ML20099C212
| ML20099C212 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 03/04/1985 |
| From: | Hukill Hj D GENERAL PUBLIC UTILITIES CORP. |
| To: | Stolz J Office of Nuclear Reactor Regulation |
| References | |
| 5211-85-2045, NUDOCS 8503110257 | |
| Download: ML20099C212 (17) | |
Text
'
?
GPU Nuclear Corporation Nuclear
=, ors:=48o Middletown, Pennsylvania 17057 0191 717 944 7621 TELEX 84-2386 Writer's Direct Dial Number:
March 4, 1985 5211-85-2045 Office of Nuclear Reactor Regulation Attn:
J. F. Stolz, Chief Operating Reactors Branch No. 4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D. C.
20555
Dear Mr. Stolz:
Three Mile Island Nuclear Station Unit 1 (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Damping Values for Conduit / Cable Tray Supports Recently a study of the seismic performance capability of Class 1E cable tray and conduit raceway systems was performed for the Systematic Evaluation Program (SEP) Owners Group of which GPUN is a member. The results of this study were published in a document prepared by URS/ John A. Blume and Associates, entitled " Seismic Evaluation of Electrical Raceway Systems". One of the highlights of the study was the investigation into the high damping effects of raceways, including their source and trends. Additionally, this report presented reconnended damping levels to be used for analysis of raceway systems.
Figure 2.15 of the attached paragraph 2.5 of the report indicates that a minimum 7% equipment damping is reconnended for conduit and unloaded cable tray raceways with multiple supports.
Section 5.2.1.2.11 of the Updated FSAR currently indicates that for assemblies and structures which are bolted or riveted, the percent of critical damping is 2.5.
For welded structures and assemblies the percent of critical damping is 1.0.
In light of this recent report, Regulatory Guide 1.61 and IEEE Standard 344-1975 (both of which indicate OBE and SSE damping values of 2 and 4% for welded steel structures and 4 and 7% for bolted structures, respectively) the FSAR values appear to be needlessly conservative. Therefore, GPUN intends to apply floor response spectra curves for OBE and SSE values with 2 and 4%
welded steel structure damping and 4 and 7% bolted steel structure damping respectively for conduit, conduit supports, cable trays and cable tray 85031 % $$$$b289 PDR PDR d y P
GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation
I t
5211-85-2045 March 4, 1985 supports.
Further, GPUN has performed a 10CFR50.59 evaluation of these damping values and determined that there are no unreviewed safety questions or Tech. Spec. changes required. GPUN plans to change the FSAR to indicate these values in the above referenced section for the 1986 update.
Sincerely, H. D.
u ill Director TMI-1 HDH/lr:1481f cc:
R. Conte J. Van Vliet 0185A l
[
I
j s
~
URS
~
ANALYTICAL TECHNIQUES, MODELS, AND SEISMIC EVALUATION OF ELECTRICAL RACEWAY SYSTEMS prepared for The SEP Owners Group Underthe Duecbon d t3AC, incorporated
% V 9 L i,D.C.
l pepared by URS/ John A. Blume
& Associates. Engineers 130 SyNan Street Darwers, Mass. 01923 l
1 e
-.--e-e.,
w
-,r y
--,-.,,,p.
yw
e plants' rod-hung raceways Is 12 in.
Allowing for the depth of the cross-member and the thickness. of the connecting nuts. the typical Intertler rod length is about 10 in. The lower dotted line in the figures corresponds to an Lg of 10 In. for the above parameter values.
For raceways with no stiffness (no end restraints and A = 0), the figures show (Intersection of dotted line 3
with solid line') that Intertler displacement can be ignored for a two-tier l
system with I greater than 19 In.
for a three-tier system with I g
g greater than 26 in.. and for a four-tier system with I greater than 35 In.
If the g
raceway system is braced (2 - > 0). the region of appilcability of the 80EF i
3 model expands. For example, for K = 1. the minimism acceptable value of I g
g decreases to 18 in. for a two-tier system. 22 In. for a three-tier system, and 25 In. for a four-tier system. Also plotted in the figures Is the line cor-responding to Lg = 20 In. As can be seen. Increasing Lg significantly reduces the region of appilcability of the assumption.
From Figure 1.1 It can be seen that 75% of the hangers surveyed have a height of 2 ft or more ( 4 I 22_ In. ).
For a three-tier systen (78% of the hangers 3
have no more than three 'lers) with a 50-ft-long end-restralned span with t
ladder trays ( GA = 20.000 lb).
'g= 0.88.
Using M gure 2.10. t k d w e parameter values, and the typical L value of 10 in.. one will find that such g
j systems can be adequately evaluated while ignoring the Intertler displacement.
i l
Although Figures 2 9 through 2.11 apply to a specific range of parameter values, they show that th's assumption is appropriate for a significant amount of realistic raceway geometries, as characterized by the hanger statistics accumulated during the plant vlsit.s and summarized in Figure 1.1.
l l
25 cas,Ing of naceway systems 1
naceways. particularly cabletray raceways, differ from typical structures in that the mass of the structure (cables) Is not Integral with the stiffness i
elements (hangars and trays). As a result, under dynamic loading there Is a l
relative displacement between the cables and the tray and hangers. That relative displacement is.not accounted for in the asstanptions under whIch the q
equations for frequency 'and mode shape were derived. However, because these ll equations have yleided analytical frequencies and mode shapes that match the i
test results, the effecti, of that relative displacement can be accounted for by
- i 4 l-As L--
6-_-
.o
o the daging value assumed for the dynamic model. The daging value will both represent >aJ'true4 damping phenomenonged appeoxjmete the,mo,r,e co%g,,
This dynamics lay 11ed by the relative displacement between cable and tray.
section discusses the damping behavior of raceway systems as developed from shaking-table testing of representative raceway systems.
The amount of apparent damping In the system is expected to be influenced by the relative motion between the cables _and,the tr,ay,, and the l
two factors:
l amount of cables in the tray (cable f{. The amount of relative motion is assumed to be a function of the level of lateral acceleration of the raceway i
That level of acceleration is quantified.as the average peak response system.
acceleration (A ) d the meway specimen as defined by Equation 2.20.
As an 3
The test example, consider shaking-table Test 46, presented in Reference 1.
specimen was a two-tier, rod-hung tray system with cable fill of So ib/f t/ tray and fixed at both ends. Analysts of the test results showed single-mode response with a frequency of 2.4 Hz, a sinusoldal mode shape, and a 17-in.
peak displacement at the center of the span. When the mode shape Is normalized to unit peak displacement, the average peak response acceleration
.I s t (2.20 f
a,,,
wy (z Mz A
=
g sin [dz 0.649
=
=
Rod-Supported Systems. The damping (d), as determined in Reference 1. Is plotted against the average peak response acceleration (A ) In N gure 2.12 for g
all rod-supported tray specimens tested, except those in Tests 38-40 of Phase I and Tests 85-89 and 95-108 of Phase 11.
The plotting symbols denote its the amount of cable fill and whether the test specimen was restrained at end.
. l Several facts are apparent from Figure 2.12.
First, unrestrained specimens are less damped than restralned specimens responding at the same level of This difference can be attributed to the tray dsformation acceleration.
N..
..m_
z._.*
..s.w n -
w
...a
-, -, - ~ ~
- n..
e e
o e
C'
.)
b i
79 G
r 3
O
.5 W W W g 5 5 5 5 3 5 5 5 5e.sead555535WWWgWvvygvW55g5WWW W W W W g 5 5 5 5" WWWWgWWWW WWWUg5555
'W U WW W WW WWWW WWW WWW cama est.menusse 1
k O
O
& a et h/t cam 8 rLL.NestaasES e e W h/E CAmE rLL.M eIRAAES e
e
.,,,,,.,gg,,,,,,,,,,
c oe somm cams,esk.insuomess,e 7
.se a
3 e
a ee a
c ese N
Je mi a
5 I
a a
E P
e s aanroe asst ups e
2-d e
e e
s' 5
ese a
if 3
1
)
3 a
e m
a-i e
e e
l D
T i
r l
e l
5 f
T I
l
(
l
- ...l........l........t........l........l........l........t........t........l........l..;
e i
e e.
es es e.4 as ee ar e.e es to i
(
MERAGE PEAK RESPONDE ACCELmm0N, As(9)
J i
s FIGURE 2.12 BEHAVIOR OF THE APPARENT SYSTEM DAMPING AS A FUNCTION OF THE AVERAGE PEAK RESPONSE ACCELERATION FOR ROO-SUPPORTED CA8LETRAY SYSTEMS
Induced by restraining the specimen, which results in higher apparent system damping. Figure 2.8 shows that for rod-hung raceways the distance between lateral supports has to exceed 75 ft before even a system with hangers as short as 2 ft approaches a true unrestrained response. Such long lengths of raceway without lateral restraint from well brackets, tray risers, or wall j
Therefore, a damping curve developed from
. penetrations do not commonly occur.
the restrained-specimen results (damping curve 1 In Figure 2.12) is more applicable to realistic systems than the lower-bound envelope of all the data i
points in the figure.
I The second conclusion that can be drawn from Figur,e 2.12 is that the damping does Indeed increase with increased A. This increase Is particularly 3Note that A incorporates both the level pronounced for restrained specimens.
g of input motion and the dynamic properties of the raceway specimen.
Therefore, a specimen with a frequency In a low-aspilfication region of the input spectrum subjected to a hlgh-level input motion could have a lower A,
3 and thus a lower level of damping, than a specimen subjected to a lower level of input motion but with a frequency in the peak ampitfication region of the input spectrum.
Finally, Figure 2.12 provides insight into the relationship between danpIng and the level of cable fill. The amount of cable fill affects the damping In An empty tray has a low level of damping. As the cable fill two ways.
increases from aero, the damping level Increases as the sliding of the cables within the tray and the cable collisions with both the tray and other cables After a certain point, increasing the cable fill dampen the dynamic response.
will serve to restrict the motion of the cables already In the tray and the l
In the extreme case, a tray packed with cable to damping level will decrease.
the point that sliding of the cable is Inyossible, a low damping level should These effects are reflected in the test data summarized in result.
Those specimens tested with cable fill of 25 lb/ft showed higher Figure 2.12.
damping than the 50-1b/ft specimens responding at the same acceleration level, thus Indicating that when the cable fill was doubled, the decrease in damping due to increased restriction on the cables' movement dominated the increase In A Ilmited number of tests dancing due to the increase 'In the amount of cable.
l were conducted with a light fill of 10 lb/ft. The desping of these specimens j
fell between the levels for the 25-1b/f t specimens and the 50-Ib/f t specimens GJfE/Blume l - -
- ~ - - - + ~, - - - _ _ _ _ _ _ _ _ _ _ _, _ _ _ ___ _
i responding at the same acceleration level. This Indicates that in decreasing YM11982I 8/Eto T0"1b'[f t. tWMeas'iiIg"effeet' on dadping o'f' ~
decreasing the mater of cables outwelghs the Increasing effect of the cables' l
greater freedom c# novement. Damping curve 1 can thus be used for the evaluation of restralned, rod-supported cabletray raceways with cable fill of 10 lb/ft or more., A nominal damping level of 5% can be conservatively used for lightly leads'(i (l'ess than 10 lb/ft) rod-supported cabletray raceways.
At some of the participating plants a number of cabletrays have_been sprayed with a fire-retardant asterial. After curing, this meterial encases the cables, resulting in a solid mass inside the tray. As described in Reference 1. shaking-table tests were conducted to Investigate the effect of the fire-retardant meterial on demping. The results are summarized in Figure 2.13 As would be expected from the discussion above, the fire retardant reduced the damping levels by significantly reducing the movement of the cables within the tray. As a result, desping curve 2 was developed for the evaluation.of rod-supported cabletray raceways with sprayed-on fire i
retardant. Although only tests with cable fl11 of 25 lb/ft were conducted. It can be concluded that damping curve 2 is appropriate for the analysis of raceways with cable fill of 10 lb/ft or more and that a constant damping value of 53 Is appropriate for IIghtly toaded systems sprayed with fire retardant.
Shaking-table tests were also conducted usin2 a rod-supported conduit raceway specimen. These tests Indicated that rod-supported condult raceways are lightly desped and that a constant danping value of 5% Is adequate for i
realistic systems.
trut-Supported Systems. The damping. d. Is plotted against the average peak response acceleration. A. In Figure 2.1% for the strut-supported specimens g
tested in Phase Ill. Also shown in the figure is damping curve 1. which is based on the damping data for restrained rod-supported tray specimens.
Restraining the rod-supported tray specimens forced the trays to carry a significant part of the seismically Induced lateral loads, thus causing the trays to deform. As discussed in the section on desping of rod-supported specimens, the deformation of the trays resulted in increased damping levels. The strut-supported tray specimens used to obtain the desping values I
represented in Figure 2.1% were supported only by their hangers. without any
. t
f D
rug''
'88'l' mg--' m 'l --
'l'
81'l'l'33'I'3'I'3'l'l'
g A e as h/9 ram t FILL.MSTansso 9 FM MidMAlff
~
(
g
=
(A) e SS hfEt CASLE FILL.MSTRASIEo.% FK M1ANb400T ~.
[
A e 36 h/IICasLE FILL,WeIESTRASIEo.T FK MTalensif :
I
( Aj e 35 h/It CABLE FILL.ameraThesimo,% FK Miafmesef ~
g 6
.r l
(6 i
4,
/
/
/
mi
/
i r
O b
I OAaristG CURVE 809I a>
d
/
~
S
?
I s
b t,
=
=.
/
a f
a
/-
L
/
A
~
j b
b
[
DAerleeG CURWE tot 2 o.
=
m 1
1 Z i
.l..
....I....
....l....
....l....
l....
....l....
....l.
....l
....l....
...li. ;
i o
os 04 os ce e.o s:
3.4 s.s i.e 3.0 7
J AVERAGE PEAK RESPONSE ACCELERATION, As(9)
?
k J
FIGURE 2.13 SEHAVIOR OF THE APPARENT SYSTEM OAMPING AS A FUNCTION OF THE AWERAGE PEAK RESPONSE ACCELERATION FOR ROO-SUPPORTED CASLETRAY SYSTEMS WITH SPRAYED-ON FIRE-RETAR0 ANT MATERIAL
.... ~
2[...._..,,'.._,
., d o!.
t f
t f
,....................--.i.-.
....i....
....i -,
....i....
....i....
....i....
....i.. 3 63o
~
~
O e TitafE2E STsun A e casmLEwEst STsIUT.Ploco WemcaL
-~
O e caWitswast sisan.Ploot westTicat.- L mem e e caspasWst sisan. Peace vs:mcaL.Psono enact 2
A
.j l 7
G G
o.30
~
3 2
d II E
M D
O
~
w o
o y
K O
o.m
}
i o
^
o onwas cuar mi j
~
t
~
)
i
...i....
....i....
....i....
....i....
....i....
....i....
....i....
....i........i...;
o o
oa o..
os o.e Ie l
(
AVERAGE PEAK RESPONSE ACCELERATIONS As(g)
J i
FIGURE 2.n BEHAVIOR OF THE APPARENT SYSTEN DANPiNG AS A FIRICTION OF 1NE AVERAGE PEAK RESPONSE ACCELt. RATION FOR STRUT-SUPPORTED CASLETRAY SYSTEMS
d end restraints. No appreciable tray deformation occurred during the shaking-table tests of these unrestrained specimens.
As with rod systems, In actual fleid situations such unrestrained strut systems are virtus11y nonexistent. Lateral restraints are usually provided at penetration points through walls by hangers braced directly to the floor, a wa11, or a structural columns and sometimes by a change of direction of the raceway system itself. These lateral restralnts would result In deformation of the tray. Damping levels similar to those observed for end-restrained rod-supported systems (daging curve 1) would then be expected.
In effect, extensive damping data for systems supported by heavily braced strut hangers were developed by ANCO Engineers and Bechtel Power Corporation In their raceway testing program.2 Because of the bracing, considerable tray deformation was observed In that testing program. After analysis of their test results.3 Sechtel recommended the curve shown In Figure 2.15, which has been found to correlate well with damping curve 1.
l The Sechtel curve relates damping to the zero period acceleration (ZPA) of the input spectrum rather than to the average peak response acceleration, A, of g
the raceway. The reason is that in the 8echtel program the test motion was tuned to the fundamental frequency of the particular specimen being tested.
As a result, each test specimen's fundamental frequency fell within the peak-ag ilfication region of the test spectrum. An approxiniste translation of the 8echtel curve from damping versus 2PA to damping versus A has been g
accomplished as described below.
The Bechtel damping curve can be represented by:
d. ( 0 562PA 2PA 4 0 36g (2.21) 0.20 2PA > 0 36g where:
d=
fraction of critical daging l
zero-period acceleration (g) l 2PA
=
l l
N,
i
e e
e e
sf 4
I 4
9 t
5011'O FU.LY LGADED TNAY j
g 16
[
G
~
12
/
0 MY & @M g
,,8 g_
.a.._.._.._.. _ _
A 4 f 0.1 0.2 0.3 0.4 0.5 0.60.70.8 0.9 1.0 INPUT FLOOR SPECTRlM ZPA(g) 9 Source: Reference 3 L
J FIGURE 2.15 RECOMMENDED DAMPlNG FOR THE DE51GN OF ltACEWAY SYSTEMS t.
[
For purposes of this study, the ratio of the peak value of an acceleration i
response spectrum. S. to its ZPA for different damping values has been i
p approximated by:
1 3
h = (W)"
(2.22) l where the exponent is a characteristic of the Input motion's frequency con-tent.
If the Input motion were a pure sine wave, y would be equal to 1.
For l
l a more broad-band excitation, such as seismic motion, y Is a value less than l
1.
The test spectra achieved during the ANC0/Sechtel test program typically had a peak-to-ZPA retto between 3 0 and 4 5 for a 5% desping value.2 This Indicates a value of y between 0.48 and 0.65 The relationship between S and A Is a fum tlon of tk shape of the test p
g specimon's fundamental mode of response (see Equation 2.20).
$1nce ANCO's test motion was tuned to the fundamental frequency of the specimen. an approx-i Imation of this relationship that is both reasonable and uncomplicated Is, sim-ply that the two parameters are equal:
(
A (2.23)
S
=
p g
Combining Equations 2.21, 2.22. and 2.23, accomplishes the desired translation of the Sechtel danping curves r
1 I 0 56(2)F,,
A, f 0 36(0.@
M A
I (2.24) d
=
g ? 0 36(0.4)*
I,0.20 A
l Equation 2.24 Is plotted in Figure 2.16 for y values of 0.48. 0 57, and 0.65 (peak-to-ZPA ratios of 3 0, 3 75. and 4 5 for a damping value of 5%). Also plotted is damping curve 1.
As can be seen, desping curve 1 correlates well with Equation 2.24 especially for y = 0.61.
Thus, the relationship between response acceleration and danping can be assumed to be the same for both end-restrained rod-supported and end-restrained strut-supported systems and all i
l 1 QSSS/Blum8
O 4
l e
..,,e,,.
.... g ;u.
f c
.s 5
-.)
u g
3 I
f... g........ g........ ;+....
.... g....
.... g........ g........ g..., 3 g
i.
E_
q E
E E
N.
- 3 o
=
m O
w i
I
! 1 i
E N
s i
s l
E N
]
N'
(
\\
I 3
o I
4 am W
sN o
W L
4 3
N E
s s
N E
~
6 N
m i
e
. e
~
s m
U e
3
- .w 9
y
- 3 4
7 e
=
- 3 I
g t'
j.
5 g
8 h
seenleien amen lenne annelsene seenInsas einilamen neseleman ananlanar.
g 3
o O
o p
11 u
'j J
e k
M t
W k
.mo
+
~
s j
effects discussed for rod-supported systems are also appilcable to strut-
]j supported systems, including the cable fill and sprayed-on fire retardants.
This is to be expected, since the sources of apparent system damping are not affected by the type of support used. The resulting damping curves for the selsmic evaluation of racewey systems are sisenartzed in Figure 2.17 l
I r
.............................................i........,................i...........A a= :
3 5
- = :
cas.sva m aa m use ume > a em cas.2 ru 4
~
d i
WWWW-Ge PW IETadmelff
=
Af8 > a em cas.2 PLL.
em i-
/
-i M:
I ' " ' ' ' ' I ' " '. ' " ' I ' ' ' '
' ' ' I ' " '. *' ' ' I ' ' ' '
' ' ' I ' ' '
" ' I ' " ' " ' I " ' ' ' ' ' I ' ' " ' ' I ' ' '
0 o
e as sa is e
m u ac.c,c= neseo e - = r a.
cg>
y FIGURE 2.17 8EHAV10R OF THE APPARENT RACEVAY $YSTEM OAMPING AS A FUNCTION OF THE AVERAGE PEAK RESPONSE ACCELERATION
. gg m>m-
-=
s REFERENCES 1.
URS/ John A. Slume s Assoelates Engineers. Shaking-M ble festing for Seismie fustustion of FEsotrimt Anosusy Systems, San Francisco, Call-fornia, April 1983.
2.
ANCO Engineers, ino., Cable fmy and Con &it Anosusy Seismio feet Pro-gms, Release 4 (Final). Norwalk, Californla, Oecoder 1978.
3 Reimer, G.
S., and P. W. Koss. Development of Analysis and Design foo4 niques freur Dynamie festing of Flectriant 4:esumy Support Systems, Re-vision 5. Teehnical Report for cable Trey and Conduit Raseway Test Pro-gram, seehtel power Corporation, San Franelseo California, July 1979 4.
URS/ John A. 81uma s Assoelates Engineers NonlineBr Stmetumt Dynamie Amlysis Prosehree for Category I Sematures, NUREG/CR-0948, prepared for the U.S. Nuelear Regulatory Commission, July 1979 5
Neowen, R. s., et al., "Plastle Capaalty of Raceway supports - Expert-menta1 EvIdenee," Proceedings, Specialty Co<ennae, CiviI Engineertng and Mkalsar Power, Knoevitte, Tennessee, September 1900 6.
Haellton, C. W., and A. H. Hadjlan, "Plastle Capaalty of Raceway Supports
- Engineering Analysis," Proceedings, Specialty COVerenos, Citrit Engi-neering and Fuclear Power, Knoevitte, Tennessee, September 1980 7
Innovation Technology, ine., Dyrussia festing of cable fmye for Nope j
Cask Cenemeing Station, Units 1 and 2. Report No. 7706-1, Mount Holly, New Jersey, May 1978.
8.
P-W Industries, Inc., fasts for Physist Properties of Cabis fmye for Alvin F. Fogtle #ualesr Plant, Report No.1027. Revision 1. Cornwells Heights, Pennsylvania, September 1979 9.
. Qualification Report for Capnoity Eustuation - Ecu Seismic Zone, Toli"To. PWE-1G01, Cornwells Heights Pennsylvania, June,1975 10
, Qualifiantion Report for Capacity Fauluation - Medium Seismia W Job Mo. PWE-1002, f,ornwells Heights, Pennsylvania, July 1975 11.
Pochester Gas and Electric Corporation. Anchomge and Seismic Support of Safety-# stated FIsotriest squipment. Final Report, Project No. EWR-2831, Rochester, New York, Decoder 1980.
12.
Beehtel Power Corporation. Fiff Report: Drilled-In Espansion Botte Under Statia and Alternating Load, BR-5853-C4, san Franelseo, California Jan-uary 1975 13 Hanford Engineering Development 1.aboratory, Qualiffation of Ispansion Ancham, PA/SSE 203, Alchland, Washington, October 1977
- 145 -
t
s
- 14. Teledyne Engineering Services Stanzry Report, Generio Response to USNRC I & E Bulletin No. 79-02, Base Plate / Concrete Expansion Anchor Bolts. TR-3501-1, Revision 1. Waltham, Massachusetts, August 1979 15 Hi1ti Fastening Systems, Hitti Architsote and Engineare Anchor and Fas-tener Design Mxnual. Stamford, Connecticut.
16.
ITT Phillips Orill Division Red Head Anchoring Systeme Catalog, Michigan City, Indiana, 1980.
I 17 Raw 1 plug Company, Inc., azwt Nasonry Anchoring Handbook, Catalog No. 40, New Rochelle, New York, 1981.
- 18. Anamet Laboratories, Inc., Fmoture Enzmination of an HTHR050 U2-in.
Threaded Rod from a Cable Tmy Nanger, Berkeley, California, June 1982.
19 Detroit Testing Laboratory, Report on Uttintite Pull-out and Shear Resistexnce Teste Conducted on Concrete Inserts and Anchors, Reports 312177-0, 312178-0, 312179-D, Ook Park, Michigan, September 1974.
e
{
"^
b
- 146 -
- - - -