Beruflich Dokumente
Kultur Dokumente
PVP2005-71014
PVP2005-71014
ABSTRACT NOMENCLATURE
ASME ANSI-OM3 standard is dedicated to the assessment α dimensional coefficient equal to 13.4 mm/s/MPa
of piping vibrations for nuclear power plants. It provides an µ ratio concentrated mass / total mass of the pipe
allowable zero-to-peak velocity, which is derived from a σ, S stress
stress/velocity relationship, where corrections factors (C1, C2K2, C0 rms-to-peak factor
C3, C4 and C5) and an allowable stress σal are introduced. In the C1 correction factor standing for the effect of concentrated
ANSI-OM3 standard, the C4 correction factor depends on the mass
pipe layout and on its boundary conditions, and is calculated for C2K2 global stress intensification coefficient as defined in
a few cases. In a former work, it was proposed to extend this Boiling and Pressure Vessels design codes or in RCC-M
factor to a larger number of pipe setups. Besides, the correction C3 correction factor for added mass effects of pipe contents
factor C1, which stands for the effect of concentrated mass, is and insulation
established on a given set-up: a clamped-clamped straight pipe C4 correction factor depending on the pipe layout and on its
span on its first vibrating mode. C1 is then supposed to be boundary condition
conservative on any piping layout. Finally, allowable velocities d displacement
derived from the ANSI-OM3 stress/velocity relationship may be D outer diameter of the pipe
very conservative. One way to reduce this conservatism is to f frequency
introduce regulatory design rules. For a larger set of pipe i intensification factor
geometries, a new set of C1 and C4 correction factors are I area moment of inertia
computed using weight and pressure designs. Using these L length
numerical results, allowable velocities can be calculated. Then, M moment
we propose here to check if a screening vibration velocity of p static pressure inside the pipe
12 mm/s rms is fulfilled. For the 181 geometries on 3708, t thickness
which do not meet the criterion, a seismic design checking is V vibration velocity
applied. Finally, by this way, 99.7% of the tested geometries,
which are supposed to be acceptable with respect to static and
seismic designs, display allowable velocities above 12 mm/s INTRODUCTION
rms and the minimum allowable vibration velocity is 11.2 Following French RCC-M design code [1], piping systems
mm/s. This screening vibration velocity of 12 mm/s commonly of nuclear power plants (NPP) are designed with respect to
used for vibration monitoring of piping systems in EDF nuclear weight, pressure, thermal and seismic considerations. RCC-M
power plants is then supported. does not stipulate any rules to prevent piping systems vibrations
in operating conditions. The only mandatory specification
concerns vibration monitoring of piping systems. ASME ANSI-
α C1C 4 0.8σ al
Vallow = (1)
C3 C2 K 2
L
straight σ limit − σ o
L-bend U-bend µ= and µ ≤ µ max
σ1 − σ o
L
Lp
Lp
Lp Dynamic computation with fluid / with mass µ in Non
L
L
⇒ modal frequency f µn
Z-bend 3D
Figure 2: standard geometries
n f µn
C1 =
f on
L Lp Loop on the modes n
Lp Lp
L L Figure 5: computational algorithm for C1 and C4
symmetrical asymmetrical
Figure 3: symmetry
Data
The data used here are identical to that of [4] (Table 1).
where the added mass correction factor C3 is used because the
The correction factors are calculated for the first four out-of-
previous dynamic computation is realized with fluid.
plane and in-plane modes of L-bend, Z-bend and U-bend pipes
The next computations (Fig. 5) are realized using a loop on
and for the eight first modes of 3D pipes.
the modes, which are divided into in-plane and out-of-plane
modes for 2D geometries (L-bend, Z-bend, U-bend). For each Table 1: numerical data
mode of order n, the static computations of the pipe with no
concentrated mass (µ = 0) and with a unit concentrated mass Young’s modulus E = 2.1011 N/m2
(µ = 1) placed in the node Non provide the maximum stresses, Poisson’s ratio ν = 0.3
respectively σo and σ1, for the determination of the coefficient µ Pipe density ρs = 7800 kg/m3
associated to an acceptable stress for weight design σlimit using Fluid density ρf = 1000 kg/m3
Eq. (4). The static stress is a linear function of µ, which reads: Allowable stress for static design σallowable = 60 MPa
Maximum coefficient µ µmax = 10
σ limit − σ o Outer diameter D = 300 mm
µ= (4) Thickness (schedule close to 10) t = 5 mm
σ1 − σ o
Radius of curvature of elbows Rc = 1,5 D
Isometric lengths L = 3 m (straight)
where σlimit is estimated using the following equation: L = 6 m (L-bend)
L = 9 m (Z-bend U-bend 3D)
pD
σ limit = σ allowable − (5)
4t
and the pressure p may be approximated by the schedule of the Sensitivity analysis
pipe: p (in bars) ≈ schedule. A sensitivity analysis of C4 with respect to the ratios
A dynamic computation with a concentrated mass thickness/diameter (t / D), radius of curvature/diameter (Rc / D)
Mconcentrated = µ (Mpipe + Mfluid) gives the modal frequency f µn , and radius of curvature/isometric length (Rc / L) has been done
in [4]. This first analysis has shown that C4 does not depend on
n
which allows calculating C1 defined by [1] as the ratio of these ratios. In addition, the influence of the ratio t / D on C1 is
frequencies with concentrated mass f µn and with no one f on : shown to be negligible in [5], which is mostly due to Eq. (5)
where the stress limit is reduced by the stress-induced pressure.
In a future work, it should be interesting to evaluate the
n f µn influence of the pipe length on the results in order to complete
C1 = (6)
f on this sensitivity analysis.
60
Discussion
It is difficult to display any particular correlation between
40 pipe setups and minimums of C1C4, which do not depend on
boundary conditions, on mode shapes (in-plane or out-of-
plane), on orientations (vertical or horizontal) and on
20 symmetries (symmetrical or asymmetrical). One observation is
that C1C4 is quite low for 2D geometries and higher for 1D and
3D geometries. For one setup, the minimum of C1C4 is also
0
0 10 20 30 40 50 60 generally reached on the lowest modes.
Vrms
−1
(mm.s )
allow
In the analysis [4], where C4 was calculated on a large
rms
Figure 6: histogram of Vallow before seismic design checking number of setups, the minimum of C4 was found to be 0.34 and
using ANSI-OM3 [2], the minimum of C1 may be estimated to
be 0.2 making µ = µmax = 10. The computation of the product
Results C1C4 here on a large number of pipe setups realistic with
For each setup (geometry, boundary conditions, orientation, respect to static design allows to increase by a factor of two the
symmetry), the coefficients C1, C4 and the product C1C4 are minimum value of C1C4: here min(C1C4) = 0.146 and before
calculated varying 80 times the ratio Lp / L in the range 0.1 to using [2, 4], min(C1C4) = 0.34x0.2 = 0.07.
10. For example, the results of the clamped-clamped horizontal A large range of standard pipe setups has been tested here.
symmetrical Z-bend pipe are presented in Fig. 9 for its first Although, more boundary conditions of pipes could be tested,
three out-of-plane and in-plane modes. For this particular case, we assume that most of the configurations encountered in real
the minimum of the product C1C4, which is 0.309 (Table 3), is piping networks are covered.
obtained for Lp / L ≈ 4.2 for the second in-plane mode. This The computed C1C4 correction factors for each type of
minimum is not obtained when C1 and C4 are separately geometries may be directly used for the calculation of allowable
minimum, which shows that the ANSI-OM3 approach is over- velocities using Eq. (7).
conservative. Although the minimum allowable vibration velocities of
The minimum of the product C1C4 for each tested geometry pipes are also increased by two following Table 3 in comparison
(straight, L-bend, Z-bend, U-bend, 3D) are presented in Table 3. with [4], 5% of the 3708 geometries computed here remain
A comparison with C4 factors calculated in [4] has been made to below the screening vibration velocity of 12 mm/s (Fig. 6).
validate the results obtained here. The minimum of C1C4 is Three types of geometries are concerned: 89 L-bend; 33 U-bend
finally 0.30 for straight pipes, 0.146 for L-bend pipes, 0.148 for and 59 Z-bend pipes. We propose here to check if those 5% of
Z-bend pipes, 0.219 for U-bend pipes and 0.295 for 3D pipes. pipe setups comply with seismic design.
Minimum values of C1C4 can be associated to minimum
allowable vibration velocities of pipes using Eq. (1), which is
modified in: SEISMIC DESIGN CHECKING
CONCLUSION
10 A method is proposed to calculate more realistic correction
factors C1 and C4 for ANSI-OM3 stress/velocity relationship. C1
and C4 correction factors are calculated for several modes on a
large number of pipe setups, which are acceptable with respect
pseudo-acceleration (g)
1 to static design.
For each type of geometry (straight, L-bend, U-bend, Z-
bend or 3D), orientation (vertical or horizontal) and symmetry
(symmetrical or asymmetrical), it is possible to derive from the
0,1 calculated C1 and C4, minimum allowable rms vibration
velocities of pipes.
This first analysis, which is based on static design
considerations, increases allowable vibration velocities for
0,01 piping systems because the correction factors C1 and C4 are
0,1 1 10 100 calculated simultaneously on more realistic pipe setups.
frequency (Hz) Nevertheless, 95% of the tested pipe setups display allowable
Figure 7: pseudo-acceleration oscillator spectrum in the horizontal vibration velocities above the screening vibration velocity of 12
directions mm/s rms used for vibration monitoring in EDF NPP.
1 In addition, a seismic design checking is realized on pipe
setups, which display allowable vibration velocities below 12
mm/s rms. This second analysis finally shows that most of these
configurations do not comply with seismic design. Finally, the
pseudo-acceleration (g)
C4
1 1
0.8 0.8
0.6 mode 1 0.6
0.4 mode 2 0.4
0.2 mode 3 0.2
0 0
0.1 1 10 0.1 1 10
L /L L /L
p p
1 1
0.8 0.8
0.6 0.6
C1
C1
0.4 0.4
0.2 0.2
0 0
0.1 1 10 0.1 1 10
Lp/L Lp/L
1.4 1.4
1.2 1.2
1 1
0.8 0.8
C1.C4
C1.C4
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0.1 1 10 0.1 1 10
L /L L /L
p p
L
Lp
L
Figure 9: C1, C4 and C1C4 correction factors for a clamped-clamped horizontal symmetric Z-bend pipe
STRAIGHT Z-BEND
min(C1C4)
Configuration Orientation min(C1C4) Configuration Symmetry Orientation
In-plane Out-of-plane
clamped-free vertical 1.32 clamped-clamped symmetric vertical 0.286 0.526
horizontal 0.79 horizontal 0.309 0.607
pinned-pinned vertical 1.32 asymmetric vertical 0.243 (0.253) 0.222 (0.253)
horizontal 0.59 horizontal 0.301 0.447
clamped-pinned vertical 0.98 clamped-pinned symmetric vertical 0.230 (0.253) 0.681
horizontal 0.36 horizontal 0.148 (0.253) 0.540
clamped-clamped vertical 0.98 asymmetric vertical 0.216 (0.253) 0.412
horizontal 0.30 horizontal 0.314 0.517
clamped-guided vertical 1.00 clamped-guided symmetric vertical 0.348 0.501
horizontal 0.64 horizontal 0.318 0.500
asymmetric vertical 0.229 (0.253) 0.286
horizontal 0.353 0.477
L-BEND U-BEND
min(C1C4) min(C1C4)
Configuration Orientation Configuration Symmetry Orientation
In-plane Out-of-plane In-plane Out-of-plane
clamped- vertical 0.232 (0.253) 0.154 (0.253) clamped-clamped symmetric vertical 0.312 0.478
clamped horizontal 0.305 0.344 horizontal 0.342 0.561
clamped- vertical 0.271 0.203 (0.253) asymmetric vertical 0.304 0.247 (0.253)
pinned horizontal 0.438 0.438 horizontal 0.314 0.529
clamped- vertical 0.278 0.557 clamped-pinned symmetric vertical 0.219 (0.253) 0.660
guided horizontal 0.146 0.387 horizontal 0.235 (0.235) 0.704
clamped-free vertical 0.288 0.672 asymmetric vertical 0.316 0.652
horizontal 0.288 0.517 horizontal 0.425 0.744
clamped-guided symmetric vertical 0.315 0.415
horizontal 0.334 0.588
asymmetric vertical 0.315 0.248 (0.253)
3D horizontal 0.443 0.578