Background

Analyzing eigenfrequencies of a structure is commonly on the to-do list when designing a component subjected to known excitations / vibrations, especially in rotating machines. However, if the structure is modifiable it can be very time consuming to study all variations.

If the eigenfrequencies are left unstudied, unexpected vibration can occur when the structure is subjected to excitation close to the natural frequency. In addition, sometimes operating the machine at the range of existing natural frequencies is intentional and thus occasional resonances are inevitable. Most of all vibration is disturbing and may also lead to fatigue failure.

Operators at the Moventas Gears Oy Ikola test rig had observed vibrations of the test setup while running tests for a certain type of windmill gearboxes. The most severe vibration occurred at the test rig itself. Moventas Gears Oy decided to study the test rig fatigue due to vibration to avoid any damage to the test rig that would lead to interrupts of the tests. In addition to the study of the current test setup, also the upcoming test setup configuration was studied.

Analysis methods

Results for the natural frequency mode shape from the eigenfrequency analysis were used as the loading for the fatigue study. Animations of natural frequency mode shapes 1 and 2 for test setup configuration 1 are shown in Videos 1 and 2. Since the natural frequency deformation shape matched the measured vibration, the amplitude of the displacements was scaled to match the measured values in order to obtain the stress ranges. The most critical location based on the stress levels was then recognized and studied more precisely, as shown in figures 1, 2 and 3. The estimated lifetime was obtained from the stress ranges for the current test setup.

For the upcoming test setup, the allowed displacement amplitude was solved. It was assumed that vibration will occur at the first or second natural frequency, as in the first test setup configuration.

Results

The results indicated that failure due to fatigue is not likely to happen when performing tests with the first test setup configuration. The allowed displacement amplitude for the second test setup configuration is something that the operators can keep an eye on when the actual tests start.

Figure 1. Finding the most severe location.

Figure 2. Studying the most critical toe line of the welds.

Figure 3. SN-curve for the effective notch stress method analysis.