PVP-Vol.

315, Fitness-for-Service and Decisions

For Petroleum and Chemical Equipment
ASME 1995

IMPROVING THE ACCURACY OF PIPING PROGRAMS WHEN

ANALYZING CLOSELY COUPLED EQUIPMENT

Michael A. Porter
Dynamic Analysis
Leawood, Kansas

Dennis Martens
The Pritchard Corporation
Overland Park, Kansas

Anthony C. Korba
The Pritchard Corporation
Overland Park, Kansas

ABSTRACT An additional concern was that in the close coupled

Standard piping analysis programs, used to determine the arrangement, an equipment nozzle that was inherently more rigid
deflections and stresses in piping systems, are often employed than another associated nozzle could result in a flawed analysis
under conditions that are not within the scope of assumptions in and possibly overload one of the nozzles.
the formulation of the programs. Such a case is the analysis of To better define the loading on the equipment nozzles it
closely interconnected heat exchangers, pressure vessels and became apparent that the nozzle stiffness would need to be
other such equipment. The major problem when using a piping included in the piping flexibility analysis. The piping flexibility
analysis program is that the nozzle connections are modeled software used was Caesar. This software accommodates
infinitely rigid rather than as an element with a finite flexibility. modeling the nozzle stiffness when it is entered as spring rates at
The results generated by such misapplication of the programs is the nozzle-to-shell node.
usually (but not always) very conservative. This paper will Since the spring rates of the nozzles were unknown, it was
demonstrate a hybrid method which employs conventional piping decided to develop the nozzle-to-shell stiffness data using Algor,
analysis software, WRC-107, WRC-297 and Finite Element (FE) a Finite Element (FE) software package. The nozzle stiffness
software to attempt to obtain a better estimate of the deflections, developed using the FE program would then be entered into the
forces, moments, and stresses. The results of the hybrid analysis Caesar model.
are then compared to a complete FE analysis and a standard To confirm the results of the Caesar flexibility analysis and
piping analysis of a sulfur recovery system. The indicated nozzle the associated WRC-Bulletin 107 and WRC-Bulletin 297 nozzle
flexibilities and stresses varied considerably depending upon the stress information, the equipment and piping system would
analysis methodology used. subsequently be modeled and analyzed using the FE software.
This would provide a data comparison platform to review and
verify the Caesar results and confirm the adequacy of this hybrid
DEFINING THE PROBLEM
approach.
The close coupling of piping and equipment for a Sulfur
Subsequent to the development of the nozzle spring rates
Recovery Unit (SRU) design raised concern that the normal
using the FE program, it was decided to determine the nozzle
methodology for studying piping flexibility would generate
spring rates using WRC-Bulletin 297 and the ASME Section III,
excessively conservative equipment nozzle loading data. The
Division I, Paragraph NB-3686.5 procedures. These spring rates
normal methodology used to model piping systems does not
were then entered into the Caesar program to compute the
include nozzle stiffness information in the input data format.
associated forces and moments.
While this is adequate for normal piping configurations (where
the piping provides the mechanism to absorb induced
displacements), it was not considered adequate for closely PIPING FLEXIBILITY (CAESAR) MODEL
coupled equipment. Normally equipment nozzles are modeled as The initial piping model consisted of a 3 bed horizontal
either rigid or standard piping junctions which results in nozzle reactor , 2 vertical reheaters and 3 condensers. The reactor was
loads that do not adequately reflect the ability of the nozzle to restrained at the center in the axial direction and both saddles of
rotate and displace. As pointed out in WRC-297 (Mershon et al., the reactor were guided. The reactor was allowed to expand from
1987) , this was expected to produce inaccurate and conservative the center outward toward the saddles. The condensers were
nozzle loadings. located horizontally under the reactor. Two saddles supported
the condensers. The saddle farthest from the reactor was
modeled as an anchor and the saddle closest to the reactor was
supported by a variable spring support. The reheaters were
located vertically above the outlet of the condensers and FINITE ELEMENT (FE) MODEL
supported by variable spring supports. All of the equipment was The FE model constructed for this analysis consisted of
connected by 20" diameter piping. (see Figure 1). Elements approximately 6750 isoparametric plate elements and 1440 beam
within the equipment were initially modeled as rigid joints. elements. The plate elements were used to model the vessel and
nozzle components. The beam elements were used for a variety
Y of purposes: to model the pipe sections connecting the vessels, to
model the tubes in the heat exchangers, to apply the load of the
catalyst to the reactor vessel, and to serve as rigid connectors
X
between the shell and reinforcement pads on the nozzles. The
Z completed model, illustrated in Figure 2 with a section cut away
so that the interior of the reactor may be seen, had approximately
41,500 degrees of freedom. The solution time on a 486-66 based
PC was slightly over 2 hours.

A Top
B Bottom

D Top
E Bottom

G Top
H Bottom

Y
FIGURE 1 - SCHEMATIC OF PIPING MODEL
X
C Bottom
It became apparent after the first run that flexibility of piping Z

as modeled would not accommodate the thermal displacements. F Bottom

The forces and moments at the connections between the pipe and
FIGURE 2 - FE MODEL WITH NOZZLES LABELED
the equipment nozzles were to be extremely high. Although
indicated forces and moments were high at the pipe to equipment
interfaces, the calculated stresses in the piping ranged between
In modeling the SRU system with the FE code, the modeling
3000 psi and 6000 psi, considerably lower than the allowable
of the springs presented somewhat of a problem. Whereas most
displacement stress of 29,000 psi allowed by ANSI B31.3.
piping analysis codes are designed to arrive at an equilibrium
In order to satisfy our concerns about the stresses developed
position through an iterative procedure, most FE codes (including
at the equipment nozzle connections, a hybrid analysis method
the code employed for this project) do not incorporate such a
was developed. The method included modifying the original
procedure. Thus it was necessary to apply balancing forces to the
model by changing the equipment elements from rigid joints to
nodes where springs were modeled to provide the equilibrium
pipe elements. In addition, a unique connecting node was placed
balance. The process employed was to start with an initial force
at the nozzle and vessel interface. At each node, six restraints
based upon the weight of each of the vessels and run the problem.
were added defining the flexibility of the nozzle (Caesar requires
The forces on the springs were then adjusted, based on the results
that all 6 fields be specifically identified as a spring rate or as
of the analysis, and the solution process was repeated. Since the
rigid). The nozzle flexibility was calculated by means of a FE
reactions of the various parts of the system were coupled, it took
model using Algor software. The nozzle flexibility used was
several iterations of this process to arrive at a balanced system.
comprised of axial translations, circumferential and longitudinal
bending. Shear and torsional stiffnesses were assumed rigid.
As a further means of comparing other methods to the FE NOZZLE STIFFNESS COMPUTATION
analysis, the model above was changed to incorporate nozzle Since a model of the entire system was constructed, the
spring rates calculated by Welding Research Council Bulletin stiffness of the various nozzles were computed using the actual
297. model geometry. If the complete system had not been modeled,
These nozzle spring rates were calculated using Ceasar's the stiffness of each of the nozzles could have been computed
built program, which calculates nozzle flexibilities based on the using a relatively simple model constructed with parametric tools
Welding Research Council Bulletin 297 procedure and such as those produced by ATDAS (Advanced Technology
automatically inserts the values into the model. Design & Analytical Services, Inc., Wethersfield, CT). Figure 3
illustrates such a model. The vessel portion of the model must be
at least long enough to satisfy the attenuation length requirement
 consistency does not exist. It is interesting to note that the
of the ASME code ( 2.5 * rt ). The length of the nozzle is, for
Mokhtarian procedure seems to work well only for nozzles A, D
the most part, not important since the stiffness being computed is and G; for the other nozzles the comparison is not good at all.
primarily that of the vessel shell. The various parameters used by the various procedures are
listed in Table 1. All of the nozzles examined appear to fall
Forces and Y within the range of applicability stated in the respective
Moments procedures, although the limits are not clearly defined. WRC-
applied to 297, for example, merely states that the procedure is good only
X
“Spider” Z
for a limited range of geometries and does not define the
geometries or the range.

COMPARISON OF FE AND PIPING PROGRAM

RESULTS
For the purpose of comparison with the piping model, the
forces, moments, deflections and stresses at 10 locations in the
FE and piping model are tabulated. These locations along with
the node numbers associated with the piping model are indicated
on Figure 1. The locations selected for reporting correspond to
the face of the flange at the various nozzle connections, as
illustrated on Figure 4.
FIGURE 3 - “PARAMETRIC” MODEL OF NOZZLE

In the parametric models, the ends of the vessel portion of the

models are fixed to prevent translations, while leaving the
rotations free. This is similar to the approach that Mershon et al.
(1987) detailed in the appendix of WRC-297. A “spider” of
beam elements is constructed to connect all of the elements at the
interior junction of the nozzle and vessel. Forces and moments
may then be applied to this rigid “spider” to compute the
resultant rotations and deflections. In order to avoid adding
artificial stiffness to the vessel shell, it is important that the rigid
“spider” elements not have a moment connection to the shell. In
addition, the radial degree of freedom between the “spider” and
the shell must not be constrained. With the FE code used, these
requirements were met with the use of “End Releases” on the
beam elements.
Since the nozzle and shell sizes are identical at both ends of
the heat exchangers, the parametric models would indicate that FIGURE 4 - TYPICAL PIPE/NOZZLE
the stiffnesses would be equal. The same would be true if the CONFIGURATION WITH NODE NUMBERING
stiffnesses were computed using the formulas in ASME 1992
Section III, Division I - NB-3686.5 Branch Connections in Tables 2, 3 and 4 present the tabulated Displacement, Force
Straight Pipe. When the complete FE model of the exchanger and Moment data computed by both the piping analysis and by
was used to compute the stiffnesses of the nozzles, they differed the FE analysis. Note that the column labeled “Orig. Pipe”
from the parametrically-derived values by approximately 50%. represents the Caesar run with the nozzles considered rigid. The
The computed stiffnesses derived from the FE analysis, the column labeled “Pipe+297” is the Caesar run with the nozzle
WRC-297 procedure and a procedure proposed as an stiffness computed with the WRC-297 procedure added. The
amplification of the WRC-297 procedure (Mokhtarian and column labeled “Pipe+FEA” is the Caesar run with the stiffnesses
Endicott, 1986) are tabulated in Table 1. The degree of computed with the FE program added, while the column labeled
agreement between the stiffness computation procedures used is “FE” is from the FE model alone.
less than encouraging. The displacements computed by the several types of analyses
Using the FE computed stiffnesses as a base, the stiffnesses are, for the most part, fairly close in value. In Table 2, however,
computed with the ASME Section II NB-3686.5 procedure vary we can see a fairly significant difference in the vertical (Y)
by a factor of somewhat less than 2 for the nozzles examined. deflection values on the reactor nozzle nodes. This difference is
These differences may be due to the apparent lack of provision due to the way in which the reactor is treated by the two different
for a reinforcing pad in the NB-3686.5 formulas. The stiffness programs types. The piping analysis program treats the reactor as
computed with the WRC-297 procedure seem to be consistently a piece of pipe supported at the centroid. The FE program takes
low by a factor of approximately 2 in the circumferential into account the fact that the vessel is supported only along the
direction. In the axial and longitudinal directions, such lower portion where it rests on the support.
 piping program tend to be conservative compared to those
reported by the FE analysis; in other cases the opposite is true.
There is a more significant variation in the moments reported by
the two programs. From this limited analysis, it is unclear if
there is a general correlation between the reported moments.
While the overall highest moments are indicated by the piping
program, there are individual location moments where the values
reported by the piping program are considerably less than
reported by the FE analysis.
The stresses in the nozzles were generally indicated to be
quite low by the FE program. Table 5 shows the Stress Intensity
indicated by the FE program at three selected nozzles on the
reactor and the heat exchanger (see Figure 2 for the nozzle
Support Support locations). Also indicated are the nozzle stresses which were
computed using the forces and moments from the piping analysis
as input to WRC-107 and WRC-297 (as would be customary for
this type of vessel). For the most part, the stresses computed
FIGURE 5 – VERTICAL DISPLACEMENT OF using either WRC-107 or WRC-297 tend to not be conservative
REACTOR when compared to the FE results unless the pad thickness is
ignored. Ignoring the pad (as proscribed in the appendix of
WRC-297) results in stresses which are quite conservative.
Additionally, it should be noted that the stresses indicated here
Figure 5 illustrates the deflected shape of the vessel with the
for WRC-297 are the shell stresses. If the WRC-297 nozzle
magnitude of the deflection in the vertical direction indicated.
stresses are examined, the degree of conservatism is even more
Here we can see that there is localized deflection in the saddle
pronounced. The apparent lack of agreement between the
support regions. This deflection is not seen by the piping
procedures seems to fall in line with the concerns recently raised
analysis program. Additionally, in the FE analysis, the
by Dekker (1994) in his comparison of stress results from WRC-
temperature variation (with elevation) in the reactor is modeled.
107 and Appendix G of BS 5500.
Since the lower portion of the reactor is hotter that the upper
portion, some of the “bowing” is due to the temperature gradient.
RESULTS
The analysis indicated that the coupled equipment
arrangement had adequate flexibility and that the equipment
nozzle stresses were acceptable. The concern that the equipment
nozzle )which was inherently more rigid than another associated
equipment nozzle) could result in a flawed pipe stress analysis
was determined to be unlikely.
The computed nozzle stiffness data developed by the
different procedures and listed in Table 1, gives the authors
concern in that the WRC-297 procedure yields nozzle-to-shell
spring rates that range from 1/2 to over 10 times the FE-
developed rates (the ratio is even higher in the axial direction).
The ASME Division III procedures yield rates that range from
60% to 160% of the FE-developed rates. This disparity of spring
rates would appear to be of such a magnitude that a piping
analysis using the WRC or ASME data would yield a
considerable variation in results.
The addition of the equipment nozzle stiffness information to
FIGURE 6 AXIAL DISPLACEMENT OF REACTOR
the piping flexibility model resulted in a general reduction of the
piping loads on the nozzles when compared to the original Caesar
run, as is indicated in Tables 3 and 4. The data scatter is
Figure 6 illustrates the deflected shape with the deflection in
considerable and is similar to the type of scatter found in the
the axial direction indicated. The axial displacement is greater at
computed spring rates.
the bottom of the vessel than at the top. This temperature effect
The deviation of the developed Caesar modeled piping forces
is not accounted for in the piping program.
and moments using the WRC-297 spring rates compared to those
The way that the two programs handle the deflections of the
developed using the FE spring rate data is considerable, but there
various components leads to a more significant difference in the
is an apparent reduction in the major nozzle forces and moments.
indicated forces and moments. As may be seen in Tables 3 and
The comparison of these results to the results obtained from the
4, the indicated forces and moments are not consistent between
FE analysis were in fair agreement, but considerable deviations
the two programs. In some cases, the forces reported by the
can be noted. The authors consider the FE model results to be the External Loadings”, Welding Research Council Bulletin No. 107,
most accurate and can attribute some of the noted deviations to August, 1965, revised March, 1979
the FE model’s ability to include equipment distortion in its
analysis.

AUTHORS’ NOTES
Constructing models which yielded consistent results using
the piping program and the FE program proved to be a difficult
task. Numerous iterations of the programs were required to
achieve the degree of agreement indicated in this paper. The
differences in the coordinate systems used in WRC-107 and
WRC-297 resulted in similar difficulties.
The authors recommend that when a vessel/piping system
containing closely coupled vessels is analyzed, the engineer
should incorporate nozzle flexibility in the analysis. The
development of nozzle spring rates is best achieved with FE type
of modeling. The use of WRC-297 or ASME Section III
methods should be limited to applications where there is
considerable margin between the calculated stresses and the
allowable stresses.
The authors suggest that the engineer consider the use of
parametric FE models to develop the nozzle stiffness data. These
same models can be used to determine the nozzle shell and
nozzle stress intensities using the forces and moments generated
by the piping program. The nozzle stresses developed by the
hybrid approach should be limited to approximately 1/2 the
allowable design stress to compensate for the inaccuracies of the
method. If it is necessary to determine the stresses more
accurately, the authors suggest that the piping and associated
vessels be modeled and analyzed using FE methods.
Finally, the authors would like to recommend that further
investigation of the WRC-197 and WRC-297 stress results, as
well as the WRC-297 and ASME Section II nozzle stiffness
results, be considered. These widely used stress computation
methods should be compared to available FE and laboratory data
to insure the engineering community that adequate results are
being obtained.

REFERENCES
ASME Boiler and Pressure Vessel Code, Sec. III, Nuclear
Power Plant Components, Div. 1, 1992, American Society of
Mechanical Engineers, New York
Dekker, C.L., 1994, “Comparison of local load stress
calculation methods for nozzles on cylinders,” Int. J. Pres. Ves. &
Piping, Vol. 58, pp 203-213
Mershon, J.L., Mokhtarian, K., Ranjan, G.V., and
Rodabaugh, E.C, 1987, “Local Stresses in Cylindrical Shells
Due to External Loadings on Nozzles - Supplement to WRC
Bulletin No. 107 - (Revision I),” Welding Research Council
Bulletin 297, September, 1987
Mokhtarian, K., and Endicott, J.S., 1986, “Sensitivity
Analysis of Flexibility at Cylinder-Cylinder Junctions,” in
Design and Analysis of Plates and Shells, G. E. O. Widera, H.
Chung, D. Hui, eds, ASME Special Publication PVP-Vol. 105,
American Society of Mechanical Engineers, New York, pp 65-69
Wichman, K.R., Hopper, A.G., and Mershon, J.L., 1979,
“Local Stresses in Spherical and Cylindrical Shells Due to
 PVP-Vol. 315, Fitness-for-Service and Decisions
For Petroleum and Chemical Equipment
ASME 1995

Table 1 - Computed Nozzle Stiffness

Λ λ Axial - lb/in Ratios
Nozzle d/D r/t L/(DT)^2 d/(DT)^.5 FE WRC-297 Mokhtarian WRC/FE Mok/FE
A 0.188 37.16 8.27 1.63 9.39E+05 8.88E+06 2.20E+06 9.46 2.34
B,E,H 0.488 13.67 5.13 2.63 4.34E+06 1.00E+10 1.48E+07 2306.00 3.41
C,F 0.488 13.67 9.72 2.63 3.03E+06 1.00E+10 1.48E+07 3298.00 4.87
D 0.139 42.67 11.61 1.29 1.73E+05 3.50E+06 1.60E+06 20.24 9.24
G 0.139 42.67 8.85 1.29 3.48E+05 5.50E+06 1.60E+06 15.82 4.60
Λ λ Circumferential - ft-lb/rad Ratios
Nozzle d/D r/t L/(DT)^2 d/(DT)^.5 FE WRC-297 Mokhtarian ASME WRC/FE Mok/FE ASME/FE WRC/ASME
A 0.188 37.16 8.27 1.63 3.50E+07 1.78E+07 2.02E+07 2.60E+07 0.51 0.58 0.74 0.68
B,E,H 0.488 13.67 5.13 2.63 3.06E+07 1.24E+07 4.09E+07 3.11E+07 0.41 13.35 1.02 0.40
C,F 0.488 13.67 9.72 2.63 2.13E+07 1.24E+07 4.09E+07 3.11E+07 0.58 19.18 1.46 0.40
D 0.139 42.67 11.61 1.29 1.88E+07 1.00E+07 1.70E+07 1.12E+07 0.53 0.91 0.60 0.89
G 0.139 42.67 8.85 1.29 1.95E+07 1.00E+07 1.70E+07 1.12E+07 0.51 0.87 0.57 0.89
Λ λ Longitudial - ft-lb/rad Ratios
Nozzle d/D r/t L/(DT)^2 d/(DT)^.5 FE WRC-297 Mokhtarian ASME WRC/FE Mok/FE ASME/FE WRC/ASME
A 0.188 37.16 8.27 1.63 6.69E+07 1.78E+07 2.02E+07 2.60E+07 0.51 0.58 0.74 0.68
B,E,H 0.488 13.67 5.13 2.63 7.75E+07 1.24E+07 4.09E+07 3.11E+07 0.41 13.35 1.02 0.40
C,F 0.488 13.67 9.72 2.63 5.52E+07 1.24E+07 4.09E+07 3.11E+07 0.58 19.18 1.46 0.40
D 0.139 42.67 11.61 1.29 5.12E+07 1.00E+07 1.70E+07 1.12E+07 0.53 0.91 0.60 0.89
G 0.139 42.67 8.85 1.29 3.18E+07 1.00E+07 1.70E+07 1.12E+07 0.51 0.87 0.57 0.89
Table 2 - Computed Displacements
Deflection - X - IN Deflection - Y - IN Deflection - Z - IN
Node # Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE
957 0.1139 0.1006 0.0888 0.0496 -0.0231 -0.023 -0.0208 -0.1177 0.4334 0.4327 0.4207 0.4519
990 0.1698 0.1351 0.1243 0.113 -0.1319 -0.1316 -0.1295 -0.2027 0.4168 0.4136 0.3234 0.4399
557 0.1127 0.0987 0.0888 0.0526 -0.0309 -0.0306 -0.0239 -0.2241 -0.0205 -0.0204 0.0196 -0.0011
590 0.1713 0.1323 0.1268 0.1419 -0.1607 -0.1604 -0.1537 -0.3236 -0.0301 -0.0306 -0.0269 0.0049
157 0.1245 0.1177 0.1033 0.0495 -0.0954 -0.0955 -0.0938 -0.1847 -0.6192 -0.6188 -0.621 -0.6072
190 0.187 0.1752 0.1574 0.1469 -0.2534 -0.2535 -0.2518 -0.3291 -0.6192 -0.6216 -0.5693 -0.6084
735 0.6594 0.5999 0.6171 0.734 0.3051 0.3064 0.3024 0.3537 0.0003 -0.0725 -0.0237 -0.0232
710 0.7325 0.6861 0.6862 0.8094 0.1885 0.1898 0.1859 0.2323 -0.0067 0.0635 -0.0272 -0.0095
335 0.709 0.6882 0.6815 0.7816 0.3705 0.3728 0.3713 0.386 -0.1629 -0.1665 -0.2272 -0.2077
310 0.7899 0.7753 0.7568 0.8488 0.2395 0.2418 0.2403 0.2508 -0.1304 -0.1344 -0.1768 -0.1674
Table 3 - Computed Forces
Force - X - LBS Force - Y - LBS Force - Z - LBS
Node # Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE
957 2,602 -3,328 -3,920 -3,867 197 1,123 1,161 115 2,373 2,117 1,608 669
990 2,602 -3,325 -3,920 3,869 540 1,861 1,899 115 2,373 -2,117 -1,608 -669
557 4,183 -4,913 -4,643 -5,140 905 1,460 1,512 1,165 276 114 143 -157
590 4,183 -4,913 -4,643 5,140 1,643 2,198 2,250 1,165 276 114 143 157
157 1,597 -2,677 -2,027 -6,718 3,328 1,903 2,067 1,889 4,961 4,542 4,263 -3,771
190 1,597 -2,677 -2,027 6,718 4,022 2,597 2,761 1,889 4,961 4,542 4,263 3,771
735 253 -596 -310 1,395 254 347 277 356 114 92 178 -422
710 253 -596 -310 -1,395 609 -516 -587 356 114 92 178 422
335 907 199 270 2,097 203 369 207 349 220 361 329 -605
310 907 199 270 -2,097 660 -494 -656 349 220 361 329 605
Table 4 - Computed Momemts
Moment - X - FT- LBS Moment - Y - FT- LBS Moment - Z - FT- LBS
Node # Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE Orig Pipe Pipe+297 Pipe+FE FE
957 20,428 18,266 13,012 -7,573 23,729 22,473 14,190 -50 22,278 -2,229 -6,609 -2,856
990 10,932 9,794 6,538 5,519 23,729 22,473 14,190 50 11,867 11,073 9,024 -9,025
557 4,859 -2,462 -1,374 -1,469 2,818 3,228 -120 -2,102 17,177 -988 -7,174 -2,684
590 3,754 -2,002 -800 1,953 2,818 3,228 -120 2,102 441 18,664 11,420 -13,100
157 45,768 -28,460 -23,843 27,858 65,663 -65,890 -53,016 56,100 19,130 5,444 10,834 -13,933
190 28,816 -12,940 -11,247 -15,600 65,663 -65,890 -53,016 -56,100 24,586 14,593 13,763 -7,910
735 3,594 -2,371 -1,884 367 2,114 3,883 2,793 3,499 6,049 1,364 -4,863 -7,828
710 2,967 -1,866 -903 1,941 2,114 3,883 2,793 -3,499 4,657 4,637 -3,157 15,458
335 10,338 -529 1,752 4,349 7,234 7,035 8,232 16,683 29,523 -14,262 -15,528 -12,000
310 11,550 1,457 3,558 -1,037 7,234 7,035 8,232 -16,683 34,505 -15,356 17,014 23,475
Table 5 - Nozzle Stress - psi, Computed using forces and moments from indicated analysis
FE Stress Orig. Pipe Pipe+297 Stiffness Pipe + FE Stiffness FE Only
Nozzle Intensity WRC-107 WRC-297 WRC-297* WRC-107 WRC-297 WRC-297* WRC-107 WRC-297 WRC-297* WRC-107 WRC-297 WRC-297*
A 13,345 5,573 6,302 33,077 3,762 4,488 33,077 3,435 4,089 19,447 4,159 5,231 23,288
B 9,013 7,801 17,074 83,079 3,726 4,773 30,065 3,867 5,231 28,736 6,272 7,104 59,589
C 6,351 4,948 7,910 40,298 2,361 3,487 18,245 2,638 3,883 20,490 3,746 5,947 29,069
WRC-297* - Reinforcement thickness not included per Appendix A-5 Nozzle Reinforcement