I was a young bearing design and application engineer for a large well-known bearing manufacturer, when our customer, the manufacturer of a new line of large integral horsepower motors, asked me to visit his customer, the end user, in order to determine why they were getting premature bearing failures. Upon arrival, the plant manager warned me that I would be seeing the maintenance foreman and his crew who were in a very bad mood as they could not remove the latest failed bearing from the armature shaft. Arriving at the scene, the foreman told me of the previous day's woes and asked me in more of a tongue-in-cheek manner if I could show them an easier way to get the bearing off. They had worked 8 hours, heating and beating with a bearing puller, all to no avail. From the smirks on the workers faces, it was obvious that they wanted the "college boy with the shirt and tie" to experience the same difficulties. The 120mm X 215mm X 76mm (4.7" X 8.4" X 3") spherical roller bearing was on a two foot long armature standing vertical in a holding fixture bolted to the floor. My first observations were directed at discovering the condition of the bearing towards determining the cause of its failure, and I was not really concerned about removal instructions. When I bent over to inspect the bearing, I grasped its O.D. much like a steering wheel so I could turn the outer race to see if the rollers would still roll. What I saw was the rollers jammed tight between the races, but the inner race turned on the shaft with the outer race. This meant it was not press-fitted to the shaft anymore, so with a mighty tug, I pulled the whole bearing right off. Well, I wish we had VCRs back then to record jaws dropping and eyes bugging, followed by all walking around in disbelief. One gentleman mumbled, "H-h-how did you do that?" To which I quickly replied,"You have to give these a little twist before quickly pulling. Do you have any more you want me to remove?"

After a few moments passed, during which they confirmed to one another what they witnessed, I told them what really happened. First, they probably had a bearing with a minimum bore size, which would have developed a tighter shaft fit than previous bearings they had successfully removed. Second, they must have heated the bearing hot enough and long enough so that the metalic structure of the steel re-oriented itself into a different phase that started later in the day and progressed overnight. This new phase produced a slightly larger volume of material, which means the whole part grew in size: enough to obviously remove the interferance fit. Third, when I came along, all I had to do was lift it off. I went on to explain that their removal procedure was basically good, and they could eliminate such a problem in the future by using two torches to heat the bearing inner race up quicker so that it stays about 200 degrees F hotter than the shaft allowing the puller to do its job.

Oh, by the way, I solved the premature failure problem by changing the motor maker's re-greasing recommendation from one ounce a month to one-half ounce every two weeks.

Bob S.
Morganton, NC
rsalter@amroll.com