on modeling uncertainty in earthquake soil structure ...sokocalo.engr.ucdavis.edu/~jeremic/ ·...
TRANSCRIPT
Motivation ESSI Illustrations Summary
On Modeling Uncertainty inEarthquake Soil Structure Interaction
Analysis
Boris Jeremic et al.
University of California, DavisLawrence Berkeley National Laboratory, Berkeley
ILEE Forum,Tongji University, Shanghai, China
December 2016
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Outline
Motivation
ESSI Illustrations
Summary
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
Long Term Motivation
I High fidelity seismic modeling and simulation forinfrastructure objects
I Development of an expert numerical modeling andsimulation tool
I Use of expert tool and high fidelity numerical models toanalyze seismic behavior of soil structure systems
I Predict and inform rather than (force) fit
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
ESSI Hypothesis
I Nonlinear interplay of the Earthquake, Soil and Structureplays a decisive role in successes and failures
I Timing and spatial location of energy dissipationdetermines location and amount of damage
I When we control and direct timing and spatial location ofenergy dissipation, we optimize soil structure system for
I Safety andI Economy
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
High Fidelity Predictive CapabilitiesI Verification provides evidence that the model is solved
correctly. Mathematics issue.
I Validation provides evidence that the correct model issolved. Physics issue.
I Prediction: use of computational model to foretell the stateof a physical system under consideration under conditionsfor which the computational model has not been validated.
I Prediction under uncertainty:I Modeling uncertaintyI Parametric uncertainty
I Predictive capabilities with low Kolmogorov Complexity
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
Modeling Uncertainty: Simplified ModelsI Simplified modeling: Features (important ?) are neglected
(6D ground motions, inelasticity)I Modeling Uncertainty: unrealistic and unnecessary
modeling simplificationsI Modeling simplifications: justifiable iff higher level
sophistication model shows that simplified features are notimportant
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
Parametric Uncertainty: Material and LoadsI Significant uncertainty in material and loadsI Need to propagate uncertainty through simulation, to give
regulators and engineers information for design, licensing...
5 10 15 20 25 30 35
5000
10000
15000
20000
25000
30000
SPT N Value
You
ng’s
Mod
ulus
, E (
kPa)
E = (101.125*19.3) N 0.63
−10000 0 10000
0.00002
0.00004
0.00006
0.00008
Residual (w.r.t Mean) Young’s Modulus (kPa)
Nor
mal
ized
Fre
quen
cy
Transformation of SPT N-value: 1-D Young’s modulus, E (cf. Phoon and Kulhawy (1999B))
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Motivation
Real ESSI Simulator System
I The Real ESSI Program is a 3D, nonlinear, time domain,parallel finite element program specifically developed forHi-Fi modeling and simulation of Earthquake Soil StructureInteraction problems for infrastructure objects
I The Real ESSI Computers are distributed memoryparallel computers, clusters of clusters with multipleperformance processors and networks
I The Real ESSI Notes are a hypertext documentationsystem (Theory and Formulation, Software and Hardware,Verification and Validation, Examples, and Applications)detailing modeling and simulation of ESSI problems
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Illustrations
Nuclear Power Plants: 6D or 1D Seismic MotionsI Assume that a full 6D (3D) motions at the surface are only
recorded in one horizontal directionI From such recorded motions one can develop a vertically
propagating shear wave in 1DI Apply such vertically propagating shear wave to the same
soil-structure system
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Illustrations
Development of Realistic MotionsI Free field seismic motion modelsI Sources will send both P and S waves
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Illustrations
Realistic Earthquake Motions, 6D vs 1D
I Danger of picking one component of motions (1D) from 3DI Excellent (forced) fit, but not a prediction, information is lost
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Illustrations
6D vs 1D NPP ESSI Response Comparison
Jeremic
Modeling Uncertainty for ESSI
Motivation ESSI Illustrations Summary
Summary
I Reduction of modeling uncertainty
I Development and expert use of high fidelity modeling andsimulation numerical tools: Real ESSI Simulator System
I Education and training of users (researchers, designers,regulators, owners) will prove essential
I Collaborators: Yang, Cheng, Jie, Sett, Taiebat, Tafazzoli,Karapiperis, Abell, Pisanò, Luo, Feng, Sinha, Lacour, Yang,Behbehani, Wang, Petrone, Wong, McKenna, McCallen
I Funding from and collaboration with the US-NRC, US-NSF,US-DOE, CNSC, LBNL, LLNL, INL, ILEE, AREVA NPGmbH, and Shimizu Corp. is greatly appreciated,
Jeremic
Modeling Uncertainty for ESSI