Improving the Reliability of a Critical Pump by Reducing Resonance Using Four Methodologies

In the largest Australian-owned Brewery, loss of production can occur due to stoppages of critical pumps. The reliability team of this plant ensures the availability of assets using vibration condition monitoring. One of the critical assets, an energy store delivery pump, suffered bearing failures between two and three times a year because of high vibration due to resonance forcing frequency. Previous attempts to resolve the problem by different contractors by adding support structures were unsuccessful leading to further review and investigation. This paper describes how we used a combination of four techniques including vibration analysis, finite element analysis, motion amplification camera, and mathematical approach to address the resonance issue. Vibration data was collected on the bearings of the motor and pump, the spectrum showed resonance at forcing frequencies of 18 Hz and 50 Hz as the root cause of the high vibration. The pipe’s supports were modelled in a CAD package, and a Finite Element Analysis program calculated the natural frequencies and their mode shapes. The natural frequencies were also calculated, using a five degree of freedom model of cantilever beams, to verify the results. A Motion Amplification Camera also measured the resonant modes of the base structure of the pump, the pipe, and the support frames. The values of the natural frequencies from the FEA, calculations, and MAC validate the results. We found that the natural frequency of the structure was close to the running speed of the pump. Data from these techniques was used to design new supports. The stiffness of the structure was increased by replacing the square section supports (80x80 mm) with I-beams (150 x 150 mm) and adding angle sections diagonally for triangulation. The overall vibration across the asset was reduced by 98% to an acceptable level. The pump bearings have not failed since the new frame was installed, increasing the reliability of the asset for production.