Earthquake Soil Structure Interaction Modeling and Simulation

Transcription

1 Motivation Real Challenges Earthquake Soil Structure Interaction Modeling and Simulation Boris Jeremić University of California, Davis Lawrence Berkeley National Laboratory, Berkeley Southern California Earthquake Center Los Angeles, May 215

2 Motivation Real Challenges Outline Motivation Real Challenges

3 Motivation Real Challenges Introduction Motivation Improve seismic design of soil structure systems Earthquake Soil Structure Interaction (ESSI) in time and space, plays a major role in successes and failures Accurate following and directing (!) the flow of seismic energy in ESSI system to optimize for Safety and Economy Development of high fidelity numerical modeling and simulation tools to analyze realistic ESSI behavior: Real ESSI simulator (aka: Stvarno Lako, Muy Fácil, Molto Facile,,, Πραγματικά Εύκολο,, Très Facile, Vistinski Lesno )

4 Motivation Real Challenges Introduction Predictive Capabilities Verification provides evidence that the model is solved correctly. Mathematics issue. Validation provides evidence that the correct model is solved. Physics issue. Prediction under Uncertainty (!): use of computational model to foretell the state of a physical system under consideration under conditions for which the computational model has not been validated. Predictive capabilities with low Kolmogorov Complexity Modeling and simulation goal Inform, not fit Predict, not diagnose

5 Motivation Real Challenges Introduction Reduction of Modeling Uncertainty Simplified modeling: Features (important?) are neglected (6D ground motions, inelasticity) Modeling Uncertainty: unrealistic and unnecessary modeling simplifications Modeling simplifications are justifiable if one or two level higher sophistication model shows that features being simplified out are not important

6 Motivation Real Challenges Friction Angle [ ] Friction Angle [ ] Friction Angle [ ] Friction Angle [ ] Undrained Shear Strength [kpa] Undrained Shear Strength [kpa] Undrained Shear Strength [kpa] Undrained Shear Strength [kpa] Introduction Uncertain Material and Loads Young s Modulus, E (kpa) E = (11.125*19.3) N.63 Normalized Frequency SPT N Value Probability Density function Field φ Cumulative Probability Density function Probability Density function Field c u 1 1 Residual (w.r.t Mean) Young s Modulus (kpa) Cumulative Probability Density function Probability Density function Lab φ Cumulative Probability Density function Probability Density function Lab c u Cumulative Probability Density function

7 Motivation Real Challenges Modeling and Simulation Issues Real Issues Seismic Motions: 6D, inclined, body and surface seismic waves (translations, rotations); Incoherency Inelastic material: soil, rock, concrete, steel; Contacts, foundation soil, dry, saturated slip gap; Nonlinear buoyant forces; Isolators, Dissipators Uncertain material and loading (above) Verification and Validation Predictions High Fidelity Models High Performance Computing Education

8 Motivation Real Challenges Modeling and Simulation Issues 6D Seismic Motions and ESSI large scale seismic field Free Field at location ESSI for an NPP

9 Motivation Real Challenges Slip Behavior No-slip Behavior Slip Behavior No-slip Behavior Slip Behavior No-slip Behavior Slip Behavior No-slip Behavior Modeling and Simulation Issues Inelastic Contact, Base Slip and Gap Acceleration (m/s2) 2-2 Acceleration (m/s2) 2-2 Acceleration (m/s2) 2-2 Acceleration (m/s2) top X top Z bottom X bottom Z 4.5s 4.6s 4.7s 4.8s 4.9s 5.s 5.1s 5.2s Energy (kj)

10 Motivation Real Challenges Modeling and Simulation Issues 1D Wave (?!), No Volume Change and Dilative Soil Lack of soil volume change data for regular G/G max and damping testing Significant influence on Seismic motion propagation Acceleration [m/s 2 ] no dilatancy vs dilatancy surf - no dil surf - dil Time [s] τ [kpa] τ [kpa] 2 z = H/ γ [-] z = H/ γ [-]

11 Motivation Real Challenges Modeling and Simulation Issues Uncertain Inelastic Wave Propagation.3 Probabilistic Elasto-Plasticity I Stochastic Elastic-Plastic Finite Element Method.25 Stress (MPa) I Mode Mean Deterministic Solution Std. Deviations Real Soil Data Conservative Guess.2 Displ acem 5 2 ent (m m) 4 Jeremic et al Real Soil Data Conservative Guess Tim 1 2 e (s 15 4 ec) CDF.2 PDF PDF Strain (%) Displacement (mm) Displacement (mm) 1 Probability of Exceedance of 5 cm (%) 8 6 Conservative Guess 4 Real Site 2 Excellently Characterized Site

12 Motivation Real Challenges Challenges Challenges Most technical challenges are actively worked on (US-NRC, US-NSF, US-DOE, CNSC projects) and results are already available (6D motions, inelasticity, uncertainty, verification, validation, &c.) Validation needs more work (new US-DOE project!) Change state of practice Change state of research Education (designers, regulators, owners, &c.) is essential

13 Motivation Real Challenges Challenges Acknowledgement Funding from and collaboration with the US-NRC, US-DOE, US-NSF, CNSC, AREVA NP GmbH, and Shimizu Corp. is greatly appreciated, Collaborators: Dr. Budnitz (LBNL), Mr. Orbović (CNSC), Prof. Pisanò (TU Delft), Prof. Sett (UB), Mr. Watanabe (Shimizu) and UCD students: Mr. Abell, Mr. Karapiperis, Mr. Feng, Mr. Sinha, Mr. Luo