Anda di halaman 1dari 42

2-D Seismic Embankment and Slope Assessment and Stability

Wednesday, August 17, 2011 12:45 PM

Lower San Fernando Dam - 1971 San Fernando Valley Earthquake, Ca.
Main Issues in Seismic Assessment of Earthen Embankments and Dam: Stability: Is embankment stable during and after earthquake? Deformation: How much deformation will occur in the dam?

Two general types of analyses needed to answer these questions:


2D Dynamic Response Analysis 2D Deformation Analysis In some approaches, these two analyses are coupled.

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 1

General Types of 2D Seismic Analysis


Sunday, August 14, 2011 3:32 PM

Pseudostatic Analysis Newmark Sliding Block Analysis Makdisi and Seed (1978) used average accelerations computed by the procedure of Chopra (1966) and sliding block analysis to compute earthquake-induced deformations of earth dams and embankments. Numerically Based Analysis FEM Quake/W Plaxis FDM FLAC

This course will focus on Pseudostatic and Newmark Sliding Block Analyses using the Makdisi-Seed (1978) Method

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 2

Effects of Liquefaction
Wednesday, August 17, 2011 12:45 PM

from:

If the embankment and foundation materials are not susceptible to liquefaction or strength reduction due to earthquake shaking, then the embankment will generally he stable and no catastrophic failure is expected (Seed, 1979). However, if the embankment or/and foundation comprise liquefiable materials, it may experience flow failure depending on post-earthquake factor of safety against instability (FOSpe). For high initial driving stress (steep geometry), the FOS will likely be much less than unity, and flow failure may occur, as depicted by strain path A-B-C. Example of this is the failure of the Lower San Fernando Dam. In this lecture we will not address the effects of liquefaction on embankment stability. This is an advanced topic taught in CVEEN 7330.

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 3

Pseudostatic Analysis
Sunday, August 14, 2011 3:32 PM

Pseudostaic apply a static (non-varying) force the centroid of mass to represent the dynamic earthquake force. Fh = ah W / g = kh W Fv = av W/ g = kv W (often ignored)

Guidance on the Selection of Kh

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 4

Pseudostatic Analysis (cont.)


Sunday, August 14, 2011 3:32 PM

Recommendations for implementation of pseudostatic analysis (Bartlett) General comment: The pseudostatic technique is dated and should only be used for screening purposes. More elaborate techniques are generally warranted and are rather easy to do with modern computing software.

Limitations of Pseudostatic Technique Representation of the complex, transient, dynamics of earthquake shaking by a single, constant, unidirectional pseudostatic acceleration is quite crude. Method has been shown to be unreliable for soils with significant pore pressure buildup during cycling (i.e., not valid for liquefaction). Some dams have failed with F.S. > 1 from the pseudostatic technique Cannot predict deformation. Is only a relative index of slope stability

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 5

Pseudostatic Analysis - Example


Sunday, August 14, 2011 3:32 PM

Example Geometry

Example Soil Properties


Layer (lb/ft3) E (kPa) (top to (kN/m3) bottom) v K (kPa) G (kPa) c (kPa) Ko Vs (m/s)

1
2 3 4

15.72
16.51 17.29 18.08

100
105 110 115

100000
100000 150000 200000

0.37
0.37 0.35 0.3

128,205
128,205 166,667 166,667

36,496
36,496 55,556 76,923

24.37
24.37 27.49 34.85

0
0 0 0

0.5873
0.5873 0.5385 0.4286

150.9
147.3 177.5 204.3

5
emban

18.08
21.22

115
135

250000
300000

0.3
0.3

208,333
250,000

96,154
115,385

34.85
34.85

0
0

0.4286
0.4286

228.4
230.9

Pasted from <file:///C:\Users\sfbartlett\Documents\GeoSlope\miscdynamic1.xls>

E = Young's Modulus = Poisson's ratio K = Bulk modulus G = Shear Modulus = drained friction angle c = cohesion Ko = at-rest earth pressure coefficent Vs = shear wave velocity

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 6

Pseudostatic Analysis - Example


Sunday, August 14, 2011 3:32 PM

Pseudostatic Results

FS = 1.252 (static with no seismic coefficient, K h)

The analysis has been repeated by selecting only the critical circle. To do this, only one radius point. This result can then be used with a Kh value to determine the factor of safety, FS.

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 7

Pseudostatic Analysis - Example


Sunday, August 14, 2011 3:32 PM

Acceleration time history


0.6 0.5 0.4 Acceleration [g] 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 0 1 2 3 4 5 6 7 8 9 Time [sec] 10 11 12 13 14 15 16

Response Spectrum for acceleration time history


1.4 1.35 1.3 1.25 1.2 1.15 1.1 1.05 1 0.95 0.9 Response Acceleration [g] 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 0 1 2 Period [sec] 3 Damp. 5.0%

pga = 0.6 g Kh = 0.5 * pga ah = 0.3 g (This is applied in the software as a horizontal acceleration).

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 8

Pseudostatic Analysis - Example


Sunday, August 14, 2011 3:32 PM

Reduce shear strength in stability model for all saturated soils to 80 percent of peak strength as recommended by the Army Corp of Engineers. This is to account for pore pressure generation during cycling of non-liquefiable soils. (See table below.) (If liquefaction is expected, this method is not appropriate.)
Layer (lb/ft3) E (kPa) 3 (top to (kN/m ) bottom) v K (kPa) G (kPa) Tan 80 percent Tan 0.3624 0.3624
0.4162 0.5571

New phi angle for analysis 19.92 19.92


22.60 29.12

1 2
3 4

15.72 16.51
17.29 18.08

100 105
110 115

100000 100000
150000 200000

0.37 0.37
0.35 0.3

128,205 128,205
166,667 166,667

36,496 36,496
55,556 76,923

24.37 24.37
27.49 34.85

0.4530 0.4530
0.5203 0.6963

5 embank

18.08 21.22

115 135

250000 300000

0.3 0.3

208,333 250,000

96,154 115,385

34.85 34.85

0.6963 0.6963

0.5571 0.5571

29.12 29.12

Pasted from <file:///C:\Users\sfbartlett\Documents\GeoSlope\miscdynamic1.xls>

The analysis is redone with Kh = 0.3 and reduced shear strength (see below).

35 0.651 24 29 32 33 34 2 6 8 10 12 14 16 18 20 22 1 23
91 81 71 61 51 41 31 21 11 1 92 82 72 62 52 42 32 22 12 2 93 83 73 63 53 43 33 23 13 3 94 84 74 64 54 44 34 24 14 4 108 101 95 85 75 65 55 45 35 25 15 5

25
149 143 150 144 138 132 126 121 117 109 110 103 97 87 77 67 57 47 37 27 17 7 127 122 116 104 98 88 78 68 58 48 38 28 18 8 151 145 139 133 140 134 152 146 154 148 142 136 130 124 114 112 105 99 89 79 69 59 49 39 29 19 9 153 147 141 135 128 119 113 107 106 100 90 80 70 60 50 40 30 20 10

26

28

27
137 131

31

30
125 120 118

11
111

129 123 115

36

1 2 3 4 5 6 7 8 9 10

102 96 86 76 66 56 46 36 26 16 6

3 5 7 9 11 13 15 17 19 21 4

The resulting factor of safety is 0.651 (too low). Deformation is expected for this system and should be calculated using deformation analysis (e.g., Newmark, Makdisi-Seed, FEM, FDM methods.)

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 9

Newmark Sliding Block Analysis


Sunday, August 14, 2011 3:32 PM

Pasted from <http://pubs.usgs.gov/of/1998/ofr-98-113/ ofr98-113.html>

Newmarks method treats the mass as a rigid-plastic body; that is, the mass does not deform internally, experiences no permanent displacement at accelerations below the critical or yield level, and deforms plastically along a discrete basal shear surface when the critical acceleration is exceeded. Thus, for slope stability, Newmarks method is best applied to translational block slides and rotational slumps. Other limiting assumptions commonly are imposed for simplicity but are not required by the analysis (Jibson, TRR 1411).
1. The static and dynamic shearing resistance of the soil are assumed to be the same. (This is not strictly true due to strain rate effects 2. In some soils, the effects of dynamic pore pressure are neglected. This assumption generally is valid for compacted or overconsolidated clays and very dense or dry sands. This is not valid for loose sands or normally consolidated, or sensitive soils. 3. The critical acceleration is not strain dependent and thus remains constant throughout the analysis. 4. The upslope resistance to sliding is taken to be infinitely large such that upslope displacement is prohibited. (Jibson, TRR 1411)
Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 10

Newmark Sliding Block Analysis (cont.)


Sunday, August 14, 2011 3:32 PM

Steps
1. Perform a slope stability analysis with a limit equilibrium method and find the critical slip surface (i.e., surface with the lowest factor of safety) for the given soil conditions with no horizontal acceleration present in the model. 2. Determine the yield acceleration for the critical slip circle found in step 1 by applying a horizontal force in the outward direction on the failure mass until a factor of safety of 1 is reached for this surface. This is called the yield acceleration. 3. Develop a 2D ground response model and complete 2D response analysis for the particular geometry. Use this 2D ground response analysis to calculate average horizontal acceleration in potential slide mass. 4. Consider horizontal displacement is possible for each time interval where the horizontal acceleration exceeds the yield acceleration (see previous page). 5. Integrate the velocity and displacement time history for each interval where the horizontal acceleration exceeds the yield acceleration (see previous page).

The following approach is implemented using the QUAKE/WTM and SLOPE/WTM.

Acceleration vs. time at base of slope from 2D response analysis in Quake/W.


Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 11

Newmark Sliding Block Analysis (cont.)


Sunday, August 14, 2011 3:32 PM

Analysis perfromed using shear strength = 100 percent of peak value for all soils (i.e., no shear strength loss during cycling).

35 1.530 24 29 32 33 34 2 6 8 10 12 14 16 18 20 22 1 23
91 81 71 61 51 41 31 21 11 1 92 82 72 62 52 42 32 22 12 2 93 83 73 63 53 43 33 23 13 3 94 84 74 64 54 44 34 24 14 4 108 101 95 85 75 65 55 45 35 25 15 5

25
149 143 150 151 145 139 133 127 122 116 110 103 97 87 77 67 57 47 37 27 17 7 152 146 140 134 129 123 115 111 104 98 88 78 68 58 48 38 28 18 8 154 148 142 136 130 124 114 112 105 99 89 79 69 59 49 39 29 19 9 153 147 141 135 128 119 113 107 106 100 90 80 70 60 50 40 30 20 10 144 138 132 126 121 117

26

28

27
137 131

31

30
125 120 118

11

109

36

1 2 3 4 5 6 7 8 9 10

102 96 86 76 66 56 46 36 26 16 6

3 5 7 9 11 13 15 17 19 21 4

Factor of Safety vs. Time


2.0

Note that the same circle is used as obtained from the pseudostatic analysis !

1.8

Factor of Safety

1.6

1.4

1.2

1.0 0 5 Time 10 15 20

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 12

Newmark Sliding Block Analysis (cont.)


Sunday, August 14, 2011 3:32 PM

Analysis repeated using shear strength = 80 percent of peak value for all soils to account for some pore pressure generation during cycling.

35 1.365 24 29 32 33 34 2 6 8 10 12 14 16 18 20 22 1 23
91 81 71 61 51 41 31 21 11 1 92 82 72 62 52 42 32 22 12 2 93 83 73 63 53 43 33 23 13 3 94 84 74 64 54 44 34 24 14 4 108 101 95 85 75 65 55 45 35 25 15 5

25
149 143 150 151 145 139 133 127 122 116 110 103 97 87 77 67 57 47 37 27 17 7 152 146 140 134 129 123 115 111 104 98 88 78 68 58 48 38 28 18 8 154 148 142 136 130 124 114 112 105 99 89 79 69 59 49 39 29 19 9 153 147 141 135 128 119 113 107 106 100 90 80 70 60 50 40 30 20 10 144 138 132 126 121 117

26

28

27
137 131

31

30
125 120 118

11

109

36

1 2 3 4 5 6 7 8 9 10

102 96 86 76 66 56 46 36 26 16 6

3 5 7 9 11 13 15 17 19 21 4

Factor of Safety vs. Time


1.8

1.6

Factor of Safety

1.4

1.2

1.0 0 5 Time 10 15 20

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 13

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 14

Makdisi - Seed Analysis (cont.)


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 15

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 16

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 17

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 18

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 19

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 20

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 21

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 22

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 23

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 24

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 25

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 26

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 27

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 28

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 29

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 30

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 31

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 32

Makdisi - Seed Analysis


Wednesday, August 17, 2011 12:45 PM

More on the yield acceleration The yield acceleration, a y, is equal to the horizontal acceleration (g) applied to the potential failure mass that produces a factor of safety of 1.0 (see example below). It can be determine from limit equilibrium or other methods. The yield acceleration for the example below varies from 0.26 to 0.31 g depending on the undrained shear strength, Su, used for the embankment properties.
The yield coefficient, ky, is equal to the yield acceleration (g) divided by g; hence it is unitless. The yield coefficient for the below example varies from 0.26 to 0.31 (unitless)

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 33

Makdisi - Seed Analysis with Deformation Analysis p. 1 of 2


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 34

Makdisi - Seed Analysis with Deformation Analysis p. 2 of 2


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 35

Advanced Numerical Methods


Wednesday, August 17, 2011 12:45 PM

Dealt with in more detail in CVEEN 7330 Based on finite difference or finite element techniques Full dynamics modeled Deformation can be estimated using elasto-plastic or other constitutive models Required advanced training

Slope geometry for analysis for FDM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 36

Advanced Techniques (cont.)


Wednesday, August 17, 2011 12:45 PM

Acceleration time history (m/s^2 ) applied at base of model

Acceleration time history (m/s^2 ) applied at crest of embankment

Horizontal displacement (m) predicted by model for weak shallow foundation layer with phi = 20 deg. at end of 35 s of strong motion
Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 37

Advanced Techniques (cont.)


Wednesday, August 17, 2011 12:45 PM

Horizontal displacement (m) predicted by model for liquefied shallow foundation layer with phi = 10 deg. at end of 35 s of strong motion

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 38

Summary of Embankment Stability Analyses


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 39

Summary of Embankment Stability Analyses (cont.)


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 40

Summary of Embankment Stability Analyses (cont.)


Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 41

Blank
Wednesday, August 17, 2011 12:45 PM

Steven F. Bartlett, 2011

Lecture 9 - 2D Dynamic Analyses Page 42

Anda mungkin juga menyukai