Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Non - Involute

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

technical

Non-Involute Gearing, Function


and Manufacturing Compared to
Established Gear Designs
Dr. Hermann J. Stadtfeld and Jasmin K. Saewe

Introduction Positive Profile Shift


Zero Profile Shift
The standard profile form in cylindri-
cal gears is an involute. Involutes are
generated with a trapezoidal rack — the
basis for easy and production-stable
Work Roll
manufacturing (Fig. 1). However, the Work Roll
resulting involute shape not only fulfills
x .m
the law of gearing for a smooth trans-
mission of rotation and torque; it also
Pitch
presents the most robust mathematical Circles
Line Work Line Work
profile function regarding center dis- of E
nga
of E
nga
gem Pitch Line gem
tance changes and other misalignments. ent ent
Involutes can be modified with a profile x .m
shift in order to avoid undercut and to
Generating Rack
provide a strength balance between pin-
ion and gear. The profile shift can also Datum Line Generating Shift x. m = 0 Pitch Line

be used to accommodate certain cen-


ter distances that are considerably afar Figure 1 Generating principle of involute gearing.
from the theoretical center distance for
a standard gearset. Involute gearing has
line contact between the two mating
members. Under load, the contact lines
change to elliptical contact areas.
The elasto-hydrodynamic of invo-
lute gears is well known and easily opti-
mized. The pitch line (pitch point in
Fig. 2) separates the addendum and the
dedendum of the tooth profiles. The
sliding and rolling velocities move the
profile contact from the top of the gear
tooth to the pitch point. The direction of
the sliding velocity changes at the pitch
point but the rolling velocity direction
remains constant. The sliding velocity
directional change at the pitch point
causes a zero sliding velocity condition Figure 2 Profile sliding and rolling of involute profiles.
in an infinitesimally small area “around”
the pitch point. In other words, only a
For Related Articles Search
relative rolling without any sliding exists
between the two mating profiles. tooth shapes
The absence of sliding presents a criti- at www.geartechnology.com
cal caused by the deteriorating hydro-
dynamic conditions around the pitch
point, or pitch line. High load and low
speed reduce the ability of the lubricant

This paper was first presented at the Gleason Corp-hosted 5th (2014) U.S./ WZL Gear Conference.

42 GEAR TECHNOLOGY | January/February 2015


[www.geartechnology.com]
to maintain a surface separating film.
Good gear design and modern, high-per-
formance oils can eliminate this pitch line
phenomenon (Ref. 1).
Scientists and inventors frequently
introduce new, non-involute gear profiles.
The major target of those profile or tooth
forms is the reduction of the relative cur-
vature between the two contacting flanks
in every roll position. Many of the pro-
posed systems seem to have a close rela-
tionship to the cycloidal tooth profile as it
is used in clocks and watches — or to the
Wildhaber-Novikov design, first intro-
duced by Ernest Wildhaber in 1910.
The main argument — often quoted
as explanation — as to why those gear
types have not been discussed and exam-
ined much earlier is the fact that current
computation technology and mechanical Figure 3 Cycloidal generating rack profile development.
machine tools of the past were incapable
of utilizing both the complex tools as meshing gears to the correct center dis- to manufacture cycloidal tooth profiles
well as the complicated machine motions tance location (utilizing the axes play of uses a pointed tool, which follows the
needed in order to create the non-invo- the gears in watches). cycloidal profile, guided either by cams or
lute profiles. A development of a cycloidal generat- interpolating axes motions.
This paper discusses the proposed ing rack tooth profile is shown in Figure The tools to manufacture cycloidal
non-involute and involute-related profiles 3. The reference line or pitch line sepa- gears are either special and, therefore,
with their advantages and disadvantages. rates the addendum and dedendum of expensive, or the manufacturing process
The criteria of the discussion are: the tooth. One roll circle, that is located is very slow. The center distance-main-
• Design calculation above the pitch line, rolls to the left and taining feature is only of interest if the
• Analysis and optimization possibility generates — with one fixed point — the axes’ position can float. In regular power
• Ease of tool manufacturing addendum of the right flank. A second transmissions, where the center distance
• Accuracy of tool geometry roll circle rolls on the pitch line from is rigid, but might vary from the theo-
• Production-stable manufacturing
below and generates the dedendum of the retical center distance, motion error and
• Robust operating performance
right flank (Fig. 3). The two profile sec- vibration are generated.
After an analysis of alternative cylin- tions meet at the pitch line, where both Wildhaber-Novikov gears. The Swiss-
drical gear tooth profiles, some new bevel cycloids have an infinitely high curvature. born Ernest Wildhaber invented in
and hypoid gear geometries that have The cycloidal tooth profile is “S-shaped,” 1926, shortly after his immigration to
been proposed during the past decade are which achieves in the rolling interaction the United States and his employment as
discussed as possible replacements for the between mating flanks a large contact Gleason scientist, a helical gear profile
traditional face hobbing and face milling area on the flank surfaces. The convex which was cut with a circular arc pro-
systems. addendum has constant rolling contact file rack cutter (Ref. 2). Thirty years later,
Involute-based cylindrical gears — as with the concave dedendum. Surface Mikhail Novikov invented in Russia a
well as bevel and hypoid gears — went stress is greatly reduced compared to similar system which featured tooth pro-
through a multiplication of their power involute gears due to this arrangement, files that are circular in the transverse
density during the past 20 years. In order while the root bending stress can poten- plane (Ref. 3). Because of the similarity
to create a fair discussion with objective tially be somewhat lower because of the of the two independent developments,
comparisons, the last part of this paper concave dedendum profile, which blends cylindrical gears with circular tooth pro-
is devoted to present the latest advance- without curvature reversal into the root file have been called Wildhaber-Novikov
ments of “traditional” involute gearing fillet radius. gears. Not only are the tooth profiles cir-
using the example of asymmetric cylin- Cycloidal gear profiles can be gener- cular, but the slots of the gear have the
drical gear designs. ated with generating rack tooth profiles, appearance of half-circles and remind in
Cycloidal gears. Cycloidal gearing has like the one shown in Figure 3.Those pro- its shape of a sprocket. Figure  4 reveals
been described as center distance main- files must be calculated and manufac- that the gear slots have profiles with
taining. The S-shaped profile of the gen- tured dependent upon the individual gear two equal radii, connected with a small-
erating rack teeth will form S-shaped pair. Standard generating profiles — like er radius in the root, which blends with
teeth and generate a force in the center the straight line in involute gearing — are the flank radii. The pinion profile radii
distance direction in order to move the not possible. Another possible process are smaller than the profile radii of the

January/February 2015 | GEAR TECHNOLOGY 43


technical

gear, and are connected at the top with a


straight line.
In order to transmit a constant ratio,
the contacting point between the two cir-
cular flank surfaces must be kept at the
same profile location, which is preferably
a point at mid-profile height with a desir-
able pressure angle. This point is called
the “profile reference point” (Fig. 4). The
different radii of pinion and gear flanks
(p1 and p2) have to be oriented normal
to the profile reference point, but with
a larger gear radius value and a smaller
pinion radius value (Fig. 4, left). Keeping
the contacting point in the same initial
profile location (Fig. 4, left) during an
incremental pinion (and gear) rotation
can only be realized with the introduc-
tion of a helix angle; the helix angle must
be defined such that for a certain pinion
rotation the profile rotates back to the
initial position. Subsequently, the gear
profile also has to rotate back into the ini-
tial position by an angular amount equal Figure 4 Wildhaber-Novikov tooth profile design.
to the angle of the pinion rotation, divid-
ed by the ratio between the two members.
The method of corrective rotation of
the tooth profile in order to transmit a
constant ratio is visualized (Fig. 5). The
first point of contact between the mesh-
ing teeth is shown in the upper graphic
as the connecting point of R1 and R2. As
the gears rotate to an advanced angu-
lar position, the radii R3 and R4 would
transmit a different ratio because the vec-
tor length — as well as the normal vec-
tor direction — changed. The cylinder in
the lower graphic demonstrates that if a
flank line with a particular helix angle is
used that rotates the contacting point for
incremental rotations into the horizon-
tal axis connecting plane, then contact
movement (Fig. 6) can be expected. The
ratio will remain constant in this case
because the radii — and the normal vec-
tors in the horizontal plane along the ref-
erence cylinders — remain constant.
A simple example: A pitch angle of 24° Figure 5 Line of engagement is kept at centerline connecting plane.
and a contact ratio of 1.0. If the pinion
is rotated by 24° and the initial contact
at the profile reference point is located at
the front face of the teeth, then the back
rotation of the profile has to occur as the
contact moves along the face width to the
back face of the teeth.

44 GEAR TECHNOLOGY | January/February 2015


[www.geartechnology.com]
The helix angle at the reference radius
is calculated: ßR = arctan (RR ∙ P/F)
Example:
Given:
Ratio i = 2
Face width F = 50 mm
Pinion pitch radius RP1 = 38 mm
Pinion reference radius RR1 = 41mm
Pinion angular pitch P1 = 24°
Gear pitch radius RP2 = 76 mm
Gear reference radius RR2 = 73 mm
Gear angular pitch P1 = 12°
Wanted helix angle along pitch line in case of
face contact ratio = 1.0:
Pinion helix angle at Ref. rad.: ßR1 = arctan
(41 ∙ 24°/180° ∙ π/50) = 18.95°
Pinion helix angle at pitch rad.: ß1 = arctan (tan
(18.95°)/41 ∙ 38) = 17.66°
Gear helix angle at Ref. radius: ßR2 = arctan
(73 ∙ 12°/180° ∙ π/50) = 17.00°
Gear helix angle at pitch radius:
ß2 = arctan (tan (17.00)/73 ∙ 76) = 17.66°

The example shows that a particular


Figure 6 Contact movement maintains constant ratio.
Wildhaber-Novikov gear design requires
a particular helix angle that depends on
the desired contact ratio and the face
width. The explanation and the calcula-
tion also show that the transverse contact
ratio of Wildhaber-Novikov gears is zero
and the modified contact ratio is equal to
the face contact ratio.
The advantages of circular tooth profile
are the large contact area. Russian studies
from the 1950s and 1960s report about
three-to-five times the load-carrying
capacity, without detrimental pitting or
wear. The reports also state that the cir-
cular wedge geometry between the con-
tacting circles that moves along the face
width, pumps the lubricant into the con-
tact area and generates oil film thickness-
es up to 10 times that of involute gear-
ing, which should also improve efficiency
(Chironis). Disadvantages include the
more complicated rack tool geometry, the Figure 7 Extended version of Wildhaber-Novikov gears.
high influence of the helix angle to the
smoothness of transmission, as well as Because of this constraint, Wilhaber- optimal is high stiffness at the root of the
the sensitivity of center distance changes. Novikov teeth consist of less than a half- teeth and elasticity from mid-dedendum
The helix angle in Wildhaber-Novikov circle, which results in tooth depths that to the tip. The elasticity contributes to a
gears is not a design freedom like in invo- are about 1.2 times the module, com- reduced entrance impact during mesh-
lute gearing, but is exactly given for each pared to 2.2 times the module for stan- ing at different loads and improves the
particular design. The operating vibra- dard involute gears. The pressure angle load sharing between consecutive tooth
tion and noise of Wildhaber-Novikov change along the profile-per-unit-length pairs. In order to account for those facts,
gears has been reported to be higher than in case of less than a half-circle is a multi- Wildhaber, as well as Novikov, mentioned
that of comparable involute gears. ple of the involute curvature change. This in their teachings the possibility to extend
The original Wildhaber-Novikov tooth is an additional reason to the non-invo- their ideas to double-circular profiles that
profiles are basically half-circles, which lute profile function for the high sensitiv- consist of a convex circle at the adden-
call for a certain pressure angle change ity to center distance changes of pure cir- dum and a concave circle at the deden-
between top and root. The end of the cular profile forms. The low-profile teeth dum (Fig. 7).
internal circle towards the top and the show a very high stiffness, which often All variations of Wildhaber-Novikov
external circle towards the root are given is falsely judged as an advantage and a gears can be manufactured by the hob-
by a pressure angle that drops below 5°. contributor to high power density. More bing and shaping processes. Depending

January/February 2015 | GEAR TECHNOLOGY 45


technical

upon whether the tooth normal profile the extended Wildhaber-Novikov profile nominal pressure angle at mid-adden-
or the tooth transverse profile should be design (Ref. 4). Nagata bases the profile dum.
of circular shape, the cutter rack profile definitions not on the gear teeth, but on Nagata shows in a theoretical evalua-
must consist of modified curves (no cir- the rack cutter or reference profile. The tion a significant reduction in center dis-
cles) to accommodate for the generating basic construction of this profile is rep- tance sensitivity of the improved profile
motion between cutter and work. Also, resented in Figure 8. The new profile has combination of circular and involute ele-
the use of circular-shaped rack cutters is a circular addendum and a dedendum ments. In a following study, Nagata tested
mentioned in the literature, which would that consists of two involutes that blend a variety of Wildhaber-Novikov-Nagata
of course generate complex non-circular at mid-dedendum. The upper dedendum gears and published the results with rec-
profile curves. is developed with a cord — unrolled from ommendations for optimal parameters
Profile development. Shigeyoshi the upper base circle — where the lower in 1985 (Nagata 85). The manufacture of
Nagata, a professor at the University of dedendum is developed with a cord that Wildhaber-Novikov-Nagata gears is pos-
Tokyo published a paper in 1981 where unrolls from the lower base circle. The sible by hobbing and shaping. All com-
he discusses a proposed improvement of addendum is circular (Fig. 8) and has a mon hard-finishing methods can also
be applied if the tool profile is formed
accordingly. Like in standard Wildhaber-
Novikov gears, the helix angle is required
and differs depending on the individual
gear design.
Convoloid gearing. Bernard Berlinger
and John Colbourne introduced in a
paper, published in 2011, a tooth form
called Convoloid (Ref. 5). The new tooth
form appears optically very similar to the
Wilhaber-Novikov-Nagata development
(Fig. 9). The addendum has a convex
shape, while the dedendum is concave.
The transition zone at the pitch point
seems to be “S-shaped” rather than the
straight section of Nagata’s development.
The authors report that the tooth profiles
are computer calculated as a point cloud
for each application case individually.
O n e i nte re st i ng c on clu s i on of
Figure 8 Wildhaber-Novikov-Nagata profile construction. Berlinger’s and Colbourne’s findings is
the fact that while involute gearing fit
well with traditional, mechanical gear
manufacturing machines, it is outdated
for today’s engineering and manufactur-
ing environment. Test rig investigations
of Convoloid gears resulted in 20%-35%
increased torque levels vs. involute gear-
ing. The center distance insensitivity of
involute gears seems not to be given, but
the inventors state that the Convoloid
gears can withstand the customary
deflections given in modern gearboxes.
An interesting aspect of Convoloid
gears is that the tooth contact can move
from root to top while maintaining the
correct ratio. This makes Convoloid gears
independent from the helix angle and
allows a choice of suitable helix angles,
depending on gearbox application
requirements. The Convoloid profile of
Figure 9 refers to the final teeth, not to
the rack profile. In order to establish the
rack cutter profile, the kinematic rela-
Figure 9 Convoloid tooth profile.

46 GEAR TECHNOLOGY | January/February 2015


[www.geartechnology.com]
tionships, e.g., of a hobbing process, have
to be employed to calculate a point-based
cutting edge definition.
S-shaped tool profile for spiral bevel
gear cutting. Stepan Lunin discussed a
Wildhaber-Novikov-style profile for a
bevel gear cutting tool in a paper pub-
lished in 2001 (Ref. 6). The profile con-
sists of radii and straight lines (Fig. 10).
The pitch line of spiral bevel gears is in all
cases of non-miter gears located towards
the pinion root (positive profile shift)
which will create profile sliding at the
location of the transition wave. Profile
sliding during the generating process will
eliminate or mutilate the transition wave,
which makes the intention of the mid-
section of the proposed profile question- Figure 10 S-shaped cutting blade profile for spiral bevel gears.
able.
The goal of using such a profile might
be the reduction of unit surface pressure
with an increase of power density. This
could be accomplished with the extended
Wildhaber-Novikov profile that would
also eliminate the complicated tool mid-
section, which is not believed to serve
any practical purpose. The paper doesn’t
mention the relationship between profile
and spiral angle. Because of the circular
profile sections, it is assumed that the
tool profile in Figure  10 requires — just
as with the Wildhaber-Novikov gears — a
particular spiral angle in order to main-
tain the correct transmission ratio during
the meshing of the mating members. This
will make the gearsets generated by the
tool profile in Figure 10 sensitive to hous-
ing tolerances and deflections.
Toroidal drive. The inventor M.R.
Kuehnle developed the idea of a compact
and high-power- density, high-reduction Figure 11 Upper-half of the toroidal “ring gear.”
planetary unit. The “heart” of the unit
is a toroid which consists of a mounted of the internal toroidal threads that ini- ponents to a high level of mechanical
upper and a lower “half-toroid” (Figure tiate a slow cage rotation (cage as out- design and manufacturability.
11 shows the upper-half of the toroid ring put shaft). The principle of planet and Kuehnle stated his toroidal transmis-
gear). The toroid has internal spherical cage rotation is shown in Figure 13. The sion will out-perform worm gear, plan-
threads that are the grooves for balls that inventive properties of the toroidal gear- etary and cycloidal transmissions for
connect the toroid with a sun gear via box are high reduction with compact medium to high ratios. Center worm
planets and planet carrier. The arrange- gearbox (high power density) and low and planet units seem straightforward in
ment between center unit (sun gear wear. manufacturing and assembly. However,
worm), planets and planet carrier motion In order to streamline the design of the spiral grooves in the split, internal
is shown in Figure 12. The sun gear unit the toroidal gearbox and verify the attri- toroidal gear components are difficult
is a multi-start, spherical worm (similar butes of its functionality with scientific to manufacture and will, post-assembly,
to a throated worm) (Ref. 7). data, the inventor consulted the Institute present a disturbance in the smooth roll-
The toroid-shaped unit is normally of Machine Elements of the Technical ing of the balls at the fitting seam.
used as the transmission housing. If the University of Aachen. The scientists and Globoidal gearing. Yakov Fleytman in
sun unit is used as an input shaft, then engineers at the institute brought all com- 1999 invented a tooth form that most-
the planets will rotate along the grooves ly applies to hypoid gears. Fleytman

January/February 2015 | GEAR TECHNOLOGY 47


technical

claims the strength of his new hypoid ment. Flank profile and tooth lead form in 2007 a bevel gear with a sinusoidally
gear design is three times that of tradi- seem different to common bevel and formed flank line with the claim of 40%
tional hypoid gears (Ref. 8). It seems that hypoid gears because no customary pro- increased load-carrying capacity (Ref. 9).
the creation of the gear tooth was done file requirements are mentioned and the The teeth of that system have some sim-
by using simple straight or twisted sur- equivalent pitch elements are not related ilarity with the “herringbone” teeth in
faces. The pinion flank surfaces might to the transmission ratio (Fig. 14). cylindrical gearing. The observer of the
have been generated with simulation soft- Globoidal gears cannot be manufac- photographs in Figure  15 notices large
ware, similar to Vericut. The gear can be tured using traditional manufacturing variations of the tooth thickness and a
defined as a tool and the axis positions methods. Prototypes have been manu- curvature inflection point at each side
of pinion and gear in a given hypoid gear factured with 5-axis machines, but no of the center. It appears that the reason
box can be used as tool and work posi- reports of evaluation results have become why the inventor anticipates a higher
tion in order to simulate the generat- public. strength could be given by the curved
ing process of the pinion flank surfaces. Cosine gears. HPG in the Netherlands, profile in the center of the face width.
Fleytman uses the kinematic coupling a company that developed 5-axis machin- The reversal with a lowering of the spiral
condition to generate the pinion teeth ing technologies for the soft and hard angle towards the ends of the teeth seems
where the gear is used as a generating ele- manufacture of bevel gears, introduced unjustified and results basically in teeth
with no spiral angle — which gives them
similar properties to ZEROL bevel gears.
Real herringbone gears are two connect-
ed helical gears with opposite helix angle
directions. The advantage is a complete
cancellation of all axial forces, combined
with a very smooth mesh characteristic
in their operation. The two opposing sec-
tions of herringbone gears are separated
by a groove in order to eliminate the sin-
gularity at the point of helix angle change.
This common herringbone gear design
takes advantage of the smooth mesh-
ing and the high strength of helical gears
without the disadvantage of rolling dis-
turbance in the transition area between
right- and left-hand sections.
It appears that the developers of the
gear in Figure 15 supposed that the sep-
aration groove of herringbone gears is
Figure 12 Worm (sun gear), planets and planet carrier cage motion. strictly the result of a manufacturing lim-
itation of the traditional manufacturing
methods and believed that the curved
flank line at the center of the face width
would increase the strength. The addi-
tional reversal of the flank line curvature

Figure 13 Toroidal, spiral motion and cage rotation. Figure 14 Globoidal, angular gear pair.

48 GEAR TECHNOLOGY | January/February 2015


[www.geartechnology.com]
towards the end of the teeth is probably The problem is not only the manu-
also only done because it wouldn’t be facturing of asymmetric spur gears. It
possible in traditional manufacturing. seems to be the fact that additional opti-
Meshing conditions of bevel gears mizations — like circular top relief — have
without a spiral angle are best with to be applied in order to make the
straight bevel gears and deteriorate with improvement visible. Fuentes developed
ZEROL bevel gears. The smooth screw- first an optimal top relief for a baseline
ing into mesh like in spiral bevel gearing design and then converted the gearset
is not possible with the cosine teeth in to asymmetric profile. However, he cre-
Figure 15. The face contact ratio is slight- ated comparable gearsets with symmetric
ly above zero and will not contribute to and asymmetric profile and was able to
improved strength. The cosine-shaped achieve contact stress reductions in the
tooth form prevents elastic deformation vicinity of 10%.
under load, which in spiral bevel gears is Several bevel gear manufacturers
used to achieve a smooth meshing and to accepted the publications’ findings and
increase the load-sharing between con- research in asymmetric spur gears and
secutive tooth pairs. Hartmuth Müller applied moderate amounts (up to 4°) of
published theoretical investigation results asymmetric pressure angles to their tra-
of bevel gears with sinusoidally shaped ditionally symmetric spiral bevel gears
flank lines and concluded that their in order to improve the load-carry-
rolling performance — as well as their ing capacity of mostly unidirectional-
strength properties — is far below any used angular transmissions. The posi-
average spiral bevel gear set (Ref. 10). tive experiences with spiral bevel gears
Asymmetric tooth gears. The two flank allow the conclusion that, not only spur
profiles of involute gearing are generally gears, but also helical gears would benefit
developed from a common base circle. from asymmetric profiles. Because of the
In this case both flank profiles have the individually ground cutting blade pro-
same pressure angle and are mirror imag- files, this technology seems to gain more
es from each other. With the develop- acceptance in bevel gearing. For cylindri-
ment of hypoid gears, Ernest Wildhaber cal gears the introduction of asymmetric
in 1930 developed the asymmetric tooth profiles would require a departure from
Figure 15 Ring gear and pinion with cosine
profile (Ref. 11). Wildhaber’s develop- standard hobs and may require additional teeth (Source: hpg-nl.com).
ment eliminated the conflict that the path profile corrections in order to work out
of engagement of the drive-side flanks the full advantages of asymmetric gear-
and the coast-side flanks was not equal ing, which seems to be a large step and
if hypoid gears were manufactured with thus precludes ready acceptance in the
equal pressure angles. industry.
Today, asymmetric spur gear profiles
that have been proposed for more than
20 years are discussed and applied to
practical applications. The goal is not to
extend the angle of engagement, as it was
for Wildhaber, on the drive-side of hyp-
oid gears. Extending the active line of
engagement would require a reduction of
the pressure angle. Alfonso Fuentes et al.
(Ref. 11) proved in an analytic compari-
son that especially flank surface stress is
reduced on the driving side by increas-
ing the pressure angle. A tooth mesh with
asymmetric pressure angles is shown in
Figure 16. Alexander Kapelevich promot-
ed asymmetric cylindrical gears for many
years (Ref. 12) and wrote calculation soft-
ware that allows design calculations for
optimized geometries.

Figure 16 Asymmetric tooth engagement.

January/February 2015 | GEAR TECHNOLOGY 49


technical

Summary to the properties of the “half-round” pro- performance advantages resulted in this
The tooth shapes discussed here can be file of Figure 4. case in load-carrying capacity similar to
sorted into those with interesting physical The Nagata development is very simi- straight bevel gears, and a rolling perfor-
properties and those without sufficient- lar to the extended Wildhaber-Novikov, mance that cannot compare to standard
ly proven functionality. It is quite true but requires rather more complicated tool spiral bevel gears.
that validating a new system is difficult geometry. The “S-shaped” profile makes Asymmetric gears build upon the
if the technical community rejects the it an additional version of the Cycloidal strength of involute gearing, with the
idea based on solely subjective reasons. gear type with similar physical proper- acknowledgement that in case of a pre-
This would prevent any private or public ties. Center distance sensibility and diffi- ferred driving direction, the properties
research funding. However, many of the cult manufacturability also prevented the of the driving flanks can be improved
new tooth profile solutions discussed in introduction of this gear type in power by taking away from the non-driving,
this paper are based on brilliant ideas that transmissions. or coasting, flanks. The practical results
failed to have a breakthrough for similar Convoloid gears, with their dedendum of transmissions with asymmetric gears
reasons, as why, for example, the Wankel transition point, which is a “flat spot” show a significant increase of power den-
engine never replaced the stroking piston in a finite profile section, cannot accept sity due to an improvement of root bend-
engine. significant center distance changes. The ing strength and higher surface durabil-
Cycloidal gears have been used for flat section of the two mating profiles ity. Investigations showed that the full
watches and clocks. Even though they must be located precisely during the roll- advantage of asymmetric gearing vs.
are center distance-sensitive, the watch ing process to separate addendum and gears with symmetric involutes can only
designer allows rather large center toler- dedendum rolling. The separation may be realized if sensible tip relief modifica-
ances in connection with axle bearing interrupt the hydrodynamic lubrication, tions are applied. This in turn creates a
play. The “S-shaped” profiles of the mat- and misalignments may cause rolling problem for many manufacturers — even
ing cycloidal gears develop a self-center- of the transition flat into addendum or those that mass-produce, say, automotive
ing force that forces the gears to roll with dedendum, which will create roll inter- transmissions. The wealth of experience
correct center distance. The advantage of ference. The transition flat prevents or that has been compiled over decades that
this design was the manufacturing cost complicates the profile generation with found its way into the international stan-
reduction due to large bearing position a reference profile on a tool like a hob. dards, as well as the material application
tolerances and large axle play. The large Convoloid profiles seem to present the tables, has no or only limited use for the
bearing play in turn would guarantee disadvantages of cycloidal profiles, with dimensioning and design of asymmetric
low friction in the sleeve bearings, which increased vulnerability around the pitch gears. Also, the cutting tools are now not
achieves one important objective of chro- line. standard tools anymore. Different pres-
nometers. Manufacturing of the cycloidal The “S-shaped” bevel gear profile is sure angle offsets for different applica-
requires specially formed tool profiles, or very similar to the Convoloid profile. tions and gear design parameters in con-
pointed tools, that move along a cycloidal However, where the Convoloid profile nection with circular tip relieves of differ-
path. The advantages cycloidal gears have is the result of an analytical mesh and ent amounts will not only make existing
in watches cannot be duplicated in indus- contact area optimization, the singular- tooling obsolete but also eliminate today’s
trial power transmissions. ity of the “S-shaped” bevel gear profile in standards in hobs and even shaper cut-
Wildhaber-Novikov gears have proven mid-profile will not permit relative pro- ters which will result in increased cutting
their increased load-carrying capacity, yet file sliding in this section; this makes this tool cost and contribute to longer tool-
show significant center distance sensitiv- proposed profile unusable in practical ing lead times. The fact that asymmetric
ity, are difficult to manufacture, require applications. gears have not had a breakthrough yet is
non-standard tools and depend on a spe- Globoidal gears use the kinematic cou- the result of those obstacles. However, the
cific, pre-determined helix angle in order pling conditions of formate gears. The symmetry offset, as well as tip reliefs or
to transmit a constant ratio. Changes in profile of the gear member is basical- other corrections, could be standardized,
center distance and axes inclinations due ly chosen to be straight and the pinion depending on module and application,
to tolerances and deflections will intro- member is generated with respect to the which would remove the major obstacle
duce significant motion errors and trans- desired ratio, shaft angle and center dis- for the broad application of asymmetric
mission vibrations, which would increase tance. The principle of interacting pitch involute gears.
the cost for shafts, bearings and/or hous- elements is not applied, which leads to
ing in order to reduce tolerances and high relative sliding and can cause partial Conclusions
deflections. Even then, the robustness mutilation to the flank surface. The motivation to change from involute
and “mechanical intelligence” of involute Cosine gears have been discussed to profiles and proven straight or curved
gearing that provides them with smooth also elaborate on flank line ideas which flank lines to alternative shapes is fueled
rolling through many million cycles do not conform to the common mathe- by the logic that, in today’s time and age,
could not be achieved. The extended ver- matical flank line functions. Cosine gears more than just incremental improve-
sion of the Wildhaber-Novikov profile might have the ability to hold higher ments should be possible, applying new
has an “S-shape” — just like the Cycloidal torques in a non-rotating, static condi- theories and sophisticated computation
profile that adds the self-centering ability tion. The attempt to design a gearset with technology. Those dramatic improve-

50 GEAR TECHNOLOGY | January/February 2015


[www.geartechnology.com]
ments would consequently lead to Nevertheless, it seems fair to say: “The
geometries that would appear exotic if involute is here to stay.”
observed with the traditional viewpoint.
The objective of this paper was to observe References
the published alternative tooth forms 1. Stadtfeld, H.J. Gleason Bevel Gear Technology
—The Science of Gear Engineering and mod-
with an open mind in order to find objec- ern Manufacturing Methods for Angular
tive answers to the question of why the Transmissions, Gleason Publication, ISBN 978-0-
interesting and sophisticated tooth form 615-96492-8, 2014.
proposals haven’t had a breakthrough in 2. Wildhaber, E. “Gears with Circular Tooth Profile
Similar to the Novikov System,” VDI Berichte,
the gear manufacturing industry. No. 47, 1961, Germany.
In angular gear drives, the absence of 3. Novikov, M.L. USSR Patent No. 109,750, 1956.
tight standards, as they exist for cylindri- 4. Nagata, S. And Y Ariga. “Development of a New
cal gears, led to many gear types with dif- Wildhaber-Novikov Gear with a Basic Rack
of Combined Circular and Involute Profile, ”
ferent profile and lead functions. The fact International Symposium on Gearing & Power
that bevel gear cutting tools have never Transmission, Tokyo, 1981.
been standardized opened the door for 5. Berlinger, B.E. and J.R. Colbourne. “Convoloid
this variety of different processes and Gearing Technology,” Gear Solutions, Media
tooth forms. The downside is that bevel Solutions, INC Publishing, February 2012. Dr. Hermann J. Stadtfeld in 1978
6. Lunin, S. “New Methods of Gear Geometry received his B.S. and in 1982 his M.S. in
gear design and manufacturing is per- Calculation,” the JSME International Conference mechanical engineering at the Technical
ceived as complicated, not straightfor- on Motion and Power Transmission, The Japan University in Aachen, Germany; upon receiving
ward — and expensive. The standards Society of Mechanical Engineering, Fukuoka, his Doctorate, he remained as a research
Nov. 15, 2001. scientist at the University’s Machine Tool
in cylindrical gears, which are based on 7. Troeder, C. “Toroid-Getriebe Laboratory. In 1987, he accepted the position
standard reference profiles and involutes, Forschungsvereinigung Antriebstechnik,” FVA of head of engineering and R&D of the Bevel
make design and manufacturing trans- Information Congress, Karlsruhe, November Gear Machine Tool Division of Oerlikon Buehrle
parent and keep the cost of cylindrical 1975. AG in Zurich and, in 1992, returned to academia
8. Fleytman, Y. Worm Gear Transmission United as visiting professor at the Rochester Institute
gear manufacturing relatively low. States Patent No. 6,148,683, November 2000. of Technology. Dr. Stadtfeld returned to the
The center distance insensitivity of 9. Schlossig H.P. Auf einfachem Weg zu guten commercial workplace in 1994 — joining The
involute gears also applies in the rela- Zähnen Werkstatt und Betrieb, Carl Hanser Gleason Works — also in Rochester — first
Publishing, Munich, April 2007. as director of R&D, and, in 1996, as vice
tionship between generating rack and president R&D. During a three-year hiatus
10. Mueller H. “Function Oriented 5-Axes
work gear and therefore makes hob- Machining of Bevel Gears” Seminar (2002-2005) from Gleason, he established a
bing a very robust process. The fact that “Innovations in Bevel Gear Technology,” WZL, gear research company in Germany while
every point along the involute has a nor- Technical University Aachen, March 2012. simultaneously accepting a professorship
11. Fuentes, A. “On the Behaviour of Asymmetric to teach gear technology courses at the
mal direction that is tangent to the base University of Ilmenau. Stadtfeld subsequently
Cylindrical Gears in Gear Transmissions” SAE-
circle gives the involute its robustness China and FISITA (eds., Proceedings of the returned to the Gleason Corporation in 2005,
during manufacturing and in operation where he currently holds the position of vice
FISTA 2012 World Automotive Congress, DOI:
president, bevel gear technology and R&D. A
(Fig. 17) (Ref. 13). All of the discussed 10.1007/978-3-642-33744-4_13, Springer-
prolific author (and frequent contributor to Gear
non-involute gears are sensitive to cen- Verlag, Berlin, Heidelberg 2013.
Technology), Dr. Stadtfeld has published more
12. Kappelevich, A. “Asymmetric Gears:Parameter
ter distance changes and require indi- Selection Approach,” Gear Technology, June/July
than 200 technical papers and 10 books on
vidual tool designs. Standardization is bevel gear technology; he also controls more
2012, Pages 48-51, Randall Publishing Inc., Elk
than 50 international patents on gear design,
not possible or difficult, which presents Grove Village, Illinois.
gear process, tools and machinery.
an additional risk in design and manu- 13. Stadtfeld, H.J. “Operating Pressure Angle,” Gear
Technology, May 2013, Pages 57–58, Randall Upon high school graduation Jasmin K.
facturing. If higher load-carrying capac- Publishing Inc., Elk Grove Village, Illinois. Saewe, from April to June 2011, completed
ity, lower noise and increased efficiency basic training in metalworking during
are goals of new gear geometries, then the her internships with German-based TRW
Automotive and Aleris Rolled Products.
absence of standards and higher design
In October 2011 she enrolled as a student
and manufacturing cost might be accept- of mechanical engineering at the RWTH
able in some cases. However, the physical Aachen University. At the Aachen University
properties of involute gearing are superi- Jasmin became a student research assistant
at the WZL, where she was involved with
or to most of the discussed non-involute practical gear cutting trials and tool life
gears. The potential of asymmetric, invo- investigations;this spurred her interest in an
lute tooth profile will allow significantly internship with Gleason. Saewe in April 2014
improved cylindrical gears by maintain- began an internship in the R&D department at
The Gleason Works in Rochester, New York. As
ing the advantages of involutes. A broad- of January 2015, Jasmin continues her studies
er interest on asymmetric gears could ini- at the RWTH-Aachen University in order to
tiate the development of new standards finish her Bachelor’s degree, while continuing
her employment as student research assistant
and would — over time — allow the gear at the WZL.
community to gain sufficient experience
in this advanced system.

January/February 2015 | GEAR TECHNOLOGY 51

You might also like