COMPARATIVE COMPOSITE AND CONVENTIONAL DRIVE SHAFT

ANALYSIS

Marko Denić 1, Zorica Đorđević 2, Vesna Marjanović 3, Nenad Petrović 4, Nenad Kostić 5

UDC:621,824

1. INTRODUCTION

The automobile industry has seen a growth in implementation of composite

materials in the past years, however it is still a slow process, as the material properties for
various load cases is still being researched. Following this trend of using lightweight
materials to replace steel components more and more parts are being produced using
composites. However composites are still most frequently found in static and cosmetic
elements. Drive shafts made of composites have most recently been featured in some BMW
M series models. Use of composite materials for propulsion components is still a developing
field of study. Drive shafts (cardan shafts) are main driving components for transferring
torque in vehicles. The use of composite materials for making drive shafts implies
decreasing the mass compared to conventional material shafts. In the use of composite
materials the orientation of fibers plays an important role in load distribution and stress
characteristics. Depending on the load case, fibers need to be placed at optimal angles to
minimize deformation of the shaft. The base of this research is to analyze and compare drive
shaft numerical calculations for steel and composite drive shafts in the same loading
scenarios.
Kaviprakash et al. [3] conducted research on fiber orientation in laminar
composites, as well as the order of layering. Their optimization was done in ANSYS for
hybrid shafts made from high strength carbon fibers, high module carbon fibers, Kevlar, and
epoxy resins. The results were compared to conventional shafts and showed improvements
in decreasing mass of approximately 79%, lowering stress as well as improving fuel
economy. Hatwar and Dalu [2] analyzed E Glass and carbon fiber shafts in combination
with epoxy resins. Static analyses conducted in ANSYS were compared to analytical results
for maximum shear strain, shear stress, equivalent stress, natural frequency, and mass, for
steel shafts, carbon, and glass shafts. The results showed a decrease of mass by over 72% for
composite shafts with similar deformation and stress characteristics. Rothe and Bombatkar
[5] tested a composite shaft made from high strength carbon fibers using static, modal, and
buckling analyses. Bhajantri et al. [1] replaced a two-part steel shaft for a single-part
composite shaft decreasing the mass by up to 50% and concluding optimal parameters for
the orientation of fibers in layering using regression analyses. Ravi [4] analyzed composite

1
Marko Denić, Ph. D. student, University of Kragujevac, Faculty of Engineering, Sestre Janjić 6,
34000 Kragujevac, [email protected]
2
Zorica Đorđević, Prof., University of Kragujevac, Faculty of Engineering, Sestre Janjić 6, 34000
Kragujevac, [email protected]
3
Vesna Marjanović, Prof., University of Kragujevac, Faculty of Engineering, Sestre Janjić 6, 34000
Kragujevac, [email protected]
4 Nenad Petrović, Assist., University of Kragujevac, Faculty of Engineering, Sestre Janjić 6, 34000
Kragujevac, [email protected]
5
Nenad Kostić, Research assoc., University of Kragujevac, Faculty of Engineering, Sestre Janjić 6,
34000 Kragujevac, nkostić@kg.ac.rs
Volume 42, Number 4, 2016
44 Marko Denić, Zorica Đorđević, Vesna Marjanović, Nenad Petrović, Nenad Kostić

shafts made from carbon fibers using a combination of tetrahedral and hexagonal finite
elements. The results were compared to conventional shafts and showed a decrease in mass
by 24%. The analysis covered hollow and full shafts concluding favourable dimension ratios
for hollow shafts. Sivakandhan and Prabhu [6] optimized fiber angles in symmetrical layers
of equal thicknesses for composite shafts. They optimized composite drive shafts achieving
decreases in mass of 15% and 72% respectively compared to conventional shafts.
This paper analyses single-part laminar composite drive shafts, for possible use in
heavy duty trucks, and gives a comparison of numerical calculations for both the composite
shaft and its steel and aluminium counterparts. Three of the most frequently used composite
materials are used in this research, as determined by literature review. Results are shown for
all parameters for both composite and metal shafts. Analysed properties are twist angle,
deflection, eigenfrequencies, and mass. Additionally variations of the number of layers, their
direction and possible combination with conventional materials are also explored.
Test example drive shafts were tested in Autodesk Helius Composites 2016. This
new software is easy to use and it is accurate in simple load cases in comparison to other
software, which use FEA method, such as ANSYS, Abaqus, and DS CATIA.

2. SELECTION OF MATERIALS

The comparative analyses conducted in this paper aim to show the difference in
performance and mass of conventional, metal, shafts and various frequently used composite
shafts. Calculations were performed in ANSYS and verified in Helius Composites for both
metal and composite shafts.
The conventional materials used for the purposes of this research are S275JR, as it
has widespread use in the automobile industry, and 6061-T6 Aluminium. Material
characteristics are taken from the standard material library in ANSYS and are given in table
1. A survey of available literature on the subject shows that the most frequently used
composite materials for drive shafts are E Glass Epoxy, High Strength Carbon, and Kevlar
(Kevlar-49). Material characteristics for E Glass and Kevlar-49 are taken from the material
library in Helius Composite and are given in table 1. HSC characteristics were used from [3]
and a material with those properties was created in the software manually.

Table 1 Characteristic of steel, aluminum and composite materials

12 13 23 ,
E11, E22, E33, G12, G13, G23, , , ,
Material [kg/
[MPa] [MPa] [MPa] [MPa] [MPa] [MPa]
[/] [/] [/] m3]
2.1∙ 2.1∙ 2.1∙ 8∙ 7.6∙ 7.6∙
Steel 0.3 0.3 0.3 7860
105 105 105 104 104 104
Alumi- 6.83∙ 6.83∙ 6.83∙ 2.62∙ 2.62∙ 2.62∙
0.33 0.33 0.33 2710
num 104 104 104 104 104 104
4.48∙ 1.24∙ 1.24∙ 5.52∙ 5.52∙ 3.60∙
E Glass 0.28 0.28 0.36 2080
104 104 104 103 103 103
1.35∙ 7∙ 9.26∙ 5.8∙ 6.15∙ 3.08∙
HSC 0.31 0.31 0.50 1580
105 103 103 103 103 103
Kevlar- 7.58∙ 5.52∙ 5.52∙ 2.07∙ 2.07∙ 1.54∙
0.34 0.34 0.47 1380
49 104 103 103 103 103 103

Volume 42, Number 4, 2016

Comparative composite and conventional drive shaft analysis 45

3. EXPERIMENT

Literature suggests general dimensions of shafts in the automobile industry as well

as their loads. This research will be conducted on a modified version of the example from
[2]. The hallow drive shaft is 1000 mm long, 100 mm in diameter, while the wall thickness
is 3.32 mm. The shaft is loaded with 3000 Nm of torque on one end, while the other end is
fixed. In order to analyze bending of the shaft (deflection) the standard procedure for
simulating bending loads was used in three points with a force of 1000 N.
The steel drive shaft was first calculated for maximal shear stresses, twist angle and
deflection of the shaft on bending analytically. In order to verify the analytical method, a
numerical analysis was conducted under the same loading and constraint conditions in
ANSYS. The finite element mesh consists of 15876 tetrahedral elements with 5 mm sides
and 108415 nodes. Calculated values as well as the mass of the steel shaft are given in table
2.
Table 2 Analytical and Numerical results comparison for steel shaft
Analytical Numerical Difference [%]
Shear Stress [MPa] 63.58 63.917 0.527
Twist angle [ ] 0.911 0.947 3.801
Deflection [mm] 0.076 0.084 9.524
Mass [kg] 7.916 7.916 0

Comparing the calculated values, it can be concluded that the differences in results
are less than 10%, and given the magnitude of the values, the numerical results are adopted
as valid.
Further calculations of the composite and aluminium shafts, conducted in Autodesk
Helius Composites 2016, will be compared to the numerical results from ANSYS for the
steel shaft.

Volume 42, Number 4, 2016

46 Marko Denić, Zorica Đorđević, Vesna Marjanović, Nenad Petrović, Nenad Kostić

4. RESULTS

Numerical calculation resaults are shown and compared graphically to best

illustrate the difference in characteristics of the examined materials of shafts. Results
attained are mass, twist angle, deflection, and eigenfrequency values for steel, aluminum, E
Glass, HSC and Kevlar-49 shafts.
The greatest benefit of using composite shafts in the automobile industry is
decreasing the mass of the vehicle. The masses of the examined shafts are given in figure 1.

Figure 1 Masses of drive shafts

The twist angle is the other important parameter calculated for the shafts. Figure 2
shows the twist angle values in degrees for all calculated drive shafts.

Figure 2 Twist angles of drive shafts

Even though drive shafts in automobiles are most frequently just subjected to
torsion, they are also calculated for the case of bending, usually by subjecting them to forces
in three points. The maximal deflection is calculated in the middle of the shaft. There are no
suggestions stipulated for deflection, however it is best to keep deflection under a few
millimetres per meter of length. As the drive shaft has enough clearance to withstand such
deformations due to stochastic changes in terrain over which the vehicle is moving these
small deflections are acceptable. Figure 3 shows maximal deflection values in the middle of
the shaft.

Volume 42, Number 4, 2016

Comparative composite and conventional drive shaft analysis 47

Figure 3 Deflection of drive shafts

Figure 4 graphically shows values of eigenfrequencies (for the 1st mod) for each
shaft. One of the main reason for practical implementation of two-part steel drive shafts is
that the single-part drive shaft exhibited unfavourable eigenfrequencies due to the length of
the shaft. Due to a drastic difference in eigenfrequencies of composite material shafts, two-
part metal shafts can be exchanged for single-part composite shafts.

Figure 4 Eigenfrequencies of drive shafts (1st mode)

All fibers for the calcul ompo n n l o 45 n o l

as suggested by literature as being optimal positioning of fibers in the layers of the
composite for more than one loading scenarios.

5. COMPOSITE VARIATIONS

Further analysis is directed towards determining the best variation of the composite
drive shafts. The variation of the number of layers by doubling the number in every
sequential iteration is conducted to determine the influence of the change in the number of
layers on the previously calculated parameters. The overall thickness of the layers is 3.32
mm and is constant. All the same calculations were conducted as in the previous section,
except that the experiment of the change in the number of layers is only conducted on the E
Glass shaft, as it is most frequently used in literature. The results of the varied number of
layers are shown in table 3.

Volume 42, Number 4, 2016

48 Marko Denić, Zorica Đorđević, Vesna Marjanović, Nenad Petrović, Nenad Kostić

Table 3 Influence of the change in the number of layers on deflection, torsion and
eigenfrequency for E Glass drive shafts
Number of layers 4 8 16 32
Deflection [mm] 1.0872 1.0739 0.0708 1.07
Angle of Twist [⁰] 6.039 5.759 5.693 5.676
st
Eigenfrequency(1 mode)
341.082 341.191 343.694 343.819
[Hz]

As the change in the number of layers has been found to not have a big influence
on shaft characteristics, the next variation in this experiment is directed towards determining
the influence of the layer order. Various orientation of the angle of fibers when the fibers in
the layers are oriented from 0⁰ to 90⁰ is tested for the four layer E Glass shaft. Two
variations are tested with layer combinations of -45⁰, 0⁰, 45⁰, and 90⁰ in variations and the
resulting characteristics presented in table 4, with the first combination being the same as in
the first comparison.

Table 4 Influence of fiber direction changes in layers

Directions of fibers combinations in 45⁰/-45⁰/45⁰/- 0⁰/- 45⁰/0⁰/90⁰/-
layers 45⁰ 45⁰/45⁰/90⁰ 45⁰
Deflection [mm] 1.0872 0.926 0.822
Angle of Twist [⁰] 6.039 8.152 9.873
st
Vibration (1 mode) [Hz] 341.082 369.584 392.366

Due to the high costs of composite materials, there is a tendency to combine

conventional with composite materials in drive shaft design to form hybrid drive shafts. This
way the best of both materials characteristics can be exploited. The idea is to start off with a
steel or aluminium base and layer over them with composites. The distribution of stress in
the layers needs to be accounted for in order to avoid having the metal base from
transmitting the greater part of the load, thereby not using the potential of the composite
layers. This research tested steel and aluminium base shafts in combination with E Glass
composite in four layers 1.66 mm thick oriented in the setup of 45/-45/45/-45. The steel and
aluminium bases have thicknesses of 1.66 in both cases. Table 5 gives the characteristics of
two hybrid shafts and compares the values to their base metal and purely E Glass shafts.

Volume 42, Number 4, 2016

Comparative composite and conventional drive shaft analysis 49

Table 5 Comparison of hybrid steel/E Glass and Aluminum/E Glass shaft to the metal and E
Glass composite drive shaft characteristics
Steel/ Aluminum
Material Steel Aluminum E Glass
E Glass /E Glass

Deflection [mm] 0.342 0.088 0.577 0.259 1.087

Angle of Twist [⁰] 2.594 0.947 4.11 2.699 6.358

Mass [kg] 5.01 7.916 2.414 2.723 2.093

Eigenfrequency
393.018 614.941 436.039 611.642 341.082
(1st mode) [Hz]

6. CONCLUSIONS

This paper analysed the use of composite materials and their variations and
compared their characteristics (deflection, torsion, eigenfrequency and mass) with those of
conventional shafts.
The use of composite materials for creating drive shafts can significantly decrease
the mass of this component. The difference in weight among the composite shafts is around
10%, and the composite with the smallest mass is Kevlar-49 with a decrease in mass of
82.36%. Other tested characteristics for composite shafts do not show such drastic
differences.
High strength carbon (HSC) showed itself as the least susceptible to torsion, with a
twist angle of 2.485⁰, which is less than that of its aluminium counterpart, but over 2.6 times
greater than that of the corresponding steel shaft. E Glass has performed the worst under
these circumstances with a twist angle of 6.358⁰ which is still acceptable.
Deflection has the greatest variance in results between the materials. Of the
composite shafts HSC performed the best, while Kevlar-49 had the greatest deflection,
however all deflection values can be considered acceptable due to their minuscule values.
Eigenfrequencies are significantly lower for composite materials than for conventional steel
and aluminum shafts, as was expected.
Further variations of the E Glass composite by changing the number of layers made
small decreases the twist angle with the increase of the number of layers, and had an overall
insignificant influence on the characteristics of the shaft in comparison to the increased
complexity of production. The changes in fiber orientation in the layers only slightly
improved deflection from the initial setup, while worsening other characteristics. Therefore
it can be concluded that the initially calculated setup of composites with 4 layers with angles
of fibers changing from +45⁰ to -45⁰ in each layer to be optimal.
Due to the costliness of composite materials a compromise in the design solution
can give favourable results compared to conventional design solutions. The combination of
E Glass with steel, and E Glass with aluminium was tested and demonstrated improvements
in deflection (around 31% increase for Glass/steel), and twist angle (around 40% decrease
for E Glass/steel), while having a greater mass and eigenfrequency compared to a purely E

Volume 42, Number 4, 2016

50 Marko Denić, Zorica Đorđević, Vesna Marjanović, Nenad Petrović, Nenad Kostić

Glass composite shaft, but a significantly lower mass than their corresponding metal
counterparts.
Any decrease of mass in automobiles is in direct correlation with the decrease of
fuel consumption. An improvement of fuel economy through further research into, and
implementation of composite materials inevitably leads to a decrease of automobiles
negative effects on the environment.

REFERENCES

[1] B j n S. V., B j n C. S., S n olk M. A., Am pu S. S., „D n n

An l o Compo D v S “, In n on l Jou n l o En n n R &
Technology (IJERT), ISSN:2321-7308, Vol. 3 Special Issue 3, May 2014.
[2] H w K. P., D lu S. R., „D n n An l o Compo D v S “,
International Journal of Science and Research (IJSR), ISSN (Online):2319-7064, Vol.
4 Issue 4, April 2015.
[3] Kaviprakash G., Lawrence I. D., Kannan C. R., Regan A. P., „D n n An l o
Compo D v S o Au omo v Appl on“, In n on l Jou n l o
Engineering Research & Technology (IJERT), ISSN:2278-0181, Vol. 3 Issue 11,
November 2014.
[4] R v A., „D n, Comp on n An l o Compo D v Shaft for an
Au omob l “, In n on l R v w o Appl En n n R , ISSN:2248-
9967, Vol. 4, Number 1 (2014), pp. 21-28.
[5] Ro G. C., Bomb k S. A., „D n n An l o Compo M l D v
S “, In n on l Jou n l o Inov v nd Emerging Research in Engingeering
(IJERT), ISSN:2394-5494, Vol. 2 Special Issue 1, 2015.
[6] S v k n n C., P b u P. S., „Op mum D n n An l o D v S n
Compo M l“, M lS n R In , Vol. 8(2), 125-130 (2011).

Volume 42, Number 4, 2016