2014-01-1780

Published 04/01/2014
Copyright © 2014 SAE International
doi:10.4271/2014-01-1780
saematman.saejournals.org

Design and Optimization of Planetary Gearbox for a Formula Student

Vehicle
Soovadeep Bakshi, Parveen Dhillon, and Teja Maruvada
Indian Institute of Technology Bombay

ABSTRACT
This paper presents the method of designing an optimized light weight, cost effective planetary gearbox for a Formula
Student vehicle. The gearbox has a high speed functioning capability, in addition to the compact size and light weight. The
iterative optimization procedure used provides a technique for selecting the best possible configuration of the gearbox.
Conventional gearboxes used for this purpose are generally two step reduction gearboxes, which are bulkier in terms of
weight and volume. Also, a review of the existing market reveals that the planetary gearboxes manufactured in India are
not capable of handling high speeds, thus rendering them futile for racing applications.

The target reduction ratio for the gearbox is a fixed parameter. The method involves design and optimization of the
gear-train with the calculated ratio. A detailed algorithm has been used, which involves developing a mathematically
modelled code for deciding gear characteristics, physical modelling using CAD, and structural testing of the same through
simulation, with repeated iterations with a target to reduce both weight and volume. The program used has been modelled
on the theoretical data acquired through perusal of available literature. The selection of bearings have been made after
calculating the forces acting on the bearings, followed by market research to select bearings with the requisite load
capability. The shafts for the gears have also been designed according to force calculations, followed by structural
analysis. The casing for the gears is constructed around the gear-train for efficient packaging and validated through static
structural simulation. A preliminary thermal analysis has been completed to verify the design according to thermal
considerations. The result of the same has been presented.

CITATION: Bakshi, S., Dhillon, P., and Maruvada, T., "Design and Optimization of Planetary Gearbox for a Formula Student
Vehicle," SAE Int. J. Mater. Manf. 7(3):2014, doi:10.4271/2014-01-1780.

INTRODUCTION DESIGN CHALLENGE

Planetary Gearboxes (or Epicyclic gearboxes) have been The system that has been designed consists of a single stage
widely used as one of the better ways to transmit power for conventional planetary gearbox (NR>NS) which can sustain the
high torque applications. With more efficiency than that of peak torque and RPM rating of the motor (54 N-m and 6000
normal series gearing, their use is becoming more and more RPM respectively) [2] with which it is to be coupled. This paper
prevalent in the automotive field, especially in the case of covers all the aspects of designing the gearbox, including
Formula Student electric vehicles. Low spatial requirement and selection of the gears, shafts, bearings and the outer casing.
a high power-to-weight ratio make these gearboxes superior, The gearbox ratio is fixed at to give a target reduction in
especially when compared to double-step single reduction between 6.2:1 and 6.5:1.
gearboxes. The planetary arrangement also keeps the motors
concentric to the driveshaft, thus providing more efficient
packaging of components. This concentricity can also be used
Efficiency Analysis of Planetary Configurations
to produce wheel hub-mounted motors, especially in Formula This paper considers three possible configurations for the
Student vehicles which are powered at all four wheels. geartrain to be used in the design, since these configurations
are capable of providing the requisite reduction ratios for speed
However, one major disadvantage of the planetary reduction. The available choices are summarized in Table (1)
arrangement is its incapability to function at higher speeds due shown below, with their respective values of efficiency [3].
to high noise and wear. Additionally, the planetary arrangement
can only be used for geartrains with a reduction ratio of at least
3:1 or greater [1].

688
 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014) 689

Table 1. Configuration Specifications. CALCULATIONS

The three main components which have to be designed are the
gears, the bearings and the shafts. In this section, the
mathematical equations involved in the design procedure of
gears, bearings and shafts will be looked at.

Gears
Gears can be of multiple types. The most commonly used gear
Here, R, S and C represent Ring, Sun and Carrier respectively. types for parallel shaft gearboxes are spur and helical gears.
The efficiency of the configuration with fixed carrier is taken to For the gearbox, spur gears have been used in order to avoid
be ηSR, which is a positive quantity between zero and one. the axial loads generated in a helical mesh, which in turn
Since the ring gear is always larger in dimension than the sun affects the bearing and casing design. All the gears present in
gear, the value of mGSR is always greater than one. the single stage planetary arrangement have similar methods
of design, barring a few alterations which have been mentioned
From simple inequality and attribution of the above mentioned below.
limits to the variables ηSR and mGSR, it can be seen that the
efficiencies of the three configurations lie in the order 3>1>2, The main equations required for gear design are the two AGMA
where the integers refer to the first column shown in Table (1). stress equations for bending stress and contact stress
respectively. The two equations are as follows [4]:
From Table (1), configuration 1, i.e., the configuration with fixed
ring gear, has a maximum reduction ratio, since NR>NS. Thus,
selection of this configuration is justified insofar as the paper
aims at optimization of the planetary gearbox in terms of weight (1)
and volume occupied. The selected configuration can provide
the requisite reduction ratio while maintaining the smallest size,
since the ratio NR/NS can be smaller than in other
configurations. The selection of configuration 3 can be easily (2)
ruled out although it is the most efficient since it cannot provide
the requisite reduction ratio. The value of tangential load acting on the gears can be
calculated from the value of rated Torque and Speed rating of
In addition to lower efficiency and higher weight and volume the gearbox, while the other factors are as mentioned in [4],
occupied as shown above, considering the fact that the which uses AGMA standards. Here, the Geometry Factor for
manufacturing of a shaft attached to a rotating annulus may Bending Strength (YJ) is approximated in Eqn. (3) as [5]:
not be economically viable, especially in order to guarantee
concentricity and eliminate eccentricity in the output shaft,
configuration 2 can be eliminated.
(3)

The input is provided at the sun gear, with the output at the This equation gives a good estimate of the values presented in
planet carrier and the ring gear remaining stationary. The the AGMA standards used for gear manufacture.
requirements are summarized in Table (2) shown below.
The selected material for the gears is carburized and case
Table 2. Design Specifications. hardened EN 353 grade alloy steel due to its economic
feasibility and ready availability with manufacturers.

The maximum stress within the material should be less than

the allowable stress values for bending and contact stresses.
These are defined as [4]

(4)
690 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014)

(5)

The values of St (Allowable Stress Number for Bending) and Sc

(Allowable Stress Number for Contact) are derived from the
AGMA standards for case hardened carburized steel alloy. The
Safety Factor values are taken as per requirement. Since the
motor used will never generate torque and RPM values which
are more than the peak ratings, the Factor of Safety can be
kept at unit value. However, the value has been chosen to be
1.2 as a precautionary measure.

The above equations are useful for the design of the Sun Gear
in the planetary arrangement. For the design of the planet
gears, the following changes have to be made:

1. The planets face bi-directional loading. Hence, the value of

St is taken to be 0.7 times the actual value [4]. This value is Figure 2. Planet gear.
recommended by AGMA, and lies in between the Goodman
and Gerber failure loci
2. The planet gears should be designed for at least [1+0.9(η1)]
times the actual load it is supposed to sustain in the
absence of a load sharing mechanism [6]
3. The gear ratio to be used for calculation does not directly
take the value of (NP/NS). The actual value to be used is (NP/
NS)(ωi-ωo)/ωi [3]

The different gears are shown in Fig. (1), Fig. (2) and Fig. (3).

Figure 3. Ring gear.

Floating Arrangement of Planetary Gears

For equal load sharing in the planet gears, there are 3 methods
as follows:

1. Completely rigid system with zero tolerances

2. System with flexibility (floating)
3. With mechanical means to adjust planets to provide load
equalisation

Figure 1. Sun gear.

Out of these three options, the second option is the cheapest
to implement, and is ensured in the following ways: i) The Sun
Gear shaft is kept flexible so that it can adjust its position
according to requirement, and ii) (NR+NS)/η should be an
integer [6].
 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014) 691

Shafts The Goodman failure criterion presents the equation:

Shafts provide a lot of scope for weight reduction, and the
design procedure requires cognition of both shear stresses due
to torsion and normal stresses due to bending. In a planetary
(14)
gearbox, there are three types of shafts:
Using Eqn. (6, 7, 8, 9, 10, 11, 12, 13, 14), we arrive at the
1. Input Shaft: Couples motor with the sun gear
following equation, which gives the relation between the inner
2. Planet Gear Shaft: Shafts supporting the planet gears. They and outer diameters of the shaft [4]:
are attached to the carrier
3. Output Shaft: Shaft from the planet carrier, coupled with the
Constant Velocity joint at the output of the gearbox

After the positions of various elements (gears, bearings) on the

shaft have been decided, the forces and torques acting on the
(15)
shaft can be calculated to determine the loading conditions. In
the case of epicyclic gearing with spur gears only, axial forces Hollowed shafts can decrease the weight by a considerable
are absent on the shaft; the only forces acting on the shaft are amount. However, since both do and di are unknown, a value
the tangential and radial loads generated at the gear mesh. for either of the two variables has to be assumed to proceed.
The fluctuating stresses due to torsion and bending are [4]: The outer diameter can be defined from the geometric
constraints due to the size of the gearbox, and from that value,
an inner diameter for the hollow shaft can be calculated.
(6)
In case of the input shaft, the inner diameter is fixed because
of direct key coupling with the shaft of the motor. Hence, a
minimum outer diameter can be calculated for which the shaft
(7) will not fail.

While designing the shafts for the planet gear, the torsional
(8) stress is absent, but the centripetal forces due to the revolution
of the planet gears about the sun have to be considered.

(9)
The CAD models for the shafts can be seen in Fig. (4), Fig. (5)
and Fig. (6).
The formula for Moment of Inertia and Polar Moment of Inertia
for a cylindrical rod is given as [7] Slots have been made for C-Clips to locate the bearings on the
shafts. Cotter pins have been avoided to prevent imbalanced
rotation.

(10)

(11)

Distortion energy failure for Von Mises stress with no axial load
gives the following equations [4]

(12)

(13)
Figure 4. Sun gear with input shaft.
692 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014)

2. Position of the bearing from the overall equilibrium of the

shaft
3. Life of bearing, i.e., the number of revolutions that the
bearing is expected to function for in total
4. Torque transferred through the shaft

From [8], the tangential and radial forces at the gear mesh are
as

(16)

Figure 5. Output shaft.

(17)

These are the forces which are present at the gear mesh. The
bearing radial load can now be easily calculated by balance of
moments, as shown in Fig. (7) [9].

Figure 6. Planet gear shaft.

Bearings
Bearings are vital components in all mechanical assemblies
which have a rotating part sliding in contact against a
stationary component. However, bearings also require
significant maintenance, and hence, proper selection of
bearings is a requisite criterion in any relevant design project.

The bearings at the planetary stage of the epicyclic geartrain

are constrained by the overall gearbox size. Thus, selection of Figure 7. Calculation of radial forces on bearings by balance of
the bearings is one of the important things to consider for moments [9].
optimized performance of the gearbox (considering size and
These values are used in the following equation [8] to calculate
weight). Three types of bearings are used in the planetary
the required Dynamic Load Rating as follows
gearbox:

1. Bearings on the input shaft

2. Bearings on the output shaft
(18)
3. Bearings on the planet gear shaft, which allow the planet to
rotate about its own axis independent of the planet carrier's Since spur gears are used, there is no axial load generated on
revolution about the sun gear the bearings, Pe = Bearing Radial Load can be used. This is an
advantage of using spur gears, since either ball or cylindrical
Bearings that support the shafts are an important part of the bearings, which are much lighter than tapered roller bearings,
final assembly as they reduce the friction between the rotating can be used.
mechanical components and constitute an important factor in
efficiency. Selection of bearing is generally performed in steps. The value of Cd is calculated, and the bearings are selected
according to the RPM and Cd rating provided by the
To select a bearing for a rotating shaft application, the following manufacturer, and the most suitable bearing is selected
are required: according to outer diameter of the shaft [8].

1. Radial (Fr) and Axial (Fa) Load on the shaft

 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014) 693

As in the case of shafts, one has to consider the centripetal Planet Carrier
forces arising due to the revolution of the planets at The planet carrier has been designed to sustain the generated
considerable speed around the sun gear while calculating the loads in the geartrain. The design of the planet carrier can be
dynamic load rating of the bearings. seen in Fig (9). The loads include the input and output torques,
the weights of the planet gears, shafts and bearings, its own
The shafts are made out of EN 36, a readily available alloy of weight, and also the centripetal forces generated due to
steel. revolution.

Casing The material used for manufacturing the planet carrier is EN 36

Steel Alloy.
The casing is designed around the gears with a cylindrical
shape selected for ease of manufacturing. The two halves of
the casing can be seen in Fig. (8a) and Fig. (8b). The material
selected for the casing is Al 6061-T6 Aluminum Alloy, since it is
readily available, and is much lighter than any steel alloy. The
heat treatment provides the material the requisite strength.

The two halves of the casing and the ring gear have reciprocal
locating steps with a rubber oil seal in between to prevent
leakage.

Figure 9. Planet carrier.

MATHEMATICAL MODELLING
The equations for gears, shafts and bearings have been
modelled as a part of a MATLAB code. The ideal configuration
is decided by the code, which iterates for different values of
facewidth and number of teeth of the sun gear for a given
module. The code calculates the approximate mass of the
entire gearbox using density of steel for the gears, shafts and
Figure 8a. Casing (Input Side View).
bearings, and the density of Aluminum for the casing.

The weight of the gears is approximated by calculating the

volume for each configuration (module and number of teeth)
with unit facewidth, and then multiplying the result of the code
to get the final volume. Shafts have been modeled as round
cylinders, not accounting for keyway slots, while the bearings
have been modeled as annular discs. The weight of the
aluminum casing has been approximated to that of a hollow
container which is cylindrical.

SIMULATIONS
The CAD model of the configuration of gears derived from the
mathematical model have been developed after which they
have been virtually tested for failure by static structural
simulation using ANSYS. The meshing conditions used are as
Figure 8b. Casing (Output Side View). shown in Fig. (10). The results for the structural analysis of the
Sun gear is shown in Fig. (11) and Fig. (12).
694 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014)

The simulations for the remaining gears have also been Static structural analyses have also been carried out for the
conducted in the same manner. shafts by calculating and applying the peak loading conditions
and constraining the shaft suitably. The analysis for the input
shaft is shown in Fig. (13) and Fig. (14).

Figure 10. Meshing used in simulation.

Figure 13. Deformation of input shaft on application of loading

conditions.

Figure 11. Deformation in sun gear on application of forces.

Figure 14. Factor of safety of input shaft on application of loading

conditions.

The analyses of the other shafts, casing and planet carrier

have been carried out following the same procedure.

LUBRICATION
The choice of correct lubricant is necessary because the
absence of a correct lubricant can lead to gear damage,
especially due to friction and wear. The lubricant also serves as
a medium of transfer of heat away from the gears to the outer
casing. Thus, the lubricant plays a major role in reducing
losses due to frictional heat generation and increases
efficiency.

Figure 12. Safety Factor in sun gear on application of forces. For geartrains with tangential velocities upto 3000 ft/min (∼15
m/s), the method of splash lubrication or oil bath can be used
[3]. For selection of gear oils for the transmission unit in the
splash lubrication method, the two important factors which
must be taken into consideration are as follows [3]:
 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014) 695

1. An optimum oil level in the oil bath required to maintain inside the gearbox has been modelled as an average of the
effective lubrication and cooling ability of the lubricant lubricating oil and air present inside the gearbox, taking an
2. Temperature limitation for the lubricant to be used, since average of the calculated heat transfer coefficients. Thermal
most lubricants fail to perform their function beyond their analysis for the gearbox has been carried out at an ambient
rated temperature temperature of 22°C.

Lubricants are generally available in standard grades which The temperature distribution of the fluid present in the gearbox
refers to their viscosities at 40°C. is shown in Fig. (16).

For the designed geartrain, Extreme Pressure (EP) grade

lubricant of suitable viscosity can be used to provide the film
between the meshed gears, which can operate at high speeds
and can sustain extreme pressure and temperature conditions.
Oils with high viscosity are preferred. The use of EP 220 grade
lubricant can be justified according to the tangential velocity,
tooth load and gear ratio present in the design [3]. However,
this grade can vary according to usage requirements and can
be finalized after researching on available standard options
and their properties.

THERMAL ANALYSIS
A preliminary thermal analysis has been completed to verify the
design, even with the minimal required operation time for the
gearbox (∼30 minutes).

The result of the analysis has been shown in Fig. (15). The
figure shows the geartrain after analysis. The model has been
described below.

Figure 16. Results of themal Analysis for gearbox fluid.

From the results, the maximum temperature of 129.8°C exists

at the gear mesh between the sun and the planets. The fluid is
expected to reach a maximum temperature of 102°C. The
casing temperature varies from 22°C (ambient temperature) to
64°C. The relatively low increase in temperatures can be
attributed to the low operation time of the vehicle, as
mentioned before.

The maximum limit on temperature of the rubber seals used in

the gearbox is approximately 80°C, beyond which the seals
start to lose their effectiveness. Since the expected maximum
temperature of the casing is lesser than the maximum
operating temperature of the seals, the design of the gearbox
is justified according to thermal considerations.

Other means of cooling which can be used are vents in the

body to direct airflow over the gearbox, or a dedicated cooling
system for the powertrain module, which can be further
incorporated into the design of the vehicle if required.

Figure 15. Results of themal analysis for gears.

SUMMARY/CONCLUSIONS
The model consists of heat generation at the contact points of
The final specifications of the gearbox are as shown in Table
the geartrain, which is considered to be 95% efficient. The total
(3). The table shows the dimensions of the components of the
time for simulation is taken to be 1800 seconds, with an
gearbox (gears and shafts), and also their material
average power transmission of 18 kW from the motor. The fluid
composition.
696 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014)

Table 3. Final Specifications. number of teeth on the planets is also an integer (82/2=41)
2. (NR+NS)/η is also an integer, where η=3, which ensures
proper load sharing
3. The number of teeth on the sun, planet and ring gears
are prime to each other, thus negating any effect of
poor manufacturing and producing sequential instead of
simultaneous meshing to reduce vibrations [6]

REFERENCES
1. Dudley, D.W., Radzevich, S.P., “Handbook of Practical Gear
Design,” Taylor & Francis, OX14 4RN, UK, ISBN 13: 978-
1566762182, 1994.
2. Agni Motors Official Website, “Agni Performance Graphs 95
and Series (Torque/Power/Eff./RPM vs. Current),” http://www.
agnimotors.com/95_Series_Performance_Graphs.pdf.
3. Merritt, H.E., “Gear Engineering,” Wiley, Hoboken, NJ, ISBN 13:
978-0470596272, 1972.
4. Budynas, R.G., Nisbett, J.K., “Shigley's Mechanical Engineering
Design (McGraw-Hill Series in Mechanical Engineering),” McGraw-
Hill Science/Engineering/Math, New York, NY, ISBN 13: 978-
0073529288, 2010.
5. Khurmi, R.S., Gupta, J.K., “Textbook of Machine Design,” S. Chand
Ltd., New Delhi, India, ISBN 13: 978-8121905015, 1996.
6. Lynwander, P., “Gear Drive Systems: Design and Application,”
Taylor & Francis, OX14 4RN, UK, ISBN 13: 978-0824718961,
1983.
7. Timoshenko, S., “Strength of Materials,” D. Van Nostrand
Company, Inc., Princeton, NJ, 1941.
8. SKF Bearings Ltd., “Roller Bearings,” http://www.skf.com/in/
From the simulation results, the results tabulated above are products/bearings-units-housings/roller-bearings/index.html
9. NTN Corporation, “Ball and Roller Bearings, 2202 IX/E,” http://
able to sustain the required torque and RPM specifications. www.ntn.co.jp/english/products/pdf/ball/pdf/Bearing_en_all.pdf
The expected weight of the gearbox is around 7 kg (derived
from CAD model). The entire gearbox assembly can be seen in ACKNOWLEDGMENTS
Fig. (17).
The authors would like to thank Prof. V. Kartik, Dept. of
Mechanical Engineering, Indian Institute of Technology
Bombay for his guidance and provision of resources. The
authors would also like to thank their teammates at IIT Bombay
Racing, especially Mr. Shashwat Gopal, for their constant
support. Mr. Satish Dharap from Transmatix Ltd. has also
provided valuable help to the authors with technical support.

DEFINITIONS/ABBREVIATIONS
Wt - Tangential Gear Load (N)
Wr - Radial Gear Load (N)
Σ - Bending Stress
Ko - Overload Factor
Kv - Dynamic Factor
Ks - Size Factor
Pd - Transverse Diameteral Pitch
B - Facewidth (narrower member) (mm)
KH - Load Distribution Factor
Figure 17. Gearbox assembly (view from output side). KB - Rim Thickness Factor
YJ - Geometry Factor for Bending Strength
This particular set of gears satisfy all the following conditions in
addition to the primary requisites: mt - Transverse Metric Module (mm)
ZR - Surface Condition Factor
1. (NR-NS)/2 is an even integer (82), thus, the theoretical ZI - Geometry Factor for Pitting Resistance
 Bakshi et al / SAE Int. J. Mater. Manf. / Volume 7, Issue 3 (June 2014) 697

ZE - Elastic Coefficient [(N/mm2)1/2] Yθ - Temperature Factor

H - Transmission Power (W) YN - Stress Cycle Factor for Bending Stress
N - Rotational Speed (RPM) YZ - Reliability Factor
Dp - Gear Pitch Circle Diameter (mm) ηSR - Efficiency of Configuration with Fixed Carrier
A - Gear Pressure Angle (degree) mGSR - Ratio of Number of Teeth of Ring to Sun
Cd - Required Dynamic Load to give Bearing Reliability of R% ZW - Hardness Ratio Factor for Pitting Resistance
N - Number of Teeth ZN - Stress Cycle Life Factor
NP - Number of Teeth on Planets SH, SF - AGMA Factors of Safety
NS - Number of Teeth on Sun mG - Gear Ratio
NR - Number of Teeth on Ring σa - Alternating Bending Stress
η - Number of Planets σm - Midrange Bending Stress
L - Life (number of revolutions) Ta - Alternating Shear Stress
Kr - Reliability Adjustment Factor Tm - Midrange Shear Stress
IF - Application Impact (depending on load and type of loading) I - Moment of Inertia
Pe - Equivalent Radial Load J - Polar Moment of Inertia
Sut - Ultimate Tensile Strength Kf - Fatigue Stress Concentration for Bending
Se - Yield Strength Kfs - Fatigue Stress Concentration for Torsion
A - Bearing Exponent (Ball Bearing=3; Roller Bearing=10/3) Tm - Midrange Torque
ωs - Speed of Sun (RPM) Ta - Alternating Torque
ωp - Speed of Planet (RPM) Mm - Midrange Bending Moment
ωi - Speed of Input (RPM) Ma - Alternating Bending Moment
ωo - Speed of Output (RPM) C - Distance from Neutral Axis
σall - Allowable Bending Stress (N/mm ) 2
do - Outer Diameter of Shaft
σc, all - Allowable Contact Stress (N/mm ) 2
di - Inner Diameter of Shaft
St - Allowable Bending Stress Number (N/mm2) SH′ - Safety Factor for Shafts
Sc - Allowable Contact Stress Number (N/mm ) 2

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE International. The author is solely responsible for the content of the
paper.