By Donald B. Shaw

As the traditional diagnostic methods failed to provide any insight into the cause of the problem I found myself focusing more and more on the comments of the operator relative to the wind. What could be the difference between the two turboblowers such that one could somehow be affected by the wind while the other was unaffected? The machines were identical. The support structures were identical. There was one difference. The intake stack for the problem turboblower was located entirely outside the building while the intake stake for the nonproblem turboblower was inside except for the last few feet. That would expose the intake stack on the problem turboblower to the wind, but both intake stacks were on separate foundations from the turboblowers. While having the intake stack exposed to the outside environment provided at least a link between the wind and the vibrations, it did not seem to account for the fact the alignment of the coupling changed with time. Also, the support structures for the stack were designed to prevent loads of the intake stake from being transmitted to the blower.

The operator's comment relative to the wind became my major focus. It was learned that the turboblower support structure was founded on piles while the intake stake was supported on a spread footing. A hypothesis was developed that over the years vibrations of the intake stack that was outside the building had dynamically consolidated the soils underlying the stack foundation allowing the stack to settle. This would result in portions of the stack load being transferred to the blower. Based on that hypothesis an expansion joint was placed in the inlet duct between the stack and the blower. When the alignment of the coupling was later checked prior to a test run on the turboblower, it was found that the alignment had changed dramatically after the expansion joint was inserted.

The expansion joint solved the problem. In retrospect, it was learned that dynamic consolidation settlement resulting from the wind induced vibrations of the intake caused stack loads to be transmitted to the turboblower causing a dynamic misalignment of the coupling that was indeed worse when the winds were blowing harder. The seemingly ridiculous observation of the operator relative to a correlation between the vibration problem and the wind turned out to be the key to the solution.

More than a lesson in engineering, I learned to always pay close attention to all observations associated with a problem and to never dismiss any observations no matter how ridiculous they may seem. The key to a problem solution is often contained in a seemingly offhand observation of those closest to the problem.