You are on page 1of 7

Proceedings of the ASME 2014 Pressure Vessels & Piping Conference

PVP2014
July 20-24, 2014, Anaheim, California, USA

PVP2014-28600

INVESTIGATION OF CHARACTERISTIC OF FLOW INDUCED VIBRATION CAUSED


BY TURBULENCE RELATING TO ACOUSTICALLY INDUCED VIBRATION
Masato Nishiguchi
Chiyoda Corporation
4-6-2, Minatomirai, Nishi-ku, Yokohama, Kanagawa Pref., 220-8765, Japan

Gaku Minorikawa
Department of Mechanical Engineering, Faculty
of Science and Engineering, Hosei University
3-7-2, Kajinocho, Koganei-shi, Tokyo, 184-8584,
Japan

Hisao Izuchi
Chiyoda Corporation
4-6-2, Minatomirai, Nishi-ku, Yokohama, Kanagawa
Pref., 220-8765, Japan

ABSTRACT
In the flare piping system, it is known that piping
vibrations occur caused by Acoustically Induced Vibration
(AIV) and Flow Induced Vibration (FIV) corresponding to high
flow rate, high pressure drop and relatively thin pipe wall
thickness. For FIV, turbulence generated at combining tee with
high fluid velocity results in low frequency piping vibration.
For AIV, large noise produced through a component with large
pressure drop results in high frequency piping vibration.
Carucci and Mueller shows the several cases with piping
failure due to AIV and most of these cases the piping failure
occurred at the combining tee. In these piping failure cases, the
velocity at the combining tee would be quite high close or
equal to sound speed and this means piping vibrations could
occur due to FIV in addition to AIV.
This paper shows the investigation results of FIV at
combining tee with 90 degrees using experimental data. The
results are compared to the previous study results for 45
degrees combining tee and the difference between 90 and 45
degrees tees are discussed in the view points of pressure
fluctuation and piping vibration. This paper also shows that the
vibration index proposed by authors is quite effective to
evaluate the vibration level caused by FIV for both of 90 and
45 degrees tees. This proposed vibration index is applied to
failure and no failure cases presented in Carucci and Mueller
paper with some assumptions and it is suggested that the
vibration indexes for failure case is relatively higher than those
of no failure cases. And this suggests that not only AIV but also
FIV could affect the piping failure reported in Carucci and
Mueller paper.

1. INTRODUCTION
In the flare piping system, the shell mode vibration caused
by large mass flow rate and thinner pipe wall thickness would
cause severe trouble such as pipe wall crack [1]. Acoustically
Induced Vibration (AIV) and Flow Induced Vibration (FIV) [2]
are known as typical mechanisms of the shell mode piping
vibration. The vibration source of AIV is the large noise
produced through a component with large pressure drop. Since
the major frequency band of large noise produced through a
component with large pressure drop is order of 1000 Hz, the
vibration frequency of AIV is relatively high. On the other
hand, the vibration source of FIV is turbulence generated at
combining tee with high fluid velocity, and it is generally said
that the vibration frequency due to FIV is relatively low.
Carucci and Mueller shows the several cases with piping
failure due to AIV and most of the piping failure occurred at
the combining tee [3]. In these piping failures, the velocity at
the combining tee would be quite high close or equal to sound
speed and this means that the turbulence energy could be so
high after the combining tee that the piping vibrations could
occur due to the FIV effect in addition to the AIV effect. There
are some proposed criteria to evaluate the AIV risk [4, 5],
however, these criteria could not consider the effect of FIV.
From this point of view, author recently had been developed an
evaluation method of the random piping vibration with the
shell mode caused by FIV for 45 degrees combining tee [2].
The 45 degree combining tee is generally used for flare header
piping system with high velocity due to relatively lower

1
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

pressure drop and reaction force, however, 90 degree


combining tee is also used in the flare piping system.
Therefore, the FIV phenomena for 90 degree tee are also
desired to be investigated.
This paper shows the investigation results of FIV at
combining tee with 90 degrees. In this investigation, the effects
of pipe size, wall thickness and branch area ratio to the main
are also investigated. This paper also shows the investigation
result of the FIV effect to AIV phenomena using the proposed
vibration index which is effective to evaluate the magnitude of
vibration stress of FIV for the failure and no failure cases in the
Carucci and Muller paper.

: Vibration Acceleration [m/s2]

g
p

: Combining Angle [deg]

limit
rms

: Vibration Stress [Pa]

: Angular Frequency [rad/s]

: Gas Density [kg/m3]


: Pipe Material Density [kg/m3]
: Allowable Limit of Vibration Stress [Pa]
: RMS Value of Vibration Stress [Pa]
: Damping Ratio [-]

3. EXPERIMENT METHOD
In order to investigate the characteristics of the shell mode
vibration caused by FIV, the pressure fluctuation of the pipe
inner surface and the shell mode vibration strain were measured
using the experimental system shown in Figure 1 to Figure 3.
Figure 1 and 2 show the experimental system for the branch
pipe diameter less and larger than 1 inch, respectively. Figure 3
shows the locations of the strain gauges at strain measure
section. Since the Computational Fluid Dynamics (CFD) which
was executed prior to this investigation indicated the high
kinetic energy occurs around 4 inches downstream from the
combining point, pressure fluctuation sensor, TPA4 was
installed at this point and the strain gauges were set along the
circumferential section close to TPA4. Table 1 shows the
experimental cases. The experimental cases were determined to
investigate the difference between 90 and 45 degrees tees, and
the effect of pipe size, wall thickness and the branch area ratio
to the main pipe. In order to measure the maximum vibration
stress of the shell mode piping vibration, 7 strain gauges were
attached to the half side of the test pipe as shown in figure 4.
The static pressure of the pipe inner surface was measured by
pressure sensors with the strain gauge type. The pressure
fluctuation was measured by pressure fluctuation sensors with
the piezoelectric type. The experimental procedure is shown
below.

: Damping Constant [kg/s]

fg

: Excitation Frequency Range of Fluid Flow [Hz]

fn
f 2 nd

: Natural Frequency [Hz]

:Transfer Function

k
m
p

: Spring Constant [N/m]


: Pressure inside Air Chamber (TP1) [Pa]

: Pressure Discontinuity through Tee in Case of

: Main Pipe Diameter [m]


: Excitation Force of Vibration [N]

: Natural Frequency of Fundamental Shell Mode [Hz]

: Mass [kg]

Critical
Condition [Pa]
: PSD of Pressure Fluctuation [Pa2/Hz]
: Non-Dimensional PSD of Pressure Fluctuation
[Pa2/Hz]
: Power Spectrum Density of Vibration
Displacement [m2/Hz]

Sw

: Overall RMS of Vibration Displacement [m]

: Cross Sectional Area of Main Pipe [m2]


: Branch Pipe Diameter [m]

Sx

X rms

: Vibration Velocity [m/s]

: Cross Sectional Area of Branch Pipe [m2]

D1
D2
F

P
P*

: Vibration Displacement [m]

&x&

2. NOMENCLATURE

A1
A2
c

x
x&

: Power Spectrum Density of Excitation Force

(1) The air inside the chamber is pressurized.


(2) The ball valve is opened and the pressurized air flows to
the downstream.
(3) The pressure of the air is reduced through the RO
(Restriction Orifice). The mass flow rate in the piping
system is almost determined by the critical flow condition
of this RO. Table 2 and Figure 5 show the specification
and configuration of the RO, respectively.
(4) During the procedure of (2) and (3), the pressure
fluctuations and the vibration strains were measured.

[N2/Hz]

t1
t2

: Branch Pipe Wall Thickness [m]

Vrms
W0

: Overall RMS of Vibration Velocity [m/s]

: Main Pipe Wall Thickness [m]


: Power Spectrum Density of Excitation Force
[Pa2/Hz]

: Flow Velocity at Tee Connection [m/s]

Vrms

: Overall RMS of Vibration Velocity [m/s]

2
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

TPA4
D2, t2

Restriction
Orifice

D1, t1

Tested Main Pipe Tested Branch Pipe


To
Atmosphere
)
6 inch

2
3

TP1

Ball Valve

TP2

3311

TPA5

200

TPA4

TPA3

6 inch

Air
Chamber

101.6 101.6

TPA2

TP3

50.8 50.8 500

Fig.1

PG

Combining
Angle

Fig.4

Strain Measured Point for Circumferential Direction

Experimental Set Up in Case the Branch Diameter is


Table 1

larger than 1 Inch


TP=Pressure Sensor, TPA=Pressure Fluctuation Sensor

D2, t2

Restriction
Orifice

D1, t1

Tested Main Pipe Tested Branch Pipe


To
Atmosphere
)
6 inch
TP2 TPA5
TPA4
TPA3

3311

101.6 101.6

200

D1

t1

D2

t2

[barg]

[mm]

[mm]

[mm]

[mm]

[-]

2 - 13

48.6

1.7

114.3

2.1

0.181

6 - 19

89.1

2.1

114.3

2.1

0.608

9 - 28

114.3

2.1

114.3

2.1

1.000

4 - 12

27.2

1.7

114.3

2.1

0.057

7.5 - 9.5

13.8

1.2

114.3

2.1

0.015

2 - 13

48.6

1.7

114.3

2.1

0.181

2 - 13

48.6

1.7

114.3

3.0

0.181

2 - 13

48.6

1.7

114.3

4.5

0.181

6 - 19

89.1

2.1

114.3

2.1

0.608

4 - 12

27.2

1.7

114.3

2.1

0.057

2 - 13

48.6

1.7

165.2

2.8

0.087

4 - 12

27.2

1.7

165.2

2.8

0.027

45

Air
Chamber

Combining
Angle

Experimental Set Up in Case the Branch Diameter is


Smaller than 1 Inch

TP=Pressure Sensor, TPA=Pressure Fluctuation Sensor


90

Flow

Strain
Measured
Section

Area
Ratio

TP3

Branch
Branch
Main
Main Pipe
Pipe
Pipe Wall
Pipe
Wall
Diameter Thickness Diameter Thickness

[-]
4inch
x
1.5inch
4inch
x
3inch
4inch
x
4inch
4inch
x
0.75inch
4inch
x
0.25inch
4inch
x
1.5inch
4inch
x
1.5inch
4inch
x
1.5inch
4inch
x
3inch
4inch
x
0.75inch
6inch
x
1.5inch
6inch
x
0.75inch

Ball Valve

TPA2

Measured
Pressure
(TP1)

PG

1.5 inch 6 inch

Main and
Branch Pipe
Sizes

[deg]

TP1

50.8 50.8 500

Fig.2

Combining
Angle

Experimental Cases

A1/A2

D1, t1

4inch

Table 2

Tested Branch Pipe

TPA5

TPA4

TPA3

TPA2

D2, t2
101.6

Fig.3

101.6

50.8

Tested Main Pipe

Experimental Set Up around Combining Point

TP=Pressure Sensor, TPA=Pressure Fluctuation Sensor

Specification of Restriction Orifice

Symbol

Dimension[mm]

Pipe Inside Diameter

151.0

tro

8.0

0.8

8.66

P.C.D. R1

39.1

P.C.D. R2

87.4

3
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

4x d
P.C.D. R 1

Flow

P* =
8xd
P.C.D. R 2

(0.5

tro

Pipe Inside
Diameter

v + p

) / (D / v )
2

(1)

D1
v

(2)

TP1 = 13barg (choking)


TP1 = 9barg (choking)
TP1 = 3barg (v=250m/sec)
TP1 = 2barg (v=197m/sec)

Fig.5

St = f

P
2

Configuration of Restriction Orifice

TP1 = 11barg (choking)


TP1 = 3.5barg (v=274m/sec)
TP1 = 2.5barg (v=224m/sec)

100
101
[kPa2/Hz]

PSD of Pressure Fluctuations

3. EXPERIMENT RESULTS
3.1 Characteristic of Pressure Fluctuation at
Combining Tee Downstream
Figure 6 shows the measurement results of pressure
fluctuation downstream of 90 degrees combining tee of 4 inch
x 1.5 inch. In this figure PSDs of the pressure fluctuation are
shown with the parameter of the pressure inside air chamber
(TP1). As shown in Figure 6, the pressure fluctuation increases
as the pressure inside air chamber increases. This means that
the pressure fluctuation increases as the turbulence energy at
the combining tee downstream increases corresponding to the
mass flow rate increase. Authors executed the experimental
investigation with the 4 inch x 1.5 inch piping system with 45
degrees tee in the previous study [2]. The results of previous
study show that it is useful to apply the nondimensionalization
of PSD of the pressure fluctuation at the combining tee
downstream and the vibration frequency using Equations (1)
and (2), respectively. Figure 7 shows the relation between nondimensional PSD of pressure fluctuation and vibration
frequency, i.e., Strouhal number. This figure shows that this
nondimensionalization with using Equations (1) and (2) is also
effective for the FIV phenomena of 90 degrees combining tee
similar to those for 45 degrees combining tee. As shown in
Equations (1), the representative pressure fluctuation in the
pipe would be proportional to the sum of kinetic energy, 0.5 x
v2 and pressure discontinuity at the tee, p. Figure 8 shows
the comparison of non-dimensional PSD of pressure fluctuation
between 90 and 45 degrees combining tees. From this figure
the level of the non-dimensional PSD of pressure fluctuation at
45 degrees combining tee downstream is almost equivalent to
that at 90 degrees combining tee. However, the RMS values of
PSD of pressure fluctuation at 90 degrees combining tee
downstream, 7.08 to 7.65 are slightly larger than those at 45
degrees combining tee, 6.54 to 6.55. From this comparison it
can be said that the pressure fluctuation at 90 degrees
combining tee downstream would be a little larger than that at
45 degrees combining tee.

102
103
104
105

100

Fig.6

1000
10000
Frequency [Hz]

100000

PSD of Pressure Fluctuation (TPA4) at 90

Degrees Combining Tee Downstream


TP1 = 13barg (choking)
TP1 = 9barg (choking)
TP1 = 3barg (v=250m/sec)
TP1 = 2barg (v=197m/sec)

TP1 = 11barg (choking)


TP1 = 3.5barg (v=274m/sec)
TP1 = 2.5barg (v=224m/sec)

101
Fluctuation [-]

Non-Dimensional PSD of Pressure

100

102
103
104
105

0.01

Fig.7

0.1
1
Strouhal Number [-]

10

Non-Dimensional PSD of Pressure

Fluctuation (TPA4) at 90 Degrees Combining Tee


Downstream

4
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

TP1 = 13barg (45deg, RMS=6.55)

stress of the 45 degrees tee. This means that the vibration stress
caused by FIV can be evaluated for both of 45 degrees tee and
90 degrees using the same proposed vibration index with the
correction factor of 1.3.

TP1 = 11barg (45deg, RMS=6.62)


TP1 = 9barg (45deg, RMS=6.54)

TP1 = 13barg (90deg, RMS=7.62)

4" x 3" (90deg)


4" x 1.5" (90deg)
4" x 0.75" (90deg)
6" x 1.5" (90deg)
4" x 1.5" (90deg) t2: 2.1mm => 3.0mm
4" x 1.5" (90deg) t2: 2.1mm => 4.5mm
6" x 0.75" (90deg)

TP1 = 11barg (90deg, RMS=7.65)


TP1 = 9barg (90deg, RMS=7.08)

Fig.8

Vibration Stress [MPa rms]

101
Fluctuation [-]

Non-Dimensional PSD of Pressure

100

102
103
104
105

0.01

0.1
1
Strouhal Number [-]

10

30
25
20
15
10
5
0
0.0

0.5

Comparison of Non-Dimensional PSD of Pressure


Fig.9

Fluctuation (TPA4) between 90 Degrees and 45 Degrees Tees

1.0
1.5
Vibration Index [m/s]

2.0

2.5

Relation between Vibration Index and Vibration


Stress in Different Branch Diameter

3.2 Vibration Stress Evaluation with Vibration Index


Authors recently had been developed the evaluation
method of the random piping vibration of the shell mode
caused by FIV [2]. Equation (3) is the proposed vibration index
which expresses the magnitude of the piping vibration stress
caused by FIV (see Annex for detail) derived from the previous
study.

rms

(Combining Angle is 90 degrees)

0.5 g A1v + pA1


F
=
= Vibration Index (3)
p D 2 2t2 f 2 nd
m fn
2

Vibration Stress [MPa rms]

The adequacy of this vibration index was already


confirmed for the piping systems of the 4 inch x 1.5 inch with
45 degrees combining tee. The effect of combining angle, main
pipe wall thickness and branch pipe area ratio to the main pipe
were experimentally investigated in this study. Figure 9 shows
the relation between vibration stress and the vibration index for
the 90 degrees tee piping systems with various conditions on
pipe wall thickness and branch area ratio to the main. As shown
in this figure the vibration stress tends to be almost
proportional to the vibration index for different pipe wall
thicknesses and branch area ratio to the main. Therefore, the
proposed vibration index would be appropriate to investigate
the magnitude of the vibration stress for wide range of pipe
wall thickness and branch area ratio to the main. Figure 10
shows the comparison of the vibration stress between 45 and
90 degrees tees. As shown in the figure the vibration stress of
the 90 degrees tee is approximately 1.3 times the vibration

4" x 3" (90deg)

4" x 3" (45deg)

4" x 1.5" (90deg)

4" x 1.5" (45deg)

4" x 0.75" (90deg)

4" x 0.75" (45deg)

6" x 1.5" (90deg)

4" x 4" (45deg)

4" x 1.5" (90deg) t2: 2.1mm => 3.0mm

4" x 0.25" (45deg)

4" x 1.5" (90deg) t2: 2.1mm => 4.5mm

6" x 0.75" (90deg)

Line for 90 degree tee

30

130%

25
20
15

100%

10

Line for 45 degree tee

5
0
0.0

0.5

1.0
1.5
Vibration Index [m/s]

2.0

2.5

Fig.10 Relation between Vibration Index and Vibration


Stress in Branch Area Ratio to the Main, Main Pipe Wall
Thickness and Combining Angle

5
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

4. EVALUATION OF AIV WITH FIV VIBRATION INDEX


In the flare piping system, it is known that piping
vibrations occur caused by AIV and FIV corresponding to high
flow rate, high pressure drop and relatively thin pipe wall
thickness. Carucci and Mueller shows the several cases with
piping failure due to AIV and most of these failures occurred at
the combining tee [3]. In these piping failures, the velocity at
the combining tee would be quite high close or equal to sound
speed and this means that the piping vibrations could occur due
to FIV in addition to AIV. From this point of view, the
proposed FIV vibration index shown in Equation (3) was
applied to the failure and no failure cases in the Carucci and
Muller paper in order to investigate the FIV effect. Figure 11
shows the calculation results of the proposed FIV vibration
index for the failure and no failure cases in the Carucci and
Muller paper. In the calculation of FIV vibration index branch
pipe diameter should be assumed since the Carucci and Muller
paper does not show the branch diameter except for the points
of A and B2 in their paper so that the Mach number at the
end of the branch pipe would be between 0.5 to 1.0. From
figure 11 the vibration indexes for failure cases are apparently
higher than those of no failure cases. This suggests that not
only AIV but also FIV could affect the piping failure for the
cases presented in the Carucci and Mueller paper. Further
investigations are needed to develop appropriate evaluation
method taking into account the combined effect of AIV and
FIV.

5. CONCLUSION
In order to investigate the effectiveness of proposed FIV
vibration index proposed by authors experiments were
executed for the combining tee with the various conditions of
different
combining angle, wall thickness and branch area
ratio to the main. In addition, the proposed vibration index was
applied to the failure and no failure cases in the Carucci and
Muller paper in order to investigate the effect of FIV. As a
result, the following conclusions are obtained:
(1) The proposed vibration index would be quite effective
to evaluate the magnitude of vibration stress for wide
range of parameters such as combining angle, wall
thickness and branch area ratio to the main.
(2)

REFERENCES
1. Husain Mohammed Al-Muslim, Nadhir Ibrahim Al-Nasri,
Mohammad Y. Al-Hashem, The Danger of Piping Failure
due to Acoustic Induced Fatigue in Infrequent Operations:
Two Case Studies, PVP2011-57133. (2011)
2.

M. Nishiguchi, H. Izuchi, I. Hayashi, G. Minorikawa,


Flow Induced Vibration of Piping Downstream of Tee
Connection 10th International Conference on FlowInduced Vibration (2012)

3.

V. A. Carucci and R. T. Mueller, Acoustically Induced


Piping Vibration in High Capacity Pressure Reducing
Systems, PVP. (1982)

4.

Guidelines for Avoidance of Vibration Induced Fatigue


Failure in Process Pipework 2nd Edition Published by the
Energy Institute (2008)

5.

F.L. Eisinger, Designing Piping Systems Against


Acoustically Induced Structural Fatigue PVP Vol.119
p.379 (1997)

6.

Fatigue Analysis for Random Vibration,


Advantage, Vol.2 Issue-3, 2008 (In Japanese)

7.

Paul H. Wirsching et al., Random Vibrations Theory and


Practice, DOVER Books (2006)

18
16

Failure
No Failure

14
Vibration Index (m/s)

The piping failure due to AIV at combining tee


reported by Carucci and Muller might be affected to
the FIV phenomena. Further investigations are
needed to develop appropriate evaluation method
taking into account the combined effect of AIV and
FIV.

12
10
8
6
4
2
0
0.0

0.2
0.4
0.6
0.8
Main Pipe Diameter (m)

1.0

Fig.11 Application Results of Vibration Index Shown in


Equation (3) to Failure and No Failure Cases in Carucci

Ansys

and Muller Paper

6
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

ANNEX
EVALUATION METHOD OF RANDOM VIBRATION CAUSED BY TURBULENCE AT COMBINING TEE
DOWNSTREAM

In order to evaluate the random vibration, statistical processing is required [6]. In this processing, the overall RMS (Root Mean
Square) of the vibration is equivalent to the standard deviation. Consider a single DOF (Degree Of Freedom) vibration system
expressed by Equation (4). If the excitation force, F(t) is assumed to be random characteristic, the PSD (Power Spectrum Density) of
the displacement can be written by Equation (5) with frequency domain.

m&x& + cx& + kx = F (t )

(4)

S x ( ) = H ( ) S w ( )

(5)

The transfer function, H () can be written by Equation (6).

H ( ) =

(6)

(k m ) + (c )
2 2

The RMS of the displacement can be written by Equation (7) which is obtained by the integration of Equation (5) [7].

X rms =
2

( )d =

H ( )

S w ( )d =

W0
(7)
4kc

The Equation (10) can be derived from Equation (7) with the definition of Equations (8) and (9).
k = (2f n ) m

(8)

c = 2 mk = 2m = 4mf n

(9)

X rms

W0 f n
m fn

W0 f g
m fn

(10)

The numerator of the right-hand side, Equation (10) expresses the excitation force. The frequency range of the excitation force fg
could be considered to be several times of fn from the view point of vibration response characteristics, and with this consideration
Equation (11) can be derived. This equation is the basis to propose the FIV vibration index.

rms V rms = 2f n X rms

W0 f g
m fn

F
m fn

(11)

7
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

Copyright 2014 by ASME

You might also like