In October 2022, the Ground Vibration Test (GVT) of the fs35 Harpyie motor glider took place at our structural dynamics laboratory in Göttingen. For two weeks the aircraft of the student research group Akaflieg Stuttgart has been tested extensively.
The fs35 has a wingspan of just below 18 m, an empty weight of about 670 kg and a 155 hp diesel engine. The motor glider was developed and built by the students of the University of Stuttgart to be able to tow gliders quickly and more efficiently to high altitudes. It is already in operation since 2019, but with a maximum permissible speed of 180 km/h, its usability is currently limited. In order to obtain certification for the fs35 with the planned maximum speed of 280 km/h, flutter calculations based on a validated FE model are necessary. For the modification of the FE model, the GVT provides the necessary experimental data.
The aircraft was lifted with bungee cords to guarantee a free-free boundary condition. This is important for two reasons. First, the bungee support is representing a defined stiffness which can be implemented into a simulation model. On the other hand, this type of suspension avoids as far as possible that the eigenfrequencies of the aircraft are influenced by the boundary conditions of the GVT on the ground. Thus, this type of soft suspension provides the possibility for direct use of the experimental modal data from the GVT in the flutter calculation.
The GVT is performed to experimentally identify the modal parameters such as the eigenfrequencies, mode shapes and damping ratios of the real structure. These parameters are necessary to adapt the FE model as a basis for a valid flutter calculation to expand the flight envelope. At 15 different positions electrodynamic shakers have been installed to excite the aircraft with a single shaker or simultaneously with two shakers. The excitation signals have been shaped by a DLR internal signal generator to generate uncorrelated random output or correlated sine sweeps. The excitation was performed on different force levels to check the structure for non-linear modes. 120 acceleration sensors placed all over the structure recorded the responses. Modal identification has been performed simultaneously to the GVT, such that the identified modal model of the glider could be used right after the GVT. An identified mode shape is shown as an example with the 3n wing bending in Figure 1.
A total of 869 modes were identified in two different configurations (control surfaces in a free and a fixed condition). Within the correlation of the results, they were assigned to 52 mode families. These include not only elastic structure modes with frequencies from 3.3 to 48 Hz, but also rigid body modes, which are in part significantly below 1 Hz. Using different excitation levels and our self-developed Correlation Tool, the experimental modal data could be analyzed for nonlinear trends. Figure 2 shows the eigenfrequency and the damping ratio of a given mode shape as a function of the amplitude of the excitation force.
The results from the GVT are now used to update the simulation model to match the characteristics of the real structure. The validated model allows a realistic calculation of the vibration behavior of the aircraft and therefore a very good estimation of whether flutter can occur in the full operating range up to 280 km/h.
Author: Carsten Thiem, DLR-Institute of Aeroelasticity, Department: Structural Dynamics and System Identification