Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Session 8: SASSI Approach to Incoherency(Simulation and Algebraic Sum Methods)
SShort4:15 - 4:45
=j• ,I I • , O~
! SASSI Incoherency Approaches
" There are four methods for considering incoherentground motion in SSI analyses by SASSI
- SASSI-SRSS
- EPRI-INCOH
- SASSI Simulation- SASSI-AS
" Treatment of the spatial modes distinguishes thesemethods
" SASSI Simulation and SASSI-AS are implemented inACS SASSI, a commercially available version of theprogram (Dan Ghiocel, GP Technologies, Inc.)
2r .......
1
I SASSI -Simulation (Randomization)
" Perform Monte Carlo simulations varying the randomphase 0 for each spatial mode
- {Ug'} = [A(Io)W] [X('o)] {lo(o)} U0 (Co)* Loop on number of simulations (5, 10, 15, 20)
- Loop on no. of SASSI solution frequencies (10s to100s)
- Randomly sample 0 for each spatial mode• At each SASSI solution frequency, form {Ug,}* Numerical techniques for smoothing & phase adjustment" Calculate SSI response for the simulation (ISRS)
" Calculate the mean of the responses
3 .. u. - 3
I SASSI Simulation Individual Results
Foundotion Z Repons du to Z Input
Motion by SASS'-SrnWultions
1 4
0.4
02
4 C-F=Iel I 'iV Ok ..~P ft....,t *)fl Ok Aa "Un
2
SASSI Simulation Mean Results
Foundation Z Response due to Z Input Motlon by SASSI-SImulItions
12.
n2•
5 rý-f=i~i I T-ýX,,- AI~4~Od
ISASSI Simulation Individual Results
Nods 229, CIS Outfigg.r Z Response due. to Z Inpu.t Notion by SASSI-Siomolotions
II
yrtlooncy (00)
~I~I2I I6
3
I SASSI Simulation Mean Results
Nods 229, CIS Outdigger R..opons. due to Z Input Motion by SASIS"Imnlations
I
10Freepn•/jz
E~I2I ~7
I SASSI-Simulation Analysis Considerations
" Selection of coherency function (1 through 5)" Specification of the number of spatial modes (no
reason to use less than all modes for this approach)" Specification of random seeds and range for spatial
mode phasing* Selection of transfer function smoothing parameter* Decide on number of simulations* Compute the mean of the response quantities of
interest
0006~~~~mr-a j:XbP.0,n0- -noo..--a -1-2
4
SASSI - AS (Algebraic Sum)
• Equivalent to one simulation of SASSI - Simulation with:- Phase angle 0 = 0 for each spatial mode
* {U9,} = [W(O)] [X(cO)I {r0(O)=1} U0(o0)" Numerical techniques for smoothing & phase adjustment" All spatial modes included" SSI response calculated directly
*X0otfld, tan. W~I o.*cn ,.
9 ~=I-E~I ooto
SASSI-AS Analysis Considerations
" Selection of coherency function (1 through 5)" Specification of the number of spatial modes (no
reason to use less than all modes for this approach)" Choice of deterministic spatial mode phasing" Selection of transfer function smoothing parameter
10 ' R nTU"
5
I ACS SASSI Numerical Techniques forIncoherency Analyses
- Numerical techniques to assure the proper relationshipof transfer function frequency to frequency- Smoothing and interpolation - using the Parzen
windowing technique
- Response phasing adjusted by limiting to range of -n/2 to +n/2 to produce higher energy response timehistories
Transfer Function Smoothing
NO& r In.YZI•NF -e8
SpuJe~us peaks usMi the oingal SASSI* unteqxiatan for 201 selected frequencies.
These qxinous peaks were not rewvered,4 ~where we selected a larger nurnber of
/ t : f/ nluencies; weconsidered 201, 224,493
10 M
2-l•2.....
6
I Phase Adjustment
I, Node 229, computed 6 dof ITF Rosette (complex plane plots)at 20Hz for 20 Z-dsaking Random Samples
10-•7y
"1 51 -5z 0 y 0
0-
SYPhMs adjustnmet tnIftthe relative phase angle
GY (avoid cancabg sifads).
-105 0 5 10 Is
I Effect of Phase Adjustment
5% D-.p~d SASM Mo d. 145 ISCV Ouftn). Y-0*.MM X4%0",APO.Abuv"Ww"..
S
I Ma
a-r-rail I-14
7
I Effect of Phase Adjustment
S. OWW M A N, 145 (.&M Y O• W.tI Y-eo,.. Z4-ih.M4100 ,"lMO.MiM IAO"d
U8
a, 10
15 =I-I- I l~EO I~
Validation
• SASSI Simulation and SASSI-AS have been validatedfor use by:- Comparison of results with other SASSI and CLASSI
methods for a validation problem- Comparison to Mita and Luco, 1986 published results
16 =-.ri I2
8
I Parameters of the Validation Problem
" Coherency functions - NAA (2005, 2006)" Free-field ground motion defined by acceleration time
histories matching site specific spectra for rock site(horizontal and vertical)
" Rock site profile
* Foundation is square - 150 ft on a side - 15 ft thick withmass properties
• Structure model based, in part, on an advanced reactorstructure stick model with eccentricities (160 modes)
" In-Structure Response Spectra (ISRS) (5% damped)
.. 0. 17 C=IýI2 I
I Coherency Function for Horizontal GroundMotion (NAA 2005,2006)
Horizontal Coherency as a Function of Frequency & Separation Distance
I1000
0,800
I 0,400
c 0400
0,00.0
0 5 10 15 20 25 30 35 40 45 50
Frequency (Hz)
0 meters - 0.5 meter - 1 meter - 5 meters - 10 meters - 25meters m eters
r _CC111 .18.........
9
I Rock Site ProfileShear Wave Velocities vs. Depth
° L Rock
2 30_
.. ,..fl 0. 0MI3.... EPI21 .......
10
I Site-Specific Response Spectra forRock Site at Ground Surface (Depth O-ft)
RG 160 H ...... t4 (ZPA.3g- - RG 1.60 v1r&Cl (ZPA-0.3g)
0.2 0001 -441
F•oq.on, ,H
210.1 10. . IN
J°,,! ,a-ra,
I Computed andfor Rock SiteTarget Response Spectra
5% Damped Free Field CEUS Rock Motion
2.0
2..
IA-
CA
0.2-
0.1 10 IN
Fqýc.y (Wo)
I -Hak-At I - HooIý 2 VellI - HwiMalbh Tbve Vertica Tau [
m f - eI I E i-Z!.122
11
I Structural Model Characteristics
• Advanced reactor structure stick model witheccentricities- 160 fixed-base modes model the dynamic characteristics of the
structure- Frequencies 3.0 Hz - 141 Hz
- Total mass (x = 92.7%, y = 92.5%, z = 93.1%)
" Three sticks are coupled at various locations - modesare coupled- ASB-3.2 Hz; SCV-5.5 Hz;CIS-13.3 Hz fixed base
* Relative mass distribution- ASB - 86%- CIS-11%- SCV-3%
24 ~ ~~
12
! Validation by Comparison of CLASSI-SASSI Methods
" CLASSlinco- Deterministic phasing
" CLASSlinco-SRSS- Structure response to each foundation input motion combined by
SRSS" SASSI-Simulation (D. Ghiocel methodology)
- Spatial modes assigned random phasing- Mean of structural response to spatial modes computed
* SASSI-SRSS (F. Ostadan)- Structural responses to each spatial mode are combined by
SRSS" SASSI-AS (D. Ghiocel methodology)
- Linear combination (algebraic sum) of spatial modes used tocompute structural response
... .. ..... .•,--.... ~ ~~25 I- ',,..,,
Results for Comparison
* Calculate three directions of seismic input and combinethe results as in the seismic design/qualification process
• Compare response spectra as calculated
- Coherent (light blue)- CLASSlinco (dark blue)
- CLASSIinco-SRSS (green)- SASSI-SRSS (yellow)- SASSI-Simulation mean (black)
- SASSI-AS (red)• Response comparisons for each direction of seismic
input separately are also discussed
26 I PIPU:,
13
I Foundation Response Comparisons
" Translations at the center of the foundation
- Incoherent response significantly less than coherentresponse
- Responses calculated by all methods are in very goodagreement
" Rotations as measured by translations on the peripheryof the foundation- Little or no reduction due to incoherency compared to
coherent response- Responses calculated by all methods are in good
agreement
27 aI-12I
I Center of Foundation Response - X- DirectionCoherent and IncoherentAll Input Directions Combined
Fdn-x Incoherent respons due to combined Input
F1quo AN AM
21 -18, - SSlSS
12 • •SASl S~ulaJ Mea
4it .• r
2O
o9.
os-- 1
o; 1.
ospk
o28
14
I Center of Foundation Response - Z- DirectionCoherent and IncoherentAll Input Directions Combined
Fdn-o Inloherenl response due to combined input
14
13 NNWl RS
100 h0
o29
I Edge of Foundation Response YY - RotationCoherent and IncoherentAll Input Directions Combined
Fdn-yy Incohernt .ot.tVon .espo-- * 75 feet due to om.bined input
12
10 lE OI• S ,',
Do
801
10 iFnýz~y( H)
15
I Edge of Foundation Response ZZ - RotationCoherent and IncoherentAll Input Directions Combined
Fdn, inco~h.-. ot~tIon mson 75 k~t don, to ooInbined Input
1I :
I~I2I I31
I Top of Shield Building -Outrigger (Node118- 75 ft.)
" Incoherent response significantly less than coherentresponse- Horizontal - frequencies greater than 12 Hz up to 30
Hz (less reductions at ZPA)- Vertical - greater than 10 Hz- Outrigger reductions somewhat less than on
centerline• Responses calculated by all methods are in goodagreement - generally within 10%
" Small increases in incoherent response over coherentresponse at peak spectral frequencies less than 10 Hzare observed - induced rotations effects
16
I Top of Shield Building Outrigger (Node 118) - X- DirectionCoherent and IncoherentAll Input Directions Combined
Nod. 118-ASB -.. po-.. d-e to combi.ed input
50 -ftA 1-RS
ýRS-
.40,
XMs
0 0nn71
10o'F ý (.) H:
331-1 -• 1•• •,i.'
Top of Shield Building Outrigger (Node 118) - Z- DirectionCoherent and IncoherentAll Input Directions Combined
Nod. 118-MB- 0 response due to ofmbined input
14
13
12
11
10
I
I
C-I raI 2I :34
17
Top of Containment Internal Structure (CIS) -Outrigger (Node 229 - 75 ft.)
* Incoherent response significantly less than coherentresponse for frequencies greater than about 12 Hz -some reductions greater than 50%
" For this high frequency structure, responses calculatedby all methods are in very good agreement - generallywithin 10%
*~~En~35 r-b, r ef 35i
I Top of CIS Outrigger (Node 229) - X- DirectionCoherent and IncoherentAll Input Directions Combined
Node 229,CIS x rspons du. to ýooblnad input
14 C• 1
1(04
2ýW
^3
18
I Top of CIS Outrigger (Node 229) - Y- DirectionCoherent and IncoherentAll Input Directions Combined
Node 229-CIS y respoase due to on ••,bId input
40 'I i 3
WtS C•mSRSS 71
337
Top of CIS Outrigger (Node 229) - Z- DirectionCoherent and IncoherentAll Input Directions Combined
Node 229-CIS z response due to 0omrbinod input
ýsAl Sa•o s~
100
FII- )
38*
19
I Top of Steel Containment Vessel (SCV) -Outrigger (Node 145 - 65 ft.)
• Incoherent response significantly less than coherentresponse- Horizontal - frequencies greater than 12 Hz (less
reductions at ZPA)- Vertical - very significant reductions for frequencies
greater than 10 Hz
" Responses calculated by all methods are in very goodagreement - generally within 10%
- M9~639 ar-rim
I Top of SCV Centerline (Node 145) - X- DirectionCoherent and IncoherentAll Input Directions Combined
Nod. 145-ICy x rmponim due to ombined input
;,5
-C.Sun~r-SýS
SO5 SASSI-SRS
,
15ý
5.0
205
o100F.1O-" (ft)
Ant.., nv ... 40 f 2 I
20
I Top of SCV Centerline (Node 145) - Z - DirectionCoherent and IncoherentAll Input Directions Combined
Node 145-SCV z response due to co-bined Input
14
F• qc1ya(0t
SM UO
10 1Fst
41f-IaI Z
I Individual Excitation Directions Independently
" Shows differences of CLASSlinco & SASSI-AS results fromthat of other methods
* Containment internal structure (CIS) outrigger (Node229) vertical response (z) is an example
OW O.O~ 00O0000 0O~ ~42 ~ ~ 2
21
I o fCSOtigr(Nd 2)-Z ieto
T op of CIS Outrigger (Node 229) - Z- DirectionIncoherent - X Input
Node 229-ClS response due to x input
10 -CLOSSI~nU,-R
9 SASSl-S, SRSC_SASS: S-umonea
-- 8 m SASSI-AS
G,.
3,3
I Top of CIS Outrigger (Node 229) - Z- DirectionIncoherent - Y Input
Node 229-CIS z -pone due to y input
10
I
j3
10Fm Y(141)
,w
44ra II 2I EI -l
22
I Top of CIS Outrigger (Node 229) - Z- DirectionIncoherent - Z Input
Node Z29-CIS. Z Mponoo due to z input
91-C SJ -SRSSwSA l-SP.$
'0 4
45=F-,-
I Top of CIS Outrigger (Node 229) - Z- DirectionCoherent and IncoherentAll Input Directions Combined
Node 229-CIS z response due to combinod input
FS=00 o.'o M-25 ýS&S-AS
ii
Is
F -mu~ (W•)
23
I Comparison with Published Results
Mita & Luco, 1986 X3
- 40 m high, 10 m radiuscylindrical structurerepresented by stickmodel X2
- Foundation has 10 m H
radius on 400 m/shalfspace
- Transfer functions &H
evaluated at base centerand edge and top centerand edge
OmS~,tI'fld, St~. W~ Mfl47 ~f I S*~ltJ
I Base Center Horizontal Response toHorizontal Input
It
I~I2I48
24
Base Center Vertical Response to VerticalInput
0 MS.~~PR.d. tnh. Mt,~0 AAA~ 49
I Top Edge Vertical Response to VerticalInput
----a. -'. i ------4 -
ta.P. ~
~ Tnffl
I~ff2I I5'50
25