1、04FTM1Gear Noise - Challenge and SuccessBased on Optimized Gear Geometriesby: Dr. F. Hoppe and Dr. B. Pinnekamp, Renk AGTECHNICAL PAPERAmerican Gear ManufacturersAssociationGear Noise - Challenge and Success Based onOptimized Gear GeometriesDr. Franz Hoppe and Dr. Burkhard Pinnekamp, Renk AGThe stat
2、ements and opinions contained herein are those of the author and should not be construed as anofficial action or opinion of the American Gear Manufacturers Association.AbstractFor gear transmissions different levels of requirements with regard to noise excitation have to be matched.Industrialapplica
3、tionsforconveyorbeltsorcementmillsarewithoutdoubtmuchlesssensitivewithrespecttonoise emission than military applications, e.g. for navy ship propulsion. However, also for industrialapplications the air borne and structure borne noise behavior more and more becomes an important feature.RENKhasbeendev
4、elopingoptimumgearunitsforallapplicationswithatransmittedpowerlevelrangingupto145 MW. This paper describes requirements and solutions with regard to noise behavior focusing onexamplestakenfromnavyapplicationsandwindturbinegeartransmissions.Theindividualapproacheshaveto be a suit-able compromise to m
5、eet the challenge of noise requirement and cost optimization without anyrestriction on gear load carrying capacity. Therefore, there is no general but individual solution for optimumdesign.Thepapercomprisesbasicconsiderationswithregardtogearnoise, noiserequirements andmeasurementsat shop and field t
6、ests in comparison to gear geometry and calculation results.Copyright 2004American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314October, 2004ISBN: 1-55589-824-61Gear Noise Challenge and Success Based on Optimized Gear Geometries Franz Hoppe, BSc, PhD: Gene
7、ral Manager Marine Gears RENK AG, Germany Burkhard Pinnekamp, BSc, PhD: Engineer, Special Programs Marine and Industrial Gears RENK Corp, Duncan, SC 1 INTRODUCTION Power generation or Navy vessel propulsions are two examples for the application of advanced gear systems. Typical applications are wind
8、 turbine mul-tiple stage step up gears, or CODOG and CODAG marine gears for up to 30 MW gas turbines and diesel engines up to 7 MW to controllable pitch propellers. In any case, the supplier of main reduc-tion gears has to provide the optimized technology for any kind of installation at utmost inter
9、face flexi-bility. This can only be achieved in close co-operation with power train suppliers, or Navies and shipyards. Furthermore, in continuous development of tech-nology enhancement, single components are sub-jected to increased refinement such as gear teeth as the heart of gear transmissions fo
10、r the benefit of optimization of the load carrying capacity and low-est noise performance. To achieve optimum noise behavior, different criteria of gear design have to be observed, such as selection of bearings (damping features) and housing (mass, noise dissipation fea-tures) and gear geometry. In
11、this respect, an ex-treme importance comes to the correct selection of macro geometry parameters as well as tooth cor-rection values, where the proper evaluation of both is supported by continuously adapted calculation methods. However, theory needs to be transferred to real operable gears with the
12、required accuracy applied, and therefore production means such as heat treatment processing and grinding tools are simultaneously to be maintained on the latest stage of technology. In completion of a modern main transmission gears, the environmental demands on control aspects require an ongoing ada
13、ptation. Today, integral pro-grammable logic control (PLC) systems provide an utmost flexibility, safety and comfort for the opera-tion of reduction gears. The gear is not just a me-chanical transmission it is developed to a system with functional sub stations interfacing with the plants operational
14、 environment. 2 GEAR NOISE BASICS 2.1 Tooth Geometry Basics for Low Noise A main power transmission gear is subjected to various external influences, such as reaction loads from adjoined external couplings, foundation distor-tion, dynamic mass forces due to heavy sea states or transient wind forces,
15、 and, not lastly, heat ex-pansion due to the power loss generated by gear teeth and bearings. In view of all these impacts, and also respecting in most marine field cases low noise generation re-quirements, the tooth design is to be focused on in particular, as pinions and gears represent “the heart
16、” of a gearbox. First priority, the decision on the basic type of gear teeth is of importance, where principally spur gears, single helical or double heli-cal gears are available. Figure 1 shows the principle coherence between tooth mesh noise excitation and overlap ratio, . With spur gears, equals
17、zero, with low single helices, values up to 2 are achievable. High heli-ces are in practical sense realized only with double helical gears, achieving 3. Apart from signifi-cant noise reduction at increased , excitation ap-pears to be minimum with integer value of overlap ratio. The fundamental resul
18、ts as depicted in Fig-ure 1 are considered as state of the art and have been confirmed throughout the past 25 years with numerous research programs, supported by experi-ence with countless applications in service. Involute gears theoretically mesh without periodical angular deviation in rotation and
19、 without dynamic excitation. However, due to manufacturing devia-tions, misalignment and elastic deformations under load, this theoretical optimum is not achieved in reality. Manufacturing and alignment can be ad-dressed by optimum quality with regard to gear grinding, assembly and commissioning. De
20、forma-tion under load cannot be avoided but addressed properly by smart design and appropriate flank modification. Above all, the macro geometry still is the decisive criterion on noise excitation. That 2means that only optimized macro geometry allows for optimum noise behavior. The basic background
21、 will be described in the following paragraphs. Gears will change their geometric position under load due to the following influences: Deflection due to Hertzian pressure, bending of teeth, elastic deflection of gear bulks and shafts. Figure 1 Noise in dB generated in tooth mesh, dependent on basic
22、layout and overlap ratio, , acc. to ref. (2) and (3). Note: Minima achieved with even values. Figure 2 Single tooth stiffness, a) spur gear mesh; b) helical gear mesh; acc. to (6) contact ratio overlap ratio start of mesh in A end of mesh in E plane of contact start of meshin A end of mesh in E rota
23、tion plane of contactrotation single tooth stiffness single tooth stiffness 3Along the path of contact, the total stiffness of all teeth in mesh will vary according to the referring bending moment arms and active face width avail-able, Figure 2. For spur gears, the active flank width is constant, fo
24、r helical gears it is low at begin-ning and end of path of contact. Therefore, the stiff-ness varies much more for helical gears than for spur gears. Even when considering deviation free tooth flanks -which is almost achievable with modern manufac-turing methods- due to this elastic deformation un-d
25、er load, the flank position will change and there-fore there will be interference between the gear teeth, see Figure 3, causing periodical noise excita-tion. This interference can be compensated for by appropriate flank modification, which only can be optimized for one load level and one stiffness v
26、alue for the mesh stiffness. Gears requiring low noise emission at different load levels, e.g. navy gears, can not be only optimized for one single load stage. Therefore, an utmost constant total mesh stiffness, c, considering all teeth being in mesh at a time, should be obtained for the complete pl
27、ane of con-tact. Figure 4 shows the course of mesh stiffness over rotational angle for a spur and a helical gear exam-ple. The mesh stiffness is the sum of the individual single tooth stiffness values for the teeth in contact shown on the bottom side of the diagram. It is obvi-ous, that with increas
28、ing number of teeth being in mesh at the same time the amplitude of the periodi-cal change in mesh stiffness decreases. The num-ber of teeth in contact is determined by the trans-verse contact ratio, , (number of teeth in mesh along the path of contact; 1-2 for most gear applica-tions) and the overl
29、ap ratio, , (number of teeth in mesh along the face width; 0 for spur gears, up to 8 for double helical gears). The decisive total contact ratio, , is the sum of and . Integer figures for further improve the mesh quality (Figure 1). Figure 4 Mesh stiffness, a) spur gear mesh; b) helical gear mesh; a
30、cc. to (6) Conclusion: Maximum total contact ratio, , as well as integer value for overlap ratio, , are the optimum basis for noise optimized gears. 2.2 Gear Lay Out with Marine Applications Coming to more distinct views on these principles, without any doubt, spur gears, with zero helix an-gle, are
31、 less suitable for high performance marine gears, neither with commercial vessels nor for smooth run of gears on Navy ships. As described before, such a design provides the tooth load trans-ferred from one tooth pair in contact to the next by a sudden load alteration over the tooth mesh causing peri
32、odical force impacts and, subsequently, in-creased generated noise. Any attempt to decrease these impacts by profile modification may only slightly improve the noise behavior on one specific load level, with the basic disadvantage at remaining lack of contact ratio and still high sensitivity for mis
33、-alignment and concurring tooth edge overload. Single helical gears, Figures 5a) and 5b) com-prise at a first glance some benefits against spur gears. Normally, low helices apply in practice, where tooth load impacts are diminished by helix Figure 3 Tooth deflection under load and sub-sequent interf
34、erence zone in follow-ing gear mesh, acc. to (6) driving geardriven gear interference zone deflected teeth under load rotationcrotationstiffnessstiffness cc: single tooth stiffness cS: actual mesh stiffness c: average mesh stiffness 4angle influenced transverse axial travel of the tooth mesh a signi
35、ficant benefit against spur gears. But, with single helical gears, major aspects are to be considered, subjected to the need of a distinct evaluation, defined with causes and countermea-sures as Limited helix angle 13 due to exceeding axial loads, and still need of separate thrust bearings Low trans
36、verse contact ratios, thus limited minimization of noise performance Axial forces cause tilting shafts in bearings, thus less predictable non-contact areas and uneven load distribution across tooth face width High values for profile and lead modification, designed for maximum continuous load, create
37、 unfavorable tooth contact pattern at lower torque causing exceeding noise at low load op-eration (e.g. cruise mode for naval vessels) Axial forces cause bending of casing structures, to be compensated with added structural weight Eventual re-adjustment of bearings aboard the vessel required to esta
38、blish a proper load share due to unpredictable casing and foundation dis-tortions In spite of the aspects as mentioned above, a single helical gear as per Figure 5a) is appropriate, if measures such as reinforced casing structures, carefully selected tooth flank corrections, and an increased demand
39、to foundation rigidity apply. They are used in single cases, at sufficient performance once the boundary conditions regarding the installa-tions interfaces are properly respected. The high helix acc. to Figure 5b), as proposed by ref. (5), would definitely show a much better noise signifi-cance comp
40、ared to low helices, due to the high overlap ratio, but generates inadmissible high axial loads not compensable in realistic views with mod-erately sized thrust bearings and reasonably de-signed casing structures. b / 2 b / 2 Driving flank Driven flank Torque =30 b b) High Helix c) Double Helix b /
41、2 b / 2 Single tooth pitch = 30 = 9 b a) Low Helix Figure 5 a) Low helix, max. overlap ratio = 3; dotted: position in bearings under load (over-scaled), b) High helix as proposed by (5), c) Double helix at overlap ratio = 5, here sole radial displacement in bearings at no shaft tilting, thus equal t
42、ooth contact over entire face width at all load conditions. Finally, the double helical gear, Figure 5c), is re-garded as the consequent resolution of the aspects above. Applied to more than 80 % in marine gear transmission technology world wide, it combines most favorably by itself centering symmet
43、ric design insensitiveness regarding external forces to casing structures, followed by a high degree of load pattern consistency throughout the full power range, with low noise performance due to maximum achievable helix angles at optimized macro geometry of the gear teeth. In summary, following asp
44、ects are to the favor of double helical gears: Maximum total contact ratio, , for the benefit of smooth tooth engagement and lowest noise performance, see Figure 1 Radial symmetric tooth forces, no axial impact to bearings, due to self centering effect, see Figure 5c) Even contact pattern throughout
45、 all loading conditions 5 No axial loads to casing structures generated, thus most light weight designs achievable, by avoidance of excessive structural reinforce-ments other than ultimately required Tooth corrections by grinding respecting just bending and torsional deflection of pinions, at no til
46、ting load or casing deflection unpredict-able impacts Double helical gears, if designed properly and with appropriately selected macro geometry parameters, present the logical extension of the high helix, Fig-ure 5b), by just adding a second tooth row, at un-changed macro geometry parameters, with t
47、odays state of machine tool technologies easily to be manufactured at highest accuracies. 2.3 Tooth Grinding Technology Upon definition of the macro-geometry of gears following the principles as discussed above, the micro-geometry is to be defined, i.e. distinct tooth flank profile and lead correcti
48、ons, as depicted with Figure 6. These consist of a fine tuned shape of the flanks both in radial and axial directions to achieve an optimized contact pattern in the gear mesh, re-specting the load profile and low vibration impacts. Figure 6 Profile correction (left) and lead correction (right) of a
49、high accu-racy double helical pinion. Once all micro-geometry parameters are set by calculation, these need to be transferred as input values to the grinding machine. It is simple to imag-ine, that finest theoretical values within a range of micrometers are realized only by highly developed machine tools, which provide accurate enough con-trols for the working piece table drive system, and the twin grinding disc support, which, in combina-tion, have to generate the tooth corrections
