ML20005B921
| ML20005B921 | |
| Person / Time | |
|---|---|
| Site: | Wolf Creek, Callaway  |
| Issue date: | 09/11/1981 |
| From: | Petrick N STANDARDIZED NUCLEAR UNIT POWER PLANT SYSTEM |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| SLNRC-81-99, NUDOCS 8109160121 | |
| Download: ML20005B921 (11) | |
Text
._.
c Q[ D 8
9 4
\\?.
SNUPPs SEP 1519815 h i
k#
Io*"fo*'s%*" "
5 choke chay Road 4
M Nicholas A. Petrick gled 20e50
\\
Executive Director v
September 11, 1981 SLilRC 81-99 FILE: 0278 SUBJ: Cable Tray and Conduit Support
/
. Harold R. Denton, Director office of Huclear Reactor Regulation U.S. Nuclear Regulatory Conmission Washington, D. C.
20555 Docket Hos. STl1 50-482, STN S0-483, and STN 50-486
Dear Mr. Denton:
In discussions with Dr. Gordun Edison, llRC project manager for the Sl4UPPS applications, it was learned that the NRC Staff required additional in-formation in order to complete their review in the subject area. Tne enclosure to this letter provides the information in the form of responses to questions that were prov'ded by Dr. Edison.
Very truly yours, t
4'Y C(\\Q 14icholas A. Petrick RLS/jdk Enclosure cc:
J. K. Bryan UE G. L. Koester KGE
- u. T. McPhee KCPL
pE 6i\\
l l
8109160121 010911 PDR ADOCK 05000482
SMUPPS Q220.2 Provide a discussion on how major cable tray test results were used in arriving at the 20% modal damping.
The discussion should assure consistency of observed data and calculations used.
RESPONSE
In a linear dynamic analysis velocity dependant forces (i.e.
viscous damping) are introduced to account for various mechanisms of energy dissipation.
These mechanisms include such things as: friction and slip-in bolted connections, hysterisis, radiation of energy away from a foundation, the effects of fluids, and no doubt other mechanisms as well.
Since these various mechanisms cannot be accounted for explicitly in a linear analysis, their effect is lumped in a single viscous damping.
Dynamic testing is used to determine an effective viscous damping, appropriate for seismic response.
This procedure is common to all structural dynamic analysis.
During the cable tray and conduit raceway test program, the random vibration of cables was identified as one of the signif-icant energy dissipating mechanisms.
This occurred because the cables represent most of the mass of the system, are able to move relative to each other, and were not rigidly attached to the supporting tray.
During the tests, this phenomenon mani-fested itself as a noticeable relative movemer.t and impact of the cables within the tray.
As is the case with other energy dissipating mechanisms, this effect was quantified in terms of an equivalent viscous damping based upon the relationship between the recorded response and the applied input to,each test specimen.
The test report entitled " Cable Tray and Conduit Raceway Seismic Test Program" provides a detailed discussion of the methods used to compute an equivalent viscous damping from the recorded results of the dynamic tests.
This discussion can be found in Section 5 with supplementary infor-mation in Appendices G, H,
and I.
The computed damping values from the various tests are tabu-lated in Appendix K of the test report.
Data was taken from l
these tables and plotted as shown in Figure 220.2-1.
On this figure, the data points of computed equivalent viscous damping i
are plotted as a function of input acceleration (floor spectrum ZPA) for over 100 tests of various braced strut hanger tray systems.
These results represent all the data from simulated earthquake inputs.
Low level sinusoidal and snap back test data are not included, since they are not directly applicable.
Since these tests represented a wide variety of tray type, connection details, struts, and cable configuration, there is a broad scatter in the data.
These data, however, do clearly show that the recorded responses of the tested tray systems are best described by a dynamic system with an equivalent viscous damping.
It should be noted that the data realistically can be Rev. 7 220.2-1 9/81
-/
SNUPPS utilized with accepted curve fitting techniques to obtain a "best-fit" curve which reflects the statistical average of the test data.
Such an approach would result in a maximum damping value far in excess of the. conservative 20 percent value.
However, in the interest of conservatism, a bilinear curve, which effectively bounds the lower end of nearly all the points, was utilized.
This curve is given in Figure ILD.2-l.
This curve represents the recommended design values of equiva-lent viscous damping.
In addition to the determination of equivalent viscous damping, as described in the test report, linear analysis was performed on finite element models of several of the ' tray system test setups.
These analyses confirmed that a very high viscous damping was required in order to predict responses similar to those recorded during the dynamic testing.
These analyses confirmed that the application of the damping values recom-mended for design in a linear analysir was consistent with the results of the test program and, therefore, would result in a conservative design of support systems.
i 1
I l
Rev. 7 220.2-2 9/81 a
e 52 g
g o
o 48 c
b 44 o
o o
a>
^
40 0
o E
36 8
E o
o e
o 3
o o
o o
g*
c 32 e c
.J o -
4 0
9 o" 0 o
o t 28
=
- =
5 a,
o o
a
=
3 4
o oo o,
o o
g oo c o
'h*o oo ou o
1 0
o o
a 20 o
C 16 oo 12 o
t o 8
oof 7'/o o o 4
_ e e
og o
O.1 0.2 0.3 0.4 0.5 0.4 0.7 0.8 0.9 1.0 1.1 1.E 1.3 INPUT (g) BRACED MANGERS Fu. z2c.2-I d
DAMPING VS. INPUT LEVEL FOR BRACED HANGER SYSTEMS l
l l
SNUPPS Q220.3 Why was cable tray test input loading applied at a 45* angle instead of simultaneous horizontal and vertical load input?
What are the implications of this testing method upon the validity of the recommended 20% damping (e.g., with respect to statistical independency requirements of different directional inputs)?
RESPONSE
The cable tray and conduit raceway test input loading was applied at 45-degree (vector biaxial) because the shake table used was limited to vector biaxial motion.
In choosing the 45-degree relationship (i.e., horizontal equals vertical), the floor response spectra of many containments and auxiliary buildings were reviewed, and this equality of horizontal and vertical motion was deemed most appropriate.
IEEE-344 and NRC regulatory guides recommend, but do not require, independent biaxial input.
In the case of raceways, the modes of vibration are symmetrical and are dominantly either horizontal or vertical and so would be adequately excited by vector biaxial motion.
As the different modes of a given raceway generally have quite distinct resonant frequen-cien, there is no problem introduced by the zero phase between horizontal and vertical loading (i.e., vertical and horizontal responses will be randomly varying in and out of phase even though the vertical and horizontal inputs are in phase).
Independent biaxial input is preferred in nonsymmetrical cases and in the possible but unusual case of testing a structure with a mode whose axis of sensitivity would be at 90 degrees to the vector biaxial input, and hence not excited.
The raceways are simple structure systems with distinct vertical, transverse, and longitudinal modes; this was confirmed during testing.
Therefore, the test results are not affected by the use of vector biaxial input.
As described above, widely spaced modes of vibration with little cross coupling were observed during the testing.
For
~
example, longitudinal swaying modes were quite low (1.8 Hz),
transverse modes followed (3.2 Hz) with tray modes following at 6.1 and 15 Hz for a typical 4'6" single tier unbraced raceway.
This data it illustrated in Figures 7.8 and 7.13 of Volume 1*
for a 100-percent cable loaded raceway of 0.10 g peak response.
Similar frequency ratios for longer strut hung raceways are illustrated in relevant data.
The purpose of the cable tray test program was essentially to verify the mathematical model ured in the analysis, not to seismically qualify the raceway systems by testing only.
4 (.4 kut eeg "cabt %g.a aa%qsmu
- w Te p ')
Rev. 7 220.3-1 9/81 J
SNUPPS l
l Q220.4 Will sprayed-on fireproofing affect cable friction and thus the damping ratios?
RESPONSE
Yes.
However, sprayed-on fire proofing is not utilized on caws, Rev. 7 220.4-1 9/81 t
1
SNUPPS Q220.5 The cable tray test conditions do not reflect the y
actual physical site situation.
Provide the rationale for extending the test results to the actual design which is different trom the test configuration.
RESPONSE
The test fixture used to test cable trays was specifically designed for this test program.
Its inverted pendulum design permitted seismic input to suspended tray support systems.
Additionally, the fixture was designed to accommodate a 40-foot-long tray system segment of up to five tiers and a hanger of up to 13 feet in length.
Sufficient width was provided in the test bay to accommodate two parallel runs, including cross connections and attached conduit.
This facility allowed for testing of long, multitiered tray systems with various bracing arrangements.
The test program included tests of a large number of varied tray types and support types in various configurations.
These test configurations were used during the testing program in order to simulate the actual field installed conditions.
Supports with or without bracing and with multitier cable trays were tested.
In addition, a combined system configuration comprised of various tray fittings such as tees, elbow, ver-tical bend, and multitiers of straight cable trev runs was tested.
In view of the scope of testing and the various test setups, it was concluded that these tests do actually simulate conditions encountered in the field and, therefore, the results of the testing would be applicable to the design of cab,le trays on the SNUPPS project.
I i
e J
Rev. 7 220.5-1 9/81
SMUPPS
'i U
j Q220.6 Specify different conditions under which different j
modal damping ratios ranging from 7-20% are used.
1 (cable tray)
I i
RESPONSE
t
)
Damping of the cable tray system is dependent on the amount of cable in the trays and the input amplitude of vibration.
Figure 220.6-1 presents the lower bound values of equivalent viscous damping as a function of input floor response spectrum ZPA and amount of cable in tray.
To be able to use the maximum value of damping, 20 percent, the instructure response spectra i
must have at least a ZPA value of 0.35 g and the tray must be at least 50 percent full by weight of cable.
e i
i 4
e i
l f
i Rev. 7 j
220.6-1 9/81 3
a
-50% TO FULLY LOADED TRAY D
[
]
i 20 A
/
M 16
/
P 12 g
-UNLOADED TRAY &
-.L N
8,,,
r- - _
y CONDUlT G
4
(%)
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 ^
1.0 INPUT FLOOR SPECTRUM ZPA Fis, 220. 6-t LOWER BOUND DAMPlNG AS A FUNCTION OF INPUT ZPA i
l f
e#
l 1
1
i SMUPPS Q220.7 It appears that the scope of the cable tray test and the number of tests may not support direct extension to SNUPPS (the appropriate project) cable tray design.
Justify that the scope of test conducted is adequate for direct design applica-tion.
a RTjSPONSE The ccope of the cable tray and conduit raceway test program included the ev61uation of a large number of variables in the design of ca!;1e trays.
Included in the test report are discus-sions of the following variables:
o Type of tray o
Type and length of hanger o
Location of splices o
Number of tiers o
Trapeze and cantilever support o
connection details, such as single clip angle double clip angle guesseted clip tray to strut type hanger o
Type and location of bracing o
Amount of cable fill o
Size and distribution of cables o
Cable ties o
Combined conduit and tray systems o
sprayed fire protection material
\\
In order to evaluate the effects of these and other variables, over 2,000 individual dynamic vibration tests were performed over a period of 11 months of testing.
As a result of these tests, over 50 volumes of raw data were generated and evaluated.
The results of the evaluation of these data form the basis for the conclusion contained in the test report and the design recommendations implemented in the SNUPPS design.
Rev. 7 220.7-1 9/81
SNUPPS In addition to the wide range of variables that were evaluated, tests were performed on tray and strut systems similar to the SNUPPS design.
As a result of the evaluation of the variables described above and the testing of hardware and support configurations similar to the SNU2PS design, a set of design recommendations was formulated.
These recommendations were developed to be generally applicable to a wide variety of hardware and specifically applicable to the support configurations used by this project and the other test program participants.
For example, the recommended damping in intermittently braced strut supported trapeze hanger syctems was determined from the data of over 100 dynamic tests on these type of systems.
Figure 220.2-1 shows the recommended damping as a function of floor accelera-tion in the form of a bilinear curve.
As can be seen from this curve, the recommended damping, for the most part, represents a lower bound of all the data obtained from the test program.
Similar conservative recommendations were formulated from the results of the test program for other aspects of design.
Consequently, it is concluded that the design recommendations formulated as a result of the cable tray and conduit raceway test program are broadly applicable to the design of strut supported raceway systems and were conservatively applied in the design of the SNUPPS raceway supports.
O Rev. 7 220.7-2 9/81
SNUPPS Q220.8 Justify the use of 7% critical damping for conduit supports for all seismic input levels.
l.
RESPONSE
During the cable tray and conduit raceway seismic test program, 1
various tests were performed on conduit runs on a trapeze raceway to determina their dynamic characteristics.
A large number of variables were considered in this test program.
The description and results of conduit raceway testing can be found in Section 8 of the test report.
The critical damping value computed from test data is 7 percent at 0.1 g input acceleration.
Higher damping value trend was observed for input acceleration higher than 0.1 g.
But at present time, for design of conduit raceway system it is recommended to use 7 percent critical damping for all levels of I
input acceleration at and above 0.1 g.
For lower input accel-eration, it is recommended to use linear interpolation from 7 percent to O percent damping for 0.1 g input to zero input acceleration.
i i
1 I
i l
Rev. 7 l
220.8-1
-9/81
%