National Academies Press: OpenBook

Pier and Contraction Scour in Cohesive Soils (2004)

Chapter: Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water

« Previous: Chapter 3 - Erosion Function Apparatus (EFA)
Page 17
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 17
Page 18
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 18
Page 19
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 19
Page 20
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 20
Page 21
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 21
Page 22
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 22
Page 23
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 23
Page 24
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 24
Page 25
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 25
Page 26
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 26
Page 27
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 27
Page 28
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 28
Page 29
Suggested Citation:"Chapter 4 - The SRICOS-EFA Method for Cylindrical Piers in Deep Water." National Academies of Sciences, Engineering, and Medicine. 2004. Pier and Contraction Scour in Cohesive Soils. Washington, DC: The National Academies Press. doi: 10.17226/13774.
×
Page 29

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.

17 CHAPTER 4 THE SRICOS-EFA METHOD FOR CYLINDRICAL PIERS IN DEEP WATER 4.1 SRICOS-EFA METHOD FOR CONSTANT VELOCITY AND UNIFORM SOIL Because cohesive soils may scour so much more slowly than cohesionless soils, it is necessary to include the scour rate in the calculations, and the SRICOS Method was developed for this purpose. The SRICOS Method was proposed in 1999 to predict the scour depth z– versus time t curve at a cylindrical bridge pier for a constant velocity flow, uniform soil, and water depth greater than two times the pier diameter. The SRICOS Method consists of the following (Briaud et al., 1999): 1. Collecting Shelby tube samples near the bridge pier, 2. Testing them in the EFA (Erosion Function Apparatus, Briaud et al., 2002) (Figure 3.1) to obtain the erosion rate z˙– (mm/hr) versus hydraulic shear stress τ (N/m2) curve, 3. Calculating the maximum hydraulic shear stress τmax around the pier before scour starts, 4. Reading the initial erosion rate z˙–i (mm/hr) correspond- ing to τmax on the z˙– versus τ curve, 5. Calculating the maximum depth of scour z–max, 6. Constructing the scour depth z– versus time t curve using a hyperbolic model, and 7. Reading the scour depth corresponding to the duration of the flood on the z– versus t curve. The maximum hydraulic shear stress τmax exerted by the water on the riverbed was obtained by performing a series of three-dimensional numerical simulations of water flowing past a cylindrical pier of diameter B on a flat river bottom and with a large water depth (water depth larger than 2B). The results of several runs lead to the following equation (Briaud et al., 1999): Where ρ is the density of water (kg/m3); v is the depth aver- age velocity in the river at the location of the pier if the bridge were not there (it is obtained by performing a hydrologic analysis with a computer program such as the Hydrologic Engineering Center—River Analysis System [HEC-RAS], 1997); and Re is where B is the pier diameter and υ the vB υ τ ρmax . log ( . )= −  0 094 1 110 4 12v Re kinematic viscosity of water (10−6m2/s at 20°C). The initial rate of scour z˙–i is read on the z˙– versus τ curve from the EFA test at the value of τmax. The maximum depth of scour z– max was obtained by per- forming a series of 43 model scale flume tests (36 tests on three different clays and 7 tests in sand) (Briaud et al., 1999). The results of these experiments, and a review of other work, led to the following equation, which appears to be equally valid for clays and sands: In Equation 4.2, Re has the same definition as in Equation 4.1. The regression coefficient for Equation 4.2 was 0.74. The equation that describes the shape of the scour depth z– versus time t curve is Where z˙–i and z–max have been previously defined, t is time (hours). This hyperbolic equation was chosen because it fits the curves obtained in the flume tests well. Once the duration t of the flood to be simulated is known, the corresponding z– value is calculated using Equation 4.3. If z˙–i is large, as it is in clean fine sands, then z– is close to z–max even for small t values. But if z˙–i is small, as it can be in clays, then z– may only be a small fraction of z–max. An example of the SRICOS Method is shown in Figure 4.1. The method as described in the previous paragraphs is lim- ited to a constant velocity hydrograph (v = constant), a uni- form soil (one z˙– versus τ curve) and a relatively deep water depth. In reality, rivers create varying velocity hydrographs and soils are layered. The following describes how SRICOS was extended to include these two features. The case of shal- low water flow (water depth over pier diameter < 2), non- circular piers, and flow directions different from the pier main axis are not addressed in this chapter. 4.2 SMALL FLOOD FOLLOWED BY BIG FLOOD For a river, the velocity versus time history over many years is very different from a constant velocity history. In z t z t zi = + 1 4 3 ˙ ( . ) max zmax . . ( . )mm Re( ) = 0 18 4 20 635

order to investigate the influence of the difference between the two velocity histories or hydrographs on the depth of scour at a bridge pier, the case of a sequence of two different yet constant velocity floods scouring a uniform soil was first considered (Figure 4.2). Flood 1 has a velocity v1 and lasts a time t1 while the subsequent Flood 2 has a larger velocity v2 and lasts a time t2. The scour depth z– versus time t curve for Flood 1 is described by: For Flood 2, the z– versus t curve is: After a time t1, Flood 1 creates a scour depth z–1 given by Equation 4.4 (Point A on Figure 4.2b). This depth z–1 would have been created in a shorter time t* by Flood 2 because v2 is larger than v1 (Point B on Figure 4.2c). This time t* can be found by setting Equation 4.4 with z– = z–1 and t = t1 equal to Equation 4.6 with z– = z–1 and t = t*. When Flood 2 starts, even though the scour depth z–1 was due to Flood 1 over a time t1, the situation is identical to having had Flood 2 for a time t*. Therefore, when Flood 2 starts, the scour t t z z t z z z i i i * ˙ ˙ ˙ ( . ) max max = + −  1 2 1 1 2 1 2 1 1 4 6 z t z t zi = + 1 4 5 2 2˙ ( . ) max z t z t zi = + 1 4 4 1 1˙ ( . ) max 18 depth versus time curve proceeds from Point B on Figure 4.2c until Point C after a time t2. The z– versus t curve for the sequence of Floods 1 and 2, follows path OA on the curve for Flood 1 then switches to BC on the curve for Flood 2. This is shown as curve OAC on Figure 4.2d. A set of two experiments was conducted to investigate this reasoning. For these experiments, a pipe with a diame- ter of 25 mm was placed in the middle of a flume. The pipe was pushed through a deposit of clay 150 mm thick that was made by placing prepared blocks of clay side by side in a tight arrangement. The properties of the clay are listed in Table 4.1. The water depth was 400 mm and the mean flow velocity was v1 = 0.3 m/s in Flood 1 and v2 = 0.4 m/s for Flood 2 (Fig- ure 4.3a). The first of the two experiments consisted of set- ting the velocity equal to v2 for 100 hours and recording the z– versus t curve (Figure 4.3c). The second of the two experi- ments consisted of setting the velocity equal to v1 for 115 hours (Figure 4.3b) and then switching to v2 for 100 hours (Fig- ure 4.3d). Also shown on Figure 4.3d is the prediction of the portion of the z– versus t curve under the velocity v2 accord- ing to the procedure described in Figure 4.3. As can be seen, the prediction is very reasonable. 4.3 BIG FLOOD FOLLOWED BY SMALL FLOOD AND GENERAL CASE Flood 1 has a velocity v1 and lasts t1 (Figure 4.4a). It is fol- lowed by Flood 2, which has a velocity v2 smaller than v1 and lasts t2. The scour depth z– versus time t curve is given by Equation 4.4 for Flood 1 and by Equation 4.5 for Flood 2. After a time t1, Flood 1 creates a scour depth z–1. This depth z–1 is compared with z–max 2; if z–1 is larger than z–max 2 then when Flood 2 starts, the scour hole is already larger than it can be Figure 4.1. Example of the SRICOS Method for constant velocity and uniform soil.

19 the program is the depth of scour versus time curve over the duration of the velocity versus time history. 4.4 HARD SOIL LAYER OVER SOFT SOIL LAYER The original SRICOS Method (Briaud et al., 1999) was developed for a uniform soil. In order to investigate the influ- ence of the difference between a uniform soil and a more real- istic layered soil on the depth of scour at a bridge pier, the case of a two-layer soil profile scoured by a constant velocity flood was considered (Figure 4.5). Layer 1 is hard and ∆z–1 thick, Layer 2 underlies Layer 1 and is softer than Layer 1. The scour depth z– versus time t curve for Layer 1 is given by Equation 4.4 (Figure 4.5a) and the z– versus t curve for Layer 2 is given by Equation 4.5 (Figure 4.5b). If ∆z–1 is larger than the maxi- mum depth of scour in Layer 1, z–max 1, then the scour process Figure 4.2. Scour due to a sequence of two flood events (small flood followed by big flood). Liquid Limit, % 34.4 Shear Strength, kPa(lab vane) 12.5 Plastic Limit, % 20.2 Cation Exchange Capacity, (meq/100g) 8.30 Plasticity Index, % 14.1 Sodium Adsorption Ratio 5.00 Water Content, % 28.5 Electrical Conductivity, (mmhos/cm) 1.20 Mean Diameter 50D , (mm) 0.0062 pH 6.00 Sand Content, % 0.0 Unit Weight (kN/m3) 18.0 Silt Content, % 75.0 Specific Gravity 2.61 Clay Content, % 25.0 TABLE 4.1 Properties of the porcelain clay for the flume experiment with Flood 2. Therefore, Flood 2 cannot create additional scour and the scour depth versus time curve remains flat dur- ing Flood 2. If z–1 is smaller than z–max 2 then the procedure fol- lowed for the case of a small flood followed by a big flood applies, and the combined curve is as shown in Figure 4.4. In the general case, the velocity versus time history exhibits many sequences of small floods and big floods. The calcula- tions for scour depth are performed by choosing an increment of time ∆t and breaking the complete velocity versus time his- tory into a series of partial flood events, each lasting ∆t. The first two floods in the hydrograph are handled by using the procedure shown in Figure 4.2 or Figure 4.4, depending on the case. Then the process advances by stepping into time and considering a new “Flood 2” and a new t* at each step. The time ∆t is typically one day, and a velocity versus time history can be 50 years long. The many steps of calculations are han- dled with a computer program called SRICOS. The output of

Figure 4.3. Multiflood flume experiment results: a) floods and flood sequence in the experiments, b) experiment results for Flood 1 alone, c) experiment results for Flood 2 alone, d) experiment results for Floods 1 and 2 sequence shown in a) and prediction for Flood 2. Figure 4.4. Scour due to a sequence of two flood events (big flood followed by small flood).

is contained in Layer 1 and does not reach Layer 2. If, how- ever, the scour depth reaches ∆z–1 (Point A on Figure 4.5a), Layer 2 starts to be eroded. In this case, even though the scour depth ∆z–1 was due to the scour of Layer 1 over a time t1, at that time the situation is identical to having had Layer 2 scoured over a time t* (Point B on Figure 4.5b). Therefore, when Layer 2 starts being eroded, the scour depth versus time curve pro- ceeds from Point B to Point C on Figure 4.5b. The combined curve for the two-layer system is OAC on Figure 4.5c. 4.5 SOFT SOIL LAYER OVER HARD SOIL LAYER AND GENERAL CASE Layer 1 is soft and ∆z–1 thick. Layer 2 underlies Layer 1 and is harder than Layer 1. The scour depth z– versus time t curve for Layer 1 is given by Equation 4.4 (Figure 4.6a), and the z– versus t curve for Layer 2 is given by Equation 4.5 (Figure 4.6b). If ∆z–1 is larger than the maximum depth of scour in Layer 1, z–max 1, then the scour process is contained in Layer 1 and does not reach Layer 2. If, however, the scour depth reaches ∆z–1 (Point A on Figure 4.6a), Layer 2 starts to erode. In this case, even though the scour depth ∆z–1 was due to the scour of Layer 1 over a time t1, at that time the situation is 21 identical to having had Layer 2 scoured over an equivalent time t* (Point B on Figure 4.6b). Therefore, when Layer 2 starts being eroded, the scour depth versus time curve pro- ceeds from Point B to Point C on Figure 4.6b. The combined curve for the two-layer system is OAC on Figure 4.6c. In the general case, there may be a series of soil layers with different erosion functions. The computations proceed by stepping forward in time. The time steps are ∆t long, the velocity is the one for the corresponding flood event, and the erosion function (z˙– versus t) is the one for the soil layer cor- responding to the current scour depth (bottom of the scour hole). When ∆t is such that the scour depth proceeds to a new soil layer, the computations follow the process described in Figures 4.5 or 4.6 depending on the case. The same SRICOS program mentioned for the velocity hydrograph also handles these calculations. The output of the program is the scour depth versus time curve for the multilayered soil system and for the complete velocity hydrograph. 4.6 EQUIVALENT TIME The computer program SRICOS is required to predict the scour depth versus time curve as explained in the preceding Figure 4.5. Scour of a two-layer soil (hard layer over soft layer).

section. An attempt was made to simplify the method to the point where only hand calculations would be needed. This requires the consideration of an equivalent uniform soil and an equivalent time for a constant velocity history. The equivalent uniform soil is characterized by an average z˙– versus τ curve over the anticipated scour depth. The equivalent time te is the time required for the maximum velocity in the hydrograph to create the same scour depth as the one created by the complete hydrograph (Figure 4.7). The equivalent time te was obtained for 55 cases generated from 8 bridge sites. For each bridge site, soil samples were collected in Shelby tubes and tested in the EFA to obtain the erosion function z˙– versus τ; then the hydro- graph was collected from the nearest gage station and the SRI- COS program was used to calculate the scour depth. That scour depth was entered in Equation 4.3, together with the cor- responding z˙–i and z˙–max to get te. The z˙–i value was obtained from an average z˙– vs τ curve within the final scour depth by reading the z˙– value that corresponded to τmax obtained from Equation 4.1. In Equation 4.1, the pier diameter B and the maximum velocity vmax found to exist in the hydrograph over the period considered were used. The z–max value was obtained from Equa- tion 4.2 while using B and vmax for the pier Reynolds Number. The hydrograph at each bridge was also divided into shorter period hydrographs, and for each period an equivalent time te was calculated. This generated 55 cases (Briaud et al., 2002). The equivalent time was then correlated to the duration of the hydrograph thydro, the maximum velocity in the hydrograph 22 vmax, and the initial erosion rate z˙–i. A multiple regression on that data gave the following relationship: The regression coefficient for Equation 4.7 was 0.77. This time te can then be used in Equation 4.3 to calculate the scour at the end of the hydrograph. A comparison between the scour depth predicted by the extended SRICOS Method using the complete hydrograph and the simple SRICOS Method using the equivalent time is shown on Figure 4.8. 4.7 EXTENDED AND SIMPLE SRICOS-EFA METHOD For final design purposes, the extended SRICOS Method (E-SRICOS) is used to predict the scour depth z– versus time t over the duration of the design hydrograph. The method proceeds as follows: 1. Calculate the maximum depth of scour z–max for the design velocity by using Equation 4.2. 2. Collect samples at the site within the depth z–max. 3. Test the samples in the EFA to obtain the erosion func- tions (z˙– versus τ) for the layers involved. 4. Prepare the flow hydrograph for the bridge. This step may consist of downloading the discharge hydrograph from a United States Geological Survey (USGS) gage t t v z e hydro i hrs years m s mm hr ( ) = ( )( ) ( )( ) ( )( )− 73 4 7 0 126 1 706 0 20 . max . . ˙ ( . ) Figure 4.6. Scour of a two-layer soil (soft layer over hard layer). Figure 4.7. Velocity hydrographs: a) constant, b) true hydrograph. Both hydrographs would lead to the same scour depth.

23 Figure 4.8. Comparison of scour depth using Extended SRICOS and Simple SRICOS Methods. Figure 4.9. Examples of discharge hydrographs: a) Brazos River at US 90A, b) San Marcos River at SH 80, c) Sims Bayou at SH 35. Figure 4.10. Velocity hydrograph and scour depth versus time curve for Bent 3 of the Brazos River Bridge at US 90A. Figure 4.11. Velocity hydrograph and scour depth versus time curve for Bent 3 of the San Marcos River Bridge at SH 80. layers involved, the velocity hydrograph v versus t, the pier diameter B, the viscosity of the water υ, and the density of the water ρw. Note that the water depth y is not an input because at this time the solution is limited to a “deep water” condition. This condition is realized when y ≥ 2B; indeed beyond this water depth the scour depth becomes independent of the water depth (Melville and Coleman, 1999, p. 197). 6. The SRICOS program proceeds by a series of time steps; it makes use of the original SRICOS Method and of the accumulation algorithms described in Figures 4.2, 4.4, 4.5, and 4.6. The usual time step ∆t is 1 day because that is the usual reading frequency of the USGS gages. The duration of the hydrograph can vary from a few days to over 100 years. 7. The output of the program is the depth of scour versus time over the period covered by the hydrograph (Fig- ures 4.10, 4.11, and 4.12). station near the bridge (Figure 4.9). These discharge hydrographs can be found on the Internet at the USGS website (www.usgs.gov). The discharge hydrograph then needs to be transformed into a velocity hydrograph (Figures 4.10, 4.11, and 4.12). This transformation is performed by using a program such as HEC-RAS (1997), which makes use of the transversed river bottom profile at the bridge site to link the discharge Q (m3/s) to the velocity v (m/s) at the anticipated location of the bridge pier. 5. Use the SRICOS program (Kwak et al., 1999) with the following input: the z˙– versus τ curves for the various

For predicting the future development of a scour hole at a bridge pier over a design life tlife, one can either develop a synthetic hydrograph (much like is done in the case of earth- quakes) or assume that the hydrograph recorded over the last period equal to tlife will repeat itself. The time required to per- form Step 3 is about 8 hours per Shelby tube sample because it takes about eight points to properly describe the erosion function (z˙– versus τ curve) and, for each point, the water is kept flowing for 1 hour to get a good average z˙– value. The time required to perform all other steps, except for Step 2, is about 4 hours for someone who has done it before. In order to reduce these 4 hours to a few minutes, a simplified version of SRICOS, called S-SRICOS, was developed. Note that this simplified method is only recommended for preliminary design purposes. If S-SRICOS shows clearly that there is no need for refinement, then there is no need for E-SRICOS; if not, an E-SRICOS analysis must be performed. For preliminary design purposes, S-SRICOS can be used. The method proceeds as follows: 1. Calculate the maximum depth of scour z–max for the design velocity vmax by using Equation 4.2. The design velocity is usually the one corresponding to the 100-year flood or the 500-year flood. 2. Collect samples at the site within the depth z–max. 3. Test the samples in the EFA to obtain the erosion func- tion (z˙– versus τ) for the layers involved. 4. Create a single equivalent erosion function by averag- ing the erosion functions within the anticipated depth of scour. 5. Calculate the maximum shear stress τmax around the pier before scour starts by using Equation 4.1. In Equation 4.1, use the pier diameter B and the design velocity vmax. 24 6. Read the erosion rate z˙– corresponding to τmax on the equivalent erosion function. 7. Calculate the equivalent time te for a given design life of the bridge thydro for the design velocity vmax and for the z˙–i value of Step 6 by using Equation 4.7. 8. Knowing te, z˙–i, and z–max, calculate the scour depth z˙– at the end of the design life by using Equation 4.3. An example of such scour calculations is shown in Fig- ure 4.13. 4.8 CASE HISTORIES In order to evaluate the E-SRICOS and S-SRICOS Methods, eight bridges were selected (Figure 4.14). These bridges all satisfied the following requirements: the predominant soil type was fine-grained soils according to existing borings; the river bottom profiles were measured at two dates separated by at least several years, these river bottom profiles indicated any- where from 0.05 m to 4.57 m of scour; a USGS gage station existed near the bridge; and drilling access was relatively easy. The Navasota River Bridge at SH 7 was built in 1956. The main channel bridge has an overall length of 82.8 m and con- sists of three continuous steel girder main spans with four concrete pan girder approach spans. The foundation type is steel piling down to 5.5 m below the channel bed, which con- sists of silty and sandy clay down to the bottom of the piling according to existing borings. Between 1956 and 1996 the peak flood took place in 1992 and generated a measured flow of 1,600 m3/s, which corresponds to a HEC-RAS calculated mean approach flow velocity of 3.9 m/s at Bent 5 and 2.6 m/s at Bent 3. The pier at Bent 3 was square with a side equal to Figure 4.12. Velocity hydrograph and scour depth versus time curve for Bent 3 of the Sims Bayou River Bridge at SH 35.

25 Figure 4.13. Example of scour calculations by the S-SRICOS Method. Figure 4.14. Location of case history bridges. 0.36 m, while the pier at Bent 5 was 0.36 m wide and 8.53 m long and had a square nose. The angle between the flow direc- tion and the pier main axis was 5 degrees for Bent 5. River bottom profiles exist for 1956 and 1996 and show 0.76 m of local scour at Bent 3 and 1.8 m of total scour at Bent 5. At Bent 5, the total scour was made up of 1.41 m of local scour and 0.39 m of contraction scour as explained later. The Brazos River Bridge at US 90A was built in 1965. The bridge has an overall length of 287 m and consists of three continuous steel girder main spans with eight prestressed concrete approach spans. The foundation type is concrete pil- ing penetrating 9.1 m below the channel bed, which consists of sandy clay, clayey sand, and sand down to the bottom of the piling according to existing borings. Between 1965 and 1998, the peak flood occurred in 1966 and generated a mea- sured flow of 2,600 m3/s, which corresponds to a HEC-RAS calculated mean approach velocity of 4.2 m/s at Bent 3. The pier at Bent 3 was 0.91 m wide and 8.53 m long and had a

round nose. The pier was in line with the flow. River bottom profiles exist for 1965 and 1997 and show 4.43 m of total scour at Bent 3 made up of 2.87 m of local scour and 1.56 m of combined contraction and general scour as explained later. The San Jacinto River Bridge at US 90 was built in 1988. The bridge is 1,472.2 m long and has 48 simple prestressed concrete beam spans and 3 continuous steel plate girder spans. The foundation type is concrete piling penetrating 24.4 m below the channel bed at Bent 43 where the soil con- sists of clay, silty clay, and sand down to the bottom of the piles according to existing borings. Between 1988 and 1997, the peak flood took place in 1994 and generated a measured flow of 10,000 m3/s, which corresponds to a HEC-RAS cal- culated mean approach velocity of 3.1 m/s at Bent 43. The pier at Bent 43 was square with a side equal to 0.85 m. The angle between the flow direction and the pier main axis was 15 degrees. River bottom profiles exist for 1988 and 1997 and show 3.17 m of total scour at Bent 43 made up of 1.47 m of local scour and 1.70 m of combined contraction and gen- eral scour as explained later. The Trinity River Bridge at FM 787 was built in 1976. The bridge has three main spans and three approach spans with an overall length of 165.2 m. The foundation type is timber pil- ing and the soil is sandy clay to clayey sand. Between 1976 and 1993, the peak flood took place in 1990 and generated a measured flow of 2,950 m3/s, which corresponds to a HEC- RAS calculated mean approach flow velocity of 2.0 m/s at Bent 3 and 4.05 m/s at Bent 4. The piers at Bent 3 and Bent 4 were 0.91 m wide and 7.3 m long, and had round noses. The angle between the flow direction and the pier main axes was 25 degrees. River bottom profiles exist for 1976 and for 1992 and show 4.57 m of total scour at both Bent 3 and Bent 4, made up of 2.17 m of local scour and 2.40 m of contraction and general scour as explained later. The San Marcos River Bridge at SH 80 was built in 1939. This 176.2-m-long bridge has 11 prestressed concrete spans. The soil tested from the site is a low-plasticity clay. Between 1939 and 1998, the peak flood occurred in 1992 and gener- ated a measured flow of 1,000 m3/s, which corresponds to a HEC-RAS calculated mean approach flow velocity of 1.9 m/s at Bent 9. The pier at Bent 9 is 0.91 m wide and 14.2 m long and has a round nose. The pier is in line with the flow. River bottom profiles exist for 1939 and 1998 and show 2.66 m of total scour at Bent 9 made up of 1.27 m of local scour and 1.39 m of contraction and general scour as explained later. The Sims Bayou Bridge at SH 35 was built in 1993. This 85.3-m-long bridge has five spans. Each bent rests on four drilled concrete shafts. Soil borings indicate mostly clay lay- ers with a significant sand layer about 10 m thick starting at a depth of approximately 4 m. Between 1993 and 1996, the peak flood occurred in 1994 and generated a measured flow of 200 m3/s, which corresponds to a HEC-RAS calculated mean approach flow velocity of 0.93 m/s at Bent 3. The pier at Bent 3 is circular with a 0.76 m diameter. The angle between the flow direction and the pier main axis was 26 5 degrees. River bottom profiles exist for 1993 and 1995 and indicate 0.05 m of local scour at Bent 3. The Bedias Creek Bridge at US 75 was built in 1947. This 271.9-m-long bridge has 29 spans and Bent 26 is founded on a spread footing. The soil tested from the site varied from low plasticity clay to fine silty sand. Between 1947 and 1996, the peak flood occurred in 1991 and generated a measured flow of 650 m3/s, which corresponds to a HEC-RAS calculated mean approach flow velocity of 2.15 m/s at Bent 26. The pier at Bent 26 is square with a side of 0.86 m. The pier is in line with the flow. River bottom profiles exist for 1947 and 1996 and show 2.13 m of total scour at Bent 26 made up of 1.35 m of local scour and 0.78 m of contraction and general scour as explained later. The Bedias Creek Bridge at SH 90 was built in 1979. This 73.2-m-long bridge is founded on 8-m-long concrete piles embedded in layers of sandy clay and firm gray clay. Between 1979 and 1996, the peak flood occurred in 1991 and gener- ated a measured flow of 650 m3/s, which corresponds to a HEC-RAS calculated mean approach flow velocity of 1.55 m/s at Bent 6. The pier at Bent 6 was square with a side of 0.38 m. The angle between the flow direction and the pier main axis was 5 degrees. River bottom profiles exist for 1979 and 1996 and show 0.61 m of local scour at Bent 6. 4.9 PREDICTED AND MEASURED LOCAL SCOUR FOR THE EIGHT BRIDGES The data for all bridges is listed in Tables 4.2 and 4.3. For each bridge, the E-SRICOS and S-SRICOS Methods were used to predict the local scour at the chosen bridge pier loca- tion. One pier was selected for each bridge, except for the Navasota River Bridge at SH 7 and the Trinity River Bridge at FM 787 for which two piers each were selected. Therefore, a total of 10 predictions were made for these eight bridges. These predictions are not Class A predictions since the mea- sured values were known before the prediction process started. However, the predictions were not modified once they were obtained. For each bridge, Shelby tube samples were taken near the bridge pier within a depth at least equal to two pier widths below the pier base. The boring location was chosen to be as close as practical to the bridge pier considered. The distance between the pier and the boring varied from 2.9 m to 146.3 m (Table 4.2). In all instances, the boring data available was studied in order to infer the relationship between the soil lay- ers at the pier and at the sampling locations. Shelby tube sam- ples to be tested were selected as the most probable represen- tative samples at the bridge pier. These samples were tested in the EFA and yielded erosion functions z˙– versus τ. Figures 4.15 and 4.16 provide examples of the erosion functions obtained. The samples also were analyzed for common soil properties (Table 4.3). For each bridge, the USGS gage data was obtained from the USGS Internet site. This data consisted of a record of dis-

charge Q versus time t over the period of time separating the two river bottom profile observations (Figure 4.9). This dis- charge hydrograph was transformed into a velocity hydro- graph by using the program HEC-RAS (1997) and proceed- ing as follows. The input to HEC-RAS is the bottom profile of the river cross section (obtained from TxDOT records), 27 the mean longitudinal slope of the river at the bridge site (obtained from topographic maps, Table 4.2), and Manning’s roughness coefficient (estimated at 0.035 for all cases after Young et al., 1997). For a given discharge Q, HEC-RAS gives the velocity distribution in the river cross section, including the mean approach velocity v at the selected pier TABLE 4.2 Full-scale bridges as case histories TABLE 4.3 Soil properties at the bridge sites

28 between the two profiles was calculated with scour being pos- itive and aggradation being negative. The net area was then divided by the Width AB to obtain an estimate of the mean contraction/general scour. Once this contraction/general scour was obtained, it was subtracted from the total scour at the bridge pier to obtain the local scour at the bridge pier. In some instances there was no need to evaluate the contraction/general scour. This was the case of Bent 3 for the Navasota Bridge (Figure 4.17). In this case, the bent was in the dry (flood plain) at the time of the field visit, and the local scour could be measured directly. Figure 4.19 shows the comparison between E-SRICOS predicted and measured values of local scour at the bridge piers. The precision and accuracy of the method appear reasonably good. Although more than 10 data points may be preferable, note that these 10 data points represent 10 full- scale, real situations. The S-SRICOS Method was performed next. For each bridge pier, the maximum depth of scour z–max was calculated by using Equation 4.2. The velocity used for Equation 4.2 was the maximum velocity, which occurred during the period of time separating the two river bottom profile observations. Then, at each pier, an average erosion function (z˙– versus τ Figure 4.15. Erosion function for San Jacinto River sample (7.6 m to 8.4 m depth). Figure 4.16. Erosion function for Bedias Creek sample (6.1 m to 6.9 m depth). Figure 4.17. Profiles of Navasota River Bridge at SH 7. Figure 4.18. Profiles of Brazos River Bridge at US 90A. location. Many runs of HEC-RAS for different values of Q are used to develop a relationship between Q and v. The rela- tionship (regression equation) was then used to transform the Q-t hydrograph into the v-t hydrograph at the selected pier (Figures 4.10, 4.11, and 4.12). Then, the SRICOS program (Kwak et al., 2001) was used to predict the scour depth z– versus time t curve. For each bridge, the input consisted of the z˙– versus τ curves (erosion functions) for each layer at the bridge pier (Figures 4.15 and 4.16), the v versus t record (velocity hydrograph) (Figures 4.10, 4.11, and 4.12), the pier diameter B, the viscosity of the water υ and the density of the water ρw. The output of the pro- gram was the scour depth z– versus time t curve for the selected bridge pier (Figures 4.10, 4.11, and 4.12) with the predicted local scour depth corresponding to the last value on the curve. The measured local scour depth was obtained for each case history by analyzing the two bottom profiles of the river cross- section (Figures 4.17 and 4.18). This analysis was necessary to separate the scour components that added to the total scour at the selected pier. The two components were local scour and contraction/general scour. This separation was required because, at this time, SRICOS only predicts local scour. The contraction/general scour over the period of time separating the two river bottom profiles was calculated as the average scour over the width of the channel. This width was taken as the width corresponding to the mean flow level (Width AB on Figures 4.17 and 4.18). Within this width, the net area

29 Figure 4.19. Predicted versus measured local scour for the E-SRICOS Method. Figure 4.20. Predicted versus measured local scour for the S-SRICOS Method. Figure 4.21. Velocity hydrograph and predicted scour depth versus time curve for pier 1E of the existing Woodrow Wilson Bridge on the Potomac River in Washington D.C. evaluate the S-SRICOS Method are the same cases that were used to develop that method. Therefore, this does not repre- sent an independent evaluation. Details of the prediction process can be found in Kwak et al. (2001). E-SRICOS and S-SRICOS described above do not include correction factors for pier shape, skew angle between the flow direction and the pier main axis, shallow water depth effects, and multiple pier effect. Chapter 5 will show how to calculate those factors. 4.10 CONCLUSIONS The SRICOS Method predicts the depth of the local scour hole versus time curve around a bridge pier in a river for a given velocity hydrograph and for a layered soil system (Fig- ure 4.21). The method described in this chapter is limited to cylindrical piers and water depths larger than two times the pier width. The prediction process makes use of a flood accu- mulation principle and a layer equivalency principle. These are incorporated in the SRICOS computer program to gener- ate the scour versus time curve. A simplified version of this method is also described and only requires hand calculations. The simplified method can be used for preliminary design purposes. Both methods were evaluated by comparing pre- dicted scour depths and measured scour depths for ten piers at eight full-scale bridges. The precision and accuracy of both methods appear good. curve) within the maximum scour depth was generated. The maximum shear stress τmax around the pier before scour began was calculated using Equation 4.1, assuming that the pier was circular (Table 4.2). The initial scour rate z˙–i was read on the average erosion function for that pier (Table 4.2). The equiv- alent time te was calculated using Equation 4.7, using thydro equal to the time separating the two river bottom profile observations, and vmax equal to the maximum velocity that occurred during thydro (Table 4.2). Knowing te, z˙–i, and z–max, the scour depth accumulated during the period of thydro was calcu- lated using Equation 4.3. Figure 4.20 is a comparison of the measured values of local scour and the predicted values using the S-SRICOS Method. The precision and accuracy of the method appear reasonably good. The 10 case histories used to

Next: Chapter 5 - The SRICOS-EFA Method for Maximum Scour Depth at Complex Piers »
Pier and Contraction Scour in Cohesive Soils Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 516: Pier and Contraction Scour in Cohesive Soils examines methods for predicting the extent of complex pier and contraction scour in cohesive soils.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!