World Academy of Science, Engineering and Technology 53 2009

Transient Stress Analysis on Medium Modules

Spur Gear by Using Mode Super Position
Technique
Ali Raad Hassan

the center of the tooth space on one side to the center of the
Abstract—Natural frequencies and dynamic response of a spur tooth space on the other side. Several different curves make
gear sector are investigated using a two dimensional finite element up the tooth outline: concentric circular arc at the tooth tip
model that offers significant advantages for dynamic gear analyses. defining the addendum circle, involutes on the two sides of
The gear teeth are analyzed for different operating speeds. A primary the tooth, and trochoides between the involutes and the bottom
feature of this modeling is determination of mesh forces using a
lands at the base of the tooth. The tooth side involutes, fillet
detailed contact analysis for each time step as the gears roll through
the mesh. ANSYS software has been used on the proposed model to
trochoides, and bottom lands were shaped to model a gear cut
find the natural frequencies by Block Lanczos technique and with a rack form cutter. Coordinates for the surface profile of
displacements and dynamic stresses by transient mode super position the tooth come from a kinematic analysis of the cutting
method. The effect of rotational speed of the gear on the dynamic process [4]. Both the rack form cutter and the resulting gear
response of gear tooth has been studied and design limits have been surface are tangent to each other at the cutting points, which
discussed. generate the gear shape from the rack shape. The involute is
generated by points on the side of the rack form, the gear
Keywords—Spur gear, Natural frequency, transient analysis, tooth fillet is generated by the tip of the rack form, and the
Mode super position technique. bottom land is generated by the top surface of the rack form
tooth. With the appropriate rotations, this slope and radius
I. INTRODUCTION locates the direction and point of application of the gear mesh

R OTATING members like shafts, pulleys, gears, etc are

subjected to dynamic loads. In gears the dynamic load
creates bending stresses at the tooth root which can lead to
force on the central tooth. For this mesh, the gear parameters
used are: module (m) 12mm, the pressure angle (φ) 20
degrees, gear ratio (i) 2, modulus of elasticity (E) 2.15 ×105
fatigue failure. One of the major concerns in the design of N/mm2, poisson’s ratio (ν) 0.3, Steel mass density ( ρ) 8.75
power transmission gears is the reduction of dynamic load. ×10-9 N.s2/mm4 and the speed (n) 1440rpm.
Research work has revealed that the basic mechanism of noise
generated from gears is due to vibration excited by the III. NATURAL FREQUENCY CALCULATION
dynamic load. The life and reliability of a gear transmission is Fig. 1 shows the gear tooth as a cantilever beam. It is easy
reduced by high dynamic load. Minimizing gear dynamic load to calculate the natural frequency of this beam by the
will decrease gear noise, increase efficiency, improve pitting following equation [12]:
fatigue life, and help prevent gear tooth fracture. [2], [3], [7],
[9], [10], [13] and [16] worked in the area of static and
dynamic analysis of spur gear tooth. In this work, three tooth
segment of a 30 tooth pinion meshing with 60 tooth gear has
been considered. Trochoid fillet at the base of the involute
tooth is generated using rack tip.

II. GEAR TOOTH GEOMETRY

The pinion has a pitch diameter of 240mm, module of
12mm and a nominal pressure angle of 20 degree. The pinion
has standard full depth teeth with addendum of 1.0 m and
Fig. 1 Gear tooth modeled as a cantilever beam
dedendum of 1.25 m. The face width of the pinion is 60mm. A
6425N load is considered along the line of action of the
model. Development of the finite element model begins with
data describing the outline of a single tooth and its fillets from k
ωn = (1)
mass
Ali Raad Hassan, PhD, Chief Engineers at Ministry of Industry and k is the stiffness of the beam and it can be calculated from
Minerals (Baghdad), Iraq (E-mail: ali_gears@yahoo.com).
the following equations:

49
 World Academy of Science, Engineering and Technology 53 2009

3EI (2) P sin φ (7)

k= σx =
l3 F .t
F ⋅t3 (3)
I=
12 P cosφ × l (8)
σy =
F .t 2
The length of the beam l is equal to the height of the tooth
6
while the width of the beam t is the tooth thickness at pitch
circle radius (π m/2). Where, t is the thickness of the tooth at the critical section
(root of the tooth), F is the face width of the tooth, P is the
mass = ρ ( l × t × F ) (4)
total load and σx and σy are the maximum stresses of the beam
Where, ρ is the mass density of Steel and F is the tooth face in X and Y axes respectively. From this formulation, the
width. By using the above formula, the natural frequency is obtained value of stress in X direction for the proposed model
calculated using the equation (1) and is equal to 89065rad/sec. is 1.94MPa, and the stress in Y direction is 25.488MPa
respectively.
IV. NATURAL FREQUENCY BY RAYLIEGH’S METHOD
VI. FINITE ELEMENT MODEL
Rayliegh’s method can also be used to find the natural
frequency of continuous systems. This method is much A model consisting of a three tooth section of a 20 tooth of
simpler than exact analysis for systems with varying 12mm module pinion was developed with the general purpose
distribution of mass and stiffness. Although the method is finite element software ANSYS. Figs. 2 show the finite
applicable to all continuous systems, it is applied to beam in element grid for the three tooth gear segment for this model.
this section; the natural frequency by this theory is [8]: Successive reflections of the coordinates for the initial tooth
generated segments of three equally spaced, identical teeth are
l
⎛ d 2u ( x) ⎞ noted. Both the tooth surface and the inside rim surface are
∫ EI ⎜⎜⎝ dx 2 ⎟⎠
⎟dx
unconstrained. The total angle subtended by the segment is 52
ωn 2 = 0
l degrees. The radial lines defining the ends of the three teeth
∫ ρA(u( x)) dx
2
segment is at ±26 degrees from the centre line of the model.
0 (5)
An eight noded iso-parametric plane stress quadratic
The shape function u(x) of cantilever beam can be taken as quadrilateral element was used to build the finite element
[12]: model inside the frameworks described above. This element
⎛ x 5 5x ⎞
has a quadratic displacement function and is well-suited for
u ( x) = C1 ⎜⎜ 5 − + 4 ⎟⎟ (6) analyzing irregular shapes. Each node in the element has two
⎝ l l ⎠
degrees of freedom - translations in the X and Y directions.
By substituting Eq. (6) in Eq. (5) and applying the required The plane stress option with unit thickness was used and
gear details of the selected model, the resulting value of the scaled to the actual model thickness of 60mm. To specify the
natural frequency is (100423rad/sec). boundary conditions, all the nodes on the two radial lines
defining the ends of the segment and the bottom rim were
V. STRESS CALCULATION selected and given zero displacement in both directions. To
The load acting during the entire period of engagement is apply the load at a node, the grid had to have a node point at
not uniform. In the beginning, at the start of engagement, two or near this loading point. Fig. 3 show the right and left sides
pairs of teeth will be in contact and it is assumed each pair of the central tooth of the model (tension and the compression
will carry only half of the load, and the contact point will be in sides), with the nodal numbers in position and stresses at these
the tip of one of these teeth, but due to rotation, during the nodes will be evaluated for comparison purposes later. The
engagement, a new stage of contact is effected when a single complete model has 320 elements, 1071 nodes, and 2142
pair of teeth is in contact near the pitch circle. This is called a degrees of freedom. The backup ratio and the rim thickness in
single tooth contact and the first point of contact of this new this model are 2.5 and 67.5mm respectively.
stage is called the highest point of single tooth contact
(HPSTC). Hence, loads equal to 6425N (total load) are
applied to the end of the cantilever beam at an angle equal to
the pressure angle of the two mating gears, namely 20
degrees. The height of the beam (l) will be assumed to be the
distance between the pitch radius and root radius of the gear.
The axial and radial components of this load are taken as the
X-direction and Y-direction components of the load at a point
corresponding to the pitch radius of the cantilever beam where
the load is full value, which are 6037.5N, and 2197.4N
respectively. The stresses in X and Y directions can be
calculated by the following equations:

50
 World Academy of Science, Engineering and Technology 53 2009

where ωn is the natural frequency of the system and {U} is the

corresponding normalized eigen vector (mode shape). The
eigen values and eigen vectors were obtained using Block
Lanczos method[14]. The first five natural frequencies and
corresponding normalized eigen vectors were calculated using
this technique. The first five natural frequencies of the gear
tooth obtained for the selected model are given in Table I.

TABLE I
FIRST FIVE NATURAL FREQUENCIES

Natural frequency of the model

Mode
cycle/sec rad/sec
1 14674 92199

Fig. 2 Finite element mesh 2 15584 97917

3 15666 98432

4 22368 140542

5 26720 167886

VIII. LOAD DISTRIBUTION ON GEAR TOOTH

Load distribution on gear tooth In order to conduct a static
stress analysis the loads have to be evaluated. The load on the
central tooth of the finite element model, which produces the
largest bending stress, is the full load acting at the highest
point of single tooth contact (HPSTC). The magnitude of load
at any point of contact on profile of gear tooth as the load
moves from root to tip of tooth depends on the contact ratio.
Fig. 4 shows the contact path, the contact ratio CR is defined
Fig. 3 Enlarged view of central tooth of the model as the ratio of length of path of contact AB to base circle pitch
Pb.
The plane stress option with unit thickness was used and
scaled to the actual model thickness of 60mm. To specify the CR =
AB (10)
Pb
boundary conditions, all the nodes on the two radial lines
defining the ends of the segment and the bottom rim were Where:
selected and given zero displacement in both directions. To
− ( rb1 + rb2 ) tan φ (11)
2 2 2 2
apply the load at a node, the grid had to have a node point at AB = ra 1 - rb1 + ra 2 - rb2
or near this loading point. Fig. 3 show the right and left sides
of the central tooth of the model (tension and the compression The contact ratio of the spur gear, for the present case is
sides), with the nodal numbers in position and stresses at these 1.63. Fig. 5 shows the magnitude of loads at various points
nodes will be evaluated for comparison purposes later. The along the path of contact. The maximum normal load P acting
complete model has 326 elements, 1097 nodes, and2194 on the gear tooth sector of the proposed model is taken as
degrees of freedom. The backup ratio and the rim thickness in 6425N based on the strength of gear tooth material. This
normal load is considered in terms of its components in radial
this model are 2.5 and 140.625mm respectively.
and tangential directions taking into account the pressure
angle (Load angle γw) at any point under consideration, and
VII. MODAL ANALYSIS
then the tangential load and radial load can be determined by
Having obtained the finite element model, a modal analysis the following equations:
has been conducted on the same. The global stiffness matrix
[K] and global mass matrix [M] was obtained by assembling Pt = P cos γw
the element stiffness and mass matrices, respectively. The Pr = P sin γw
natural frequencies of the gear tooth were obtained by solving
the eigen value problem given by the following equation:
[ K ] {U } = ω n 2 [ M ] {U } (9)

51
 World Academy of Science, Engineering and Technology 53 2009

The moving load during the time of contact is considered in 9

intervals. At each time interval, the position and magnitude of
the load is different and shown in Table II. The static stress on
all the elements for each load is obtained by using ANSYS
Finite Element software and the static bending stresses on the
critical nodes (Node 92) on the compression side are listed in
the same table.

TABLE II
STATIC BENDING STRESS OF THE MODEL FOR EACH LOAD STEP
Load Load Applied Stress at node 92 (MPa)
Load
Step on Nodes σX σY
1 48 Half 1.732 20.399

Fig. 4 Contact path 2 68 Half 1.534 17.863

3 66 Half 1.331 15.259

4 64 Full 2. 231 24.989

5 62 Full 1.799 19.539

6 60 Full 0.694 7.236

7 58 Half 0.480 4.765

8 56 Half 0.268 2.605

9 54 Half 0.086 0.987

Fig. 5 Magnitude of loads along the path of contact X. TRANSIENT ANALYSIS

After computing the natural frequencies and the mode
shapes the forced response is obtained using the modal
superposition technique [12], [14]. The method is
computationally efficient particularly for a large sized
problem. The governing differential equation of gear tooth
considering the damping is:

{ }
[ M ] δ D D + [C] {δ D } + [k] {δ } = { P (t )} (13)

Where [M], [C] and [K] are global mass, damping and
Fig. 6 Load on various nodes of the selected model stiffness matrices of size (n×n), respectively. The {P (t)} is the
external time varying load of size (n×1).The symbols {δ},
Fig. 6 show the right side (Tension side) of the profiles of {δ°}, {δ°°} are the displacement, velocity, acceleration
the central tooth of the selected model, shown in Fig. 3, vectors of size (n×1). Eq.(13) has to be recast such that m
stretched out into straight line with the nodes indicated for the uncoupled equations in a single degree may be obtained. The
purpose of application of loads. The load that have been recast equation can be obtained by substituting for {δ}, as {δ}
applied for the various runs of static analysis are indicated on = [U] {p}, where [U] is a matrix (size n×m) of the first m
the respective nodes starting from the extreme left end (node eigen vectors (m<<n), and {p} is a generalized displacement
48) to the final point of contact (node 54) of the model[1]. vector of size (m×1). Pre multiplying Eq. (13) by [U]T , gives:

{ } { }
[U ]T [ M ][U ] p °° + [U ]T [C ][U ] p ° + [U ]T [ K ][U ]{p} = [U ]T { P (t )} (14)
IX. STATIC BENDING STRESS ANALYSIS
After computing the natural frequencies and the mode Here [U]T [M] [U] is a unity diagonal matrix of size (m×m)
shapes the forced response is obtained using the modal and [U]T [C] [U] is a diagonal matrix of size (m×m). The
superposition technique [13]. The method is computationally diagonal elements of these matrices are 2ξi ωi where ξi are the
efficient particularly for a large sized problem. The governing damping ratio and ωi are the natural frequencies for i= 1, 2, 3,
differential equation of gear tooth considering the damping is: ……m. The [U]T [K] [U] is also a diagonal matrix with the
square of the natural frequencies (ω12, ω22, ω32…, ωm2) as the
diagonal terms. The resulting decoupled set of equations are
[K] {δ } = [P ] (12)
solved as standard single degree of freedom system and the

52
 World Academy of Science, Engineering and Technology 53 2009

resulting displacement vector is transformed back to represent to its distance from either node, to the two nodes under
the gear tooth displacements. consideration as shown in Fig. 6. The Initial conditions for
displacements and velocities for the gear tooth in the proposed
⎡ξ1ω1 0 ⎤ model are taken as zero for all degrees of freedom. The third
⎢ 0 ξω ⎥
mode is selected for the mode superposition technique. At
[U]T [C] [U] = 2⎢ 2 2 ⎥ (15)
⎢ 0 ⎥ each time interval, the acting load is calculated and fed into
⎢ ⎥ the mode superposition part in ANSYS software and the
⎣ 0 ξ mω m ⎦
corresponding deformation and stress are thus obtained. The
In order to obtain the diagonal elements of the [U]T [C] [U] dynamic displacements and dynamic stresses in the X and Y
given in Eq. (15), the damping matrix is expressed as a linear directions in the central tooth root portion of the selected gear
combination of [M] and [K] (Rayleigh damping), that is: model, which are at node 92, are plotted in condition of
1440rpm. The dynamic analysis is also carried out for the
[C] = α [M] + β [K] (16) following four speeds of gear namely, 360rpm, 720rpm,
Where α and β are constants to be determined. Then, 1440rpm and 1800rpm.
2 ξ1ω1 = α + ßω1 , 2 ξ 2ω2 = α + ßω2 ,....,2 ξ mωm = α + ßωm
2 2 2

XI. RESULTS
(17)
Figs. 7 and 8 show the dynamic displacements in the X and
From the finding of Mohammad et al. [6] the values for mild
Y directions of the selected model at node 92. Figs. 9 and 10
steel material, ξ1 ω1=2.0 and ξ3 ω3=3.2, where ω1 and ω3 are
show the dynamic stresses at X and Y directions of the model
the first and third natural frequencies in rad/sec. Using these
two expressions, the damping ratio for other modes are at node 92. Figs. 11 and 12 show the dynamic shear stress
calculated by substituting in Eq. (17). Since for most (τxy) and dynamic von Mises stress of the model at node 92.
industrial problems (ω3 >> ω1) including the first three natural
frequencies to compute the dynamic transient response is
fairly accurate. After substituting the values of ω1 and ω3 of
the model from Table I in Eq. (17), the values of α and β are
obtained as 13.2 and 8×10-8 respectively.
A. Time Steps
The time of contact T of gear tooth depends on the
rotational speed of the gear. If the gear is assumed to run at a
speed of n-rpm, the time taken for one revolution of the gear
will be (60/n) sec. In one revolution, Z number of teeth will
get engaged and disengaged.
1
ΔT = (18)
10ω m
Fig. 7 Displacement of the model in X-Direction
Where Z is gear number of teeth. Then the time taken for one
pair of teeth in engagement will be (60/ n. Z) sec, which will
be (2.09×10-3 sec) in the selected model. The time taken for
one pair of teeth in engagement can be divided into required
number of steps (NTS - number of time steps). One time step
∆T can be calculated by considering the number of modes,
which are expected to contribute to the dynamic response. So,
the third frequency of the model is (15666 cycle/sec) which is
taken from the finite element modal analysis results in Table
1; the time interval obtained is (6.38×10-6 sec) and is shown in
Fig. 6, if the natural period of the mth mode is Tm, a choice of
∆T equal to Tm/10 should give a reasonable dynamic
response up to mth mode [13]. Therefore,
60
TNTS = (19)
ΔT .N .Z
Where, TNTS is the total number of time steps or intervals.
So, the total number of intervals (time steps) for the model is Fig. 8 Displacement of the model in Y-Direction
326 which mean that there are about 40 intervals between
each two nodes shown in Fig. 6. At any time, two nodes are
considered to calculate the load vector for that time interval.
The actual load at a point is distributed in inverse proportion

53
 World Academy of Science, Engineering and Technology 53 2009

Fig. 9 Dynamic Stress of the model in X-Direction Fig. 12 Dynamic von Mises Stress of the model

Table III lists the Maximum values of σX, σY, τxy and von
Mises stresses of the model at node 92 for different speeds.
Figs. 13, 14 and 15 show the relation curves of σX, σY and von
Mises stresses with the rotational speeds, respectively.

TABLE III
MAXIMUM DYNAMIC STRESSES FOR EACH SPEED OF THE MODEL
Spee τxy von Mises
σX σY
d (MPa) (MPa)
(MPa) (MPa)
(rpm)

360 2.119 20.343 5. 723 21.701

720 2.623 28.127 7.51 29.867

Fig. 10 Dynamic Stress of the model in Y-Direction. 1440 3.512 39.006 10.441 41.78

1800 3.918 44.217 12.022 47.225

Fig. 11 Dynamic Shear Stress of the model

Fig. 13 Curve Relation of the Speed and Maximum Dynamic Stress

in X-Direction

54
 World Academy of Science, Engineering and Technology 53 2009

value. The natural frequency obtained from Eq. (1) is more

reasonable and recommended. The first natural frequency
obtained from finite element modal analysis was
92199rad/sec; there is a good agreement between this value
and the recommended natural frequency value obtained from
Eq. (1). It was observed from the plots (shown in Figs. 9 and
10) that the dynamic stresses were varied with a time period of
6.96 ×10-5sec whereas the time period of the fundamental
natural frequency was 6.814 ×10-5sec. So, the dynamic
stresses are varying periodically with fundamental bending
vibration time period of the model. The dynamic stress plots at
tooth root where (shown in Figs. 9 and 10) and deformation at
the tooth root were (shown in Figs. 7 and 8), three important
distinguishable regions were observed. In the first region the
Fig. 14 Curve Relation of the Speed and Maximum Dynamic Stress displacements and dynamic stresses during starting of
in Y-Direction engagement of the gear tooth were zero, and both the pairs of
teeth were contact and share normal load equally. In the
second region, displacements and dynamic stresses were high
compared to all other regions, in this region, a single pair of
gear tooth will be in contact and full load act on the gear pair
under study. In the third region again, two pairs of gear teeth
were in contact and share load equally. From the dynamic
analysis results, the dynamic stresses at the root are lesser
when the load was at the tip, and were high when the load was
at the mid height [that is at HSPTC- the highest point of single
tooth contact] at this point the full load was acting. From the
static analysis, the maximum values of σX and σY were at node
92 of the model lie in the root at left side (compression side)
of the central tooth. For the actual load distribution on the gear
profile obtained by finite element method the stresses σX and
σY were 2.231MPa and 24.989MPa, respectively, when full
load is at HSPTC point (node 64). By beam gear design
Fig. 15 Curve Relation of the Speed and Maximum Dynamic von approach the values of σx and σY calculated from Eqs.7 and 8
Mises and Shear Stresses
were 1.94MPa and 25.488MPa, respectively, this shows good
agreement between these values if the differences between the
A. Checking for Plot Fluctuation tooth and beam shapes are considered. From the dynamic
The correctness of the results in Figs. 7 to 12 were verified stress analysis stress plots at the root of the model (node 92;
by the following procedure, the number of peaks for a the maximum σx, σy, τxy and von Mises stress were 3.512MPa,
selected time period was calculated for each of the above 39.296MPa, 10.441MPa and 41.78MPa respectively.
plots. The frequency of the response, in cycles per seconds, Referring to Figs. 9 and 10, it was observed that the
can be calculated by dividing the number of peaks for a magnitude of static stresses (σx and σy) at the point of
selected time period by the time taken for the same. For application was half of the enveloping curves of the dynamic
example; at Fig. 10 (which shows the dynamic stress of the stresses (mean of the maximum and minimum dynamic
model in Y-direction) for time period (0 to 0.418×10-3 sec) the stresses). The amplitude of the dynamic stress goes on
number of peaks was 6, so; the resulting frequency was increasing as the speed increases with its major frequency of
calculated by dividing 3 on 0.417×10-3 sec, which yields variation corresponding to the fundamental frequency of the
(14354 cycle/sec). It was clear that this value matches with gear tooth. This may be observed from Fig. 16. It is observed
fundamental natural frequency of the model (first natural from Table III that the maximum magnitudes of σx , σy , τxy,
frequency of the model in Table I). and von Mises stress at the tooth root for various speeds of the
selected gear model (in the range of 360rpm to 1800rpm) were
XII. DISCUSSION increasing gradually from 2.119MPa to 3.918MPa, from
The natural frequency of the tooth obtained by Rayliegh’s 20.343MPa to 44.217MPa, from 5.723MPa to 12.022MPa and
method (Eq. 5) was 100423rad/sec. The natural frequency from 21.711MPa to 47.225MPa respectively, for increase in
obtained by beam formulation (Eq.1) was 89065rad/sec. It speed from 360rpm to 1800rpm. The variation is shown in
was noted in Rayliegh’s method that the assumed shape u(x) Figs. 13, 14 and 15. Therefore, the dynamic stresses at the
unintentionally introduces a constraint on the system (which root of the low modules spur gear tooth were increasing with
amounts to adding additional stiffness to the system), and increase in the rotational speed of the gear.
hence the frequency given by Eq. (5) is higher than the exact

55
 World Academy of Science, Engineering and Technology 53 2009

REFERENCES
[1] Ali Raad Hassan, Thanigaiyarasu G. and Ramamurti V. (2008). ‘Effects
of Natural Frequency and Rotational Speed on Dynamic Stress in Spur
Gear’, Proceeding of World Academy of Science, Engineering and
Technology, Volume 36, pp. 1279-1287.
[2] ANSI/AGMA (2001-B88) (1988). ‘Fundamental Rating Factors and
Calculation Methods for Involute Spur and Helical Gear Teeth’,
American Gear Manufacturers Association, Arlington, VA.
[3] Eiff von H., Hirschmann K.H., Lechner G. (1990). ‘Influence of gear
tooth geometry on tooth stress of external and internal gears’, ASME
Journal of Mechanical Design 112 (4) 575-583.
[4] Hefeng B., Savage M., Knott R.J. (1985). ‘Computer modeling of rack
generated spur gears’. Mechanism and Machine Theory 20 (4) 351-360.
[5] Khurmi R.S., Gupta J.K. (2006). ‘A Textbook of Machine Design’,
Eurasia Publishing House (PVT.) LTD., New Delhi, pp 1035-1038.
[6] Mohammad D., Khan N U, Ramamurti V. (1995). ‘On the role of
Rayleigh damping’. Journal of Sound and Vibration 185 (2) 207.
[7] Oda S., Nagamura K., Aoki K. (1981). ‘Stress analysis by thin rim spur
gears by finite element method’. Bulletin of the Japanese Society of
Fig. 16 Stress response of the 1st model in Y-Direction
Mechanical Engineers 24 (193) 1273-1280.
[8] Rao Singiresu S. (2004). ‘Mechanical Vibrations’ 4th Edition, Pearson
A. Food for Thought Education (Singapore) Pte. Ltd., India Branch, Delhi.
[9] Ramamurti V., Reddy K. R. K. (2001). ‘Dynamic analysis of spur gear
Modal super position technique assumes that any transient tooth’. The Institution of Engineering (India) Journal 82 33-40.
disturbance of member was the sums total of the disturbances [10] Ramamurti V., Gupta L. S. (1979). ‘Dynamic stress analysis on spur
due to the first few modes of vibration. In reality the gear teeth’. ASME Conference 38, DET.
contribution from the higher modes die down very fast in [11] Ramamurti V., Vijayendra Nayak H., Sujatha C. (1998). ‘Static and
dynamic analysis of spur and bevel gears using FEM’, Mechanism and
comparison with the 1st mode. This has been experimentally Machine Theory 33 (8) 1177-1193.
observed to be true. This is the basis of subsections 7 to 9 of [12] Ramamurti V. (2002). ‘Mechanical Vibration Practice with Basic
this paper. Theory’, Narosa Publishing House, New Delhi.
Referring to section 14.8 (Dynamic Factor) of [15], [13] Ramamurti V. (1998). ‘Computer Aided Design in Mechanical
Engineering’, McGraw Hill Publishing Co., New York.
dynamic stress is obtained by multiplying by the Lewis form [14] Seshu P. (2004). ‘Text Book of Finite Element Analysis’, Prentice-Hall
factor cv (Equation 14.27): of India, New Delhi.
[15] Shigley J.E. and Charles R.M. (2003). ‘Mechanical Engineering
(i). This factor does not take into account gear face width Design’, Tata McGaraw-Hill, New Delhi, 6th edition.
[16] Wilcox L., Coleman W. (1973). ‘Application of finite elements to the
(F). analysis of gear tooth stresses’. Journal of Engineering for Industry 95
(4) 1139-1148.
(ii). It is assumed that the static stress is multiplied by a
factor involving peripheral velocity of the gear[5].
(iii). It is well known that when the face width is increased Ali Raad Hassan, Assistant Professor of Technical
Institutes (Iraq), Chief Engineers at Ministry of Industry
the gear tooth becomes stronger and, hence the and Minerals (Baghdad),Iraq. He has been awarded 5
magnitude of dynamic stress is smaller than the gear times for his researches and developments on gear
with smaller face width, but this has not been taken into manufacturing machines. Currently he got his PhD in
account in this formula. mechanical engineering from Department of Mechanical
Engineering, Anna University (India). He has more than
(iv). During the transient phase the gear tooth is expected to 10 years experience in Industry.
vibrate in one or more of its natural frequencies, which
has not been assimilated or taken into account in the
existing code.
In the present approach these aspects have been taken into
account.

XIII. CONCLUSION
Modal super position technique assumes that any transient
disturbance of member is the sum totals of the disturbance due
to the first few modes of vibration. In reality the contribution
from the higher modes die down very fast in comparison with
the first mode. This has been experimentally observed to be
true. It is found that as the rotational speed of gear increases,
the bending stress of dynamic stress also increases. For the
moving load, the ratio of maximum dynamic bending root
stress to maximum static bending root stress is increases.

56