1、- 99FTM14 The Effect of Material Defects on Gear -Performance - A Case Study by: R.J. Drago, Mechanical Power Transmission Systems Consultants and A.F. Filax, University of Chile TECHNICAL PAPER COPYRIGHT American Gear Manufacturers Association, Inc.Licensed by Information Handling ServicesThe Effec
2、t of Material Defects on Gear Performance Raymond J. Drago, Mechanical Power Ransmission System Consultants and Alejandro Font Filax, University of Chile The statements and opinions contained herein are those of the author and should not be construed as an official action or opinion of the American
3、Gear Manufacturers Association. Abstract The quality of the materia1 used for highly loaded critical gears is of primary importance in the achievement of their full potential. Unfortunately, the role which material defects play is not clearly understood by many gear designers. The mechanism by which
4、 failures occur due to material defects is often circuitous and not readily apparent. In general, however, failures which are associated with material defects show characteristics which point to the source of the underlying problem, the mechanism by which the failure initiated, and the manner in whi
5、ch it progressed to failure of the component. In this case study, the authors examine the failure of a medium sized pinion used in a mining application. The mode of failure was rather catastrophic in nature but did not follow any of the typically understood mechanisms such as tooth bending, surface
6、distress, wear, etc. Often, as wasthe case for the subject pinion, material defects do not manifest themselves in these more typical tooth orientedfailure modes, though the initial presentation of the failure often suggests more classical origins. A complete shaft fracture was the ultimate cause for
7、 the pinions removal from service. Initial inspection of the failed pinion indicated the presence of cracking in the toothed area. This cracking appeared to have progressed through the pinion and resulted in the shaft fracture which caused the pinion to cease transmitting torque. In order to avoid a
8、 recurrence of the problem in this very critical application, it was very important to understand the failure fully including its cause and progression mechanism. This paper presents a summary of the failure, its investigation, and the methods proposed for its resolution. The data is presented in tu
9、torial format so that the basic effect of the identified material defects can be better understood and used in future designs. Copyright O 1999 American Gear Manufacturers Association 1500 King Street, Suite 201 Alexandria, Virginia, 22314 October, 1999 ISBN: 1-55589-752-5 COPYRIGHT American Gear Ma
10、nufacturers Association, Inc.Licensed by Information Handling ServicesThe Effect of Material Defects on Gear Performance - A Case Study - Raymond J. Drago. P. E. Chief En, uineer Drive Systems Technology. Inc. Mechanical Power Transmission Systems Consultants 24 Marlborough Lane Glen Mills, PA 19342
11、-1519 Professor Alejandro Font Filax. PhD Department of Mechanical Engineering University of Chile Beaucheff 850 - 5 Piso Casilla 2777 Correo 2 1 Santiago. Chile The quality of the material used for highly loaded critical gears is of primary importance in the achievement of their full potential. Unf
12、ortunately, the role which material defects play is not clearly understood by many gear designers. The mechanism by which failures occur due to material defects is often circuitous and not readily apparent. In general, however, failures which are associated with material defects show characteristics
13、 which point to the source of the underlying problem, the mechanism by which the failure initiated, and the manner in which it progressed to failure of the component. In this case study, the authors examine the failure of a medium sized pinion used in a mining application. The mode of failure was ra
14、ther catastrophic in nature but did not follow any of the typically understood mechanisms such as tooth bending, surface distress, wear, etc. Often, as was the case for the subject pinion, material defects do not manifest themselves in these more typical tooth oriented failure modes, though the init
15、ial presentation of the failure often suggests more classical origins. A complete shaft fracture was the ultimate cause for the pinions removal from service. Initial inspection of the failed pinion indicated the presence of cracking in the toothed area. This cracking appeared to have progressed thro
16、ugh the pinion and resulted in the shaft fracture which caused the pinion to cease transmitting torque. In order to avoid a recurrence of the problem in this very critical application, it was very important to understand the failure fully including its cause and progression mechanism. This paper pre
17、sents a summary of the failure, its investigation, and the methods proposed for its resolution. The data is presented in tutorial format so that the basic effect of the identified material defects can be better understood and used in future designs. COPYRIGHT American Gear Manufacturers Association,
18、 Inc.Licensed by Information Handling ServicesIntroduction In 1992, two new single helical gear sets were installed as part of a new mill system at a mine site. The pin ion and gear sets are used to drive essentially identical ball mills used for processing ore from the mine. After some relatively m
19、inor start up problems, the gear sets were placed into normal service and performed without significant incident until some pitting was discovered during a visual examination conducted iate iii 1994. At that time, we noted only very minor surface distress, as Figures I and 2 show. The distress was l
20、ocated at about the middle of the face width aiid had been slightly ground out at the time of iny November 1994 visit. The pinion shown in Figures 1 and 2 was in service until removal in August 1995 because of severe surface damage. It was reinstalled in mid 1998 and failed catastrophically in 1998,
21、 approximately three working years after startup. The number of cycles accuinulated by the pinion at the tiine of failure was something lower than 3 X IO. This. therefore, was certainly not a low cycle failure, but one third of the total design life. The failed pinion was replaced and the ndl restar
22、ted. During a visual inspection conducted in ea-ly 1999, tlie new, replacement pinion was found to be exhibiting pitting very siinilar in both extent and location as the original pinion. This pitting was hand dressed out and was found to have not progressed significantly during a visual inspection w
23、hich was conducted during April 1999, as Figure 3 shows. I Figure 1 - Distress Observed on Ball Mili Pinion During I994 Visual Evaluation I Figure 3 - Currently Installed Ball Mill #2 Pinion Showing Region (Near Mid Face) Which Has Been Ground Out to Remove Pitting and Surface Cracks and Then Restar
24、ted I Figure 2 - Close View of Distress Observed on Ball Mill Pinion Durino 1994 Inmeclion During an inspection in April 1999 we did not observe any cracks on this pinion (we were unable to dye check it) but it was reported to have some cracks in the pitted region which were discovered during a prev
25、ious dye check inspection. The cracks were ground out and the pinion was returned to service. At the time of our inspection, it appeared that only minor progression of the pitting was occurring. ,I Second Helical Gear Set As noted above, there are two identical inill gear systeins on line at this si
26、te. The second inill was running at the tiine of our visit and we were unable to shut it down for a good look at the teeth. We were able to visually examine the teeth through the inspection port with the aid Page 1 of 7 COPYRIGHT American Gear Manufacturers Association, Inc.Licensed by Information H
27、andling Servicesof a strobe which allowed the pinion and gear rotation to be “stopped“ artificially. As Figure 4 shows, this pinion exhibited moderate surface distress in the formof pitting. I Figure 4 - Pitting Distress on Second Mill Pinion Teeth We could not detect any cracks on the surface of th
28、is pinion, however, it would be almost impossible to do so under the conditions of this inspection. The regions of pitting observed were siinilar in size to those observed on the first (failed) Mill Pinion but they were in a somewhat different in location along the race width. In the profile directi
29、on, the pitting was located near the pitch line on both mill pinions. I Figure 5 - Ball Mill Pinion #1 Teeth Which I Exhibit No Surface Distress Lubricant distribution on both gear sets appeared to be adequate and uniform across the face width. There was no indication of any lubrication related dist
30、ress on any of the pinions or gear examined. The distress is located about 2/3 of the face width froin the motor end of the face width. While Figure 4 shows typical distress, it is important to note that not all pinion teeth exhibit this distress. Figure 5 shows another group of pinion teeth on the
31、saine pinion which exhibit virtually no distress at all. Failed Mill Pinion The pinion which failed in August 1998 was available for our visual examination, however, as Figure 6 shows, it was quite rusted. Still, our visual evaluation provided valuable clues regarding the cause of the failure. I Fig
32、ure 6 - Failed Pinion Removed From #2 Ball I Mill in Auauct 1998 It was apparent that a crack propagated through tlic pinion. The cracl was at about the divided the iace width at about the 2/3 - 1/3 point. The crack propagated through about 75% of the shaCt tliickiiess, Figure 7, but did not complet
33、ely sever the pinion. This failure was discovered during normal maintenance and no problem had previously been noted during operation. At shut down, maintenance personnel found il piece missing from one tooth and noted the crack extending about 75% around the circumference of the pinion. ,I It was r
34、eported that this pinion had been run in both directions and the contact apparent on the tooth fianks confirms this. It was also reported to me that this pinion had been running in one direction only for “. many years .“ The pitting observed was reported to have occurred generally at four locations,
35、 positioned about 90 degrees apart. The pitting which I observed on this pinion, Figure 8, was not sufficient, of and by itself, to account for this massive crack propagation failure. Note also that this level of pitting is not very different froin Page 2 of 7 COPYRIGHT American Gear Manufacturers A
36、ssociation, Inc.Licensed by Information Handling Servicesthat which I observed on this same pinion when I last visually evaluated it during my November 1994 visit (see Figures 6 and 7 above). Close examination of Figure 8 clearly shows that there was no real progression of the pitting distress and t
37、hat the fracture does not appear to originate in the pitted region. relatively smooth in appearance. It is clear that multiple crack branches have propagated in the region of the fracture. This topography is suggestive of a fatigue crack propagation mechanism. Scanning electron microscopy of this re
38、gion showed crack growth rates of 10.“ indcycle, that is, 80 min in 3 days of operation. I Figure 7 - Pinion Fracture Face (Light area at top of pinion is crack propagation area. Dark area at bottom of the pinion is area which was cut after the failure.) I Figure 8 - Pitting Observed on Fractured Pi
39、nion (Note ground out condition of pitted area and lack of new progression) I Figure 9 - Fractured Tooth Surface Suggestive of a Subsurface Flaw Based on the view of the fracture face shown in Figure 10, it appears that the original failure had multiple origins and that it spread froin the tooth roo
40、t asea down into the shaft section. I Figure 1 O - Cracks Progressed From the Tooth I Roots The surface characteristics of the failure, as showii iii Figure 9, suggest the presence of a subsurface defect which initiated the crack which propagated to failure. Despite the rusted nature of the fracture
41、 face, it is still Current Pinion Condition The condition of the pinions and gears currently installed in both mills are very similar to each other. At this time neither pinion exhibits any critical probleins, however, i Page 3 of 7 COPYRIGHT American Gear Manufacturers Association, Inc.Licensed by
42、Information Handling Servicesthe overall condition of the pinions are very similar to the condition of the fractured pinion which failed in 1998. This is, of and by itself, some cause for concern. The pitting which is apparent on these pinions is somewhat unusual in both its location and its appeara
43、nce. The fact that the pitting is largely mid face indicates that it will not be practical to realign the gear sets to favor the undamaged tooth surfaces. The appearance of the pitting suggests an over crowned condition on either the pinions or their mating gears or, perhaps. both. I have no indicat
44、ion of the actual geometry of these pinion and gear sets (i.e., lead and profile inspection charts or other similar measurements) which might identify the specific cause. considering the experience of the failed pinion. however, and the similarity of the appearance of the pitting on all three pinion
45、s, it is quite possible that the two currently installed pinions also exhibit some of the material anomalies which afflicted the failed pinion and. as described below, ultimately led to its fracture. If this pitting is the result of material anomalies as was the fractured mill pinion described below
46、, dressing the pitted area will lower the stress level in the region where the inclusions appear to be prevalent thus reducing the possibility that they will cause a crack initiation and progression mechanism similar to that which led to the pinion failure. Failed Pinion Shaft Analysis The fracture
47、failure of this pinion shaft is an extreme example of the highly deleterious effects of material anomalies on the load capacity, reliability, and life expectancy of large gears. The specific mode of failure experienced was via initiation of cracks near the top of one tooth with propagation occurring
48、 across approximately 75% to 85% of the shaft cross section, as Figures 11, 12. and 13 which show the mating fracture faces. clearly indicate. The cross hatched areas in these figures is the region which was cut to expose the fracture faces while the unhatched regions show the region over which the
49、cracks progressed. Note especially the origin of the failure in the tooth top at the top of both figures and the extensive branched crack network below the origin. Despite this extensive propagation, no indication of the failure of the shaft was observed during normal operation of the mill and the fracture was only noted during a maintenance check. This crack propagation took place in less than three months of operation. Figure 11 - Pinion Fracture Face. Mating Surface I Shown in Figure 12 The condition of the tooth flanks and the location of the fracture (almost mid face) also indicat
