The following situation occurred at a Municipal Power Plant. During a planned outage, the on-site team was utilizing a Rotalign® Pro with Boralign® to assess the position of the bearing bores after the removal of the rotor on an Allis Chalmers power generation steam turbine. The results were varying far more than normal and the customer wanted to understand why. I was called in to make sure there were no errors in the use of the Boralign system. The tool was being used properly and there were no problems with the tooling. The customer had already set the unit up on a surface plate and established that there was no appreciable drift with the tool. I suspected excessive vibration, so I used my Vibscanner® to measure overall vibration in displacement. No significant level of vibration could be detected; certainly not enough energy to cause the level of non-repeatability they had been observing in the readings. I thought if the movement of the lower shells had a very low frequency that was too slow to measure utilizing traditional vibration sensors and methods, I might be able to measure it with the Rotalign Pro Straightness program. Both the laser and receiver were mounted on standard magnetic dial stands with brackets that are part of the Straightness hardware set. A series of points were established on the lower shells to measure between, both along the rotational axis and from side to side. Setting the Rotalign Pro Straightness program to the shortest sample time and observing the X-axis values over a 2 to 4-minute period, I was actually able to see a very slow movement between various points on the lower shells. While some areas of the lower shells were reasonably stable, there was movement between some points that was as much as .024”, but at a very low CPM. While this process was somewhat like ODS, there was no real-time phase data, so I could not tell how the various points were moving relative to each other. It was still very helpful to the rotating engineer to understand how much the various areas were moving. Better yet, with a fairly simple program and only 6 hours invested in collecting the data, the rotating engineer had a good idea of the magnitude of the problem.

Filed under:
, , by Keith Van Hentenryck