76 Int. J. Simulation and Process Modelling, Vol. 11, No.

1, 2016

Formational principle and accurate fitting

methodology for a new tooth surface of the spiral
bevel gear

Rui Zhan and Xieeryazidan Adayi*

School of Mechanical Engineering,
Xinjiang University,
Xinjiang Urumqi 830046, China
Email: zrwudi1989@163.com
Email: adayxj@126.com
*Corresponding author

Han Ding
State Key Laboratory of High-performance Complex Manufacturing,
Central South University,
Changsha 410083, China
Email: dinghan0204@163.com

Abstract: Distinguished with some basic theory and approaches of traditional tooth surface
design, the spiral bevel gear with new tooth surface, namely spherical involute tooth surface, is
investigated. Firstly, it improved its formational principle based on the spherical involute theory.
As a key part, the parametric equation of the generating line was made a detailed derivation.
To this end, fast and accurate solutions of the boundary curves and the tooth profile curve
family are accomplished, respectively. Then, taking advantages of modelling techniques of cubic
non-uniform rational B-spline (NURBS) curve and surface in CAD/CAM, a reconstruction
method for the spherical involute tooth surface was proposed to enhance the precision. At last,
related optimisation schemes associated with constructed NURBS tooth surface are utilised to
obtain the parameterisation of tooth surface data and higher-accuracy fitting. Given numerical
examples indicate that the accuracy of the tooth surface is obviously enhanced and is enough to
provide gear finite element (FEM) analysis the data and the basic model in its digitised design
and manufacturing.

Keywords: spiral bevel gear; spherical involute tooth surface; generating line; cubic NURBS;
reconstruction method; higher-accuracy fitting.

Reference to this paper should be made as follows: Zhan, R., Adayi, X. and Ding, H. (2016)
Formational principle and accurate fitting methodology for a new tooth surface of the spiral
bevel gear, Int. J. Simulation and Process Modelling, Vol. 11, No. 1, pp.7684.

Biographical notes: Rui Zhan is a Master degree candidate from the School of Mechanical
Engineering, Xinjiang University, Urumqi City, Xinjiang Uygur Autonomous Region, China.
His areas of research include the special electrochemical processing, digitised design and
manufacturing of the spiral bevel and hypoid gear.

Xieeryazidan Adayi is a Professor in School of Mechanical Engineering at the Xinjiang

University, Urumqi, China. He received his PhD in Mechanical Engineering from Dalian
University of Technology, Dalian, China. He has contributed about 30 research papers in journals
and conferences at home and abroad level. His areas of research include the special
electrochemical processing and advanced design theory and method, digitised design and
manufacturing of the spiral bevel and hypoid gear.

Han Ding is currently a PhD student at State Key Laboratory of High-Performance Complex
Manufacturing, Central South University, China. He has contributed about ten research papers in
journals and conferences at home and abroad level. His area of research includes manufacture
method for the spiral bevel and hypoid gear.

Copyright 2016 Inderscience Enterprises Ltd.

 Formational principle and accurate fitting methodology for a new tooth surface of the spiral bevel gear 77

1 Introduction This paper firstly improves a fundamental design

principle to achieve the spiral bevel gear with a kind of new
With the development of modern mechanical industry, the
tooth profile named the spherical involute. Additionally, in
spiral bevel gear has become an increasingly important
order to the tooth flank of gear model after simulation
component for power transmission between intersecting
process has a very high accuracy and provide a base model
axes. Recently, the latest and highest level of process
for the follow-up gear FEM analysis, such as TCA,
technology has been represented by the Gleason
LTCA (De-vaujany et al., 2007; Guingand et al., 2004), this
PHOENIX Series five-axis CNC machine tools equipped
paper investigates some optimised design schemes by
with the advanced manufacture expert system (GEMS)
combining the unique advantages with the characteristics
(Fan and Ron, 2005). However, gear tooth profile designed
of the non-uniform rational B-spline (NURBS) curves
and machined out is still the approximate one based on
and surfaces in CAD/CAM. In ISO STEP standards
the meshing principle and all kinds of machining methods
promulgated, NURBS method as an excellent overall
are the approximate spherical involute (Litvin et al.,
smooth function of curve fitting (Piegl and Tiller, 2001) and
2006; Argyris et al., 2002). It has some characteristics and
local regulation characteristics is a kind of mainstream
drawbacks:
mathematical method to define the geometrical shape
1 In design of the tooth surface, it is too difficult to obtain of the industrial products (Weiss and Andor, 2002). With
the accurate equation expression to get a fast application of the optimised extraction of proper data points
parameterisation model because of numerous and which are made from anti-calculation for completion of
complex derivations (Wang and Fong, 2006; Gosselin cubic NURBS curve fitting, the reconstruction method for
et al., 1995, 2002). initial tooth flank and relevant optimal approaches are
introduced to improve precision of simulation machining
2 In process of the tooth surface, it can not maintain the
modelling (Roman and Marek, 2015), respectively.
correct state of meshing gear drives where
instantaneous transmission ratio is not constant to cause
noise, heat deformation, uneven distribution of contact
2 The proposed theory and its inference
pressure and other phenomena. Consequently, tooth
contact analysis (TCA) or loaded tooth contact analysis Compared with the traditional spiral bevel gear design, the
(LTCA), tooth flank modification and other vital geometric expression of new tooth profile is more complex.
technologies in manufacture (Chen and Fong, 2015; In the previous design, the fillet tooth surface was often
Vimon, 2014) have to be applied to reduce this error. ignored (Shunmugam et al., 1998; Li et al., 2015). In the
formation of tooth flank, the vital generating line is replaced
Moreover, in the actual transmission procedure of the spiral
by a straight line which can just form the straight bevel gear
bevel gear drives, the tooth profile can always exist on the
(Hu and Yang, 2011). Therefore, taking into account
spherical surface whose centre is the base cone vertex.
previous researches, this paper gives an improved
Theoretically speaking, the tooth profile curve is a spherical
identification of the spherical involute tooth flank and then
involute curve (Zeng, 1989). This kind of tooth profile has
describes an improved formational principle. Moreover, it
many advantages, some of which can made a good
derives the generating line which determines the initial
compensation for the drawbacks of the conventional
points in the formation procedure. As for the parametric
approximate tooth profile as the result of a constant
modelling process of the whole tooth surface, it is similar to
transmission ratio. With the development of state-of-the-art
the methodologies in related references (Li et al., 2015).
CNC process technology, there are always some researchers
continuing advancing its corresponding researches.
Shunmugam et al. (1998) introduced the exact spherical 2.1 Composition of the spherical involute tooth
involute of spiral bevel gear and investigated normal surface
deviation from the idealised surface. Suh et al. (2003, 2001) Strictly speaking, a whole tooth surface of the spiral bevel
and Dacak and Aangeles (1994) carried out a detailed gear includes a top, two side and transverse surfaces as
modelling and process investigation on the spherical shown in Figure 1. The tooth top is surrounded by the two
involute tooth surfaces. In the reference (Giorgio and Jorge, pieces of top line and top circle of the toe and heel. Two
2005), the exact spherical involute tooth profile of bevel side surfaces refer to the tooth concave and tooth convex in
gears and crown rack is obtained through the pure-rolling symmetrical shape which can respectively contain three
movement of a great circle of the fundamental sphere on the parts namely working tooth surface, tooth root surface and
base cone. Meanwhile, the research team of Jilin University fillet curve surfaces between the above two. The integral
in China has created a new method named generating-line tooth profile of transverse surfaces, namely the gear toe or
cutting method and published some experimental research the gear heel, can cover tooth top arc, working tooth profile,
papers (Hu and Zhang, 2011). Li et al. (2015) built a fillet curve and tooth root curve. It is worth remarking that
mathematical model of the spherical involute tooth surface the spherical tooth surface is enclosed by the working tooth
of a spiral bevel gear whose generating curve is a circle arc. profile of toe and heel, tooth fillet curve and tooth top arc.
They finally got the contact patterns in meshing of spiral Incidentally, the work tooth surface is typical spherical
bevel gear drive by imposing the TCA. involute one. From a practical point of view, tooth surface is
78 R. Zhan et al.

referred to as the work tooth surface. Now, for brevity, surface. The equation of spherical involute K0Kt can be
working tooth surface as a main subject is made a detailed expressed by the spherical deflection angle k at any point K
description, because each part of the whole tooth surface as follows:
has a good agreement. It is worth noting that the tooth
surface means the work tooth surface in rest of this present
paper.
=
arc cos ( cos b )
sin b (
arc cos tan b
tan ) (1)

Figure 1 Composition of spherical involute tooth surface Figure 2 Formation of the spherical involute tooth surface
(see online version for colours) (see online version for colours)

(a)

2.2 The proposed theory of forming the spherical

involute tooth surface
It is worth remarking that, in the transmission of spiral bevel
gear drives, the distance from each tooth point to the pitch
cone vertex is always the same, it means that the tooth
profile curve is the spherical one (Zeng, 1989). In other
words, the ideal tooth profile is a spherical involute curve.
From this perspective, the formational principle of the
spherical involute tooth surface can be depicted: the (b)
spherical involute tooth surface is accomplished by
In the structured spherical coordinate system, the spherical
enveloping tooth profile from one transverse surface toward
involute can be expressed as:
another one. As demonstrated in Figure 2(a), when a certain
plane is purely rolling on a conical surface, this flat K = K ( , , ) = K ( R , , ) (2)
trajectory of a point on the plane is called the spherical
involute. Conical surface is base cone Ok0O1 that is tangent where k represents the cone angle at a point of the
to the circle plane O at the bus bar Ok0. And the curve K0Kt corresponding spherical involute, the = Rk [R B, R],
in both the plane O and the base circle cone is named = k [f, a], where B is gear face width, a is top cone
generating line. Selecting the equal division points kn (n = 0, angle, f is root cone angle and b is base cone angle. They
1, , t) along the generating line, after the point of the heel are the basic geometric parameters for gear design.
kn = 0 completes one motive trajectory of tooth profile curve
by rotation angle , straight line Ok0 as the tangential 2.3 Solution of equation of the generating line
during pure rolling automatically turns a certain angle n,
continually the next point kn = 1 makes a space sphere In the formation procedure, the generating line k0kt is the
movement to form an piece of spherical involute tooth key to achievement of the accurate spherical involute tooth
profile once again. Making such repeated motion, every surface. According to tooth surface characteristics of the
point respectively completes a space motion when taking a spiral bevel gear (Zeng, 1989), generating line can be
related point from starting point k0 to end point kt, all the represented by the base cone spiral curve which is an
trajectory curves can constitute a whole side of the spherical Archimedean spiral curve, namely equidistant spiral curve.
involute tooth surface, as shown on the right in Figure 2(b). As displayed in Figure 3, in the light of definition of the
Besides, this principle can also be described as: changing spiral curve, assuming that the cone distance of moving
the radius R of the circle plane namely the tangent line by a point kn from point O is Rn, the spiral curve can be
certain angle n, its transverse point kn (n = 0, 1, ..., t) makes expressed by equation as follows:
the completion of one space trajectory, finally all the
trajectory curves are evolved the spherical involute tooth
 Formational principle and accurate fitting methodology for a new tooth surface of the spiral bevel gear 79

x = Rn sin b cos n And then, the spiral rotation angle n is solved by the
following formula:
y = Rn sin b sin n (n = 0,1, 2...t ) (3)
z = R cos tan n
n b
n = (5)
sin
Moreover
where represents pressure angle of the spiral bevel gear
R0 Rn Rt (4)
and n represents spiral angle at any points on the toe and
where R0 and Rt are distance from the heel and toe from the heel that can be obtained by the following approach (Zeng,
vertex on base cone surface, respectively. 1989). As tooth trace of the spiral bevel and hyoid gear
denoted in Figure 4, Figure 4(a) is its formation process in
Figure 3 Solution of the generating line the machining state, and Figure 4(b) is a schematic view of
its geometry.
1
( Rn ) + 2rc R p sin R p2
2
n = arc sin (6)
2 Rn
where rc means the radius of the cutter head, Sd indicates
cutter location, Rp signifies cone distance in correspondence
with the nominal spiral angle .

3 NURBS fitting of spherical involute tooth

surface
Cubic NURBS surface can be represented as:
n m

Figure 4 Tooth trace of the spiral bevel gear (see online version
for colours)
N
i =0 j =0
i ,3 (u ) N j ,3 (v )Wi , jVi , j

S (u , v) = n m
(7)
N
i =0 j =0
i ,3 (u ) N j ,3 (v )Wi , j

where u, v [0, 1], Vi,j are control vertices, and Wi,j

are weighting factors. Node vectors at u-direction are
expressed as:
U = [0, 0, 0, 0, u4 , , un ,1,1,1,1] [0,1] (8)

Similarly, node vectors at v-direction are consistent with

the u-direction ones. Ni,3(u) and Ni,3(v) represent the cubic
B-spline base function at u-direction and v-direction,
respectively. They can be solved by the following recursive
formula:

(a) 1, ui u ui +1
Ni ,0 (u ) =
0, the others
u ui u u
Ni ,3 (u ) = N i ,2 (u ) + i + 3+1 N i +1,2 (u ) (9)
ui + 3 ui ui + 3 +1 ui +1
0 = 0
0

Based on cubic NURBS surface modelling, this paper

employs the skinning method by extracting boundary of the
tooth surface and the internal cross-section curve family, in
order to structure the spherical involute tooth surface. The
main steps are:

(b)
80 R. Zhan et al.

Step 1 Construction of an internal cross-section curve Wi(i = 0, 1, , n) is weighting factors, and Bi,3(u)
family. is the cubic B-spline base function. NURBS curve
interpolation can be performed with anti-calculated
Firstly, combined with proposed equation of the
control vertices Vi(i = 0, 1, ..., n) and weighting
generating line, the graphics were drawn through
factor Wi(i = 0, 1, ..., n) that must meet the
the relevant CAD software (Luis et al., 2014).
conditions, so as to complete fitting the boundaries.
Then, it needs to gain some points with equal
Their specific anti-calculation procedure has been
divided segments onto generating line. Here, the
given a detailed answer in the reference (Piegl and
number of points with equal length n can be solved
Tiller, 1997).
according to the following formula:
Step 3 Solution of the node vector.
R0 Rt
n= (m = 0,1,..., t 1) (10) In this paper, a modified radial accumulated chord
Rm Rm +1
length parameter method (Piegl and Tiller, 1997)
After getting equal division points kn, they were was applied to solve them through the formula:
submitted into related graphics software. In
Si 4 Si 5 i 4
addition to, after accomplishing the positional
relationship of circle plane and the base cone, 3 Si 4 Si 5 Si 3 Si 4
based on ration function selected in software, a ki 3 = 1 + 2 Si 4 Si 5 i 3
serial of tooth profile curves can be obtained in +

light of spherical involute tooth formational S i 3 S i 4 S i 2 Si 3

principle. These tooth profile curves are depicted ki 3 S i 3 S i 4
as an internal cross-section curve family of the ui 3 = (12)
n +1

tooth surface. To facilitate the NURBS fitting
procedure, cross-sectional family of curves is



i =4
ki 3 S i 3 S i 4

denoted as
v-direction curve family S(v). i 3 = min Si 4 Si 3 S2 ,
2
Step 2 Identification of the main boundary curves.
S1 S0 = Sn 1 Sn 2
As shown in Figure 5, main boundaries of a
spherical involute tooth surface should be where S presents u-direction node, i = 1,2, , n.
constituted by the working tooth profile S(v = 0) of Step 4 Unified each node vector.
the heel and S(v = 1) of the toe, the tooth top arc
S(u = 1) and the generating line namely fillet curve Each section curve vector (set as the u-direction)
S(u = 0). Among them, the starting one of needs to do some numerical operations. Nodes can
u-direction curve family is the generating line be inserted new ones in order to ensure the shape,
S(u = 0) which is solved out. And S(u = 1) has been then each section curve to calculate their control
explicated to make cubic NURBS fitting. After vertices again to get a new unified one.
getting a tooth profile with a computer program, it Step 5 Solution of the u-direction control vertices.
needs to simulate space motion trajectory of
aliquots points at the generating line. Data at the According to solution of u-direction control points
other end points are extracted to make fitting and of the transition fillet curve, the average data of
written as a set of data points Pn. As a result of data uniform node vectors is regarded as their new node
points aliquots, their weighting factors hn(n = 0, 1, vector. Then, the required control points as data
..., t) may be selected sequentially and points, one by one anti-calculations of the
equivalently. u-direction control vertices are made. Resulting
vertices after their process are shown in Figure 6.
Therefore, a cubic NURBS curve can be expressed
as: Step 6 Skinning operation

n Based on boundary curves, internal cross-section

B
i =0
i ,3 (u )WiVi
curve family after the NURBS fitting and
the whole control vertices obtained through
P (u ) = , u3 u un +1 (11)
n anti-calculation, the NURBS reconstruction of the
B
i =0
i ,3 (u )Wi spherical involute tooth surface is executed by
applying the skinning method (Piegl and Tiller,
Here, the node vector U = [u0, u1, , un+3+1] 1997). As shown in Figure 7, it represents the final
[0, 1] is similar with equation (8), so there are u0 NURBS tooth surface.
= u1 = u3 = u3 = 0 and un+1 = un+2 = un+3 = un+4 = 1.
And Vi(i = 0, 1, , n) indicates control vertices,
 Formational principle and accurate fitting methodology for a new tooth surface of the spiral bevel gear 81

Figure 5 Internal cross-section curve family and boundary existent minimum distance from the surface. In this
mapping process, the data points need to satisfy the
following equation:

{ }
( P ( ui , vi ) + J ( ui , vi )( ui , vi ) ) Pi Pu ( ui , vi ) = 0


{ }
( P ( ui , vi ) + J ( ui , vi )( ui , vi ) ) Pi Pv ( ui , vi ) = 0 (13)
s.t. ( ui , vi ) <

here J = (P , P ) is Jacobian matrix of the tooth

u v
surface to be operated, Pi is point matrix and is threshold
value of the optimised tooth flank error.
The above formula can be solved by using Newton
iterative method, because this algorithm has a good global
convergence and an excellent computational efficiency
Figure 6 Anti-calculation of the control vertices (see online (Park, 2001). Firstly, the grid is divided into interleaved
version for colours)
equal parametric curves, and then coordinates and
parameter values of each grid point were calculated.
Simultaneously, parameter value of the nearest grid to the
data point node was as the initial value for iterative search
operation, accurate parametric of the related data points on
the spherical tooth surface can be completed.

4.2 Accurate fitting for data points on tooth surface

Taking into account the singular or ignored data points in
the above described extraction (Pavel and Zdenek, 2015) an
optimisation model can be constructed to achieve adequate
smoothness of the reconstructed tooth flank. To this end,
Figure 7 NURBS reconstruction of tooth flank (see online taking the control vertices of base surface as degrees of
version for colours)
freedom, the set approximation error of the original data
points of tooth surface is the most optimal model of the
objective function for finishing solving the best surface
approaching. The nearest distance of the original data points
to the base surface is approximation error, optimisation
model can be built as:
g
min E (Vij ) = P (u , v ) P
2
k k k
k =1

= minn {|| WBV P ||2 } (14)

V R
g
s.t. C (Vij ) = (V Vij0 ) <
2
ij
k =1

where k is the number of fitting curves on the base surface,

4 Optimisation for NURBS-based fitting tooth g is the number of related data points, represents the
surface permitted error value and P represents data points.
4.1 Parameterisation of data points on tooth surface Moreover, W = diag(w1, w2, , wg) Rgg, B = (Bi,3(uk),
Bj,3(vk))h,i,j Rg(m+1)(n+1), V = Vi, j R(m+1)(n+1)3, and P = (P1,
In above described anti-calculation of control points, , Pg)T Rg3.
there will be some control points that may be uneven or This optimal model involves a non-constrained and
irrational. For this problem, this paper introduces a tooth nonlinear quadratic programming problem which can be
flank approximation utilising the least squares method to transformed into the set of ultra-linear equation. According
extract tooth flank points (Park, 2001). While making to the fitting accuracy requirement of given weighting
parameterisation on the base surface after the fitting, data of factor, it can quickly get a stable solution with householder
corresponding projection point from control point was transformation method (Sun et al., 2004) after rendering
usually selected as the parameter value because of the
82 R. Zhan et al.

corresponding weights. As represented in Figure 8, data based on the universal machine tools are 40.9 m on
points after accurate fitting are in good smoothness. concave and 33.8 m on convex respectively. In the
reference (Kawasaki and Tsuji, 2010), the tooth flank with
Figure 8 Data points on tooth surface after accurate fitting error 24 m is used to TCA. With the comparison above
(see online version for colours) results, it clearly indicates a higher-accuracy after
application of the proposed methodologies.
As shown in Figure 11, in establishment of provided
FEM model, it utilises hexahedral mesh to make operation
to reflect the quality of the real model and reduce the size of
the analysis model, in the premise of meeting the accuracy
of the analysis. Then, in the guarantee of the accuracy and
efficiency of the FEM analysis, it provides 128,316 meshes
which is less than the quarter of completely consistent
model with a total of 524,433 meshes, and the size of the
tooth surface element in the mesh area is about 1mm. At
last, gear drive model after reconstruction and optimisation
are submitted into the FEM software, which can indicates it
is enough applicable to make some FEM analyses such as
TCA, LTCA and so on in follow-up studies.
To this end, with comparison of after and before
optimisation, the model with this error accuracy can meet
5 Numerical examples the requirement of gear applications, and the proposed
With application of the formational principle of the methodologies about gear reconstruction and optimisation
spherical involute surface with basic parameters in Table 1, has a good validity.
the model of spiral bevel gear can be obtained. The
Table 1 Main parameters of the modelling of the spiral bevel
modelling error is conventionally evaluated by the root
gear
mean-squared error (RSME) of each tooth surface. As
shown on the right in Figure 9, tooth error with 5 9 data Normal Pitch Outer
Tooth
points after cubic NURBS reconstruction is represented, Module pressure cone cone
number
where the left is the error flank, and the right is the error (m /mm) angle angle distance
(z)
distribution. The maximum error is 4.875 m, the minimum (/) (/) (Ra/ mm)
error is 34.473 m and RSME of tooth error is 18.274 m. 46 8.220 20 71.9395 152.40
Some larger error values are mainly distributed in the two Nominal
ends. As shown in Figure 10, it is the tooth error after Tooth Base cone Root cone Top cone
spiral
width angle angle angle
optimisation and has been significantly reduced. The angle
(B/mm) (b/) (f/) (a/)
maximum error is 12.418 m, the minimum error is (/)
3.6324 m and RSME is 8.4363 m. In the reference 28.573 16.5908 35 68.6585 73.4881
(Shih and Chen, 2012), RMSEs before tooth modification

Figure 9 Tooth error after NURBS reconstruction, (a) error surface (b) error distribution (see online version for colours)

(a) (b)
 Formational principle and accurate fitting methodology for a new tooth surface of the spiral bevel gear 83

Figure 10 Tooth error after optimisation (see online version References

for colours)
Argyris, J., Fuentes, A. and Litvin, F.L. (2002) Computerized
integrated approach for design and stress analysis of spiral
bevel gears, Computer Methods in Applied Mechanics and
Engineering, Vol. 191, Nos. 1112, pp.10571095.
Chen, S.H. and Fong, Z.H. (2015) Study on the cutting time of the
hypoid gear tooth flank, Mechanism and Machine Theory,
Vol. 84, pp.113124.
Dacak, M. J. and Aangeles, J. (1994) The modeling of bevel gears
using the exact spherical involute, ASME Journal of
Mechanical Design, Vol. 116, No. 2, pp.364368.
De-vaujany, J.P., Guingand, M. and Remond, D. (2007)
Numerical and experimental study of the loaded transmission
error of a spiral bevel gear, ASME Journal of Mechanical
Design, Vol. 129, No. 2, pp.195200.
Fan, Q. and Ron, D.F. (2005) Gleason expert manufacturing
system (GEMS) opens a new era for digitized manufacturing
of spiral bevel and hypoid gears, World Manufacturing
Engineering & Market, Vol. 79, No. 4, pp.8793.
Figure11 Model provided for the FEM analysis (see online
Giorgio, F. and Jorge, A. (2005) Algorithms for involute and
version for colours)
octoidal bevel-gear generation, ASME Journal of Mechanical
Design, Vol. 127, No. 4, pp.664672.
Gosselin, C., Cloutier, L. and Nguyen, Q.D. (1995) A general
formulation for the calculation of the load sharing and
transmission error under load of spiral bevel and hypoid
gears, Mechanism and Machine Theory, Vol. 30, No. 3,
pp.433450.
Guingand, M., De-Vaujany, J.P. and Icard, Y. (2004) Fast
three-dimensional quasi-static analysis of helical gears using
the finite prism method, ASME Journal of Mechanical
Design, Vol. 126, No. 6, pp.10821088.
Hu, Y., Zhang, X.C. and Yang, Z.J. (2011) Precision cutting of
spiral bevel gear with spherical involute tooth profile,
Journal of Beijing University of Technology, Vol. 37, No. 5,
pp.641647 (in Chinese).
Kawasaki, K. and Tsuji, I. (2010) Analytical and experimental
tooth contact pattern of large-sized spiral bevel gears in
6 Conclusions cyclo-palloid system, ASME Journal of Mechanical Design,
Vol. 132, No. 4, p.041004.
1 In this paper, a new type of tooth surface and the Li, H.T., Wei, W.J. and Liu, P.Y. (2014) The kinematic synthesis
formation principle of spherical involute tooth surface of involute spiral bevel gears and their tooth contact analysis,
of the spiral bevel gears are investigated. It is capable Mechanism and Machine Theory, Vol. 79, pp.141157.
of fast and accurate obtainment of the tooth surface Litvin, F.L., Fuentes, A. and Hayasaka, K. (2006) Design,
parametric expression. manufacture, stress analysis, and experimental tests of
low-noise high endurance spiral bevel gears, Mechanism and
2 Exploiting fitting technology of NURBS curve and Machine Theory, Vol. 41, No. 4, pp.83118.
surface, accurate reconstruction for the spherical Luis, M.S.D., Guilherme, A.B.P., Pavel, V. and Oliveira, J.A.
involute tooth surface is proposed to provide a uniform (2014) Layout and process optimisation: using
standard for the standardised geometric representation computer-aided design (CAD) and simulation through an
of spiral bevel gear products. It is more efficient to integrated systems design tool, International Journal
of Simulation and Process Modelling, Vol. 9, Nos. 1/2,
provide information and analysis model for the relevant
pp.4662.
mapping and FEM analysis software, especially in the
Park, H. (2001) An approximate lofting approach for B-spline
vital subsequent research such as TCA and LTCA and
surface fitting to functional surfaces, The International
so on. Journal of Advanced Manufacturing Technology, Vol. 18,
3 With application of tooth surface NURBS design for No. 7, pp.474482.
the spiral bevel gear, it can provide an access to its Pavel, R. and Zdenek, U. (2015) Comparison of optimisation
NURBS milling process. For an example, the tool path methods tested on testing functions and discrete event
simulation models, International Journal of Simulation and
calculation and program, and NURBS interpolation Process Modelling, Vol. 10, No. 3, pp.279293.
processing and its optimisation in the NURBS
Piegl, L.A. and Tiller, W. (1997) The NURBS Book, 2nd ed.,
machining, can be expanded and improved to execute Springer, Berlin, Herdelberg.
higher accuracy manufacturing.
84 R. Zhan et al.

Piegl, L.A. and Tiller, W. (2001) Parameterization for surface Sun, Y., Liu, H. and Liu, J. (2004) Research on the method of
fitting in reverse engineering, Computer Aided Design, accurate NURBS surface fitting to scattered points, Chinese
Vol. 33, No. 8, pp.593603. Journal of Mechanical Engineering, Vol. 40, No. 3, pp.1014
Roman, . and Marek, S. (2015) Production function (in Chinese).
implementation in an agent-based simulation, Intrtnational Vogel, O., Griewank, A. and Br, G. (2002) Direct gear tooth
Journal of Simulation and Process Modelling, Vol. 10, No. 2, contact analysis for hypoid bevel gears, Computer
pp.156162. Methods in Applied Mechanics and Engineering, Vol. 191,
Shih, Y.P. and Chen, S.D. (2012) Free-form flank correction in No. 36, pp.39653982.
helical gear grinding using a five-axis computer numerical Wang, P.Y. and Fong, Y.H. (2006) Fourth-order kinematic
control gear profile grinding machine, ASME Journal synthesis for face-milling spiral bevel gears with modified
of Manufacturing Science and Engineering, Vol. 134, No. 4, radial motion (MRM) correction, ASME Journal of
p.041006. Mechanical Design, Vol. 128, No. 2, pp.457467.
Shunmugam, M.S., Rao, B.S. and Jayapakash, V. (1998) Weiss, V. and Andor, L. (2002) Advanced surface
Establishing gear tooth surface geometry and normal fitting techniques, Computer Aided Design, Vol. 19, No. 1,
deviation, Mechanism and Machine Theory, Vol. 33, No. 5, pp.1942.
pp.525534. Yang, Z.J., Li, L.N. and Wang, Y.K. (2011) Basic principle and
Simon, V.V. (2014) Optimal machine-tool settings for the mathematical model of cutting hypoid gears by generating
manufacture of face-hobbed spiral bevel gears, ASME J. line method, Advanced Materials Research, Vol. 154155,
Mech. Des., Vol. 136, No. 8, p.081004. pp.113118.
Suh, S.H. Jih, W.S. and Hong, H.D. (2001) Sculptured surface Zeng, T. (1989) Design and Manufacture of Spiral Bevel Gears,
machining of spiral bevel gears with CNC milling, Harbin Institute of Technology Press, Harbin (in Chinese).
International Journal of Machine Tools & Manufacture,
Vol. 41, No. 6, pp.833850.
Suh, S.H., Jung, D.H. and Lee, E.S. (2003) Modeling,
Implementation and manufacturing of spiral bevel gears
with crown, The International Journal of Advanced
Manufacturing Technology, Vol. 21, pp.775786.