Mechanism

and
Mechanism and Machine Theory 42 (2007) 698–726
Machine Theory
www.elsevier.com/locate/mechmt

Effects of machining errors, assembly errors and

tooth modifications on loading capacity, load-sharing ratio
and transmission error of a pair of spur gears
Shuting Li *

Nabtesco Co. LTD., Oak-hills No. 202, Heki-cho 7028-2, TSU-shi, Mie-ken 514-1138, Japan

Received 30 December 2005; received in revised form 27 April 2006; accepted 5 June 2006
Available online 20 July 2006

Abstract

This paper quantitatively investigates effects of machining errors (ME), assembly errors (AE) and tooth modifications
(TM) on loading capacity, load-sharing ratio (LSR) and transmission error (TE) of a pair of spur gears by using special-
developed finite element method (FEM) software in a personal computer. Tooth surface contact stress (TSCS) and tooth
root bending stress (TRBS) of a pair of spur gears with ME, AE and TM are analyzed by the FEM software. Effects of
ME, AE and TM on TSCS and TRBS are also defined as ‘‘Influence factors’’ and the influence factors of ME, AE and TM
on TSCS and TRBS are calculated quantitatively. It is found that misalignment error of gear shafts on the plane of action
of the spur gears exerts great effects on TSCS and TRBS while misalignment error of gear shafts on the vertical plane of the
plane of action almost exerts no effects on TSCS and TRBS. It is also found that ME and lead crowning have greater
effects on TSCS and TRBS. When the maximum tooth surface machining error is 40 lm, influence factor of ME on TSCS
and TRBS are about 2.2 and 1.5. When quantity of lead crowning is 40 lm, influence factor of lead crowning on TSCS and
TRBS are about 2.0 and 1.4. These influence factors can be used as references when ISO standards are used for tooth sur-
face contact strength and root bending strength calculations of a pair of spur gears with ME, AE and TM.
Methods for LSR calculation of a pair of spur gears with ME, AE and TM are also presented in this paper. Effects of
ME, AE and TM on LSR are analyzed quantitatively. It is found that ME has greater effect on LSR. When the maximum
tooth surface machining error is greater than 10 lm, double pair tooth-contact shall become single pair tooth-contact for
sake of the effect of ME. This means that when ME on tooth surface is large, outer limit of the single pair tooth-contact
area can not be used as the so-called ‘‘worst load position’’. It is safe and reasonable to use tooth tip as the ‘‘worst load
position’’ when to perform tooth bending strength calculation.
Methods for TE calculation of a pair of spur gears with ME, AE and TM are also presented in this paper. Effects of
ME, AE and TM on TE of the gears are analyzed quantitatively. It is found that ME not only changes waveform of TE
easily, but also makes TE greater and greater when ME is increased. AE and TM also make TE greater and greater when
they are increased, but do not change waveform of TE.
Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Gear strength; Load-sharing ratio; Transmission error; Machining errors; Assembly errors; Tooth modifications

*
Tel./fax: +81 059 2566213.
E-mail address: [email protected]

0094-114X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mechmachtheory.2006.06.002
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 699

1. Introduction

It is well known that ME, AE and TM of gears exert great effects on TSCS and TRBS of gears. But so far, it
has not been available that there is an effective method to be able to conduct precise analyses of TSCS and
TRBS of a pair of gears with ME, AE and TM. ISO (International Organization For Standardization) made
standards to calculate TSCS and TRBS of spur gears and helical gears. In ISO standards, application factor,
dynamic factor, face load factor and transverse load factors are introduced to consider the effects of ME, AE
and TM on TSCS and TRBS calculations of gears. But since it depends on experiences and experimental
results to determine these factors and it is almost impossible to get correct values of these factors, the reality
is that ISO standards cannot precisely calculate TSCS and TRBS of a pair of gears with ME, AE and TM. In
recent years, lightweight and compact size design for gears have been required so strictly that gear designers
must precisely know the real TSCS and TRBS of a pair of gears with ME, AE and TM. Also they must know
the effects of ME, AE and TM on gear’s transmission performances quantitatively in theory.
It was started very early to investigate the effects of ME, AE and TM on tooth surface contact strength and
root bending strength by experiments. Hertel [1–3], Rademacher [4], Nieman and Richter [5–7] and Timmers
[8,9] did a lot of experiments to investigate the effects of lead crowning and misalignment errors of gear shafts
on tooth-contact strength and root bending strength. Their researches showed that lead crowning and mis-
alignment errors weakened the tooth contact strength and root bending strength. Nieman and Richter [10–
12], Kalkert [13], Feltkamp and Rademacher [14,15] also investigated the effects of gear machining errors
on tooth-contact strength by experiments. They firstly made gears with four different methods: grinding, shap-
ing, hob cutting and lapping, then tested tooth contact strength of the gears. Their test results showed that ME
weakened tooth-contact strength of the gears. Fijita and his colleagues [16] also studied the effects of lead
crowning on TRBS. But their measured root stress distributions along the lead are not agreement with the
shape of lead crowning curves well. Though so many researches have been conducted so far, an effective
theoretical method that can precisely calculate the effects of ME, AE and TM on gear’s loading capacity,
load-sharing ratio and transmission error cannot be found now.
This paper is the second step of the study to make out the effects of ME, AE and TM on gear’s loading
capacity and transmission performances in theory. At the first step, special finite element methods and the
responsive FEM software [17] are developed to conduct loaded tooth contact analyses (LTCA) and TSCS
and TRBS calculations of a pair of spur gears with ME, AE and TM. As the second step, this paper uses this
FEM software to investigate the effects of ME, AE and TM on loading capacity, LSR and TE of a pair of spur
gears quantitatively in theory. TSCS and TRBS of a pair of spur gears with ME, AE and TM are analyzed
precisely. Influence factors of ME, AE and TM on TSCS and TRBS are also defined and calculated quanti-
tatively. Influence factors calculated by this paper can be used as references when ISO standards [18–20] are
used for tooth surface contact strength and root bending strength calculations for a pair of spur gears with
ME, AE and TM.
Methods for LSR calculation of a pair of spur gears with ME, AE and TM are also presented in this paper.
Effects of ME, AE and TM on LSR are analyzed quantitatively. It is found that ME has great effect on LSR.
Methods for TE calculation of a pair of spur gears with ME, AE and TM are also presented in this paper.
Effects of ME, AE and TM on TE of gears are also analyzed quantitatively. It is found that ME not only
changes waveform of TE easily, but also makes TE greater when ME is increased. AE and TM also make
TE greater when they are increased, but do not change waveform of TE.

2. Definitions of assembly errors, machining errors and tooth modifications

2.1. Assembly errors (AE)

Fig. 1 is used to define AE of a pair of parallel-shaft gears. In Fig. 1(a), A0B0 and CD are shaft positions of
a pair of parallel-shaft spur gears (Gear 1 and Gear 2) without AE. A0, B0, C, and D are four bearing positions
to support the shafts A0B0 and CD. In Fig. 1(b), AB is the new position of A0B0 when there are AE. In order to
define AE of the gears, a three-dimensional (3D), coordinate system (A0–XYZ) is built. XA0Y plane is a hor-
izontal plane that passes through A0B0 and CD. It is called S-plane here. YA0Z plane is a vertical plane that
700 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

b
V

Y
Z
B0
B
Φ

S D

ZE A0
A θ

XE
C
X

Fig. 1. Definition of assembly errors: (a) positions of gear shafts without assembly errors, (b) positions of gear shafts with assembly errors
and (c) other coordinate system used to define the misalignment errors ‘‘e2’’ and ‘‘e3’’.

passes through A0B0 and is perpendicular to the S-plane. It is called V-plane here. A0 is the origin. In the coor-
dinate system (A0–XYZ), AE can be defined as center distance error XE on S-plane, center distance error ZE
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 701

on V-plane, misalignment error h on S-plane and misalignment error / on V-plane as shown in Fig. 1(b).
When XE, ZE, h and / are given, AB is positioned in the coordinate system (A0–XYZ).
Fig. 1(c) is the other 3D coordinate system used in this paper. The plane of action of the pair of spur gears is
used as a horizontal coordinate plane and misalignment error of the gear shafts on this plane is expressed as
‘‘e3’’. The vertical plane of the plane of action is used as the vertical coordinate plane and misalignment error
of the gear shafts on this plane is expressed as ‘‘e2’’. Effects of ‘‘e2’’ and ‘‘e3’’ on TSCS, TRBS, LSR and TE
shall be investigated in this paper.

2.2. Machining errors (ME)

ME of gears defined as 3D shape deviation of tooth surfaces in this paper. That is to say, ME of the gears is
the 3D difference between a machined tooth surface and the theoretical one in the direction of the line of
action. This 3D shape deviation (difference) includes profile deviation, helix deviations and pitch deviations
of the gears. Fig. 2 is an example of the measured 3D ME of the wheel as shown in Table 1. This wheel is
cut by a hob under accuracy requirement of JIS 5th grade. In Fig. 2, the horizontal axis is number of mea-
surement points within the lead, the longitudinal axis is number of measurement points within the tooth pro-
file and the vertical axis is 3D shape deviation of ME distributed on entire the tooth surface of the wheel. The
maximum value of the 3D deviation on entire the tooth surface is expressed as ‘‘MEmax’’. In Fig. 2,
MEmax = 23 lm.
Fig. 2 shall be used to investigate the effects of ME on TSCS, TRBS, LSR and TE of the pair of spur gears
given in Table 1. ME in Fig. 2 is enlarged and shortened by multiplying a coefficient k. When k = 10/23, 20/23,
30/23 and 40/23 are used separately, Fig. 2 becomes Fig. 3(a)–(d). In Fig. 3(a)–(d), MEmax becomes 10, 20, 30
and 40 lm separately. Fig. 3(a)–(d) shall be used in LTCA to investigate the effects of ME on TSCS, TRBS,
LSR and TE.
Fig. 4 is an image of the 3D shape deviation in a 2D section plane that is parallel to the end face of a gear.
In Fig. 4, if the machined profile is on the outside of the theoretical profile, then ME is defined as positive,
otherwise negative.

2.3. Tooth modifications (TM)

Lead crowning and end relief are usually used for tooth modifications of a gear. In Fig. 5, one kind of lead
crowning curve and two kinds of end relief curves are given. An arc is used for lead crowning. One arc with
two straight lines and three straight lines are used for end relief. ‘‘Q’’ is used to express the maximum quantity
of lead crowning and end relief. Effects of the quantity Q of lead crowning on TSCS, TRBS, LRS and TE of
the gears shall be investigated in this paper.

Hob-cut gear

ME max=23 μ m

Tip
35
e
25
30
o fil
pr
Shape deviation μ m

25
20 in
20 i th
15 15 tsw
o in
10 10 tp
5 5 en
em
0 Root s ur
5 10 15 20 25 30 35 ea
M
Measurement points within lead
Fig. 2. Measured machining error (hob-cutting gear).
702 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

Table 1
Parameters of gears used as research objects
Pressure angle = 20°, module m = 2 mm, Contact ratio = 1.502, Edge radius of cutter = 0.375 m, Material = SCM415
Involute spur gears Pinion Wheel
Tooth number 16 47
Shifting coefficient +0.5 0.5
Face width (mm) 12 12
Accuracy (JIS) 0 grade 5th grade
Torque (kg m) 1.0 2.94
Heat-treatment Carbonizing, Hardness = 630 HV

a ME max=10 μ m
b ME max=20 μ m

35 35
e ile
40 30 o fil 40 30
of
Shape deviation μ m

Shape deviation μ m
pr 25 pr
25 in in
30 th 30 th
20
s wi 20
s wi
20 15 i nt 20 15 int
10 t po 10 t po
10 en 10 en
5 em 5 re
m
s ur su
0
5 10 15 20 25 30 35 ea 0
5 10 15 20 25 30 35 ea
M M
Measurement points within lead Measurement points within lead

c d ME max=40 μ m
ME max=30 μ m

35 35
f ile ile
30
ro 40 30 of
Shape deviation μ m

Shape deviation μ m

40
p pr
25 in 25 in
30 i th 30
20 i th
sw sw
20
nt nt
20 15
poi 20 15
p oi
10
e nt 10
e nt
10 m 10 m
5
u re 5
u re
0 e as 0 e as
5 10 15 20 25 30 35 M 5 10 15 20 25 30 35 M
Measurement points within lead Measurement points within lead

Fig. 3. Machining errors used in LTCA (a) MEmax = 10 lm, (b) MEmax = 20 lm, (c) MEmax = 30 lm and (d) MEmax = 40 lm.

3. Theoretical methods for LTCA, stresses, LSR and TE calculations

3.1. Face-contact model of a pair of gears

Fig. 6 is a face-contact model used for LTCA of a pair of spur gears in this paper. Engagement of a pair of
spur gears (Gear 1 and Gear 2) on the geometric contact line is shown in Fig. 6(b). Fig. 6(a) is a 3D view of the
engaged tooth surface of Gear 1. In Fig. 6(a), loaded tooth contact is assumed to be on a reference face with a
contact width Width. Of course, this reference face is a part of the tooth surface of Gear 1 and the geometric
contact line is located at the center of this reference face. In Fig. 6(a), many lines parallel to the geometric
contact line are made artificially. These lines (including the geometric contact line) are called reference lines
and many points on the reference lines are made artificially. These points are called contact reference points,
or simply say reference points. These reference points shall be used as contact points in LTCA. That is to say,
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 703

Fig. 4. Definition of machining errors of gears.

Face width
Straight line
End relief (1)
End relief (2)
Crowning
(Modified Quantity)
Q

Arc

Arc

Fig. 5. Definitions of lead crowning and end relief.

a b

Fig. 6. Face-contact model for loaded tooth-contact analysis.

704 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

tooth-contact on the reference face of Gear 1 shall be replaced by the contacts on the reference points when
FEM is used to perform LTCA. The responsive contact points on tooth surface of Gear 2 are found geomet-
rically according to the positions of the reference points on Gear 1 [17].
At the first step of this research [17], this face-contact model and special FEM were presented to conduct
LTCA of a pair of spur gears with ME, AE and TM. As the second step of the research, this paper shall use
this model and FEM to investigate the effects of ME, AE and TM on TSCS, TRBS, LSR and TE of the gears.

3.2. Mathematical programming method for LTCA

Mathematical programming method is used to conduct LTCA and obtain tooth load distributions of a pair
of spur gears with ME, AE and TM. The following is mathematical model of the mathematical programming
method. Since gear transmission has a very high efficiency (about 95–98%), friction between the contact tooth
surfaces is ignored in LTCA.

3.2.1. Mathematical model of mathematical programming method

Objective function
Z ¼ X nþ1 þ X nþ2 þ    þ X nþn þ X nþnþ1 ð1Þ
Constraint conditions
 ½SfF  þ dfeg þ ½IfY g þ ½IfZ 0 g ¼ feg ð2Þ
T
feg fF g þ X nþnþ1 ¼ P ð3Þ
where

½S ¼ ½S kj  ¼ ½akjð1Þ þ akjð2Þ ; k ¼ 1; 2; :::; n; j ¼ 1; 2; . . . ; n

fZ g ¼ fX nþ1 ; X nþ2 ; . . . ; X nþn gT
0

T
fF g ¼ fF 1 ; F 2 ; . . . ; F k ; . . . ; F n g
fY g ¼ fY 1 ; Y 2 ; . . . ; Y k ; . . . ; Y n gT
T
feg ¼ fe1 ; e2 ; . . . ; ek ; . . . ; en g
F k P 0; Y k P 0; ek P 0; d P 0; k ¼ 1; 2; . . . ; n
X nþm P 0; m ¼ 1; 2; . . . ; n þ 1

In above equations, akj(1) and akj(2) are deformation influence coefficients of the pairs of contact points on en-
gaged tooth surfaces of Gear 1 and Gear 2 separately. They are calculated by 3D FEM. P is total load of a
pair of spur gears along the line of action. P can be calculated with Eq. (4). In Eq. (4), rb is radius of gear base
circle. {e} is gap array that consists of all pairs of contact points on the reference face as shown in Fig. 6(a). {e}
can be calculated geometrically. {F} is a array of contact loads between the pairs of contact points. d is relative
deformation of the pair of gears along the line of action. When akj(1), akj(2), {e} and P are known, {F} and d
can be calculated by solving the Eqs. (1)–(3) with the modified simplex method of mathematical programming
principle. Detailed procedures about the LTCA can be found in Ref. [17].
P ¼ Transmitted Torque=rb ð4Þ

3.3. Tooth-contact stress and root bending stress calculations

3.3.1. Calculations of TSCS

Hertz formula is often used to calculate TSCS of gears when tooth load is known. But it is difficult to be
used here for a pair of gears with AE, ME and TM. So this paper calculates the TSCS of gears with a ‘‘Unit
Force’’ method. It means to calculate tooth load distributed on unit contact area of a tooth surface. That is to
say, dividing the tooth load with a contact area on the reference face as shown in Fig. 6(a).
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 705

Double pair tooth-contact area

Outer limit
Single pair tooth-contact area
Inner limit

Double pair tooth-contact area

Fig. 7. Tooth-contact areas and limits.

For the pair of gears as shown in Table 1, TSCS at the outer limit of the single pair tooth-contact area as
shown in Fig. 7 is greater than the one at the inner limit [17]. So, TSCS at the outer limit is calculated when the
effects of ME, AE and TM on TSCS are investigated in this paper.

3.3.2. Calculations of TRBS

When tooth load distributions are obtained in LTCA, TRBS are calculated with 3D FEM. For gears hav-
ing contact ratios in the range 1 < e < 2, the maximum tensile root stress happens at outer limit contact of the
gears. Tooth engagement at the outer limit as shown in Fig. 7 is often called the ‘‘worst load position’’ and this
position is used for TRBS calculation of the gears. Usually, a so-called tangency point of 30° angle on the root
fillet at tensile side of the tooth has the maximum tensile bending stress. This point is called critical stress point
and the tensile bending stress of this point along the surface of the root fillet is called critical root stress (CRS).
CRS is used to evaluate tooth-bending strength of a pair of spur gears. In this paper, CRS is also used to eval-
uate tooth bending strength when effects of ME, AE and TM on TRBS are investigated.

3.4. Calculations of LSR and TE

In order to calculate tooth LSR and TE of the pair of gears as shown in Table 1, one meshing period of the
pair of gears are divided into 12 engagement positions as shown in Table 2 based on the contact ratio of the
pair of gears (contact ratio = 1.502). In Table 2, 1–6 are double pair tooth-contact positions and 7–12 are sin-
gle pair tooth-contact positions. Relative time of every position within one meshing period is also given in
Table 2. In Table 2, positions 6 and 7 have the same time, but different pairs of tooth-contact. LTCA are

Table 2
Tooth engagement positions used for LTCA
Engagement position Time within one meshing period Pairs of tooth-contact
1 0 Double pair tooth-contact Tooth root of the wheel
2 0.1004 Double pair tooth-contact
3 0.2008 Double pair tooth-contact
4 0.3012 Double pair tooth-contact
5 0.4016 Double pair tooth-contact
6 0.5020 Double pair tooth-contact Inner limit of the wheel
7 0.5020 Single pair tooth-contact Inner limit of the wheel
8 0.6016 Single pair tooth-contact
9 0.7012 Single pair tooth-contact
10 0.8008 Single pair tooth-contact
11 0.9004 Single pair tooth-contact
12 1 Single pair tooth-contact Outer limit of the wheel
706 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

performed at all the 12 positions separately when ME, AE and TM are considered. TE are calculated by Eq.
(5) when d in Eq. (2) is available in LTCA:
Transmission error ¼ d=rb ð5Þ
Total load on each tooth surface is calculated when 3D tooth load distribution on each tooth surface is
available in LTCA at double pair tooth-contact positions 1–6. Then tooth LSR is calculated by Eq. (6):
Tooth load sharing ratio ¼ Total load on each tooth=P ð6Þ

4. Comparisons between calculated results with the measured ones

As introduced in Ref. [17], special FEM software was developed and following experiments have been done
to prove reliability of the special FEM software.
Tooth-contact pattern and root strains of a pair of spur gears as shown in Fig. 8 are measured when assem-
bly errors h = 0.42°, / = 0.04°, XE = 2.1 mm and ZE = 0.2 mm are given to the gear shaft A0B0 as shown
in Fig. 1(b). These gears are ground under the accuracy requirement of JIS 1st grade. A so-called power-

Gear 1 Gear2
Fig. 8. A pair of spur gears (Z1 = Z2 = 50, m = 4, X1 = X2 = 0, a = 20°).

Fig. 9. Test gearbox in power-circulating form gear test rig.

 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 707

circulating form gear test rig is used to do the tests at a very load speed (1.65 rpm) under torque T = 294 N m.
Fig. 9 is the test gearbox used in the test rig. In Fig. 9, a removable support is used to give AE to the shaft of
Gear 1 by changing the position of this support. Fig. 10 is tooth-contact pattern measured. Fig. 11 is the
tooth-contact pattern calculated under the same conditions. In Fig. 11, A is the areas of double pair tooth-
contact and B is the area of single pair tooth-contact. It is found that calculated contact pattern is agreement
with the measured one well. Root strains are also measured. Fig. 12 is comparisons between the measured root
strains and the calculated ones at measurement positions of stain gauges 1–4 as shown in Fig. 13. It is also
found that calculated root strains are agreement with the measured ones.

Fig. 10. Tooth-contact pattern measured (h = 0.42°, / = 0.04°, XE = 2.1 mm and ZE = 0.2 mm; Torque = 294 N m).

Fig. 11. Tooth-contact pattern calculated.

200

0
Root strains ε

-200
Measured

-400 Calculated
Gauge 1
Gauge 2
-600 Gauge 3
Gauge 4
-800
0 200 400 600 800 1000
Measurement point number when gear is rolling

Fig. 12. Root strain comparisons.

708 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

Fig. 13. Strain gauge positions.

Tooth-contact lengths and contact stresses of a pair of spur gears with lead crowning were also calculated
by the special FEM software and compared with measurement results in Ref. [17]. It was also found that the
calculated results were agreement with the measured ones well.

5. Gears used as research object

Gears used as research objects are given in Table 1. In Table 1, the wheel is cut by a hob under accuracy
requirement of JIS 5th grade. Fig. 2 is the measured 3D machining errors distributed on entire the tooth sur-
face of the wheel. The pinion is ground under accuracy requirement of JIS 0 grade. So machining errors of the
pinion are ignored in the calculations. Torque conditions are also given in Table 1. These torques are used for
all the calculations in the following.

Fig. 14. FEM models.

 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 709

Fig. 14 is FEM models of the pair of gears used in LTCA. Fig. 14(a) is FEM mesh-dividing pattern of the
wheel. The pinion is also divided similarly. Fig. 14(b) is FEM model in the part of tooth-contact. Four pairs of
teeth are shown. The reference face as shown in Fig. 6(a) is fine divided with 48 meshes within the contact
width ‘‘Width’’ and 20 meshes within the face width. Nodes on inner hole surface (hub of the gears) as shown
in Fig. 14(a) are fixed as the boundary conditions of FEM when deformation influence coefficients and TRBS
are calculated.

6. Effects of ME, AE and TM on tooth-contact strength

As it has been stated that the outer limit of the single pair tooth-contact is used as the ‘‘worst load position’’
to do LTCA and TSCS calculations of the pair of gears as shown in Table 1. Fig. 15 is contour lines of cal-
culated TSCS when there are no ME, AE and TM. This TSCS is distributed on the reference face as shown in
Fig. 6(a). In Fig. 15, the horizontal axis (X-axis) is face width of the gears and the vertical axis (Y-axis) is con-
tact width Width as shown in Fig. 6(a). Geometrical contact line of the gears is located at the center of contact
width (Y = 0). Y < 0 is the side from the geometrical contact line to the root, Y > 0 is the side from the geo-
metrical contact line to the tip. From Fig. 15, it is found that contact stress distributions are almost the same
along the lead when there are no ME, AE and TM.

-0.15 No ME, AE and TM

-0.10
Contact width mm

-0.05 1.6E2 2.4E2 3.2E2

80
4.8E2 4E2 6.4E2
7.2E2 5.6E2
0.00 7.2E2
5.6E2
6.4E2 4E2 4.8E2
3.2E2 2.4E2
0.05 80 1.6E2

0.10 Geometrical contact line

Contact stress (max)=768 MPa
0.15

0 2 4 6 8 10 12
Face width mm

Fig. 15. Contour lines of tooth-contact stress.

MEmax=10 μ m

- 0.15
Machining error μ m

40 - 0.10
- 0.05 m
30 m
0.00 th
id
20 0.05
tw
0.10
n tac
10 0.15 Co
0 2 4 6 8 10 12
Face width mm
Fig. 16. Machining errors on reference face.
710 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

6.1. Effect of ME on TSCS

LTCA and TSCS calculations of the pair of spur gears as shown in Table 1 are performed at the outer limit
separately when MEmax = 0, 10, 20, 30 and 40 lm. Figs. 16 and 17 are ME of the wheel distributed on the
reference face as shown in Fig. 6(a) when MEmax = 10 and 40 lm. Figs. 18 and 19 are contour lines of TSCS
when MEmax = 10 and 40 lm. It is found that ME exerts great effect on TSCS distribution when Figs. 18 and
19 are compared with Fig. 15. Fig. 20 is a relationship between the maximum TSCS and MEmax. It is found
that the maximum TSCS becomes greater when MEmax is increased. Influence factor of ME is introduced
here to evaluate the effect of ME on TSCS quantitatively. So, it can be calculated through dividing the max-
imum TSCS of the gears when there are ME by the maximum TSCS of the gears when there are no ME, AE
and TM. Relationship between the influence factor of ME and MEmax is also shown in Fig. 20. From Fig. 20,
it is found that the influence factor of ME arrives at 2.2, a quite large number, when MEmax = 40 lm. Fig. 20
can be used as references when ISO 6336/2 [19] is used to calculate TSCS of a pair of spur gears with ME.

6.2. Effect of AE on TSCS

6.2.1. Effect of e3 on TSCS

LTCA and TSCS calculations of the pair of spur gears in Table 1 are performed at the outer limit separately
when misalignment error e3 = 0°, 0.01°, 0.02°, 0.03° and 0.04°. Figs. 21 and 22 are contour lines of TSCS

- 0.15
Machining error μ m

40 - 0.10
- 0.05 m
30 m
0.00 th
id
20 0.05
tw
0.10 t ac
MEmax=40 μ m n
10 0.15 Co
0 2 4 6 8 10 12
Face width mm
Fig. 17. Machining errors on reference face.

- 0.15 MEmax=10 μ m

- 0.10
Contact width mm

- 0.05 2.4E2
1.1E3
1.2E2 4.8E2 6E2
0.00 8.4E2 9.6E2
7.2E2 4.8E2
1.2E2 3.6E2
0.05

0.10 Geometrical contact line

Contact stress (max)=1101 MPa
0.15

0 2 4 6 8 10 12
Face width mm
Fig. 18. Contour lines of tooth-contact stress.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 711

- 0.15 MEmax=40 μ m
5.4E2

- 0.10
1.1E3

Contact width mm
9E2 3.6E2
- 0.05 1.6E3
1.3E3
7.2E2 1.4E3
0.00 1.8E2

0.05
Geometrical contact line
0.10
Contact stress (max)=1709 MPa
0.15
0 2 4 6 8 10 12
Face width mm
Fig. 19. Contour lines of tooth-contact stress.

1800
2.4
1600 2.2
Contact stress (max) MPa

1400 2.0

Influence factor
1.8
1200
1.6
1000 1.4

800 Contact stress 1.2

Influence factor 1.0
600
0 10 20 30 40
Max machining error (MEmax) μ m

Fig. 20. Contact stress and influence factor versus MEmax.

- 0.15 Misalignment error e3=0.02°

- 0.10
Contact width mm

1.1E2
2.2E2
- 0.05 3.3E2

0.00 4.4E25.5E2
6.6E2
7.7E2
8.8E2 9.9E2
2.2E2
0.05 1.1E2 3.3E2

0.10
Geometrical contact line
0.15 Contact stress (max)=1013 MPa

0 2 4 6 8 10 12
Face width mm
Fig. 21. Contour lines of tooth-contact stress.

when e3 = 0.02° and 0.04°. It is found that e3 has great effect on TSCS distribution when Figs. 21 and 22 are
compared with Fig. 15. Fig. 23 is a relationship between the maximum TSCS and e3. It is found that the max-
imum TSCS becomes greater when e3 is increased. Influence factor of e3 is introduced here to evaluate the
effect of e3 on TSCS quantitatively. So it can be calculated through dividing the maximum TSCS of the gears
712 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

- 0.15 Misalignment error e3=0.04°

- 0.10

Contact width mm
4.8E2
- 0.05 1.2E2

0.00 7.2E2 8.4E2 9.6E2 1.1E3

2.4E2 3.6E2 6E2
0.05

0.10
Geometrical contact line
0.15 Contact stress (max)=1198 MPa
0 2 4 6 8 10 12
Face width mm
Fig. 22. Contour lines of tooth-contact stress.

1200 1.6

1.5
Contact stress (max) MPa

1100

Influence factor
1.4
1000
1.3

900 1.2

Contact stress 1.1

800
Infulence factor
1.0
700
0.00 0.01 0.02 0.03 0.04
Misalignment error e3

Fig. 23. Contact stress and influence factor versus misalignment error e3.

when there are e3 by the maximum TSCS of the gears when there are no ME, AE and TM. Relationship
between influence factor of e3 and e3 is also shown in Fig. 23. It is found that the influence factor is about
1.6 when e3 = 0.04°. Fig. 23 can also be used as references when ISO 6336/2 is used to calculate TSCS of a
pair of spur gears with the misalignment error e3.

6.2.2. Effect of e2 on TSCS

LTCA and TSCS calculations of the pair of spur gears as shown in Table 1 are performed at the outer limit
separately when the misalignment error e2 = 0°, 0.1°, 0.2°, 0.3° and 0.4°. Here e2 is 10 times greater than e3.
Figs. 24 and 25 are contour lines of TSCS when e2 = 0.1° and e2 = 0.4°. It is found that e2 has little effect on
tooth-contact pattern and the maximum TSCS when Figs. 24 and 25 are compared with Fig. 15. Fig. 26 is a
relationship between the maximum TSCS and e2. It is found that the maximum TSCS is increased very little
with the increment of e2. Influence factor of e2 is introduced to evaluate the effect of e2 on TSCS quantita-
tively. So it can be calculated through dividing the maximum TSCS when there are e2 by the maximum TSCS
when there are no ME, AE and TM. Relationship between the influence factor of e2 and e2 is also shown in
Fig. 26. From Fig. 26, it is found that the influence factor of e2 almost has no change even if e2 = 0.4°. So,
effect of the misalignment error e2 on TSCS can be ignored when tooth surface contact strength calculations
are performed.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 713

- 0.15 Misalignment error e2=0.1°

- 0.10

Contact width mm
- 0.05
4.8E2 3.2E2 2.4E2 1.6E2 80 4E2
6.4E2
5.6E2
0.00 7.2E2
5.6E2 6.4E2 7.2E2
4.8E2
2.4E2 4E2 1.6E2 3.2E2
0.05 80

0.10
Geometrical contact line
0.15 Contact stress (max)=772 MPa

0 2 4 6 8 10 12
Face width mm
Fig. 24. Contour lines of tooth-contact stress.

- 0.15 Misalignment error e2=0.4°

- 0.10
Contact width mm

- 0.05 2.4E2
80
5.6E2 4E2 1.6E2
0.00 3.2E2
7.2E2 6.4E2 4.8E2
5.6E2
0.05 6.4E2
3.2E21.6E2
80 2.4E2 4E2
4.8E2
0.10
Geometrical contact line
0.15 Contact stress (max)=774 MPa

0 2 4 6 8 10 12
Face width mm

Fig. 25. Contour lines of tooth-contact stress.

1.010
774
1.008
Contact stress (max) MPa

773
Influence factor

772 1.006

771
1.004
770
Contact stress
1.002
769 Influence factor

768 1.000

0.0 0.1 0.2 0.3 0.4

Misalignment error e2

Fig. 26. Contact stress and influence factor versus misalignment error e2.

6.3. Effect of TM on TSCS

LTCA and TSCS calculations of the pair of spur gears as shown in Table 1 are performed at the outer limit
separately when the quantity Q of lead crowning are 0, 10, 20, 30 and 40 lm. Figs. 27 and 28 are contour lines
714 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

of TSCS when Q = 10 and 30 lm. It is found that lead crowning has great effect on TSCS distribution when
Figs. 27 and 28 are compared with Fig. 15. Fig. 29 is a relationship between the maximum TSCS and Q. From
Fig. 29, it is found that the maximum TSCS becomes greater when Q is increased. Influence factor of lead
crowning is introduced here to evaluate the effect of lead crowning on TSCS quantitatively. So it can be

- 0.15 Lead crowning Q=10 μ m

- 0.10
Contact width mm
- 0.05
1.2E2
1.1E3 8.4E2
0.00 9.6E2
4.8E2
0.05 6E2 7.2E2 3.6E2
2.4E2

0.10
Geometrical contact line
0.15 Contact stress (max)=1171 MPa
0 2 4 6 8 10 12
Face width mm
Fig. 27. Contour lines of tooth-contact stress.

- 0.15 Lead crowning Q=30 μ m

- 0.10
Contact width mm

- 0.05
8E2
0.00 1.4E3 4.8E2

0.05 1.1E3 1.3E3 3.2E2

9.6E2
1.6E26.4E2

0.10
Geometrical contact line
0.15 Contact stress (max)=1448 MPa
0 2 4 6 8 10 12
Face width mm
Fig. 28. Contour lines of tooth-contact stress.

1600
2.0
1400
Contact stress (max) MPa

1.8
Influence factor

1200 1.6

1000 1.4

Contact stress 1.2

800
Modification factor
1.0
600
0 10 20 30 40
Quantity of lead crowning μ m

Fig. 29. Contact stress and influence factor versus lead crowning.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 715

calculated through dividing the maximum TSCS when there are lead crowning by the maximum TSCS when
there are no ME, AE and TM. Relationship between the influence factor and Q is also shown in Fig. 29. From
Fig. 29, it is found that influence factor of lead crowning is over 2.0 when Q = 40 lm. Fig. 29 can also be used
as references when ISO 6336/2 is used to calculate TSCS of a pair of spur gears with lead crowning.

7. Effects of ME, AE and TM on tooth root bending strength

7.1. Effect of ME on TRBS

LTCA and TRBS calculations of the pair of spur gears as shown in Table 1 are performed at the ‘‘worst
load position’’ separately when MEmax = 0, 10, 20, 30 and 40 lm. Fig. 30 is the results of critical root
stress (CRS) distributions along tooth longitudinal direction. In Fig. 30, it is found that ME has great effect
on TRBS. Fig. 31 is a relationship between the maximum CRS and MEmax. From Fig. 31, it is found that
the maximum CRS becomes greater when MEmax is increased. Influence factor of ME is introduced here
to evaluate the effect of ME on TRBS quantitatively. So it can be calculated through dividing the maxi-
mum CRS of the pair of gears when there are ME by the maximum CRS of the gears without ME,
AE and TM. Relationship between the influence factor of ME and MEmax is also shown in Fig. 31. It
is found that the influence factor of ME on TRBS is about 1.5 when MEmax = 40 lm. Fig. 31 can also
be used as references when ISO 6336/3 [20] is used to calculate the maximum CRS of a pair of spur gears
with ME.

120

100
Root bending stress MPa

80

60
MEmax=0 μ m
40 MEmax=10 μ m
MEmax=20 μ m
20 MEmax=30 μ m
MEmax=40 μ m
0
0 2 4 6 8 10 12
Face width mm
Fig. 30. Root bending stress.

120
1.5
Root bending stress (max) MPa

110 1.4
Infulence factor

1.3
100

1.2
90
Root stress 1.1
Influence factor
80 1.0

0 10 20 30 40
Max machining error (MEmax) μ m

Fig. 31. Root bending stress and influence factor versus MEmax.
716 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

7.2. Effect of AE on TRBS

7.2.1. Effect of e3 on TRBS

LTCA and TRBS calculations of the pair of spur gears as shown in Table 1 are performed at the ‘‘worst
load position’’ separately when the misalignment error e3 = 0°, 0.01°, 0.02°, 0.03° and 0.04°. Fig. 32 is the
results of CRS distributions along tooth longitudinal direction. From Fig. 32, it is found that e3 has great
effect on TRBS. Fig. 33 is a relationship between the maximum CRS and e3. From Fig. 33, it is found that
the maximum CRS becomes greater when e3 is increased. Influence factor of e3 is introduced here to evaluate
effect of e3 on the maximum CRS quantitatively. So it can be calculated through dividing the maximum CRS
of the pair of gears with misalignment error e3 by the maximum CRS of the gears without ME, AE and TM.
Relationship between the influence factor of e3 and e3 is also shown in Fig. 33. It is found that influence factor
of e3 is about 1.75 when e3 = 0.04°. Fig. 33 can also be used as references when ISO 6336/3 is used to calculate
the maximum CRS of a pair of spur gears with the misalignment error e3.

7.2.2. Effect of e3 on TRBS

LTCA and TRBS calculations of the pair of spur gears as shown in Table 1 are performed at the ‘‘worst
load position’’ separately when the misalignment error e2 = 0, 0.1, 0.2, 0.3 and 0.4. Here e2 is 10 times greater
than e3. Fig. 34 is the results of CRS distributions along tooth longitudinal direction. From Fig. 34, it is found
that e2 almost has no effect on TRBS. Fig. 35 is a relationship between the maximum CRS and e2. From
Fig. 34, it is found that the maximum CRS becomes very little greater when e2 is increased. Influence factor

160 Misalignment error e3

Root bending stress MPa

140 e3=0
e3=0.01°
120 e3=0.02°
100 e3=0.03°
e3=0.04°
80
60
40
20
0
0 2 4 6 8 10 12
Face width mm
Fig. 32. Root bending stress.

150 1.9
Root bending stress (max) MPa

140 1.8
1.7
130
1.6
Infulence factor

120 1.5
110 1.4

100 1.3
1.2
90
Root stress 1.1
80 Influence factor 1.0
70 0.9
0.00 0.01 0.02 0.03 0.04
Misalignment error e3 degree

Fig. 33. Root bending stress and influence factor versus misalignment error e3.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 717

80

Root bending stress MPa

75

70
Misalignment error e2
65 e2=0
e2=0.1°
60 e2=0.2°
e2=0.3°
55 e2=0.4°

50
0 2 4 6 8 10 12
Face width mm
Fig. 34. Root bending stress.
Root bending stress (max) MPa

79.6
1.015
79.4

Infulence factor
79.2 1.010

79.0

78.8 1.005

78.6 Root stress

1.000
78.4 Influence factor

0.0 0.1 0.2 0.3 0.4

Misalignment error e2 degree

Fig. 35. Root bending stress and influence factor versus misalignment error e2.

of e2 is introduced here to evaluate effect of e2 on the maximum CRS quantitatively. So it can be calculated
through dividing the maximum CRS of the pair of gears with e2 by the maximum CRS of the gears without
ME, AE and TM. Relationship between the influence factor of e2 and e2 is also shown in Fig. 35. It is found
that influence factor of e2 is about 1.015 even if e2 = 0.4°. So, effect of the misalignment error e2 on TRBS can
be ignored when tooth root bending strength calculations are performed.

7.3. Effect of TM on TRBS

LTCA and TRBS calculations of the pair of spur gears as shown in Table 1 are performed at the ‘‘worst
load position’’ separately when the quantity Q of lead crowning are 0, 5, 10, 20, 30 and 40 lm. Fig. 36 is the
results of CRS distributions along tooth longitudinal direction. From Fig. 36, it is found that lead crowning
has great effect on TRBS distributions even if Q = 5 lm. Fig. 37 is a relationship between the maximum CRS
and Q. From Fig. 37, it is found that the maximum CRS becomes greater when Q is increased. Influence factor
of lead crowning is introduced here to evaluate the effect of lead crowning on the maximum CRS quantita-
tively. So it can be calculated through dividing the maximum CRS of the pair of gears with lead crowning
by the maximum CRS of the gears without ME, AE and TM. Relationship between the influence factor
and Q is also shown in Fig. 37. It is found that the influence factor of lead crowning is over 1.4 when
Q = 40 lm. Fig. 37 can also be used as references when ISO 6336/3 is used to calculate the maximum CRS
of a pair of spur gears with lead crowning.
718 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

120
110

Root bending stress MPa

100
90
80
70
Crowning=0
60 Crowning=5 μ m
50 Crowning=10 μm
Crowning=20 μm
40
Crowning=30 μm
30 Crowning=40 μm
20
0 1 2 3 4 5 6 7 8 9 10 11 12
Face width mm

Fig. 36. Root bending stress.

120
1.5
Root bending stress MPa

110
1.4

Influence factor
100 1.3

90 1.2

Root bending stress 1.1

80 Influence factor
1.0
70
0 5 10 15 20 25 30 35 40
Quantity of lead crowning μm

Fig. 37. Root bending stress and influence factor versus lead crowning.

8. Effects of ME, AE and TM on load-sharing ratio

8.1. Effect of ME on LSR

LTCA and tooth load calculations are performed separately at all the 12 engagement positions as shown in
Table 2 when MEmax = 0, 10, 20, 30 and 40 lm. Then LSR are calculated separately. Fig. 38 is a section view
of 3D, FEM model of the pair of gears at the engagement position 3. Since the Position 3 is a double pair
tooth-contact position as shown in Table 2, two pairs of teeth share the total load P of the gears as shown
in Fig. 38. Figs. 39 and 40 are contour lines of contact stress distributions on tooth 1 and tooth 2 separately
when MEmax = 0. Figs. 41 and 42 are contour lines of contact stress distributions on tooth 1 and tooth 2
separately when MEmax = 10 lm and Figs. 43 and 44 are contour lines of contact stress distributions on
tooth 1 and tooth 2 separately when MEmax = 30 lm. By comparing Figs. 41 and 42 with Figs. 39 and
40, also Figs. 43 and 44 with Figs. 39 and 40, it is found that ME has great effect on tooth-contact pattern.
Fig. 45 is calculated tooth LSR. In Fig. 45, the abscissa is tooth engagement position. Relative time as shown
in Table 2 is used to express tooth engagement positions. From Fig. 45, it is found that ME exerts great effect
on tooth LSR too. In Figs. 45, 1 and 2 should be the double pair tooth-contact positions, but they become
single pair tooth-contact positions when MEmax is greater than 10 lm and only tooth 2 shares the total load
P of the gears. This means that the outer limit of single pair tooth-contact area cannot be used as the ‘‘worst
load position’’ for tooth bending strength calculation when ME is large. It is reasonable and safe to use tooth
tip as the ‘‘worst load position’’ when ME is large. Fig. 46 is a relationship between LSR and MEmax at Posi-
tion 3. From Fig. 46, it is found that LSR is decreased from 0.55 to 0.18 when MEmax = 40 lm.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 719

Fig. 38. Teeth engagement at Position 3.

-0.10
Tooth 1
MEmax=0
Position 3
-0.05
55 1.1E2
Contact width mm

2.2E2 1.7E2 2.8E2

3.9E2 3.3E2 4.4E2
5E2
0.00 5E2
4.4E2
3.3E2 3.9E2
2.2E2 1.1E2 2.8E2
55 1.7E2
0.05
Geometrical contact line
Contact stress (max)=511 MPa
0.10
0 2 4 6 8 10 12
Face width mm
Fig. 39. Contact stress distribution on tooth 1.

Tooth 2
-0.10 MEmax=0
Position 3
Contact width mm

-0.05 45 90
1.8E2 1.4E2
2.3E2 2.7E2
3.2E2
3.6E2
0.00 4.1E2 4.1E2
3.2E2 3.6E2
2.3E2 2.7E2
1.8E2 1.4E2 90 45
0.05
Geometrical contact line
0.10 Contact stress (max)=400MPa

0 2 4 6 8 10 12
Face width mm
Fig. 40. Contact stress distribution on tooth 2.
720 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

-0.10
Tooth 1
MEmax=10 μ m
-0.05 Position 3

Contact width mm
Geometrical contact line
4E2
2.4E2 80
0.00 3.2E2 6.4E2
4.8E2
5.6E2
1.6E2 5.6E2 7.2E2
0.05

Contact stress (max)=747 MPa

0.10
0 2 4 6 8 10 12
Face width mm
Fig. 41. Contact stress distribution on tooth 1.

Tooth 2
-0.10 MEmax=10µm 1.4E2
4.2E2
Position 3
Contact width mm

6.3E2
4.9E2
-0.05 70
5.6E2 5.6E2
2.8E2

0.00 3.5E2
70
2.1E2
0.05
Geometrical contact line
Contact stress (max)=694MPa
0.10

0 2 4 6 8 10 12
Face width mm
Fig. 42. Contact stress distribution on tooth 2.

Tooth 1
- 0.10 MEmax=30 μm
Position 3
Contact width mm

- 0.05
Contact stress (max)=908 MPa
0.00
1E2

Geometrical contact line 6E2 4E2

0.05 2E2
8E2
5E2
0.10 3E2
7E2

0 2 4 6 8 10 12
Face width mm
Fig. 43. Contact stress distribution on tooth 1.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 721

- 0.2 Tooth 2 1E2

MEmax=30 μm 4E2 3E2 9E2
7E2
Position 3

Contact width mm
6E2 5E2
- 0.1 8E2
2E2
6E2

0.0

0.1 Geometrical contact line

Contact stress (max)=981 MPa
0.2

0 2 4 6 8 10 12
Face width mm
Fig. 44. Contact stress distribution on tooth 2.

Double pair Single pair Double pair

1.0
0.9 7
8 9 10 11
0.8
Load-sharing ratio

12
0.7 3
0.6 2
0.5 5 6
4 MEmax:
0.4 1 0
0.3 10 μ m
0.2 20 μ m
30 μ m
0.1
40 μ m
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 45. Tooth load-sharing ratio.

0.60
0.55
Position 3
0.50
Load-sharing ratio

0.45
0.40
0.35
0.30
0.25
0.20
0.15
0 10 20 30 40
Max machining error (MEmax) μ m

Fig. 46. Load-sharing ratio versus MEmax.

8.2. Effect of AE on LSR

LTCA and LSR calculations are also conducted separately when the misalignment error e3 = 0°, 0.01°,
0.02°, 0.03° and 0.04°. Fig. 47 is the calculation results of LSR and Fig. 48 is a relationship between LSR
722 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

1.0 Misalignment
error e3:
0.9
0

Load-sharing ratio
0.8 0.01°
0.02°
0.7 0.03°
0.04°
0.6

0.5

0.4

0.3
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 47. Load-sharing ratio.

0.585

0.580
Load-sharing ratio

0.575

0.570 Engagement position 3

0.565

0.560

0.555
0.00 0.01 0.02 0.03 0.04
Misalignment error e3 degree

Fig. 48. Load-sharing ratio versus misalignment error e3.

and e3 at Position 3. From Figs. 47 and 48, it is found that the misalignment error e3 almost has no effect on
LSR.

8.3. Effect of TM on LSR

LTCA and LSR calculations are also conducted separately when the quantity Q of lead crowning are 0, 10,
20, 30 and 40 lm. Fig. 49 is the calculation results of LSR and Fig. 50 is a relationship between LSR and Q at

1.0
Lead crowning:
0.9 0
10 μ m
Load-sharing ratio

0.8 20 μ m
30 μ m
0.7 40 μ m
0.6

0.5

0.4
0.3
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 49. Load-sharing ratio.

 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 723

0.585

0.580

Load-sharing ratio
0.575 B

0.570

0.565

0.560

0.555

0 10 20 30 40
Lead crown μ m

Fig. 50. Load-sharing ratio versus lead crowning.

Position 3. From Figs. 49 and 50, it is also found that the quantity Q of lead crowning has very little effect on
LSR.

9. Effects of ME, AE and TM on transmission error

9.1. Effect of ME on TE

LTCA and TE calculations are conducted separately at all the 12 engagement positions as shown in Table 2
when MEmax = 0, 10, 20, 30 and 40 lm. Fig. 51 is the calculation results of TE when radius of base circle of
the pinion is used in Eq. (5). In Fig. 51, the abscissa is tooth engagement position. Relative time as shown in
Table 2 is used to express tooth engagement positions. From Fig. 51, it is found that TE become larger and
larger when MEmax is increased. It is also found that waveform of TE has been changed even if
MEmax = 10 lm.

9.2. Effect of AE on TE

LTCA and TE calculations are conducted separately at all the 12 engagement positions as shown in Table 2
when the misalignment error e3 = 0°, 0.01°, 0.02°, 0.03° and 0.04°. Fig. 52 is the calculation results of TE.
From Fig. 52, it is found that TE become larger and larger when e3 is increased. Fig. 53 is relationships of
the value A and B as shown in Fig. 52 with e3. From Fig. 53, it is found A and B is increased when e3 is
increased and B is suddenly increased when e3 is greater than 0.02°. B is often used to evaluate vibration

0.011
0.010 ME=40 μ m
Transmission error degree

0.009
0.008
ME=30 μ m
0.007
ME=20 μ m
0.006
0.005
ME=10 μ m
0.004
0.003 ME=0

0.002
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 51. Transmission error.

724 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

0.011
Misalignment error e3
0.010

Transmission error degree

0.009 e3=0.04°
0.008 e3=0.03°
0.007 e3=0.02°
0.006
e3=0.01°
0.005
0.004
0.003 B
A
e 3 =0
0.002
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 52. Transmission error.

0.010 0.0021

0.009 0.0020

0.008 0.0019

Value B
Value A

0.0018
0.007
0.0017
0.006
0.0016
0.005 Value A
Value B 0.0015
0.004
0.0014
0.00 0.01 0.02 0.03 0.04
Misalignment error e3 degree

Fig. 53. Value A and B versus misalignment error 3.

0.010
Crowning=40 μ m
Transmission error degree

0.009
Crowning=30 μ m
0.008 Crowning=20 μ m

0.007 Crowning=10 μ m

0.006

0.005

0.004
B
0.003 A
Crowning=0
0.002
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Root Tip
Engagement position of a tooth

Fig. 54. Transmission error.

and noise level of gears roughly. Maybe it can be said that vibration and noise level of gears shall become large
suddenly when e3 exceeds 0.02°.

9.3. Effect of TM on TE

LTCA and TE calculations are conducted separately at all the 12 engagement positions as shown in Table 2
when the quantity Q of lead crowning are 0, 10, 20, 30 and 40 lm. Fig. 54 is the calculation results of TE.
 S. Li / Mechanism and Machine Theory 42 (2007) 698–726 725

0.010
0.0030
0.009

0.008
0.0025

Value A

Value B
0.007

0.006 0.0020
Valute B
0.005 Value A
0.0015
0.004

0 10 20 30 40
Quantity of lead crowning μm

Fig. 55. Value A and B versus lead crowning.

From Fig. 54, it is found that TE become larger and larger when Q is increased. Fig. 55 is relationships of the
value A and B as shown in Fig. 54 with Q. From Fig. 55, it is found A and B is increased when Q is increased.

10. Conclusions

(1) In recent years, it has been strictly required to design a gear with lightweight and compact size. This
makes it necessary for gear designers to know the effects of gear machining errors, assembly errors
and tooth modifications on gear strength design and transmission performances exactly. Since the effects
of machining errors, assembly errors and tooth modifications on gear strength design and transmission
performances have not been made out quantitatively in theory, this paper investigates the effects of
machining errors, assembly errors and tooth modifications on loading capacity, load-sharing ratio
and transmission error of a pair of spur gears quantitatively in theory by using special-developed finite
element method software in a personal computer.
(2) Tooth surface contact stress and tooth root bending stress of a pair of spur gears with machining errors,
assembly errors and tooth modifications are analyzed. Influence factors of machining errors, assembly
errors and tooth modifications on tooth surface contact stress and tooth root bending stress are also
defined and calculated quantitatively. When the maximum tooth surface machining error is 40 lm, influ-
ence factors of machining error on tooth surface contact stress and tooth root bending stress arrive at 2.2
and 1.5. When quantity of the maximum tooth lead crowning is 40 lm, influence factors of the lead
crowning on tooth surface contact stress and tooth root bending stress arrive at 2.0 and 1.4. Influence
factors given in this paper can be used as references when ISO standards are used for tooth surface con-
tact strength and root bending strength calculations of a pair of spur gears with machining errors, assem-
bly errors and tooth modifications. It is suggested that misalignment error of gear shafts on the plane of
action should not be greater than 0.01°; quantity of the maximum tooth lead crowning should not be
greater than 20 lm and machining error on tooth surface should not be greater than 10 lm if you want
your gear with high strengths.
(3) Methods for load-sharing ratio calculation of a pair of spur gears with machining errors, assembly errors
and tooth modifications are also presented in this paper. Effects of machining errors, assembly errors
and tooth modifications on load-sharing ratio are analyzed quantitatively. It is found that machining
errors has great effect on load-sharing ratio of teeth. When the maximum tooth surface machining error
is greater than 10 lm, double pair tooth-contact maybe become single pair tooth-contact for sake of the
effects of machining errors. This means that when machining errors on tooth surface is large, outer limit
of the single pair tooth-contact area cannot be used as the so-called ‘‘worst load position’’. It is safe and
reasonable to use tooth tip as the ‘‘worst load position’’ when to perform tooth bending strength
calculation.
(4) Method for transmission error calculation of a pair of spur gears with machining errors, assembly errors
and tooth modifications are also presented in this paper. Effects of machining errors, assembly errors
726 S. Li / Mechanism and Machine Theory 42 (2007) 698–726

and tooth modifications on transmission error of gears are analyzed quantitatively. It is found that
machining errors not only change waveform of the transmission error easily, but also make transmission
errors greater and greater when the machining errors are increased. Assembly errors and tooth modifi-
cations also make transmission error greater and greater when they are increased, but do not change
waveform of the transmission error.

References

[1] M. Sennba, Gear Errors and Strength, The Nikkan Kogyo Shimbun Press, 1974, pp.70–148 (in Japanese).
[2] H.-G. Hertel, D. Kagerl, D. Greven, Fertigungsgeometrische Aussagen des Tuschiertragbilds an geradverzahnten Stirnradgetrieben,
Fertigungstechn. Betr. 18 (1968) 538–544.
[3] G. Hertel, D. Greven, Informationsgehalt des Tuschiertragbilds von Stirnradgetrieben, bezüglich Achsneigungs-, Achsschränkungs-
und Flankenrichtungsabweichungen, Fertigungstechn. Betr. 17 (1967) 689–695.
[4] J. Rademacher, Ermittlung von Lastverteilungsfaktoren für Stirnradgetrieben, Ind., -Anz. 89 (1967) 331–337.
[5] G. Niemann, W. Richter, Versuchsergebnisse zur Zahnflanken-Tragfähigkeit, VI Abhängigkeit der Zahnflanken-Tragfähigkeit von
Tragbild und Sprungüberdeckung, Konstruktion 12 (1960) 272–278.
[6] G. Niemann, D. Reister, Einseitiges Breitentragen bei geradverzahnten Stirnrädern, Konstruktion 18 (1966) 95–101.
[7] D. Reister, Der Tragfehler bei Geradzahn-Stirnrädern und Möglichkeiten zu seiner Verringerung, antriebstechnik, 6 (1967), 95–101.
[8] J. Timmers, Der Einfluss fertigungstechnisch-und lastbedingter Achsversetzungen in Stirnradgetrieben auf die Zahnverformung, Ind.,
-Anz. 87 (1965) 1771–1778.
[9] H. Opitz, J. Timmers, M. Bosch, J. Rademacher, Möglichkeiten zur Verbesserung des Geräuschverhaltens von Zahnradgetrieben,
Forschungsberichte des Landes Nordrhein-Westfalen (5) (1976) 1867–1873.
[10] G. Niemann, W. Richter, Der Einfluss der Umfangsgeschwindigkeit auf die Zahnflanken-Tragfähigkeit, Konstruktion 12 (1960) 191–
197.
[11] G. Niemann, W. Richter, Einfluss von Modul, Schrägungswinkel und Profilverschiebung auf die Zahnflanken-Tragfähigkeit,
Konstruktion 12 (1960) 236–242.
[12] G. Niemann, W. Richter, Versuchsergebnisse zur Zahnflanken-Tragfähigkeit, V Minderung der Zahnflanken-Tragfähigkeit durch
Verzahnungsfehler, Konstruktion 12 (1960) 269–275.
[13] W. Kalkert, Untersuchungen über den Einfluss der Fertigungsgenauigkeit auf den Zahnkraftverlauf und die Flankentragfähigkeit
ungehärteter Stirnräder, Ind., -Anz. 85 (1963) 175, 353, 513.
[14] F. Feltkamp, J. Rademacher, Fertigungsgenauigkeit und Laufverhalten von Zahnradgetrieben, Ind., -Anz. 90 (1968) 1769–1775.
[15] J. Rademacher, Einfluss von Flankenrichtungsfehlern und Breitenballigkeiten auf die Tragfähikeit von Stirnradgetrieben, Ind., -Anz.
90 (1968) 915–921.
[16] K. Fujita, F. Obata, K. Miyanishi, Method of root stress calculation for a crown gear, Trans. JSME, Part 3 39 (322) (1973) 1968–
1976.
[17] S. Li, Finite element analyses for contact strength and bending strength of a pair of spur gears with machining errors, assembly errors
and tooth modifications, Mech. Mach. Theory, in press, doi:10.1016/j.mechmachtheory.2006.01.009.
[18] International Standard ISO 6336/1, Calculation of load capacity of spur and helical gears—Part 1: Basic principle, introduction and
general influence factors, 1993, pp. 1–100.
[19] International Standard ISO 6336/2, Calculation of load capacity of spur and helical gears—Part 2: Calculation of surface durability
(pitting), 1993, pp. 1–28.
[20] International Standard ISO 6336/3, Calculation of load capacity of spur and helical gears—part 3: Calculation of tooth strength,
1993, pp. 1–72.