METHOD AND TESTING DEVICE FOR TESTING ROTOR BLADES
20220010780 · 2022-01-13
Assignee
Inventors
- Moritz Batge (Bremen, DE)
- Malo Rosemeier (Bremen, DE)
- Bernd HALLER (Rastede, DE)
- Kwangtae Ha (Bremen, DE)
- Willi Wroblewski (Lilienthal, DE)
- Florian Sayer (Bremen, DE)
Cpc classification
F03D13/30
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
F03D17/00
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
G01M7/022
PHYSICS
F05B2260/83
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
G01M5/0058
PHYSICS
Y02E10/72
GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
International classification
F03D13/30
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
Abstract
A method for testing a rotor blade of a wind turbine may include predefining a setpoint bending moment distribution. At least two active load-introducing means may be provided which each engage on a load frame. A first of the at least two active load-introducing means may be configured for introducing load in a pivot direction of the rotor blade and a second of the at least two active load-introducing means may be configured for introducing load in an impact direction of the rotor blade. Also provided is at least one passive load-introducing means. A cyclic introduction of load is effected by the at least two active load-introducing means, where a load introduction frequency of the first active load-introducing means and a load introduction frequency of the second active load-introducing means are selected such that the ratio thereof is rational. A testing device for carrying out the method is also provided.
Claims
1-21. (canceled)
22. A method for testing a rotor blade of a wind turbine, wherein a setpoint bending moment distribution is predetermined, comprising at least the steps: clamping the rotor blade in a clamping device so that a longitudinal axis of the rotor blade extends from a clamping point of the rotor blade, attaching one or more load frames to the rotor blade, providing at least two active load-introducing means which each engage on one of the load frames, wherein a first of the at least two active load-introducing means is set up for introducing load in a pivot direction of the rotor blade and a second of the at least two active load-introducing means is set up for introducing load in an impact direction of the rotor blade, providing at least one passive load-introducing means engaging one of said load frames, wherein for a system comprising said rotor blade and said at least one passive load-introducing means, a system natural frequency for the pivot direction and/or for the impact direction is changed by said at least one passive load-introducing means, a cyclic introduction of load is effected by the at least two active load-introducing means, wherein a load introduction frequency of the first active load-introducing means and a load introduction frequency of the second active load-introducing means are selected such that the ratio thereof is rational, and wherein the load frames and the at least two active load-introducing means and the at least one passive load-introducing means are arranged at previously determined positions which are selected in such a way that a bending moment distribution introduced during cyclic load introduction does not fall below the setpoint bending moment distribution and does not exceed it by more than 20%.
23. The method according to claim 22, wherein the at least one passive load-introducing means is arranged such that the system natural frequency in the impact direction matches the load introduction frequency in the impact direction or, for applying an additional load, deviates no more than 10%, from the system natural frequency in the impact direction.
24. The method according to claim 21, wherein the at least one passive load-introducing means is arranged such that the system natural frequency in the pivot direction coincides with the load introduction frequency in the pivot direction or, for introducing an additional load, deviates no more than 10%, from the system natural frequency in the pivot direction.
25. The method according to claim 21, wherein the at least two active load-introducing means are controlled or regulated for adjusting and maintaining the ratio between the load introduction frequency of the first active load applying means and the load-introduction frequency of the second active load-introducing means.
26. The method according to claim 21, wherein at least one passive load-introducing means is provided for adjusting the system natural frequency in the impact direction and at least one passive load-introducing means is provided for adjusting the system natural frequency in the pivot direction.
27. The method according to claim 21, wherein the rotor blade is clamped such that the longitudinal axis of the rotor blade extends substantially horizontally and the pivot direction is oriented horizontally.
28. The method according to claim 21, wherein the at least two active load-introducing means and/or the at least one passive load-introducing means are externally anchored.
29. The method according to claim 21, wherein the at least one passive load-introducing means is selected from fixed masses, decoupled masses and elastic elements.
30. The method according to claim 21, wherein at least one of the at least one passive load-introducing means is chosen as a decoupled mass or as an elastic element.
31. The method according to claim 21, wherein the ratio between the load introduction frequency of the first load-introducing means and the load introduction frequency of the second load-introducing means is 1:1, 2:1 or 1:2.
32. The method according to claim 21, wherein a distribution of moment vector paths occurring in the field is determined, via an aeroservoelastic multi-body simulation, and a setpoint middle bending moment is set by the passive load-introducing means by the decoupled masses and/or by a bias of the elastic elements, such that a midpoint of moment vector paths generated by means of the active load-introducing means lies within the distribution of moment vector paths occurring in the field.
33. The method according to claim 32, wherein the moment vector paths generated by the active load-introducing means are adapted to a contour of the distribution of the moment vector paths occurring in the field by adjusting the rotor blade clockwise or counterclockwise about its longitudinal axis.
34. The method according to claim 33, wherein the moment vector paths generated by the active load-introducing means are matched to a contour of the distribution of the moment vector paths occurring in the field by adjusting a phase angle between the load introduction in the pivot direction and the load introduction in the impact direction.
35. The method according to claim 21, wherein the ratio between the load introduction frequencies is 1:1 and a phase angle between the load introduction in the pivot direction and the load introduction in the impact direction is between 0° and 180°.
36. The method according to claim 21, wherein the ratio between the load introduction frequencies is 1:2 or 2:1 and a phase angle between the load introduction in the pivot direction and the load introduction in the impact direction is 0° or 180°.
37. The method according to claim 21, wherein a minimum distance between adjacent load frames is 1 m.
38. A testing device for testing a rotor blade, comprising: a clamping device for clamping a rotor blade at a clamping point of the rotor blade, so that a longitudinal axis of the rotor blade extends substantially horizontally starting from the clamping point, one or more load frames to be attached to the rotor blade, at least two active load-introducing means, each of which is attached to one of the load frames, at least one passive load-introducing means attached to one of the load frames, wherein at least one of the at least two active load-introducing means is configured for introducing loads into the rotor blade in a pivot direction, and at least one of the at least two active load-introducing means is configured for introducing loads into the rotor blade in an impact direction, and wherein the at least one passive load-introducing means is externally anchored and comprises an elastic element and/or a decoupled mass and is adapted to adjust a system natural frequency in the pivot direction and/or in the impact direction, for a system comprising the rotor blade and the at least one passive load-introducing means.
39. The testing device of claim 38, further comprising a control device adapted to control a load introduction frequency of the first active load-introducing means and a load introduction frequency of the second active load-introducing means such that a ratio between these two load introduction frequencies is rational.
40. The testing device according to claim 38, wherein the active load-introducing means is in the form of a hydraulic, or pneumatic or electric actuator and is externally anchored.
41. The testing device according to claim 38, wherein the elastic element is designed as a spring or torsion spring or beam or leaf spring and is externally anchored.
42. The testing device, according to claim 28, wherein the elastic element comprises a fibre composite comprising glass fibres and/or carbon fibres.
Description
[0064] In the following, the invention is explained in more detail by way of example with reference to figures.
[0065] In the drawings:
[0066]
[0067]
[0068]
[0069]
[0070]
[0071]
[0072]
[0073]
[0074]
[0075]
[0076]
[0077]
[0078]
[0079]
[0080]
[0081]
[0082] An active load-introducing means 5 is arranged on the rotor blade, which is designed as a hydraulic, pneumatic or electric actuator. The actuator is connected by means of joints 11A, 11B to a load frame 4A attached to the rotor blade and also to a ground 3 via an additional rod. The actuator acts as an active load-introducing means in the impact direction of the rotor blade, thus substantially vertically, although at least when the rotor blade is vibrated, the actuator can deviate from vertical thanks to the joints 11A, 11B and continue to introduce load into the rotor blade.
[0083] A further load frame 4B is arranged further out on the rotor blade 1 and holds a fixed mass 15 as a passive load-introducing means, by which the oscillation behaviour of the rotor blade 1 is influenced. Due to the fixed mass 15, on the one hand, the bending moment acting on the rotor blade is changed and, on the other hand, the natural frequencies of the oscillating rotor blade 1 are modified. That is, a system natural frequency for a system comprising the rotor blade 1 and the passive load-introducing means is modified with respect to the natural frequency of the rotor blade 1 in the pivot direction and in the impact direction due to the passive load-introducing means formed as a fixed mass 15. The actuator may be operated so that its load introduction frequency matches the system natural frequency in the impact direction in which the actuator is acting. However, it can also be operated with a detuning of, for example, less than 1% with respect to the system natural frequency in order to act as a so-called actuator mass or actuator spring and introduce an additional load into the rotor blade.
[0084] The structure shown in
[0085] In addition to the elements shown in
[0086] It may be, for example, that the design shown in
[0087]
[0088]
[0089]
[0090] At a first load frame 4A arranged closest to the clamping point 1′, an elastic element configured as a spring 7A is configured as a passive load-introducing means so as to act in the impact direction of the rotor blade 1. The spring acts on the load frame 4A via a lever arm 9A, which pivots about a hinge 10A, and a vertical rod 8A, which is connected to the lever arm, and acts to distribute the bending moment in the impact direction.
[0091] A second load frame 4B has a similar device arranged thereon as in the case of the first load frame 4A, in which a spring 7B acts as a passive load-introducing means. The spring 7B acts on one end of a lever arm 9B. At the other end of the lever arm 9B, it sits on a hinge 10B. Above the hinge 10B, an angled beam 16A is also provided extending upwardly at right angles to the lever arm 9B. The angled beam is in turn connected to the load frame 4B by a horizontally extending rod 8B. That is, the passive load of the spring 7B is redirected by the device such that the bending moment distribution in the vicinity of the second load frame 4B is modified in the pivot direction. Designs in which a lever arm movably mounted on a joint is equipped with an angled beam and rod are suitable for deflecting the acting forces into the horizontal, i.e. in the present configuration into the pivot direction of the rotor blade 1. Such a device can also be used in conjunction with other load-introducing means, such as active load-introducing means or decoupled masses, as will be further described below.
[0092] An actuator 5A is arranged on a third load frame 4C as an active load-introducing means, which is configured as shown in
[0093] A further actuator 5B is arranged on a fourth load frame 4D, which acts in the pivot direction. This is achieved, analogously to the case of the load frame 4B, by connecting the further actuator 5B to a lever arm 9C movably mounted on a hinge 10C. The lever arm is connected to the fourth load frame 4D via an upwardly directed angled beam 16B and a horizontal rod 8C, and the force applied by the further actuator 5B is thus redirected in the pivot direction.
[0094] Further towards the tip of the rotor blade 1, two more load frames 4E, 4F are arranged, each connected to a decoupled mass 6A, 6B. The decoupled mass acting on the load frame 4E acts in the impact direction, just as in the case of
[0095] In the present case, the elastic elements are located closest to the clamping point, followed by active load-introducing means and finally the decoupled masses located furthest out. This configuration can be advantageous for producing bending moment distributions typical of rotor blades.
[0096] By means of the passive load-introducing means provided, on the one hand the bending moment distributions in the pivot direction and in the impact direction are adapted to the setpoint bending moment distribution in the respective direction in such a way that the bending moment distribution during testing is not more than 10% or not more than 5% above the setpoint bending moment distribution and at the same time does not fall below it. On the other hand, the system natural frequencies for the rotor blade 1 with the load-introducing means arranged thereon are changed.
[0097] The direction of action of the rods 8B, 8C, 8E for the passive and active load-introducing means in the pivot direction is selected to coincide with the direction of pivoting movement at the load introduction point in the rest state. The load-introducing means can then act decoupled from the impact moment. Exactly the opposite is true for the rods 8A, 8D of the load-introducing means for the impact direction. Specifically, this may mean that at the various sections where the load frames are located, the pivot and impact modes and their respective directions of movement are determined for the system with the passive load-introducing means located thereon, and the rods are arranged as described based on the determined directions of movement. During the test, the rods are then deflected equally in both directions from their resting state.
[0098] Now, when the setup described in
[0099] The passive load-introducing means adjust the system natural frequencies in the impact direction and in the pivot direction so that they do not deviate by more than 10% from the load introduction frequency in the respective direction.
[0100] In particular, it may be that excitation is exactly at the system natural frequencies, i.e. the load introduction frequencies correspond to the system natural frequencies in the respective direction. This allows the test duration and energy consumption to be further reduced.
[0101] However, it is also possible that the load introduction frequencies are detuned with respect to the system natural frequencies prevailing in the respective direction, i.e. for example a deviation of at most 10% or at most 5% or at most 2% or at most 1% with respect to the system natural frequency is set. This allows an additional load to be applied to the rotor blade 1.
[0102] The actuators 5A, 5B of the active load-introducing means are designed as hydraulic, pneumatic or electric actuators. In the present example, they are externally anchored, whereby the described possible additional load introduction can be advantageously achieved by detuning the excitation frequencies. However, it is also possible to design the active load-introducing means as mass exciters, that is, as actuators fixed to the rotor blade 1 which cyclically move a mass attached thereto.
[0103] In the proposed method or associated testing devices, typically at least one of the passive load-introducing means is configured as a decoupled mass or an elastic element and is connected to the ground or wall. Furthermore, it is typical in the method or testing devices that at least one passive load-introducing means is set up for load application in the pivot direction and at least one passive load-introducing means is set up for load introduction in the impact direction.
[0104] To illustrate the invention, another example using specific numerical values will be given: For example, it may be that the rotor blade 1 in the unprepared state has a natural frequency in the impact direction of 0.7 Hz and a natural frequency in the pivot direction of 0.9 Hz. Then, for example, an adjustment of the natural frequencies can be made via the attachment of the passive load-introducing means, in which the system natural frequency in the impact direction is raised to 0.9 Hz in order to produce a frequency ratio of f.sub.impact:f.sub.pivot=1:1. However, the system natural frequency can also be raised to 1.4 Hertz in the pivot direction to produce a frequency ratio of f.sub.impact:f.sub.pivot=1:2.
[0105] Then the load introduction can be carried out with the corresponding frequency ratio in resonance, i.e. the load introduction frequency can be selected equal to the system natural frequency in each case. The cycles can then be easily observed and counted. In the case of the frequency ratio 1:1, for example, 3 million cycles can be performed for each direction. For example, at the frequency ratio of 1:2, 1 million cycles can be performed in the impact direction and 2 million cycles can be performed in the pivot direction, and the test is preferably completed simultaneously for both directions. A frequency ratio of 2:1 is therefore advantageous, for example, if twice as many cycles are provided in one direction as in the other for applying the desired load.
[0106] To adjust and maintain the frequency ratio, the test rig has a control and/or regulating device. This allows the blade to perform movements corresponding to closed Lissajous figures throughout the test period.
[0107]
[0108]
[0109]
[0110] In the arrangement shown in
[0111] The structure shown in
[0112]
[0113]
[0114]
[0115]
[0116]
[0117] In the embodiments of
[0118]
[0119]
[0120]
[0121]
[0122]
[0123] The embodiment illustrates how a load-introducing means can be provided for each of the two directions on a single load frame. Other types of load-introducing means can also be combined with each other. For example, any of the load introducing means of
[0124] In the example shown, the decoupled mass 6 causes a preload in the impact direction, whereas the beam 13 does not cause a preload in the pivot direction.
[0125]
[0126]
[0127]
[0128] In the optimization process, constants or constraints K1-K5, which are given and cannot be changed, as well as certain target values Z1-Z4, which are to be achieved as well as possible by means of the optimization, are used as input values. In a high-dimensional optimization procedure, design variables D1-D7 are then adjusted to arrive at a setup and experimental design that achieves the target values as well as possible.
[0129] The constants or constraints K1-K4 include: [0130] K1: the frequency ratio between the load applied in the impact direction and the load applied in the pivot direction. It is limited to f.sub.impact:f.sub.pivot=n1:n2, where n1 and n2 are integers. Preferably, f.sub.impact:f.sub.pivot=1:2 or fStroke:f.sub.pivot=1:1. [0131] K2: permissible installation area of the load frames. For each rotor blade 1, there is typically a minimum distance from the clamping point 1′ and a minimum distance from the blade tip. Design-critical areas can also be defined in between, in which no load frame may be located. In addition, a minimum distance between two adjacent load frames is usually specified. [0132] K3: maximum actuator deflection. Due to the design of the actuator used, but also due to the dimensions of a hall in which the testing device is placed, the maximum deflection that cannot be exceeded is predetermined. [0133] K4: maximum actuator force. The force of the actuator is limited by its design as well as by load limits of the rotor blade. [0134] K5: Phase angle between the load introduction in the impact direction and the pivot direction. The phase angle can be adjusted to modify a superposition of the load in the impact direction and the pivot direction. It may be envisaged that the phase angle, for example in the case of f.sub.impact:f.sub.pivot=1:1 is chosen as 90°, so that when the maximum load is applied in one of the directions, no load is applied in the other. With f.sub.impact:f.sub.pivot=1:2, for example, 0° or 180° can be selected as the phase angle.
[0135] Quantities that are to be achieved as accurately as possible by means of optimization are the target values Z1-Z3: [0136] Z1: Bending moment distribution. First and foremost, it should be achieved that the bending moments introduced in the process correspond as closely as possible to the setpoint bending moment distribution. Falling below the setpoint bending moment distribution is completely avoided. Furthermore, for example, a local exceeding of the nominal bending moment of maximum 10% can be aimed at. Target value Z1 can be prioritized over other target values. [0137] Z2: Energy consumption. Energy consumption should typically be kept as low as possible. Energy consumption can be optimized by aligning the load introduction frequencies and the system natural frequencies. [0138] Z3: Test duration. The test duration should also be kept as short as possible, which can sometimes be achieved by aligning the load introduction frequencies and the system natural frequencies. [0139] Z4: setpoint middle bending moment. If the field conditions of the rotor blade are known from a measurement or a simulation, a setpoint middle bending moment adapted to this can be set (see also
[0140] To achieve the target values Z1-Z3, the design variables D1-D6 are available as adjusting screws of the optimization, which can be varied: [0141] D1: load frame position. The position of the load frames can be changed in the frame defined by the constant K2, so that additional load frames can be arranged in areas of large deviations from the setpoint bending moment distribution, over which loads are then introduced by means of the load-introducing means (see D3-D5). [0142] D2: number of load frames. The number of load frames can be changed, again taking into account the constant K2. Typically, care is taken to keep the number of load frames as small as possible, since in the immediate vicinity of a load frame, the rotor blade 1 is not loaded or not loaded strongly enough. For example, 2 or 3 load frames may be desired. [0143] D3: Elastic elements. Elastic elements, as described in the figures above, represent a possible passive load-introducing means with which the bending moment distribution can be influenced and the system natural frequency can be changed. They can be used in such a way that the parameters are only changed in the impact direction or in the pivot direction. They can be attached to one or more of the load frames. They can be designed with or without preload or mean load. [0144] D4: fixed masses. Fixed masses can also be used to adjust bending moments and system natural frequencies. They cause a change in bending moments and system natural frequencies in both directions and are typically used when an adjustment is needed that requires it. The fixed masses can be attached to one or more of the load frames. [0145] D5: Decoupled masses. Like elastic elements, decoupled masses can be used specifically for an adjustment in the impact direction or in the pivot direction. The decoupled elements can have a reversed direction of action compared to the elastic elements and can accordingly be used for opposite corrections. The decoupled masses can be attached to one or more of the load frames. They can be designed with or without preload or mean load. [0146] D6: Frequency offset of the load introduction frequencies. As mentioned, the load introduction frequencies can be detuned relative to the system natural frequencies to better achieve the setpoint bending moment distribution. This can be at the expense of energy consumption. [0147] D7: Angle of attack of the rotor blade. The angle of attack (also called pitch angle), can be modified to adjust the load introduction (see also
[0148] The optimisation process then results in a concrete arrangement of load frames and of active and passive load-introducing means arranged thereon, for example the structure shown in
[0149] At this point, an example is given for illustrative purposes.
[0150] For an exemplary rotor blade, a natural frequency in the impact direction can be 0.5 Hz and a natural frequency in the pivot direction can be 1 Hz in an unprepared state. In this example, in order to achieve the target load, at least 1 million cycles are required in the impact direction and 3 million cycles in the pivot direction. This would result in a test duration of 23.1 days for the impact direction and 34.7 days for the impact direction according to conventional methods, which would correspond to a total test duration of 57.8 days. According to the method presented here, the natural frequencies can be converted into system natural frequencies. These can have a rational ratio of 1:1, approximately both 0.75 Hz. The amplitude of the load introduction in the impact direction can be changed in such a way that the desired damage is also achieved in this direction after 3 million cycles. The test is then carried out in both directions simultaneously, is completed for both directions at the same time and takes a total of 46.3 days. This reduces the test duration by 20% compared to the conventional procedure.
[0151]
[0152]
[0153] By means of static middle bending moments in the impact direction and pivot direction (vector L1), which are introduced with the decoupled masses or pre-stressed spring elements described in this document, the centre of the ellipse L5 achieved in the test procedure described here, which was displaced along L2 under the influence of gravity, can now be further displaced along the vector L1 towards a point L8, which represents a centre of the probability cloud L7. This corresponds to the setting of the setpoint middle bending moment. Alternatively or additionally, by pitching the rotor blade, the effective direction of the gravitational vector L2 relative to the ellipse L5 can be influenced so that it coincides with the effective direction of L1. A phase angle adjustment of greater/smaller 90° allows the ellipse to be rotated counterclockwise/clockwise. If the pivot amplitude and impact amplitudes are further scaled, it is possible to transform the ellipse L5 into the ellipse L6 and thus to approximate the moment cloud L7 occurring in the field.
[0154] The advantage of this method is a more realistic (closer to the most frequently occurring field conditions) loading of the rotor blade. Furthermore, the shift of the center of the ellipse changes an occurring stress ratio R, which can lead to a reduction of the test cycles if the setpoint bending moment distribution is adjusted due to the changed setpoint middle bending moment, so that a uni-form target damage of the load bearing areas occurs along the areas of the blade to be tested (see for example the publication Rosemeier et al. “Benefits of subcomponent over full-scale blade testing elaborated on a trailing-edge bond line design validation”, Wind Energ. Sci., 3, 163-172, 2018, https://doi.org/10.5194/wes-3-163-2018).
[0155]
[0156]
[0157]
LIST OF REFERENCE NUMERALS
[0158] 1 rotor blade [0159] 1′ clamping point of the rotor blade [0160] 2 clamping device [0161] 3 ground [0162] 4 load frame [0163] 5 actuator [0164] 5′ actuator point of attack [0165] 6 decoupled mass [0166] 7 spring [0167] 8 rod [0168] 9 lever arm [0169] 10 hinge [0170] 11 joint [0171] 12 torsion spring [0172] 13 elastic beam [0173] 14 leaf spring [0174] 15 fixed mass [0175] 16 angled beam
[0176] If several elements of a category are present in a figure, these are additionally marked with a letter to distinguish them.
[0177] Constants: [0178] K1 frequency ratio impact direction/pivot direction [0179] K2 permissible installation area of the load frames [0180] K3 maximum actuator deflection K4 maximum actuator force K5 phase angle between load introduction impact direction/pivot direction
[0181] Design Variables: [0182] D1 load frame position [0183] D2 number of load frames [0184] D3 elastic elements [0185] D4 fixed masses [0186] D5 decoupled masses [0187] D6 frequency offset of the load introduction frequencies to the system natural frequencies [0188] D7 angle of attack (pitch angle) of the rotor blade
[0189] Target Values: [0190] Z1 setpoint bending moment distribution [0191] Z2 minimum energy consumption [0192] Z3 minimum test duration [0193] Z4 setpoint middle bending moment (center point)