proposed that the most important soil properties affecting resistance to erosion were the percentage of clay and silt, the soil plasticity, and the grain size distribution. A summary of his findings in correlating erodibility parameters with soil properties is presented in the Chapter 3. This jet apparatus was conceptually similar to the JET developed later by Hanson (1990a, 1990b) and, therefore, has advantages and drawbacks, similar to those listed for the JET. 2.1.8 Submerged Jet Test at the University of Texas Moore and Masch (1962) developed a submerged jet test at the University of Texas. In the proposed test, the change in the scour depth of the sample was obtained for different jet velocities, but the hydraulic shear stress at the soil surface was not calculated. Moore and Masch used a cylindrical sample with a diameter of 127 mm (5 in.) and height of 101.6 mm (4 in.). The jet velocity was kept constant for more than 1 h; meanwhile, the eroded weight of the sample was recorded every 10 min. The change in the depth of the hole was calculated by using the volume of the soil removed. The same procedure was repeated for higher velocities, and the data were compared. It was inferred that the depth of the hole in the sample can be affected by the following parameters: velocity of the submerged jet, diameter of the impinging jet, head pressure at the jet nozzle, viscosity of the eroding fluid, and the scour resistance of the sediments (Moore and Masch 1962). Figure 13 shows a schematic diagram of the vertical jet scour test. This jet apparatus was conceptually similar to the JET developed later by Hanson (1990a, 1990b) and, therefore, had advantages and drawbacks similar to those listed for the JET. Moore and Masch defined a variable called the scour rate index, Ks. This parameter is the slope in versus ave 0 2 S h t d     µ r    

Existing Erosion Tests 33 where Save = depth of scour hole for a specific jet velocity, h0 = distance from jet orifice to soil surface, t = time of scour for a specific velocity, µ = viscosity of eroding fluid, r = eroding fluid density, and d = nozzle diameter. Moore and Masch could observe that measured Ks was linearly correlated with the Reynolds number. The work of Moore and Masch was studied and used by Hanson for his JET analysis. 2.1.9 Rotating Cylinder Apparatus Developed at University of Texas In addition to trying a submerged jet test, Moore and Masch (1962) developed a new scour- testing apparatus that worked by subjecting the sample to a rotating flow around the side of the cylinder. The main purpose for implementing this device was that, by that time, other available tests for predicting the erodibility of soil [i.e., the jet apparatus developed by Dunn (1959)] could not accurately measure the hydraulic shear stress on the surface of the soil. A cylinder of cohesive soil 76.2 mm (3 in.) in diameter and 76.2 mm (3 in.) in length was used as the test specimen. The testing apparatus included a larger translucent cylinder that had the option to rotate around the vertical axis. The maximum rotation speed that the apparatus could handle was 2,500 rpm. The test specimen was then placed into the cylinder coaxially, and the residual space between the sample and translucent cylinder was filled with the eroding fluid. The fluid rotation would apply the shear stress onto the soil surface. Figure 14 shows a schematic diagram of the rotating cylinder testing device. As shown in Figure 14, the soil sample was constructed around a flexure pivot, which could help calculate the torque directly applied to the sides of the sample. The induced torque on the sample was then calculated and, thus, the shear stress could be backcalculated. One of the challenges that Moore and Masch had was to choose the right liquid as the shear transmitter in the annular space of the cylinder. They found that the use of some liquids, such as glycerin, would form a resistive layer on the soil surface. Finally, they decided to use water in the annular Figure 13. Schematic diagram of vertical jet scour test developed by Moore and Masch (1962).

34 Relationship Between Erodibility and Properties of Soils space to transmit the shear stress onto the side surface of the sample. The test procedure is as follows: 1. Place the cylindrical 76.2-mm-diameter sample in the apparatus. 2. Fill the empty space with water. 3. Increase the rotating speed until the occurrence of surface scour can be observed. 4. Record the reading of the torque. Use the calculated torque to measure the shear stress. It is worth mentioning that to calibrate the test, the operator needed to induce a known torque on a dummy sample. Then the rpm required to rotate the sample was recorded. In this way, a plot of different rpms and different torques would be obtained. Moore and Masch recommended choosing the rate of mass removal during a particular time period as the ero- sion rate. The rotating cylinder test apparatus was later used and modified by some researchers. Masch et al. (1965) worked more on the recommendations made by Moore and Masch (1962) and subsequently developed the original guideline for the rotating cylinder test. Arulanandan et al. (1973) slightly modified the previous version of the rotating cylinder and studied the effect of clay mineralogy on the erodibility of the soils by conducting some tests on Yolo loam. Kandiah and Arulanandan (1974) also used the rotating cylinder test and compared their results with the results of a flume test on the Yolo clay loam. Arulanandan et al. (1975) used the modified rotating cylinder to study the effect of pore fluid composition and also the concentra- tion of salt in the eroding fluid. The soil samples they used were remolded saturated soils. A summary of the results of Arulanandan and colleagues is presented in Section 3.1 of this report. Some of the advantages and drawbacks of the early version of the rotating cylinder test are as follows: Advantages 1. Contrary to most erosion tests, a very small amount of water is needed. 2. The shear stress can be directly estimated by using the induced torque on the side surface of the specimen. 3. The test can generate very high shear stresses. 4. The influence of the physicochemical properties of the eroding fluid (i.e., pH, salinity) on erosion rate can be easily studied. Figure 14. Schematic diagram of rotating cylinder test developed by Moore and Masch (1962).

Existing Erosion Tests 35 Drawbacks 1. Owing to the existence of the shaft within the soil sample in the apparatus, the test can only be conducted on remolded samples. 2. The samples need to be cohesive and strong enough to stand under their own weight; therefore, testing of coarse-grained soils and soft clays and silts is not possible. 3. There is no direct measurement of torque, as the induced torque is calibrated on the basis of the results from dummy samples. 2.1.10 Improved Rotating Cylinder Test Chapuis and Gatien (1986) used the same principles of the rotating cylinder apparatus developed by Moore and Masch (1962) and improved the testing technique in order to be able to test not only remolded samples but also intact samples. Before this, the rotating cylinder test could only run erodibility tests on reconstructed clays and recreated mixtures in the lab. As mentioned earlier, the clay samples were reconstructed around a center shaft that made it possible to measure the torque on the sample. In the apparatus developed by Chapuis and Gatien (1986), no flexure pivot was in the middle of the sample; therefore, the intact sample could be placed between the upper and bottom ends of the device. Figure 15 shows a photograph of the testing apparatus. The other advantage of this version of the rotating cylinder compared with earlier versions was that it could directly measure the torque through a pulley weight system. The weight system Figure 15. Photograph of improved rotating cylinder test: (1) rotating outer cylinder, (2) soil sample, (3) eroding fluid in annular space, (4) guiding shaft for installation, (5) torque measurement system, (6) head, (7) base, (8) access for cleaning, and (9) gravity drainage (Chapuis and Gatien 1986).

36 Relationship Between Erodibility and Properties of Soils included masses ranging from 0 to 40 g and had a precision of 0.1 g. The device could also pro- duce a maximum 1,750 rpm rotational speed. It was observed that the roughness of the side surface of the soil was constantly changing during a test, and therefore was affecting the shear stress measurements. The step-by-step procedure of the improved rotating cylinder test was as follows (Chapuis and Gatien 1986): 1. Place the clay sample (75 mm in diameter; height 89 mm) into the cylindrical cell of the apparatus. 2. Depending on the purpose of the test, fill the annular space around the sample with the eroding liquid (water or other chemical liquids). 3. Induce a stationary torque to the sample by using the pulley weight system. 4. On the same rpm, record the induced shear stress for stages of 10 to 30 min. 5. After each stage, collect the eroded samples, oven dry, and record the eroded mass for that particular time under that particular shear stress. Advantages 1. A small amount of water is enough as compared with most other erosion tests. 2. High shear stresses can be generated. 3. The test can be conducted on remolded and intact samples. 4. The test directly measures the torque through a pulley weight system. Drawbacks 1. The samples need to be cohesive and strong enough to stand under their own weight; therefore, testing of coarse-grained soils and soft clays and silts is not possible. 2. The test device is expensive to build (more than $30,000). 2.1.11 Rotating Erosion Test Apparatus Following a similar concept, Kerr (2001) and Sheppard et al. (2006) modified the previous versions of the rotating cylinder test. The modified version, which is called the rotating erosion testing apparatus (RETA), was constructed at University of Florida. This test was modified to be used only for stiff clays and hard rocks such as sandstone and limestone. It basically holds the same constraint that exists in previous versions of the rotating cylinder test: use of a self-support- ing sample. The RETA can be run for several days, which is required for more erosion-resistant materials such as rock. The samples tested with RETA can be both 61 mm (2.4 in.) and 101.6 mm (4 in.) with a height of 101.6 mm (4 in.). The test apparatus is equipped with a torque transducer at its base and a load cell to record the weight of the sample. The apparatus is also equipped with a water cooling system to reduce the temperature for long tests (more than 72 h for rocks). The central shaft still exists; therefore, intact samples are not usable unless a center hole can be drilled through them. After the hole is drilled, the sample is oven dried and placed in the device to saturate. During the saturation, the device applies a very small torque to the sample for at least a day to remove any loose material. Sheppard et al. (2006) believed that the results of shear stress would be unexpectedly large if the loose material was not removed before the test was performed. After the sample is saturated, it is inserted into a sleeve and placed in the RETA cylinder for the test. Figure 16 shows a schematic diagram and photograph of the testing machine. Advantages 1. Can be run for very long time (more than 72 h for rocks). 2. Can generate very high shear stress values.

Existing Erosion Tests 37 3. The apparatus is equipped with a torque transducer. 4. The flow temperature can be controlled during the test. Drawbacks 1. The apparatus is modified to be used only for stiff clay and hard rock such as limestone. 2. Intact samples are not usable unless a center hole can be drilled through them. 3. As with earlier versions, samples need to be self-supporting. 4. The test device is bulky and expensive to build (more than $30,000). 2.1.12 Pinhole Erosion Test Sherard et al. (1976a) developed a laboratory test to qualitatively measure the erodibility of fine-grained soils. In this test, distilled water is passed through a drilled hole under a pressure head of 51 mm in the center of the sample, and the erosion resistance of the soil is observed. The hole punched in the center of the sample has a diameter of 1 mm. The test was particu- larly designed to study and simulate the leakage effect in both dispersive and nondispersive fine-grained soils, which was the case in most earth embankments. The pinhole test was later standardized as ASTM D4647. The test consists of compacting a 38-mm-long soil sample in a plastic cylinder with an inside diameter of 33 mm. A truncated jet nozzle with a diameter of 1.5 mm directs the water through the 1-mm diameter punched hole in the center of the cylindrical specimen. Figure 17 depicts a schematic diagram of the test apparatus. In observing the appearance of the flowing water and final size of the hole in the tested sample, Sherard et al. (1976) noticed considerable differences in the behavior of dispersive and non- dispersive clays subjected to water flow. With some limitations, the test could be done on intact field samples. (a) (b) Figure 16. RETA: (a) schematic diagram and (b) photograph at the University of Florida (Bloomquist et al. 2012).

38 Relationship Between Erodibility and Properties of Soils The procedure for the pinhole erosion test is as follows: 1. Create a 38-mm-long sample by compacting the soil in the test cylinder above the coarse sand space, which is covered by a wire screen (Figure 17). 2. Make sure that the soil specimen is representative of the field conditions in terms of moisture content and dry unit weight. 3. Push the truncated cone jet nozzle into the center of the cylindrical sample. 4. Using the test wire punch, punch a 1-mm-diameter hole in the center of the sample. 5. Remove the wire punch, place a wire screen on the top of sample, and fill the remaining space with coarse sand. 6. Start the test by shooting the jet into the hole with a pressure head of 51 mm. 7. Continue the test up to 5 min. Depending on the cloudiness of the effluent and the measured flow rate, decide to continue the test or not for higher pressure heads, as described in the ASTM standard. Table 2 is used to interpret the results of the pinhole erosion test. On the basis of the criteria defined in Table 2, the erosion resistance of soil is classified as one of the nine defined categories. Advantages 1. Simulates the leakage in dispersive and nondispersive fine-grained soils. 2. Standardized as ASTM. 3. Directly measures the dispersibility of clay soils. 4. Costs relatively less to build the device as compared with other erosion tests (less than $10,000). Drawbacks 1. Qualitative results. 2. Mostly applicable for remolded samples; however, can be run on intact samples with some limitations. 3. Not applicable for coarse-grained soils (i.e., coarse sand and gravel). Figure 17. Schematic diagram of pinhole erosion test apparatus (ASTM D4647).

Existing Erosion Tests 39 2.1.13 Drill Hole Test Lefebvre et al. (1985) developed a new technique for predicting the internal erosion resis- tance of natural clays. The concern was the erodibility of the natural clays in the eastern part of Canada. The test was inspired by the earlier version of the pinhole erosion test. The testing apparatus uses a 10-cm-long cylindrical sample with a diameter of 35.5 mm. At the center of the sample, a 6.35-mm hole is drilled. Schematic diagrams of the test are shown in Figure 18. The test is conducted by circulating water through the bored hole into the sample. The pressure drop through the sample is measured with a differential manometer connected to both sides of the sample. A tank is used to produce a 143-cm pressure head and thereafter direct the flow through the sample. The flow is adjusted with a valve and measured by a flow meter. The adjusting valve is used to increase the flow velocity 0.5 m/s every 15 min. The deposited sediment in the reservoir tank is dried and weighed to measure the mass removal rate for that particular velocity. The average diameter of the hole is recorded after each step. The shear stress is calculated according to the pressure drop measured by the manometers. The results are reported as removed mass versus velocity or shear stress. Lefebvre et al. (1985) also observed that the change in roughness of the hole during the test can be interpreted by using the Moody diagram (Moody 1944) when the friction factor and the Reynolds number are known. Erosion at the clay particle level is accompanied by an increase in the smoothness of the hole (decrease in relative roughness), while erosion of lumps of clay particles leads to an increase in hole roughness. Advantages 1. Has a larger hole diameter than the pinhole erosion test (6.35 mm vs. 1 mm), which minimizes the soil disturbance. 2. Yields quantitative erodibility results. 3. Costs relatively less to build the device as compared with other erosion test devices (less than $15,000). Drawbacks 1. Shear stress is indirectly measured by using Moody charts, which might not be accurate. 2. The test is mostly applicable for remolded samples; however, it can be run on intact samples with some limitations. 3. The test is not applicable for coarse-grained soils (i.e., coarse sand and gravel). Table 2. Interpretation of the results of the pinhole erosion test (ASTM D4647).

40 Relationship Between Erodibility and Properties of Soils (a) (b) Figure 18. Schematic diagrams of (a) entire drill hole test assembly and (b) sample setup of drill hole test (Lefebvre et al. 1985).

Existing Erosion Tests 41 2.1.14 Hole Erosion Test The hole erosion test (HET) is a laboratory erosion test that evolved from the older pinhole erosion test and can be credited to Robin Fell in Australia (Wan and Fell 2002; Wahl et al. 2009a, 2009b; Benahmed and Bonelli 2012). The test consists of drilling a hole 6 mm in diameter through a soil sample and forcing water to flow through the hole at a chosen velocity while recording the rate of mass removal per unit area as a function of time to obtain an erosion rate (kg/s/m2) (Figure 19). The soil is compacted in a standard compaction mold 100 mm (4 in.) in diameter. As in the drill hole test, the sample is connected to a tank that can maintain a variable head ranging from 50 to 800 mm. The flow is also controlled through a valve. The cost of building the device ranges from $20,000 to $35,000. The rate of mass removal per unit area is calculated as 2 d d d t r × φ where dφ = change in diameter of the hole, rd = dry unit weight of the sample, and dt = change in time. Because the diameter of the hole cannot be monitored during the test, this value is indirectly predicted by using the measured flow rate, the hydraulic gradient, and Equations 13 and 14 (in SI units). Equation 13 refers to laminar flow conditions, and Equation 14 refers to turbulent flow. Turbulent flow is associated with a Reynold’s number higher than 5,000. 16 (13) Laminar, 1 3Q f gs t t t w t φ = × × πr     64 (14) 2 Turbulant, 2 1 5Q f gs t t t w t φ = × × π r     where Qt = flow rate at time t, fLaminar,t and fTurbulant,t = estimated friction factors at time t, rw = water unit weight, g = ground gravity acceleration, and st = hydraulic gradient obtained from manometers at both ends of sample. In these equations, the friction factors are estimated by using the recorded hole diameter before and after the test. The test results link the rate of mass removal per unit area to the net shear stress above critical; the shear stress on the wall of the hole is estimated with Equation 15. This equation is obtained after consideration of force equilibrium on the body of the eroding fluid along the preformed hole at a particular time, t. 4 (15)g sw tt = r × × × φ

42 Relationship Between Erodibility and Properties of Soils (a) (b) Figure 19. HET: (a) schematic diagram and (b) photograph of sample setup.

Existing Erosion Tests 43 where rw = density of water, st = hydraulic gradient across the hole, and φ = diameter at time t. The equation used for the erosion function is linear:  (16)m Ce c( )= t − t where Ce (s/m) is the erosion coefficient and m . is the rate of mass removal. The erosion rate index is then defined as log (17)I CHET e( )= − Wan and Fell (2002) went on to propose some erosion categories based on IHET (Table 3). Advantages 1. There is direct similitude with piping erosion in earth dams. 2. A wide range of pressure heads, and therefore a wide range of hydraulic shear stress at the soil–water interface, can be applied. Drawbacks 1. The sample needs to be cohesive and strong enough to stand under its own weight. Therefore, the test cannot be run on cohesionless samples. 2. The test is very difficult to run on intact samples in Shelby tubes from the field. It is better for remolded samples in the lab. 3. Preparation of the test is difficult and time consuming (up to days). 4. The erosion process cannot be directly monitored. The erosion rate needs to be extrapolated and inferred. 5. The hydraulic shear stress is inferred rather than directly measured. 6. The data reduction process is subjective. 7. The flow within the eroded hole and at the soil boundary is complex and difficult to analyze. 2.1.15 Slot Erosion Test The slot erosion test (SET) was also developed by Wan and Fell (2002) in Australia. The concept of the SET is very similar to that of the HET, except that the sample in the SET is different. The test (Figure 20) consists of drilling a 2.2-mm-wide, 10-mm-deep slot at the surface of a 1-m-long rectangular soil sample and having water within the slot at a chosen Group No. Erosion Rate Index, IHET Description 1 <2 Extremely rapid 2 2–3 Very rapid 3 3–4 Moderately rapid 4 4–5 Moderately slow 5 5–6 Very slow 6 >6 Extremely slow Table 3. HET: Fell’s classification according to the erosion index (Wan and Fell 2002).

44 Relationship Between Erodibility and Properties of Soils (a) (b) PUMP ERODING FLUID SUPPLY TANK WIRE MESH 20mm GRAVELS FLOW METER PRESSURE GAUGE 50mm DIA. PIPE 25 00 m m 1000mm 30 0m m 15 0m m 100mm 8mm THICK ALUMINIUM COVER PLATE FRAME MADE OF 100×50×6mm ALUMINIUM CHANNELS 16mm THICK PERSPEX COVER PLATE PRE-FORMED SLOT 10mm × 2.2mm PRE-FORMED SLOT 10mm×2.2mm SOIL SPECIMEN PRESSURE GAUGE TO DRAIN SECTION X-X X X Figure 20. SET: (a) schematic diagram and (b) photograph of sample setup.

Existing Erosion Tests 45 velocity while the rate of mass removal per unit area as a function of time is recorded to obtain an erosion rate (kg/s/m2). The cost of building the device ranges from $20,000 to $35,000. As with the HET, Equation 15 is used to calculate the shear stress. In the SET, the hydraulic diameter slotA Pw     is used instead of 4 φ in Equation 15. Aslot refers to the cross-sectional area of the preformed slot, and Pw is the wetted perimeter. ISET is also calculated with the same procedure used to calculate IHET (Equation 17). All the described tests that study internal erosion require that the soil be self-supporting (cohesive). Cohesionless soils cannot preserve an open hole or slot. Therefore, in cases in which noncohesive soils form a high-stress portion of the embankment, the HET, SET, drill hole, and pinhole tests cannot appropriately simulate actual field conditions. The advantages and drawbacks of the SET are same as those identified for the HET. For that reason, new internal erosion test devices are being developed. Some of these devices are described below. 2.1.16 Stress-Controlled Erosion Apparatus Chang and Zhang (2011) developed a new test for studying internal erosion in soils at Hong Kong University. They ran some tests on a man-made cohesionless soil. A schematic diagram and photograph of this test are shown in Figure 21. The test consists of a triaxial system that is fed by a water supply system and controlled by a computer. The porous stone used in this apparatus is modified to accommodate the high permeability of the soil tested in this experiment. The soil sample is 10 cm in diameter and 10 cm high and is mounted on a hollow base with a 10-mm-thick, 95-mm-diameter perforated plate. Water flow seeps through the hollow base and the perforated plate and the soil sample. Before the internal erosion testing is begun, a 10-kPa confining pressure is applied to the sample. Then, de-aired water is injected slowly into the specimen from the bottom base to saturate the sample. During the erosion test, the vertical deformation of the sample is measured with a linear variable differential transformer (LVDT), and the radial deformations can be measured with a video camera. The test is controlled by adjusting the hydraulic gradient of the seepage water through the sample. A soil collection system is placed at the bottom of the triaxial system. Each hydraulic gradient is maintained for a 10-min period, and the eroded mass of soil is collected, dried, and weighted. Advantages 1. Has direct similitude with internal erosion in earth dams. 2. Works properly for cohesionless samples. 3. Uses advanced measurement techniques for axial and radial deformations. 4. Can control multiple criteria (i.e., confining pressure, hydraulic gradient) during the test. Drawbacks 1. Requires costly setup (more than $30,000 to build). 2. May not be efficient for very low-permeability clays and rocks. 3. Hard to test on intact samples, owing to the size of the sample.

46 Relationship Between Erodibility and Properties of Soils (b) (a) Figure 21. Stress-controlled erosion apparatus: (a) schematic diagram and (b) photograph (Chang and Zhang 2011).

Existing Erosion Tests 47 2.1.17 True Triaxial Piping Test Apparatus Richards and Reddy (2010) developed a true triaxial piping test apparatus (TTPTA) at the University of Illinois at Chicago to study the internal erosion in both cohesive and cohesion-less soils. The test consists of applying a wide range of confining pressures, with measurements of pore pressure and hydraulic gradient in a true triaxial cell (Figure 22). The results of this test give the critical hydraulic gradient as well as the critical velocity. Advantages 1. Has direct similitude with internal erosion. 2. Works properly for both cohesionless and cohesive samples. 3. Uses advanced measurement techniques for deformations. 4. Can control multiple criteria (i.e., confining pressure, hydraulic gradient) during the test. Drawbacks 1. Relatively complicated and costly setup (more than $30,000 to build). 2. May not be efficient for very low-permeability clays and rocks. 3. Hard to test on intact samples because of the size of the sample. 2.1.18 Constant Gradient Piping Test Apparatus Fleshman and Rice (2013) developed a new test apparatus to evaluate the hydraulic conditions required for starting piping erosion. The testing apparatus is shown in Figure 23. The sample is held in a sample holder, while a constant hydraulic gradient is imposed throughout the sample. During the test, the differential head is increased and the pore pressure and the soil behavior are monitored. Fleshman and Rice (2013) used this testing device on various sandy soils with different grain size distribution, specific gravity, and gradation, and recorded the critical hydraulic condition (i.e., critical gradient) in which the piping erosion was initiated. It was observed that the initiation of the erosion occurs in four stages: (1) the first observable movement of particles, (2) progression of heave, (3) boil formation, and (4) final or total heave. Advantages 1. Has direct similitude with piping erosion. 2. Includes automated saturation instrumentation and advanced measurement of flow and head. Drawbacks 1. Used only to evaluate the initiation of piping rather than to give the entire erosion function. 2. No measurement of soil axial and radial deformations. 3. Efficient only for sandy soils. 4. Sample preparation is relatively difficult. The test is not designed for intact samples. 2.2 Field Erosion Testing 2.2.1 Pocket Erodometer Test The pocket erodometer test (PET) was developed by Briaud et al. (2012) at TAMU. The pocket erodometer is a regulated mini jet impulse–generating device in the form of a plastic water gun. The jet is aimed horizontally at the vertical face of the sample. The jet velocity is calibrated to be always 8 m/s, and the nozzle is kept 50 mm from the soil surface. The depth of the hole in the surface of the sample created by 20 impulses of water is recorded. The eroded

48 Relationship Between Erodibility and Properties of Soils (a) (b) Figure 22. TTPTA: (a) schematic diagram and (b) photograph (Richards and Reddy 2010).

Existing Erosion Tests 49 depth is compared with an erosion chart to determine the erosion category of the soil, which helps the geotechnical engineer with preliminary design of erosion projects. Many different options were considered during the development of the pocket erodometer, including the most appropriate device, velocity range, direction of application, distance from the face of the sample, and repeatability from one person to another. Figure 24 shows the schematic diagram of the PET, along with a photograph from the test (Briaud et al. 2012). The original device selected for the pocket erodometer is 105 × 77 × 18 mm. The diameter of the nozzle is about 0.5 mm. The jet velocity of 8 m/s was chosen because it eroded most tested specimens. The pocket erodometer needs to be calibrated before the test to reproduce the velocity of 8 m/s at the nozzle. The following equation is used to calibrate the velocity of the impinging jet at the nozzle: 2 (18)0.5v x H g =     where x and H are as shown in the schematic diagram of the PET depicted in Figure 24 and v is the initial horizontal velocity right at the nozzle. Figure 23. Schematic diagram of the constant gradient piping test apparatus (Fleshman and Rice 2013).

50 Relationship Between Erodibility and Properties of Soils The height of the erodometer (shown as H in Figure 24) must be kept constant during the calibration process. Also, external forces such as wind should be avoided. The PET can be done with any type of apparatus that can meet the requirements of this test and reproduce 8 m/s velocity at the nozzle with ±0.5 m/s with an impulse time period of near 0.15 s. Briaud et al. (2012) con- ducted the PET on many samples from different levees and compared the results with EFA results that had been obtained. The comparisons resulted in an erosion category chart based on the PET depth ranges (Figure 25). The step-by-step process of the PET is as follows (Briaud et al. 2012): 1. Place the sample horizontally, either by laying it on a flat surface or by holding it in your hand. (Note: The test must not be run with the jet pointed vertically.) 2. Smooth the surface to remove any uneven soil. Beginning with a smooth and vertical surface makes it easier to measure the erosion depth. 3. Point the pocket erodometer at the smooth end of the sample, 50 mm away from the face. 4. Keeping the jet of water from the PET aimed horizontally at a constant location, squeeze the trigger 20 times at a rate of 1 squeeze per second, forming an indentation in the surface of the sample. Each squeeze should fully compress the trigger, and then the trigger should be fully released before it is compressed again. (a) (b) Figure 24. PET: (a) schematic diagram and (b) photograph (Briaud et al. 2012). 100,000 10,000 1,000 100 10 1 0.1 0.1 1.0 10.0 100.0 Erosion Rate (mm/hr) Velocity (m/s) Figure 25. Erosion depth ranges of PET depicted on erosion categories proposed by EFA.

Existing Erosion Tests 51 5. Using the end of a digital caliper or an appropriate measuring tool, measure the depth of the hole created. 6. The test should be repeated at least three times in different locations across the face of the sample, and an average should be used to ensure a good estimate. 7. Use Figure 25 to determine the erosion category. Advantages 1. The cost is very low ($25 per test). 2. The device is very handy and simple to operate, both in the field and in the laboratory. 3. The test gives a quick and crude estimate of soil erodibility. Drawbacks 1. The PET is on a very small scale. 2. The test gives only the erosion category, with no measurement of critical shear stress or critical velocity. 3. The test is useful only for preliminary field evaluation and not good for design purposes. 2.2.2 In Situ Erosion Evaluation Probe Gabr et al. (2013) developed an erosion testing device called the in situ erosion evaluation probe (ISEEP) at North Carolina State University. The test is conducted by advancing a vertical jet probe into the subsurface soil and measuring the rate of advancement. As discussed earlier, all other in situ tests were limited to evaluating the scour potential of the soil solely on the ground surface (e.g., JET, SEDflume), and the EFA was the only test that could evaluate the erodibility of the natural soil associated with a particular depth. The flow velocities of the ISEEP are normally much less than the imparted velocities in the EFA. ISEEP can investi- gate the erodibility of any soil at any depth, provided the probe can penetrate by erosion. The results of ISEEP are reported on the basis of the concept of stream power, which was first presented by Annandale and Parkhill (1995), who believed that this concept would better represent the erodibility potential of an eroding fluid as compared with velocity or shear stress. Annandale (2006) used Equations 19 and 20 to define stream power, P. (19)P qHw= γ (20)0P U= t where P = stream power (watts per unit area), γw = water unit weight, q = flow discharge in unit area, H = energy head, t = shear stress, and U0 = velocity. Equation 20 shows that the stream power is a function of both the shear stress and the velocity. Figure 26 shows a photograph of the ISEEP device, which was tested at the North Carolina State University lab before being used in the field. The jet nozzle is a truncated cone probe that is guided into the soil as the soil is being eroded by the impinging jet. The eroded material moves up through the annulus space between the probe and the wall of the hole created by erosion. The water velocity at the nozzle is controlled digitally by a variable speed pump on the ground surface. The rate of advancement into the soil

52 Relationship Between Erodibility and Properties of Soils subsurface is measured and represents the erosion rate. The body of the probe is divided into sections so that the length can be adjusted for deep locations in the field. The orifice of the jet nozzle is 19 mm (0.75 in.) long, and the nozzle velocity can go up to 12 m/s. During penetration, the advancement is recorded with a video camera. The results of this test are reported as the rate of advancement (penetration) versus the stream power value, P. The bed shear stress is obtained from Equation 21 on the basis of Julien’s (1995) study: (21)2C Uwt = r where t = bed shear stress, U = jet velocity, rw = water density, and C = diffusion coefficient that varies depending on flow condition. Some advantages and drawbacks of the ISEEP are as follows: Advantages 1. The ISEEP can evaluate the erodibility of any soil at any depth with a wide range of jet velocity. 2. There is no need for sample extraction and procurement. Drawbacks 1. The penetration may be limited if the probe fails to erode the soil. The ISEEP is better for use in sandy soils. (a) (b) Figure 26. ISEEP: (a) truncated cone probe and (b) apparatus prototype at North Carolina State University (Gabr et al. 2013).

Existing Erosion Tests 53 2. The use of the stream power makes it difficult to compare this device with other erosion devices. 3. The test device is fairly expensive to build (more than $20,000), and difficult to interpret. 2.2.3 Borehole Erosion Test The borehole erosion test (BET) is an in situ test developed by Briaud at TAMU (Briaud et al. 2017a). The purpose of this test is to quantify the erodibility of the soil layers as a function depth, as follows: 1. Drill a hole into the ground, say 100 mm in diameter, 10 m deep. 2. Remove the drilling rods and measure the initial diameter of the borehole with a borehole caliper. 3. Reinsert the rods to the bottom of the hole and circulate water down the rod and up the outside annulus of the hole for a given time, say 15 minutes. 4. Remove the rods and measure the diameter of the hole with the borehole caliper. 5. Calculate the erosion rate. The increase in diameter of the borehole at a certain depth given by the calipers divided by the flow time is the erosion rate of the soil at that depth for the flow velocity applied during the test. Profiles of erosion rate for different velocities can be prepared in this fashion. Figure 27 shows a schematic diagram of the BET and field work photographs. The advantages and drawbacks of the BET areas follows: Advantages 1. Only typically available field equipment (i.e., common drilling rig for wet rotary boring, flow meter in line with the drilling rig pump, and borehole caliper) is used to perform a BET; therefore, this test can be performed in many drilling projects. (b) (a) DIAMETER, D WATER FLOW VELOCITY, V WATER FLOW Figure 27. BET: (a) schematic diagram and (b) photographs of the test at the Riverside campus at TAMU (Briaud et al. 2017).

54 Relationship Between Erodibility and Properties of Soils 2. Each test gives the erosion function for all layers traversed, since a complete borehole diameter profile is obtained from the caliper. Many tests on many samples would be required if laboratory tests were to be conducted. 3. The BET has two component tests: the lateral erosion test associated with the increase in diameter of the borehole and the bottom erosion test associated with the increase in depth below the bottom of the drilling rods during the flow. The latter is much like an in situ JET. 4. The BET can be used in any soil or rock in which a hole can be drilled. Drawbacks 1. The shear stress is obtained from the Moody chart. 2. The test is limited by the pump flow available on the drill rig. 3. The hole needs to be capable of staying open. 4. In sand boreholes, the addition of bentonite during drilling needs to be controlled, so as not to affect the erosion resistance. 5. Because of gravity, the velocity over a given distance changes and needs to be taken into account. 6. Erosion of layers above the tip of the pipe could be altered by the sediment transport from the lower layers. For example, Sheppard (2002) showed that equilibrium scour depths are reduced by the presence of suspended fine sediments. 7. Borehole measurements with the caliper may not be accurate as a result of material slough- ing off. 2.2.4 In Situ Scour Testing Device The in situ scour testing device (ISTD) is the most recent field erosion test device that is cur- rently under development by FHWA. Zinner et al. (2016) presented the concept behind the ISTD and its application in pier scour studies. This device has a cylindrical shape and can be used in a boring test rig and fit into the steel casing of a hollow stem auger to evaluate the erodibility of soil at any depth. The ISTD generates a horizontal flow at the bottom of the borehole. Figure 28 shows a diagram of the cylindrical ISTD concept. So far, the ISTD is applicable only for soils that are below the groundwater table and that have a maximum N-value of 30. Also, the ISTD is generally limited to upper soil layers, as the mechanical parts are limited. Development of the ISTD by FHWA is on-going. Figure 28. Schematic diagram of the cylindrical ISTD concept (Zinner et al. 2016).

Existing Erosion Tests 55 2.3 Summary A summary of all the erosion tests reviewed in this chapter is presented in Table 4 in terms of their application in the field or the lab. Table 5 shows some of the most common and important erosion tests with information regarding their ability to measure shear stress, the soil type that can be tested, and the cost associated with them. Erosion Test Range of Soil Types That Can Be Tested Range of Shear Stress (Pa) That Can Be Applied Cost of Devicea Reliability of Results Lab JET Sands to clays <100 Low Good In situ JET Sands to clays <500 Medium Good EFA Sands to clays <165 High Good HET Clayey soils ≤800 High Good SET Clayey soils ≤400 High Medium RETA Clayey soils <100 High Medium PET Sands to clays <20 Very Low Medium ISEEP Sands to clays <650 High Good BET Sands to clays <600 Medium Good aVery low: <$5,000; low: $5,000 to $15,000; medium: $15,000 to $30,000; high: ≥$30,000. Table 5. Some erosion tests with information about their application. Laboratory Erosion Tests In Situ Erosion Tests • Lab jet erosion test (JET) • Hole erosion test (HET) • Pinhole erosion test • Drill hole erosion test • Slot erosion test (SET) • Rotating cylinder test (RCT) and improved versions • Rotating erosion testing apparatus (RETA) • EFA and similar versions of it (e.g., SEDflume, SERF, ESTD) • Stress-controlled erosion apparatus • True triaxial piping test apparatus (TTPTA) • Constant gradient piping test apparatus • Field jet erosion test (JET) • In situ erosion evaluation probe (ISEEP) from North Carolina State University • Borehole erosion test (BET) • Pocket erodometer test (PET) • Adjustable Shear Stress Erosion and Transport (ASSET) flume • In situ scour testing device (ISTD) • Field flume tests Table 4. Summary of all types of erosion tests in terms of their application.