Professional Documents
Culture Documents
NON-LINEAR TIME HISTORY ANALYSIS OF LARGE CONCRETE DAM CONSIDERING NEAR FIELD EFFECT
Presented By: Ravi Sharma Bhandari M. Sc. Structural Engineering Thesis Defense November 12, 2013
PRESENTATION OUTLINE
Introduction Research Objective Research Methodology Near-Fault Earthquake Characteristics Ground Motion and Their Characteristics Finite Element Modeling of Gravity Dam Structural Performance and Damage Criteria Linear Time History Analysis Technique Nonlinear Time History Analysis Techniques Conclusion and Recommendation
2
Several numbers of large hydropower projects are planned in Nepal for near future. Proper Analysis of Seismic forces is a matter of concern before the construction of these dam. Since the Northridge and Kobe earthquakes, there has been much discussion about the adequacy of design practice of concrete dams. Two significant damage to concrete dam occurred in the 1960s: Hsinfengkiang in China and Koyna in India. The damage was severe enough but no complete collapse, required major repairs and strengthening.
3
INTRODUCTION
RESEARCH OBJECTIVES
Performance Study of concrete gravity dams under near field earthquake pulse and comparison to the far field earthquake effects. To illustrate the application of linear and nonlinear time history methods to earthquake response analysis of gravity dams. To assess stability condition of the dam. Locations of occurrence of probable cracks on dam during Earthquake.
4
RESEARCH METHODOLOGY
Data collection and review of literatures Preparation of Finite Element model of dam (SAP2000 V14) Perform Linear time history analysis of Dam Perform non linear time history analysis of Dam Interpretation of results Evaluation, conclusion and recommendation
Source: The 14th World Conference on Earthquake Engineering, October 12-17, 2008, Beijing, China Identification of Near-Fault Earthquake Record Characteristics Ch.A. Maniatakis, I.M. Taflampas and C.C. Spyrakos
11
12
Equation Of Motion:
14
15 Figure : Basis for upper limit demand-capacity ratio and cumulative inelastic duration
Load Combinations:
Seismic Loads Case 1 2 3 4 Vertical (V) + + Horizontal (H) + + Static Loads (Stress) + + + +
16
17
Load Case: V + H (FP) + Static Load S11,heel = 11.37 MPa S22,heel = 6.35 MPa
18
2.4 MPa
4.8 MPa
Time, Seconds
Figure: Time history of major principal stress at the heel of dam for Imperial Valley-06 (Recording station: El Centro Array #6) Earthquake (FN + V)
20
Figure: Time history of major principal stress at the heel of dam for Imperial Valley-06 (Recording station: El Centro Array #6) Earthquake (FP + V)
21
Figure: Time History of horizontal displacement at top of the dam due to Imperial Valley-06 (Recording station: El Centro Array #6) Earthquake, Left: FN-Component, Right: FP Component
22
Area [%]
Figure : Comparison of cumulative inelastic duration of stress cycles with acceptance stresses at the heel of the dam
Crest Displacement and Stresses due to Near Field Earthquake (Envelop FP) Crest Displacement (mm) Horizontal Stress (S11) Vertical Stress (S22) NGA# x y Heel Toe Heel Toe 181 92.68 49.93 11.37 4.5 6.35 2.3 182 68.23 -25.98 7.62 1.61 3.98 0.84 779 123.54 -41.78 13.7 6.77 7.38 3.19 821 98.04 -33.87 8.65 2.49 4.39 1.08 825 206.73 -72.82 17.53 9.6 9.37 4.47 828 115.21 -35.78 11.08 4.56 5.76 2.03 879 86.86 -35.29 8.37 2.76 4.31 1.2 1044 131.98 -49.49 11.48 5.05 6.2 2.44 1063 148.65 -60.69 14.3 7.39 7.94 3.6 1084 137.59 -45.12 14.51 6.86 7.83 3.16 1085 116.26 -40.2 12.03 5.37 6.33 2.46 1086 89.9 -32.52 7.7 1.85 3.84 0.79 -64.24 16.3 9.13 8.76 4.26 1106 212.44 1119 292.93 -80.81 22.64 14.6 12.43 6.89 26 1120 191.27 -61.9 17.62 10.08 9.44 4.67
Figure: Time history of instantaneous factor of safety for near field earthquake [NGA#1084]
27
Figure 7. 95 Envelops of maximum stresses (N/mm2), Left: Horizontal Stress, Right: Vertical Stress
Maximum horizontal stress at heel = 5.24 N/mm2 and at toe = -0.25 N/mm2 Maximum vertical stress at heel = 2.47 N/mm2, and at toe = -0.21 N/mm2 Maximum top displacement = 49.9 mm to right and -17.69 mm to down.
28
-2
Figure: Time history of major principal stress at the heel of dam for Imperial Valley-06 (Recording station: Coachella Canal #4) Earthquake
Crest Displacement and Stresses due to Far Field Earthquake
Crest Displacement (mm) NGA# Case Horizontal Stress, N/mm2 (S11) Vertical Stress, N/mm2 (S22)
x 166 826 832 948 1105 Envelop Envelop Envelop Envelop Envelop 49.9 47.43 42.64 42.55 49.4
that of the concrete. The nonlinear response in the form of tensile cracking is likely to start at the base. Gap-friction elements are introduced at the base of the dam to simulate cracking and sliding response of the dam. For nonlinear model time-history analysis (FNA), only the nonlinear behavior of the Link/Support elements is included.
Where is the stiffness matrix for the linear elastic elements (all elements except the Links/Supports); C is the proportional damping matrix; M is the diagonal mass matrix; is the vector of forces from the nonlinear degrees of freedom in the Link/Support elements; u, , and are the relative displacements, velocities, and accelerations with respect to the ground; and r is the vector of applied loads.
30
31
Figure: Time history of sliding displacements of nodal points (heel and toe) at the base of the dam (NGA#1120)
Figure: Deflected shape at the time of maximum displacement. 31 gap elements out of 51 experienced opening and sliding (NGA#1120)
CONCLUSION
Smooth curved profile perform better than sharp profile changed
section. This study conclude that acceleration only may not be sufficient to cause damage to the structures, pulse type velocity time history is also may be the major factor for destruction. Thus cannot be ignored in design of large structures. This research conclude that while performing analysis of large concrete dam in tectonically active country like Nepal, it would be better to use near-fault earthquake records rather than code based or far-field records. In general Fault-normal component cause more damage than Faultparallel component. Result shows that, entire base will undergo sliding with a permanent displacement offset of about 40 mm in the downstream direction.
33
nodes and also for shell elements. Soil-Structure interaction can be modeled in different ways and compared to present study. Fluid-structure interaction can be modeled in different ways and compared to the present study by Westergaard Added Mass method. Cracked at other places than base also can be checked by Discrete crack modeling or smeared crack modeling.
34
ACKNOLODGEMENTS
Dr. Hari Ram Parajuli Prof. Dr. Prem Nath Maskey and All Faculty Members
of M.Sc. Structural Engineering and Civil Engineering Department Dr. Roshan Tuladhar Dr. Krisnna Prasad Dulal and Er. Sandip Upreti Er. Mahesh Acharya, NEA Er. Soyuz Gautam
35
36