This application is the US National Stage of International Application No. PCT/EP2014/051322 filed Jan. 23, 2014, and claims the benefit thereof. The International Application claims the benefit of European Application No. EP13153956 filed Feb. 5, 2013. All of the applications are incorporated by reference herein in their entirety.
The invention relates to a method for detuning a rotor-blade cascade.
A turbomachine has rotor blades which are arranged in rotor wheels, which may be regarded as firmly clamped at their blade roots and can oscillate during operation of the turbomachine. Depending on the operating state of the turbomachine, oscillation processes may occur in which oscillating states with high and critical stresses in the rotor blade occur. In the event of long-term loading of the blade by critical stress states, material fatigue takes place which can ultimately lead to a lifetime reduction of the blade, necessitating replacement of the rotor blade.
Because of the centrifugal forces acting on the rotor blade during operation of the turbomachine, a prestress is generated in the rotor blade. Owing to this and the high temperature of the rotor blade during operation, the natural frequencies of the rotor blade during operation differ from the natural frequencies of the cold rotor blade at rest. As a quality-assurance measure during manufacture, only the natural frequencies when the turbomachine is at rest can be measured, although for the configuration of the rotor blade it is necessary to know the natural frequencies under the centrifugal force, so that the oscillation processes in which the oscillation states with high and critical stresses in the rotor blade occur can be avoided.
EP 1 589 191 discloses a method for detuning a rotor-blade cascade.
It is an object of the invention to provide a method for detuning a rotor-blade cascade of a turbomachine, the rotor blades having a long lifetime during operation of the turbomachine.
The method according to aspects of the invention for detuning, in particular rotor-dynamically detuning, a rotor-blade cascade, comprising a multiplicity of rotor blades, of a turbomachine, has the steps: a) establishing for each of the rotor blades of the rotor-blade cascade at least one setpoint natural frequency νF,S which the rotor blade has for at least one predetermined oscillation mode during normal operation of the turbomachine under the effect of centrifugal force, such that the oscillation load of the rotor-blade cascade under the centrifugal force lies below a tolerance limit; b) compiling a value table νF(m, rS) with selected discrete mass values m and radial center-of-mass positions rS, which result from variations of the nominal geometry of the rotor blade, and determining the respective natural frequency νF of the predetermined oscillation mode under the centrifugal force for each selected value pair m and rS; c) measuring the mass mI and the radial center-of-mass position rS,I of one of the rotor blades; d) determining an actual natural frequency νF,I of the rotor blade under the centrifugal force by interpolation of the measured mass mI and the measured radial center-of-mass position rS,I in the value table νF(m, rS); e) in the event that νF,I lies outside a tolerance around νF,S, selecting from the value table νF(m, rS) a value pair mS and rS,S such that νF,S at least approximates νF,S, and removing material of the rotor blade in such a way that mI and rS,I correspond to the value pair mS and rS,S; f) repeating steps c) to e) until νF,I lies within the tolerance around νF,S.
By measuring the mass mI and the radial center-of-mass position rS,I and by interpolating these values in the value table νF(m, rS), the natural frequency νF,I under the centrifugal force can advantageously be determined with a high accuracy. With the method according to the invention, it is likewise advantageously possible to adjust this natural frequency νF,I with a high accuracy and approximate it to the established setpoint natural frequency νF,S. The oscillation load of the rotor blade during operation of the turbomachine can therefore be reduced, so that the lifetime of the rotor blade is extended. Furthermore, the method can be carried out straightforwardly because, for accurate determination of the actual natural frequency νF,I, it is surprisingly sufficient to measure mI and rS,I of the rotor blade without its full geometry. Furthermore, mI and rS,I are quantities which are simple to measure; for example, mI can be measured with a balance.
The predetermined oscillation modes are particularly selected in such a way that the natural frequencies νF,S associated with the oscillation modes are equal to or of lower frequency than a multiple harmonic of the rotor rotation frequency, in particular the eighth harmonic, a value table νF(m, rS) respectively being compiled for a multiplicity of or all of the oscillation modes, the actual natural frequency νF,I being determined for each value table and the value pair mS and rS,S being selected in such a way that the determined νF,I are at least approximated to the established νF,S.
The method according to the invention for detuning, in particular rotor-dynamically detuning, a rotor-blade cascade, comprising a multiplicity of rotor blades, of a turbomachine, has the steps: a) establishing for each of the rotor blades of the rotor-blade cascade at least one setpoint natural frequency νF,S which the rotor blade has for at least one predetermined oscillation mode during normal operation of the turbomachine under the effect of centrifugal force, such that the oscillation load of the rotor-blade cascade under the centrifugal force lies below a tolerance limit; b) compiling a value table νF(m, rS) and a value table νS(m, rS) with selected discrete mass values m and radial center-of-mass positions rS, which result from variations of the nominal geometry of the rotor blade, and determining the respective natural frequency νF of the predetermined oscillation mode under the centrifugal force and the respective natural frequency νS with the rotor blade at rest for each selected value pair m and rS; c) measuring the mass m1 and the radial center-of-mass position rS,I of one of the rotor blades; d) determining an actual natural frequency νF,I of the rotor blade under the centrifugal force by interpolation of the measured mass mI and the measured radial center-of-mass position rS,I in the value table νF(m, rS); e) in the event that νF,I lies outside a tolerance around νF,S, selecting from the value table νF(m, rS) a value pair mS, rS,S such that νF,I at least approximates νF,S, and removing material of the rotor blade in such a way that mI and rS,I correspond to the value pair mS, rS,S; f) in the event that material has been removed, measuring an natural frequency νS,I of the rotor blade at rest; g) repeating steps e) to f or c) to f) until νF,I lies within the tolerance around νF,S and νS,I lies within a tolerance around νS,S corresponding to the tolerance.
By the additional measurement of the natural frequency νS j, the actual natural frequency νF,I under the centrifugal force can advantageously be determined with an even higher accuracy. It is also possible to use the measurement of the natural frequency νS,I at rest in order to monitor the removal, without repeating the measurement of m1 and rS,I.
The predetermined oscillation modes are particularly selected in such a way that the natural frequencies νF,S associated with the oscillation modes are equal to or of lower frequency than a multiple harmonic of the rotor rotation frequency, in particular the eighth harmonic, respectively a value table νF(m, rS) and respectively a value table νS(m, rS) being compiled for a multiplicity of or all of the oscillation modes, the actual natural frequency νF,I and the actual natural frequency νS,I being determined for each value table and the value pair mS and rS,S being selected in such a way that the determined νF,I are at least approximated to the established νF,S and the natural frequencies νS,I being measured for the predetermined oscillation modes.
The variations of the nominal geometry may comprise thickening and/or thinning of the rotor blade in each radial section or in radial sections. It is advantageous for the variations of the nominal geometry to comprise a linear variation of the thickness of the rotor blade over the radius. It is advantageously possible to combine the value table using the thickening and thinning of the nominal geometry with an accuracy sufficient for determining the natural frequencies νF and νS.
The setpoint natural frequencies νF,S are particularly established in such a way that rotor blades arranged next to one another in the rotor-blade cascade have unequal setpoint natural frequencies νF,S, and that the setpoint natural frequencies νF,S are different to the rotor rotation frequency during normal operation of the turbomachine up to and including a multiple harmonic of the rotor rotation frequency, in particular the eighth harmonic of the rotor rotation frequency. This prevents an oscillating rotor blade being able to excite a rotor blade next to it in an oscillation, and coupling of the rotation of the rotor-blade cascade with the oscillations of the rotor blades taking place. The oscillation loads of the rotor blades are therefore low and their lifetime is long.
It is advantageous for the measurement of the mass mI and of the center-of-mass position rS,I to be carried out relatively in a different difference measurement with respect to a reference blade which has been three-dimensionally measured, in particular by a coordinate measuring device and/or by an optical method. The accuracy of a measurement depends on the size of the measurement range, a larger measurement range resulting in a lower accuracy. By carrying out the measurement of mI and rS,I relative to a reference blade, a small measurement range with a high accuracy can be used. It is therefore necessary only to take a single rotor blade as thereference blade and to characterize it once by a cost-intensive three-dimensional method, so that mI and rS,I of all the other rotor blades can also be measured with the high accuracy.
It is advantageous for the value pairs mS and rS,S to be selected in such a way that the unbalance of the rotor is reduced and/or that the outlay for the removal is minimal. Knowledge of the value pair mS and rS,S is sufficient for an unbalance of the rotor, so that detuning and balancing of the rotor-blade cascade can be carried out in a common method step by the removal of the material. The removal of the material may also be carried out in such a way that the amount of material to be removed is minimized.
The predetermined oscillation mode is particularly selected in such a way that the natural frequency νF,S of the predetermined oscillation mode is equal to or of lower frequency than a multiple harmonic of the rotor rotation frequency, in particular the eighth harmonic. The natural frequencies νF and/or νI are particluarly determined computationally, in particular by a finite element method.
It is advantageous that, during the measurement of the frequency νS,I, the rotor blade is clamped at its blade root, and the oscillation of the rotor blade is excited and measured. The oscillation is particularly measured by oscillation transducers, acceleration sensors, strain gages, piezoelectric sensors and/or optical methods. This constitutes a simple method for determining the natural frequency.
Adaptation of the model for determining the natural frequencies νF and νS is particularly carried out by a comparison of the measured natural frequency νS,I with an actual natural frequency determined by interpolation of mI and rS,I in the value table νS(m, rS). In this way, influences of the material on the natural frequencies can advantageously be taken into account as well.
The invention will be explained in more detail below with the aid of the appended schematic drawings, in which:
The second rotor blade shows variations 6, 7 of the nominal geometry 5, in which, starting from the nominal geometry 5 the mass m is varied but the radial center-of-mass position rS of the rotor blade is not. In the first variation 6, the mass m is increased by uniformly thickening the second rotor blade at each radial distance r from the rotation axis, and in the second variation 7 the mass m is reduced by radially thinning the second rotor blade at each radial distance r.
In the variations 8, 9 of the third rotor blade, starting from the nominal geometry 5 the thickness of the rotor blade is varied linearly over the radius r in the circumferential direction and/or the axial direction. According to the third variation 8, starting from the nominal geometry 5 the rotor blade is thickened at its blade root 2 and thinned at its blade tip 3, and according to the fourth variation 9, starting from the nominal geometry 5 the rotor blade is thinned at its blade root 2 and thickened at its blade tip 3. Because of this, in the third variation 8, the radial center-of-mass position rS is displaced radially inward and in the fourth variation 9 it is displaced radially outward, although the mass m does not change. The variations 8, 9 may, however, be carried out in such a way that both the mass m and the radial center-of-mass position rS are varied. Furthermore, it is possible to carry out the mass m and the radial center-of-mass position rS by thickening and/or thinning the rotor blade 1 in selected radial sections.
A multiplicity of variations of the nominal geometry 5 are carried out, and for each variation the natural frequency νS of the lowest-frequency bending oscillation of the rotor blade 1 clamped at its blade root 2 and at rest is calculated by a finite element method. Furthermore, for each variation the natural frequency νF of the same bending oscillation is calculated, the centrifugal force acting on the rotor blade 1 during operation of the turbomachine being taken into account. Optionally, an elevated temperature and material properties therefore varying may be taken into account in the calculation of νF. For a given rotor-blade cascade, it is advantageously possible only to carry out the variations of the nominal geometry once.
Subsequently, for each variation of the nominal geometry 5, the mass m and the radial center-of-mass position rS of the rotor blade 1 are determined and a value table νS(m, rS) with value triplets νS, m, rS and a value table νF(m, rS) with value triplets νF, m, rS are compiled. The value table νS(m, rS) is represented in the left-hand graph of
Subsequently, for each setpoint natural frequency νF,S, a corresponding setpoint natural frequency νS,S, which the rotor blade 1 has for the lowest-frequency bending oscillation of the rotor blade 1 firmly clamped at its blade root 2 at rest, is determined 15. Following this, as described above, the value table νS(m, rS) and the value table νF(m, rS) are compiled 16 using the variations of the nominal geometry 5.
After manufacture 18 of the rotor blade 1, its mass m and radial center-of-mass position rS are measured 19. Subsequently, the actual natural frequency νF,I of the rotor blade 1 under the centrifugal force is determined 17 by interpolation of the measured mass mI and the measured radial center-of-mass position rS,I in the value table νF(m, rS).
An actual/setpoint match 21 is carried out by comparing νF,I with νF,S. In the event that νF,I lies outside a tolerance around νF,S, a value pair mS and rS,S is selected from the value table νF(m, rS) such that νF,I at least approximates νF,S, and material is removed 24 from the rotor blade 1 in such a way that mI and rS,I correspond to the value pair mS and rS,S. As can be seen from the right-hand graph of
In order to monitor the removal 24, the natural frequency νS,I of the rotor blade 1 at rest may be measured 20. To this end, the rotor blade 1 is clamped at its blade root 2, the oscillation of the rotor blade 1 is excited, for example by impact, and the sound emitted by the rotor blade 1 is measured. As an alternative, in order to monitor the removal 24, the mass m and the radial center-of-mass position rS of the rotor blade 1 may be measured 19. The monitoring can be carried out with a particularly high accuracy by measuring both the natural frequency νS,I 20 and the mass m and the radial center-of-mass position rS 19.
It is also possible to measure both the mass m and the radial center-of-mass position rS 19 and the natural frequency νS,I 20 already before the removal 24 of the material, so as to measure the actual natural frequency νF,I with a particularly high accuracy. By a comparison of the measured natural frequency νS,I with an actual natural frequency determined by interpolation of mI and rS,I in the value table νS(m, rS), adaptation of the model for determining the natural frequencies νF and νS can be carried out.
In the event that νF,I lies inside a tolerance around νF,S, method steps 22 may optionally be carried out on the rotor blade 1, for example removal of a coating. The rotor blade 1 is subsequently installed in the rotor-blade cascade 23.
Although the invention has been illustrated and described in detail with reference to the preferred exemplary embodiments, the invention is not restricted by the examples disclosed and other variants may be derived therefrom by the person skilled in the art without departing from the protective scope of the invention.
Number | Date | Country | Kind |
---|---|---|---|
13153956 | Feb 2013 | EP | regional |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2014/051322 | 1/23/2014 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2014/122028 | 8/14/2014 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
4108573 | Wagner | Aug 1978 | A |
4924706 | Moore | May 1990 | A |
5988982 | Clauer | Nov 1999 | A |
6042338 | Brafford | Mar 2000 | A |
6814543 | Barb et al. | Nov 2004 | B2 |
7252481 | Stone | Aug 2007 | B2 |
7500299 | Dupeux | Mar 2009 | B2 |
7941281 | Jammu et al. | May 2011 | B2 |
7997873 | Slepski et al. | Aug 2011 | B2 |
20020064458 | Montgomery et al. | May 2002 | A1 |
20050249586 | Dupeux et al. | Nov 2005 | A1 |
20120148401 | Kulathu et al. | Jun 2012 | A1 |
20130177427 | Kayser | Jul 2013 | A1 |
Number | Date | Country |
---|---|---|
101762385 | Jun 2010 | CN |
103119248 | May 2013 | CN |
102009033618 | Jan 2011 | DE |
1589191 | Oct 2005 | EP |
1640562 | Mar 2006 | EP |
S54114619 | Sep 1979 | JP |
S5618764 | May 1981 | JP |
S5993901 | May 1984 | JP |
S59150903 | Aug 1984 | JP |
H01106902 | Apr 1989 | JP |
2000265803 | Sep 2000 | JP |
2002257672 | Sep 2002 | JP |
2004211705 | Jul 2004 | JP |
2007537385 | Dec 2007 | JP |
2010230006 | Oct 2010 | JP |
2012137054 | Jul 2012 | JP |
2382911 | Feb 2010 | RU |
Entry |
---|
JP Office Action dated Jul. 15, 2016, for JP application No. 2015-555656. |
Number | Date | Country | |
---|---|---|---|
20160010461 A1 | Jan 2016 | US |