Evaluation of viscous damper on the vibration transmission to the piping system and the support structure exposed to sinusoidal vibration in vertical, horizontal, and lateral directions.

Viscous dampers have been shown to be a potential solution to addressing piping vibration problems. The primary objective of this study is to confirm the ability of viscous dampers in addressing piping vibration problem in a piping system at different supporting structural stiffness. The piping systems were fabricated and arranged horizontally, vertically, and axially with respect to the direction of the input vibration force. Input force was applied to the piping system using an electromagnetic shaker. A sinusoidal vertical vibration at a 5 Hz frequency (21 mm/s RMS) was applied at one end of the piping system. The use of viscous damper was varied from no viscous damper to one, two and three dampers. Results showed that having three viscous dampers installed caused the piping vibration at point P2 showed the maximum reduction in vibration of up to 80% for three axes, respectively. For the case of structural vibration, the measurement results showed that point S1 experienced the highest vibration exposure before the installation of the viscous dampers, in all axes. With the three viscous dampers installed, the maximum vibration reduction for the sinusoidal vibration test case at point S1 were 86% and 91% for the x- and z-axes, respectively. In conclusion, viscous dampers able to reduce both the piping and structural vibration in multiple axes. This study also highlights the importance of positioning dampers in reducing piping vibration and vibration transmitted to the structure, with reduction up to 90% amplitude for sinusoidal input and 80% for random input.


Introduction
Piping system is a crucial element in many industrial facilities and the occurrence of problem in the piping system can potentially cause the whole facility from performing its intended function as designed.Therefore, it is important for the lifespan of the piping system to be maintained as much as possible.The lifespan of a piping system lifespan depends on the behaviour and characteristics of the pipe dynamic.The pipe dynamic gives rise to piping vibration, which is a normal phenomenon due to the normal operation of industrial facilities be it in a nuclear power plant, conventional power plant, and oil and gas refineries.Piping vibration can also be induced due to other phenomena such as earthquake, pipe rupture, pressure pulsation, valve actuation, slug, water hammer and flow induced vibration [1].
An internal piping vibration can occur during normal operation due to large changes in fluid velocity and unsteady fluid flow [2].These phenomena can occur anywhere along the piping system but are typically seen at piping attachments such as tees, elbows, and reducers [3].Whereas external piping vibration phenomenon can occur when the natural frequency of any applied external force is the same or within 20% of to the operating frequency of any equipment connected to the pipe or the natural frequency of the piping system itself [2,4] The task determining the root cause of piping vibrations is typically challenging due to piping system in industrial facilities such as power plant and chemical facilities are complex in configurations and have been connected to multiple routing that can lead to various excitations and loads on the piping system.This includes dynamic loads such as reciprocating equipment and fluid behaviours which plays an important role in affecting piping vibration [1,5] An example of piping vibration occurring due to normal operation in industrial facilities can be seen in the case of the Loviisa power plant, the first nuclear power plant in Finland [6].Excessive vibration in the piping system was observed when the electric power production was increased by 10%.The required increase in power production was achieved by increasing the flow velocity in the feed and steam piping system resulting in excessive vibration in the system [1] To address these piping vibration issues, extensive research has been carried out to analyse the causes of the vibration and mitigation steps that can be carried out to reduce the piping vibration.It has been reported that visco-damper was able to successfully address piping vibration issues in a variety of cases when compared to the use of snubbers, to address the same vibration issues [7,8].This is because the visco-damper are more flexible, or 'softer', when compared to snubber [4].In snubber, the reaction force from the snubber goes from zero to a finite value in a very short time.Meanwhile, the reaction force for the piping damper shows a more gradual increase with time [9].For a visco-damper to be successful in addressing piping vibration issues, there is a need to identify the potential damper locations and the type of damper to be used in the piping system [10].
The main objective of this study is to investigate the effectiveness of a XX visco-damper in reducing piping vibrations.The effectiveness the dampers were investigated experimentally for a specified piping system.The effectiveness of the visco-damper was tested for various vibrating conditions.

Experimental setup
The piping systems were fabricated and arranged horizontally, vertically, and axially with respect to the direction of the input vibration force as shown in the Figure 1.The component of the pipe was connected using weld neck flanges in accordance to ASME B16. 5 Standard [11].The piping system was designed to represent actual configurations of piping system available at the operating plant.Steel I-beam is used as a support structure.Details of the piping components and the support structure are shown in the Table 1.

Viscous Damper
A total of three XX visco-damper were tested in this study -i) RHY-10, ii) RHY-20 and iii) RHY-40 as shown in the Figure 2. RHY-10 and RHY-20 were installed on the main structure whereas RHY-40 was installed on the individual support structure.In general, RHY-40 has the highest stiffness in all directions (horizontal and vertical directions) when compared to the RHY-10 and RHY-20 types.The parameters of the damper are important in positioning the damper to get optimum effectiveness in vibration reduction.Although a viscous damper (VD) is able to react simultaneously in all directions within its allowable degree-of-freedom, the VD shows a higher damper resistance in horizontal direction as compared to the vertical direction.The trend is consistent for all RHYs type.

Equipment and signal processing
Vibration signals was generated using an electrodynamic shaker from ETS Solution which is connected to the vibration controller (LMS SCADAS mobile) and amplifier.The shaker was used to generate vibration in the y-axis (see Figure 1).
Vibration measurements were acquired using three CTC accelerometers with 100 mV/g sensitivity to measure the vibration in the x-, y-and z-axes, simultaneously.Signals from the accelerometers were obtained using ADASH VA4Pro Vibration Analyser.The post processing of the collected experimental data was done using MATLAB.The signals were obtained and conditioned at 512 samples per second.The signals were averaged over three times with Hanning window applied.The acceleration signals were low-pass filtered at 50 Hz [15] 2.4.Test condition 2.4.1.Signal test condition Energy Institute Guidelines for Avoidance of Vibration Induced Fatigue Failure in Process Pipework [2] has demonstrated that the vibration assessment criteria for piping systems are as shown in Figure 2. If the vibration level is in the 'Problem' criteria, there is a high risk of fatigue damage occurring to the piping system.Thus, to investigate the effectiveness of the XX visco-damper, the shaker is set to excite for both sinusoidal and random vibration falling within the 'Problem" criteria.Table 2 shows the signals test conditions.The reason for using two types of signal excitations (sinusoidal and random) for testing is because the sinusoidal vibration has the advantage in controlling the overall excitation of the shaker, and hence able to improve the general understanding of the VD effectiveness, while random vibration represents more realistic vibration loading that would be encountered at site [16].

Location of measurement on support structure
The vibration transmission to the structure was measured for the condition with and without the VD, similarly to the pipe measurement.The point locations of the supporting structure are shown in Figure 4.

Vibration transmission to the piping system
Results for pipe vibration in x-, y-and z-axes are shown in Table 3. Point P2 recorded the highest vibration transmission for all three axes, while point P1 recorded the lowest vibration transmission for all three axes.The most prominent axis was x-axis for points P2 and P3 while point P1 was affected the most at the y-axis.This is because, point P1 is directly attached to the main shaker that excites at y-axis resulting is a response that is different to the response observed at points P2 and P3.The percentage reduction of the pipe vibration is shown in Figure 5.In general, the dampers are able to reduce the pipe vibration for all axes with the average reduction of 60% for one VD.By installing two VD, the effect of the vibration reduction improves to 74%.For the case of three VD installed, the average percentage reduction is 76%.Although in general having three VD have the highest percentage of the reduction of vibration, the improvement in average vibration reduction is not as significant when compared to installing one and two VD.
The time waveform graph for points P1 -P3 for all axes are shown in Figure 6.. From the results, it was observed that the pipe vibration is dominant in the x-axis for points P2 and P3 while low vibration was measured in z-axis.In general, the VD is able to reduce piping vibration, but the percentage of the vibration reduction varies.

Vibration transmission to the support structure
The vibration transmitted to the support structure at various points are measured.The vibration measurements for sinusoidal excitation at 5 Hz for x-, yand z-axes is shown in Table 4.The comparison of the reduction percentage between points is shown in Figure 7.It was found that the installation of the VD resulted in reduction of the vibration on the support structure for all the axes, except at S8.At point S8, when there is no VD installed, the point structure exposed to a very low vibration because it is not connected to the main pipe (see Figure 8).In this study, with only one VD was installed, the structure vibration was reduced up to 74% at the S1 in x-axis.This indicates the importance of having the dampers in reducing the overall vibration, although there is a possibility of additional vibration transmitted at the location of the damper installed, for example in this case at S8.

Point (i) No VD
(ii) with VD Some of the points, for example y-axis at S9 or z-axis at S2 have almost negligible vibration.By only relying on the percentage of vibration reduction might lead to inaccurate information, especially when the vibration value is small.To have a better understanding on the ratio velocity for overall point structure, time waveform graph with has been plotted as shown in the Figure 9. From the graph plotted, it is obvious that the vibration is prominent in the x-axis, for S1, S2, S4, S5 and S6 while the vibration in y-axis and z-axis is almost negligible.
Generally, installation of one XX viscous damper has significant impact on reducing the transmitted vibration by up to 80%, particularly at locations where the vibration exposure is high (i.e. at point S1 in x-axis).Installing two VD, further improved the reduction of transmitted vibration to the structure by an additional 10%.Having three VD have mixed result on the reducing the transmitted vibration.The third VD has been installed at S5 as shown in the Figure 10.At S5, having the additional third XX viscous damper caused decreasing in the reduction percentage, which indicate more vibration transmitted when compared to two dampers installed.At S5, when there was no VD installed, the velocity recorded was 70.2 mm/s RMS, which indicates that point structure has been exposed to severe vibration.Thus, by installing the damper at the structure that is severely vibrates, will reduce the performance of the damper.This condition proved that although having more dampers will generally improve the structure vibration, the location of the damper installation is vital for optimum result.In addition, it is anticipated that the vibration from the pipe is transmitted to the structure through the damper.Previous studies have shown that the support structure for damper must be at least five times higher than the stiffness of the damper at fundamental vibration frequency [2].

Conclusions
From the experimental study, the results showed that for the sinusoidal vibration test case and three viscous dampers installed, the piping vibration at point P2 has the maximum vibration reduction by up to 86%, 89% and 90% in the x-, y-and z-axes, respectively.For the random vibration test case and with three XX viscous dampers installed, the piping vibration was reduced by up to 45%, 72% and 80% in the x-, y-and z-axes, respectively.
For the case of structural vibration, the measurement results showed that point S1 experienced the highest vibration exposure before installation of the viscous dampers, in all axes.The maximum vibration reductions for the sinusoidal vibration test case at point S1 were 86% and 91% for x-and zaxes, respectively, with three XX viscous dampers installed.For the random vibration test case, the maximum vibration reduction was observed to be at point S1 and the reductions were up to 37%, and 70% for the x-and z-axis, respectively.In the case of the y-axis, the vibration is very low (less than 2 mm/s RMS when the viscous dampers were installed).

Figure 1 .
Figure 1.The location of the XX visco-damper installed.

Figure 4 .
Figure 4. Point measurement at the structure.

Figure 5 .
Figure 5.Comparison of the vibration percentage reduction at P1, P2 and P3.

Figure 7 .
Figure 7.Comparison of the reduction percentage reduction at all points structure.

Figure 8 .
Figure 8. XX visco-damper at Point 8 with (i) No XX condition, (ii) with XX condition at Point 8.

Figure 10 .
Figure 10.Location of the third XX visco-damper.

Table 1 .
Components of the piping system and the support structure.

Table 3 .
Pipe measurement in sinusoidal vibration in x, y and z-axis.

Table 4 .
Vibration measurement at 5 Hz sinusoidal vibration in x-, y-and z-axis.