Test bench and procedure for testing a storage arrangement

Abstract

Method for testing a storage arrangement (2) comprising the following features: - Environmental data relating to external influences affecting a plant including a storage arrangement (2) are provided, - Based on the environmental data, different environmental scenarios are classified, which differ from each other in terms of their influence on the system, - Using the environmental scenarios that are relevant with regard to the service life of the bearing arrangement (2), data relating to various realistic load scenarios (BS1, BS2, BS3) of the bearing arrangement (2) during the estimated service life of the bearing arrangement (2) are determined, - the proportion of the different load scenarios (BS1, BS2, BS3) to the estimated service life of the bearing arrangement (2) is determined, - Regarding each load scenario (BS1, BS2, BS3), load parameters influencing lifespan (nr, Br, B) are determined. ave , B min , B max ) determined, - at least for a subset of the load scenarios (BS1, BS2, BS3) schemes for modifying the load parameters (B ave , B min , B max ) created, whereby the modified load parameters (P ave , P min , P max ) a mapping of the realistic load scenario (BS1, BS2, BS3) onto a shortened test scenario (PS1, PS2, PS3) results, - Data on the decrease in the dependence of the modified load parameters (P ave , P min , P max The test results obtained from realistic loads and realistic load durations, showing the degree of shortening, are determined for various load scenarios (BS1, BS2, BS3). - For different load scenarios (BS1, BS2, BS3), specific shortening factors are determined, which indicate the ratio between the duration of the test scenario (PS1, PS2, PS3) and the duration of the load scenario (BS1, BS2, BS3). - the bearing arrangement (2) is subjected to a test with the determined test scenarios (PS1, PS2, PS3), which together determine the loads (B) that occur during the estimated service life (Tnutz) and affect the service life of the bearing arrangement (2). ave , B min , B max ) depict, subjected to, whereby the percentage temporal composition of the different test scenarios (PS1, PS2, PS3) differs from the temporal composition of the associated realistic load scenarios (BS1, BS2, BS3).

Description

[0001] The invention relates to a method for testing a bearing arrangement, in particular a rotor bearing for a wind turbine. The invention also relates to a test rig designed for testing a bearing arrangement.

[0002] German patent DE 10 2010 017 456 B4 discloses a test rig designed for testing components of wind turbines. The test rig comprises a load application frame, the distance between which and the test specimen is adjustable.

[0003] Another device for testing the load-bearing capacity of a large bearing is disclosed in DE 10 2014 219 727 A1. This device is also intended for testing components of wind turbines. One test procedure includes varying the drive speed and generating torsional vibrations.

[0004] Another bearing testing device is disclosed in DE 10 2009 026 943 B3. The testing device has a torsional oscillating chain which includes a flywheel and a torsionally flexible torsion element.

[0005] US Patent 2009 / 0 198 470 A1 describes a test rig for accelerated lifetime testing of a bearing assembly. The concept requires regular test rig stops for manual bearing disassembly and visual inspection, which increases time and costs.

[0006] Document CN 103 246 821 A discloses a test rig in which a bearing assembly under test is cyclically subjected to radial and axial loads by means of hydraulic cylinders. The load profiles are fixed and programmable, allowing only limited amplitude and frequency variation. Sensor signals are used only for recording, not for adaptive real-time control of the test parameters, meaning that unforeseen bearing wear progression can only be detected retrospectively.

[0007] Document CN 103 674 546 A describes a test rig for fatigue testing of large rolling bearings, which uses two counter-rotating disks to apply combined bending moments. The mechanical structure features rigid table supports, meaning that bending and tilting moments can only be generated in two dimensions.

[0008] CN 106 021 928 A discloses a method for the endurance testing of bearing assemblies in which temperature and lubricant viscosity are kept constant during the test to achieve statistically reproducible results. However, this approach ignores critical influencing factors such as thermal gradients, lubricant aging, and contamination, which in practice significantly contribute to bearing failure.

[0009] Escobar and Meeker (A Review of Accelerated Test Models, 2006, in: Statistical Science, 2006, Vol. 21, No. 4, 552–557. DOI: 10.1214 / 088342306000000321 [accessed on September 28, 2018]) provide an overview of statistical models for accelerated fatigue testing and demonstrate methods for extrapolating test results to operating conditions. However, the discussion focuses on mathematical evaluation and does not include a concrete implementation on a mechanical test bench.

[0010] German patent DE 10 2012 024 273 A1 describes a method for adjusting load-dependent processes in wind turbines. The method includes determining the individual remaining service lives of wind turbine components.

[0011] For the design of bearings, especially rolling bearings, their fatigue life can generally be calculated. However, there are limits to the computational accuracy of factors that affect a bearing's service life. Against this background, application-specific bearing tests are common practice in many industries. Naturally, the effort required for these tests increases with the size of the bearing. This is not only due to the size of the required testing equipment, but also to the fact that large bearings, such as those used in wind turbines, are typically designed for a significantly longer service life compared to smaller bearings. For example, testing a bearing assembly designed for a service life of 20 to 25 years should take no more than three to six months.Within this test period, components of the bearing arrangement should be tested, for example, with regard to failure mechanisms listed in the standard ISO 15243:2017 (Rolling bearings - Damage and failures).

[0012] High-speed endurance tests with significantly shortened test durations compared to the designed service life are also known as HALT (Highly Accelerated Lifetime Test). In such tests, the loads acting on a bearing are increased compared to the loads encountered in real-world operation to such an extent that the failure events observed during test operation occur with a comparable, ideally identical, frequency to those occurring in actual bearing operation. The loads acting on the bearing include, in particular, forces, moments, and rotational speeds. Design loads for wind turbines, for example, can be found in the IEC 61400-1 standard.

[0013] The test bench operation should simulate the real-world operation of a bearing arrangement as realistically as possible in accelerated time. However, the more drastically the test duration is shortened compared to the estimated service life, while simultaneously increasing the loads, the greater the likelihood that the test will lead to damage mechanisms that would never occur in real-world operation. This creates a conflict between shortening the test duration on the one hand and ensuring the test's reliability on the other.

[0014] The invention is based on the objective of mitigating this conflict of objectives when testing bearing arrangements.

[0015] This problem is solved according to the invention by a method for testing a bearing arrangement with the features of claim 1. Likewise, the problem is solved by a test rig according to claim 11. The embodiments and advantages of the invention explained below in connection with the device, i.e., the test rig, also apply mutatis mutandis to the testing method and vice versa.

[0016] The testing procedure includes the following features: Environmental data relating to external influences acting on a system, including a storage arrangement, are provided; the environmental data includes, in particular, wind data. - Based on the environmental data, different environmental scenarios are classified, which differ from each other in terms of their influence on the system, - Using environmental scenarios relevant to the service life of the bearing arrangement, data relating to various realistic load scenarios of the bearing arrangement during its estimated service life are determined. - The proportion of the various load scenarios to the estimated service life of the bearing arrangement is determined, whereby the load scenarios to be considered subsequently typically do not add up to 100% of the estimated service life, since scenarios that are irrelevant with regard to the service life of the bearing arrangement were already eliminated during the classification of the environmental scenarios. - For each load scenario, load parameters affecting service life are determined, - at least for a subset of the load scenarios, schemes for modifying the load parameters are created, whereby the modified load parameters result in a mapping of the realistic load scenario to a shortened test scenario, - Data on the decrease in the dependence of test results obtained with modified load parameters on results obtained with realistic loads and realistic load duration on the degree of shortening are determined for various load scenarios, whereby without shortening the test duration the closest coupling between events observed in real operation and events observed in test bench operation is given, - Specific shortening factors are determined for different load scenarios, which indicate the ratio between the duration of the test scenario and the duration of the load scenario. - The bearing arrangement is subjected to testing with the determined test scenarios, which together represent the loads occurring during the estimated service life that affect the service life of the bearing arrangement, whereby the percentage temporal composition of the various test scenarios differs from the temporal composition of the associated realistic load scenarios.

[0017] Initially, it is assumed that not all environmental influences necessarily have a relevant impact on the service life of a system. In the case of a wind turbine, wind naturally plays the primary role, and its probability of occurring at different speeds varies. Therefore, a selection can be made at this stage, creating the conditions for a shortened test even before a precise understanding of the system to be erected. The invention further considers that various parameters can be varied to shorten the test duration compared to the actual service life of a bearing arrangement. The modification of these parameters, compared to actual operation, affects both the achievable time-acceleration effect and potential distortions of the test result in different ways.

[0018] To minimize the distortion of the test results compared to the design-compliant operation over the entire service life of the bearing, the categorizations according to the design can be adopted in the test operation with regard to the differentiation of different load scenarios.

[0019] Key load parameters that play a role in the various scenarios include forces, rotational speeds, and rates of change of at least one of these parameters. The load parameters occurring in a test scenario can encompass mean values ​​and distributions, particularly minimum and maximum values, of loads. Furthermore, different load parameters may be modified in various ways during testing, i.e., within a specific test scenario, relative to the corresponding loads expected in realistic operation.

[0020] The stress scenarios include, for example, the following scenarios: - Normal operation of the storage arrangement, optionally divided into different sub-scenarios, - Start-up operation, - Speed ​​reduction to a standstill, i.e., shutting down the system, - Idle oscillation.

[0021] As a special case of shutting down the system, including the bearing assembly, to a standstill, an emergency stop can also be considered. It has been shown that separate test scenarios are not required for the "emergency stop" and "regular shutdown" scenarios. Rather, the load scenario "speed reduction to standstill" is preferably covered by the "emergency stop" scenario, whereby in the corresponding test scenario, the angular acceleration of the bearing assembly component being decelerated is increased by a factor greater than two and less than five, in particular by a factor of three, compared to a realistic load scenario resulting from the regular speed reduction to standstill.In a particularly preferred, realistic procedure, during braking in the "emergency stop" test scenario, the sign of an axial load acting on the bearing arrangement changes, whereby at the end of the emergency stop process the axial load has the same sign as at the beginning of the process.

[0022] In the simulated load scenario "normal operation of the bearing arrangement" during testing, the mean value of a load, in particular a tilting moment, is preferably increased compared to the corresponding realistic load scenario, while simultaneously reducing the range of the load acting on the bearing arrangement, i.e., the difference between maximum and minimum load. Alternatively, it is possible to keep the mean value of a load constant in the test scenario while simultaneously increasing the amplitude and / or frequency of the load.

[0023] In the "start-up operation" load scenario, when testing the bearing arrangement, the angular acceleration is preferably increased by a factor greater than two and less than five, in particular by a factor of three, compared to the realistic load scenario, while the number of start-up operations is increased by a factor greater than 80 and less than 1000, in particular at least 200 and at most 300, for example 3 5 The value is reduced to 243. Similarly, the aforementioned factor can range from 100 to 200.

[0024] In the test scenario, which simulates the "idle oscillation" load scenario, the swivel angle between the relative rotatable components of the bearing arrangement is preferably limited to a value of at least 1 degree and at most 15 degrees, and in particular not higher than 12 degrees. In this test scenario, the mechanical loads are not increased compared to the realistic load scenario, but the same angular ranges of the components of the bearing arrangement are always subjected to load during each oscillation. The swivel angle to be selected for the test can also be determined based on the contact width b (in Hertz) in the raceway contact, i.e., the width of the contact between the rolling element, usually a roller, and the raceway. The area swept during the swivel movement is preferably at least 2 2 ×b and a maximum of 2 5 × b.

[0025] By selecting a defined swivel range, which means specifying a particular area that is repeatedly subjected to loads, any changes that do not yet indicate component damage, but already provide corresponding indications, for example of brinelling, can be detected with a small number of load cycles compared to real operation.

[0026] The various load scenarios to be simulated by the test scenarios can be combined into several test modules, wherein a first test module comprises the normal operation of the bearing arrangement including a fluctuating tilting moment, the start-up operation, and the speed reduction to standstill, a second test module includes a load on the bearing arrangement with higher, constant loads compared to the first test module, namely axial and radial loads as well as a higher tilting moment, and a third test module represents the idling oscillation with a defined angular relationship between the various bearing components. Here, the third test module is preferably divided into a plurality of sections, which are inserted intermittently between sections of the second test module.In each of these sections, which together form the third test module, the set of rolling elements is positioned in a defined manner at the beginning of the respective section.

[0027] The test bench according to the invention comprises a control and evaluation device designed for carrying out the method according to claim 1. The control and evaluation device can be implemented by any number of data processing units, which do not necessarily have a spatial relationship to the mechanical components of the test bench.

[0028] According to the invention, the various load scenarios are not uniformly shortened in test bench operation in relation to the actually expected operation, i.e., in comparison to the design scenario. Rather, a specific time compression is applied for each load scenario.

[0029] The parameters that can be adjusted in the individual load scenarios to shorten the test duration are typically different for the various load scenarios, with individual parameters being relevant in several or all load scenarios.

[0030] The set of parameters that can influence the test duration typically varies for different stress scenarios. If a specific stress scenario can be streamlined for testing purposes by modifying the stress parameters involved, those parameters should be selected and prioritized that offer the greatest time reduction relative to the associated loss of test results. This process is performed separately for each stress scenario. Thus, each stress scenario has its own specific scheme for modifying the stress parameters. The accuracy of the mapping from a stress scenario to a test scenario varies across the different schemes assigned to the various stress scenarios.This inconsistency is exploited to achieve an optimized acceleration of the test by setting the factor determining the acceleration effect differently for each load scenario. The sequence of scenarios comprising the test does not necessarily correspond to the sequence of load scenarios according to the design specifications. In any case, the temporal weighting of the test scenarios within the overall test differs from the temporal weighting of the corresponding load scenarios within the design operation.

[0031] The procedure for testing a bearing arrangement, particularly a rolling bearing for a wind turbine, is distinguished by its reliance on natural factors that fundamentally influence the operation of the entire system, especially the wind turbine. Based on these factors, environmental scenarios are first defined. These scenarios, insofar as they affect the system's service life, are used to generate load scenarios specific to the bearing arrangement. Various load scenarios are then mapped to abbreviated test scenarios in different ways, thereby minimizing the overall test duration and simultaneously reducing any undesirable, distorting effects of streamlining the test procedure.

[0032] An embodiment of the invention is explained in more detail below with reference to a drawing. The drawing shows: Fig. 1. A test bench including the bearing arrangement to be tested, shown in a schematic representation. Fig. 2. A frequency distribution of wind speeds to be taken into account for the design of a wind turbine, Fig. 3 the ones from Fig. 2. derived cumulative frequency of wind speeds, Fig. 4. The dependence of the power output of a wind turbine on wind speed, Fig. 5. The dependence of the rotor speed of the wind turbine on the wind speed, Fig. 6. the dependence of an effective mechanical load acting in a bearing arrangement of the wind turbine on the wind speed, Fig. 7 fundamental relationships between different load scenarios that occur during the operation of the wind turbine and test scenarios used to test the bearing arrangement of the wind turbine, Fig. 8 to 11 different possible relationships between realistic loads on the bearing arrangement during operation of the wind turbine and loads in associated test scenarios, Fig. 12 to 14 different possible relationships between the degree of shortening of a test scenario compared to the assigned realistic stress scenario and the associated loss of significance of the test results, Fig. 15 to 18 loads acting within the bearing arrangement at different wind speeds, Fig. 19 the frequency dependence of occurring tilting moments in the bearing arrangement, Fig. 20 the time dependence of a tilting moment acting under realistic conditions as well as under test conditions, Fig. 21 the time dependence of the rotor speed of the wind turbine during the start-up process, Fig. 22 and Fig. 23 the time dependence of axial or radial loads during the start-up process, Fig. 24 in a diagram analogous Fig. 22 and Fig. 23 the time dependence of a tilting moment acting in the bearing arrangement during the start-up process, Fig. 25 the course of the angular acceleration during the test scenario “start-up process”, Fig. 26 the increase in rotor speed during the test scenario “start-up process”, Fig. 27 in a diagram analogous Fig. 21 the decrease in rotor speed when the system is shut down until the rotor comes to a standstill, Fig. 28 to 30 in diagrams analogously Fig. 22 to 24 the temporal change of various loads during the process after Fig. 27, Fig. 31 in a diagram analogous Fig. 25 the angular acceleration in the test scenario “emergency stop”, Fig. 32 in an idealized representation the decrease in rotor speed in the load scenario “emergency stop”, Fig. 33 realistically the decrease in rotor speed in the test scenario “emergency stop”, Fig. 34 the course of an axial load acting in the bearing arrangement in the test scenario “emergency stop”, Fig. 35 and Fig. 36 the course of a radial load or a tipping moment during the test scenario “emergency stop”.

[0033] In Fig. Figure 1 schematically sketches a section of a test rig designated as 1. Test rig 1 contains a bearing arrangement 2 to be tested, which comprises two rolling bearings, in this case angular contact roller bearings 3 and 4. A test object 5, namely a hub of a wind turbine, is supported by the bearing arrangement 2. Forces are introduced into the test object 5, and thus also into the bearing arrangement 2, via a test rig bearing 6, which here is designed as a double-row roller bearing, by means of an actuator arrangement 7, which is part of test rig 1. Instead of the Fig. In addition to the arrangement sketched in Figure 1, other bearing arrangements can also be tested using test rig 1. Examples include fixed-floating bearing arrangements consisting of a spherical roller bearing acting as a fixed bearing and an additional spherical roller bearing acting as a floating bearing. Double-row bearings, such as bearings constructed like the auxiliary bearing used in this case as test rig bearing 6, can also be tested. Testing bearing arrangements consisting of three bearings using test rig 1 is also possible. This could, for example, be an axial-radial axial bearing arrangement comprising two axial cylindrical roller bearings and one radial cylindrical roller bearing.

[0034] The actuator arrangement 7 is connected via data technology to a control and evaluation device 8, which is further coupled to a sensor arrangement 9 that detects loads acting on the bearing arrangement 2 as well as state variables, for example bearing temperatures and axial positioning.

[0035] In the Fig. In the coordinate system shown in Figure 1, the x-axis points in the axial direction of the test object 5, that is, translated to real-world operation, typically in the direction of the wind, apart from an inclination angle, which is, for example, in the range of 4° to 6°. The y-axis is vertical, the z-axis horizontal.

[0036] The design of the wind turbine, which is to include bearing arrangement 2, must be based on natural environmental data, particularly concerning the wind at the turbine's location. The wind speed distribution typically follows a Weibull distribution, as described in the Fig. 2 and Fig. Figure 3 is shown. In the present case, the expected wind speed at the given location is not higher than 22.7 m / s for 97.72% of the total time. Fig. 2 and Fig. WS denotes the wind speed in m / s. Fr represents the relative frequency and cFr the cumulative frequency of the wind speed.

[0037] It is realistic that the wind turbine to be built will exceed the capacity of the existing wind turbines. Fig. Figure 4 shows the relationship between wind speed WS and relative power Pr. Detailed knowledge of the wind turbine's construction is not required at this stage of consideration.

[0038] The normalized rotor speed nr of the wind turbine is, as shown from Fig. As shown in Figure 5, the rotor speed remains constant over a wide range of wind speeds WS. At very high wind speeds WS, the rotor speed nr decreases, but even in this range, as long as the wind speed WS is not too high, the power output of the wind turbine should remain constant. At low wind speeds WS, in this case below 10 m / s, the aim is also to maintain the power output of the wind turbine at the highest possible level. Therefore, at particularly high and particularly low wind speeds, the wind turbine must be operated, in particular by appropriately selecting the pitch angle of the rotor blades, in such a way that an increased rotor torque is generated to compensate for the reduced rotor speed. This increased rotor torque is reflected in an increased relative load Br, which must be absorbed by the bearing arrangement 2, as shown in Figure 5. Fig. 6 emerges.

[0039] The in Fig. 5 and Fig. Figure 6 visualizes the load scenario for bearing arrangement 2 and relates to the normal operation of the wind turbine. Further load scenarios will be discussed in more detail later.

[0040] In Fig. Seven different load scenarios, designated BS1, BS2, and BS3, which occur during the operation of the wind turbine and have different effects on the bearing arrangement 2, are presented. The load scenarios BS1, BS2, and BS3 are each to be mapped to test scenarios PS1, PS2, and PS3, which are drastically shortened in duration. Fig. 7 is strongly distorted and indicated.

[0041] Various ways to modify the loads occurring in a test scenario compared to realistic loads are described in the Fig. Figures 8 to 11 illustrate this. In each of these figures, the left scale refers to the realistic load scenario BS1, BS2, BS3, and the right scale to the corresponding test scenario PS1, PS2, PS3. In the realistic load scenario BS1, BS2, BS3, the average load is B. av , the minimum load B min and maximum load B max The assigned loads in the test scenarios PS1, PS2, PS3 are denoted by P av , P min and P max designated. According to Fig. 8. When transferring the realistic load scenario BS1, BS2, BS3 to the test scenario PS1, PS2, PS3, the average load of B will be av on P av increased, while the range of loads, that is, the difference between B max and B min or the difference between P max and P min , remains constant.

[0042] According to Fig. In contrast, 9 remains the average load B av = P av constant, while the range of loads in test scenarios PS1, PS2, PS3 is significantly increased. In contrast, according to Fig. 10 the maximum load B max = P max kept constant while the range of loads decreases, resulting in an increase in the mean load of B av on P av This is accompanied by such an increase in the average load of B av on P av is also according to Fig. 11 is given, whereby in this case the range of loads also increases.

[0043] Each of the in the Fig. The schemes outlined in Figures 8 to 11 for modifying the load parameters allow, depending on the damage effect, a reduction in the test duration compared to the estimated service life of the bearing arrangement 2. Depending on the type of load and the design of the component of the bearing arrangement 2 being tested, a reduction in the test duration can affect the validity of the test results in different ways.

[0044] In the diagrams according to Fig. In sections 12 to 14, ZR denotes the period during which a specific load scenario BS1, BS2, BS3 occurs. The period ZR is always shorter than the estimated service life Tnutz of the bearing arrangement 2. If the bearing arrangement 2 were tested in each load scenario BS1, BS2, BS3 for a test duration corresponding to the respective period ZR, this would be optimal with regard to the quality of the test results. In this case, the closest possible correlation would be achieved between events observed in real-world operation and those observed on the test bench.

[0045] After the entire period ZR, a number of significant observations designated NSB is expected. These significant observations do not necessarily represent damage events; rather, they may also be changes detectable in bearing arrangement 2 that merely indicate potential damage events under increased loads and / or extended operating time.

[0046] The more the test duration is shortened compared to the actual expected load duration in various test scenarios PS1, PS2, PS3, the more the test results lose their significance. In other words, there is a correlation between the test results obtained with modified load parameters on the one hand and the results obtained with realistic loads and realistic load durations on the other. This correlation is less pronounced the greater the degree of reduction, i.e., the quotient between realistic load duration and assigned test duration. A reduction in test duration is always achieved by increasing the loads acting on the bearing arrangement 2.

[0047] The decrease in the reliability of the test results with shorter test duration is reflected in the various constellations according to the Fig. 12 to 14 in an area that widens to the left in the respective diagram. In the diagrams, nE denotes the number of identified events classified as relevant. According to Fig. 12. Even a moderate reduction in the test duration can lead to either significantly too few or significantly too many relevant events being triggered. Here, the lower edge of the hatched area marks the minimum number and the upper edge of this area the maximum number of events expected with a shortened test duration. In the upper branch of the diagram, according to Fig. Figure 12 shows a clear kink, indicating that with shorter test durations and increased loads, damage mechanisms occur that are irrelevant in real-world operation. Therefore, there are strict limits to reducing the test duration. In this case, only a reduction of the test duration to 80% of the duration of the associated realistic load scenario is acceptable. The shortened test period is denoted by VR.

[0048] Also according to Fig. In section 13, there is a clear, but less pronounced, correlation between shortening the test duration and the loss of informative value of the test results. In this case, a reduction of the test duration to 60% is still acceptable. The weakest correlation between a shortened test duration and a wider range of test results is found in the scenario described in section 13. Fig. 14. In this case, the test duration can be reduced to 20% without having to accept unacceptable changes in the test result.

[0049] Based on the varying relationships between increased test loads, shortened test durations, and the decrease in the reliability of the test results, different reduction factors are defined for the various stress scenarios BS1, BS2, and BS3. These factors indicate the ratio between the test duration and the corresponding realistic stress duration. In this way, a significant reduction in test duration can be achieved while minimizing the loss of reliability in the test results.

[0050] The following section explains various mechanical stresses that occur during the operation of a wind turbine, including the bearing arrangement 2. Fig. Figures 15 to 18 show radial loads, namely shear forces Fq and vertical forces Fv, which occur at wind speeds of 4 m / s ( Fig. 15), 10 m / s ( Fig. 16), 20 m / s ( Fig. 17) and 30 m / s ( Fig. 18) are to be absorbed by the bearing arrangement 2. In all cases, these are loads during normal operation. The maximum loads to be absorbed at the different wind speeds do not occur with constant values, but with a clear frequency dependence, as shown in Fig. 19 is recognizable. Here, the value 1 corresponds to the frequency generally designated as BF, the rotor speed, that is, the frequency at which the test object 5 rotates. Increased loads acting on the bearing arrangement 2 at this frequency FR = 1 can be explained, for example, by angular errors between individual system components. An increased load, which occurs at three times the frequency of the rotor speed, results from the three rotor blades swinging past the mast of the wind turbine. In the Fig. The load shown in Figure 19 is a tilting moment KM, which loads the test object 5. The magnitude of this tilting moment KM fluctuates in Fig. 20 (lower curve) shown and represented as a realistic tipping moment MK r This is referred to as... To implement this load in a test scenario, the load is applied in test rig 1 according to the upper curve in... Fig. 20, meaning a tilting moment MKP is generated during the test. This type of transformation from a realistic load scenario to a test scenario corresponds to the scheme according to the scheme. Fig. 10. Thus, there is an increase in the average workload while the range of the workload decreases. The frequency with which the maximum workload occurs is increased compared to the actual pattern.

[0051] The Fig. Sections 21 to 26 relate to the start-up operation of the bearing arrangement 2. This shows Fig. 21 the increase in the rotational speed n of the test object 5 during start-up. The Fig. 22, Fig. 23 and Fig. Figure 24 shows the course of axial load, that is, axial force F. ax , vertical load F v and tilting moment MK in the same period, namely in a total period of 600s. To simulate these loads with a shortened test duration, the test object 5 is subjected to an angular acceleration aw according to Fig. 25. This results in an increase in rotational speed n according to Fig. 26. In comparison to realistic starting operation, the angular acceleration aw is increased to three times the value in the test scenario.

[0052] The Fig. Figures 27 to 30 show the same in diagrams. Fig. 21 to 24 describes the reduction in speed until the test object 5 comes to a standstill, i.e., the shutdown of the system. The regular shutdown is not simulated separately, but exclusively by the "emergency stop" mode, to which the Fig. References are 31 to 36. It is clearly evident that the axial load F ax During the rapid deceleration of test object 5, the value briefly entered the negative range, while otherwise a positive sign was given. Reference symbol list 1 test bench 2 Storage arrangement 3 rolling bearings, angular contact roller bearings 4 rolling bearings, angular contact roller bearings 5 Test object 6 test bench bearings 7 Actuator arrangement 8 Control and evaluation device 9 Sensor arrangement aw angular acceleration b Contact width in Hertz B av , P av average load BF Load frequency B min , P min minimal load Bmax Pmax maximum load Br relative load BS1, BS2, BS3 load scenarios cFr cumulative relative frequency F ax Axial force F q Shear force F rad Radiative power Fr relative frequency F v Vertical force MK tilting moment MK r realistic tipping moment MKP tilting moment during testing nE Number of relevant events n rotational speed nr normalized speed NSB Number of significant observations Per relative power PS1, PS2, PS3 test scenarios Tnutz usage duration VR shortened period WS wind speed ZR period

Claims

[1] Method for testing a storage arrangement (2) comprising the following features: - Environmental data relating to external influences affecting a plant including a storage arrangement (2) are provided, - Based on the environmental data, different environmental scenarios are classified, which differ from each other in terms of their influence on the system, - Using the environmental scenarios that are relevant with regard to the service life of the bearing arrangement (2), data relating to various realistic load scenarios (BS1, BS2, BS3) of the bearing arrangement (2) during the estimated service life of the bearing arrangement (2) are determined, - the proportion of the different load scenarios (BS1, BS2, BS3) to the estimated service life of the bearing arrangement (2) is determined, - Regarding each load scenario (BS1, BS2, BS3), load parameters influencing lifespan (nr, Br, B) are determined. ave , B min , B max ) determined, - at least for a subset of the load scenarios (BS1, BS2, BS3) schemes for modifying the load parameters (B ave , B min , B max ) created, whereby the modified load parameters (P ave , P min , P max ) a mapping of the realistic load scenario (BS1, BS2, BS3) onto a shortened test scenario (PS1, PS2, PS3) results, - Data on the decrease in the dependence of the modified load parameters (P ave , P min , P max The test results obtained from realistic loads and realistic load durations, showing the degree of shortening, are determined for various load scenarios (BS1, BS2, BS3). - For different load scenarios (BS1, BS2, BS3), specific shortening factors are determined, which indicate the ratio between the duration of the test scenario (PS1, PS2, PS3) and the duration of the load scenario (BS1, BS2, BS3). - the bearing arrangement (2) is subjected to a test with the determined test scenarios (PS1, PS2, PS3), which together determine the loads (B) that occur during the estimated service life (Tnutz) and affect the service life of the bearing arrangement (2). ave , B min , B max ) depict, subjected to, whereby the percentage temporal composition of the different test scenarios (PS1, PS2, PS3) differs from the temporal composition of the associated realistic load scenarios (BS1, BS2, BS3). [2] Method according to claim 1, characterized by , that load parameters (P) assigned to a test scenario (PS1, PS2, PS3) ave , P min , P max) Mean values ​​and distributions, especially minimum and maximum values ​​(P min , P max ), of loads, including various load parameters (P ave , P min , P max ) are modified in different ways in relation to the associated loads expected in realistic operation. [3] Method according to claim 1 or 2, characterized by , that the load parameters (B ave , B min , B max ) Forces (F q , F rad , F v ), rotational speeds (n) and rates of change of at least one of these parameters (F q , F rad , F v , n), include. [4] Method according to any one of claims 1 to 3, characterized by , that the load scenarios (BS1, BS2, BS3) include the following scenarios: - Normal operation of the storage arrangement (2), - Start-up operation, - Speed ​​reduction to standstill, - Idle oscillation. [5] Method according to claim 4, characterized by , that the load scenario “speed reduction to standstill” is covered by a scenario “emergency stop”, wherein in the corresponding test scenario (PS1, PS2, PS3) the angular acceleration (aw) of the component of the bearing arrangement (2) to be decelerated is increased by a factor greater than two and less than five, in particular by a factor of three, compared to the realistic load scenario (BS1, BS2, BS3), which is given by the regular speed reduction to standstill, and wherein during deceleration in the test scenario “emergency stop” the sign of an axial force (F) acting on the bearing arrangement (2) ax ) changes. [6] Method according to claim 4 or 5, characterized by , that in the load scenario “normal operation of the bearing arrangement” compared to the realistic load scenario (BS1, BS2, BS3) the mean value (B ave) of a load, in particular a tilting moment (MK), is increased and at the same time the bandwidth of the load acting on the bearing arrangement (2) is reduced. [7] Method according to claim 4 or 5, characterized by , that in the load scenario “normal operation of the bearing arrangement” compared to the realistic load scenario (BS1, BS2, BS3) the mean value (B ave ) a load, in particular a tilting moment (MK), is kept constant and at the same time at least one of the quantities amplitude and frequency of the load is increased. [8] Method according to any one of claims 4 to 7, characterized by, that in the load scenario “start-up operation” the angular acceleration (aw) is increased by a factor greater than two and less than five, in particular by a factor of three, compared to the realistic load scenario (BS1, BS2, BS3), and the number of start-up operations is reduced by a factor greater than 80 and less than 1000, in particular at least 100 and at most 300. [9] Method according to any one of claims 4 to 8, characterized by, that in the load scenario “idle oscillation” the swivel angle between the components of the bearing arrangement (2) that are rotatable relative to each other is limited to a value that is at least 1 degree and at most 15 degrees, in particular not higher than 12 degrees, wherein in the test scenario (PS1, PS2, PS3) the mechanical loads are not increased compared to the realistic load scenario, but the same angular ranges of the components of the bearing arrangement (2) are always loaded in each oscillation process. [10] Method according to any one of claims 4 to 9, characterized by, that the various load scenarios (BS1, BS2, BS3) depicted on test scenarios (PS1, PS2, PS3) are combined into several test modules, wherein a first test module comprises the normal operation of the bearing arrangement (2) including a fluctuating tilting moment (MK), the start-up operation, and the speed reduction to standstill, a second test module loads the bearing arrangement (2) with higher, constant loads compared to the first test module, namely axial and radial loads (F ax , F rad ) and a higher tilting moment (MK), a third test module that reproduces idling oscillation, and wherein the third test module is divided into sections which are intermittently inserted between sections of the second test module. [11] Test rig (1) for testing a bearing arrangement (2) for a wind turbine, comprising a control and evaluation device (8) designed to carry out the method according to claim 1.

Images