damper(barutzki).pdf

9
5/27/2018 damper(barutzki).pdf-slidepdf.com http://slidepdf.com/reader/full/damperbarutzkipdf 1/9 Plim + Plex 2003, New Orleans, USA Improving the Reliability and Life Expectancy of Piping Systems through the use of Viscous Dampers Frank Barutzki GERB Schwingungsisolierungen GmbH & Co. KG Roedernallee 174-176, Berlin 13407, Germany  Abstract Piping systems in power plants and chemical facilities are complex dynamic structures that are subject to various loads and excitations. Vibrations are often the cause of failure and damage - sometimes with catastrophic results. Viscous fluid dampers can significantly reduce both vibration amplitudes and dynamic stresses. The system natural frequencies are lowered and the piping is less sensitive to dynamic excitations. Material fatigue and failure are reduced, and the operating life of the pipe system is increased. The result is not only cost savings for the user, but also additional safety during both normal operation, and during abnormal, potentially catastrophic events. The successful installation of viscous fluid dampers requires a realistic evaluation of critical vibrations, and an optimization of the dampers in both size and mounting location. The paper describes the dynamic charac- teristics of viscous dampers, the design, selection and installation criteria, as well as a procedure for reducing operational vibrations in existing plants. Introduction Operational experience in power and chemical plants often shows that the reliability and life expectancy of piping systems are largely determined by their dynamic characteristics and behavior. Dynamic loads are experienced during normal, continuous operation and during abnormal, potentially devastating situa- tions [1]. During normal operations the following dynamic excitations may occur: Internal Excitations (Vibration caused by inter- nal pressure pulsations during unsteady fluid flow.) Non-stationary fluid flow in pipes and valves may result in measurable pulsations. Fluid flow is controlled intentionally by opening and clo- sing valves. Unsteady fluid flow can also be caused by the piping arrangement itself, e.g. the number and location of elbows, tees and reducers, especially in case of two-phase flow. Pressure pulsations may exceed the maximum permissible pressure rating, or fall below the fluid vapor pressure resulting in cavitation. External Excitations (Vibration of the entire pipe or individual sections through connected equipment such as pumps or turbines.) Unacceptable pipe motions usually occur only when the natural frequency of the piping system matches the operating frequency of a connected piece of equipment. Even small excitation forces may cause large motions due to resonance effects, not only close to the exci- tation source, but also at greater distances. Usually, low damped, flexible piping systems

Upload: alenmarkaryan1978

Post on 17-Oct-2015

31 views

Category:

Documents


0 download

TRANSCRIPT

  • Plim + Plex 2003, New Orleans, USA

    Improving the Reliability and Life Expectancy of Piping Systems through the use of Viscous Dampers

    Frank Barutzki

    GERB Schwingungsisolierungen GmbH & Co. KG Roedernallee 174-176, Berlin 13407, Germany

    Abstract Piping systems in power plants and chemical facilities are complex dynamic structures that are subject to various loads and excitations. Vibrations are often the cause of failure and damage - sometimes with catastrophic results. Viscous fluid dampers can significantly reduce both vibration amplitudes and dynamic stresses. The system natural frequencies are lowered and the piping is less sensitive to dynamic excitations. Material fatigue and failure are reduced, and the operating life of the pipe system is increased. The result is not only cost savings for the user, but also additional safety during both normal operation, and during abnormal, potentially catastrophic events. The successful installation of viscous fluid dampers requires a realistic evaluation of critical vibrations, and an optimization of the dampers in both size and mounting location. The paper describes the dynamic charac-teristics of viscous dampers, the design, selection and installation criteria, as well as a procedure for reducing operational vibrations in existing plants.

    Introduction Operational experience in power and chemical plants often shows that the reliability and life expectancy of piping systems are largely determined by their dynamic characteristics and behavior. Dynamic loads are experienced during normal, continuous operation and

    during abnormal, potentially devastating situa-tions [1].

    During normal operations the following dynamic excitations may occur:

    Internal Excitations (Vibration caused by inter-nal pressure pulsations during unsteady fluid flow.)

    Non-stationary fluid flow in pipes and valves may result in measurable pulsations. Fluid flow is controlled intentionally by opening and clo-sing valves. Unsteady fluid flow can also be caused by the piping arrangement itself, e.g. the number and location of elbows, tees and reducers, especially in case of two-phase flow. Pressure pulsations may exceed the maximum permissible pressure rating, or fall below the fluid vapor pressure resulting in cavitation.

    External Excitations (Vibration of the entire pipe or individual sections through connected equipment such as pumps or turbines.)

    Unacceptable pipe motions usually occur only when the natural frequency of the piping system matches the operating frequency of a connected piece of equipment. Even small excitation forces may cause large motions due to resonance effects, not only close to the exci-tation source, but also at greater distances. Usually, low damped, flexible piping systems

  • Plim + Plex 2003, New Orleans, USA

    are characterized by closely spaced natural frequencies, which may be easily activated by one of the excitation frequencies. Therefore, attempts to decrease operational vibrations by adding or changing low-damped supports and restraints are usually not very successful. Since removing the excitation source is often not possible, or at the least, very costly, the addition of effective damping to the piping system may the better means to decrease operational vibrations.

    Dynamic deflections may also be caused by abnormal, potentially devastating events, such as

    Earthquake Plane crash Explosion (blast) Pipe breakage

    Operational vibrations usually show only small displacements and stress. Yet they can lead, on a long-term basis, to pipe fatigue and vibra-tion crack corrosion. Alternating stresses that may be below the static yielding point of the pipe material can also lead to micro slips, which cause submicroscopic cracks near the top surface. Due to crack propagation and unification, technical cracks may develop with a large stress peak at their tip. And finally, under continuously alternating loads, fatigue fractures may appear.

    As a result, operational vibrations are often the cause of pipe damage. Material fatigue increa-ses with vibration velocity. The amplitude and frequency of the vibration are determinant fac-tors causing pipe damage.

    Evaluation of Pipework Vibrations A main problem in the evaluation of operating vibrations in piping systems is the lack of internationally accepted and consistent criteria.

    Based on the particular standard, displace-ment or velocity amplitudes are assessed depending on the frequency. Peak- or RMS-values are sometimes used as acceptable vibration limits. Some examples of evaluation criteria:

    According to the method of R. Gamble and S. Tagart [2], which is based on the experience and the error analysis of 400 piping systems in American nuclear power plants, the maximum amplitudes are determined to be

    0.50 mm for frequencies up to 10 Hz and 0.25 mm for frequencies between 10 Hz

    and 40 Hz.

    In France, the vibration velocity limit for feed water lines in nuclear power plants with capacities of 1300 MW [3] is an RMS value of 12.0 mm/s. Russia uses the Standard PTM 38.001-94, with the following classifications:

    (I) Damage is not possible, (II) Damage is improbable (III) Improvement is required and

    damage is possible.

    Permissible displacement amplitudes are spe-cified in m for the individual areas depending on the frequency.

    Frequency Hz

    Area 2 4 6 8 10 20 30 40 50 60

    Vibration Velocity in m

    I 250 230 200 180 165 120 95 85 75 70

    II 500 450 400 360 330 230 180 145 135 130

    III 1250 1100 950 800 750 500 420 350 320 300

    Currently the ANSI/ASME OM3-1982 Proce-dure 1 [4] has become widely accepted.

    223

    3el41

    allow KCC1064.3)S8.0(CCV

    =

    C1 Factor for mass distribution C2, K2 Parameter for stress condition C3 Factor for pipe contents and insulation C4 Factor for restraints (0.8 Sel) Fatigue limit

    The bases for the use of the specified formula are the measured or calculated values of the velocity, displacement, and corresponding fre-quency. By using of the physical correlations

  • Plim + Plex 2003, New Orleans, USA

    for a beam between bending moment and elongation, these values are assigned to stress levels. Factors reflect the geometry, installation condition, load distribution, and stress concen-tration of the piping system and/or section.

    The pipe sections with unacceptably high vibrations must be analyzed dynamically with the goal of reducing the vibrations to accep-table values. An attempt should be made to improve the source of vibrations.

    Dynamic Restraints for Piping Systems Several types of dynamic restraints are used in power and chemical plants:

    Mechanical and hydraulic snubbers Elastic-plastic absorbers or stoppers Axial shock absorbers High viscous fluid dampers

    Dynamic restraints should provide the following features:

    High damping capacity for any dynamic excitation (seismic, shock, vibration)

    Negligible forces under thermal expansion No delay under dynamic loads Long service life Easy inspection and maintenance Overload ability without losing functionality

    Although snubbers are widely used, there are a number of shortcomings. For example, they are not suitable for damping operational vibra-tions.

    Viscous Fluid Dampers as Pipework Dampers Pipework dampers consist of a damper pot, containing a highly viscous damping fluid, and a damper piston, which is immersed in the damping medium. The piston can move in all directions, short of contact with the damper pot, figure 1. Therefore, the damper is effective in all six degrees of freedom.

    The damping forces result from the shearing and displacing of the damping fluid. They are approximately proportional to the relative

    velocity, v, between the damper piston and damper case. The proportionality factor is called the damping resistance, r.

    v)f(rvrF ==

    In order to assure the proper function of the damper, one damper component, either the piston or the damper pot, must be fixed. For practical applications, this means that a sufficiently stiff mounting support is required. Then, the absolute velocity of the moving part can be used for the design calculations.

    Figure 1: Basic design of Pipework Damper

    With the ideal viscous damper, the damping resistance, r, is frequency-independent, figure 2. Therefore, the damper force is ideally pro-portional to the velocity. In addition, when harmonically loaded, the phase angle between the damper force and the displacement would be 90.

    Figure 2: Frequency dependency of the damping resistance

  • Plim + Plex 2003, New Orleans, USA

    In reality, viscous dampers have phase angles between 60 and 80, since there is always an elastic component of the damper force, in addition to the viscous component. Therefore, the phase angle may be used as a measure of the quality of a viscous damper [5].

    Figure 3 shows the standardized time history of force and displacement, as well as the resulting hysteresis loop for a phase angle of 70 between force and displacement.

    -1

    -0,5

    0

    0,5

    1

    0 2 4 6 8 10Time [s]

    F/Fmax S/Smax

    -1

    -0,5

    0

    0,5

    1

    -1 -0,5 0 0,5 1

    S/Smax

    F/Fm

    ax

    Figure 3: Time history and force-displacement loop

    The area of the stationary hysteresis loop is a measure of the damping effect, and cor-responds to the dissipated energy per cycle. Ideal damping behavior, with a 90 phase shift between damper force and displacement, would result in a circle.

    The achievable damping depends on the damping medium, the internal design, and the

    damper load. Static loads are not supported due to the velocity proportional behavior of the damper.

    Slow movements, like thermal expansions of the pipe, cause only minor resistance forces. The viscous elastic qualities of the damper can be described with rheological models, which are formed from the combination of ideal springs and dampers, figure 4. The Voigt-Kelvin-Model is well known and often used for the description of vibration problems.

    When describing basic damper behavior, the generalized Maxwell-Model suits well, since it has ideal relaxation qualities. It is able to describe the viscous elastic qualities of the damper for harmonious excitations, as well as for sudden shock-type loads over a large frequency range. However, the larger the frequency range, and the more variables there are to be considered, the more complex the mechanical models have to be.

    Figure 4: Rheological Models

    Different parameters are used to select dampers for specific tasks. These parameters are determined experimentally for each Pipework damper [11, 12]. They may be characterized by the following:

    The vertical and horizontal damping resistance [kNs/m]

    The vertical and horizontal equivalent stiffness [kN/mm]

    The nominal load [kN] The permissible vertical and horizontal

    displacements [mm]

    The damping resistance is primarily used for operational vibrations. It is determined experi-mentally, assuming ideal viscous behavior,

  • Plim + Plex 2003, New Orleans, USA

    from the dynamic amplitudes of force and vibration velocity over a large frequency range. Figure 5 shows the vertical damping resistance of the aptitude-tested damper series VES. The frequency influence is clear: the damping resis-tance decreases with increasing frequency.

    The equivalent damper stiffness is an auxiliary parameter, which may be used for computa-tional programs that cannot work with velocity proportional damping forces acting in single spots. During intermittent excitation, the damper is handled like an elastic spring, which otherwise is not existent.

    Figure 5: Vertical damping resistance (VES type)

    For this purpose, the equivalent stiffness is defined according to the stiffness definition of a snubber. The equivalent stiffness must be measured, similar to the damping resistance of each damper, figure 6. This parameter must not be mistaken for the elastic stiffness component of the damper force. The equivalent stiffness should only be used for emergencies, and not for normal operational vibrations, since the energy dissipating qualities and the phase shift between force and displacement don't come into play.

    Pipe damper design and selection can also be made on the basis of the nominal force, FN [6]. The rated load is the three-dimensional, dynamic force, which is approved as the maximum damping force at operating tempera-ture.

    Figure 6: Vertical equivalent stiffness (VES type)

    Dynamic impacts should always be below this load limit, which is also determined experimen-tally for every damper, and which is mainly determined by the qualities of the damping medium.

    If the dampers are loaded above the rated nominal load, the damping medium may be sheared off, and no longer in contact with the damper piston (F > 1,7 x FN). However, this process is reversible, and after a short time the damper is again fully functional [7]. A replace-ment of the damping medium is not required.

    The permissible displacement is the sum of all straight movements, i.e. the thermal expansion of the pipeline, the operating oscillations and the impulse response.

    In the case of large thermal expansion, the dampers may be preset in all three directions. With increasing temperature and thermal expansion, the damper piston moves toward the center position.

    The damping behavior of some damping fluids depends strongly on temperature. Therefore, for these dampers, the damping effect depends on the proper determination of the working temperature in the damping medium during continuous operation, and the proper selection of the damping fluid. It is understood that the operational temperature is the highest temperature inside the damping medium during continuous operation. This temperature is influenced by the ambient temperature, the

  • Plim + Plex 2003, New Orleans, USA

    medium temperature inside the pipe and the potential heat transfer.

    Installing insulating plates or spacer construc-tions between the damper and pipe can further reduce the heat transfer into the damper, thus reducing the operating temperature.

    Pipework dampers should be mounted at the locations where experience or detailed calcula-tions show that the largest displacements (anti-nodes) will occur. Considering the first natural frequency and mode shape, this is often also the place where the largest thermal expansion occurs. In most cases, it is better to employ several smaller dampers instead of one big damper, and to distribute them over several points of support [8]. As a result, more mode shapes can be effectively dampened. Even in cases where dampers are installed close to nodal points with no linear displacements, Pipework dampers do affect the piping, as they also provide rotational damping resistances. In order to avoid the transfer of moments to the pipe, they may be used symmetrically in tandem arrangement as shown in figure 7. Dampers must always be installed upright. They can be mounted below, above, or beside the pipe. Due to velocity proportional behavior, they do not support static loads. These loads have to be supported by other components, for example, pipe hangers, sliding bearings, or constant hangers.

    Figure 7: Damper arrangements

    The most important qualities of the Pipework dampers can be summarized as follows:

    Effectiveness in all 6 degrees of freedom.

    High damping forces with shock-type excitations. At great load rates that occur in emergency cases, the Pipework dampers develop high resistances forces. As a result, unacceptable deflections, e.g. during earthquakes, aircraft crash or pressure pulse, are suppressed.

    Damping of operational vibrations. Pipework dampers increase the overall damping of piping systems. They are effective in emergencies, as well as during operational vibrations.

    Immediate response without delay, time lag or minimum response shift. The piston is always in contact with the damping medium, so that the damper responds immediately as a dynamic restraint.

    Small resistance forces during slow move-ments. Pipeline movements due to thermal expansions are not hindered.

    Maintenance free Pipework dampers are virtually mainte-nance free, since they are simply designed, have no wearing parts, and the damping media are not susceptible to aging.

    Damper Design and Selection Dampers may be selected with or without numerical verification.

    If no proof or verification is required, the damper selection will be made for a linear system with one degree of freedom, based on the nominal load of the damper. Shock loads must be smaller than the nominal load of the respective damper. If shock loads are unknown, the so-called 1g-criterium is applied. Under the assumption that no acceleration

  • Plim + Plex 2003, New Orleans, USA

    larger than 1g occurs in the significant frequency range up to 40 Hz, the weight of the pipe or pipe section is used as the shock load. If a numerical verification is needed, FEM programs are available to represent the piping system with all of its components. Some programs are able to work with velocity pro-portional damping forces, which act on indivi-dual points of the structure. In those cases, the energy-dissipating characteristics of the damper can be taken into account. Examples of those programs are ANSYS and dPipe.

    Therefore, viscous dampers can be considered during the design phase of a piping system. However, they can also be installed into existing piping.

    Procedure for the Reduction of Opera-tional Vibrations The procedure for the reduction of operational vibrations is depicted in figure 8. The evalua-tion is performed in accordance to ANSI / ASME requirements. The structural analysis is carried out using a computation model that was adapted, as well as possible, to real measurements [9]. Unreliable load parameters are determined as conservatively as possible.

    Figure 8: Procedure for the reduction of operational vibrations [10]

    The goal of the damper selection is to optimize the introduced damping in such a way that the "decisive", mostly low-frequency modes receive the maximal possible modal damping. Because of the energy dissipation, the modes that play the more substantial part in the dynamic response are significantly reduced. Deflections are unable to build up, and resonance effects are softened.

    By installing as many dampers as necessary, damping can be selectively inserted into the structure at optimal positions. Several dange-rous modes are effectively reduced, and resonance effects are eliminated. This practice reduces metal fatigue of the piping, and therefore, increases the service life of all related pipe components.

    In existing plants, this practice must be complemented by site inspections to find a compromise between the optimal, calculated mounting points, and the installation options feasible on site. The use of viscous fluid dampers to reduce operational vibrations was applied with great success on feed water lines at the NPP PAKS, Hungary, figure 9. The reduction of stress and deflections are shown in figures 10 and 11.

    Figure 9: NPP PAKS, Hungary, Feed-water piping

    The subsequent installation of dampers increases the service life of the piping system. In addition, earthquake safety is improved. Figure 10 depicts how the dampers effectively reduce the stress values in the pipe. It is also clear that the success of the measure depends on the optimal selection of the dampers.

  • Plim + Plex 2003, New Orleans, USA

    0

    20

    40

    60

    80

    100

    120

    140

    160

    180

    200

    220

    24016

    9

    301

    110

    228

    169

    228

    110

    361

    361 39 110

    228

    361 39 110

    228

    361 39 301

    228

    Nodal point

    Stre

    ss [N

    /mm

    ]

    without dampers

    VES 20(1)/10(2)/2.5(3)

    VES 20(1)/10(2)/5(3)

    VES20(1)/20(2)/10(3)

    VES30(1)/20(2)/10(3)

    allowable 2 x Sa = 170 N/mm2

    Figure 10: Comparison Stress levels

    Effective Values of Displacement 15,15 m Level

    0

    0,2

    0,4

    0,6

    0,8

    1

    1,2

    1,4

    1,6

    1,8

    2

    2,2

    2,4

    15 2 3 5 6 5 7 7 4 9 3 1 15

    1 27

    1 41

    1 52

    1 74

    1 86

    1 96

    2 04

    2 11

    2 45

    2 65

    2 75

    2 90

    3 03

    3 18

    3 33

    3 35

    3 44

    3 78

    3 83

    4 26

    4 32

    5 12

    5 16

    6 16

    6 26

    6 56

    6 66

    6 96

    7 06

    7 33

    7 46

    7 86

    9 26

    Nodal point

    Dis

    plac

    emen

    t [m

    m] without dampers

    with dampers

    Figure 11: Comparison Displacements

  • Plim + Plex 2003, New Orleans, USA

    Figure 11 shows again the success of the procedure by comparing the deflections with and without dampers. The deflections could be reduced to 10% of their initial values.

    Conclusions For decades, Pipework Dampers have suc-cessfully protected piping and components against impermissible operational vibrations, shocks, and earthquakes. They are able to provide selectively punctual damping, and are inherent to the design pipe support concepts in power and chemical plants.

    Optimal results can be achieved when exact data of the operating temperature and thermal expansions are available, and when proper support points are found. Measurements or piping calculations usually provide sufficient information for proper damper design and optimal damper location.

    Dampers can be considered during the design phase of a plant or subsequently in the event of unforeseen vibration problems.

    Bibliography [1] Schwahn, K.- J.: Nachweis der Reduzie-

    rung von Strukturschwingungen mittels viskoser Dmpfer. VGB Kraftwerks-technik 69, Heft 10, Okt. 1989.

    [2] Gamble, R.M.; Tagart, S.W.: A Method to Assign Failure Rates for Piping

    Reliability Assessment. PVP-Vol. 215, Fatigue, Fracture, and Risk, ASME

    1991. [3] Seligman, D.; Guillou, J.: Flow induced

    vibration in a PWR piping system. Transactions of the 13th SMIRT, Porto

    Alegre, Brazil, August 13-18, 1995. [4] Requirements for Preoperational and

    Initial Start-up Vibration Testing of Nuclear Power Plant Piping Systems.

    ANSI/ASME OM 3-82. [5] Reinsch, K.-H.; Barutzki, F.: Technischer

    Bericht Rohrleitungsdmpfer, GERB Schwingungsisolierungen, Aus-

    gabe 1997.

    [6] Reinsch, K.-H.; Barutzki, F.: Dmpfung von Schwingungen in Rohrleitungs-systemen. Handbuch Rohrleitungs-technik 6. Ausgabe, 1994, S. 142-147, Vulkan-Verlag Essen.

    [7] Kuitzsch, W., Delinic, K.; Zerrmayr, F.: Die Reduzierung von Rohrleitungs-schwingungen im Betrieb und in Strfall. VDI Berichte Nr. 603, S. 263-292, 1986.

    [8] Delinic, K.: Eigenschwingungsverhalten von Strukturen bei Einsatz von Dmp-fern. VDI Berichte Nr. 627, S. 375-401, 1987.

    [9] Katona, T.; Ratkai, S; Zeitner, W.; Richter, G.; Delnic, K; Reinsch, K.-H.:

    Reduktion der Betriebsschwingungen der Speisewasserleitung des KKW

    Paks. 20 MPASeminar, Stuttgart 1994. [10] Zeitner, W.; Katona, T.: Ratkai, S.;

    Reinsch., K.-H.: Reduction of operational Vibrations. ENS TOPSAFE 95, Budapest 1995.

    [11] Guideline KTA 3205.3: Component Support Structures with Non-integral Connections. Part 3: Series-Production Standard Supports. Carl Heymanns Verlag, Kln, 1989.

    [12] TV Hannover / Sachsen-Anhalt e.V.: TV performance Test of VISCO-DAMPERS. Manufactured by GERB, Berlin. Test Certificate No. T08-91-12, Rev. 2, January 1999.