A 7316 Angular Contact and NUF 216 had ran for only seven hour prior to failure in large industrial machine. By viewing application drawings it is clear that the NU216 Cylindrical Roller bearing was used to carry radial loads and the Angular Contact bearing at the opposite end of the application was intended to hold the shaft in position axially. The user was convinced that the cylindrical roller was faulty as most of the apparent damage was seen on that bearing.
Approximately 50% of the outer ring from the NU216 was provided for infestation along with four severely damaged rolling elements. Clear evidence of tarnishing was noted on the outside diameter surface of this outer ring. The raceway was fatigued as were the sides of the roller retaining ribs. All surfaces were seen to be a deep grey blue in colour; such discolouring suggested that the bearing had experienced very high temperatures during service. From the condition of the raceway and ribs it was concluded that the heat generation had occurred within the bearing.
The four rollers were heavily distressed and deformed and had also been affected by heat generation
The 7316 angular contact bearing was returned whole condition with ample amounts of lubricant inside it. The bearing was very rough to rotate and the lubricating grease found to be highly contaminated with metallic debris possibly from a bearing raceway. No moisture was detected in the residual grease.
The cage was found to be intact with, very little wear was seen in the ball pockets. The back surface of the cage had clearly been rubbing and wearing against a solid surface. By further review of the application it was identified that only an end cover would come close to the back face of the cage.
The outside surface of the outer ring showed only normal fitting marks. No rotation had occurred and no signs of tarnishing due to heat were present. At points on the large side face small witness mark were identified running circumferentially around the ring adjacent to the bearing chamfer. The large side face was shown in the drawing abutting a full face washer that should not produce witness mark of the type seen.
The balls inside an angular contact bearing are designed to carry both radial and axial load in a single direction, the normal ball running track should therefore be on the ground and superfinished side of the outer ring.
It was observed that the ground and superfinished side of the outer ring raceway was in an as machined condition and that no loaded contact had occurred. The flat portion of the outer ring raceway, a portion of the raceway not designed to carry load, shoed clear evidence of damage caused by contaminants within the lubricant being rolled onto the raceway by the load carrying elements.
The inner ring inspection revealed evidence of fitting marks and rotation within the bore, and a small thickness side face showing wear over its full surface. This was considered abnormal as the drawing shows the shaft abutment shoulder only contacting approximately 50% of the ring surface area not 100% as was the case.
The raceway of the inner ring showed no signs of ball running contact at any point, the raceway surfaces being in an as produced condition. At the edge of the turned portion of the inner ring, adjacent to the small diameter sideface, an area of metal fatigue and flaking was found running around the circumference. From the damage seen it was clear that the rolling elements had been running off the correct load carrying surfaces of the bearing.
The damage seen inside the bearings was consistent with the assembly having been running with the 7316 bearing carrying the thrust load in the opposite direction to that intended.
The witness marks on the outer ring and lack of marks on the side face of the inner ring clearly showed that the angular contact bearing was fitted in the opposite direction to that shown on the drawing. If this is the case then clearly the 7316 had been running in a direction that would result in the trust faces not carrying load.
The NU216 side ribs had been forced to carry all the axial load, not a force they are designed to withstand and consequently a catastrophic failure of the cylindrical roller bearing resulted.
Care needs to be taken to ensure that where an assembled bearing is designed to carry loads in a particular direction that the installation team do indeed fit the bearing the correct way out. The original assumption by the end user was that the cylindrical roller was faulty, the true root cause was incorrect fitting of the angular contact bearing.