PSI - Issue 65

D.G. Solomonov et al. / Procedia Structural Integrity 65 (2024) 275–281 D.G. Solomonov and M.Sh. Nikhamkin / Structural Integrity Procedia 00 (2024) 000–000

277

3

The structure under study is exposed to vibration under operation. It causes cyclic bending of the shell and may result in flange detachment during operation as texted by Solomonov et al. (2023) and Solomonov et al. (2022). Figure 2 shows a scheme of sample fatigue testing, which implements such loading. The study was carried out on an experimental installation based on an electrodynamic vibration stand. The sample is fixed to the flange on the table of the vibration stand. It is loaded by the inertia force of the load attached to it with flexural-mode resonant harmonic vibrations (Fig. 2). This testing scheme ensures a symmetrical cycle of sample loading (cycle asymmetry coefficient R = 1). The control system of the experimental installation has an automatic power control. It maintains the resonant oscillation frequency and the predetermined load on the sample. The deformation of the sample is recorded continuously by a strain gauge located on the surface of the sample in the most loaded zone A (see Figs. 1 and 2). The frequency of the resonant vibrations of the sample is about 80 Hz. The self-heating of the sample at this frequency does not exceed several degrees, and it does not affect the mechanical properties of the carbon fiber. The current number of loading cycles and the resonant oscillation frequency of the sample are also recorded during the testing process. A more detailed description of the experimental setup was given by Solomonov et al. (2023). The temperature field on the surface of the sample is recorded in a non-contact mode during the test by means of a precision infrared camera. A TH9100WR IR camera with a spectral range of 8–14 μ, a sensitivity of 0.03 °C, and a temperature measurement range of 40...+500 °C was used. To reduce thermography errors, a constant room temperature is maintained. Efforts are being made to exclude the influence of extraneous sources of infrared radiation. The surface of the sample is covered with black matte paint with an emission factor of about 1. Each sample was checked by ultrasound to ensure that there were no initial defects before testing. In accordance with the idea of the IRT method, sample loading is a sequence of blocks of cyclic deformation with a constant load within a block. The load level was set by strain amplitude ε a . It was measured by a strain gauge located in the most loaded zone A (Fig. 1). Block loading of the sample included 17 blocks. The strain amplitudes increased from block to block in the range (0.05–0.41) ε st (ε st is the limiting strain under static loading of the shell material). The duration of each loading block is 40 to 50 thousand cycles. The temperature field was recorded every 5000–10000 cycles. A break was taken to cool the sample to the initial temperature after each loading block. The loading frequency was the same in all blocks, 80 Hz. Figure 3 shows, as an example, thermograms recorded at the end of two loading blocks. The sample heating area is also localized in the area of connection between the flange and the shell (zone A, Fig. 3). The temperature in this zone increases with the strain amplitude in the block. The test results were processed as follows. The maximum and average temperature values T max and T mid over heating zone A (Fig. 3) were recorded depending on the cyclic operating time N in each loading block. Next, the temperature increases θ max = T max T 0 and θ mid = T mid T 0 were determined (T0 is the temperature of the sample at the beginning of the block). An example of the dependences θ max (N) and θ mid (N) is shown in Fig. 4. Due to self heating, the temperatures T max and T mid gradually increase and reach some stabilization values T (stab) ^max and T stab ^mid . The heat release in the sample is balanced by heat removal to the environment at these temperatures. Each loading block is characterized by strain amplitude ε a and the stabilization values of the maximum and average temperature increments (Fig. 4): θ ���� � �� =� � ��� � � � � � and θ ���� ��� =� � ��� � �� � � . (1) In addition, the sample heating characteristic in each block is the heating rate at the beginning of the block, θ ��� = � � � θ ��� (�) , where N = 0, and θ ^��� = � � � θ ��� (�) , where N = 0. (2) 3. Results and discussion