The requirements for wind turbine test benches are continuously increasing and demand, in turn, investments in test bench expansions or new test benches. Fraunhofer IWES has developed a hybrid test procedure to expand the capacity of existing testing facilities and render them fit for future testing requirements with the aid of digital methods
The validation of new wind turbine designs and technologies is an essential foundation to assure the reliable operation of wind turbines. To this end, multiple test benches for wind turbine drivetrains have been built and put into operation around the world in recent years. Each of these test benches, which are usually complex and associated with high investment costs, was designed based on the demands of turbine sizes foreseeable at the time. As a result of the continuing trend toward larger turbines and higher rated powers, there is also a need for larger test facilities to allow validation of new developments under realistic operating conditions. This is a challenge, because both the development and construction of larger test benches are associated with high costs. In addition, within a few years, existing testing facilities are often no longer powerful enough for the latest turbine ratings, with the result that it is no longer possible to apply the required extreme loads, for example. One approach to solving this problem is the increased use of simulation methods, with the help of which virtual tests can be performed. The important aspect here is for the simulation to build upon the results of physical tests, since even state-of-the-art simulation tools can only deliver results with insufficient accuracy if the model parameterization is uncertain.
Combining test bench and simulation
The results of the research project “VirtGondel – Development and validation of a virtual representation of the nacelle test bench for the deployment of advanced test methods and more efficient test campaigns” provide an answer to this trend. Integrating state-of-the-art simulation tools helps us to develop suitable test scenarios for wind turbine drivetrains that are above the nominal power or load capacity of existing test facilities. The concept is simple: the majority of the tests below the maximum load can still be carried out on the test bench. For the lacking share exceeding the test bench capacity, the scientists then employ state-of-the-art simulation methods to determine the required measurements. In this way, physical tests under partial load and simulation-based virtual tests are combined to create a “hybrid test”.
In the scope of the VirtGondel research project, we have further developed this method with the name “hybrid testing” and applied for a patent as an innovative test procedure. Our research focuses on the mechanical testing of wind turbine drivetrains on a nacelle test bench such as the Dynamic Nacelle Testing Laboratory (DyNaLab), but the concept is transferable to a wide range of other test scenarios both within and beyond the wind energy sector. First, the test specimen undergoes a nacelle testing campaign in the DyNaLab. However, due to the limitations of the test bench, it is not possible to approach all maximum loads of the test specimen during the physical test campaign, but only up to a level of, e.g., 70% of the target load. The hybrid test procedure picks up right there and combines the physical partial load tests with virtual tests. For this purpose, we create one (or more) detailed simulation model(s) of the test specimen, which can reproduce the relevant measurement variables (e.g., strains, deformations, vibrations, and accelerations). It is subsequently crucial to optimize and validate the simulation model(s) on the basis of the extensive measurement data from the physical testing campaign – a process that goes beyond conventional virtual testing. Only with this step is sufficient accuracy of the simulation model achieved, thus allowing it to be subsequently utilized to determine the remaining measurements from test scenarios above the test bench capacity by means of virtual testing.
Figure 1: Distribution of test results between virtual (simulation) and physical tests in a conventional full-load test (top) and the new hybrid test procedure (bottom) © Fraunhofer IWES
The advantages of the new hybrid test procedure
The procedure makes it possible to test drivetrains with increasing rated powers on test benches that are nominally no longer an ideal fit. In this way, high costs for new testing facilities can be saved. In addition to the cost-efficient further use of the existing test infrastructure, the often many years of experience in the operation of a test facility represent another advantage. While new test benches with higher nominal power ratings can still suffer from teething problems, especially in the first few years, the industry can benefit from the comprehensive expertise of existing test benches. The procedure also provides a validated simulation model of the test specimen. Compared with purely simulation-based virtual testing, the method offers the great advantage of being based on real physical tests. Simulation models are often not precise enough without the feedback of real measurements – here, the optimization and validation of the simulation model based on the physical tests under partial load are the decisive steps towards valid simulation results.
Figure 2: Comparison of the process diagrams of a conventional test campaign under full load with the new hybrid test method. © Fraunhofer IWES
The next steps for the hybrid test procedure
For the scientists at IWES, the procedure represents an important step toward the digitalization of wind turbine testing. It allows us to offer cost-effective testing of the latest turbine designs without having to rely on ever new and larger system test benches, which are associated with high costs and risks. In addition, we make validated simulation models of the test specimens available to customers. A first demonstration of the method focusing on parasitic loads on the drivetrain was presented at this year’s Conference for Wind Power Drives (CWD 2023) in Aachen, Germany . Our current research focus is on investigating further applications of the method as well as quantifying possible limitations, for example with regard to nonlinearities or measurement variables that are difficult to model. In addition, we are investigating how far the nominal test bench capacity can be exceeded in the hybrid test in order to ensure satisfactory accuracy of the simulation results. IWES is striving to further develop the method within the framework of future public research projects.
 M. O. Siddiqui, A. R. Nejad und J. Wenske, „On a new methodology for testing full load responses of wind turbine drivetrains on a test bench,“ Forschung im Ingenieurwesen, 2023.