Mechmine had recently the opportunity to get its hand on vibration data and the corresponding bearings. Read here why.
| Customer situation: A pump operator (regional utility) was recommended by a pump expert to overhaul five water pumps. There was no audible improvement after the overhaul of the first pump P5. An independent assessment was commissioned before the recommended overhauls of the four remaining pumps. The cost of several tens of thousands of Euros for this work was not budgeted. |
 |
| Pump vibration measurements: Measurements were taken at three measuring points: electric motor (towards the pump), at the first bearing (motor side) and at the second pump bearing (near the impeller). The comparative measurements showed that there was no urgent need for action. Instead of pump P1, which got classified as critical, Mechmine felt that pump P2 should be the next one to be repaired because its vibration data showed an increased noise level. This noise was not observed with the standardized ISO RMS (root mean square) measurement of the vibration velocity. Its visible clearly in the frequency range far above 1000 Hz. The graph below shows the sequential measurements of all 5 pumps over two measurement campaigns, before and after the repair of pump P2. The first measurement campaign comprised 8 individual measurements and the second campaign 5 individual measurements per pump. The respective 1st individual measurement is marked by an “M”. One can see how the measured signal level of pump P2 fell back to the level of the other four pumps, including the already repaired pump P5. However, the positive effect of the bearing replacement can be seen more easily in the noise level in a frequency spectrum. |
 |
| The 4-fold averaged envelope spectrum (graph below) shows the noise level and highlights rollover frequencies also. The decrease in the noise level is clearly visible, compare green and blue curve. The blue curve equals now the noise level of all the other pumps, including the first repaired pump. The first blue peak at 50 Hz comes from the third harmonic of the rotational frequency and the peak at 100 Hz stems from the impeller passing frequency. Note that the green peak at 48.5 Hz has completely disappeared after the bearing replacement. Thanks to the high frequency resolution, such peaks, spurs, can be distinguished from each other. |
 |
| Fortunately, a visual damage analysis of the three old pump bearings could be carried out after the repair of P2 to complement the vibration data analysis. The three photos below show different damage patterns: perpendicular scratches across a rough running surface of the bearing outer ring (left), material chipping on a rolling element (center), and foreign particles and heavy wear marks on another bearing roller (right). One possible and likely cause could be the lack of lubricant, which leads to increased wear due to metal-to-metal contact. Which got confirmed by the operator. |
 |
| Customer benefit: The pump operator decided not to repair any more pumps after the pump overhaul of P2 and the vibration data comparison. This enabled him to save an unbudgeted expense of several tens of thousands of Euros in the current budget year without increasing the perceived risk of another pump failing. Furthermore, two out of five pumps have been overhauled, which also benefits redundancy. In the long term, high-resolution vibration measurements, either annually or every two years, will provide information on the condition of all five pumps. Thanks to the stored vibration raw-data in the Mechmine Cloud, a.k.a. mmViewer, trend curves can be generated quickly and effectively in the future, and pumps can be cost-effectively examined for machine faults and damage using the integrated diagnostic functions in the mmViewer, e.g. from the office. |