ML20207E382

Topical Rept Evaluation of SL-5159(P), Methodology & Verification of Gapp Program for Analysis of Piping Systems with E-Bar Supports. Staff Finds Topical Rept Acceptable for Referencing in Licensing Applications

ML20207E382
Person / Time
Issue date:	03/02/1999
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20207E377	List: ML20207E371 ML20207E382
References
NUDOCS 9903100251
Download: ML20207E382 (7)

Text

!

\

OFFICE OF NUCLEAR REACTOR REGULATION MECHANICAL ENGINEERING BRANCH DIVISION OF ENGINEERING TOPICAL REPORT EVALUATION l

TITLE: Report SL-5159 (P)," Methodology and Verification of the GAPP Program j for Analysis of Pip.ng Systems with E-BAR Supports"  ;

ORGANIZATION: Sargent and Lundy L.L.C. (S&L) and LISEGA, Inc. I

1.0 BACKGROUND

S&L and the German organization LISEGA, Inc., have developed proprietary, patented, piping

! support devices known as E-BARS. These devices are proposed as replacements for snubbers, to be used for restraining piping systems during seismic or other transient events.

l E-EARS have creset gaps that permit free thermal expansion under normal operation. Under j sufficiently large displacements experienced during abnormal dynamic conditions, the gaps  ;

close and the E-BARS act as elastic or elastic-plastic axial restraints. The load-deformation l characteristics of these devices are therefore nonlinear in nature. l The nonlinear analysis of piping systems with E-BARS is considered impractical for production design since it is highly complex and requires experienced analysts. S&L and LISEGA have l therefore developed the proprietary computer program GAPP for the static and dynamic  !

analysis of piping systems containing E-BARS. The GAPP program calculates equivalent linear l elastic load-deformation characteristics of the E-BARS, thus permitting the application of l standard linear-elastic dynamic piping analysis techniques under all ASME Section ill, Service I Levels B, C and D loading conditions.

2.0 EVALUATION An E-BAR is equivalent to a pin-connected linear support with large gaps. During a seismic event or other abnormal transient condition, if the local dynamic displacement exceeds the gap i distance, the gap closes and the support acts as a pipe restraint. The device is designed to '

absorb the impact energy through plastic deformation of the outer metal housing of the device, thus limiting the pipe displacements and stresses to acceptable values.

2.1 E-BAR Construction The E-BAR consists of an outer tube with slightly flared ends, and two wedge-shaped thick discs, mounted on an inner shaft which slides within the tubular body. The outer tube is attached to the pipe and the inner shaft is attached to the structural support, or vice-versa. The shaft has a fixed and an adjustable stops. The distance between the stops is adjustable to any l prescribed gap. When the gap is closed under the relative motion between the pipe and the  !

supporting structure, one of the stops presses against the corresponding disc. The tube I deforms e!astically by the wedging action of the disc into the flared end of tne tube, until the disc expands the tube and the tube yields at a certain axial load, thus deforming plastically and Enclosure

]

9903100251 990302 PDR TOPRP EECSALE C PDR

o 1 i l I

I absorbing the energy of the irnpact. This load is termed the working load, and remains constant during the plastic deformation. The load deformation curve of this stage corresponds to that of an elastic, perfectly plastic material. This is also the load acting on the pipe and the supporting structure. The inelastic deformation is limited by an inner stop on the housing, after which the device acts like an elastic strut untilits design capacity is reached. The maximum load is considerably larger than the working load. The complete load-deformation curve is therefore highly nonlinear, 2.1.1 E-BAR Design

. E-BARS are designed and fabricated in accordance with applicable provisions of ASME Section Ill, Subsection NF (Reference 1), and ASME Code Case N-420 (Reference 2), which is accepted by the staff in USNRC Regulatory Guide (RG) 1.84 (Reference 3). The material properties used in tiie design of the E-BAR components correspond to those of ASME Section 4

II,1989 Edition (Reference 4) and ASME Code Case N-240 (Reference 5), which is accepted by the staff in USNRC RG 1.85 (Reference 6).

Although LISEGA stated that the design of E-BARS is based on all editions of ASME Section 111, Subsection NF, since 1974, the staff recommends that the design of E BARS be based on the Winter 1985 Addendum to the 1983 Edition of ASME Section 111, Subsection NF, and later

editions. This Addendum specifies the rules of ASME Section lli Appendix F (Reference 7) for the design of component supports under Service Level D loading.

1 2.1.2 E-BAR Oualification l The topical report describes a representative qualification test report for an E BAR model,

, labeled Type 32. LISEGA performed static tests on the components (tube and cone, and tube and travel stop) of various sizes of this model, to determine their working loads (the load at -

which a tube deforms in the plastic range) and the safe maximum loads. For the size with the largest capacity (78600 lb.) the maximum safe load, determined as the load required to deform a built-in travel stop, was at least three times the working load.

. LISEGA also performed cyclic dynamic tests to evaluate the E-BAR performance under dynamic conditions. Two types of tests were performed:

• Displacement controlled endurance tests up to 2x10E6 cycles at 0.04 inch amplitude, to 2 determine the wear of the guide bore. Results showed no wear in the guide bore or change in diameter.

.

• Dynamic cyclic tests, consisting of low frequency tests (1-3 Hz) and higher frequency tests (> 10Hz) until the working load was reached. The purpose of these tests was to determine the load-displacement characteristics of the energy absorbers under dynamic conditions.

. Static overload tests were also performed to determine the ultimate staSc loads. These tests

, were performed on the units used for the endurance tests. The test loads corresponded to the

loads determined from analysis for the various ASME Service Levels. Service Level D corresponded to 1.72 times the working load. The tests were terminated when the test loads reached four times the working loads. No failures were reported under these loads.

2

i I r j The staff has evaluated the test methodology and the results of the tests, and finds the

methodology and the results reasonable and in accordance with current engineering practice, and therefore acceptable.

4 2.2 Program GAPP i The dynamic analysis of piping systems under all ASME Service Levelloading conditions is j crdinarily perfarmed on an elastic beh, using either response spectrum methodology or time-

} history enalysis. Both procedures are based on the modal superposition approach, which

requires that modai responses be combined by various approaches to obtain total responses.

Currently acceptable methods for combining modal responses under ASME Service Level

loading conditions are specified in USNRC RG 1.92 (Reference 8). S&L has specified a number of modal combination methods used in GAPP, including those currently specified in the j Regulatory Guide. The staff finds this acceptable.

2 With installed E-BAR devices, the behavior of a piping system is nonlinear under both static and l

j dynamic loads. S&L has adopted an equivalent linearization procedure found in the literature for 4 analyzing nonlinear structures (References 9 and 10), which permits the inclusion of gapped

, supports in a standard linear elastic analysis of piping under transient loading. This procedure

forms the basis for GAPP, applicable to problems where the conditions of linear elasticity and j small changes in geometry are approximately valid.

1 l GAPP has the capability of performing static analysis, uniform support acceleration response j spectrum analysis, and time history analysis under support acceleration or displacement time history inputs and force inputs. These analyses are based on modal superposition principles.

1 Time history analyses are based on a combination of modal superposition, the pseudo-force method (Reference 11) and numericalintegration of the dynarnic modal equations to determine I

the system displacements. Since the time history methodology is capable of performing analyses with independent support excitation, it should conform with the applicable conditions j for application stated on pp. 2-2 and 2-3 of Reference 12.

T( 'ccount for the nonlinear characteristics of the E-BARS, GAPP calculates an equivalent linear stiffness by an iterative technique, which depends on the displacements of the piping.

Similar procedures were previously evaluated by the staff and its contractor, the Brookhaven s

National Laboratory, in connection with the review of a similar device and a similar computer

program (References 13 and 14), and found to be acceptable. However, this similar device does not depend on plastic deformation of its con.ponents for energy absorption. An extensive

! discussion of linearization techniques, and a comparison of calculations and test results is also

] presented in Reference 15.

2.2.1 GAPP Damping ,

)

S&L stated that the analysis of piping systems with E-BARS based on the GAPP Time History 1 l method of analysis will use damping stipulated in USNRC RG 1.61 (Reference 16). The staff finds this acceptable, provided this RG is applicable to the particular plant for which the use of

- E-BARS is being considered.

1 5

The uniform response spectrum methodology of GAPP also requires that damping be specified. I S&L has stated that the proper damping values to be used in the analysis of piping systems 3  ;

i o

i with E-BARS are those based on ASME Code Case N-411-1 (Reference 17). This Code case j was approved by the staff, subject to certain limitations stipulated in USNRC RG 1.84 l (Reference 3). One such stipulation is that this Code case is not appropriate for analyzing the dynamic response of piping systems using supports designed to dissipate energy by plastic deformation. The staff, therefore, recommends that the analysis of piping systems by GAPP

with E BARS should be based on damping values no larger than those specified in USNRC RG 1.61, provided this RG is applicable to the particu!ar plant for which the use of E-BARS is being

! considered.

2.2.1 GAPP Verification S&L performed limited benchmarking of the GAPP program by comparing calculated results l

with measured results from tests described in the EPRI report NP-644219eference 18), and by comparison with calculations performed with industry accepted piping ana ysis computer programs with similar analytical capabilities as those in GAPP.

i

! 2.2.1.1 Comparison with Test Results i EPRI NP-6442 describes shaker table tests with seismic time history input of two simple piping l l systems with gapped supports (not E-BARS), labeled System 1 end System 2. These supports i were designed to deform only in the elastic range. System 1 consisted of a 3" diameter schedule 80 straight pipe, simply supported at the ends, containing a vertical gapped support at a

midspan. The fundamental frequency for this system was observed as 6.4 Hz, and the damping ratio was reported as .25%. System 2 represented a simple three dimensional 3" diameter schedule 80 piping configuration called a "Hovgaard Bend." The configuration contained a vertical gapped support on a horizontal leg, and a horizontal gapped support at one i location on a verticalleg. The first mode frequency was observed to be 4.9 Hz. No damping j ratio was reported. However, based on a recorded strain time history shown in EPRI NP-6442, the staff determined an approximate damping ratio of 1.1%. The staff has evaluated the tests

described in this report, and finds the results of these tests acceptable for referencing in

comparisons of calculated and test results, j Using GAPP, S&L calculated support loads and stresses for the two systems described above, i and compared these values with the measured values. Using a response spectrum analysis

based on a .25% damping acceleration response spectrum, S&L calculated the highest support

loads and stresses for 10 tests of System 1. A comparison with the test results indicates that GAPP-support loads overestimated the test support loads in 9 out of the 10 tests. The bending stress was underestimated in 8 of the 10 tests Mthin 10% or less, except in one test. where the bending stress was underestimat . .i by about 20% S&L also performed a resp %se spectrum analysis of System 2, for 3 tests o 1 different gap configurations, again using a .25% damping acceleration response spectrum. S&L calculated selected maximum bending moments, and compared these with measured moments. The comparison indicates that the calculated i moments were overestimated. However, there is no assurance that the these moments woe'd  ;

have been overestimated M response spectra based on higher damping values had been used ,

l in the calculations of the w..onts.  !

2.2.1.2 Comparison with Other Calculations S&L also performed verifications of GAPP by comparing GAPP solutions with those calculated 4

-, , , . , - - - - ,.-.,,,.g. , y , ,

. - ~ . -- - -- .- . _ _ . . --- -. -.- - - - - - - - . -

i d

I l using a similar program, RLCA-GAP, and the commercial general purpose program ANSYS

- (Reference 19), for the test systems described in EPRI NP-6442. The later solutions were also

- based on .25% damping acceleration spectra. S&L also performed verification of the time history analysis features of GAPP, by comparing GAPP solutions and those calculated using the commercial non-linear program ADINA (Reference 20), for the *Hovgaard Bend" contiguration. The pseudo-force approach to time history analysis was also verified by comparing GAPP solutions with those of Reference 11.The GAPP solutions were found to be  :

reasonably close to those obtained using the other programs (which have been accepted by the staff in other applications), and therefore the comparisons are acceptable.

The staff has examined the comparison of the GAPP analyses, reported tests described in 2

EPRI Report NP-6442 and other calcolated results, and concludes that the GAPP results correlate reasonably well wlh the test data shown in the EPRI report. These results also support the staff conclusion that the damping used in GAPP calculations for systems with E-BARS should not exceed the damping values stated in RG 1.61.

3.0 CONCLUSION

> \

} Based on its review, the staff finds, this topical report acceptable for referencing in licensing l applications, subject to the following conditions:

1. The design of E-BARS should be based on the Winter 1985 Addendum to the 1983 i Edition of ASME Section lil, Subsection NF, and later editions. 1 I

2. The daniping values used in GAPP dynamic response calculations should be those j approved by the staff for piping analysis of the specific facility, but not to exceed those j listed in USNRC RG 1.61 (Reference 16).

I j 3. The methods of Uniform Response Spectrum modal combination in GAPP should i correspond to those in USNRC RG 1.92 (Reference 8).

4. For independent support motion analysis, the modal and directional components should be combined in accordance with rules listed on pp. 2-2 and 2-3 of NUREG 1061,

! Volume 4 (Reference 12).

i 4

This concludes our review of the subject topical report.

4 5

i

-i l

, . i l

, .1 1 l

4.0 REFERENCES

i

1. Subsection NF," Component Supports," of the ASME Boiler and Pressure Vessel Code f (B&PV) Section lil, Division 1, all valid editions through the latest edition listed in 10 CFR Part 50, Q50.55a

2. ASME B&PV Code Case N-420," Linear Energy Absorption Supports for Subsection NF, Classes 1,2, and 3 Construction, Section lil, Division 1" i 3. USNRC RG 1.84, " Design and Fabrication Cod 9 Case Acceptability, ASME Section ill Division 1," Revision 30, October 1994

4. ASME B&PV Section ll, latest edition accepted by the staff

5. ASME B&PV Code Case N-249, " Additional Material for Subsection NF, Classes 1,2,3 ,

and MC Component Supports Fabricated Without Welding," Section Ill, Division 1"

6. USNRC RG 1.85," Materials Code Case Acceptability, ASME Section lli Division 1,"

Revision 30, October 1994

7. ASME B&PV Section Ill, Appendix F," Rules for Evaluation of Service Loadings with Level D Service Limits," 1986 edition to latest edition accepted by the staff

8. USNRC RG 1.92, " Combining Modal Responses and Spatial Components in Seismic '

Response Analys:s," Revision 1, February 1976

9. Iwan, W. D.," Predicting the Earthquake Response of Resiliently Mounted Equipment with Motion Limiting Constraints," World Conference on Earthquake Engineering, Proceedings of the 6th Conference, New Delhi, India,1977

10. Iwan, W. D., " Application of Nonlinear Analysis Techniques,"in Applied Mechanics in Earthquake Engineering, AMD-Vol. 8, American Society of Mechanical Engineers,1974

11. Molnar, A. J., et al.," Application of Normal Mode Theory and Pseudo Force Methods to Solve Problems with Nonlinearities," Journal of Pressure Vessel Technology, ASME, May 1976

12. NUREG 1061, Volume 4, " Report of the US Nuclear Regulatory Commission Piping Review Committec, Evaluation of Other Dynamic Loads and Load Combinations,"

December 1984

13. Letter of April 11,1995, from B. W. Sheron, USNRC, to R. L. Cloud, Robert L. Cloud &

Associates, Inc., with enclosed evaluation of the topical report "A Topical Report on the Methodology, Verification and Applications of Computer Program GAPPIPE,"

RLCA/P94/04-94/009, June 1,1994 i

14. Letter of May 21,1990, from P. C. Shemansky. USNRC, to T. J. Kovatch, Commonwealth Edison Company, with enclosed Safety Evaluation Report 6

a r

4 i 15. NUREG/CR-6559,"Large Scale Tests of Main Steam and Feedwater Piping Systems j with Conventional and Energy-Absorbing Supports," prepared by Y.-J. Park, et. al.,

Brookhaven National Laboratory, for the USNRC, August 1998 ]

' )

16. USNRC RG 1.61," Damping Values for Seismic Design of Nuclear Power Plants,"

October 1973 l 17. ASME B&PV Code, Code Case N-411-1, " Alternative Damping Values for Response 5

Spectra Analysis of Classes 1,2,3 Piping, Section 111, Division 1," dated February 20, 1986

18. EPRI Report NP-6442,"A Simplified Piping Support System with Seismic Limit Stops,"

July,1989 i

ANSYS Structural Analysis System, Swansen Analysis Systems, Houston, Pennsylvania 19.

i

! 20. ADINA - Automatic Dynamic incremental Nonlinear Analysis, ADINA R&D, Inc.,

Watertown, Massachusetts j

i 1

i i

e f

1 s

7 i ,

J e - -, w--- - 6