analysis and design of seismically isolated nuclear ... and design of seismic... · analysis and...
TRANSCRIPT
Analysis and design of seismically isolated nuclear structures in the
United States
Andrew Whi8aker, Ph.D., S.E. Director, MCEER
University at Buffalo Annie Kammerer, Ph.D., P.E
US Nuclear Regulatory Commission Michael ConstanGnou, Ph.D.
University at Buffalo Michael Salmon, P.E.
Los Alamos NaGonal Laboratory
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Acknowledgments
• US Nuclear Regulatory Commission • Dr. Robert Budnitz, LBNL • Professor Yin-‐Nan Huang, NaGonal Taiwan University • Manish Kumar2 • ASCE Standard 4 Commi8ee • LBNL/USNRC oversight commi8ee – Nilesh Chokshi – Antonio Godoy – James Johnson – Robert Kennedy – Don Moore
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Outline
• Isolators for US pracGce • Performance expectaGons • Earthquake ground moGons for design • Analysis of nuclear structures • Design consideraGons
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Isolators
• Addressed for US pracGce – Low damping natural rubber – Lead-‐rubber – Spherical sliding (FP) bearing
• Others acknowledged in the NUREG/ASCE 4-‐** – Elastomeric
• High-‐damping rubber • SyntheGc rubber (neoprene)
– Sliding • EradiQuake
– 3D isolaGon system
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Isolators for the nuclear industry
• High-‐damping rubber – Compound + cure – Grant et al. (2004)
• Phenomenological model • BidirecGonal horizontal response • Calibrated to measured responses
– ElasGc force by 5th order polynomial – Nonlinear damping funcGon – Scragging and Mullins effects
• No rate dependence – Path forward for a HDR isolator
• One compound, complete cure – Documented process by isolator geometry – Thermo-‐chemical-‐mechanical analysis
• USNRC 6-‐step process • Develop V&V models without calibraGon
Journal of Earthquake Engineering,Vol. 8, Special Issue 1 (2004) 161–185c© Imperial College Press
BIDIRECTIONAL MODELLING OFHIGH-DAMPING RUBBER BEARINGS
DAMIAN N. GRANT
European School for Advanced Studies in Reduction of Seismic Risk (ROSE School),Universita degli Studi di Pavia, 27100 Pavia, Italy
GREGORY L. FENVES
Department of Civil and Environmental Engineering,University of California, Berkeley, CA 94720, USA
ANDREW S. WHITTAKER
Department of Civil, Structural and Environmental Engineering,State University of New York at Buffalo, NY 14260, USA
High-damping rubber (HDR) bearings are used in seismic isolation applications forbuildings and bridges, although no models are currently available for the accurate de-scription of the shear force–deformation response under bidirectional loading. A strainrate-independent, phenomenological model is presented which effectively represents thestiffness, damping, and degradation response of HDR bearings. The model decomposesthe resisting force vector as the sum of an elastic component in the direction of the dis-placement vector and a hysteretic force component parallel to the velocity vector. Theelastic component is obtained from a generalised Mooney–Rivlin strain energy function,and the hysteretic component is described by an approach similar to bounding sur-face plasticity. Degradation is decomposed into long term (“scragging”) and short term(“Mullins’ effect”) components. Calibration is carried out over a series of bidirectionaltest data, and the model is shown to provide a good match of slow strain-rate experimen-tal data using a unique set of material parameters for all tests. A testing protocol andcalibration of the model for use in design of structures with HDR bearings are discussed.
Keywords: High-damping rubber bearings; seismic isolation; mathematical model.
1. Introduction
Seismic isolation is widely used in buildings and bridges to protect them from theeffects of strong ground motion. Flexible isolation bearings are placed between theprimary mass of a structure and the support motion, effectively using inertia andincreased flexibility to limit structural deformation in critical components. In thismanner, buildings are isolated from their foundations, and the superstructures ofbridges are isolated from the piers.
High-damping rubber (HDR) bearings are a type of seismic isolator used inbridge and building construction and retrofit. As with other elastomeric isolation de-vices, HDR bearings are composed of layers of an elastomeric compound, reinforced
161
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
USNRC and DOE commonaliGes
• First Onset of Significant InelasGc DeformaGon – Applied at the component level – Developed for convenGonal nuclear structures – Adopted in principle for isolaGon NUREG
• Risk-‐oriented framework of ASCE 43 – FOSID at MAFE = E-‐5
– How applied to nonlinear isolaGon systems? – DBE = DF * UHS at E-‐4 = GMRS – 1% NEP for 100% DBE shaking – 10% NEP for 150% DBE (EDB GMRS) shaking
– Plant level HCLPF for USNRC-‐regulated NPPs – 1% NEP for 167% GMRS shaking
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Table 2.1 Performance expectations for seismically isolated safety-related nuclear structures (ASCE forthcoming)
Isolation system Superstructure Other SSCs
Hazard Use Isolation system displacement Performance Acceptance
criteria Performance Performance Umbilical lines Hard Stop or
Moat
DBE Response
spectrum per Chapter 2
Production testing of isolators. Design loads for isolated superstructure. In-structure response spectra (ISRS).
Mean and 80th percentile isolation system displacements.
No damage to the isolation system for DBE shaking.
Production testing of each isolator for the 80th percentile isolation system displacement and corresponding axial force. Isolators damaged by testing cannot be used for construction.
Conform to consensus materials standards for 80th percentile demands. Greater than 99% probability that component capacities will not be exceeded. Greater than 99% probability that the superstructure will not contact the moat.1
Conform to ASME standards for 80th percentile demands; adjust ISRS per Section 6.2.3. Greater than 99% probability that component capacities will not be exceeded.
- -
BDBE 150% of DBE
Prototype testing of isolators. Selecting moat width (or Clearance to Stop).
90th percentile isolation system displacement.2
Greater than 90% probability of the isolation system surviving BDBE shaking without loss of gravity-load capacity.
Prototype testing of a sufficient3 number of isolators for the CS displacement and the corresponding axial force. Isolator damage is acceptable but load-carrying capacity is maintained.
Greater than 90% probability that the superstructure will not contact the moat. Achieved by setting the moat width equal to or greater than the 90th percentile displacement. Greater than 90% probability that component capacities will not be exceeded.
Greater than 90% probability that component capacities will not be exceeded.
Greater than 90% confidence that all safety-related umbilical lines and their connections, shall remain functional for the CS displacement by testing, analysis or a combination of both.
Clearance to Stop (CS) or moat width equal to or greater than the 90th percentile displacement. Damage to the moat is acceptable in the event of contact.
1. Can be achieved by satisfying the requirement for BDBE shaking.
2. 90th percentile BDBE displacements may be calculated by multiplying the mean DBE displacement by a factor of 3.
3. The number of prototype isolators to be tested shall be sufficient to provide the required 90+% confidence.
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Earthquake input for RH analysis
• Basic representaGon is a spectrum – PSHA – Geometric mean UHRS
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Earthquake input for RH analysis
• Basic representaGon is a spectrum, but which one? – Uniform hazard spectrum (Cornell, 1968) – CondiGonal mean spectrum (Cornell and Baker, 2006) – CondiGonal spectra (Baker et al., 2011)
Earthquake input for RH analysis
• Ground moGons consistent with a spectrum – Spectrum compaGble – Maximum/minimum
Earthquake input for RH analysis • Ground moGons consistent with a spectrum – Column 1: 30 spectrum compaGble – Column 2: Max/min; preserve geometric mean
Earthquake input for RH analysis
• CorrelaGon of ground moGon components – USGS Technical Report by Huang et al. • WUS Far field • WUS Near field
• CEUS – MoGons for analysis should be appropriate • Recover characterisGcs of recorded ground moGons
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Earthquake input for RH analysis
• CorrelaGon of ground moGon components
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Earthquake input for RH analysis Statistics of the correlation coefficients for 147 WUS near fault motions
Type of correlation coefficient Mean Standard
deviation Median Ninetieth percentile
Cumulative probability for
Maximum value 0.24 0.15 0.21 0.44 0.74
FN-FP 0.14 0.12 0.12 0.31 0.89
Statistics of the correlation coefficients for 165 WUS far-field motions
Type of correlation coefficient Mean Standard
deviation Median Ninetieth percentile
Cumulative probability for
Maximum value 0.21 0.13 0.18 0.38 0.79
FN-FP 0.14 0.11 0.11 0.29 0.91
Siler InternaGonal Workshop, Rome June 18 and 19, 2013
Analysis methods
• Site independent – CerGfied plant design – Surface mounted LLWR or SMR – Need to quanGfy input at the foundaGon, SIDRS • Three translaGonal components • Three rotaGonal components?
– Dependent on facility geometry and soil properGes
• Site independent – Three components of input to soil domain
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Site dependent analysis
• Three methods of analysis • Fully coupled, nonlinear Gme domain • Fully coupled, frequency domain • MulGstep
• Fully coupled, nonlinear Gme-‐domain – Soil (LB, BE, UB), isolators, SSCs – ABAQUS, LS-‐DYNA, NRC ESSI – Used for all types of isolators – 3D soil domain, domain reducGon method
• Full coupled, frequency domain per ASCE 4 – LB, BE, UB soil properGes – SASSI or similar – LDR bearings
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Site dependent analysis
• MulG-‐step – Frequency domain analysis to compute SIDRS; equivalent linear isolators
– Ground moGons matched to SIDRS – Nonlinear analysis of isolated superstructure – Nonlinear models of isolators
– How many sets of ground moGons as input? – Mean response: 5 sets of moGons (15 analyses) – PercenGles (90, 99): 10 sets of moGons OR use factors derived by Huang et al. (2009)
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
Design consideraGons
• Basemat and foundaGon – Loss of isolator – Capacity design of pedestal and connecGons
• External events – Flooding – Aircrao impact, IED detonaGon
• Fire suppression system • OperaGng temperature • Isolator QA/QC • Prototype and producGon tesGng
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
MCEER research underway
• Earthquake engineering – Modeling seismic isolators under extreme loadings – Small modular reactors
• Issues related to embedment – Design procedures for hard stop – Time-‐domain SSI and SSSI analysis of NPPs – RC and SC walls, including fragility funcGons
• Blast and impact engineering - Aircrao impact on isolated nuclear structures
SILER InternaGonal Workshop, Rome June 18 and 19, 2013
www.csee.buffalo.edu www.mceer.buffalo.edu
SILER InternaGonal Workshop, Rome June 18 and 19, 2013