While working as an on-site thermagrapher at a facility, our PDM dept was having problems with a multi-stage blower (Hoffman). The problem we were having was that no matter how the millwrights completed the alignment, as the blower ran, the vibration data indicated an increasing misalignment between the motor and the blower. Our PDM dept had grown to the point to where we were taking on problem solving roles using all the tools at our disposal; infrared, vibration analysis, MCE, MCA, oil analysis, RCA, RCM, and NDT. We had developed a high level of cooperation between our group and the trades and had lots of feedback and positive recommendations to correct reoccurring problems.
The millwrights and our group had determined that some kind of offset alignment was needed to correct the alignment on these blowers (there were six of these on-site) but they were unwilling to just guess as to how much offset was required. Can’t blame them, the “poke and hope” method is not a part of the Precision Maintenance method nor does it look very good to the customer or to upper management!
If you are not familiar with the blower type that I am talking about here is an image of one:
Each “coil” in the blower is another compression stage from one end to the other. The temperature increases from end to end. The issue that we were having with the alignment is that any expansion below the center line of the shaft effects the alignment. This means all materials, the base, the motor base, motor frame, blower frame, motor housing, blower housing, bearing mounts, and housings. All of these materials expand at different rates according to their operating temperatures.
I took the task on to build a thermal map. Using a very low emissivity targets (aluminum tape), a grid was laid out from the shaft center line to the floor on all sides. I used one-half inch wide strips of aluminum tape to building out six inch squares. The mechanical engineer had requested the six-inch grid (something about the math being easier). We set up for a cold background for the reflection to make the grid stand out in the image. Lucky for us, it was winter in the Midwest and we simply removed the insulation on the wall and shut down the heaters, giving us a nice uniform cold background.
An image of the blower assembly at ambient temp was saved as a control. The blower was started up and images were recorded at two-minute intervals until a steady stat was reached from all sides. We used a Mikron 7515 with a remote view/control panel to minimize the effect of unwanted reflections in the grid of aluminum tape we had created. Using the recorded time stamped images, the engineering group and the PDM group were able to determine the “growth” of all the materials below the shaft center line of the blower assembly to determine the correct amount of offset for a “cold alignment” to achieve correct alignment while the unit is operating at “normal” temperature.
Yes, this did take some time to complete, starting the blower, taking the images, shutting down, moving the tripod, letting the blower cool to ambient, recording the control from that view, and starting the process again (four times). But it prevented the “loss” of blower bearings and rotors due to misalignment (we were losing on average one every two months) and at 3600 RPM a loss was catastrophic! The point of this is that if you put a little effort in on the front end, you can successfully achieve the results you want. This thermal growth study is proof of that and why you should make a plan and execute. You’ll find that you have accurate results and more success than with the “poke and hope” method!