ML20100J877

From kanterella
Jump to navigation Jump to search
Structural Review of Piping Analysis Including Effect of Heavy Elbows
ML20100J877
Person / Time
Site: Beaver Valley
Issue date: 10/31/1983
From:
DUQUESNE LIGHT CO.
To:
Shared Package
ML20100J856 List:
References
B2-12241-65, NUDOCS 8412100487
Download: ML20100J877 (150)


Text

Attachment 2 STRUCTURAL REVIEW OF PIPING ANALYSIS INCLUDING EFFECT OF HEAVY ELBOWS OCTOBER 1983 1

1 i

l l

4 L

8412100487 841025 1

PDR ADOCK 05000412 G PDR B2-12241-65

TABLE OF CONTENTS Section Title Page 4

1 INTRODUCTION 1-1 2 CODE REQUIREMENTS AND NRC REGULATORY GUIDES 2-1 2.1 CODE REQUIREMENTS 2-1 2.2 NRC REGULATORY GUIDES 2-2 3 ANALYTICAL METHODS 3-1 3.1 LINEAR ANALYSIS METHODS 3-1 3.2 RESPONSE SPECTRUM MODAL SUPERPOSITION 3-1 3.3 TIME-HISTORY METHODS 3-2 3.3.1 Time-History Modal Superposition 3-2 3.3.2 Direct Time-History Integration Method 3-3 3.4 NONLINEAR ANALYSIS METHODS 3-3 i 3.5

SUMMARY

OF ANALYTICAL VARIABLES 3-4 4 EARTHQUAKE MOTION DEFINITION 4-1 4.1 INPUT TO BUILDING 4-1 4.2- PEAK SPREADING 4-1 4.3 ENVELOPING 4-2 4.4 RESPONSE COMBINATIONS 4-2 4.5 CONSERVATISMS OF RESPONSE SPECTRA ANALYSIS 4-2 l

5 DAMPING 5-1 5.1 TYPES OF DAMPING 5-1 5.2 DAMPING LIMITS 5-1 5.3 VARIATIONS IN DAMPING 5-2

l. 5.4 COMPARISON OF DAMPING VALUES 5-2 6 COMPARISON OF ANALYTICAL METHODS 6-1 I.

l f-

.B,2-12241-65 i

TABLE OF CONTENTS Section Title Page 7 RESULTS OF GENERIC STUDY 7-1 7.1 MODELS CHOSEN 7-1 7.2 ANALYTICAL METHODS 7-7.3 RESULTS FOR MODEL 1 7-7.4 RESULTS FOR MODEL 3 7-7.5 RESULTS FOR MODEL 7 7-I 7.6 RESULTS FOR 24" PIPING STUDY MODEL 7-8*

SUMMARY

AND CONCLUSIONS 8-1 9 REFERENCES B2-12241-65 ii

c

]

Il

.l i

SECTION 1 INTRODUCTION The structural evaluation of nuclear power plant piping involves considera-tion of loading conditions - ranging from thermal expansion and weight of the piping and its contents to complex dynamic loads caused by earthquakes or fluid transient loads (Figure 1.1). The stress and loads developed in the piping system are dependent upon a number of different variables (Figure 1.2). Piping characteristics such as inherent flexibility, the location and types of supports, damping characteristics of the piping system and its supports, pipe support gaps and linkage free play, along with material nonlinearities are some of'the factors that affect stress levels developed in piping systems (FIGURE 1.3). This complex material and structural behavior is typically analyzed using linear elastic analysis methods.

J The codes that govern the verification of piping structural integrity recognize the nonlinear material behavior and have developed a simplified

" Design By Rules" method to limit stress and strains by elastic methods.

These limits were developed based on sound engineering experience anil con-servative assumptions. In developing these rules the code takes into con-sideration the available industry standards that control the geometry and dimensions of pipes and pipe components. These standard components have i~ been used for many years in piping systems including nuclear power plants I' and have, in general, provided satisfactory performance. These standards e.

- provide minimum dimensional requirements for structural adequacy and allow j manufacturing tolerances to these-nominal dimensions. The code recognizes that these tolerances in dimensions could lead to stress intensification and B2-12241-65 1-1 i

i 1

- flexibility factor values different from those.provided in the code for fit-

' tings with nominal dimensions.

However, many conservative assumptions such as load inputs, load combina-tions, and low damping values, to name a few, coupled with low allowable stresses, provide adequate assurance that piping systems meet code require-ments for the specified condition of design.

The objective of this paper is to identify the conservatisms inherent in the piping analysis methods presently employed by SWEC, to review the existing piping design process, and to a limited . degree, quantify the effects of changes in some of the variables. In the end, it will be shown that refine--

ments in one or two variables used in the present linear analysis methods are not warranted or justified when considered in the context of predicting complex nonlinear behavior with conservative linear elastic methods.

4 e

B2-12241-65 1-2

SECTION 2.

CODE REQUIREMENTS The ' codes that govern piping design in nuclear power plants have evolved' over the last 70 years. The first code, adopted in 1915, addressed the con-l struction of boilers and consisted of about 100 pages. The present day ASME

, code consists of more than 7,000 pages. In the evolutionary process, the requirements for piping analysis have continued-to grow (Figure 2.1).

The earlier editions of _ the B31.1 power piping code basically required an evaluation of thermal stress and pressure stress and provided guidance for

-locating pipe supports. Later editions provided guidance for the evaluation of seismic and fluid transient induced stress. In 1969, the B31.7 nuclear

. piping code was published for use on what is now known as ASME Class 1 pip-ing. This code required a much more in-depth evaluation of piping than was specified in B31.1. Code requirements had extended beyond simple checks placed on deadweight, pressure, and gross thermal expansion to a rather com-1 plex fatigue evaluation considering all moment, pressure, and local thermal effects (Figure 2.2).

i-The code provides specific design and seismic stress limits to assure struc-tural - integrit.y. The design condition, actual operating conditions, and

! postulated off-normal conditions are treated by rules that require the l

L evaluation of stress and strain. The limits contained in the code were set by establishing a relationship between calculated quantities and failure.

I modes that lead to the definition %f stress categories and required design l

l g ,

margins. The code essentially prevents three failure modes (Figure 2.3):

l B2-12241-65 2-1

1. .Burstin'g, gross distortion, and elastic instability.
2. Progressive distortion.
3. Fatigue failure.

The rules begin with the identification.of these different failure modes and then establish the. corresponding elastic stress categories as shown in Figure 2.4. These categories are primary stress, which can cause cata-strophic or instantaneous failure, secondary stress, which causes no instantaneous failure but could lead to incremental plastic collapse, and peak stresses, such as those which occur at discontinuities, that control

' fatigue crack initiation and propagation.

. The limit state of these failure modes is associated with strain well above the proportional limit. The code provides a means of evaluating these in-elastic failure modes by introducing a sequence of conservative assumptions to develop simplified rules for the admittedly complex, nonlinear material and structural behavior. This simplification is achieved-by the application of limit load and shakedown concepts for gross effects and by the develop-ment of stress intensification factors and stress indices for local stress effects. Figures 2.5, 2.6 and 2.7 show code allowables and appropriate equations.

As a result, the present design rules require no detailed analysis (or knowledge) of the actual elasto plastic behavior of the component. By ex-ploiting these nonelastic concepts, the developers of the code provided simplified means for inelastic strain evaluation by utilizing traditional techniques of linear elastic an. lysis (Figure 2.8).

B2-12241-65 2-2

r__

Figure 2.9 summarizes the code design requirements relating to elbows.

Essentially, the code states that nominal dimensions should be used in the analysis and in the calculation of stress indices, stress intensification and flexibility factors. Both the code' and standards provide requirements for minimum dimensions but neither contains an overtolerance on wall thick-ness or weight. Thus, an elbow or tee with varying thickness greater than the minimum required would meet the industry standards. The implication is that modeling elbows or tees lighter than those actually used ' result in conservative piping systems.

(

i I -

i B2-12241-65 I 2-3 .

SECTION 3 ANALYTICAL METHODS The _ analytical tools available for use by engineers for the evaluation of structural response of piping systems can be categorized into two major analysis methods, Linear Analysis and Nonlinear Analysis (Figure 3.1).

Linear analysis methods are used for production design, and the nonlinear analysis method is usually used for study or investigation.

3.1 LINEAR ANALYSIS METHODS 1Two distinctly different linear analytical methods are recognized as accept-able by U.S. NRC Standard Review Plan 3.7.2 (Ref. 2) for the seismic analy-sis of piping systems. They are:

1. Response Spectrum Method
2. Time-History Method 3.2 RESPONSE SPECTRUM MODAL SUPERPOSITION Response spectrum is the most widely used method for calculating dynamic re-sponse of a piping system. In this method, dynamic input to the piping analysis is in the form of an amplified response spectra (ARS) obtained from

'the civil or structural analysis. Normal mode analysis is utilized and significant ' contributions are assumed to occur for all frequencies up to 33 Hz. The seismic inertia calculation is performed on a modal basis for a three-dimensional earthquake. For Beaver Valley Unit-2 Power Station.

B2-12241-65 3-1

the horizontal res ponses are combined by square root sum of squares (SRSS) and conservatively added ab so lut el y to- the vertical response before the individual modal responses are combined by SRSS or modifications to SRSS as suggested in NRC

~

Regulatory Guide 1.92 (Ref. 3), to obt ain total. system dynamic response. For these generic studies, the horizontal and vertical responses will be combined by SRSS.

3.3 TIME-HISTORY METHODS In this method, seismic excitation time-histories calculated from the building floor response time-histories are applied to the piping system. Separate time-history analyses are pe r-formed fo r each of the three directional seismic-excitations.

3.3.1 Time-His tory Modal S upe rpos it ion This . method calculates modal response retaining the time and phase relationships between the individual modes. Then the modal contributions can be algebraically superposed at each I

time increment to obtain tot al modal responses fo r each of the three directions of seismic excitations. Res po ns es f r om each g direction are further combined by the SRSS method. Again, this method is ap p li c ab le only to linear ela s t ic structures.

B2-12241-65 3-2

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

3.3.2 Direct Time-History Integration Methods This method utilizes time-history seismic excitations similar to the pre- +

vious method except-that the results are not defined in terms of structural

modes. This method involves a step-by-step analysis using very short time e

increments. The analysis procedure involves determination of the response history for each time interval and . takes into consideration the response i

history at the beginning of each time interval. . By successive iteration, the total responses over the entire Ieriod can be calculated. During each I' time . increment, the structure is assumed to be linearly elastic; however, between increments, the properties are modified in accordance .with the cur-J rent condition of deformation. Thus, this method can be utilized to take

! into account the nonlinearities by calculating the nonlinear responses as a sequence of linear responses of successively differing systems.

1- t 3.4 NONLINEAR ANALYSIS METHODS i

j.

The nonlinear analysis methods are not normally used to predict piping re-

! sponse due to the typically high costs associated with this method. However,

! it is available for use to determine a more accurate response of a nonlinear system including gaps and free play at restraints and snubbers. The direct time-history integration method described above is used, but with nonlinear boundary conditions.

f I

}

l i

B2-12241-65 3-3

3.5

SUMMARY

OF ANALYTICAL VARIABLES All of the methods discussed above have a number of significant variables which directly affect the analytical result obtained. In the following two sections specific variables relating to earthquake motion definition and system damping are considered.

A comparison of these analytical methods complete with referenced test data is presented in Section 6.

Results of a piping parametric study that - considers damping and elbow thickness as variables is presented Section 7.

?

i

/

B2-12241-65 3-4

SECTION 4 EARTHQUAKE MOTION DEFINITION 4.1 RESPONSE SPECTRA The. time history response at different elevations of structure is calculated (Figure 4.1) and applied - to a family of single degree of freedom models as -

shown in Figure 4.2 to generate the amplified response spectra curves to ~ be used in the piping analysis. Typical ARS curves are shown on Figures 4.3 and 4.4 for the Diesel Generator Building of Beaver Valley 2 site. The tab-ulation in Figure 4.5 illustrates the significant effect that equipment damping has upon the maximum seismic response. The low level of damping assumed in the analysis and the enveloping of the highest elevation within the building relative to the particular piping system and applying the envelop spectra to the entire stress problem provide significant con--

servatism in the seismic analysis, d

4.2 PEAK SPREADING i

The peaks in amplified floor response spectra are broadened to' account for variation in material properties and approximations in modelling. Fig-ure 4.6 depicts an ARS curve with peak spread. Peak broadening is intended to reflect a range of uncertainty in the precise location of the resonant E peak of the response curve, and not to indicate that multiple peak resonant

. response is likely within a broadened range. What typically exists is a

't i family of resonant curves, each having only one point of maximum resonant l

l response.

B2-12241-65 4-1

It would be more precise to analyze systems and components for a number of unbroadened spectra which are members of the~ broadened family of possible response spectra. Since there is usually only one significant peak frequ-ency for a given system, the use of peak broadened floor response spectra is a conservative analytical expediency that results in an additional margin of safety for systems, components, and supports.

4.3 ENVELOPING As was shown in Figure 4.6, response curves are simplified by enveloping the curves with straight lines. For piping systems supported at two or more locations in one or more buildings, the ARS curve is taken as the upper bound envelope of all the individual response spectra for these locations.

The envelope spectrum is then applied uniformly to all points in the system, resulting in additional conservatism for most regions of the piping system.

4.4 RESPONSE COMBINATIONS Closely spaced and separated modal responses are combined conservatively in accordance with the rules of U. S. NRC Regulatory Guide 1.92 (Ref. 3).

4.5 CONSERVATISM 5 0F RESPONSE SPECTRA ANALYSIS There is much conservatism in the development of amplified response spectra for seismic analysis. Figure 4.6 shows the effects of peak spreading and enveloping. Other factors, such as a single equivalent low system damping r

B2-12241-65 4-2

value being assigned and applied to all frequencies and the uniform applica-tion of enveloped spectra accelerations to the entire piping system, add to the conservatisms inherent in this analysis method. The effect of damping

'is discussed separately in the following section.

'/

r, 8

B2-12241-65, 4-3

SECTION 5 DAMPING 1

Damping, as it is normally defined, is the means by which the response motion of a structural system is reduced as a result of energy ~ losses. In the performance of a dynamic structural analysis, this energy dissipation is usually accounted for by specifying a damping value that would result in energy dissipation in the analytical model equivalent to that expected to occur as a result of material' or structural damping in the real structure.

5 .1 ' TYPES OF DAMPING Figure 5.1 summarizes the types of damping usually considered as follows:

1. Structural damping is due ' to pipe slippage on supports, joint slippage, and frictional effects internal to . the pipe supports.

Typically, structural damping is estimated to be 5 to 10 percent of critical damping.

2. Impact damping includes impact or closing of gaps in supports and
changes in geometry. This damping is typically estimated between 5 and 10 percent of critical for steel piping.

i.

3. Material damping due to hysteresis energy loss, and nonlinear or detuning effects due to changes in boundary conditions as the piping -is vibrated are typically only 0.04 to 0.2 percent of.

l- "

l critical damping for steel piping.

B2-12241-65 5-1

5.2 DAMPING LIMITS The historic development of damping limits used in seismic analysis of piping is summarized in Figure 5.2. The damping value, 0.5 percent of cri-tical, suggested by Housner was used extensively from 1963 to 1968. The values proposed by Newmark in 1969 ranged from 0.5 to 5.0 percent depending on stress level. NRC Regulatory Guide 1.61 (Ref. S') issued in 1973 limits, as an interim lower bound, range from 1.0 to 3.0 percent damping depending on pipe size and type of earthquake (OBE or SSE). The values proposed by Newmark and Hall in 1978 are also in the 1.0 to 3.0 percent range.

5.3 VARIATIONS IN DAMPING I

Damping is not linear as assumed in the stress analysis but increases with stress level. As a result, the higher vibratory loads at resonance are l

partially compensated for by higher damping and lower amplifications than assumed.

The damping supplied by snubbers, which is not considered in the analysis, varies with frequency of vibration.

5.4 COMPARISON OF DAMPING VALUES Subsequent testing of piping systems has indicated that actual damping far exceeds the NRC Regulatory Guide 1.61 limits.

6 B2-12241-65 5-2

In Figure 5.3, the damping values used in the BV2 analysis are compared with the NRC Regulatory Guide 1,61 values and test values from several sources.

For the BV2 analysis, 0.5 percent of critical damping is used for OBE and 1.0 percent for SSE. These damping values are low compared even to the Regulatory Guide 1.61 limits. Typically, a change of 0.5 to 2.0 percent damping reduced seismic loads in a resonant region by a factor of two or more. Figure 4.5 shows typical data for 0.5, 1, 3, and 4 percent damping illustrating this trend.

Vibration tests on piping systems show that actual system damping is signi-ficantly higher than the BV2 values and those stated in NRC Regulatory Guide

1. 61. Vibration testing results on the Joyo heat transport system (Ref. 6) indicate damping values from 3.3 to 23.6 percent of critical, with the highest damping at the lowest frequencies, which are the frequencies more likely to respond to seismic excitation. The testing of small bore piping (Ref. 7) with both struts and snubbers at intermediate supports indicated damping between 10 and 20 percent. Westinghouse Reference 8 snapback tests indicate damping of 4 to 6 percent for uninsulated piping and 5 to 11 per-cent for insulated pipe. J. D. Stevenson (Ref. 9) projects damping values to be in the range of 10 to 12.7 percent of critical.

More recently, the subject of damping has been investigated by the damping subcommiteee of the PVRC Technical Committee on Piping Systems. Although their investigation is not complete their preliminary recommendation is to use 5 percent damping to 10 Hz linearly decreasing to 2 percent at 20 Hz and B2-12241-65 5-3

held ~ constant at.2 percent to 33 Hz for both OBE and SSE independent of pipe diameter. This recommendation is sh'own in Figure 5.4 along with NRC.

Regulatory Guide 1.61 recommended values.

0 9

B2-12241-65 5-4

r-SECTION 6 COMPARISON OF ANALYTICAL METHODS Piping systems of nuclear power plants receive a detailed structural evalua-

' tion comparing ~ the results obtained from piping analyses to the acceptance criteria described in the FSAR. The analysis of piping systems has tradi-tionally been carried out under assumed linear conditions, utilizing the relatively inexpensive and conservative response spectrum technique. This technique ignores nonlinear dynamic characteristics present in a piping sys -

tem. As the state-of-the-art in piping analysis has expanded into other u .

techniques of dynamic analysis capable of considering no slinearities, many authors have undertaken investigations to study the effect of these non-linearities in the piping responses. These studies provide a comparison of piping responses obtained by linear and ' nonlinear analytical methods. In some cases test results have been included.

G. M. Hubert, Reference 11, compares the dynamic responses of an LMFBR cool-ant loop piping calculated by two different methods, response spectrum and time-history modal superposition. Figure 6.1 shows the piping system nodel i utilized in the analysis. Both methods assume linear conditions. The 4

response spectrum method utilizes peak broadened response spectra while the time history analysis utilizes time histories with seismic excitation time scales expanded and contracted by 110 percent, analogous to peak broadening i

in the response spectrum method. The modal responses in'the three different directions in the response spectrum method were combined in accordance with U.S. NRC Regulatory Guide 1.92 (Reference 3). In the case of the time-history method, maximum responses were obtained utilizing four possible l

< B2-12241-65 6-1

p unique algebraic combinations of the three directional earthquake. -Figure 6.2 shows the maximum loads at reactor nozzle calculated using the two ana-lytical= methods. The modal superposition time history results are between 11 percent and 42 percent lower than the response spectrum modal analysis results. This conservatism in the response spectra method is attributed to the different methods utilized for modal response combination as well as to the difference between the developed seismic excitations used in the two analytical methods. While. seismic input in both methods introduces the same conservatism for peak broadening, the seismic input in the response spectra method introduces additional conservatism, namely the removal of valleys from the response spectra (Figure 6.3). Hubert studies the effect of this additional conservatism.by examining the analytical results for a particular mode. Figure 6.4 shows contributions to nozzle . loads from the 12.31 Hz mode. The results indicate that response spectrum analysis modal contribu-tions are 24.5 percent larger than those - from time-history analysis. The same percentage difference exists in the seismic input, suggesting that this conservatism is due to the removal of valleys in the spectra.

The differences between calculated responses become more significant when the actual nonlinearities present at pipe supports are introduced. L . K .-

Severud et~ al, Reference 7 demonstrate this difference in a small bore pip-l ing test. The authors subjected a 1 in. dia stainless steel piping system with several bends, risers, and supports (see Fig. 6.5 for piping model) to j a dynamic test simulating seismic accelerations. The test results are compared to results obtained " by linear and nonlinear analytical . methods.

. Linear response spectra analysis utilized three different modal combina-1 tions, namely absolute, U. S. Regulatory Guide 1.92, and SRSS (square root-l B2-12241-65 6-2 .

L.,

u -

sum of squares) ' methods. The . nonlinear time-domain analysis.is' performed for .various values of gap and damping. Figure 6.6 shows the comparison of test results . to thcse obtained by the 'two analytical methods. Figure 6.7 highlights the trends depicted in Figure 6.6. The results clearly demon-t strate _ that when a more exact piping model, including nonlinearities, is introduced,-it produces significantly' lower responses than those obtained by the response spectra method utilizing simple linear assumptions. In . fact, the test results show the conservatism present even in the complex time-domain nonlinear analysis method.

.Similar conclusions have been arrived at by other. investigators. In Ref. 12, D. A. Barta et al perform a more detailed study utilizing three piping models (Figures 6.8, 6.9, and 6.10) analyzed by response spectra and nonlinear time history methods. Results comparison ~ shown in Figures 6.11, 6.12, and 6.13 clearly reinforces the conclusions arrived at by other inves-tigators that the linear response spectra method predicts more conservative 1*

!I results.

l One of the factors that contributes to the conservative results in the linear elastic analysis is damping as discussed in Section 5.4. Reference 6

' reports the results of in-situ vibration testing of the Joyo heat transport piping system. Figure 6.14 shows a schematic view of the Joyo primary hot l.

l leg piping system that was tested dynamically. Figure 6.15 presents the l

piping tested on the primary loop. Figure 6.16 gives the in-situ vibration

.g

. test summary with analytical data. The test results indicate that damping I- values range from 3.3 percent to 23.6 percent of critical.

l B2-12241-65 6-3 e g

SECTION 7

-RESULTS OF GENERIC STUDY 7.1 MODELS CHOSEN

. Three _ benchmark piping models were chosen from NUREG/CR-1677 (Ref.14) as part ~ of this generic study. Problems No. 1, 3 and 7 were selected and modified as described below and shown in Figures 7.1,_7.2, and 7.3, respec-tively. In addition' _ a 24. in. piping study model shown in Figure - 7.4 was

- selected.for its realistic. pipe size.

Problem No. 1 is a three directional piping model made up . of only three in-line straight pipe elements with two pipe bend elements all between two fixed anchors. Since this hypothetical model of 7.28 in. dia pipe had five dia. bends, they were modified for this study to long radius elbows. This problem was ' selected for its simplicity allowing ready checking of key results.

Problem No. 3 was , chosen because of its relative simplicity, its mathemati-cal characteristics, and the large number of on-axis restraints acting in

. three global coordinate axes. Since it is described in NUREG/CR-1677 as a hypothetical model of 7.28 in. dia. pipe with five dia. bends developed to verify ' different methods of ' combining modal results using manual) calcula-tions . as - well as to compare different computer program characteristics, capabilities, and results, it was modified for this study to contain long radius elbows instead of the five diameter bends.

1 J

B2-12241-65 7-1 /

. Problem No. 7 . was chosen - because of its realistic geometry. The model represents ' an actual multi-branch header with different pipe diameters (4 in./3 in.) and includes valves, springs, restraints, and an 'in-line anchor, as well as several elbows and tees.

+

The 24 in. piping study model was chosen ' for its realistic pipe size. It-contains many elbows, valves, anchors, and tee connection. The model which is standard weight pipe has several relatively long straight piping runs with long radius elbows to a.new direction. It is conservatively supported for thermal conditions, using snubbers as required, anl it lends itself well

'for these studies. The piping is modeled at 350*F for evaluation of thermal

-expansion loading.

7.2 ANALYTICAL METHODS Each model was analyzed for loadings, due to thermal exapansion, deadweight and seismic inertia loading. The results of the thermal- expansion analysis

~;.

are based on an operating temperature of 350*F for all models except Model 7 - -

which is based on 400*F. The results of the deadweight analysis are based' , I on the uniformly distributed weight of the pipe or , lumped w'eight obtained from NUREG/CR-1677 for the benchmark models.

l

=

The results of the seismic analysis are based on a response spectrum modal i

. superposition method with the intra-modal results being combined by SRSS and the inter-modal results being combined by the " grouping method" suggested in Reference 3. All four modals were analyzed using specific BV// 2 ARS for 1 percent and 3 percent-of critical damping. No effort to modify the various B2-12241-65 7-2 a .

a

support schemes to suit the BVPS-2 ARS was undertaken and as can be seen by reviewing the results - for Models 1 and 24 in., there is little or no response exhibited. To simulate the SSE intensity while ignoring the fre-quency content of these models, flat response spectra were developed and utilized. The flat response spectrum of 5g's was used to represent 1 per-cent of critical damping and 3g's was used to represent 3 percent of criti-cal damping.

The rationale employed in the current design basis for piping at BVPS-2 uses 1 percent of critical damping for SSE with standard weight fittings. When evaluating the effects of the thicker fittings a more realistic, though still conservative, 3 percent of critical damping value for the SSE is used, as explained in Section 5 of this report.

It is recognized that the extra heavy fitting case, where the fittings are three times as thick as the standard fittings, is a very extreme condition.

All comparisons will be based on the standard vs heavy fitting cases which are more ,likely to occur. The results of the extra heavy fittings analysis is provided for information and to verify the anticipated trends.

7.3 RESULTS FOR MODEL 1 Figure ~ 7.5 gives the natural frequency response of a Model I considering s tar.da rd , heavy and extra heavy fittings where the heavy fittings are twice as thick as the standard fittings and the extra heavy fittings are three times thicker than standard fittings.

l l

l

'B2-12241-65 7-3

The results of the stress comparisons for Model 1 are presented in Figure 7.9 for ' loadings due to thermal expansion and deadweight. The results of the thermal expansion- are based on an operating temperature of 350*F, and results of the deadweight analysis are based on the lumped weights obtained .

from NUREG/CR-1677.

The ' stress results of the seismic analyses for Model 1 are presented in Figures 7.10 and 7.11. Model I was analyzed using specific BVPS-2 ARS for'l percent and 3 percent critical damping, and for flat repsonse spectra of-Sg's used to represent 1 percent of critical damping and 3g'o used to repre-sent 3 percent of criticial damping.

The pipe support loads for Model 1 due to thermal expansion and deadweight are presented in Figure 7.12. The seismic pipe support loads for Model 1, which are base on the falt 3g and Sg response spectra, are shown on Figure 7.13. The total design loads for Model 1 are shown in Figure 7.14 and represent the sum of thermal expansion, deadweight and seismic loadings.

7.4 RESULTS FOR MODEL 3 Figure 7.6 gives the natural frequency response for Model 3 considering standard, heavy, and extra heavy fittings.

These results of the stress comparisons for Model 3 are presented in Figure-

-7.15 for loadings due to thermal expansion and deadweight. The results of the thermal expansion are based on an operating temperature of 350'F, and B2-12241-65 7-4

results of the. deadweight analysis are based on the lumped weights obtained from NUREG/CR-1677.

The results Lof the seismic stress analyses for Model 3 are presented in Figure 7.16 and are based on specific BVPS-2 ARS having 1 percent and 3 per-cent of critical damping.

The pipe support loads for Model 3 due to thermal expanion and deadweight are presented in Figure 7.17 for intermediate supports and Figure 7.20 for anchors. The seismic pipe ~ support loads for Model 3, which are based on specific BVPS-2 ARS, are shown in Figure 7.18 for intermediate supports and 7.21 for anchors. The total design loads for Model 3 are shown in Figure 7.19 for intermediate supports and 7.22 for anchors, and represent the sum of thermal expansion, deadweight, and seismic loadings.

7.5 RESULTS FOR MODEL 7 Figure 7.7 gives the natural frequency response for Model 7 considering standard, heavy, and extra heavy fittings.

The results of the stress comparisons for Model 7 are presented in Figure i 7.23 for loadings due to thermal expansion and deadweight. The results of i

the thermal expansion are based on an operating temperature of 400*F, and results of the deadweight analysis are. based on the lumped weight obtained F.#

from NUREG/CR-1677.

l' B2-12241-65 7-5 N ';

The results of the seismic stress analyses for Model 7 are presented in Figure 7.24 and are based on the specific BVPS-2 ARS having 1 percent and 3 percent of critical damping.

f The pipe support loads for Model 7 due to thermal expansion and deadweight are presented in Figure 7.25 for intermediate suppports and Figure 7.26 for anchors. The seismic pipe support loads for Model 7, which are based on specific BVPS-2 ARS, are shown in Figure 7.25 for intermediate supports and 7.27 for anchors. The total design loads for Model 7 are shown in Figure 7.25 for intermediate supports and Figure 7.28 for anchors, and represent the sum of thermal expansion, deadweight, and seismic loadings.

7.6 RESULTS FOR 24" PIPING STUDY MODEL Figure 7.8 gives the natural frequency response of the 24" Piping Study Model considering standard, heavy and extra heavy fittings.

The results of the stress comparisons for Model 1 are presented in Figures 7.29 for loadings due to thermal expansion and deadweight. The results of the thermal expanion are based on an operating temperature of 350*F, and results of the deadweight analysis are based on the uniformly distributed weight of the piping.

The stress results of the seismic analysis for the 24" Piping Study Model are presented in Figures 7.30 and 7.31. This Model was analyzed using specific BVPS-2 ARS for 1 percent and 3 percent critical damping, and for flat response spectria of Sg's used to represent 1 percent of critical damp-B2-12241-65 ,

7-6

ing and 3g's used to represent 3 percent of critical damping.

The pipe ' support loads for the 24" Piping Study Model due to thermal expan-sion and deadweight are presented in Figure 7.32 for intermediate supports and Figure 7.36 for anchors. The seimic pipe support loads for the 24" Piping Study Model, which -are based on the flat 3g.and Sg response spectra, are shown in Figure 7.33 for intermediate supports, Figure 7.35 for inter-mediate snubbers, and Figure. 7.38 for anchors and represent the sum of thermal expansion, deadweight, and seismic loadings.

l l

I

?

l B2-12241-65 7-7 l

L

T' SECTION 8

SUMMARY

AND CONCLUSIONS The analysis of a piping system both for static and dynamic loads has inher-ent in it a number of simplifying assumptions. The process is to model and analyze admittedly complex material and structural behavior with economical linear elastic analysis techniques. Obviously, the approach requires assumptions to be on the conservative side. The investigation documented in this report indicates that excessive conservatism exists in a number of the analyses variables.

Investigation of analytical methods supports 'the conclusion of many authors that the response spectra method of analysis is quite conservative and nor-mally results in over-supported piping systems. Much of the conservatism is attributed to enveloping, peak spreading, and the considerations of spatial and modal components.

Investigation of damping reveals that test results from many different authors support the use of larger damping values than those contained in NRC Regulatory Guide 1.61 and presently used in analysis. Damping in the range of 4 or 5 percent of critical appears to be more representative for piping in the frequency range applicable to seismic excitation (Figure 5.4). The effect of increased damping is to reduce piping response. Typically, a change from 1/2 to 2 percent or from 1 to 4 percent reduces response by a factor of 2 or more (Figure 4.5).

B2-12241-65 8-1

Other areas of conservatism exist in the calculation of dynamic stress.

Stress limits based upon material under static loads are somewhat conserva-tive in terms of limiting dynamically induced stress because strain energy effects are neglected.

Many investigators have supported the conservatisms of present-day piping analysis methods as discussed in Section 6 of this study. Many have quanti-fied the effects of changing different variables inherent to the analysis process. Figure 6.7 illustrates the overall conservatism presented in a typical piping system by comparing linear response spectra analysis and non-linear analysis with neasured test results. More recently, an investigation into the probability of piping fracture of a reactor coolant loop included a sensitivity study. The results of this study (Reference 15) concluded that their best estimate of the state of stress as compared to that calculated in the design process was about an order of magnitude less.

The generic study described in Section 7 presents the effect of varying thickness of pipe fittings, namely, elbows and tees. Analyses were per-formed for standard, two times, and three times the standard fitting thick-ness for thermal, deadweight, and seismic loading conditions. Since the

~

linear analysis methods presently employed by SWEC conservatively use 1 per-cent of critical damping for SSE ARS and nominal piping dimensions as stated in the ASME Code Section III, the effects of thicker fittings were evaluated using a more realistic but still conservative 3 percent of' critical damping for SSE ARS. To simulate a uniform SSE intensity for the models which experienced little seismic response, a flat response spectrum of Sg's was B2-12241-65 8-2

t used to represent 1 percent of critical damping and 33's to represent 3 per-cent of critical damping.

The effects of thickened fittings on the system natural frequencies is mini-mal as all . natural frequencies have remained approximately the - same or

' increased slightly above those of the respective standard fitting models (Figures 7.5 though 7.8). It can be demonstrated that increases in the sys-tem natural frequencies .will usually have a positive effect on reducing system response.

. The effects of thickened fittings on the system stress levels are generally favorable in that the peak stresses are usually reduced. The ~ maximum stresses in a piping system are usually at the fittings which are reduced due to decreases in the stress intensification factors and increases 'in the section. moduli associated with the thicker fittings. Figure 7.9 shows a

' maximum thermal stress of 15,237 psi (at an elbow) for the standard fitting -

case and a maxi:num thermal stress for only 13260 psi (at an anchor) for the heavy fitting case for Model 1. Simarlarly, Figure 7.23 for Model 7 shows a thermal stress reduction of 6,212 psi to 3,675 psi, both at an elbow. The use of more realistic damping values for the SSE ARS and the reduced intensification factors for the thicker fittings generally overshadow the increased weight of the heavier fittings and result in lower seismic stresses. For Model 7 there is a dramatic decrease in the seismic stress of 46,219 psi for the standard fitting case with I percent SSE ARS to 9,843 psi for the heavy fitting case with 3 percent SSE ARS (Figure 7.24). Simular

. trends may be observed for the other models (Figures 7.*10, 7.16, and 7.30).

B2-12241-65 8-3

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

P 1

< Therefore, pipe stresses are conservatively predicted when using nominal standard weight fitting and 1 percent of critical damping.

{

The effect of thickened fittings on pipe support loads vary for specific piping systems. The results of Model 1 showed increases in the total design loads for ' both , anchors, but this model is not representative of typical piping systems because there are no intermediate supports. Model 3 results i

indicate one support with a moderate increase, one with a slight increase ,

1 and three with a decrease in total design load (Figure 7.19). Of the anchors on Model 3, only one of the three anchors shows increases of two of the components for the total design loads (Figures 7.22). The results of Model 7 show a reduction in restraint design loads (Figure 7.25), and a general reduction in anchor loads (Figure 7.28). The results of the 24 in.

Piping Study Model indicate that of 43 intermediate restraints, - only two show a modest increase in the total design loads while all others reduce as shown on Figure 7.35. The anchors on the 24 in. Piping Study Model all show a reduction in design loads for the heavy fittings with 3 percent of criti-4 cal damping (Figure 7.38). Based on this study of selected piping models, j it' can te concluded that only a small percentage of the . supports are likely to experience a moderate increase in the total design load. Therefore, pipe i support loads are generally conservatively predicted when using nominal standard weight fittings and I percent of critical damping for SSE ARS.

l '

The data presented in this study provides a sufficient bases to. conclude

] that the current design methods, which use SSE ARS having 1 percent of cri-i tical equipment damping and nominal standard weight fittings, will yield

! conservative pipe stress results and conservative pipe support loads. A B2-12241-65 8-4 eeva--c.-*-r-m~--e

r-similar conclusion can be inferred for current design methods using opera-tion basis earthquake (OBD) ARS having 1/2 percent of critical equipment

. damping and nominal weight fittings.

l i

I 1

I i

n B2-12241-65 8-5 .

A

.~.

SECTION 9 REFERENCES

1. American Society of Mechanical Engineers (ASME) " Boiler and Pressure Vessel Code,"Section III, 1971 Edition
2. United States Nuclear Regulatory Commission, (U. S. NRC) Standard Review Plan 3.7.2
3. U. S. NRC Regulatory Guide 1.92, " Combination of Modes and Spatial Components in Seismic Response Analysis" dated February 1976.
4. U. S. NRC Regulatory Guide 1.60 " Design Response Spectra for Seismic Design of Nuclear Power Plants" dated December 1973.
5. U. S. NRC Regulatory Guide 1.61 " Damping Values for Seismic Design of Nuclear Power Plants" dated October 1973.
6. Insitu vibration tests Joyo Plant Japan. Proceedings of U.S. DOE /PNL specialist exchange meeting on seismic piping test held at Advanced Reactors Division dated September 20-21, 1982.
7. L. K. Severud, D. A. Barta, and M. J. Anderson "Small Bore Piping Seismic Test Findings" ASME Pressure Vessel and Piping Conference June-July 1982, Vol. 67 "Special Applications in Piping Dynamic Analysis" pp 29-41. ,,
8. Seismic Design Technology monthly technical progress report - Materials and Structures, November 1982, Westinghouse Task DE-AT02-80CH94049.
9. J. D. Stevenson, Structural damping values as a function of dynamic response stress and deformation levels, 5th International Conference on Structural Mechanics in Reactor Technology Berlin (West) August 13-17, 1979.

B2-12241-65 9-1

10. U.S. NRC NUREG/CR-2137, " Realistic Seisiac Design Margins of Pumps, Valves, and Piping," June 1981, E. C. Rodabaugh, Battelle Columbus Labs., and K. D. Desai, US NRC
11. G. M. Hubert " Comparison of LMFBR piping response obtained using response spectrum and tine-his t.ory methods: ASME Piping and Pressure Vessel Conference, Denver, Colorado June 15-16, 1981.
12. D. A. Barta, S. N. Huang, and L. K. Severud " Seismic Analysis of Piping With Nonlinear Supports" ASME Pressure Vessel and Piping Conference,

. August 1980, Vol. 40, " Effects of Piping Restraints on Piping Integr-ity" pp 5-25.

13. M. J. Anderson, D. A. Barta " Seismic Analysis of Pipe Lines Supported by Seismic Snubbers -

Recommended Guidelines" Task No. HE-1.2, December 1981, Hanford Engineering Development Laboratory.

14. U.S. NRC NUREG/CR-1677 " Piping Benchsaark Problems, Dynamic Analysis Uniform Support Motion Response Spectrum Method" dated August 1980.
15. NUREG/CR-2189 " Probability of Pipe Fracture in the Primary Coolant Loop of a Power Plant" prepared for USNRC by Lawrence Livermore Laboratory, 1981.

B2-12241-65 9-2

LOADS APPLIED TO PIPING STATIC PRESSURE THERMAL EXPANGION LOCAL THERMAL EFFECTS DEADWEIGHT EARTHQUAKE DISPLACEMENTS BUILDING SETTLEMENT DYNAMIC FLUID TRANSIENTS l EARTHQUAKE INERTIAL NYDRODYNAMIC 4

I l

l -

L l

\ -

FIGURE 1.1 i

e a . - ,--- , - -

,.w._emw

4 PIPING DESIGN PROCESS FACTORS THAT EFFECT PIPING RESPONSE 1

CO,DE REQUIREMENTS U.S. REGULATORY REQUIREMENTS LOADING CONDITIONS ANALYTICAL METHODS EARTHQUAKE MOTION DEFINITION DAMPING RESULTS OF PARAMETRIC STUDIES 3

4 i

t FIGURE 1.2

l l

l-FACTORS THAT EFFECT PIPING RESPONSE l

INHERENT FLEXIBILITY

  • LOCATION AND TYPES OF SUPPORTS

! COMPONENT DIMENSIONAL VARIATIONS / TOLERANCES DAMPING CHARACTERISTICS OF PIPING & SUPPORTS

!! ASS DISTRIBUTION SMALL GAPS & FREEPLAY IN SUPPORT SYSTEMS PIPING GEOMETRY

(

I l

, FIGURE 1.3 l

l

CODE REQUIREMENTS CODE EVOLUTION FAILURE MODES STRESS CATEGORIES STRESS COMBINATIONS & LIMITS DESIGN MARGINS CODE APPROACH TO PIPING ANALYSIS e

r FIGURE 2.1

~

l l

CODE DEVELOPMENT I

1935 - ASA B31.1 " CODE FOR PRESSURE PIPING" 2C  :

^

1967 - ANSI B31.1 " POWER PIPING CODE"  ;

i 1969 - ANSI B.31.7 "N UCLEAR POWER PIPING CODE" -

1971 "ASME BOILER AND PRESSURE VESSEL CODE SECTION III" l

l

'?~"*" . _ _ _ - - _ _ _ _ _ _ _

STRUCTURAL FAILURE MODES CONSIDERED BY ASME l

SECTION HI i

N g BURSTING, GROSS DISTORTION, AND ELASTIC INSTABILITY PROGRESSIVE DISTORTION FATIGUE FAILURE 80 -17980

i l.

STRESS CATEGORIES

~

PRIMARY STRESS j- CAUSES BURSTING, OR TENSILE IN-

_ t STABILITY NOT SELF LIMITING SECONDARY STRESS '

CAUSE RATCHETTING OR INCREMENTAL '

COLLAPSE IS SELF LIMITING

~

PEAK STRESS CONTRIBUTES TO FATIGUE FAILURE

==-~~ _ - _ _ _

CODE STStESS ALLOWABLES

~

ASME CLASS 1

~

~2 / 3 S E S = THE SMALLER OF <

Y

M ~

1/3 S u

ASME CLASS 2/3 &B

~

31.1 5/8 Sy S = THE SMALLER OF 4

. s 1/4 Su 33-70414 .

ASME U CLASS 1 STRESS & FATIGUE ANALYSIS REQUREMENTS PER NB3650 C ND S PRIMARY STRESS INTENSITY P Do (EQUATION 9) b 2t

  • b 21)Mi $ 1.5 S. $ 2.25 Sm $ 3.OS, PRIMARY + SECONDARY S = Cf )+C(2 )Mi + 2(1-F) Eel AT,l STRESS RANGE N/R N/R (EQUATION 10) + C3 Ede,T - e nb T! $SSe 1

$ PEAK STRESS RANGE Sp= Ki cf )+ Kg2 C (hMi + 2(1-r)M/ AT,l iii (EQUATION 11) g N/R N/R l g + K3 C#eble,T,-e nTnl + i yEelATal THERMAL EXPANSION RANGE (EQUATION 12) 8=C(h)Mg e g $ 35 N/R N/R PRIMARY + SECONDARY g -

Cf ,1p * ~N 2 g' +

MEMBRANE, + BENDING '

STRESS (EQUATION 13) $ 33,

, ALTERNATING STRESS i (EQUATION 14) Sa m" '/2 K.S, N/R N/R ACTUAL NOL CYCLES USAGE FACTOR O E U $ 1.0 N/R N/R ECMS

~ _ _ _ - ,

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

1 l

1

! ASME NCLASS 2 & 3 STRESS ANALYSIS REQUIREMENTS PER NC3650 & ND3650 1

NORMAL ANO UPSET CONOfTIONS EMERGENCY FiMJLTED CONDITIONS CONDITIONS

, 3 I

h SUSTAINED LOADS P D. (0.75i) Ma IS y (EQUATION 8) 41 E h N/R N/R '

w ,

OCCASIONAL LDADS g , (0.751)Ma (0.751) Me (EQUATION 9) 4t E E $ 1.2 S h $1.8Sh 52.4S h '

THERMAL EXPANSION LOADS iMc

! (EQUATION 10) E $SA N/R N/R

$ f(1.25Sc+ 0.25Sn SUSTAINED + PD.

. THERMAL EXPANSION LDAOS 4, + (0.75i)Ma g + .g (i)Mc

$Sa+Sh N/R N/R (EQUATION 11)

CODE APPROACH TO PIPING ANALYSIS

, DEVELOP COMPUTER MODEL -

s i

INCLUDE FLEXIBILITY OF VARIOUS PIPING COMP.ONENTS CALCULATE FORCES, MOMENTS, DEFLECTIONS & ROTATIONS CALCULATE NOMINAL STRESS AND MODIFY BY USING STRESS j INTENSIFICATION FACTORS OR STRESS INDICES i

i _ " - " ' . -

t

<3 4

ASME SECTION III, 1971 ED. TO WINTER 1972 .

SE1.ECTED CODE REQUIREMENTS RELATING TO ELBOWS CODE PARACRAPH DESCRIPTION OF ITEM NB/NC-3641.1 Requires that manufacturer's tolerances be .,

considered in calculating minimum wall thickness.

NB/NC-3642.1 Thickness af ter bending curved segments of pipe must meet minimum wall thickness requirements.

N.5-3642.2 Elbows manufactured in accordance with ANSI B16.9 and 516.28 are acceptable for the same nominal thickness of pipe except the crotch region of short radius elbows (B16.28) must be 20Z thicker them minimum usil thickness.

NB/NC-3650 Stress evaluation equations use nominal wall thickness.

NB-3672.5 ^ & States that expansion stress calculations shall be based "NC-3672.9 upon the least cross-sectional area of the pipe or fitting, using nominal dimensions.

NC-3673.2(e) States that nominal dimensions shall be used for flexibility calculations, s

NS-3683.1 States in part that for ANSI 316.9 and B16.28 piping prod-ucts nominal dimensions for the equivalent pipe are used for diameter and well' thickness terms.

FIGURE 2.9

k ANALYTICAL METHODS LINEAR ANALYSIS METHODS STATIC DYNAMIC RESPONSE SPECTRA MODAL ANALYSIS TIME HISTORY MODAL SUPPERPOSITION DIRECT TIME HISTORY INTEGRATION NON-LINEAR ANALYSIS METHODS l

4 6

! FIGURE 3.1 i

,e v , - - --- ,y.,--,w,4g,... - , , ,-,--r- ,-,--.mr,.--m-,% y_,--- - . . - -,------v_,-, , - - ~- ,-c-.-~,-,-- - . - . ,-, -r--w-- , , ..r.,

1 i

GENERATION OF ARS t

)

n r ak,

& A k > a-

. i: T A = a lA

, l I -

I 3 a -

T a h 'g

  • Mh

. BUILDING i

BUILDING RESPONSE

} MODEL SPECTRA I

'l i

i '

1 :-}

GENERATION OF AN AMPLIFIED RESPONSE SPECTRUM '

$ ARS MASS AMPLIFIED -

RESPONSE .

s '

! DAMPER 1 SPRIN G

! is f2 Is I4 Is in , ROOR

! 2 FREQUENCY E

, g

,8,,, ,

$$'Ibif.T*N

' INPUT ACCELERATION 5 nURDING l DAMPED SINGLE DEGREE-l [ OF-FREEDOM OSCILLATOR i

0 is fa

'I r ,

e FLOOR 15 SEC.

l c  :.

MAX."g"

-,-.-- - X.L.

w-r- 4"v -

l r2 c._a x1 j/ i EARTHQUAKE

1 -

i s - . .

i i a

i . .

I i i I j 4 4

,1 y ,

, .  ! i .

i 6 9

i I 5 t

i i i i

i

. . I I i e i

"- , 5

' ' i .

t e i

' s i e M gut 8

i w y e 4 5 .

e . I E a mm. I ,

I e .

I G  !

E 8 t i

~ '" i 1 .

1 .

i r oi i i .

T a w I I as

  • i I W 9, D I _

i .-

w q -

, i i E *

' i

  • t . . e i I I I t E I i 3

l

+ i i

a 3

- w E i f I 2

- , n 6 i .

' l

' t 3

  • 1

' _ i I r l i 8 ==

_ t W I

+ l  ! ,

V i K i 6 C 3 i e s

i 3 Jr D i

I . WX I 1 I g - j l E 5 ' 6 II i es i

. us . us g .

3l l t M I C iW 3

' ' 19 ' I e s.a i- --

  • me im _i "" ,

F - = 10B 1"

" ' EI te ~

8 n e3

R t a in n i w e e i i y- ,,,

i- ,

' m a .'s I i _ EI E. , i Ln 6 D--

. =

gu - ,

e i'

E i i MI  %*

- r R 5 i .Ea 8 i

,e , , , 2 w 1.

' #1  ! W-W F <r T  ! af LJ-

, M "s :E .E ~a i i _M

(

2 O w -

i=

.. . 'm.-

3 i e

, , ,M i re a .. , an . .tu

, = ,

i , f- , na is in .e r n-

I . . f 1

! F '" _2 G l

' J a# )  ?

k. .  ; -- i i p..- . ,

i " am 8 I

  • E iT * .#2 f t i,E -- a, x -
s , ,

, e .= . , i . i i

= >

2 s w  ;. - i i i ' i

, . , i

x - r in . .

.s .m i i i

= , t

> e a. i. i i

" " " .. E . .

  • 5 O

'- E i i 7 'M i 98 ,

i i in a ' s---

2

-, i O. ---i l . -

,m_ . ._ _, ,

- x m .. >

.c < -. ., _w i x -

i 1 InB I 1J W .>= i l N. 5

'* + L_ t_

" T a e s .-

  • 5 i i i _ _ , m .' W I '

N === eut 'as E ** M C If I - ~ ~~~~

% C 6

iv

" "_ 7 M !E

\1s "n n G,

- i- a i i

. Ti

  • i

= - z , , .

i e= E i = W I I i\\ \

l I*

  • i B K i l r 5  ;

1 1 I j i i e i ! _ t r i +

I l

_3 i

'

  • i i -

. . . K, r .

..s. a. . h ,

ww w

= = . . . ..

vv == v, _ ww A _ . . . .w a_ a_. _ _ _ w w _

  • wa =

' t

' i "J NLJI1BM fI4 1IH i i i l

  • 1 i l s l
  • I
  • l  ! i ,

l i

\

4

. . l

?

i a

FIGURE 4.3 -

I i

7 i

=

.-e-l l

l

_ ___ ,_.m. _ . _ . . _ _ _ . . _ _ __ _ , _ _ __

, - , - -. 9 - - ,

7 p-- ,

.. ., i i i i ,

{

i T j 4 6 6 . i e i e

I I y 9 l l i  ! t 3

6 i  ; g ,

1 . ,

I g s s i i 6  :

e -

I i 4  :

l i sii i 3 .= . ..

  1. 99  !

y e .

I E -

r e  ! i i 6

i i '

O I E  !

i i

- = ,

ur .

i '  !  ! En i a m i ' S i I

  • I e <O 3 i 1 i w: I w .
  • 4 l --

~

l l i 4p i 7 i d i i i

% I i.

i i

! a p r

3 E i i

i ,

3 i k s'E I t e + i

, i l

. 8 i e

l se  !

I

' I a e

e ~

I i

r. .

I U a Y g3 i

i

  • l F r i

i

_Z:  ! _ i e

me w 5 1 op . i e ,= I . t ,

i um I i w y W I as t

. .m a_m i" . *

. w sg i i w i 1 . -y a E .

rik ; '

i .s

= i f i r

3 s4-

. h t. e

, 8 i i W 5 I- e t J = lE 6

i 3" g

, . . . W.

I

-l

  • im - - ,

W lM a e 6

'O  ;>. uns

, i t i i , i La~

r y e s ee -

i <

i G EE 't Z a e i * ' ! 'i +=e -

s s i; d. e nf" i t p == <D e i5 , i

'. t a -

= - ,_ i

' . . -en n

, P- tE 8 i 1 i

! he iX e i 6 4 0 - W _

EE lb

- L C 3- i t

  • g .  := j i e i * . al - '

i F lb ' 9" E '

r '" 1E , i i

  • f __
  • i t

y *y  !

i i 4 D

.J i E B i f r

i 1

.p- i $

, ' y . es '

7 3 J-

' r a

5 el t t

ye aw

    • r i -  ! #: .

1 es sa qu em y

( i _#

l

, x i

De

  • e r ff
  • I . I i i

i

,.

  • i se x 3 E si I . _ ,

l )I t W i

"" - E ,, . _ _ a3 }

i er E in E -

' ' f '

f W .A hs e  ! *- +

I 1: i Ja y a __ e 1

! l

" i i L e. = ;. _

D i i GI i a == ha sg iy s I ==

a sa- en a .e A "a t i E. . E *  ! "' t .

1 ' i I s en - u aan ]  !

i !T I 4 p- e,, e i F- -

=

2- =. -

i i

1 a f

l

\

L e

, l l T ,

i 6

I a i ' t 7 I  !

e .

e a e e __ m

)

vu 6 UW'6 U6 6 ys U yr U UU U

. . Us 6 Ur 6 1

1 fl ~NLIIIMMJ I J'1 IM t

i i t I i i i

1 I f I r

i I t f

i

  • I 1 l, , ,

, i, ,

, i i - i

,  ?

n i

4 FIGURE 4.4 ,

1 i

l 3  ! r I *

[ l i

i i e r

g I , I

MAXIMUM SSE SEISMIC RESPONSE SPECTRA FOR DIESEL GENERATOR BUILDING ELEVATION 733-775 EQUIPMENT DAMPING EAST-WEST VERTICAL NORTH-SOUTH (X-AXIS) (Y-AXIS) (Z-AXIS) 1/2 % 6.64 g 4.70 g 6.22 g .

4 1% 4.93 g 3.54 g 5.23 g 3% 3.00,g 2.05 g 3.17 g l

l' l- 4% 2.49 g 1.75 g 2.71 g l

I o .

l FIGURE 4.5 l

l l

l

~

G -

N "2 -

Base Spectra Horizontal North-South

'*

  • lll Broadening (Hg)

Envelope g 3 / Calculated from g Time-History Na l Floor 1 Spectra Horizontal North-South

, r Floor 2

, 3 I

G l Floor 1 l

l .

l l

Ng Envelope Spectra Horizontal North-South CONSERVATISM IN SEISMIC RESPONSE SPECTRA FIGURE 4,6

_ ~.

7 ..

TYPES OF DAMPING

- STRUCTURAL DAMPING (5 - 10% CRITICAL)

SLIP 1' AGE ON SUPPORTS, JOINT SLIPPAGE AND FRICTIONAL EFFECTS.

- IMPACT DAMPING (5 - 10% CRITICAL)

IMPACT OR BANGING IN THE CLOSING OF GAPS IN SUPPORTS.

- MATERIAL DAMPING (0.04 - C.2% CRITICAL)

HYSTERSIS ENERGY LOSS FIGURE 5.1

HISTORIC DEVETAPMENT OF DAMPING PERCENT OF SOURCE YEAR CRITICAL DAMPING LIMIT 4

HOUSNER 1963 0.5 -

NElMARK 1969 0.5 <1/4 Sy 2 0.5 TO 1.0 <1/2 Sy i @ 2.0 4. Sy g 5.0 > Sy

+

. ui g REG. CUIDE 1.61 1973 1.0 OBE - PIPE 6 12" (INTERIM LOWER 2.0. OBE- > 12" 2.0 SSE - PIPE f 12" 3.0 SSE- 7 12" NEWMARK AND 1978 1.0 TO 2.0 $ 1/2 Sy
HALL 2.0 TO 3.0 f Sy 1

4

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

l

, COMPARISON OF DAMPING VALUES.

IN PIPING SYSTEMS i

l

, SOURCE PEkCENT OF CRITICAL DAMPING l

i j OBE SSE j y ( 67 YIELD) (.9 YIELD) 3 s '

O

. C j g BEAVER VALLEY 2 ANALYSIS 0.5 1.0 l w

REC. CUIDE 1.61 j LARCE PIPING (>12") 2.0 3.0 l SMALL PIPING 1.0 2.0 1

3 l MEASURED + NORMALIZED 10.0 12.7 i REF. 2 - STEVENSON'S PAPER 1

I MEASURED 3.3 TO 23.6 --

i REF.11 - JOYO TEST MEASURED 10 TO 20 ---

REF. 9 - SMALL BORE TEST l MEASURED I

REF. 10 - SNAPBACK TEST 5 to 11 --

i l

l i

4 f

i 3M G E Q

PVRC TG on DAMPING Ym OE3 En i; 3=

(Consen sus Pos erso n from E APlelL. Jl )12. ,1983 Meenm) 3 1 -

! 5 '

1 i

ReconssoATia4 - Revision I

! _ $3E. ano GBE - Au Diam ETER.S I R.tG. Guipe. l.Gil 4

g

'. SSE - D y it.*u. D

! h O m p T I E c

al $

N U R I E

y. q
c. .

g, 3 3

I REG. GusDE I.Q m

SSE - D c \'t '* j OBE - D 7 It'"- N

[t a

i E REG. Guapg!L j.C l 1 C osE. - o a it 'a-I $

i ______ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ .

SSE- SAFE M Dotald E stru ou m a.

OBe - OPEKMTNG Basus Emp.vHQomen.

D - Pi PE. De AME.TE A O -

I o 5 10 15 Zo ZS 30 l Fft EQu EMY (D - I4&. g (viz, te>s

y y::- e .

iP 2 # NX (

. n

. / k, i *#5 I e ee #

<#a 6 x.

d v u

! r )

t E Q 4 0 INX t

/

j REACTOR VESSEL k "

e DENOTES ELEAIENT N00E POINT o e OENGTES LUAIPE0 RIASS - ' "

~ OENGTESSNUSSER W OENGTES RIGIO ROD FIGURE!-IDEALIZATION OF PROTOTYP!C P!P!NG LEG l

G. M. Hubert - LMFBR Study Ref. (11)

FIGURE 6.1

a e TABLE 1. MAXDEUM REACTOR VISSEL NOZZLE LOADS CALCULATED '

USING TWO ANALYTICAL METHOOS Namde Lead Ammiyded F. F, F. .% M, M,

Meshed Es Es Es Es es Ee as Ka a Response speanus Assiysis s74o sco slas 3:30s os30s s9100 Medal Superpeessies $000 6940 47ll0 3::700 54400 73800 l Tlass Missory Aanlysis

% Di5srense Between Methods 42 is 42 IS 42 II Response spectra for response spectrum analysis are smoothened and peak broadened.

Time histories for modal superposition time history analysis include 10% expansion and contraction.

l G. M. Hubert - IJ4FBR Study Ref. (11)

FIGURE 6.2 1

1

- ,---,--------e -- ,-

e,-,,,+-- my- y--4,o _,,__,wmyp_ m,- w wwwwwwwmmmm -~ w, , mm m me www

1000 000 -

000 -

ge 700 -

PIPNIS NATURAL PREGUENCIES e

800 g

i = -

{ DEmsN RESPONSE SPECTRA

= 400 -

5 consrette REsetest 3 age _ srtcTRA

, A_ ,

m -

100 g i i i I I I I I I O 5 II 15 20 25 30 35 40 45 50 FREQUENCY,NZ

-COMPARISON OF DESIGN RESPONSE SPECTRA AND COMPUTED RESPONSE SPECTRA OF VERTICAL EARTHQUAKE SHOCK FIGURE 6.3 MODAL CONTRIBUTION FROM 12.3I Ha MODE TO REACTOR YESSEL NOZZLE RL8PONSE i

Nossie Loed Analysiest F. F, F, M. M, M, Method Kg Kg .Kg Kg.m Ka m Es-m Response Spectrum Analysis 1880 4420 99 990 900 39900 Medal t;- , Jh 1410 3330 75 750 600 30000 Tune History Analyss

% DiNemace Between Methods 25 25 24 24 24 25 G

0%

FIGURE 6.4

r-l t .

I

!'

  • R
  • RIGID STRUT

-+s - astenasseAs.senman

  • ~ 8ecoAL asAss g j

s a

35 g S* #

S d

8 Yi '

A R 39 !3 .

SMALL BORE PIPING SEISMIC TEST FINDINGS

- L. K. SEVERUD ET AL REF. (7) l l

l I

FIGURE 6.5 l

I

- - - - , . . . - . ~ . .

. . , _ _ , - . , , , - - , , _ _ , , _ _ . _ , . .__.-,.,.n-

Table 3: Setemte Suppers Imede wth ammbhare. Ias I

E43155 EME M EFECTRA T3M DWetIB ABE2EIS TWF E0. 30. RE 1 2 3 h 5 6 7 8 E - 11 k-I 109 7 39 7 55.0 90 18.3 18.9 18.6 6.8 10 7-Y hh.7 26.1 17.9 .o 5.7 6.5 6.0 h.6 10 7-2 270 5 262.6 236.1 115.9 uT.5 114.7 123.8 33 3 9 9-1 233.2 232.6 212.8 96 3 98.h 105.2 103.6 h5 0 8 11 - 3 537 7 320.8 228.9 123.8 132.7 152.8 1h2 5 58.8 7 16 - I 337.1 187.8 1h7.3 19 1 31.6 67.6 34.7 1.5 6 17 - E 671.7 h55.1 389 7 166.3 2h1.7 227.7 265.9 115.6 5 20 - 2 286.5 231 5 101.k 9.6 12.0 21.0 12.6 8.h  !

5 30 - T 318.9 230.5 166.1 .o 8.h 19.9 10.1 3.2 h 23 - 2 112 9 72.7 65 3 .o 25 5 51.5 31 3 19.1 2 ap - Y 213.5 145 7 79 9 15.8 13.8 16.3 16.h 25 2 39 - 689 3 shh.h 3o1.8 las.7 las.o laf.h 1s6.5 50.0 1 31 - Y 215.8 176.7 75 7 18.0 m.2 28.1 23 9 118 5* i E- 1 31 - 1 335 5 3R3.8 1h5 5 143.6 165.6 160.6 15s.3 76.1

1. .e se - e.al,es...m.m.ie.e,ab.nl.a. - et ent .
2. Respamse spectro ens 3pois - medal reopeases e w assereias to U. S. EC Begnaaeer. Omide 158.
3. Boopenas speetes emnlyste - medal responsee combinet ty SW aothee.
h. ftmo esmata aanlysis - 38 eyesen despias, sashbor esmpias, emebber seps = .R5 insh.
5. Same as ama 6 emosyt emmbber gope = 0.
6. Taae esmata emelysis - samhbor seps = 0. 5 systen emoptag, emubber descans = 58 et tasal.
7. ,15se esenta esalpets - emmbhor gaps = 0. 38 ersten deeptag omshher deessas = 308 et tesel.
8. sees resnite (vita sembbere).
  • ante response umaerprestated ny eentrees. cause umbanus, se he 1--- _c *----'

in funere testieg ens aanlyses.

1 L8 = 0.22481 N SMALL BORE PIPING SEISMIC TEST FINDINGS

- L. K. SEVERUD ET AL

  • REF. (7)

FIGURE 6.6  :

400 --

350 -

i j 300 -

.f E r[ T250 -

.a $ in , TEST RESULTS U 03 k -

E N8 O 200

RESPONSE E E m $PECTRA ANALY.

e na e sRss cosa p5 o g i 150 -

MAX NON UNEAR

.A 3 q E ANALY.

gj k .

wo - -

g 50 -

o - l O .L L .L .L _

2 S 6 8 9

$UPPORT NO.

. .. ... .a ,

4 FIFIBG STMDS AgALIZED Three sizes of pipias systems were analFsed e b is. (10 cm) pipe 11ae, see Figure k e16 in. (k1 as) e 28 is. (71 cm) pipelias, see Figure 5 pipeline see Figure 6.

Se b is. (10.es)' pipe sFoten has ik seismic supportet 7 rigid vertical supports and T horisostal restraints using mall snubbers. The 16 in. (k1 cm) pipe system has a total of 19.seisnio supports with medium to large size asehanical saubbers. At same inestions, two saubbers are used to provide re-straint is a stasle direction. The 28 in. (71 cm) piping has 13 seismic sup-ports and the saubbers range from meditas to large in size.

l l

! u.ev l net I e ,M 7V Y IU'I

l. 11 n.ag l 3 MW 4

1 3 #

N.gr # 17 o W 2 m ISOUTMI B

g g,3y M N 2Y

w k N

yB H-3E M *g H-1Y g 3 i g L M.1X l

  • M ,,

l

=

.j, 3

- ~

i .

/

Figare k. b Inch Pipeline Modei Seismic analysis of piping with non-linear supports -

l (D. A. Barta et al Ref. 6) '

FIGURE 6.8 I

I E. ,

y n'

=

  • w a v n

m 11 8 8EX #

, Ms v a

. cuv Maa v ',

7/ ' a see cuw wez , '*

wa.x -

a m.x n a

4 i g yam

  • 9e l

"* isovs'n g- FIGURE 6.9

_ ,. e_.

l n eere 5. 16 Zach Pipeline Model i H-7Y H 5Y d M-7X BELLOWS a H 9Y U H-12Z "g O H-11X H-ex @

Z

  • ym/m E

moLATion VALVE x

g pgg LEGEND:

O - aoolau=BEa gg y Q = ELBOWNUMBER VE8SEL 11 28 Inch Pipeline Model l

l t

l l

Seismic analysis of piping with non-linear supports (D. A. Barta et al Ref. 6)

FIGURE 6.10

TAsta 3. 4 INCE FIFELINE

'm s La. (1 13r = 6,648 s)

EA303 puu 7 Ituu 8 RUu 9 Rtal 10 M 11 RUN 12 pus 13

30. RUN 1 puli 2 puu 3 aus b muu 5 aus 6 39 38 - 39 39 39 38 39 28 5 91 9h 13h 107 39 39 0 16 0 0 0 0 0 0 72 109 E - 8E 62 168 105 66 50 51 51 50 51 53 5 TY TT 192 lbh 51 51 51 k80 518 89 151 70 70 65 79 I . 6E Se9 966 h31 TO 107 26 39 30 33 3h b3 I - ST 105 125 91 33 33 33 33 68 23 0 0 0 0 198 199 5 - 58 h39 M8 200 0 33 22 22 25 35 E ho 67 152 113 23 m 22 22 19 25 0 128 21b 0 13 0 0 0 0 0 I ha 191 a69 156 31 25 36 29 31 31 36 E sr 73- 175 115 31 31 31 se 9 13 0 0 0 0 97 9h E 35 St 176 99 o h4 38 55 - h6 67 55 63 5 . Er 36 270 all h6 he he 0 0 0 0 0 0 84 113 E - EE 79 183 1he o 13 l

ho 65 15 23 16 26 210 261 E - 11 1ht 359 180 30 60 a 18 22 21 21 h5 5.u Sb 131 8e at ta at 25 laCRTION MIDESS FIFINO STM

  • PWI (1 Es! = 6.89 es) 12.1 33 505.* 1 50fM 1 FIFWD emalreis with rigid supporte med notas envelopias design setemic spectre of Figures 10 eed 11.
3. FIP M emelysis with support est elamp flesittlity and uslag envelopieg setente spectra.
3. ANTE undel with support and elemp flesittlity same as la NUE 2 tut using celeslated seisele spectra.
h. Ben 11aser, time demain aanlyste with support est elamp flesittlity same. es in NUN 3 and contined with I

seehher Stiffnese end demping eheretteristics.

Smehter sepe = .030 ta. (.wm em).

S. same es aus b not with sambhar gepe = .005 io. (.013 cm).

i '

6. Same es Run h tut with sadher gape = .015 to. (.038 es).

l I

T. See es M h tot with emeMer damples reduced 505.

8. See as Bug b tut with eW stiffmoos doeressed 305.
9. Some es aus k tot with elamp stiffuses inerseeed 305
10. same es suu h but with time history eaapressed 105.

u . same es aus b not wi m time hantery empended 108.

13. Some es M h but wi e emetter de W eg sealed la propertion to the emerer aeroes the enutter.
13. Struetural parametere same as la M 3 but with spettal eempeeeste of eeteste action applied seperately e a e,.41.t res,sesse e mtaea eseeeetag to seem1** err Ge14e 192. Artiele 2.2.

Seismic analysis of piping with non-linear supports (D. A. Barta et al Ref. 6)

FIGURE 6.11

"-'* -- =n-- ww ,.mm- _ _ , _ , , _ _

F Sagts 4. 16 Ise Pts m aus asISEC MIase suppear taas s ta. (1 tar e 6.668 m) spron? LIFRam acuLIssaa e, galyggs a.a. we M1 RUB f INE 3 Mh aus 5 333 6 mis ?

E.12 6861 91 95 87 M 131 1300 3 47 3367 104 1% 113 153 135 TM s.22 5076 969 1200 961 13e7 lap last

, E.3E 10706 569 4D 517 449 68h 907 e.37 3645 3s4 603 317 32 395 812 3.b2 29512 3165 uf0 3r*6 2036 2002 m.67 75456 3579 353 376 boe 366 396 5 5Y San aW3 0 0 0 0 0 N 51 12966 Pl?R tha 2139 3962 2173 2190 3968 a . Sar 60e6 0 0 17 4 15 See s . 541 9357 3337 3s62 3t36 3989 stop 5 63 3s6#

36525 3096 3F96 3719 4157 3-7I 6150 h16e 77312 1536 143 that 17e? 1672 4378 a.7Y 10661 1983 3E19 2189 3361 3 . TAE 89006 223r 3R14 3r75 33eh 3158 3B12 3 pop Sees 5 742 16440 aste a187 2153 3803 ass, b673 a . les 31664 de7a 99h5 6 sob 6168 6167 6881 FIFtBB FIFIBS em s pet (1 N1 e 6.8p sym)

E Mit 50.

1 16185 3 67 37W 6 138hh 33s7 3 sat 3eb3 3767 fila (b3 3573 3739 3714 4ess 8 38963 133s6 15353 14876

. 9 1946e 13hu 196d6 5880s mest andhb state se131 sanat 35pf6 11 6336 7931 kg6 7779 7989 2 87 918p 13 37teh 3ge37 33737 3her3 zw? 3 373 3sege as M sus 1. Firm entrees utth suggert tissualty, estas af W W estasse sportra est 15 mothed ament emmuties.

Em a . num14amme. time testeer emmapets esta asucher staffemme and easynes ease.

' estartseses. sumane sup e .ese sa (.896 en).

I asureen e& amp se6 femmes. BM tems Meter.

M 3. Sus as M 3 tut utth time Maampy sangremond 1M.

W 4. Amme as M S tut utth less bestery empmatet 3M.

M S . Bume ao M 3 tua utth alansma e& amp st&ffumes.

M 6. See as M 3 tut atta ammame elsup settflamme.

M 7 4m.mte.m

.t m . rensesses mar no1fWen asse.e um P.6,ese sue tv e= 4.n esse a. mages samataus elle w a,u.ie s.a.

a , sea.1.w.

i l

l

! seismic analysis of piping with non-linear supports (D. A. Barta et al Ref. 6)

FIGURE 6.12 l

EacEA S. 20 1 M F19eLIIB Rasem IAAD s Lee. (1 IM b.bbe e)

Ensem Lierge Ae4L?ste BSM.IseAs AAAL78te W. MS 1 MS 3 em ; MB b nit S les 6 mm 7 RW 8 BUB 9 Mm 10 pue it mei 12 5=2I 3M64 16eba 5806 Selb 8e59 Set 6 S431 330e 6100  !)387 115 # 10266 5 8I 9he6 9769 bS93 h667 6119 4583 b6ee 4376 begg 19752 8336 7062

! E . ahE. Shet6 13 eft Sehe Seat Spet Sof9 Seat Se39 5133 16M3 10773 11076

, E . att Oppe 13067 36 3 3bg6 7606 3630 3660 3364 3See 10619 6776 65 %

E=31 3eeps 13e1 M188 $675 36bt 5473 54N Self 56s6 32Ta$ 23603 26462 5.SY bebe 89t6 3657 MM 6491 3669 3666 3653 seat essa 5965 5676 5 65 79E9 fMAS S$3h 96bs utb3 3538 S$3s Sa66 5783 29300 22739 22216 5 7I 9996  !? Set 3371 23N b773 se66 3419 2158 3665 16eSS $962 SeTo

( 5 7T TSN Then its 17te 3fSe 1667 last 19th 14e1 Sehe 6h60 hatt i

5-93 h654 sestt 170s 18h6 3060 1719 1718 16eb 1811 19917 5415 79e6 l 5 9Y $98L 366h6 19e3 18T4 70eb 1931 1S8e lhes 1697 10W4 9033 4h92 a . us h8N 16stS sega sees este sete sepe 3669 3836 16135 11100 4066 a .15 SSee se670 3905 38f1 6973 3379 16ef 3616 3756 19809 12332 11536 j masu

  • Ptrise mass emees
  • NT (1 ret = 6.99 ePe) m.

l 1 70ge 30338 7513 TSee - 7110 7515 7764 76S1 3033e 15163 16712 I S 16814 175et 8617 836b - 863e eset tat 6 8776 21093 16730 10631 4 letts 147tp 63N 634 - SWa 6Ian Sepe 6565 293e6 16383 15338 7 8765 30512 S300 SBS - 3335 13ee Sept $16e 31106 13308 10861 le 13e8 seles 6865 Test 6866 6061 76be 766a 34536 19976 12603 12 80$b 193e8 Se37 hees _- Se33 Se37 hep $ Selb 16919 12354 106he NetBes reb 1. Ftrms 11aser emetrate with rooster veneel ledes flestM11ty but with riote pipe supporte. ~

M i . PtFM llaner amm1 pose with remeter veneet ledes flestM14ty and with flesitte pipe supporte.

aus 3. assus mantener emelpese with reester vessel ledes fleutbility, esernes elamp estrfeses, seusher test data et attfrases and damples eterestertettee, samther espe e 0.038 te (0.076 es).

M 4. Same se M 3 tut with emubber empo = 0.e11 le (e.830 cm).

M S . eens as 25 3 tut with M $ suggert ettffmesses and soukhor dampleen e 300 lb.eee/le (350 e=ese/en).

M 6 . Same as a s 3 but wath utsesem M ental enemy stiffheeses et s.7 and e-9.

M 7. Suas as M 3 tut with mentam Westet elemy estrfensees et 5 7 and e-9.

M S . Same me FW 3 but with Stes hietary e-pressed 108.

83 9 Same as mW 3 tut with tlas hietary empensed 108.

Seu Is emme es ses 3 but with enra damptes and 1.0 tiene the herf eestel aseelerettes/ ties history.

l aeroes the seuther.

M u.le.

as emmeem.se.Meen 10 but utthwith u == eeubhor damples opeust eseled w er ses-sedeum-ue.

Le properties .to the "e*Ty ewis 4 espe t. and egetsu ev.

- e= Mod eso.etes t. es elm y easde 1.w artiste z.n.

l 1

i Seismic analysis of piping with non-linear supports l (D. A. Barta et al Ref. 6) l l FIGURE 6.13 l

l l

r r

/

-~ -

f\

REACTOR VESSEL

,N. INTERMEDIATE HEAT i EXCHANGER ,

/ ,l / ,)

', ,/,[, i PIPE HANGER h' ,/// -

. /..

y' /s la!  !

L a \ '.

- = gl (. _

./

s y ,, ' * , 1

    • .. fkf
  • Q ,)

\ ,y' ~. 3 s  !!j

'/ b:!. j' '

PfPE SNUBBER

f i!! ;h (%(','
%

/ i 3ELows ...s f -

4.s, EXPANSiJN -

lQ M.)S. .

@ NT T,'I[;hs,,'

, ( j .

'-. . N., ,

s f(..

t 4

N.,.s'P-l N

\.y .[ - g ,

s

(

l

. f jf  %

se /

soo.uu LEAM

' :. s.'  %

DETECTOR  %-s.'N.. .'t N' %. , i l_

~. . -

SCHEMATI'C' VIEW OF THE JOYO PRIMARY HOT-LEG P PING SYSTEM Ref. 9 - In Situ Vibration Tests Joyo Planc, Japan FIGURE 6.14 r

. Z lHX Inlet Nonle

, Reactor Vessel Outlet Nonle @

x 4 , ,

C<-

q .tC CM 1 X,Y,Z X,Y,Z y Inner Pipe 20 inches Outer Pipe 22 inches q]

w 41 4( "

5 C l

!C D

2 X,Y,Z Jk g

p Cy X(,Y,Z v'

! X,Y,Z Ce#

l h wConstant Hanger f

q% Electro-Magnetic

@ Mechanical Snubber J

+ Lug Point Exciter f'

+ Spacer Point e outer Pipe sellows Sensor

. t PIPING TESTED ON THE PRIMARY LOOP lef. 9 - In Situ Vibration Tests Joyo Planc, Japan FIGURE 6.15 -

! Test Result Summary of PHTS In-situ Vibration Test With Analytical' Data Eigenfrequency Hz .._ _ .._ .. . . . _.__ __. Damping F actor, % .

.... . Analytical Result _."_3 _ . . _ _ . _ . . . . _

Dynamic Spring Constant of Snubber. kg/aun r* .i No Design'8

. Test . Design Mode Result Struts K=50t. K=je00. K=2g10. K=3000. K=4000 Rigid yalue. . _ Test Result __ yalue. Serial ft 7

li' --

8.40.835.03 6.26 7.40 7.94 8.27 13.45 - -

I.0 I j y -

8.8 0.93 5.08 6.89 8.77 9.52 9.85 14.42 - -

2 gg -

9.2 2.01 5.46 7.22 9.68 11.29 12.32 15.98 -

10.5~15.9 3 l 2R --

10.0 2.59 7.07 8.69 10.76 12.05 12.88

  • 16.90 -

5.4~23.6 4

-n 8 --

10.8 4.52 8.43 10.87 12.21 13.53 l4.72 18.68 -

8.9~15.9 5 i 1 12.8 6.10 9.14 11.1613.85 15.0015.6219.11 10.79 4.6~ 9.2 6 h2 13.6 7.01 9.7011.8915.0216.11 17.07 26.0513.90 5.7~ 9.3 7

.[ --

14.4 11.08 11.49 12.51 15.27 17.08 17.97 27.56 -

4.4~10.1 8 g 3 14.8 I4.01 14.6314.9816.1417.9519.46 35.8516.17 4.9~ 8.8 9 k4 16.815.4216.1317.0318.70 20.7122.03 39.1616.34 5.6~ 7.4 10 1 -

18.8 18.24 18.57 18.94 19.79 20.83 22.69 39.86 - -

11 5 19.218.6519.0019.52 21.13 23.07 24.12 42.3819.28 3.3~ 4.7 12 20.4 19.88.20.29 20.81 22.06 23.31 25.27 48.79 1.0 .. _ .

13 .

Note) *1 220'C. RUN 10, *2 500'C

Y d \ h 4.51Y

( ):

I X l' 4,32g l 4.m z' l

l l

) () 8 3,31 8 4-te v

1 0* "

i.28-2.21 2 0

2.27' BENCm(ARK PROBLDi NO.1 P ELBOW STUDY-NOVGARD BEND MODIFIED MODEL

i FIGURE 7.1

..,,,_,..--.-..-,_._.,.r.-m--- , , , . . _ , . . _,,s, , , , ,,,,,,,,_,,,,.m.- , , . - . , . . . , _ - , _-,,_m,_..m,_m

r t

4 Z/\X d

' b &s 5 W

3r.,

MD AX, AY, AZ 6 6 MD d h .

s  !$ MD dY ,"

m t- unen '

h 4 4 hg W

9 RGD AX ow & MD AY e e @- #2 v.ssa

  • 4

. MD AX

,,4c ,

" ANCHOR

^2 ,

~

NJ

  • v' ,,

G

a. ,

b MO($X,$Y,4Z w ..

S 4MnN4 K

l DIMENSIONS IN FEET MODIFIED MODEL NO.3 FIGURE 7.2 i

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

1 aY i >

i m vx -

i z

!4 i

4 a

iz4

'28 ist). 84 4,

44 4u g

ns y es ,

4r og s ,

! )!,

srmm my 5fm" m77 ' " "

, 's '

4" M

'5 y

wg 2-2 3 Q o

n

!

  • n i n

,34 i its @ @ 3 i a m is

~ . RE . _

g its _

is ,

ses sa' % 6 **

sogse so @ g zrgto , is ,,

! D#

RGD AX, AY, Al N DA f 4x,4Y,,4E i

i BENCHM ARK PROBLEM N2 7 I

FIGURE 7. 3 l

I

_, . .StV M#

~ A A G./

%._ . m. 4. Z.. A Z. .L..i ._ .

32,0 g (wg) ,

so 6'-o" 4'- E" 130 A - S '- o" iss g em no i

ni- .SS -

87 0 d

_ ,e its m

89 0 405

- 5 m

= , .

e 8-4.%

i

.Ja lf aso' *

\ am

  • J. .

.. 7. . .

4045 4 44sEn 'sio fik Mo

]sas ses 'd' "t$ ' 'sd l 32'* O geo< >

WWW e,

  • y ti rio ~ 4-2" '
  • u EL-St-p m <

h

,cm , ,

8 8 250 aos iges a ssa 93 seesb # w 8 sos. 6 808 #6-C 28 6" ear tosa p 3.o. ,

eso f / 8'-O" l 3a0 4 '.g 0 325 m41 un us o m

34cdb I 34sdb eos 3 son a eso i n m eso ses 360<>

365

37) o" 3es e *as sw a ,, , ,

[,,**s 24" PIPING STUDY MODEL

=o*3 e estagrN '

l g iso 3esen 4.o' l

l FIGURE 7.4 l

I COMPARISON OF NATURAL FREQUENCIES (HZ)

MODIFIED BENCHMARK MODEL NO. 1 MODE STANDAP.D FITTINGS HEAVY FITTINGS EXTRA HEAVY 1 11.58 11.59 11.09 2 19.25 21.10 21.33 3 27.18 27.07 26.07 4 59.44 60,91 61.17 5 60.05 62.73 63.35 6 90.37 93,43 94.63 7 90.59 96.61 99.69 NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

Figure 7.5 c

r COMPARISdNOFNATURALFREQUENCIES(HZ)

MODIFIED MODEL NO. 3

/

-MODE STANDAP.D FITTINGS HEAVY FITTINGS EXTRA HEAVY l- 3.38 3.69 3.71 2 4.19 5.02 5.26 3 9.55 10.71 10.97 4 13.59 14.33 14.45 5 15.01 17.10 17.47 6 17.90 17.96 17.87 7 19.78 22.19 22.68 8 38.73 42.24 42.91 9 48.32' 48.02 48.22

~

NOTES: Heavy fittings are twice as thick as standard fittings.

l Extra heavy fittings are three times as thick as standard fittings.

FIGURE 7.6 l

i

. . - . . . . _ _ _ . . . _ _ ,m . _ . . . _ , . . _ , . . ,,- ._ . -

i

-]

COMPARISON OF NATURAL FREQUENCIES (HZ)

BENCHMARK MODEL NO. 7 MODE STANDARD FITTINGS HEAVY FITTINGS EXTRA HEAVY ,

l 1 5.02 6.05 6.22 2 7.80 8.44 8.46 3 8.18 8.92 8.96 4 8.96 9.99 10.04 5 9.30 10.20 10.25 6 9.88 10.77 10.80

- 7 13.19 15.16 15.35 8 14.90 15.91 15.79 9 15.03 16.02 16.06 10 17.72 18.30 18.23 11 18.05 21.94 22.74 12 22.83 24.05 24.03 13 24.96 26.17 26.00 14 25.62 26.77 26.80 15 26.82 28.73 28.83 16 28.05 .30.58 30.61 17 30.10 32.25 32.42 18 35.11 36.59 36.44

-19 36.99 38.34 38.20 20 42.49 44.70 44.42 21 44.24 48.25 48.95 22 48.00 49.72 49.58 NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

FIGURE 7.7

24" PIPING STUDY COMPARISON OF NATURAL FREOUENCIES (HZ)

MODE STANDARD FITTINGS HEAVY FITTINGS EXTRA HEAVY 1 6.816 7.294 7.544 2 9.838 10.666 10.598 3 10.415 11.191 11.510 4 11.362 11.739 11.788 5 13.282 13.372 13.231 6 13.525 13.545 13.435 7 13.719 13.790 13.708 8 14.866 15.466 15.576 9 15.570 15.736 15.678 10 15.634 15.912 15.888 NOTES: Heavy fittings are twice as thick as standard fittings.

I Extra heavy fittings are three times as thick as standard fittings.

Figure 7.8 i

,..,-,,-.,-,,,--,e--- ~ - ' * ' " "

  • I l

4 COMPARISON OF SELECTED THERMAL AND DEADWEIGHT STRESSES (PSI)

MODIFIED BENCHMARK MODEL NO. 1 MEMBER STANDARD HEAVY EXTRA LOADING NODE TYPE FITTINGS FITTINGS HEAVY 1 ANCHOR 10547 l 13260 l 14466

201 RUN 1504 3996 5223 5 ELBOW l15237l 9367 6053 U

5 RUN 6020 11133 13412 9 ELBOW 11669 5619 3401 1 ANCHOR 66 101 126 201 RUN 107 l 174} 217

.E GHT 5 ELBOW l 179 l 105 88 5 RUN 94 =166 216 9 ELBOW 64 56 49 l

l NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

l Figure 7.9 e

i I

COMPARISON OF SELECTED SEISMIC STRESSES (PSI)

MODIFIED BENCHMARK MODEL NO. 1 FLAT ARS MEMBER EQUIPMENT STANDARD HEAVY EXTRA NODE TYPE DAMPING FITTINGS FITTINGS HEAVY l' ANCHOR I9093l 9073 9671 201 RUN 1955 2360 2691 5 ELBOW 1% 3943 1752 1330 5 RUN (5G) 2077 2777 3261 9 ELBOW 5029 2718 2146 1 ANCHOR 5456 l 5444 l 5803 201 RUN 3g 1173 1416 1614 5 ELBOW 0 '

(3g) 5 RUN 1246 1666 1957 9 ELBOW 3017 1631 l 1288 i l NOTES:

Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

e Figure 7.10 i

k _

___=

COMPARISON OF SELECTED SEISMIC STRESSES (PSI)

MODIFIED BENCHMARK MODEL 1 BVPS-2 ARS (FIG. 4.3 & 4.4)

MEMBER EQUIPMENT STANDARD HEAVY EXTRA NODE TYPE DAMPING FITTINGS FITTINGS FEAVY 1 ANCHOR 848,' 841 908 201 RUN 175 205 234 5 ELBOW 362 157 119 5 RUN 191 248 292 9 ELBOW 467 242 190 1 ANCHOR 815 1808l 877

'201 RUN 169 198 227 5 ELBOW 3% 348 151 115 5- RUN 183 239 283 l

9 ELBOW 449 235 185 NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

l-Figure 7,11

! _ _ ~, ._ - . _ . _ . , . . . , _ . - - . _ . _ . . . . - . . . _ . _ _ _ _ _ _ _ . _ . _ . - _ _ . - . _ _ _ _ _ _ _ _ _ . . _ . . _ - - _

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODIFIED BENCIE! ARK MODEL NO.1, NODE 1 FORCES, LBS. MOMENTS, FT-LB ON X Y Z k b Z STANDARD -1044 -924 -338 -2593 902 7512 THERMAL HEAVY -1544 -1517 -527 -3922 793 9224

-1763 -1762 -612 -4483 724 9986 i

STANDARD -12 -309 3 -27 -23 36 DEAD HEAW -21 -347- 4 -42 -28 58 WEIGHT

-27 -377 4 -52 -32 74 NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD F;TTINGS.

Figure 7 12 (1 of 2) o 9

c 3 l

l l

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODIFIED BENCHMARK MODEL NO. 1, NODE 11 LOADING FITTING FORCES, 135. MOMENTS, FT-L3 CONDITION TYPE I

X Y Z b b Z STANDARD 1044 924 338 4458 -4577 -3221 THERMAL HEAVY 1545 1517 527 7320 -5981 -4239 1763 1762 612 8430 -6559 -4555 STANDARD 12 -177 -3 -697 -82 -20 l DEAD HEAVT 21 -188 -4 -732 -141 -56 l

WEIGHT l

EXTRA t HEAVY 27 -202 -4 -798 -177 -77 l

l NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

1 EXTRA HEAVY FITTINGS ARE THREE TI.ES AS THICK AS STANDARD FITTINGS.

Figure 7.12 (2 of' 2)

-.._........_.,___....____m,.,_ . , _ _ . . _ . - , _ _ . _ . _ _ , . _ _ . _ _ _ _ _ , . - . . . . _ . . , . . _

7-I COMPARISON SEISMIC ANCHOR LOADS MODIFIED BENCHMARK MODEL NO. 1, NODE 1

. FLAT ARS FORCES, LBS. MOMENIS, FI-LB g

Fg T 7

Fg g y Mg -l STANDARD 482 288 798 6217 1099 2774 1%

EEAVY 544 359 815 6001 1399 3061 (SG) 595 419 858 6318 1590 3367 i

STANDARD 289 173 479 3730 659 1664 3%

9 0 839 1837 (3g)

EXTRA 357 251 515 3791 954 2020 l

j HEAVY l

NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7.13 (1 of 2) t

F=

COMPARISON OF SEISMIC ANCHOR LOADS MODIFIED BENCIH4 ARK MODEL NO.1, NODE 11 FLAT ARS FI FORCES, LBS. M0MENTS, FT-L3 E

I I

F 7

Fg g y Mg STANDARD 1289 658 535 3546 6681 305 1% HEAVT 1432 727 707 3980 7010 424 J

(SG)

EXTRA 1579 808 809 4475 7619 526 HEAVT

. STANDARD 774 395 321 2128 4009 183 HEAVT 859 436 424 2388 4206 254 (3g) _ _- - - ---J EXTRA 947 485 486 2685 4571 316 HEAVY NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7.13 (2 of 2)

J e

9

-.---v. - - - - - . - - , - . - - - , - - , - . - - - ,.-.,,-m,.--w.cr-r_-.c.,-.%w..ee,rwww,.,,w.-,w -,1--,,. v..

COMPARISON OF ANCIIOR DESIGN LOADS MODIFIED BENCHMARK MODEL NO. 1, NODE 1 (DL + THER + FLAT ARS) .

NG FORCES, LBS. MOMENTS, FT-LB DAMPING Fg F 7

Fg g y Mg I

STANDARD -1538 -1521 -1133 -8837 1978 10322 1% HEAVY -2109 -2223 -1338 -9965 2164 12343 (5g)

-2385 -2558 -1466 -10853 2282 13427 STANDARD -1345 -1406 -814 -6350 1538 9212 3% HEAVY -1892 -2080 -1012 -7564 1604 11119 (38)

EXTRA -2147 -2390 -1123 -8326 1646 12080 HEAVY l

i NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

1 I

l Figure 7.14 (1 of 2) l l

l

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

F COMPARISON OF ANCHOR DESIGN LOADS MODIFIED BENCHMARK MODEL NO. 1, NODE 11 (DL + THER + FLAT ARS)

FORCES, LBS. MOMENTS, FT-LB D

Fg F T Z b b Z I 2345 1405 870 7307 -11340 -3546 STANDARD 1% REAW 2997 2056 1230 10568 -13132 -4719

, (SG)

EETRA 12107 -14355 -5158 3369 2368 1417 HEAW l

' -3424 4

1830 1142 656 5889 -8668 STANDARD 1765 947 8976 -10328 -4559 3% gEAW 2424 (3g)

EXTRA 2737 2045 1094 10317 -11307 -4948 i

HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD TITTINGS.

l EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STNiDARD TI* TINGS.

O i

Figure 7.14 (2 of 2) 4 O

.,..,n-- , - - - . - -

7 COMPARISON OF SELECTED THERMAL AND DEADWEIGHT STRESSES (PSIl MODIFIED MODEL NO. 3 MEMBER STANDARD HEAW EXTRA LOADING NODE TYPE FITTINGS FITTINGS HEAVY 1 ANCHOR 303 464 521 3 ELBOW 355 261 213 THERMAL 11 RUN l 2965 l 3159 3233 19 ELBOW 2382 1391 903 1 ANCHOR 163 250 305 DEAD 3 ELBOW 767 416 274 WEIGHT 11 RUN l 1306 l 1283 1338

'19 ELBOW 423 236 171 i

NOTES: Heavy fittings are twice as thi,ck as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

l

! FIGURE 7.15 i

1 l

m; CONPARISON OF SELECTED SSE SEISMIC STRESSES (PSI)

MDDIFIED MDDEL NO. 3 BVPS-2 ARS (FIG. 4.3 & 4.4).

MEMBER EQUIPMENT STANDARD HEAW EXTRA NODE TYPE DAMPING FITTINGS FITTINGS HEAW 1 ANCHOR 12615 12132 12175 3 ELBOW l 13573 l 5398 3404 11- RUN 8363 9402 9911 19 ELBOW 5513 3395 2262 1 ANCHOR 7715 l 7537 l 7567 3 ELBOW 3g 8386 3366 2120 11 RUN 5221 5893 6218 19 ELBOW 3364 2108 1405 NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three tiros as thick as standard fittings.

FIGURE 7.16 i

9

RESTRAINT THERMAL & DEAD WEIGHT LOADS (lbs.)

FOR MODIFIED MODEL NO. 3 LOADING FITTING TYPE F F F F F y y x z x STANDARD -396 -505 1172 -2004 -220 THERMAL HEAW -/21

-606 +1496 -2850 -252 EXTRA -881 -651 1655 -3268 -268 HEAW DEAD STANDARD -6 33 -627 27 -14 WEIGHT HEAW -10 60 -625 -20 -20 EXTRA HEAW -13 75 -653 -36 -23 NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDAPS FITTINGS.

EXTRA HEAW FITTINGS ARE THPEE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.17

s COMPARISON OF SEISMIC [SSE]

RESTRAINT LOADS [1bs.] MODIFIED MODEL No. 3 EVPS-2 ABC (FIC. 4.3 & 4.4)

EQUIPMENT FITTING NODE NODE NODE NODE NODE DAMPING TYPE 7 9 11 13 15 F F, x y y STANDARD 2175 1103 2251 1355 86 1%

BEAW 2368 1405 2612 1641 92 2536 1549 2766 1753 103 STANDARD 1333 680 1421 853 59 3%

"" '1474 877 1645 1031 62 EXTRA 1579 966 1745 1103 68 BEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

- EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.18

F RESTRAINT DESISN LOADS [1bs.]

FOR MODIFIED MODEL No. 3 (DL + ten + ssrT)

EQUIPMENT FITTING NODE NODE NODE NODE NODE D.4MPING TYPE 7 9 11 13 15 t

F, P, F F F,

, 7 y STANDARD 2577 1575 2878 3332 320 1

1%

HEAW 3099 1951 3483 4511 364 EXTRA 3430 2125 3768 5057 394 HEAW STANDARD 1735 1152 2048 2830 293 3%

HEAW 2205 1423 2516 3901 334 2473 1525 2747 4407 359 H

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l i

FIGURE 7.19 l

l i

I L

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODIFIED MODEL NO. 3 NODE 1 ggg;,, rgg c FO cES t.S. MOME rS, FT-to F

X Y Z b b Z STANDARD

- - - -60 -221 -12 THERMAL HEAVY - - - -57 -347 -12

-53 -392 -12 STANDARD -29 -47 111 DEAD HEAW -20 -44 183 WEIGHT EXTRA -16 -45 226 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

e FIGURE 7.20 (10F 3)

CONFARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODIFIED MODEL No. 3 NODE 17 FORCES, LBS. MOMENTS, FT-LB 0 ON j

Fg Fy Fg M X b Z STANDARD 19 831 362 - - -

! THERMAL HEAW 23 1353 747 - - -

r EETRA 25 1612 939 REAW i

! STANDARD 16 -354 -7 1

DEAD HEAW 22 -375 -7 WEIGHT EETRA 24 -396 7 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

l EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.20 ( 2 0F 3) l l

p- ,

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODIFIED MODEL NO. 3 NODE 21 ggg;,, rgg"o m CES, uS. M - S W-o I

X Y Z k b Z l

, STANDARD 705 1 34 4 3890 -20 f' THERMAL HEAVY 836 2 -26 4 4143 -26 EXTRA 894 2 -58 7 4227 -27 HEAW I

i STANDARD -9 -271 3 1046 -41 -13 l

DEAD HEAW -14 -288 5 1094 -57 -61 l

l WEIGHT EXTRA -18 -306 7 1173 -67 -83 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

l l EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

I FIGURE 7.20 (3 0F 3)

r- .

COMPARISON OF (SSE) SEISMIC ANCHOR LOADS MODIFIED MODEL NO. 3 NODE 1 BVPS-2 ARS i i

i FORCES, LBS. MONEKIS, FT-LB D

F X Y Z b b Z l STANDARD - - - 5885 6498 3827 1% HEAW - - -

6197 5113 4481

- r EXTRA - - - 6297 4836 4710 BEAW 4

STANDARD - - - 3584 3940 2420 3% REAW - - -

3816 3160 2822

! IKTRA - - - 3898 2989 2966 REAW NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURC 7. 21 (10F 3) i

r COMPARISON OF (SSE) SEISMIC ANCHOR LOADS MODIFIED MODEL NO. 3 NODE 17 BVPS-2 ARS EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYFE F

X Y Z b b Z STANDARD- 72 123 99 - - -

1% HEAW 86 144 112 - - -

EXTRA 92 153 118

, BEAW i STANDARD 47 75 71 - - -

3% HEAW 56 90 78 - - -

EXTRA 59 95 80 REAW NOTES: HEAW FITTINGS ARE IWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7. 21 (2 0F 3)

r COMPARISON OF (SSE) SEISMIC ANCHOR LOADS MODIFIED MODEL NO. 3 NODE 21 BVPS-2 ARS E@I NT FIMING FORCES, LBS. MOMElfrS, FT-LB DAMPING TYPE F

X Y Z b b Z STANDARD 1052 379 390 2047 4795 261 1% HEAW 1413 385 505 2035 5608 250 EXTRA 1592 403 589 2111 5952 256 HEAW STANDARD 645 247 238 1330 2943 164 3% HEAW 879 252 314 1328 1489 160 i

EXTRA 988 270 366 1404 3696 167 HEAW I

, NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

I i

FIGURE 7. 21 (3 0F 3) l t

e

b COMPARISON OF ANCHOR DESIGN LOADS MODIFIED MODEL NO. 3 NODE 1 (DL + THER + SSEI)

T F NG FORCES, LBS. MOMENTS, FT-LB X Y Z b b Z STANDARD - - -

5974 6768 3938 ,

1% HEAW - - - 6274 5504 4664 IITRA - - -

6366 5273 4924 i

HEAW STANDARD - - -

3673 4208 2531 l 3% REAW - - -

3913 3551 3005 l

EXTRA - - -

3967 3426 3180 NEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

i f

FIGURE 7.22 (1 0F 3) i

- - ..-. _z--,._..-,r _ , , _ - _ _ , w . . , . _ . - , , - - , . . , _ _-.......,_....-_..-,..r,_-.-__..


,,._,es-,.,_,_r,

4 3

COMPARISON OF ANCHOR DESIGN LOADS MODIFIED MODEL NO. 3 NODE 17 (DL + THER + SSEI)

EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYPE F

X Y Z b b Z STANDARD 107 600 454 1% HEAW 131 1122 852 EXTRA 141 1369 1064 NEAW STANDARD 82 522 426 3% HEAW 101 1068 818

! EXTRA 108 1311 1026 REAW i

i

. NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

O

[ FIGURE 7.22 (2 0F 3) k

COMPARISON OF ANCHOR DESIGN LOADS MODIFIED MODEL NO. 3 NODE 21 (DL + THER + SSEI)

FORCES, LBS. MOMENTS, FT-LB D G b b X Y Z Z STANDARD 1607 650 427 3097 8644 294

1% HEAW 2235 671 526 3133 9694 337 EXTRA 2468 707 640 3291 10112 366 HEAW STANDARD 1677 527 327 2358 7484 241 3% HEAW 1701 538 335 2426 7575 247 I

I' EXTRA 1864 574 417 2584 7856 277 l NEAW l

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l FIGURE 7.22 (3 0F 3) i

3 COMPARISON OF SELECTED THERMAL AND DEAD WEIGHT STRESSES (PSI), BENCHMARK MODEL NO 7 MEMBER STANDARD HEAVY EXTRA LOADING NODE TYPE FITTINGS FITTINGS HEAVY 2 ELBOW 4995 3189 2873 11 ELB0W 4297 2783 2474 41 ELBOW 6212 l 3675l 3276 60 RUN 173 202 212 81 ELBOW 3155 1786 1578-121 ELBOW 1533 925 828 2 ELBOW 225 157 154 DEAD 11 ELBOW l 307 l 162 145 41 ELBOW 295 147 131 WEIGHT 60 RUN 166 l 188 l 194 81 ELBOW 185 113 102 121 ELBOW 71 45 42 l

NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard i fittings.

l l FIGURE 7. 2 3 l

i i

l L-

c COMPARISON OF SELECTED SSE SEISMIC STRESSES (PSI)

BENCHMARK MODEL NO. 7 BVPS-2 ARS (FIG. 4.3 & 4.4)

MEMBER EQUIPMENT STANDARD HEAVY EXTRA NODE TYPE DAMPIUG FITTINGS FITTINGS HEAVY 2 ELBOW l 46219 l s843 8381 11 ELB0W 13059 2770 2364 41 ELBOW 1% 6959 1545 1324 60 RUN 8140 2564 2358 81 ELBOW 1611 835 747 121 ELBOW 6117 2812 2474 2 ELBOW 28379 l 6126l 5329 11 ELBOW 8009 1716 1496 41 ELBOW 3% 4283 1008 883 60 RUN 4989 1590 1495 81 ELBOW 1079 592 510 121 ELBOW 1727 3616 1521 NOTES: Heavy fittings are twice as thick as standatd fittings.

Extra heavy fittings are three times as thick as standard fittings.

FIGURE 7.24

COMPARISON OF RESTRAINT LOADS (LBS)

BENCHMARK MODEL NO. 7 THERMAL DEAD WEIGHT FITTING

'. -TYPE NODE 20 NODE 110 NODE 20 NODE 110 Fy F Fy Fy Y

STANDARD -141 13 -343 -165 HEAW -222 14 -359 -171

-243 14 -372 -175 HEA SEISMIC (SSE) DESIGN EQUIP. FITTING NODE 20 NODE 110 NODE 20 NODE 110

! F y F F Fy Y Y STANDARD 250 192 734 357 1%

l HEAW 223 158 804 329 243 162 858 323 HA l

l STANDARD 161 125 645 290 3%

HEAW 146 104 727 275 m RA 776 268 161 107

[' HEAW i-NOTES: Heavy fittings are twice as thick as standard fittings.

l l Extra heavy fittings are three times as thick as standard fittings.

I FIGURE 7.25

,-w-=e-+- -:-v,

., ,,,-, ,v..+v, -, --.,w,, . , . - _+,..,p ..:,,,,-----_.,,y, y.

l I

l l

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS ,

MODEL NO. 7 NODE 1 I

O ON Fg F 7

Fg g y Mg STANDARD 308 187 -48 565 -668 1113 THERMAL HEAW 404 271 -79 806 -988 1504 EXTRA 429 293 -87 871 -1075 1608 HEAW 4 STANDARD -11 -69 3 -50 28 6 DEAD HEAW -13 -76 4 -82 36 7 WEIGHT EXTRA -14 -82 4 -97 .39~ 7 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.26 (10F 5)

r-COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODEL NO. 7 NODE 55 L0 E NG F m ING FORCES, LBS. MOMENTS, FT-LB CONDITION TYPE F

X Y Z b b Z STANDARD 169 -47 -317 -768 -388 -360 THERMAL HEAW 215 -58 -364 -872 -349 -449 EXTRA 225 -60 -372 -886 -340 -466 HEAVY STANDARD 0 -92 -2 7 -5 3

i. DEAD HEAVY 0 -97 -3 7 -6 5 WEIGHT EXTRA 0 -100 -3 7 -6 6 HEAVY NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

e FIGURE 7.26 (2 0F 5)

I l

1 i

l COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODEL NO. 7 NODE 65 L0 E NG FI m NG MOMENTS, FT-LB FORCES, LBS.

CONDITION TYPE X Y Z b b Z STANDARD -419 25 648 74 -233 -110 f

, THERMAL HEAW -547 34 772 81 -392 -101 EXTRA -580 36 796 82 -430 -99 HEAW STANDARD 10 -378 -1 46 5 -188 DEAD HEAW 12 -385 -1 51 7 -172 WEIGHT EXTRA 13 -397 -2 54 8 -168 HEAW NOTES: HEAW FITTINGS ARE 'IVICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.26 (3 0F 5) k'

-.m., -, - , . - _ , . , . . . .y .. ... ,.. __,,.-.~ _.....,,,. ___.,.,, , ,,~... .,,,e, ..,...._.~..,,,,,.m.-,,_._

1 l

i l

l l

1 COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODEL NO. 7 NODE 95 ggg, rgg"o FORCES, u S. x0 mms, W-2 X Y Z b b Z l STANDARD 10 251 -611 -598 45 THERMAL HEAW 12 291 -720 -674 54 l EXTRA 12 298 -736 -687 56 HEAW STANDARD 0 -91 0 -8 2 18 DEAD REAW 0 -96 0 -9 2 25  !

WEIGHT l I

EXTRA 0 -99 0 -9 2 26 ,

HEAW l l

l NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS. .

- l

, EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS. ,

FIGURE 7.26 (4 0F 5)  :

l i

- l

( . \

! COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS MODEL NO. 7 NODE 140 T ON F

X Y Z b b Z STANDARD -68 -9 -33 -31 -352 222 THERMAL HEAW -83 -9 -38 -48 -439 268 I

EXTRA -86 -9 -39 -50 -454 276 HEAW 8

STANDARD 1 -104 0 -30 4 46 i-f DEAD HEAW 1 -111 0 -33 3 59 WEIGHT l EXTRA 1 -116 0 -34 4 63 HEAW NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l-FIGURE 7.26 (5 0F 5) li

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

COMPARISON OF (SSE) SEISMIC ANCHOR LOADS BENCHMARK MODEL NO. 7 NODE'1 BVPS-2 ARS (FIG. 4.3 & 4.4)

F NG

, FORCES, LBS. MOMENTS, FT-LB X Y Z b b Z STANDARD 498 1480 1678 11687 2577 1105 l

il 1% HEAW 176 751 828 5901 893 440 EXTRA 170 768 841 6018 845 421 HEAW STANDARD 303 909 1029 7162 1566 674 4

3% HEAW 144 469 513 3645 681 305 EXTRA 144 489 533 3800 673 304 HEAW NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE NO. 7. 27 (1 0F 5)

,, .- .-n.-. , , - , - ,-- ..~,,n.., , , . , . , ,,_,,,,,,.,,,,,,.,--,,n-.

,,,-v --,n.,,,,,,_.,+.,,.,,.,._,.,_,,.,.,.,,n--.-,,, - , _ , , , , , _ , , - . . , - , . ,

COMPARISON OF (SSE) SEISMIC ANCHOR LOADS BENCHMARK HODEL NO. 7 NODE 55 BVPS-2 ARS FIG. 4.3 & 4.4) f FORCES, LBS. MOMENTS, FT-LB X Y Z b b Z l

STANDARD 187 212 670 1934 1682 412 i

l 1% HEAW 86 127 300 878 706 392 EXTRA 85 ' 127 297 869 689 405 HEAW STANDARD 123 146 411 1187 1031 337 l

I 3% HEAW 72 ' 101 192 560 444 339 EXTRA 72 103 195 566 443 350 HEAW l

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l FIGURE NO. 7.27 (2 0F 5) l e ..w.cc= --,..+ve-** -v-e ,--w*.-- - ,,...-----,--mw,ww-,v+-- sr-..e--- -e++w-s----g----+-e-m ae - w - v e-en , - -~- +

l I

1 d

COMPARISON OF (SSE) SEISMIC ANCH0R LOADS BENCIDfARK MODEL NO. 7' NODE 65 BVPS-2 ARS (FIG. 4.3 & 4.4)

EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYPE F

X Y Z b Z STANDARD 303 170 557 1288 4741 811 1% HEAVY 209 140 268 417 1852 639 EXTRA 214 145 269 385 1795 656 HEAW STANDARD 190 125 347 797. 2911 572 3% HEAW 146 105 185 277 1184 459 EXTRA 149 109 188 263 1170 473 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE NO. 7.27 (3 0F 5) 4

- - - - + , , - , , - , , , , - ,,-.n -n,e,,,,-,-- , ,, , , - - - - , , , - - - - , - , , , ,_,.~.-----,----,-,,w,,-,-www---,--,w---,www,-.

-m-,-,,_--m,nw-ve---nanm.,e,wwv,,w,,

I J

COMPARISON OF (SSE) SEISMIC ANCHOR LOADS BENCHMARK MODEL NO. 7 NODE 95 BVPS-2 ARS (FIG. 4.3 & 4.-)

EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYPE F

X Y Z b b Z STANDARD 94 175 101 262 151 614 1% HEAW 92 148 107 276 147 565 EXTRA 94 150 111 286 15 0 580 i HEAW STANDARD 80 132 68 176 96 470 l

3% REAW 78 119 74 189 96 454 l EXTRA 79 121 76 195 99 467

' HEAW l

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

I EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDAPS FITTINGS.

L

! FIGURE NO. 7. 27(4 CF 5) l t

l

COMPARISON OF (SSE) SEISMIC ANCHOR LOADS BENCHMARK MODEL NO. 7 NODE 140 BVPS-2 ARS (FIG. 4.3 & 4.4)

F ING

FORCES, LES. MOMENTS, FT-LB X Y Z k b Z STANDARD 125 252 495 1345 943 1018 1% HEAW 129 195 468 1306 855 923 EXTRA 134 199 476 1334 867 953 HEAVY STANDARD 87 165 292 791 555 649 3% HEAW 95 123 287 800 522 583 EXTRA 99 127 292 818 530 603 HEAVY NOTES
HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

i EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

1 FIGURE NO. 7. 27 (5 0F 5)

9 COMPARISON OF ANCHOR DESIGN LOADS BENCHMARK MODEL NO. 7 NODE 1 (DL + THER + SSEI)

EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYPE F

X Y Z b b Z STANDARD 795 1596 1723 12202 3217 2244 1% HEAW 567 950 903 6625 1845 1951 EXTRA 585 979 924 6792 1881 2036 HEAW STANDARD 600 1025 1074 7677 2206 1813 3% HEAW 535 668 588 4369 1633 1816 EXTRA 559 700 616 4574 1709 1919

HEAW NOTES
HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FIITINGS.

FIGURE N0'. 7. 28 (1 0F 5)

COMPARISON OF ANCHOR DESIGN LOADS BENCHMARK MODEL NO. 7 NODE 55

~

(DL + THER + SSEI) ,

FORCES, LBS. MOMENTS, FT-LB D 3 X Y Z b b Z STANDARD 356 351 988 2695 2075 775 1% HEAW 301 282 667 1743 1061 836 EXTRA 310 287 672 1748 1035 865 HEAW L

t STANDARD 242 285 729 1948 1424 700 l 3% REAW 287 256 559 1425 799 783 EXTRA 297 263 570 1445 789 810 l HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINCO.

l l EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE NO. 7.28 (2 0F 5)

COMPARISON OF ANCHOR DESIGN LOADS BENCHMARK MODEL NO. 7 NODE 65 (DL + THER + SSEI)

EQUIPMENT FITTING FORCES, LBS. MOMENTS, FT-LB DAMPING TYPE N

X Y Z b b Z STANDARD 712 523 1204 1334 4969 1109 1% HEAW 744 441 1039 549 2237 912 EXTRA 781 506 1963 521 2217 923 HEAW STANDARD 599 478 994 843 3139 870 l

1 3% HEAW 681 456 956 409 1569 732 l

EXTRA 716 470 982 399 1592 740 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l FIGURE NO. 7. 28 (3 0F 5) 1

l l

COMPARISON OF ANCHOR DESIGN LOADS BENCHMARK MODEL NO. 7 NODE 95 (DL + THER + SSEI)

EQUIPMENT FITTING FORCES, LBS. MGMENTS, FT-LB DAMPING TYPE F

X Y Z b b Z STANDARD 104 293 352 881 747 677 1% HEAW 104 274 398 1005 819 644 EXTRA 106 280 409 1031 835 662 HEAW STANDARD 90 250 319 795 692 533 4

3% HEAW 90 245 365 918 768 533  !

EXTRA 91 251 374 940 784 549 REAW  ;

I NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

L -i l-l FIGURE NO. 7.28 (4 OF 5)

l COMPARISON OF ANCHOR DESIGN LOADS BENCHMARK MODEL NO. 7 NODE 140 (DL + THER + SSEI)

FORCES, LBS. MOMENTS, FT-LB F

X Y Z b b Z STANDARD 192 365 528 1406 1291 1286 1% HEAW 211 315 506 1387 1291 1250 l

l EXTRA 219 324 515 1418 1317 1292 HEAW l

i STANDARD 154 278 325 852 903 917 l

3% HEAW 177 243 325 881 958 910 ,

EXTRA 184 252 331 902 980 942 HEAW i

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS,

! EXTRA HEAW FITIINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

L l e l

l FIGURE No. 7.28 (5 0F 5) l l

l

24" PIPING STUDY COMPARISON OF SELECTED THERMAL AND DEAD WEIGHT STRESSES (PSI)

MEMBER STANDARD HEAVY EXTRA LOADING NODE TYPE FITTINGS FITTINGS HEAVY 290 ELBOW l22,181l l19,053l 14,261 335 RUN 15,742 18,410 20,241 380 ELBOW 1,844 1,938 1,436 469 TEE 10,291 3,509 1,878 575 RUN 11,395 12,582 13,277

- 600 ELBOW 12,070 8,821 6,110 290 ELBOW 627 258 159 335 RUN 718 778 843 380 ELBOW l 6,652 } l2,077l 1,070 GHT 469 TEE 156 61 35 575 RUN 894 1,235 1,508 600 ELBOW 174 361 317 l ' NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard l fittings.

I' I

i i

i Figure 7.29 l

l

24" PIPING STUDY COMPARISON OF SELECTED SEISMIC STRESSES (PSI) - FLAT ARS I

MEMBER EQUIPMENT STANDARD " HEAVY EXTRA NODE TYP,E DAMPING FITTINGS FITTINGS HEAVY 290 ELBOW 4,167 1,544 912 335 RUN 5,316 5,604 5,616 380 ELBOW 1% 24,291 5,796 2,551 469 TEE (5G) 9,279 3,493 1,992 575 RUN 7,534 7,864 8,151 600 ELBOW 11.879 3,417 2,115 4

290 ELBOW 2,500 _926 547 335 RUN ,

3,189 3,363 3,370 380 ELBOW 3% 14,575 3,478 1,530 469 TEE 5,568 2,094 1,195 (3g) 575 RUN 4,520 4,719 4,890 600 ELBOW 7,128 2,050 1,269 t

NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard i

fittings.

Figure 7.30

24" PIPING STUDY COMPARISON OF SELECTED SEISMIC STRESSES (PSI)

BVPS-2' ARS' (FIG.~4.3'&'4.4)

MEMBER EQUIPMENT STANDARD ' HEAVY EXTRA NODE TYP,E DAMPING FITTINGS FITTINGS HEAVY 290 Elbow 379 130 74 335 Run 450_ 453 456 380 Elbow gg l 2.996 l 566 238 469 Tee 755 283 161 575 Run 697 716 750 600 Elbow 1,298 356 230 290 Elbow 349 124 72 335 Run 429 439 441 380 Elbow 3% 2,388 486 208 469 Tee 742 279 . 159 575 Run 637 1 651 J 678 600 Elbow 1,085 303 190 NOTES: Heavy fittings are twice as thick as standard fittings.

Extra heavy fittings are three times as thick as standard fittings.

Figure 7. 31

, - . _ - _- . _ _ _ _ _ _ _ . __. . .- _. __._ _ _ _ . . _ _ . ~ . _ _ _

24" PIPING STUDY RESTRAINT THERMAI, & DEAD WEIGHT LOADS [lbs.]

SUPPORT TYPE - RIGID LOADING FITTING TYPE NODE NODE NODE N0DE NODE NODE 25 45 65 150 170 230 1

F Fz Fz Fx Fx Fx STANDARD -3,794 6,368 -3,932 3,785 -3,070 -4,325 THERMAL REAVT -4,662 7,827 -5,270 5,051 -3,728 -5,648 EXTRA -5,087 8,539 -5,924 5,669 -4,050 -6,442 BEAVT l

DEAD STANDARD 0 0 0 0 0 1,066 WEIGHT-HEAVY 0 0 0 0 0 1,664 N O O O O O 2,090 EEAVY l NOTEi: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE IEREE TIMES AS THICK AS STANDARD FITIINGS.

l l FIGURE 7. 32 (10F 3) i l

24" PIPING STUDY

, RESTRAINT THE10fAI. & DEAD WEIGHT I.OADS [lbs.}

SUPPORT TYPE - RIGID TYPE NODE NODE NODE NODE NODE. NODE 230 250 270 315 335 335 Fy Fy Fy Fx Fx Fz STANDARD 2,215 -7,480 13,087 9,633 - 11,039 16,424 THERMAL HEAVT 2,784 -9,082 19,310 11,659 13,735 22,501 EITRA 3,164 -10,181 23,627 12,952 - 15,450 26,716 HEAVT STANDARD -3,978 -2,431 -3,813 1,545 -2,015 19 DEAD WEIGHT l

l HEAVY -4,395 -2,492 -3,754 1,663 -2,197 17 EZIRA -4,786 -2,505 -3,770 1,802 -2,386 16 l.

uEAvT NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE THES AS THICK AS STANDARD FITTINGS.

l FIGURE 7. 32 (2 0F 3) l l

6 T-

- .-._.__e,.---_-.-,-,-.,-y

,,,y,---,ew,-wm,v-%-_,,,,,w-.v,~,.,y,y- ve. ,% , y.---..~we-ne-rw-,_ney,.,y- ,,w-

24" PIPING STUDY RESTRAINT THERMAI, & DEAD WEIGHT LOADS (lbs.]

o SUPPORT TYPE - RIGID LOADING FITTING NODE NODE NODE NODE NODE N0DE T'fPE 355 385 505 525 525 640 F, Fy Fy Fy Fz Fy STANDARD -9;493

-7,497 987 567 2,329 -3,255 THERMAL l

i IIAVT -11,086 -12,163 2,469 476 2,698 -5,119 I

I M -12,182 -15,477 3,381 390 2,915 -6,215 l EIAVY DEAD STANDARD -5 -19,092 1,495 -20,579 -76 -6,543 WEIGHT ,

! HEAVY -4 -19,963 946 -21,137 -122 -6,273 l

l

-4 -20,797 651 -21,815 -152 -6,124 i

NOTES: HEA9T FITTINGS ARE TWICE AS THICK AS STAND RD FITTINGS.

I EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7. 32 (3 0F 3) l I

l I

24" PIPING STUDY MC RESTRAINT LOADS [1bs.]

FLAT ARS SUPPORT TYPE - RIGID EQUIPIENT FITTING DAMPING TYyg NODE NODE NODE NODE NODE NODE 25 45 65 150 170 230 F2 Fz Fg Fx Fx Fx l

STANDARD 11,695 16,203 13,833 10,361 9,547 17,381 i

f 1% HEAVT 12,439 17,343 13,766 12,154 10,686 16,902 l (5G) l g 12,109 17,354 12,307 12,406 10,614 18,834

( STANDARD 7,071 9,722 8,300 6,216 5,728 10,429 t

3% EEAVT 7,463 10,406 8,260 7,292 6,412 10,141 (38)

EKTRA 7,266 10,412 7,384 7,444 6,368 11,300 '

EEAVT t

NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD TITTINGS.

i EK!RA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.  !

l FIGURE 7. 33 (10F 3)

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

24" PIPING STUDY -

MC RESTRAINT LOADS [1bs.]

FLAT ARS '

SUPPORT TYPE - RIGID DAMPING TYPE NODE NODE NODE NODE NODE NODE 230 250 270 315 335 335 Fy Fy Fy Fx F Fz x

StaunARn 8,823 6,878 14,490 10,664 17,899 11,906

! 1% HEAYT 9,745 6,109 13,857 11,611 19,506 11,849

(5G)'

N 9,464 6,671 12,354 11,108 18,226 13,329 EEAVT

(

! STANDARD 5,294 4,127 8,694 6,399 10,740 7,143 3% EIATT 5,847 3,665 8,314 6,967 11,704 7,110 (3g)

M 5,678 4,003 7,412 6,665 10,935 7,997 EEAVT NOTES: HEAVT FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVT FITTINGS ARE THREE TDES AS THICK AS STANDARD FITTINGS.

FIGURE 7. 33 (2 0F 3)

4 24" PIPING STUDY NC RESTRAINT LOADS [1bs.]

FLAT ARS p

SUPPORT TYPE - RIGID DA!! PING TTFg NODE NODE NODE NODE NODE NODE 355 385 505 525 525 640 F3 Fy Fy Fy F2 Fy 1

STANDARD 12,568 66,831 16,699 17,401 13,250 11,200 1% EEAVT 13,022 61,365 17,718 20,034 14,795 10,072 (SG)

REA 14,935 60,121 17,933 21,569 16,262 10,978 STAllDARD 7,541 40,099 10,019 10,440 7,950 6,720 3% IIATT 7,813 36,819 10,631 12,021 8,877 6,043 (3g)

EXTRA 8,961 36,073 10,760 12,941 9,757 6,587 EEAVY NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TI2ES AS THICK AS STANDARD FITTINGS.

FIGURE 7.33 (3 0F 3)

24" PIPING STUDY RESTRAINT LOADS {1bs.]

(DW + THER + FLAT ARS)

SUPPORT TYPE - RIGID DAMPING TYPE NODE NODE NODE NODE NODE NODE 25 45 65 150 170 230 Fz Fg Fz Fx Fx Fx STANDARD 15,489 22,571 17,765 14,146 12,617 20,640 1% EEAYT 17,101 25,170 19,036 17,205 14,414 20,886 (SG) 17,196 25,893 18,231 18,075 14,664 23,186 t

STANDARD 10,865 16,090 12,232 10,001 8,798 13,688

3% EEAYT 12,125 18,233 13,530 12,343 10,140 14,125 (3g)

EEIRA 12,355 18,951 13,308 13,113 10,418 15,652 EEAVT ,

1 NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

FIGURE 7.34 (1 0F 3) 4

-- .,e,-mm-mm---,we~nr

24" PIPING STUDY COMPARISON OF DE51GN RESTRAINT LOADS {1bs.]

(DL + THER + FLAT ARS)

SUPPORT TYPE - RIGID DAMPING TYPE NODE NODE NODE NODE NODE NODE 230 250 270 315 335 335 Fy Fy Fy Fx Fx Fz STANDARD 12,801 16,789 23,764 21,842 30,953 28,349 l

l 1% EEAYT 14,140 17,683 29,413 24,933 35,438 34,367 (3G) 14,250 19,357 32,211 25,862 36,062 40,061 m

STANDARD 9,272 14,038 17,968 17,577 23,794 23,586 3% EEAYT 10,242 15,239 23,870 20,'289 27,636 29,628 (3g)

M 10,464 16,689 27,269 21,419 28,771 34,729 l

BEAVT i

l NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EKTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS. .

FIGURE 7. 34 (2 0F 3)

24" PIPING STUDY COMPARISON OF UE51GN RESTRAINT LOADS [lbs.}

(DL + TILER + FLAT ARS)

SUPPORT TYPE - RIGID EQUIPMENT FITTING DAMPING UFE NODE NODE NODE NODE NODE NODE

~

355 385 505 525 525 640 Fg Fy Fy Fy Fz Fy STANDARD 22,066 93,420 19,181 37,980 15,503 20,998 1

1% EEAYT 24,112 93,491 21,133 41,171 17,371 21,464 (5G) g 27,121 96,395 21,965 43,384 19,025 23,317 STANDARD 17,039 66,688 12,501 31,019 10,203 16,518 3% EEAYT 18,903 68,945 14,046- 33,158 11,453 17,435 (3g)

M 21,147 72,347 14,792 34,756 14,150 18,926

HEAYT 1

f i

NOTES:

HEAVT FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRAHEAVYFITTINGSARETHREETIMESASTHICKASSTANDARDFJTTINGS.

l FIGURE 7.34 (3 0F 3)

24" PIPING STUDY COMPARISON OF UE51GN/ SEISMIC RES11 TAINT LOADS [1bs. } FOR SNUBBERS (DL + THER + FLAT ARS)

~

EQUIPMENT FITTING DAMPING TYFg NODE NODE NODE NODE NODE NODE 85 130 315 355 485 6051 Fz Fx Fz Fx Fz Fy STANDARD 11,957 14,285 15,750 17,497 11,426 35,904 1% 3EAVY 13,439 15,086 14,792 18,686 11,486 36,837 (SG) 14,420 14,880 14,626 18,262 11,718 38,375 t

STANDARD 7,174 8,571 9,450 10,498 6,856 21,542 l

3% EEAVT 8,063 9,052 8,875 11,212 6,892 22,102 (3g)

EXTRA 8,652 8,928 8,775 10,957 7,031 23,025 EIAVT l

NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l FIGURE 7. 35

t 24" PIP 1f!G STUDY COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS

^

(TERMINAL ANCHOR, NODE 5)

~

T ON X T Z b b Z STANDARD -1421 0 - 20 0 185 0 THERMAL HEAW -2178 0 -

25 0 228 0 C EXTRA -

2547 0 -27 0 249 0 HEAW STANDARD 0 -1554 0 743 0 -7952 DEAD HEAW 0 -1554 0 651 0 -7952 WEIGHT EXTRA 0 -1554 0 643 0 -7952 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7. 36 (1 of 4)

I w--r- . - - , , , , -------,m---... ,,-.w-., ,--w.,. n-,------, -s,--..-,,- ,.m,------- n

24" P1 PING STUDY COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS (INLINE ANCHOR, NODE 190)

LOADING FITTING MOMENTS, FT-LB FORCES, LBS.

CONDITION TYPE X Y Z k k M z

STANDARD 942 137 -8274 -

734 -3153 -7 THERMAL HEAVY 1107 171 -12,308 -

916 -4044 89 EXTRA 1185 194 -15,255 -1037 -4489 142

. HEAW STANDARD 7 -3023 -11 121 75 -911 DEAD HEAW 11 -3030 -8 194 112 -795 WEIGHT 13~ -3034 -6 248 138 -844 NOTES: HEAW FITTINGS ARE IWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

4 Figure 7.36 (2 of 4) 4

-* , , , , ,-,+,----,,,,-nv..-.n.,,,,,--,,.-,,--,-,-n,,-,, ----.,--,,.,-w.--,,-,,.,,,,,._,+n,-,..,,,_,----,,,,.--,,,.nn,,,

1 1

1 l

l

)

24" PIPII.'G STUDY _,  ;

1 i

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS (TERMINAL ANCHOR, NODE 445)

ORCES, US. MMS, mu ON E X Y Z k b Z STANDARD 1355 10 - 16 1820 -129 -78 THERMAL HEAW 2014 -17 -

19 2069 -158 143 g 2430 -31 - 22 2269 -176 252' STANDARD 455 -1396 0 -4 -1 6301 DEAD HEAW 513 -1397 0 0 -1 6303 l WEIGHT I EXTRA 0 0 6287 561 -1395 1 HEAVY l.

NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

4 Figure 7. 36 (3 of 4) 1 l

,w , e-,--w,,-y--- e,,-ov.ww>-,,,,v-v-, -en,,,,w,,-,,g,w-.,--w,-,~m-m,e-~_.,,en~w-, ,wnn--m ww ww, , -- ---.g-,n ,-

'24" PIPING STUDY ,

COMPARISON OF ANCHOR THERMAL & DEAD WEIGHT LOADS (TERMINAL ANCHOR, NODE 660) ggg;,, rgg=o FORCES, oS. M-S, FT-u F

X F

7 Fg g y Mg STANDARD 80 1895 4165 15,509 -656 2960 t THERMAL HEAW 94 2668 5999 21,826 -768 3515 i EXTRA 100 3123 7069 25,553 -819 3780 HEAW STANDARD -11 -48 -261 7,865 111 -252 DEAD REAW -15 -184 -486 6,444 158 -428

,, WEIGHT l

EXTRA -18 -254 -630 5,701 193 -547 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS. .

EXTRA HEAW FIITINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7. 36 (4 of 4)

24" PIPING STUDY COMPARISON OF SEISMIC ANC110R LOADS (TERMINAL ANCHOR, NODE 5)

FLAT ARS FORCES, LBS. MOMENTS, FT-LB I

I T Z b b Z STANDARD 65,'682 5,940 4,759 2,046 23,368 29,208 1% EEAW 53,564 5,959 4,955 1,881 24,333 29,302 (5G)

EXTRA BEAVY 50,490 5,967 4,656 1,988 22,861- 29,341 STANDARD 39,409 3,564 2,855 1,228 14,021 17,525 3% EEAVY 32,139 3,575 2,973 1,129 14,600 17,582 (3g)

EXTRA 30,294 3,580 2,793 1,193 13,716 17,605 EEAVY l NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD TITTINGS.

Figure 7. 37 (1 of 4) l l

24" "IFING STUDY COMPARISON OF SEISMIC ANCHOR LOADS (INLINE ANCHOR, NODE 190)

FLAT ARS FORCES, LBS. MOMENTS, FT-LB D

X Y Z b b Z STANDARD 9,312 8,594 54,715 31,413 48,850 4,798 1% 10,536 7,524 56,124 32,670 EEAVY 52,705 6,552 (5G)

BEA 11,020 7,608 58,203 35,679 53,245 5,921 STANDARD 5,587 5,156 32,829 18,848 29,310 2,879 3% HEAVT 6,322 4,514 33,675 19,602 31,623 3,931 (3g)

EXTRA 6,612 4,565 34,922 21,407 31,947 3,553 HEAVT

-NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA, HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7. 37(2 of 4) m

24" PIPING STUDY COMPARISON OF SEISMIC ANCHOR LOADS (TERMINAL ANCHOR, NODE 445)

! ' FLAT ARS EQUIPMENT FITTIyc DAMPING FORCES, LBS. MOMENTS, FT-LB TYPE I T Z b b Z STANDARD 9,937 5,434 6,993 3,120 35,962 30,725

! 1% HEAVY 15,089 6,229 6,597 2,761 32,945 34,993 (5G)

EETRA BEAVT 17,826 5,045 7,385 2,785 36,204 28,739 STANDARD 5,962 3,261 4,196 1,872 21,577 18,435 3% HEAVY 9,054 3,737 3,958 1,656 19,767 20,996 (3g)

. EXTRA 10,696 3,027 4,431 1,671 21,722 17,244 HEAVY i

-NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

l Figure 7.37 (3 of 4) i

24" PIPING STUDY COMPARISON OF SEISMIC ANCHOR LOADS (TERMINAL ANCHOR, NODE 660)

FLAT ARS E

FORCES, LBS. M0MENTS, FT-L3 D E Fg F T

Fg y y M Z

STANDARD 7,710 7,209 9,960 43,265 42,396 3,394 1% HEAVY 7,731 7,283 13,351 42,542 41,616 4,241 (SG)

HEA 8,087 8,265 16,429 47,372 44,509 6,143 STANDARD 4,626 4,326 5,976 25,959 25,438 2,036 l

3% HEAVY 4,639 4,370 8,010 25,$25 24,970 2,544 (3g)

EXTRA 4,582 4,959 9,857 28,423 26,705 3,686 HEAVY l

I l

NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7. 37 (4 of 4) ,

.. ., .-.--.--,,~--.-~.--.,n,,,-._------_..,,...--..-,--,,-,-n,.._,,n,-n,,,c.,n-,-,...---...,--_.,,--,-~.,~,e-

r l

24" PIFING STUDY ,

COMPARISON OF ANCHOR DESIGN LOADS (TERMINAL ANCHOR, NODE 5)

(DL + THER + FLAT ARS)

FORCES, LBS. MOMENTS, FT-LB D E Fg F 7

Fg g y Mg STANDARD 67,103 7,494 4,779 2,789 23,553 37,160 1% EEAW 55,742 7,513 4,980 2,532 24,561 37,254 (SG) 53,037 7,521 4,683 2,631 23,110_ 37,293 STANDARD 40,830 5,118 2,875 1,971 14,206 25,477 3% HEAW 34,317 5,129 2,998 1,780 14,828 25,534 (3g)

EXTRA 32,841 5,134 2,820 1,83 6 13,965 25,557 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7. 38 (1 of 4) i

[

24" PIPING STUDY COMPARISON OF ANCHOR DESIGN LOADS (INLINE ANCHOR, NODE 190)

(DL + THER + FLAT ARS)

F NG FORCES, LBS. MOMENTS, FT-LB r

Fg F Y

Fg g y My STANDARD 10,261 11,617 63,000 32,026 51,930 5,716 l 1% HEAVY 11,654 10,554 68,440 33,392 56,637 7,347 (5G) l

' EETRA 12,218 10,642 73,464 36,465 57,596 6,765 HEAW i

STANDARD 6,536 8,179 41,114 19,461 32,390 3,797 l

l l

3% HEAW 7,440 7,544 45,991 20,324 35,555 4,726 i

(3g)

EXTRA 7,810 7,599 50,183 22,193 36,298 4,397 ,

HEAW l

l NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7.38 (2 of 4)

.....wr, .-,e.w.... ..,,or,. .,,,,___._..___..._,,.--__.,_,,_..,m_.,m.,,.,____,,__,..,,___,.m,_..%_,,,_,_m___ _-, _

24" PIPING STUDY COMPARISON OF ANCHOR DESIGN LOADS (TERMINAL ANCHOR, NODE 445)

(DL + THER + FLAT ARS)

EQUIPMENT FITTING DAMPING FORCES, LBS. MOMENTS, FT-LB

TYPE Fy F 7 F Z y y Mg STANDARD 11,747 6,830 7,009 4,936 36,092 37,026 1% BEAVY 17,616 7,643 6,616 4,830 33,104 41,439 (5G) 20,817 6,471 7,407 5,055 35,180 35,278 4

STANDARD 7,772 4,657 4,212 3,688 21,707 24,736 i

3% HEAVY 11,581 5,15 1 3,977 3,725 19,926 27,442 (3g) I i

EXTRA 13,687 4,453 4,453 3,941 21,898 23,783 HEAVY i

l NOTES: HEAVY FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAVY FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

Figure 7.38 (3 of 4) i I

i i

24" PIPING STUDY

~

COMPARISON OF ANCHOR DESIGN LOADS (TERMINAL ANCHOR, NODE 660)

(DL + THER + FLAT ARS)

E I F E FORCES, LBS. MOMENTS, FT-LB Fy F 7

Fg g y Mg STANDARD 7,779 9,056 13,864 66,639 42,941 6,102 L

1% EEAW 7,810 9,767 18,864 70,812 42,226 7,328

-s s (SG) y EXTRA 8,169 11,134 22,868 78,626 45,135 9,376 EEAW STANDARD 4,695 6,173 9,880 49,3'33 25,983 4,744 3% HEAW 4,718 6,854 13,523 53,795 25,580 5,631 (3g)

/

EXTRA 4,664 7,828 16,296 59,677 27,331 6,919 HEAW NOTES: HEAW FITTINGS ARE TWICE AS THICK AS STANDARD FITTINGS.

EXTRA HEAW FITTINGS ARE THREE TIMES AS THICK AS STANDARD FITTINGS.

e Figure 7.38 (4 of 4)

< .