There is an interesting story behind a long series of randomly occurring thrust-bearing failures in one particular type of slurry pump in service at a South American bauxite mine. Apparently the thrust bearings would sometimes fail after a few days or, at other times, after a few weeks of operation. Before the mechanics produced a crossectional view similar to the simplified version depicted in Figure 9-4, the visiting troubleshooter had been told that it was often necessary to rebush and line-bore the bearing housing. The relevance, accuracy, or importance of this verbal description becomes evident only when the drawing is examined in detail.
Figure 9-4. Cross-sectional view of slurry pump with failure-prone thrust bearing configuration.
With the impeller inverted so as to reduce the differential pressure across the shaft packing area, it is immediately shown that the primary thrust is from right to left. The two angular contact bearings on the extreme left are correctly oriented to take up the predominant load. However, their outer rings are completely unsupported because the fabricator had somehow decided to overbore the housing in the vicinity of these two bearings. Consequently, the entire radial load acting on the coupling end of the pump had to be absorbed by the remaining third angular contact thrust bearing. This bearing was thus overloaded to the point of rapid failure and was, of course, prone to rotate in the housing. Using a double row spherical roller bearing at the hydraulic end of the pump would normally make for a sturdy, well-designed pump. In this case, however, the spherical rotation or compliance feature tended to further increase the radial load transferred to the one remaining outboard bearing. The basic agent of the component failure mechanism was, of course, force.
An equally serious burden was imposed on this pump by the well-intentioned person who, in an effort to link the spare parts requirements of the North and South American plants of this major aluminum producer, added to the drawing the parts list partially reproduced in the lower left-hand corner of Figure 9-4. Having left off the appropriate alphanumeric coding behind the bearing identification number 7312, this plant and its sister facilities would receive thrust bearings in other than matched sets. A quick look at the bearing manufacturer's dimension tables (see insert, Figure 9-4) shows simple type 7312 bearings to have a width that may differ from the next bearing by as much as 0.010 inches. Mounting two such bearings in tandem may cause one to carry zero to 100% of the load, while the other one would simultaneously carry 100 to zero % of the load. On the other hand, matched sets intended for tandem mounting would be precision-ground for equal load sharing (50/50%) and would be furnished with code letter suffixes to indicate this design intent.
Did we go through our seven cause categories to identify the above root causes? Frankly, no. When both the fabrication sketch and the procurement documentation— “information processing”—show two very obvious errors, it is reasonable that rectification of these deviations should be a prerequisite to further fine-tuning. Which is just another way of saying that if it looks like a duck, walks like a duck, and quacks like a duck, we ought to call it a duck and dispense with further research into the ancestry of the bird.