AGMA 13FTM09-2013 Investigations on Tooth Root Bending Strength of Case Hardened Gears in the Range of High Cycle Fatigue.pdf

1、13FTM09 AGMA Technical Paper Investigations on Tooth Root Bending Strength of Case Hardened Gears in the Range of High Cycle Fatigue By Dr. N. Bretl, S. Schurer, Dr. T. Tobie, Dr. K. Stahl and Dr. B.-R. Hhn, Gear Research Centre (FZG)2 13FTM09 Investigations on Tooth Root Bending Strength of Case Ha

2、rdened Gears in the Range of High Cycle Fatigue Dr. N. Bretl, Stefan Schurer, Dr. Thomas Tobie, Dr. Karsten Stahl and Dr. Bernd-Robert Hhn, Gear Research Centre (FZG) The statements and opinions contained herein are those of the author and should not be construed as an official action or opinion of

3、the American Gear Manufacturers Association. Abstract Tooth root load-carrying capacity is one of the determining factors in gear design. In addition to the strength of the material itself, the existing state of stress significantly influences tooth root load-carrying capacity. Based on extensive ex

4、perimental investigations of gears, the beginning of the fatigue strength range is generally set 3106load cycles, which common calculation methods, like ISO 6336, also take into account. According to this, standard test methods for tooth root bending endurance strength usually assume a load cycle li

5、mit of 3-6106. However, current as well as completed studies on tooth root load-carrying capacity show tooth root fractures with relatively high numbers of load cycles in a range of general fatigue strength and above. Analysis of these fracture surfaces shows that these late breakages are often init

6、iated by small inclusions or microstructural defects in the material. These tooth fractures that initiate with cracks under the surface have a negative effect on the tooth root load-carrying capacity in the range of high cycle fatigue. Therefore, experimental investigations regarding high cycle fati

7、gue have been carried out in a pulsator test rig on gears of various sizes, materials and residual stress conditions. As a result, depending on the existing residual stress condition, there are different levels of tooth root load carrying capacity, different failure behaviors in high cycle fatigue a

8、nd different types of damage. Especially for test variants with high residual stresses, the size of the gear and the cleanness of the material have an impact on the tooth root load-carrying capacity and the damage pattern. This paper discusses the different fracture modes by means of examples. Furth

9、ermore, it presents the influence of residual stresses, size and material cleanness on the tooth root load-carrying capacity and on the type of tooth root fractures with crack initiation on and under the surface. These influences will be additionally confirmed by examples of experimental test result

10、s. Copyright 2013 American Gear Manufacturers Association 1001 N. Fairfax Street, Suite 500 Alexandria, Virginia 22314 September 2013 ISBN: 978-1-61481-066-7 3 13FTM09 Investigations on Tooth Root Bending Strength of Case Hardened Gears in the Range of High Cycle Fatigue Dr. N. Bretl, Stefan Schurer

11、, Dr. Thomas Tobie, Dr. Karsten Stahl and Dr. Bernd-Robert Hhn, Gear Research Centre (FZG) Introduction Gears are one of the critical components which determine the capability and reliability of drive systems. Thus the tooth root load-carrying capacity is one of the determining factors in gear desig

12、n. Continuous demand for higher efficiency, increased load-carrying capacity and endurance life, while at the same time, ensuring smaller size and low costs, increasingly often make the solid expertise of fatigue mechanisms of gears indispensable. In addition to the strength of the material itself,

13、the existing state of stress can significantly influence the tooth root load-carrying capacity and the associated fracture mode. According to the current state of the art, case hardened gears for industrial gear box applications often are peened for the cleaning process. Along with the effects of cl

14、eaning, peening processes also substantially affect the tooth root load-carrying capacity. Additionally, controlled shot-peening processes lead to an increase of the tooth root load-carrying capacity up to more than 15% 3 compared to blast cleaned gears. Based on extensive experimental investigation

15、s on gears different kinds of fracture modes can be observed depending on the present state of stress. According to the current state of knowledge it must be distinguished between breakages with an initiation of fatigue cracks on and under the surface. Especially those breakages with an initiation o

16、f cracks under the surface have a negative effect on the tooth root load-carrying capacity in the range of high cycle fatigue. The scope of this research project was to increase the knowledge of the tooth root load-carrying capacity of cylindrical gears with different residual stress conditions, esp

17、ecially in the range of high numbers of load cycles up to 100 106. Therefore, an extensive program of gear tests in pulsator test rigs has been carried out to verify on the one hand the different types of tooth root fracture modes and, on the other hand, the associated dominating influences on high

18、cycle fatigue of gears. Internal fracture mode in the tooth root Due to common experience of tooth root bending tests in pulsator test rigs, the endurance limit of the tooth root fatigue strength is usually considered to be 3 106load cycles. As a result, suitable standard tooth root bending tests se

19、t the endurance limit at which a test is meant to be graded as fatigue resistant to 3-6 106cycles. In current and completed studies on tooth root load-carrying capacity at FZG, especially on case hardened gears, more and more frequent tooth root breakages were found that occurred in and above the ra

20、nge of these ultimate numbers of cycles. Analysis of corresponding fracture surfaces show that these damages in the range of high numbers of load cycles often initiate from small inclusions or microstructural defects in the material. Similar damage patterns determined on simple specimens exhibiting

21、a small bright spot at the crack origin are well known from technical literature, the so called “fish-eye” (Figure 1). Figure 1. Example of tooth root breakage with “fish-eye” failure 4 13FTM09 Extensive studies 8, 9 on the initiation and propagation of internal cracks in gears or simple specimens a

22、lready exist. Most of them describe the same propagation process of fatigue crack initiation and essentially specify the three different areas of the fish-eye failure according to Figure 2 9. The inclusion in the center surrounded by the GBF (Granular Bright Facet), an area of multiple microcracks,

23、and the surface crack, as a result of the subsurface cracks, form the typical appearance of these failures. Surface crack initiation occurs at high stress levels and relatively low numbers of load cycles, whereas subsurface crack initiation is mainly observed at lower stress levels and in the high c

24、ycle range. As a consequence, failures with a crack initiation under the surface tend to decrease the tooth root load-carrying capacity in regions of high numbers of load cycles. As a result, the normally applied S-N curve has to be converted schematically to a stepwise S-N curve (Figure 3), which t

25、akes into consideration the possible reduction in load-carrying capacity in the high cycle range. Relating to the different gear standards, ISO 6336 7 takes a decrease of the tooth root load-carrying capacity in high cycle areas into account. The corresponding life factor is based on AGMA 2001 1 sta

26、ndard and considers various influences, such as cleanness of material, residual stresses or signs of fatigue. This factor is the result of assessments and experiences and is insufficiently secured by systematic experimental results in the high cycle range. Influence of the existing state of stress i

27、n the tooth root Gears are exposed to different types of stresses during their manufacture, for example from milling, heat treatment, grinding, shot peening. These stresses in the material overlap with applied stresses caused by the existing load. The different states of stress affect the load-carry

28、ing capacity of gears in different ways. To evaluate the load-carrying capacity of gears it is necessary to know the local stress situation and the associated strength of the material itself. The local load induced stresses mainly depend on the current load situation and the size of the gear, wherea

29、s the local strength of the material is primarily affected by base material, previous heat treatment, peening conditions or irregularities in the material itself (according to Figure 4). Figure 2. Schematic illustration of subsurface crack growth 9 Figure 3. Schematic illustration of stepwise S-N cu

30、rve 8 5 13FTM09 Figure 4. Schematic distributions of load stress, residual stress and hardness in tooth root area of case hardened gears depending on material depth Common calculation methods for tooth root bending stress, such as ISO 6336-3 7, apply the point of contact of the 30 tangent and the to

31、oth root fillet to maximum local tooth root bending stress. The maximum local tooth root bending stress generally occurs on the surface of the root fillet due to slip deformations, surface defects and additional notch effects. Depending on the current state of stress, different fracture modes can be

32、 determined on gears. Experimental investigations 3, regarding high cycle fatigue on gears of various sizes, materials and residual stress conditions showed that the main factors on high cycle fatigue of external gears are residual stresses, tooth size and the material cleanness. These influences ar

33、e regarded more closely in the following. Residual stresses The state of the art in industrial practice is often a cleaning process after previous heat treatment to remove scale layers and impurities of case hardened gears with the help of blast cleaning techniques. Besides the mentioned cleaning ef

34、fects, an additional increase of the tooth root load-carrying capacity caused by residual compressive stresses can be noticed. This positive stress situation on and just below the tooth root surface counteracts the tensile load stresses and leads, especially on the surface and in the near surface la

35、yer, to an increase of the local fatigue strength. Therefore, the values of the tooth root load-carrying capacity for case hardened gears, material quality MQ, published in the standard ISO 6336-5 7, depend on gears that have been professionally blast cleaned under industrial standards. Furthermore,

36、 ISO 6336-5 7 attributes case hardened and shot-peened gears an increase of the tooth root load-carrying capacity of 10% compared to blast cleaning. Concerning the major influence of peening in general, extensive experimental investigations in diverse studies 6, 10, 12 on gears and other components

37、have been carried out. Figure 5 and Figure 6 contain exemplary test results of tooth root bending tests on unpeened and peened gears of different sizes and materials. The increase of tooth load-carrying capacity of peened gears can be observed. It should be noticed that these test results are based

38、on standard tooth root bending test methods with an endurance limit of 6 106load cycles. 6 13FTM09 Figure 5. Increasing tooth root load-carrying capacity because of peening process 10 Figure 6. Experimental test results showing the influence of peening process on tooth root load-carrying capacity of

39、 case hardened gears 12 In addition to the positive influence of peening on the tooth root load-carrying capacity, based on standard test methods, the resulting residual compressive stresses also affect the high cycle fatigue. Typical distributions of residual stresses due to different peening opera

40、tions are shown in Figure 7. Residual compressive stresses influence the surface and near surface layer and subside very quickly into the depth of the material. As a result the residual compressive stresses counteract the highest prevailing load stress conditions on and just below the surface, and m

41、ay prevent an initiation of cracks from the surface. However, in deeper parts of the material the local residual stresses subside much more quickly than the load stresses and cannot significantly contribute to the local material strength. As a consequence, the local stress can exceed the local endur

42、ance strength of the material, especially with the presence of defects in the material, which may lead to an initiation of cracks under the surface. These cracks propagate slowly, depending on the current load, and may influence the high cycle fatigue. Therefore one objective of this study was to in

43、vestigate the fatigue behavior of case hardened gears of various sizes, materials and peening conditions. The following model representation was established based on the experimental results in order to interpret the test results in consideration of the current state of the art. 7 13FTM09 Figure 7.

44、Typical residual compressive stress progressions of non-peened, blast cleaned and shot-peened gears The evaluation of the experimental investigations resulted in the conclusion that each breakage, with crack initiation on and under the surface, has its own S-N curve corresponding to the respective f

45、racture mode. The position of these two S-N curves depends mainly on the current residual stress situation and the material cleanness. Thereby residual stresses primarily influence the position of the S-N curve for surface fatigue, material cleanness the position of the S-N curve for subsurface fati

46、gue. Due to the experimental results of unpeened spur gears with residual compressive stresses 900 N/mm. Based on the induced residual stresses, the surface and the surface layer experience a further strengthening and therefore the S-N curve for surface fatigue is additionally lifted up to a higher

47、stress level (Figure 10). The result is that the difference of the stress level between the two S-N curves varies widely, depending on the material cleanness, the residual compressive stress and the load stress. Furthermore the appearance of tooth root breakages with crack initiation under the surfa

48、ce are promoted and at the same time leads to a more or less sharp decrease of the tooth root load-carrying capacity in high cycle fatigue. Based on the test results, shot-peened gears with a high material cleanness did not show breakages with crack initiation under the surface. Following the establ

49、ished model representation, the S-N curve for the internal fracture mode is lifted up to a higher stress level and breakages with crack initiation under the surface can be avoided (Figure 11). As a result, the desired increase of the tooth root load-carrying capacity can be used even at high numbers of cycles. Figure 10. Schematic S-N curves for shot- peened gears Figure 11. Schematic S-N curves for shot-peened gears and influence of high cleanness of the material 9 13FTM09 Tooth size The influence of the tooth size on the fatigue strength is e