1
Overview on BNL Assessment of Seismic Analysis Methods for Deeply Embedded NPP Structures
J. Xu1, C. Costantino1, C. Hofmayer1, H. Graves2
1Brookhaven National LaboratoryUpton, New York 11973-5000, USA
2U.S. Nuclear Regulatory CommissionWashington, D.C. 20555-0001, USA
Presented at19th International Conference on
Structural Mechanics in Reactor Technology
August 12-17, 2007Toronto, Canada
Brookhaven Science AssociatesU.S. Department of Energy
2
BACKGROUNDBACKGROUND
Need for the Research:
Several Conceptual Designs of Advanced Reactors Proposed Deeply Embedded and/or Buried (DEB) Features for NPP Structures
Understanding of Effects of DEB Features on Seismic Analysis Methods
Evaluation of Current Seismic Criteria for Soil-Structure Interaction (SSI) Pertaining to DEB Structures
3
PROGRAM OBJECTIVE AND TASKSPROGRAM OBJECTIVE AND TASKSObjective:
Investigate and Evaluate the Applicability of Existing Seismic SSI Practice, Regulatory Guidance and Computer Codes for DEB NPP Structures. Develop insights to be used for the safety evaluation of deeply embedded and/or buried (DEB) structures
Tasks:
Perform Literature Review of Existing Practices and Guidelines and Identify Potential Seismic Issues Requiring Further Study
Evaluation of Simplified vs. Detailed Methodologies for SSI Spectrum Computations (CARES versus SASSI).
Evaluation of Methods for Computing Seismic Induced Pressures for DEB Structures
4
SCOPE OF LITERATURE REVIEWSCOPE OF LITERATURE REVIEWReviewed the open literature to determine the state-of-the-art for performing seismic analyses of DEB NPP structures
Performed search on a broad range of literature using keywords such as: SSI, Kinematic, Wall pressures, Embedded, Earthquake measurements.
The search yielded hundreds of abstracts related to the subject
Reviewed more than 100 references
Brookhaven Science AssociatesU.S. Department of Energy
5
LITERATURE REVIEW OF SEISMIC ANALYSES AND PRACTICE
LITERATURE REVIEW OF SEISMIC ANALYSES AND PRACTICE
SSI phenomenon: kinematic and inertia interaction effects are well studied for embedded structures:
• Analytical efforts (authors such as Pais and Kausel, Iguchi, Roesset, Luco, etc.)
• Numerical efforts: SASSI, CLASSI, FLUSH, CARES
Seismic induced pressures
• Mononobe-Okabe (M-O) method for earth retaining walls
• Elastic solution by wood (ASCE 4)
• Others such as studies by Veletsos, Ostadan, etc., for seismic-induced
pressure calculations of buried structures.
• Japanese studies using field tests and shake table tests.
6
KEY ISSUES IMPORTANT TO DEB SSIKEY ISSUES IMPORTANT TO DEB SSI
Key Issues Attributes of Importance Current Computational Capability Kinematic Interaction
(KI)
Rigid foundation modifies free-field motion
Incorporated into codes such as SASSI SASSI for linear SSI response.
Free-Field Seismic Motions
P, SH or SV waves in free field motions need to be defined for computing SSI response
All computer codes typically utilized to address SSI response issues can address the issue of free-field input motion characteristics associated with various wave types.
Wall Pressures and Other Nonlinear
Effects
Nonlinear effects are important in determining seismic wall pressures.
The SASSI Code can only treat the equivalent linear problem. Nonlinear effects can currently only be treated in time-domain codes having this capability (LS-DYNA, ABAQUS). The codes require the definition of various input parameters to properly incorporate these effects within the model.
Sidewall Interaction
Kinematic interaction effect needs to be considered in simplified lump mass SSI models.
The SASSI Code can treat this problem correctly for a given free-field configuration and input motion. Simplified SSI codes (e.g., CARES) make use of parameters from a library of solutions available in the literature.
Ground Motion
Incoherence
High frequency (greater than 10 Hz) SSI response, which is important for facilities sited on hard rock.
The codes typically available to evaluate these effects are currently in their developmental stage. In addition, the data used to develop the incoherence parameters is relatively restricted.
7
EVALUATION OF SIMPLIFIED VS. DETAILED METHODOLOGIES FOR SSI
EVALUATION OF SIMPLIFIED VS. DETAILED METHODOLOGIES FOR SSI
Evaluation of Simplified vs. Detailed Methodologies for SSI Spectrum Computations for DEB StructuresCARES was Used for Simplified Method (NUREG/CR-6922)SASSI for Detailed Method (SASSI 2000, UC Berkeley)
8
STRUCTURAL CHARACTERISTICSSTRUCTURAL CHARACTERISTICS
9
STRUCTURAL CONFIGURATIONSTRUCTURAL CONFIGURATION
10
CARES AND SASSI ModelsCARES AND SASSI Models
SSI node
Rigid wall
11
Free Field ResponseFree Field Response
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
Frequency (cps)
Acce
lera
tion
(g)
Ground surface
Depth=11.5m
Depth=23m
Depth=34.5m
Depth=46m
Depth=80m
Rock outcrop
Free Field Soil Column Free Field Response Spectra
rock
Soil column80m
12
Soil ProfilesSoil Profiles
-80
-60
-40
-20
0200 300 400 500 600 700 800 900 1000
Shear Wave Velocity (m/s)
Dep
th (m
)
Quadratic distribution
Discrete distribution (Col. C)
Uniform lower bound (Col. A)
Uniform upper bound (Col. B
13
Rock Outcrop MotionRock Outcrop Motion
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.1 1 10 100
Frequency (cps)
Spe
ctra
l Acc
eler
atio
n (g
)
14
Comparison of Response Spectra DOB = 23 m, Soil Column C
Comparison of Response Spectra DOB = 23 m, Soil Column C
Base Roof
0
0.5
1
1.5
0.1 1 10 100
Frequency (cps)
Ac
ce
lera
tio
n (
g)
Free field motion
SASSI prediction
CARES prediction
0
0.5
1
1.5
2
2.5
3
3.5
4
0.1 1 10 100
Frequency (cps)
Ac
cele
rati
on
(g
)
SASSI prediction
CARES prediction
15
Comparison of Response Spectra DOB = 46 m, Soil Column C
Comparison of Response Spectra DOB = 46 m, Soil Column C
Base Roof
0
0.5
1
1.5
0.1 1 10 100
Frequency (cps)
Ac
ce
lera
tio
n (
g)
Free field motion
SASSI prediction
CARES prediction
0
0.5
1
1.5
2
0.1 1 10 100
Frequency (cps)
Ac
ce
lera
tio
n (
g)
Free fie ld motion
SASSI pred iction
CARES predic tion
16
Performance Evaluation for SSI Analyses, Soil Column C--- Area Differences
Performance Evaluation for SSI Analyses, Soil Column C--- Area Differences
17
Performance Evaluation for SSI Analyses, Soil Column C--- Peak Spectral Differences
Performance Evaluation for SSI Analyses, Soil Column C--- Peak Spectral Differences
18
SEISMIC SOIL PRESSURE CALCULATIONS SEISMIC SOIL PRESSURE CALCULATIONS
Evaluation of Methods for Computing Seismic Induced Pressures for DEB StructuresSASSI LS-DYNA: Explicit Finite Element ProgramAssumption of Linearity applied in both analyses
19
CALCULATION OF SOIL PRESSURES (Cont’d)CALCULATION OF SOIL PRESSURES (Cont’d)
Soil pressures computed for four depths of burial (11.5m, 23m, 34.5m, and 46m)Circumferential variation of pressures are found to be either sine or cosine as expectedVertical distribution of pressures shows strong SSI effect which is predominantly inertia interaction for shallow embedded structures to principally kinematic interaction for deeply embedded structures
20
LS-DYNA MODEL FOR 50% DOBLS-DYNA MODEL FOR 50% DOB
21
VERTICAL DISTRIBUTION OF SOIL PRESSURE FOR 25% DOB
VERTICAL DISTRIBUTION OF SOIL PRESSURE FOR 25% DOB
0
20
40
60
80
100
-200 -100 0 100 200 300
Soil Stress Components (kN/m^2)
Dep
th: (
z/E
)x10
0
Srr (SASSI)
Srt (SASSI)
Srz (SASSI)
Srr (LS-DYNA)
Srt (LS-DYNA)
Srz (LS-DYNA)
22
VERTICAL DISTRIBUTION OF SOIL PRESSURE FOR 100% DOB
VERTICAL DISTRIBUTION OF SOIL PRESSURE FOR 100% DOB
0
20
40
60
80
100
-200 -150 -100 -50 0 50
Soil Stress Components (kN/m^2)
Dep
th: (
z/E
)x10
0Srr (SASSI)
Srt (SASSI)
Srz (SASSI)
Srr (LS-DYNA)
Srt (LS-DYNA)
Srz (LS-DYNA)
23
Fourier Spectra of Soil Pressure for 75% Depth of Burial
Fourier Spectra of Soil Pressure for 75% Depth of Burial
0.00
0.20
0.40
0.60
0.80
1.00
1.20
0.1 1 10 100Frequency (cps)
Pre
ssur
e (k
N/m
2)
SASSI solution
LS-DYNA solution
24
ConclusionsConclusionsAssessed the state-of-the-art (current analyses methods acceptable for ground motions that do not result in non-linear SSI response).For SSI spectrum computation, both CARES and SASSI compare comparably; simplified approach mostly predicted slightly higher responses For wall pressure calculation, reasonably close comparisons were shown between SASSI and LS-DYNA (linear domain). When non-linearity is modeled (no discussion was made in this presentation; detailed analysis can be found in NUREG/CR-6896), such as contact area between foundation and soil, and nonlinear soil models, the wall pressure distributions would be vastly different from linear analysis.