Software for Operational Modal Analysis
ARTeMIS

 

 

New Upcoming Feature - Reduce Peaks tool for ARTeMIS Modal Pro

 

In the upcoming ARTeMIS Modal Pro 5.2 a new tool for the reduction of undesired peaks from the structural response spectra is introduced. This newsletter briefly explains the new feature and applies it with using two examples.

 

The Reduce Peaks tool can be used to reduce the influence of e.g harmonic peaks originating from a rotating parts acting on the structure during a test. Likewise, it can be deployed to remove undesired modes. This particular application is beneficial in case of very strong modes being mixed with weak modes. In order to better estimate the weak modes the strong ones can simply be removed from the response data before conducting the modal analysis.

 

The idea behind the tool is to predict the response only of the peaks defined by the user. Once estimated, the original measurements are orthogonally projected onto the predicted response of the selected peaks. The projected measurements become blind to the peaks, and as such are reduced significantly.

 

The reduction require several steps:

 

1. Estimate a range of state space models using Stochastic Subspace Identification.

2. Select the optimal state space model containing the eigenvalues corresponding to the selected peaks.

3. Predict the modal response of user-selected peaks from the modal decomposed state space model.

4. Orthogonally project the original measurements onto the predicted response of selected peaks.

 

It might, in some cases, be necessary to run the reduction several times, by cleaning the measurements recursively.

 

 

Example 1: Plate with Harmonics

The first example is a simple, rectangular, aluminium plate excited both with a broad-band random and a deterministic harmonic excitation from a shaker. The response data is recorded by a 16 channel acquisition system and sampled with 4096 Hz. Harmonic peaks are present at 374, 748 and 1496 Hz. More detailed, full-length analysis of this example can be seen in this YouTube video.

 

As a base case, the three harmonic peaks are removed as in the figure illustrated below. In the top diagram the original measurements are represented by the Singular Values of the spectral densities. The diagram at the bottom shows the same data after the reduction.

 

Removal of the three harmonic peaks. The top diagram presents the Singular Values of the spectral densities of the original measurements. The bottom diagram shows the modified data after reduction.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


 

 

 

 

 

 

 

 

 

 

In the next example, the third mode is selected and removed. This demonstrates how an arbitrary mode can be selected and reduced/removed if required.

 

Removal of the third mode. The top diagram presents the Singular Values of the spectral densities of the original measurements. The bottom diagram shows the modified data after reduction.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Example 2: Full-scale ferry with resonance problems

 

Second example presents the industrial use of the tool on an operating data from a ferry subjected to resonance. The ferry is excited by a random environmental load (wind and waves) with harmonic interference from a rotating machinery on-board. The dynamic test is undergone under fully operational conditions; the responses are recorded using a 16 channel acquisition system and sampled with 128 Hz. Here the data is analyzed up to 12.8 Hz as there are no modes of interest above this frequency.

 

The first user-selected peak is, in fact, a harmonic peak very close to the resonance of one of the fundamental modes of the structure. Additional orders of this harmonic frequency are also selected for the reduction. A higher order mode at 8.2 Hz is masked by the harmonics of the same frequency. The aim of this analysis is to remove the harmonics that render the operational modal analysis of this data less effective. 

 

The results of the harmonic reduction are illustrated on the figure below. More detailed, full-length analysis of this example can be seen in this YouTube video.

 

 

Removal of the six harmonic peaks. Five of them have been manually selected for reduction. The last one is reduced automatically. The top diagram presents the Singular Values of the spectral densities of the original measurements. The bottom diagram shows the modified data after reduction.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Conclusion

The new Reduce Peaks feature provides an effective mean to reduce/remove unwanted peaks or modes from the structural responses. Developed methodology is robust towards the real-life problems such as: non-stationary harmonic excitation like e.g. a burst sine, or systems with varying/changing rpm, where for both it is capable of reducing/removing selected peaks. The tool will be available in ARTeMIS Modal Pro 5.2 to be released in October 2017.

 

For more information please contact us at info@svibs.com .

 

 

 


Structural Vibration Solutions A/S • Tel: +45 9635 4422 • sales@svibs.com

To unsubscribe: Write Unsubscribe to mo@svibs.com