Engineering Science and Technology, an International Journal 59 (2024) 101849

Contents lists available at ScienceDirect

Engineering Science and Technology,

an International Journal
journal homepage: www.elsevier.com/locate/jestch

Full Length Article

Research on the top cutting features on face gears using normal to tooth
profile method
Wei Sheng a, Zhengminqing Li a,b,* , Xiaofeng Yu a
a
College of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
b
State Key Laboratory of Helicopter Aeromechanics, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China

A R T I C L E I N F O A B S T R A C T

Keywords: In the field of face gear tooth profile research, while the phenomena of undercutting and pointing are widely
Face gear recognized, the occurrence of tooth top overcut is often overlooked. This paper employs the Normal to Tooth
Normal to tooth profile method (NTPM) Profile Method (NTPM) to construct a mathematical model for the face gear tooth surface, encompassing the
Top cutting
working surface, fillet surface, and top cutting surface. The tooth top overcut phenomenon is elucidated by
Chamfering
means of tooth surface visualization and simulated gear cutting. The reliability of the proposed model is also
compared with traditional models through the use of finite element simulations and tooth contact experiments.
The study also analyzes the influence of design parameters on the amount of overcut and provides chamfering
recommendations to eliminate it. The findings reveal that tooth top cutting is particularly significant in face gear
pairs with high gear ratios, necessitating special attention when designing power-split transmission systems.
Furthermore, the proposed NTPM is a more effective method for examining face gear tooth profile
characteristics.

1. Introduction also found that correcting the small gear side pressure angle can avoid
contact between the inner and outer edges of the end face gear teeth. Wu
Gears are an indispensable component in the field of machinery, et al. [9] put forward a strategy of uniformly arranging points on the
which is closely linked to the advancement of human science and tooth surface of face gears, which facilitated the establishment of a stress
technology. Face gear transmission is highly favored in aviation appli- analysis method for face gear generation design based on CAD/CAE
cations due to its exceptional transmission performance, particularly in integration. The tooth surface morphology created by the face gear
power-split transmission systems [1,2]. Currently, tooth profile char- grinding process was predicted by Cai et al. [10]. Chang et al. [11]
acteristics of face gear such as undercutting and pointing have been explained the changes in contact paths and transmission errors caused
acknowledged by the majority of scholars, while top cutting in gear by assembly errors. Tsay and Fong [12] suggested a gear tooth modifi-
teeth is frequently disregarded. Top cutting compromises the integrity of cation method that can control the meshing trajectory of helical gear
gear teeth and thus requires adequate attention. pairs. Lu et al. [13] established a two-variable optimization model and a
The face gear transmission technology has a strong theoretical search algorithm for the purpose of calculating contact trajectories,
foundation in practical engineering thanks to the work of Litvin et al. which solved the high nonlinearity problem of the general meshing
[1,3–5] who conducted extensive research on the principle of generating equation of face gear transmission. Andersson [14] disclosed a contact
tooth surfaces of face gears in face gear research, derived the tooth ratio calculation method for avoiding tooth separation near a specific
surface equations of face gears and cylindrical gears, and analyzed the critical rotational speed. Based on the recognized standard for cylin-
conditions for undercutting and pointing [6]. Zschippang et al. [7] drical gear contact stress calculation, Hochrein et al. [15] developed two
questioned the applicability of the general formula for calculating the contact stress calculation methods for face gear transmission. Liu et al.
inner radius of face gears since actual manufacturing tools are not [16,17] studied the calculation methods for the centerline contact ratio
infinitely long. In addition to proposing a simple method for deter- and point contact ratio during the formation process of face gears and
mining the minimum and maximum radii of face gears, Feng et al. [8] provided a meshing stiffness model. Based on TCA and LTCA methods,

* Corresponding author at: College of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China.
E-mail address: lzmq_cmee@nuaa.edu.cn (Z. Li).

https://doi.org/10.1016/j.jestch.2024.101849
Received 18 May 2024; Received in revised form 4 September 2024; Accepted 24 September 2024
Available online 10 October 2024
2215-0986/© 2024 The Author(s). Published by Elsevier B.V. on behalf of Karabuk University. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Fig. 1. The principle of generating face gear tooth surface: (a) Coordinate system for generating face gears; (b) Shaft section tooth profile of shaper.

Zhou et al. [18] found that small wheel tapering can not only adjust the contact imprint tests were conducted on a rolling test rig to validate the
meshing clearance of face gear transmission but also alleviate tooth edge effectiveness of the improved model. Finally, a study was performed on
contact. A model for predicting friction loss in offset face gears was tooth profile parameters related to top cutting, and rational recom-
developed by Zschippang et al. [19,20], which is based on the relative mendations for avoiding tooth top overcutting were proposed.
sliding relationship between tooth surfaces. Zhu and Dai [21] proposed
a model for predicting the efficiency loss of face gear pairs under splash 2. Mathematical model of tooth surface
lubrication conditions. Zhu et al. [22] investigated the dynamic char-
acteristics of polymorphic meshing, including side tooth meshing, side 2.1. Coordinate system of the model
tooth disengagement meshing, side tooth meshing, and alternating
meshing between single and double tooth pairs. Peng and Zhou [23] The process of machining face gears typically involves the slotting
suggested using unloading characteristic control methods such as method, which operates in accordance with the principle of the enve-
fourth-order transmission error, contact ellipse length, and contact path lope. Fig. 1 (a) depicts the static coordinate systems for the shaper and
direction in order to achieve active modification of face gear pairs as face gear, denoted as Ss0(os0, xs0, ys0, zs0) and Sf0(of0, xf0, yf0, zf0)
well as vibration reduction effect. respectively. Their fixed linkage coordinate systems, Ss(os, xs, ys, zs) and
In terms of tooth surface shaping, face gears can be made using a Sf(of, xf, yf, zf), coincide with the origin of all four coordinate systems.
variety of techniques, which consist of grinding with a disc wheel During machining, the shaper rotates at an angle fs around the zs0-axis,
[24,25], multi-step grinding using a worm [26], milling with a disc while simultaneously, the face gear rotates at an angle ϕf around the zf0-
cutter [27], grinding with a grinding tool [28], shaving method [29], axis. These angles ϕs and ϕf satisfy the gear ratio relationship: ϕs /ϕf = ifs
CNC radial milling [30,31], rapid prototyping with a hob [32], elec- = 1/ isf, where ifs represents the gear ratio.
trochemical machining [33], and so on. Tooth surface accuracy testing is As illustrated in Fig. 1 (a), the process of producing face gears
essential for evaluating machining quality. Wang et al. [34] established through tool gear cutting entails transforming coordinate points on the
the relationship between the measurement coordinate system and the tooth surface of the cutting tool gears, according to the meshing rela-
design coordinate system. They then applied B-spline fitting technology tionship, into the face gear coordinate system to generate its corre-
to reconstruct the measurement model so as to obtain an accurate sponding tooth surface. This transformation involves converting the
measurement model. Tao et al. [35] established a measurement error vector coordinate system from Ss to Sf, and the transformation matrix Mfs
compensation model to adapt to the detection of gear tooth surface for this coordinate conversion can be defined as follows:
accuracy on coordinate measuring machines. In addition to the study of
traditional tooth profiles, various functional tooth profile designs have Description of symbols
also emerged. Fu et al. [36] studied the geometric characteristics of an
offset, non-orthogonal, and profile-shifted face gear. Mo et al. [37] ai (i = 1, coefficient of the ifs, isf gear ratio and its reciprocal
…, n) meshing path function
analyzed non-orthogonal asymmetric helical face gears with good tooth
width characteristics. A novel face gear transmission was proposed by Ai (i = 1, tooth top boundary IAi (i = position of the face gear teeth
…, n) point 1, …, n) in the tooth height direction
Mu et al. [38], which consists of a conical enveloped cylindrical worm
α0, αi pressure angles at the IAa, IAd tooth top and tooth root of face
and a face gear. Liu and Cai [39,40] proposed a gear group consisting of dividing circle and the gear
planar non-circular gears and specific face gears that can transmit var- pitch circle
iable angular velocity. Sheng et al. [41–43] presented a design method Bi (i = 1, tooth root boundary IAs instantaneous rotation axis
…, n) point
for low sliding face gear pairs based on the correlation between sliding
Bf face gear tooth width φ1 shaper rotation angle of
ratio and meshing path. meshing point relative to pitch
Numerous scholars have studied the principles of face gear tooth point
generation, transmission performance, processing, and inspection c* clearance coefficient κ center angle corresponding to
methods, achieving three-dimensional modeling of tooth surfaces based half of gear teeth
D1, D2 chamfer parameters Li position parameters in the
on a general modeling approach. Nonetheless, there is a lack of aware-
width direction of face gear
ness of the phenomenon of top cutting in the reliability research of the tooth
3D model. This paper reveals the possibility of top cutting at the inner Di, Do, Dt inner, outer radius, and m gear module
end of gear teeth by proposing a tooth profile-generating approach top cutting radius of face
called NTPM. This approach combines envelope simulation and finite gear
Dtc top cutting tooth width, Mfs transformation matrix of the
element analysis to build the mathematical model of face gears. Subse- and Dtc = Dt – Di coordinate system Ss to Sf
quently, physical models of both the traditional and improved models
(continued on next page)
were obtained through numerical control machining, and tooth surface

2
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Description of symbols (continued ) ⎛ ⎞

− ys + ifs zs cosϕs
ai (i = 1, coefficient of the ifs, isf gear ratio and its reciprocal ⎜ xs − ifs zs sinϕs ⎟
…, n) meshing path function v(s,f)
s = ωs ⎜ ⎟
⎝ ifs (ys sinϕs − xs cosϕs ) ⎠ (2)
Es equation of involute n(w)
s , n(f)
s normal vector of the involute 1
profile based on work and fillet tooth surfaces
traditional method
Es(w), equation of involute N s , N p , Nf number of teeth for shaper,
where ωs denotes the angular velocity of the shaper, and (xs, ys, zs)
Es(f) gear tooth profile based pinion, and face gear represents the coordinate of the shaper tooth surface.
on NTPM
Ef(w), equation of face gear θ0 , θs spreading angle parameter of 2.2. Traditional mathematical model
Ef(f), tooth profile based on the involute profile
Ef(t) NTPM
Σ contact path between Pi (i = 1, pitch point Fig. 1 (b) displays the involute profile of the shaper. In general, the
rack and gear …, n) radial vector of the involute profile in coordinate system Ss can be
Σi (i = 1, contact path between rs, rb, ra radius of base circle, dividing expressed as:
…,n) shaper and face gear circle and addendum circle of ⎛ ⎞
shaper ±rb [sin(θs + θ0 ) − θs cos(θs + θ0 )]
Σ L, Σ R shaper tooth surfaces on Rsi (i = 1, radius of pitch circle ⎜ rb [cos(θs + θ0 ) + θs sin(θs + θ0 )] ⎟
both sides …, n) Es = ⎜ ⎟ (3)
⎝ hs ⎠
fs(w), fs(f) meshing equation Ss0, Ss static and linage coordinate 1
between the shaper and systems of shaper
the face gear
ϕs, ϕp, ϕf rotation angle of shaper, Sf0, Sf static and linage coordinate where θ0 indicates the spreading angle of the involute profile, θs denotes
pinion and face gear systems of face gear the angle between the involute profile and the base circle, rb reflects the
γ, γ1, γ2 axial angle vs(s, f) relative velocity between base circle radius, hs shows the tooth width of the shaper, and “±”
shaper and face gear represent the tooth surfaces Σ L and Σ R.
ha * addendum coefficient ωs angular velocity of shaper
hs range of shaper tooth xa, xb parameters of the shaper tooth
Furthermore, the meshing relationship between the shaper and the
width top and root face gear can be expressed as
H addendum height of xt parameters of face gear tooth
rack root f (w)
s = n(w) (s,f)
s ⋅vs = ifs hs cos(θs + θ0 − ϕs ) − rb = 0 (4)
ηi phase difference NTPM normal to tooth profile method
where ns(w) represents the normal vector of the involute tooth working
⎛ ⎞
cosϕs cosϕf − sinϕs cosϕf sinϕf 0 surface.
⎜ − cosϕs sinϕf sinϕs sinϕf cosϕf 0⎟ According to the meshing relationship and coordinate trans-
Mfs = ⎝
⎜ ⎟ (1)
− sinϕs − cosϕs 0 0⎠ formation relationship, the equations for the working and fillet tooth
0 0 0 1 surface of the face gear can be obtained as follows:
⎧
The motion relationship between the shaper and the face gear can be ⎨ Ef = Mfs ⋅Es
(w)

described using the relative velocity vs(s, f), which is expressed in the (5)
⎩ f = n(w) ⋅v(s,f) = 0
(w)
coordinate system Ss as: s s s

Fig. 2. The proposed mathematical modeling of face gear pair teeth: (a) Coordinate systems of rack and pinion; (b) Meshing relationship between gears and racks; (c)
Instantaneous rotation axis; (d) Pressure angle at different sections; (e) Critical meshing points; (f) Phase angle of tooth profiles.

3
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

⎧
found in Appendix B.
⎨ Ef = Mfs ⋅Es
(f)
⎪
⎪
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅2 (6)
⎪
⎪ r2a − rb 2.3.2. Face gear model
⎩ θs =
rb When the face gear tooth surface is generated, the rotational move-
ment between the shaper and the gear being generated is transmitted
where ra indicates the addendum circle radius. between two intersecting axes. The angle between these two axes, zs0
The Eqs. (5) and (6) outlined above might be considered somewhat and zf0, is denoted as γ, as displayed in Fig. 2 (c). The meshing line of the
complex and can be explained from three aspects. Firstly, the clarity in angular velocity, ωsf, during the relative motion between the shaper and
determining the limiting parameter θs in the tooth height direction for face gear is the instantaneous rotation axis IAs. The direction of IAs
the face gear tooth surface equation is lacking, and its correlation with relative to the rotation axis of the two gears is determined by the angles
the rotation angle ϕs involves complex trigonometric functions, making γ 1 and γ2, which are represented as follows:
it inconvenient to solve. Secondly, there are established calculation ⎧
methods exist for determining the undercutting and pointing to limit the sinγ
⎨ tanγ 1 = ifs + cosγ
⎪
⎪
⎪
tooth width parameter hs, but it requires solving a system of equations.
(11)
Thirdly, the range of angle fs corresponding to different parameters hs is
⎪ tanγ = sinγ
⎪
⎪
also unclear. While three-dimensional working and fillet tooth surfaces ⎩ 2
isf + cosγ
have been achieved under the conditions of clearly defined radii to
eliminate undercut as well as pointing, the top cutting at the inner end of here γ=90◦ is studied in this paper.
the tooth has been overlooked. In order to address these complicated In Fig. 2 (c), within the effective tooth width range, all points (such
issues, this paper introduces a novel model for tooth surfaces using as P0, P1, P2) on the instantaneous rotation axis are referred to as pitch
NTPM. This model is designed to streamline the description of param- point, wherein the relative motion at these points is pure rolling, while
eter value ranges and their interrelations, which offers a solution to the meshing points in other areas of the tooth surface entail sliding-
these challenges. rolling motion. The meshing path equation between the shaper section
and the face gear corresponding to each point can be expressed as:
2.3. Improved mathematical model using NTPM
y = − xtanαi (12)
2.3.1. Shaper model
here αi represents the pressure angle at each pitch circle, as shown in
For involute gears, the fundamental machining method involves
Fig. 2 (d). The relationship between the pitch circle radius Rsi of the
rack-type tool machining, as depicted in Fig. 2(a). The meshing rela-
shaper and the pressure angle can be expressed as:
tionship between the teeth can be found in Fig. 2(b).
The equation of the meshing line Σ between the involute gear and Rsi cosαi = rb (13)
rack as follows:
On the basis of the pure rolling characteristics at the pitch point, the
y = − xtanα0 (7) axial position (such as L0, L1, L2) of the cross-section of the shaper can
also be determined, as shown below:
here α0 denotes the pressure angle at the dividing circle.
According to the NTPM described in Appendix A, the work surface Li = ifs Rsi (14)
equation of the shaper in Fig. 2(b) can be calculated as follows:
⎛ ⎞ Using the NTPM and coordinate transformation, the equations for the
∓[xcos(φ1 + κ) + (y + rs )sin(φ1 + κ)] working and fillet tooth surface of the face gear can be derived as fol-
⎜ − xsin(φ1 + κ) + (y + rs )cos(φ1 + κ) ⎟ lows:
E(w) =⎜ ⎟ (8)
s ⎝ hs ⎠
⎛ ⎞
1 ∓(xcosψ + Li sinψ )
⎜ − xsinψ + Li cosψ ⎟
E(w) =⎜ ⎟ (15)
where φ1 represents the shaper rotation angle of the meshing point f ⎝ xtanαi − Rsi ⎠
relative to the pitch point, κindicates the center angle of the circle cor- 1
responding to half of the gear teeth, and rs reflects the dividing circle ⎛ ⎞
radius. ∓[xb cosφs + (Rsi − xb tanαi )sinφs ]cosψ + Li sinψ
⎜ − [xb cosφs + (Rsi − xb tanαi )sinφs ]sinψ + Li cosψ ⎟
⎧
− (1 + tan2 α0 )x E(f) =⎜ ⎟ (16)
⎪
⎪
f ⎝ xb cosφs + (Rsi − xb tanαi )sinφs ⎠
⎨ φ1 =
rs 1
(9)
⎪
⎪ π
⎩ κ= with
2Ns
ψ = ifs (ϕs + η + κ) (17)
here Ns denotes the gear tooth number.
Similarly, the fillet surface equation of the shaper can be derived as where ϕs and φs are expressed in Eqs. (18) and (19), phase difference ηi
follows can be found in Appendix B.
⎛ ⎞
∓[xcos(φ1 + κ) + (rs − H)sin(φ1 + κ)] − (tan2 αi + 1)x
⎜ − xsin(φ1 + κ) + (rs − H)cos(φ1 + κ) ⎟ ϕs = (18)
E(f) ⎜ ⎟ (10) Rsi
s = ⎝ Li ⎠
1 (tan2 αi + 1)(x − xb )
φs = (19)
Rsi
where H represents the addendum height of rack.
Compared to Eq. (3), the utilization of Eqs. (8) and (10) to depict the The shaper fillet surface is machined to obtain the top cutting surface on
tooth surface model offers several advantages. Not only does it provide a the radial inner side of the face gear. Utilizing the meshing relationship
clearer representation of the working and the fillet tooth profile, but it equation and the fillet profile equation, the equation for the top cutting
also reveals the relationship between the rotation angle φ1 and the in- surface can be derived as follows:
dependent variable x. The value of the independent variable x can be

4
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Table 1 equation also provides tooth width parameters of the face gear, such as
Design Parameters of Gear Pairs. undercutting, pointing and top cutting, the details of which are given in
Parameters name Values Appendix C.
Utilizing tooth surface Eqs. (15), (16) and (20), MATLAB software
Modulus m /(mm) 2.5
Shaper tooth number Ns 25 was employed to conduct a three-dimensional visualization analysis of
Pinion tooth number Np 23 the tooth surface. The gear parameters are listed in Table 1, and the
Face gear tooth number Nf 100 corresponding results are shown in Fig. 3. There are two dividing lines:
Pressure angle α0 / (◦ ) 20 the boundary Lwf separates the working and fillet surfaces, and the
Addendum coefficient ha* 1
Clearance coefficient c* 0.25
boundary Lwt separates the working and top cutting surfaces, which is
the focus of this paper. It is evident that the top cutting phenomenon
occurs at the inner tooth top, and is obtained from the shape of the tool
⎧
E(t) fillet tooth surface. This previously unmentioned phenomenon, which is
f = Mfs Es
⎨ (f )
(20) crucial when studying inner gear tooth shapes, requires consideration of
⎩ f (f ) = n(f) ⋅v(s,f) = 0
s s s top cutting interference.

where ns(f) represents the normal vector of involute tooth fillet surface.
The independent variable x in the working tooth surface equation
can be analyzed in conjunction with Fig. 2 (f). Appendix B is the refer-
ence for the tooth top boundary point (such as A0, A1, A2) and the tooth
root boundary point (such as B0, B1, B2). Moreover, the tooth surface

Fig. 3. Visualization of face gear full tooth surface.

Fig. 4. The process of gear cutting simulation.

5
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Fig. 5. Finite element analysis of gear tooth contact: (a) Finite element analysis model of face gear pair; (b) Result of traditional face gear model; (c) Result of
proposed face gear model.

3. Simulation physics model of tooth surface accordance with the transmission ratio with the cylindrical gear. In
order to visually demonstrate interference that occurred during the
3.1. Tooth surface forming simulation cutting process of face gears, the cylindrical gear is initially modeled in
zones, as depicted in Fig. 4. The working tooth surface is shown by the
CAD software simulation of tooth cutting facilitates a rapid under- red area, the fillet surface by the blue area, the tooth top surface by the
standing of the tooth surface shape [44]. CATIA software is used for the black area, and the tooth root surface by the yellow area. Additionally,
cutting simulation of face gears, with the simulation parameters derived the overall color of the gear blank on the opposite side is marked,
from Table 1. The inner as well as outer radii of a face gear are deter- differing from that of cylindrical gears and adopting a silver-gray hue.
mined by specific formulas. In the coordinate system, the positioning of Upon completing the model establishment, the cylindrical gear and face
the cylindrical gear tooth width is smaller on one side than the inner gear blanks are named in the face gear model file, along with naming the
radius of the face gear and larger on the other side in comparison to the corresponding axes. Finally, a macro program that predominantly uti-
outer radius of the face gear. Based primarily on the gear-shaping lizes loop statements and Boolean operations is written for automatic
principle of face gears, the key difference lies in replacing the original simulation of gear cutting. The step size of the movement angle of the
bowl-shaped gear-shaping cutter with a cylindrical gear. The cylindrical cylindrical gear is adjusted according to the program operation status,
gear rotates solely around the axial direction without any axial feed with a given value of 1.2 degrees. As a result, the face gear’s step size is
motion, whereas the face gear blank rotates around its own axis in set at 0.3 degrees. Upon running the macro program, the gear-cutting

Fig. 6. Gear surface rolling inspection test: (a) Design of face gear blank. (b) Tooth shape machining. (c) Testing. (d) Meshing marks.

6
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Table 2 surface at the top of the inner tooth of the face gear is cut from the fillet
Parameters of experimental gear pairs. tooth surface of the cylindrical gear. This confirms the occurrence of
Parameters name Values overcutting on the top surface of the inner end teeth of the face gear.
Modulus m /(mm) 4
Shaper tooth number Ns 27
3.2. FEM Validation of mathematical models
Pinion tooth number Np 25
Face gear tooth number Nf 81
Pressure angle α0 / (◦ ) 20 To compare the validity of the models in Sections 2.2 and 2.3, finite
Addendum coefficient ha* 1 element analysis of the static meshing process of face gear pairs was
Clearance coefficient c* 0.25 carried out through the use of finite element software. The geometric
parameters of the gear pairs are derived from Table 1. After assembling
process commences, and upon completion, the resultant gear surface is the two gears into a fixed coordinate system in accordance with the
displayed in the figure. The resulting tooth surface of the face gear discrete angles ϕp and ϕf, the model was imported into ABAQUS soft-
makes it clear that the color marking of the tooth root surface of the face ware. Fig. 5(a) specifies the internal loads and boundary conditions.
gear aligns with the tooth top plane of the cylindrical gear, while the Points of and op are located on the rotation axes of the face gear and the
working tooth surface and transition surface of the face gear correspond pinion, respectively. The material properties of the pinion and the face
to the working surface of the cylindrical gear. Notably, the red marked gear throughout the preprocessing stage are the same: Young’s modulus
of 206 GPa and Poisson’s ratio of 0.3. When setting up the contact

Fig. 7. The impact of a single design parameter on top cutting width: (a) Modulus; (b) Pressure angle; (c) Addendum coefficient; (d) Head clearance coefficient.

7
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Fig. 8. The influence of design parameter on critical value for top cutting: (a) Modulus; (b) Pressure angle; (c) Addendum coefficient; (d) Head clearance coefficient.

surfaces, frictionless contact is selected, with the face gear tooth surface expected. However, due to interference between the pinion’s tooth root
designated as the slave surface and the pinion tooth surface as the and the face gear’s tooth top at the inner end, the gears in area I fail to
master surface. Boundary conditions and loads are set up through the mesh correctly, with distinct markings at the interference points. In
following steps: all degrees of freedom of the spur gear are fixed, the addition to indicating certain deficiencies in traditional theoretical
rotation degrees of freedom around the zp axis of the pinion are released, modeling, these phenomena also demonstrate the reliability of the
and a torque of 75 N.m is applied. model proposed in this paper.
Figs. 5(b) and (c) depict the contact stress and contact ellipses pro-
ducted by the finite element method. As illustrated in Fig. 5(b), the 4. Results and discussion
model constructed using conventional approaches undergoes contact
interference, which leads to a stress concentration at the tooth top. 4.1. The influence of design parameters on top cutting
Conversely, the tooth surface constructed using the mathematical
modeling method proposed in this paper exhibits contact deformation Using the following fundamental parameters, analyze the impact of
only at the theoretical contact position, as shown in Fig. 5 (c). design parameters on the top cutting tooth width of the face gear:
modulus m = 4 mm, number of teeth of the gear cutter Ns = 20–150,
number of teeth of the face gear Nf = 20–150, pressure angle α0 = 20◦ ,
3.3. Experimental verification tooth top height coefficient ha*=1, and head clearance coefficient
c*=0.25. The influence of various design parameters on the top cutting
To further validate the models, verification was conducted using the tooth width Dtc is illustrated in Fig. 7.
method illustrated in Fig. 6. The following are the specific steps: Fig. 7 (a) demonstrates the effect of modulus on the top cutting tooth
Step 1: Apply the parameters in Table 2, and calculate the inner and width: when the number of teeth between the gear cutter and the face
outer radii of the face gear based on the tooth width limit conditions (see gear is constant, an increase in modulus leads to a rise in the top cutting
Appendix C) to determine the blank’s dimensions. tooth width. Similarly, augmenting the pressure angle or addendum
Step 2: Divide the face gear into two regions for gear modeling: one coefficient can yield similar outcomes, as depicted in Fig. 7(b) and (c).
region utilizes the mathematical model from Section 2.2, while the other However, as shown in Fig. 7 (d), an increase in the top clearance coef-
region adopts the method proposed in this paper, as outlined in Section ficient has the opposite effect. Lastly, it is evident from this figure that
2.3. the larger the transmission ratio, the wider the top cutting tooth width.
Step 3: Complete the gear shaping of the end face gear on a CNC On the basis of Fig. 7, further insights into the effect of design pa-
machine. rameters on the critical values of the tooth number and transmission
Step 4: Mount the machined face gear and cylindrical gear onto a ratio for top cutting were obtained, which is depicted in Fig. 8. The
rolling inspection machine. Uniformly apply blue lead oil on the contact number of tooth pairs with top cutting above and below the line is
surface of the face gear, making sure that area II meshes with the pinion indicated by the data line and critical line on the graph, respectively. In
in its initial installed state. Manually drive the face gear while observing Fig. 8(a), it is evident that changing the modulus has no effect on the
the contact marks. critical transmission ratio for a fixed number of teeth pairs with top
The gear meshing characteristics of the two regions obtained from cutting. Nevertheless, with an increased pressure angle, the critical
the experiment are depicted in Fig. 6 (d). The gear surfaces in area II are transmission ratio is smaller when the shaper tooth number is small. On
seen to be in good meshing condition, with contact positions aligning as

8
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Fig. 9. Schematic diagram of chamfering at the tooth top of face gear.

width, consequently diminishing the bending resistance of the teeth.

Table 3
With regard to the tooth top chamfer, the performed chamfer here is
Results of chamfer design parameter.
denoted as D1 × D2, as illustrated in Fig. 9.
Parameters name Results of Results of The parameters D1 and D2 are directly associated with the maximum
Table 1 Table 2
inner radius Dt at which top cutting occurs and the minimum inner
Inner radius without root cutting Di / 119.31 155.78 radius Di at which undercutting does not occur, expressed as
(mm) {
Inner radius without top cutting Dt / (mm) 121.85 157.93 D1 = Dt − Di
(21)
Chamfer parameters D1, D2 / (mm) 2.54, 0.64 2.15, 0.72 D2 = ifs D1

In order to verify the rationality of chamfering, it is recommended to

the other hand, for a larger tooth number of shapers, top cutting is more
round up the values of D1 and D2 derived from Eq. (21) to obtain in-
likely at lower transmission ratios, as indicated in Fig. 8(b). Fig. 8(c)
tegers. The tooth profile parameters in Table 1 and 2 were used to verify
clearly illustrates that raising the coefficient of tooth height produces the
the rationality of the chamfer design, and the results are listed in
opposite effect of increasing the pressure angle. Moreover, augmenting
Table 3.
the clearance coefficient results in a larger critical value for the critical
Combined with MATLAB visualization analysis, as shown in Fig. 10,
transmission ratio, as demonstrated in Fig. 8(d).
the chamfer line Lct is clearly below the boundary line Lwt, successfully
avoiding top cutting.
4.2. Suggestions for chamfering
5. Conclusion
Accurate modeling is paramount for ensuring the adaptability of face
gears to additive manufacturing, powder metallurgy, casting, and other In this paper, through theoretical derivation and experimental veri-
processes. As a result, it is imperative to provide relevant model refer- fication, it was found that there is a tooth top overcutting phenomenon
ences, which mainly consist of two schemes: the configuration of the top during the machining process. An improved model of the face gear was
cutting surface and the chamfering of the tooth top. Within this frame- put forth to describe the top-cutting characteristics. The following are
work, the main objective of this work is to provide designs for tooth top the primary conclusions:
chamfering schemes, as the first method results in a reduction in tooth

Fig. 10. Feasibility analysis of tooth chamfering: (a) Results of Table 1; (b) Results of Table 2.

9
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

(1) A mathematical model of the face gear tooth surface was con- Software, Methodology, Formal analysis. Zhengminqing Li: Supervi-
structed using the normal to tooth profile method, which includes sion, Project administration, Funding acquisition, Data curation. Xiao-
the working tooth surface, fillet surface and top cutting surface. feng Yu: Writing – review & editing, Conceptualization.
In particular, the domain of the function of the relevant param-
eters is given.
(2) A simulation program for face gear envelope cutting was devel- Declaration of Competing Interest
oped using 3D software. This program offered a mathematical
model for the over-cutting part of the tooth surface as well as a The authors declare that they have no known competing financial
visual representation of the phenomenon of over-cutting at the interests or personal relationships that could have appeared to influence
inner tooth top of the face gear. the work reported in this paper.
(3) Two types of gear teeth were formed using CNC machine tools,
and the interference phenomenon in traditional tooth shape Data availability
mathematical modeling methods was revealed on a rolling in-
spection machine, which also confirmed the rationality of the Data will be made available on request.
proposed mathematical model.
(4) Suggestions were made for chamfering the inner end of the face Acknowledgments
gear blank to avoid gear tooth interference and to provide a
reliable three-dimensional model for processing processes such as The authors acknowledge the financial support from the National
face gear additive manufacturing, powder metallurgy, and Science and Technology Major Project (Grant No. 2019-VII-0017-0158),
casting. National Natural Science Foundation of China (Grant No. 52175053),
and the National Key Laboratory of Science and Technology on Heli-
CRediT authorship contribution statement copter Transmission (Grant No. HTL-A-22G11). The authors also
sincerely appreciate the comments and suggestions for modifications
Wei Sheng: Writing – original draft, Visualization, Validation, made by the editors and anonymous referees.

Appendix A

This chapter provides a comprehensive overview of utilizing the normal to tooth profile method (NTPM) to define gear tooth shapes, with a
particular emphasis on obtaining conjugate tooth profiles via predetermined contact paths. Using the derivation of the rack tooth profile Σ 1 model as a
case study, the establishment of a coordinate system, illustrated in Fig. A1, is detailed: the origin of the fixed coordinate system oxy is aligned with the
pitch point P, while the coordinate system o1x1y1 is affixed to the rack as the moving coordinate system.

Fig. A1. Modeling of the rack tooth profile using the NTPM.

The equation describing the contact path Σ in the coordinate system oxy is assumed to be as follows:
y = f(x) (A.1)

Assuming the initial meshing position is at position I, where the meshing point is designated as A. As the rack moves from its initial position to the left,
reaching the dashed line position II and intersecting the meshing line at point B1(x, y), the corresponding point on the tooth profile at the initial
position is denoted as B(X, Y). In the fixed coordinate system, the normal lines at points B and B1, which are parallel, satisfy the following relationship:
dX y
= − (A.2)
dY x

Since the vertical coordinates of B1 and B are equal, Eq. (A.2) can be further expressed as follows:

10
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

∫
y
X= − dy + C (A.3)
x
where C represents the integration constant. In most cases, taking the pitch point as the initial contact point may facilitate the calculation of the
integration constant.
According to the meshing path equation, the equation for the rack tooth profile can be expressed as:
⎧ ∫
⎨ X = − y dy + C
x (A.4)
⎩
Y=y

Above all, the profile of cylindrical gears can also be obtained by employing this approach.

Appendix B

This chapter mainly outlines the parameter values of the tooth surface equation of face gear pairs. The range of values for the independent variable
x in Eqs. (8) and (10) are x ∈ [xa, xb] and x ∈ [0, xb], respectively, which can be solved from the following equations:

x2a + (rs − xa tanα0 )2 = ra2 (B.1)

− xb tanα0 = − H = − m(h*a + c* ) (B.2)

The value of the independent variable x in Eq. (15) can be analyzed together with Fig. 2 (f). The tooth top boundary point (such as A0, A1, A2) follows
Eq. (B. 3), and the tooth root boundary point (such as B0, B1, B2) follows Eq. (B. 4).
xtanαi = − mh*a − Rsi + rs (B.3)

x2 + (Rsi − xtanαi )2 = ra2 (B.4)

The range of values for the independent variable x in Eq. (16) is x ∈ [xb, xt], where xt at the dedendum circle satisfies the following Eq. (B.5).
( )
(tan2 αi + 1)xb − xb (tan2 αi + 1)xt
+ arctan = (B.5)
Rsi Rsi − xb tanαi Rsi
The angular difference ηi between two pitch points, such as P1 and P2 in Fig. 2 (e), satisfies the following equation.
⎧
⎪
⎪ x2 + (rs − xtanα0 )2 − R2si = 0
⎪
⎪
⎪
⎪
⎪
⎨ − (tan2 α0 + 1)x
ϕs =
rs (B.6)
⎪
⎪
⎪
⎪
⎪ xs = xcosϕs + (rs − xtanα0 )sinϕs
⎪
⎪
⎩
ys = − xsinϕs + (rs − xtanα0 )cosϕs

Then the angular difference ηi can be denoted as:

( )
x y
ηi = s arccos s (B.7)
|xs | Rsi

Appendix C

The formula for resolving the tooth width limit parameters of face gear, including undercutting, pointing, and top cutting is provided in this
chapter. Based on the tooth top parameters of the face gear mentioned in Eq. (B.4) above and the undercutting calculation method proposed by Litvin,
the section parameters Rsi of the shaper to avoid undercutting are calculated as follows:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
R2si (R2si − rb2 ) ± i2fs rb2 r2a − r2b R2si − r2b ∓ i2fs (ra2 − rb2 )rb2 = 0 (C.1)

Therefore, the minimum inner radius to avoid undercutting can be expressed as

√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
x2b + L2i = Di (C.2)

There is a phenomenon of pointing at the outer end of the face gear tooth, and the section parameters that correspond to the point at the total tooth
height satisfy the following equation.
xcos[ifs (ϕs + η + κ)] + Li sin[ifs (ϕs + η + κ)] = 0 (C.3)

11
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

Hence, the maximum outer radius to avoid pointing can be expressed as:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
x2 + L2i = Do (C.4)

According to the inner and outer radii, the face gear tooth width can be calculated as:
Bf = Do − Di (C.5)

To avoid top cutting, the following equation should be met.

r2 rb2
rb cosαi ⩾rs − mh*a ⇒ b ⩾rs − mh*a ⇒Rsi ⩽ (C.6)
Rsi rs − mh*a

Furthermore, the following formula can be used to determine the face gear’s radius without top cutting:
rb2
Li = (C.7)
ifs (rs − mh*a )

References [19] H.A. Zschippang, S. Weikert, K. Wegener, Face-gear drive: meshing efficiency
assessment, Mech. Mach. Theory 171 (2022) 104765, https://doi.org/10.1016/j.
mechmachtheory.2022.104765.
[1] F.L. Litvin, J.C. Wang, R.B. Bossler, et al., Application of face-gear drives in
[20] H.A. Zschippang, N. Lanz, K.A. Küçük, et al., Face-gear drive: assessment of load
helicopter transmissions, J. Mech. Des. 116 (3) (1994) 672–676, https://doi.org/
sharing, transmission characteristics and root stress based on a quasi-static
10.1115/1.2919434.
analysis, Mech. Mach. Theory 151 (2020) 103914, https://doi.org/10.1016/j.
[2] S. Mo, X. Tang, K. Chen, et al., Analysis of vibration characteristics of face gear
mechmachtheory.2020.103914.
powering-split transmission system, Commun. Nonlinear Sci. Numer. Simul. 130
[21] X. Zhu Y. Dai Zhu X, Dai Y. Development of an analytical model to predict the
(2024) 107750, https://doi.org/10.1016/j.cnsns.2023.107750.
churning power losses of an orthogonal face gear. Engineering Science and
[3] F.L. Litvin, Y. Zhang, J.C. Wang, et al., Design and geometry of face-gear drives,
Technology, an International Journal, 2023. 41: 101383. doi: 10.1016/j.
J. Mech. Des. 114 (4) (1992) 642–647, https://doi.org/10.1115/1.2917055.
jestch.2023.101383.
[4] F.L. Litvin, G.P. Ignacio, A. Fuentes, et al., Design, generation and stress analysis of
[22] L. Zhu, J. Shi, X. Gou, Modeling and dynamics analyzing of a torsional-bending-
face-gear drive with helical pinion, Comput. Methods Appl. Mech. Eng. 194
pendular face-gear drive system considering multi-state engagements, Mech. Mach.
(36–38) (2005) 3870–3901, https://doi.org/10.1016/j.cma.2004.09.006.
Theory 149 (2020) 103790, https://doi.org/10.1016/j.
[5] F.L. Litvin, A. Nava, Q. Fan, et al., New geometry of face worm gear drives with
mechmachtheory.2020.103790.
conical and cylindrical worms: generation, simulation of meshing, and stress
[23] X. Peng, J. Zhou, Optimization design for dynamic characteristics of face gear drive
analysis, Comput. Methods Appl. Mech. Eng. 191 (27–28) (2002) 3035–3054,
with surface-active modification, Mech. Mach. Theory 176 (2022) 105007, https://
https://doi.org/10.1016/S0045-7825(02)00235-9.
doi.org/10.1016/j.mechmachtheory.2022.105007.
[6] Benz A. Cylkro® face gears: Dutch design and Swiss ingenuity cause transmission
[24] J. Tang, F. Yin, X. Chen, The principle of profile modified face-gear grinding based
breakthrough, in International Gear Conference 2014: 26th–28th August 2014,
on disk wheel, Mech. Mach. Theory 70 (2013) 1–15, https://doi.org/10.1016/j.
Lyon, France. 2014, 1184-1187. doi: 10.1533/9781782421955.1184.
mechmachtheory.2013.06.013.
[7] H.A. Zschippang, S. Weikert, K.A. Küçük, et al., Face-gear drive: geometry
[25] H. Guo, X. Peng, N. Zhao, et al., A CNC grinding method and envelope residual
generation and tooth contact analysis, Mech. Mach. Theory 142 (2019) 103576,
model for face gear, Int. J. Adv. Manuf. Technol. 79 (2015) 1689–1698, https://
https://doi.org/10.1016/j.mechmachtheory.2019.103576.
doi.org/10.1007/s00170-015-6915-7.
[8] G. Feng, Z. Xie, M. Zhou, Geometric design and analysis of face-gear drive with
[26] Y. Zhou, J. Tang, H. Zhou, et al., Multistep Method for Grinding Face-Gear by
involute helical pinion, Mech. Mach. Theory 134 (2019) 169–196, https://doi.org/
Worm, J. Manuf. Sci. Eng. 138 (7) (2016) 071013, https://doi.org/10.1115/
10.1016/j.mechmachtheory.2018.12.020.
1.4033387.
[9] Y. Wu, Y. Zhou, Z. Zhou, et al., An advanced CAD/CAE integration method for the
[27] Y. Wang, L. Hou, Z. Lan, et al., Precision milling method for face-gear by disk
generative design of face gears, Adv. Eng. Softw. 126 (2018) 90–99, https://doi.
cutter, Int. J. Adv. Manuf. Technol. 89 (2016) 1545–1558, https://doi.org/
org/10.1016/j.advengsoft.2018.09.009.
10.1007/s00170-016-9189-9.
[10] S. Cai, Z. Cai, C. Lin, Modeling of the generating face gear grinding force and the
[28] Y. Wang, Z. Lan, L. Hou, et al., An efficient honing method for face gear with tooth
prediction of the tooth surface topography based on the abrasive differential
profile modification, Int. J. Adv. Manuf. Technol. 90 (2017) 1155–1163, https://
element method, CIRP J. Manuf. Sci. Technol. 41 (2023) 80–93, https://doi.org/
doi.org/10.1007/s00170-016-9447-x.
10.1016/j.cirpj.2022.11.022.
[29] Y. Wang, G. Su, X. Chu, et al., A finishing method for the continuous generation of
[11] S.H. Chang, T.D. Chung, S.S. Lu, Tooth contact analysis of face-gear drives, Int. J.
spur face gears with shaving cutters, Int. J. Mech. Sci. 190 (2021) 106020, https://
Mech. Sci. 42 (3) (2000) 487–502, https://doi.org/10.1016/S0020-7403(99)
doi.org/10.1016/j.ijmecsci.2020.106020.
00013-2.
[30] J. Tang, X. Yang, Research on manufacturing method of planing for spur face-gear
[12] Tsay MF, Fong ZH. Novel profile modification methodology for moulded face-gear
with 4-axis CNC planer, Int. J. Adv. Manuf. Technol. 82 (2016) 847–858, https://
drives. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of
doi.org/10.1007/s00170-015-7417-3.
Mechanical Engineering Science, 2007. 221(6): 715-725. doi: 10.1243/
[31] X. Yang, J. Tang, Research on manufacturing method of CNC plunge milling for
0954406JMES32.
spur face-gear, J. Mater. Process. Technol. 214 (12) (2014) 3013–3019, https://
[13] X. Lu, Y. Zhou, D. He, et al., A novel two-variable optimization algorithm of TCA
doi.org/10.1016/j.jmatprotec.2014.07.010.
for the design of face gear drives, Mech. Mach. Theory 175 (2022) 104960, https://
[32] I. Dudás, S. Bodzás, Production geometry analysis, modeling, and rapid
doi.org/10.1016/j.mechmachtheory.2022.104960.
prototyping production of manufacturing tool of spiroid face gear, Int. J. Adv.
[14] A. Andersson, An analytical study of the effect of the contact ratio on the spur gear
Manuf. Technol. 66 (2013) 271–281, https://doi.org/10.1007/s00170-012-4323-
dynamic response, J. Mech. Des. 122 (4) (2000) 508–514, https://doi.org/
9.
10.1115/1.1320819.
[33] L. Huang, Y. Cao, F. Jia, et al., Research on spur face gear by electrochemical
[15] J.F. Hochrein, M. Otto, K. Stahl, Face gear drives: nominal contact stress
machining based on the complex surface mesh, Chaos Solitons Fractals 130 (2020)
calculation for flank load carrying capacity evaluation, Mech. Mach. Theory 195
109443, https://doi.org/10.1016/j.chaos.2019.109443.
(2024) 105573, https://doi.org/10.1016/j.mechmachtheory.2024.105573.
[34] S. Wang, Y. Zhou, J. Tang, et al., Digital tooth contact analysis of face gear drives
[16] L. Liu, L. Zhu, X. Gou, Calculation of line and point contact ratio for orthogonal
with an accurate measurement model of face gear tooth surface inspected by
spur-face gear drive, Int. J. Mech. Sci. 236 (2022) 107758, https://doi.org/
CMMs, Mech. Mach. Theory 167 (2022) 104498, https://doi.org/10.1016/j.
10.1016/j.ijmecsci.2022.107758.
mechmachtheory.2021.104498.
[17] L. Liu, L. Zhu, X. Gou, Modeling and analysis of load distribution ratio and meshing
[35] J. Tao, J. Zhou, S. Wang, et al., An efficient and accurate measurement method of
stiffness for orthogonal spur-face gear drive under point contact, Mech. Mach.
tooth flank variations for face gears, Measurement 221 (2023) 113486, https://doi.
Theory 182 (2023) 105239, https://doi.org/10.1016/j.
org/10.1016/j.measurement.2023.113486.
mechmachtheory.2023.105239.
[36] X. Fu, Z. Fang, Y. Cui, et al., Modelling, design and analysis of offset, non-
[18] R. Zhou, N. Zhao, Z. Tan, et al., Computerized generation and meshing simulation
orthogonal and profile-shifted face gear drives, Adv. Mech. Eng. 10 (9) (2018)
of a backlash-adjustable face-gear drive with a tapered involute pinion, Mech.
1–12, https://doi.org/10.1177/1687814018798250.
Mach. Theory 140 (2019) 479–503, https://doi.org/10.1016/j.
mechmachtheory.2019.06.017.

12
W. Sheng et al. Engineering Science and Technology, an International Journal 59 (2024) 101849

[37] S. Mo, B. Luo, W. Song, et al., Geometry design and tooth contact analysis of non- [41] W. Sheng, Z. Li, H. Zhang, et al., Geometry and design of spur gear drive associated
orthogonal asymmetric helical face gear drives, Mech. Mach. Theory 173 (2022) with low sliding ratio, Adv. Mech. Eng. 13 (4) (2021) 1–12, https://doi.org/
104831, https://doi.org/10.1016/j.mechmachtheory.2022.104831. 10.1177/16878140211012547.
[38] S. Mu, Y. Zhao, J. Cui, et al., Meshing theory of face worm gear drive with [42] J. Zhao, W. Sheng, Z. Li, et al., Study on the lubrication characteristics of spur gear
hardened cylindrical worm, Mech. Mach. Theory 185 (2023) 105323, https://doi. pairs with low sliding ratio under mixed elastohydrodynamic lubrication, J. Tribol.
org/10.1016/j.mechmachtheory.2023.105323. 144 (7) (2022) 071604, https://doi.org/10.1115/1.4052939.
[39] D. Liu, T. Ren, X. Jin, Geometrical model and tooth analysis of undulating face [43] W. Sheng, J. Zhao, Z. Li, et al., Geometric design of a face gear drive with low
gear, Mech. Mach. Theory 86 (2015) 140–155, https://doi.org/10.1016/j. sliding ratio, Appl. Sci. 12 (6) (2022) 2396, https://doi.org/10.3390/
mechmachtheory.2014.12.004. app12062936.
[40] C. Lin, Z. Cai, Research on the Discrete Algorithm of tooth surface for a curve-face [44] L.V. Mohan, M.S. Shunmugam, CAD approach for simulation of generation
gear, J. Mech. Des. 142 (5) (2020) 053301, https://doi.org/10.1115/1.4044434. machining and identification of contact lines, Int J Mach Tool Manu 44 (7–8)
(2004) 717–723, https://doi.org/10.1016/j.ijmachtools.2004.02.013.

13