Ghiocel Predictive Technologies Inc. AP1000 NI SSI Model. Westinghouse Co.

Transcription

1 EPRI AP1000 NI Model Studies on Seismic SSSI Effects For Hypothetical Site Condition That Includes 40 ft Backfill Soil Over A Hard Rock Formation Dan M. Ghiocel Ghiocel Predictive Technologies Inc. AP1000 NI SSI Model Dali Li, Nicholas T. Brown and Jennifer J. Zhang Westinghouse Co. OECD NEA SSI Workshop Ottawa, Canada, October 6-8,

2 Presentation Purpose: To present the results of a on-going research study on seismic structure-soil-structure interaction (SSSI) including the effects of motion incoherency. Specifically, the seismic SSSI between the AP1000 NI complex (EPRI AP1000 stick model, TR# ) and a generic adjacent Annex Building (AB) structure are investigated. Discuss the ACS SASSI-ANSYS integration for refined stress analysis and computation of seismic soil pressures including wall-soil separation and uplift 2

3 Presentation Content: 1. AP1000 Stick SSSI Research Study 2. ACS SASSI-ANSYS Integration for Refined Stress Analysis and Computation of Seismic Soil Pressure 3. Conclusions 3

4 1. AP1000 Stick SSSI Research Study Stick SSI Model NI20 SSI Model 150ft x 250ft Accurate structural modeling up to 30 Hz Accurate structural modeling up to 100 Hz 4

5 Seismic Incoherent SSI Approaches Stochastic simulation approach similar to Monte Carlo simulation used for probabilistic analyses. It is based on performing statistical SSI analyses for a set of random field realizations of the incoherent free-field motion. Respects in detail the SSI physics. Compute mean incoherent SSI responses and their scatter. Recommended for both simple and complex SSI models with rigid or flexible foundations. Deterministic approximate approaches based on simple rules for combining the incoherency modes (AS approaches) or modal SSI responses (SRSS approaches). Approximates the mean incoherent SSI responses. Recommended for simple stick models with rigid basemats. Stochastic Simulation validated by EPRI (TR# , Nov 2007) and endorsed by US NRC (ISG-01, May 2008). Reference approach for validating deterministic approaches. 5

6 AP1000 NI Complex and Annex Bldg Configurations Top View Y Lateral View Z Nuclear Island (NI) X NI AB Annex Building (AB) Backfill Soil,Vs=1,000fps Coherent SSI Analysis: - Seismic input defined by RG 1.60 input - Control motion at the top of the hard-rock (NI foundation) Incoherent SSI Analysis: Hard-Rock Formation Vs=8,000fps - Stochastic Simulation with 8 simulations (EPRI TR# ) Abrahamson coherency model for hard-rock sites (EPRI TR# ) 6

7 Coherent vs. Incoherent SSI Inertial Forces COHERENT Motion Amplitude Symmetric Structure Non-symmetric Rigid Structure Non-symmetric Flexible Structure INCOHERENT Motion Amplitude Symmetric Structure Non-symmetric Rigid Structure Non-symmetric Flexible Structure Kinematic SSI is important 7

8 Isolated NI and Coupled NI-AB SSI Models NI Model NI-AB Model AB 8

9 Isolated AB Founded on Soil and Rock Deposits. Coherent and Incoherent ATF Plots in Y-Direction Basemat (El.100ft) High Elevation (EL.154ft) Soil Rock Soil Rock NOTE: 40 ft backfill layer significantly amplifies the AB Seismic Input and SSI Response 9

10 Isolated AB Founded on Soil and Rock Deposits. Coherent 5% Damp ISRS Plots at AB Basemat Center X-Direction Y-Direction Rock Rock Soil Soil 10

11 AB and Coupled NI-AB Coherent and Incoherent SSI. ATF Plots at the AB Basemat Center (El. 100ft) X-Direction Y-Direction AB NI-AB AB NI-AB 11

12 AB and Coupled NI-AB Coherent and Incoherent SSI. 5% Damp ISRS Y-Dir at AB Basemat Corner (El. 100ft) X-Direction Y-Direction NI-AB NI-AB AB AB 12

13 NI and Coupled NI-AB Coherent and Incoherent SSI. 5% Damp ISRS at NI Complex - Top of SCV (282 ft) X-Direction Y-Direction NI NI NI-AB NI-AB 13

14 NI and Coupled NI-AB Coherent and Incoherent SSI. 5% Damp ISRS at NI Complex - Top of ASB (333 ft) Y-Direction Z-Direction NI-AB NI-AB NI NI 14

15 AB and Coupled NI-AB Coherent and Incoherent SSI. Shear Forces in the AB Structure X-Direction Y-Direction Lower Frequency Input (Filtered by Soil) High Frequency Input (Due to Large KSSI) 15

16 NI and Coupled NI-AB Coherent and Incoherent SSI. Shear Forces in the NI Complex Structures for X-Dir LONG 16

17 NI and Coupled NI-AB Coherent and Incoherent SSI. Shear Forces in the NI Complex Structures for Y-Dir TRANS ANIMATIONS

18 2. ACS SASSI ANSYS Interface for Refined Seismic Stress Analysis Exported SSI Model ANSYS Structural Model Automatically Converted From ACS SASSI Using PREP Module Refined Model ANSYS Refined Structural Model Using EREFINE command or ANSYS GUI (rank 1-6) COPYRIGHT 2010 OF GP TECHNOLOGIES 18

19 Computing Structural Stress/Forces Seismic SSI Analysis Selected Critical Time Steps for Maximum Stresses To be Used for Equivalent Static Structural Analysis Structural Element Stress tk Time t1 SSI Solution Time Frames As Equivalent Static Structural Loading at Critical Time Steps EQS Forces + BC Springs EQS Forces + BC Displacements EQS Relative Displacements tj Submodeling Mesh Refinement Variable Accuracy C-by-C Exact COPYRIGHT 2010 OF GP TECHNOLOGIES Exact 19

20 ACS SASSI ANSYS Interface for Seismic Soil Pressure Analysis ANSYS Soil FE Model Is Automatically Generated by SOILMESH Module Embedment mesh is extended. User controls extension size and mesh density. Can use EREFINE. Contact surfaces automatically added By ACS SASSI SOILMESH module COPYRIGHT OF GP TECHNOLOGIES - ACS SASSI-ANSYS INTEGRATION 20

21 Computing Seismic Soil Pressures Seismic SSI Analysis Selected Critical Time Steps for Maximum Stresses To be Used for Equivalent Static Structural Analysis Structural Element Stress tk Time t1 SSI Solution Time Frames As Equivalent Static Loading at Critical Time Steps EQS Forces Linear & Nonlinear tj EQS Relative Displacements Linear (Welded) separation COPYRIGHT 2010 OF GP TECHNOLOGIES 21

22 Linear Seismic Soil Pressure Analysis LINEAR (WELDED) - This option provides for a basic soil pressure analysis assuming there is no separation possible between the structure and the soil - Displacements from the interaction nodes of the structure are applied directly to the soil FE model. The structural FE model is not required for this case COPYRIGHT 2010 OF GP TECHNOLOGIES 22

23 Embedded Concrete Box Model (UHS Type) 23

24 Example of Soil FE model Created Automatically by New SOILMESH Module for A Box Structural Model COPYRIGHT 2010 OF GP TECHNOLOGIES 24

25 Equivalent-Static Stress Analysis for Structure- Soil System Model (Generated by SOILMESH) 25

26 ACS SASSI Seismic Soil Pressures for X-Input (Frame 903) SXX SYY SZZ Linear SSI Analysis 26

27 Seismic Soil Pressures for X-Input (Frame 903) ACS SASSI Element Center Stress SXX Linear SSI Analysis ANSYS Element Node Stress SSI Displacements wrt Free-Field SSI Seismic Forces COPYRIGHT 2010 OF GP TECHNOLOGIES 27

28 ACS SASSI Seismic Soil Pressures for X-Input (Frame 903) ACS SASSI Element Center Stress SYY Linear SSI Analysis ANSYS Element Node Stress SSI Displacements wrt Free-Field SSI Seismic Forces COPYRIGHT 2010 OF GP TECHNOLOGIES 28

29 ACS SASSI Seismic Soil Pressures for X-Input (Frame 903) ACS SASSI Element Center Stress SZZ Linear SSI Analysis ANSYS Element Node Stress SSI Displacements wrt Free-Field SSI Seismic Forces 29

30 Nonlinear Seismic Soil Pressure Analysis NONLINEAR CONTACT (FOUNDATION SEPARATION) - This option allows for the structure to separate from the soil using surface to surface contact elements in ANSYS - Both the structural elements and the soil elements are required. Both APDL files written from SOILMESH must be loaded into ANSYS. -Inertial Force should be applied to the structure. - Contact and target surfaces are included in the soil FE model COPYRIGHT 2010 OF GP TECHNOLOGIES 30

31 Nonlinear Seismic Soil Pressure Analysis COPYRIGHT 2010 OF GP TECHNOLOGIES 31

32 Element Stresses Von Misses Compression Side View Tension Side View COPYRIGHT 2010 OF GP TECHNOLOGIES 32

33 Effects of SSI Soil Separation on Soil Pressures for X-Input ANSYS Equivalent-Static Seismic Force Loading Option for Given Time Solid Element Center Stresses SXX, SYY, SZZ Foundation Wall Separation COPYRIGHT 2010 OF GP TECHNOLOGIES Foundation Uplift 33

34 3. Conclusions For the particular case investigated herein that assumed that NI complex is founded on a bedrock and embedded in a soft backfill layer and the neighbor AB is founded on the backfill soil (with no soil improvement or lean concrete underneath), the effects of SSSI for the coupled NI-AB model indicates - Relatively minor effects, mostly favorable, for the NI complex structures on ISRS and structural forces - Significant effects for the AB structure forces. The SSSI effects reduce significantly the shear forces. Also, change the ISRS frequency content. NOTE: The alternate Subtraction/Flexible Interface method provides a more accurate SSI solution than standard Subtraction/Flexible Interface method at the penalty of a significant increase of the SSI analysis run time. 34

35 Thank you! 35