alinik
۱۴ مهر ۱۳۸۸, ۰۰:۱۵
Although there are many studies dedicated to the problem of vortex-induced vibration (VIV) of marine risers, VIV
experiments with internally flowing fluid have not been carried out before. In order to investigate this area, the present
experiment with an internally flowing fluid and external current was designed. The riser was towed in the water flume
with varying internal flow speeds. The tests in still water and in a current were conducted successfully. Various
measurements were obtained including the frequency responses and the time-domain tracing of in-line and cross-flow
responses. The experimental results exhibit several valuable features. First, with an increase in internal flow speed, the
response amplitude increases while the vibration frequency decreases. Secondly, internally flowing fluid lessens the
correlation of the vibration between different sections. In addition, by plotting both in-line strain and cross-flow strain
simultaneously, a figure-of-eight for bending strain is also observed, and the trajectories in different cycles are more
concordant with the increase of internal flow speed.
r 2007 Elsevier Ltd. All rights reserved.destructive effect on marine risers. Vortex-induced forces may excite the riser in its normal mode of transverse
vibration. When the vortex shedding frequency approaches the natural frequency of a marine riser, the cylinder takes
control of the shedding process causing the vortices to be shed at a frequency close to its natural frequencies. This
phenomenon is called vortex shedding ‘‘lock-in’’ or synchronization. Under ‘‘locking in’’ conditions, large resonant
oscillations occur. Large responses give rise to oscillatory stress. If these stress values persist, significant fatigue damage
may occur.
The vortex-induced vibration (VIV) response of a marine riser is a complicated process involving both the hydrodynamic
and the structural properties of the riser. Model testing has given valuable insight into VIV. Different types of
experiments have been done, e.g. forced motion with rigid cylinders in a uniform flow (Gopalkrishnan, 1993), spring
supported rigid cylinders in uniform flow (Vikestad, 1998) and scaled riser models in uniform and sheared flows (Lie
and Vandiver, 1998). Blevins (1990) gives a comprehensive introduction to the phenomenon of VIV in general, while
Vandiver (1998) gives an account of the state-of-art when it comes to the VIV of marine risers.
Although some work has been done for VIVs, a system with the inclusion of internal flow inside the pipe has rarely
been considered. When the internal fluid travels inside the curved path along the deflected riser, it experiencescentrifugal and Coriolis accelerations, respectively, due to the curvature of the riser and the relative motion of fluid to
the time-dependent riser motion. Those accelerations experienced by the riser, in turn, affect the dynamic behavior of
the riser and cause additional vibrations (Moe and Chucheepsakul, 1988; Paı¨doussis et al., 2002; Lopes et al., 2002;
Semler et al., 2002). In addition, the riser vibration could be ‘‘locked-in’’ with the flow-induced vortex shedding such
that they resonate together to produce large deflection and stress. This phenomenon could lead to the failure of the riser
system prematurely. Therefore, it remains to be solved how internal flow affects VIV responses. Chen (1992) explored
this topic in his paper. Hong (1994) investigated the effect of internal flow on VIVs by using a simple oscillator model.
Hong and Huh (1999) developed a mathematical model for the analysis of VIV with the inclusion of internal flow and
examined the effect of internal flow on VIVs.
However, up to now, VIV experiments considering an internally flowing fluid and an external marine environment
have not been properly carried out. In order to investigate the VIVs of risers more thoroughly, we designed an
experiment simultaneously involving internal fluid flow and external current. The instrumented riser with a length of
1.2m is made of rubber and has fixed ends. Its natural frequency can change with varying top tension. The pipe was
towed vertically in the water flume with varying internal fluid speed. Impact tests in still water and tests in different
current magnitudes were conducted successfully. Various measurements were obtained from the strain gauges placed on
the pipe, and the effect of internal flow on VIV was investigated.
2. Experiment
The present experiment was conducted in a wind-wave-current flume, which is 65m long, 1.2m wide and 1.75m deep.
The experiment used a pipe made of rubber with a smooth surface. The pipe had an outer diameter of 14mm and a
thickness of 2 mm. Fig. 1 shows in detail the set-up of the experiment. The instrumented pipe was fixed vertically and
allowed oscillations in both the in-line direction (X-direction) and the cross-flow direction (Y-direction). Its effective
length was 1.2 m, with 0.7m below the water surface. For the internal flow, a water pump circulated the water into the
riser model through an input plastic pipe at a given speed, and a plastic output pipe drained the water into the flume.
Strain gauges were mounted on the riser to measure the responses. Three locations were selected to place strain
gauges, identified as locations 1#, 2# and 3#. At each location, four strain gauges were placed as shown in Fig. 2.
The two strain gauges in the X-direction were used to measure the in-line vibrations, while the other two gauges in the
Y-direction were used to measure the cross-flow vibrations. After installing the strain gauges on the outer surface of the
pipe, the pipe was covered with glue for water proofing.
In order to measure the natural frequency of the riser model, impact tests were conducted on the instrumented pipe in
still water with varying internal fluid velocities. From the strain signals, the power spectra of the riser strain were
obtained. The natural frequency of the riser system can be inferred from the power spectrum of the strain.0.04 m/s. This range of speeds corresponds to Reynolds numbers from 2.24103 to 8.4103. In this range of the
Reynolds number, a fully turbulent vortex street is formed in the wake (Blevins, 1990). For each current, the internal
fluid speed varied from 0, 3.7, 5.5 to 7.2 m/s. In order to represent better the relative significance of internal flow, the
internal flow speed is expressed as a fraction of the current velocity, that is v ¼ V/U, where v is the relative internal flow
speed, V is the internal flow speed, and U is the current speed.
The sampling frequency was 200 Hz, which allowed us to measure signals up to 100 Hz. The highest shedding
frequency was expected to be approximately 9 Hz, and thus the sampling frequency was high enough to avoid aliasing.
The length of time for recording was all 2.56 s. From the strain signals, various measures such as the power spectra,
time-domain traces of in-line and cross-flow direction motions were obtained. The vibration frequency can be inferred
from the power spectrum of the strain.
experiments with internally flowing fluid have not been carried out before. In order to investigate this area, the present
experiment with an internally flowing fluid and external current was designed. The riser was towed in the water flume
with varying internal flow speeds. The tests in still water and in a current were conducted successfully. Various
measurements were obtained including the frequency responses and the time-domain tracing of in-line and cross-flow
responses. The experimental results exhibit several valuable features. First, with an increase in internal flow speed, the
response amplitude increases while the vibration frequency decreases. Secondly, internally flowing fluid lessens the
correlation of the vibration between different sections. In addition, by plotting both in-line strain and cross-flow strain
simultaneously, a figure-of-eight for bending strain is also observed, and the trajectories in different cycles are more
concordant with the increase of internal flow speed.
r 2007 Elsevier Ltd. All rights reserved.destructive effect on marine risers. Vortex-induced forces may excite the riser in its normal mode of transverse
vibration. When the vortex shedding frequency approaches the natural frequency of a marine riser, the cylinder takes
control of the shedding process causing the vortices to be shed at a frequency close to its natural frequencies. This
phenomenon is called vortex shedding ‘‘lock-in’’ or synchronization. Under ‘‘locking in’’ conditions, large resonant
oscillations occur. Large responses give rise to oscillatory stress. If these stress values persist, significant fatigue damage
may occur.
The vortex-induced vibration (VIV) response of a marine riser is a complicated process involving both the hydrodynamic
and the structural properties of the riser. Model testing has given valuable insight into VIV. Different types of
experiments have been done, e.g. forced motion with rigid cylinders in a uniform flow (Gopalkrishnan, 1993), spring
supported rigid cylinders in uniform flow (Vikestad, 1998) and scaled riser models in uniform and sheared flows (Lie
and Vandiver, 1998). Blevins (1990) gives a comprehensive introduction to the phenomenon of VIV in general, while
Vandiver (1998) gives an account of the state-of-art when it comes to the VIV of marine risers.
Although some work has been done for VIVs, a system with the inclusion of internal flow inside the pipe has rarely
been considered. When the internal fluid travels inside the curved path along the deflected riser, it experiencescentrifugal and Coriolis accelerations, respectively, due to the curvature of the riser and the relative motion of fluid to
the time-dependent riser motion. Those accelerations experienced by the riser, in turn, affect the dynamic behavior of
the riser and cause additional vibrations (Moe and Chucheepsakul, 1988; Paı¨doussis et al., 2002; Lopes et al., 2002;
Semler et al., 2002). In addition, the riser vibration could be ‘‘locked-in’’ with the flow-induced vortex shedding such
that they resonate together to produce large deflection and stress. This phenomenon could lead to the failure of the riser
system prematurely. Therefore, it remains to be solved how internal flow affects VIV responses. Chen (1992) explored
this topic in his paper. Hong (1994) investigated the effect of internal flow on VIVs by using a simple oscillator model.
Hong and Huh (1999) developed a mathematical model for the analysis of VIV with the inclusion of internal flow and
examined the effect of internal flow on VIVs.
However, up to now, VIV experiments considering an internally flowing fluid and an external marine environment
have not been properly carried out. In order to investigate the VIVs of risers more thoroughly, we designed an
experiment simultaneously involving internal fluid flow and external current. The instrumented riser with a length of
1.2m is made of rubber and has fixed ends. Its natural frequency can change with varying top tension. The pipe was
towed vertically in the water flume with varying internal fluid speed. Impact tests in still water and tests in different
current magnitudes were conducted successfully. Various measurements were obtained from the strain gauges placed on
the pipe, and the effect of internal flow on VIV was investigated.
2. Experiment
The present experiment was conducted in a wind-wave-current flume, which is 65m long, 1.2m wide and 1.75m deep.
The experiment used a pipe made of rubber with a smooth surface. The pipe had an outer diameter of 14mm and a
thickness of 2 mm. Fig. 1 shows in detail the set-up of the experiment. The instrumented pipe was fixed vertically and
allowed oscillations in both the in-line direction (X-direction) and the cross-flow direction (Y-direction). Its effective
length was 1.2 m, with 0.7m below the water surface. For the internal flow, a water pump circulated the water into the
riser model through an input plastic pipe at a given speed, and a plastic output pipe drained the water into the flume.
Strain gauges were mounted on the riser to measure the responses. Three locations were selected to place strain
gauges, identified as locations 1#, 2# and 3#. At each location, four strain gauges were placed as shown in Fig. 2.
The two strain gauges in the X-direction were used to measure the in-line vibrations, while the other two gauges in the
Y-direction were used to measure the cross-flow vibrations. After installing the strain gauges on the outer surface of the
pipe, the pipe was covered with glue for water proofing.
In order to measure the natural frequency of the riser model, impact tests were conducted on the instrumented pipe in
still water with varying internal fluid velocities. From the strain signals, the power spectra of the riser strain were
obtained. The natural frequency of the riser system can be inferred from the power spectrum of the strain.0.04 m/s. This range of speeds corresponds to Reynolds numbers from 2.24103 to 8.4103. In this range of the
Reynolds number, a fully turbulent vortex street is formed in the wake (Blevins, 1990). For each current, the internal
fluid speed varied from 0, 3.7, 5.5 to 7.2 m/s. In order to represent better the relative significance of internal flow, the
internal flow speed is expressed as a fraction of the current velocity, that is v ¼ V/U, where v is the relative internal flow
speed, V is the internal flow speed, and U is the current speed.
The sampling frequency was 200 Hz, which allowed us to measure signals up to 100 Hz. The highest shedding
frequency was expected to be approximately 9 Hz, and thus the sampling frequency was high enough to avoid aliasing.
The length of time for recording was all 2.56 s. From the strain signals, various measures such as the power spectra,
time-domain traces of in-line and cross-flow direction motions were obtained. The vibration frequency can be inferred
from the power spectrum of the strain.