6868 Ijmme Ijens
6868 Ijmme Ijens
6868 Ijmme Ijens
Abstract-- Today’s request for reducing the fuel consumption fuel decreasing. The main cause of truck aerodynamic drag is
of Heavy vehicles is one of the most interesting issues within the due to form drag (pressure drag). Pressure drag on trucks due
automotive industry. Together with the increased fuel price, the to flow separation establishes more than 80% of the total
development of more fuel efficient vehicles has intensified. aerodynamic drag, while frictional drag establishes for the
Recent research about fuel-saving technologies for trucks remaining 20%. Thus, decreasing aerodynamic drag is
displayed that aerodynamic improvement is one of the most
significant for the fuel saving, [4] and [5].
essential technologies when it comes to reducing fuel. The Main
objective of this study is to determine the aerodynamic impact The fuel consumption is an important issue in the road
(drag force) for various profiles in the truck - cabin shape and transport industry; the Centre for Transportation Analysis
Aerodynamic devices added in truck cabin such as, [ Cap of calculated that medium and heavy trucks consumed 6012
truck (with different angle), Gap device (with different length)].
trillion BTU (British thermal unit) in the US during 2014,
To measure the aerodynamic drag produced by the truck,
numerical model studies are undertaken using a 1/50 scale which is the 23% of Domestic Transportation Energy (DTE),
model of standard heavy truck. In this research, a numerical being the second largest consumers after the light vehicles
validation procedure by ANSYS FLUENT ®, computational category, as shown in fig. 1. If the average of drag losses of
fluid dynamics software with various turbulence models is
described for estimation aerodynamic characteristics. It is heavy vehicles were 20%, the drag losses would represent
observed that at the present work a good agreement between the 4.6% of the DTE, which would be equivalent to 1204 trillion
numerical study and the experimental work with the Realizable BTU. This means that any improvement in aerodynamic will
k-ε model with maximum error is about 8%. represent an important fuel saving, [6].
Then, computational fluid dynamic (CFD) investigation is
utilized for each case to compare with respect to coefficient of
drag, Turbulence Kinetic Energy contours, pressure contours,
velocity contours, 3D streamlines and velocity vectors between
a standard 3D truck model with and without aerodynamic
profiles and devices. The results show that the front and mid
fillet radius profile has a significant drag coefficient reduction is
noticed by about 17.75% with optimum dimensions. The top
fillet radius profile has an opposite effect on the drag coefficient
due to the Coandă Effect without cap of truck. When top fillet
radius is utilized with a cap of the truck, the drag coefficient
improvement with an optimism cap truck angle by about 9.92%.
By adding Gap device with different lengths, the drag coefficient
decreasing by about 8.36%. Finally, by using all aerodynamic Fig. 1. Percentages of domestic consumption of transportation energy in the
profiles and devices on the truck - cabin studied at the same time US during 2014, [6].
the improvement in drag coefficient is about 36.03% from a
standard 3D truck model. From [7], It is investigated numerous aerodynamic
Index Term-- Aerodynamics, CFD, Drag reduction, Truck – retrofitting techniques to decrease heavy vehicle fuel
Cabin profiles, fuel consumption. consumption. The numerical models are used to simulate
realistic on-road operations to show the effect of retrofits on
1. INTRODUCTION various vehicle weights and driving cycles. The results are
shown that the fuel economy improvement could be
The heavy commercial vehicles have high fuel consumption,
accomplished from less than 1% to almost 9% of annual
in comparison to other ground vehicles, due to high
mileage. From [8], a numerical simulation is investigated
aerodynamic drag, [1], [2] and [3]. The fuel reduction
unsteady aerodynamic flows affecting the fuel consumption
technologies for trucks show that aerodynamic development
of Class 8 trucks and is validated their results with
is one of the most significant technologies when it comes to
comparison to experimental data. from [9], The
1
Mechanical Engineering Departement. Faculty of Engineering, Albaha University KSA, on Leave from Institute of
Aviation Engineering and Technology, Giza , Egypt, [email protected].
computational fluid dynamic (CFD) analysis is used to cost for the same. Computational fluid dynamics (CFD) has
investigate the effects of adding append devices like become tool for developing, supporting, optimizing,
deflector, cab vane corner, Cab/trailer gap, Front fairing, innovating, verifying and, especially, for validating
Back vane and Base flap on heavy commercial vehicle drag procedures. Verifying and validating forerun steps for the
reduction. It is found about 41% drag reduction by installing code calibration procedure is investigated by [13]. In this
all supplementary parts at their optimized positions. From research, it is taken a 1/50 scale detailed model of a standard
[10], the numerical models with Shear tress transportation heavy truck and it is measured the drag with velocity
(SST) turbulence model is used to investigate both experimentally and numerically. It is clarified the
unmodified and modified profile of the truck-trailer. The aerodynamic impact of various profiles in the cabin truck
results showed that the profile modification cases decrease in shape and Aerodynamic devices added in truck cabin such as,
aerodynamic drag up to 21 %, which decreases the fuel [Cap of truck (with different angle), Gap filling (with
consumption by 4 liters for 100 km for the diesel-powered different length)].
truck. From [11], it is carried out a CFD study in order to
design and optimize the cabin geometry and its various parts 2.1. Experimental Set up
for drag decrease including the side deflectors, the mirrors,
The institute of aviation engineering and technology has an
and the sun visor. From [12], a 1/10 the scale semi-trailer
open-circuit suction wind tunnel in the aerodynamics
truck model is used in a wind tunnel at various speeds and
laboratory. The tunnel ducting is composed of four major
yaw angles using different combinations of fairings to find
duct components. The components are the settling chamber,
drag reduction. It is found that any improvement in the front
contraction cone, the test section, the diffuser and then fan
area of the truck has the most significant effect on drag. It is
housing. The wind tunnel test section is 24 inch length and
obtained a 26% drag reduction for the best combination. It is
12 × 12- Inch cross-section. The maximum velocity that can
calculated that the cab roof fairing alone can reduce about
create around 30 m/s with 9 blade fan, as shown in Fig. 2.
17% of drag, and if joined with the fairing for the tractor-
The wind tunnel has three-component sting balance. The
trailer gap, the drag reduction can be up to 25.5%.
sting balance is used to measure all two forces (drag, lift) and
one moment at a time. The reading of the sting balance is
2. VALIDATION PROCEDERS
indicated in the inductor panel, behind the panel signal
The main object of this research is to measure the conditioning cards, which convert the volt signals from the
aerodynamic drag force of the truck by experimental and sting balance and pressure transducers to forces, moment,
numerical work. However, the experimental techniques are pressure, and velocity, as shown in fig. 2.
quite laborious and surely cost more than CFD techniques
In this research, only aerodynamic drag force (D) data and its dimensionless parameter drag coefficient (CD) are
presented. The CD is calculated by using the following formula:
D
CD (1)
0.5 v 2 A
Where: D is the drag force, A is project area of the truck, ρ A one-fifty scale detailed model of a standard heavy truck is
and v are the density and air velocity used as a baseline truck, as shown in fig. 3(b). From figure 3
(a), the model is placed in the test section of wind tunnel and experienced and the results are averaged for minimizing the
it is measured drag forces with variable velocities (from 12 to further possible errors in the raw experimental data.
25 m/s). Several data collections are collected at each speed
Velocity Inlet
1.5 cm
5.5 cm
24 cm 6.3 cm
10 cm
1.5 cm 3.7 cm
three-component Tunnel floor
3 cm
sting balance
41.8 cm
(a) A standard heavy truck model in side the tunnel (b) A standard heavy truck model with dimensions
2.2. Numerical Analysis same properties in the experimental work are used in the
present simulation to validate the code. The free stream
According to a standard heavy truck in experimental temperature is 300 K, which is the same as the environmental
method, it is designated the same model dimensions, as temperature. The density of the air at the given temperature
shown in Fig. 3(b). The flow field, temperature, pressure and is ρ=1.225kg/m3, the pressure is 101325 Pa and the viscosity
velocity inlet in the numerical model of 3D truck, are solved is μ=1.7894×10-5 kg/m s. A segregated, implicit solver is
using ANSYS FLUENT@19. utilized (ANSYS FLUENT® processor) Calculations are
done for velocity ranging from 10 m/s to 25 m/s (Re from
2.2.1. Boundary Conditions and Meshing 350000 to 650000). The pressure outlet (number 2) is 101325
Pa as the environmental pressure. The Standard heavy truck
At first, a standard heavy truck is modeled as two profile (number 3) as shown in Fig 4, is considered adiabatic
boxes by using Design modular and boundary conditions are and no slip wall. The two vertical and horizontal planes
defined. The Reynolds number (Re) for the velocity inlet around 3D truck are considered adiabatic and no slip wall, as
boundary (number 1) is 5.75231x105, as shown in Fig 4. The shown in Fig 4.
2- Pressure outlet
1- Velocity Inlet
2 4
4 3
1
3- 3D Truck profile
5- Two Horizontal walls
In order to obtain precise the drag force on a standard and 17], it is presented a number of grid generation methods
heavy truck, grids near the 3D truck volume must be dense to construct high quality single- and multi-block structure
sufficient and computed fields must be large sufficient to grid for complex shape. In present work using multi-block
satisfy higher accuracy and reducing time. Using 3D unstructured grid to increase grids near 3D truck volume by
unstructured tetrahedral mesh, increasing mesh cells, it is creating block around truck and making body inflation, as
needed super computer and more time to solve the problem. shown in fig 5. It is created face inflation and face sizing in
Many researchers developed various methods and tools to truck volume faces which reduce the computational time and
overcome the problems described above. From [14, 15, 16 obtain accurate drag, as shown in fig 5.
(A) (B)
Fig. 5. A standard 3D truck of the meshed control volume (a) Multi Block Method for meshing the truck to generate body sizing (b) Mesh
face sizing on truck surface.
The mesh is generated using ANSYS FLUENT MESH® truck walls and it is processed a growth rate of 1.02 and a
19 as a pre-processor and mesh generator. The body sizing is maximum element size of 0.1 m. The grid is of the
applied with a minimum element size of 1×10-4 m at the unstructured tetrahedral-hybrid grid type. The meshed
control volume of block around 3D tuck volume is shown in results. The appropriate number of nodes is determined by
Fig 5. The total number of cells in the full grid is about increasing the number of cells until the mesh is sufficiently
3000000 cells. fine, so that further refinement does not change the results.
To examine the independence of the results to cell number,
2.2.2. Grid dependency check and Verification of six kinds of mesh are generated. Figure 6 shows the effect of
Numerical Model the number of grid cells in coefficient of drag at velocity inlet
20 m/s. It is taken 2500000 cells to reducing time of solution.
The first step in performing a CFD simulation should be
to investigate the effect of the mesh size on the solution
0.9100
0.9000
0.8900
0.8800
0.8700
CD
0.8600
0.8500
0.8400
0.8300
0.8200
0.8100
0 1000000 2000000 3000000 4000000
NUMBER OF GRID CELLS
Fig. 6. Curve of drag coefficient at velocity inlet 20 m/s against number of grid cells.
A similar Numerical Model a standard heavy truck of the drag coefficient (CD) with velocity from 12 to 25 m/s
same previously-mentioned grid size and type is developed, (Reynolds number from 350000 to 650000) of numerical and
for verifying numerical model with the experiment model experimental studies, plotted on the same axes and scale for
study with the same boundary conditions, as shown in Fig 7. comparison. It is found that the Realizable k – ε model with
Compare the results of the numerical model by Realizable k enhanced wall treatment model with a maximum error about
– ε model with enhanced wall treatment, the standard model 8% more accurately than k–𝜔 and Spalart-Allmaras
k–𝜔 model and Spalart-Allmaras model to those of the turbulence models.
experimental measurements. The results show a good
agreement of a drag coefficient with the corresponding values
in the experimental model measurements. Figure 7 shows the
3. MODELING, RESULTS AND DISCUSSION 3.1. Effect of Front Fillet Radius Ratio of
The aerodynamic profiles and devices considered here are cabin (FFRR)
defined as for cabin and gab between cabin, and container. At First, it is investigated front fillet radius ratio of cabin
first, a standard 3D Truck is modeled with aerodynamic (FFRR). The FFRR is defined the ratio of fillet radius (R1)
profiles and devices with variable dimensions by Design and the height of front cabin (L1), as shown in figure 8. The
modular, then the meshing is performed and boundary FFRR is created with different ratio [0, 0.2, 0.4, 0.58 and 0.8]
conditions are defined. The vehicle is assumed to have speed and it is illustrated the effect of it on the drag reduction.
of 20 m/s (Re= 575321) at 300 K. The density of the air at the
given temperature is ρ=1.225kg/m3, the static pressure is
101325 Pa and the viscosity is μ=1.7894×10-5 kg/m s.
Fig. 8. A standard heavy truck model with FFRR definitions as FFRR = [R 1/L1].
The following figure 9 shows velocity vectors (m/s) on the figure 9(B, C, D), the standard 3D truck with FFRR=0.2, 0.4
standard 3D truck without and with FFRR at velocity 20 m/s. and 0.8 the vortices in front, around cabin and above
it is increases FFRR from 0 to 0.8, as shown in Figure 9(A) container are decreased that is indicated a reduction in
to Figure 9(D). At figure 9(A), the standard heavy truck amount of drag force. The smallest vortices investigate on
without FFRR the vortices in front, around cabin and above front, around cabin and above container with FFRR= 0.8 that
container are significant that is indicated high drag force. At is indicated a minimum amount of drag force.
Fig. 9. Velocity vectors of A standard 3D truck model (A) FFRR=0, (B) FFRR=0.2,
(C) FFRR=0.4, (D) FFRR=0.8.
The standard 3D truck without aerodynamic profiles and devices at velocity 20 m/s has a drag coefficient about 0.82. Then, it
is considered the effect of FFRR on drag force. The Results show CD decreasing with increasing FFRR, as shown in figure 10.
The improvement in CD is about 17% at FFRR= 0.8.
Fig. 10. Drag coefficient of a standard 3D Truck with FFRR [0, 0.2, 0.4, 0.58 and 0.8].
Fig. 11. A standard heavy truck model with MFRR = [R2/L1], [0, 0.2 and 0.4] and FFRR=0.8.
The following figure 12 shows a Velocity contours (m/s) and velocity vectors on the standard 3D truck with FFRR=0.8 and
varies MFRR (0, 0.2 and 0.4) at velocity 20 m/s. The high intensity blue area (stagnation points and vortices regions in truck
volume) located on the front and above surface of cabin due to sharp edge in front of cabin of the truck due to flow separation
that is increasing a drag force. The following figure 12(A) show a static Velocity contours (m/s) and velocity vectors on the
standard 3D truck with FFRR=0.8 and MFRR=0. The high intensity blue area is located in mid fillet. Figure 12(B) show a
velocity contours (m/s) and velocity vectors on the standard 3D truck with FFRR=0.8 and MFRR=0.2. The high intensity blue
area is decreased on mid fillet that is decreased the drag force. Figure 12(C) show a static Velocity contours (m/s) and velocity
vectors on the standard 3D truck with FFRR=0.8 and MFRR=0.4. The high intensity blue area is increase on mid fillet that is
increased the drag force. The effect of MFRR is very small in Drag reduction.
Fig. 12. Velocity contours and vectors of A standard 3D truck model with FFRR=0.8
(A) MFRR=0 (B) MFRR=0.2 (C) MFRR=0.4.
Then, It is illustrated the effect of MFRR with previous FFRR=0.8. The CD is decreasing to optimum value and then increase
with increasing MFRR up to optimum value. The improvement in C D is about 0.75% by optimum MFRR=0.2 from previous
cases, as shown in figure 13.
Fig. 13. Drag coefficient of a standard 3D Truck with FFRR=0.8 and MFRR [0, 0.2, 0.4, 0.58 and 0.8]
Fig. 14. A standard heavy truck model with TFRR = [R3/L2], [0, 0.2, 0.3, 0.6 and 0.8] MFRR=0.2 and FFRR=0.8.
The following figure 15 shows a Velocity contours (m/s) on the standard 3D truck with FFRR=0.8, MFRR=0.2 and varies
TFRR [0, 0.2, 0.3, 0.6 and 0.8] at velocity 20 m/s, as shown in figure 15. From figure 15(A), It is illustrated the flow attached
on top fillet with highest velocity and the flow lift off over above the container of truck due to Coandă effect. It is observed the
high intensity blue area increasing on front and above container that is increasing the drag force. From figure 15 (B, C and D),
It is increased TFRR that it observed the high intensity blue area decreasing on front and above container that is decreasing the
drag force compare to case (A). The vortex is illustrated front and above truck container by using TFRR that is caused increasing
in drag force.
(B) (A)
Significant Recirculation Significant Recirculation
(C) (D)
Fig. 15. Velocity contours of A standard 3D truck model with FFRR=0.8 and MFRR=0.2
(A) TFRR=0.2, (B) TFRR=0.3, (C) TFRR=0.6, (D) TFRR=0.8.
Then, It is illustrated the effect of TFRR with previous FFRR=0.8 and MFRR=0.2. The CD is increasing from previous case
with increasing TFRR due to Coandă effect that’s increasing recirculation above the container. The TFRR has opposite effect
on drag coefficient, as shown in figure 16.
Fig. 16. Drag coefficient of a standard 3D Truck with FFRR=0.8 and MFRR [0, 0.2, 0.4, 0.58 and 0.8]
Fig. 17. A standard heavy truck model with TFRR = [R3/L2], [0, 0.2, 0.3, 0.6 and 0.8] MFRR=0.2 and FFRR=0.8.
The following figure 18 shows a Velocity vectors (m/s) and force. When it is increasing CTAR above the optimum value,
pressure contours (Pa) on the standard 3D truck with the flow separation is increasing. Figure 18(C) shows CTAR
FFRR=0.8, MFRR=0.2, TFRR=0.2 and varies CTAR [0.805, as 0.916 Cap truck angle as 165). It illustrates high vortices
0.833, 0.861, 0.889 and 0.916] at velocity 20 m/s. from figure above container that is increasing the drag force. The high
18(A), It is taken CTAR as 0.805 (Cap truck angle as 145). intensity red area (stagnation points in truck volume) is
The angle is low and the flow lift off over truck container increased in front of cabin and container that is increased the
surface and it illustrates significant vortices. It is caused drag force.
reverse pressure and flow separation that’s increase the drag According to Figure 19, the best angle of cab truck is
force. The pressure is increasing in front of cabin truck and obtained at certain TFRR and when the angle exceeds or
cab that is increasing the drag force. The high intensity red lessens this angle, the drag coefficient increases because of
area (stagnation points in truck volume) value is 245.76 Pa. disturbing the airflow. At TFRR as 0.2, the best CTAR is
When it is increasing CTAR, the flow separation is 0.861 and the drag coefficient is 0.6488. At TFRR as 0.3, the
decreasing. Figure 18(B) shows the optimum CTAR as 0.861 best CTAR is 0.805 and the drag coefficient is 0.6534. At
Cap truck angle as 155). It illustrates small vortices above TFRR as 0.6, the best CTAR is and the drag coefficient is
container that is decreasing the drag force. The high intensity 0.6672.the best case is the truck at TFRR=0.2 and
red area (stagnation points in truck volume) is decreased as CTAR=0.861that is made improvement in the drag about
243.7 Pa in front and cap truck that is decreased the drag 9.92%.
Fig. 19. Drag coefficient of a standard 3D Truck with FFRR=0.8, MFRR=0.2 and variable TFRR and CTAR.
Fig. 20. A standard heavy truck model with FFRR=0.8, MFRR=0.2, TFRR =0.2, CTAR=0.861 and variation of GLDR.
The following figure 21 shows a Velocity vectors (m/s) on front and top view of the standard 3D truck with FFRR=0.8,
MFRR=0.2, TFRR=0.2, CTAR 0.861 and GLDR of [0, 0.5 and 1] at velocity 20 m/s. from figure 21(A), It is taken GLDR as
zero. It illustrates two significant vortices between cabin and container and above container that is increased the drag force.
When it is increasing GLDR to 0.5, the vortices between cabin and container and above container are decreasing that is
decreased the drag force, as shown in figure 21(B). From figure 21(C), the GLDR is one. The vortices between cabin and
container and above container is dissipated that is decreasing the drag force.
According to Figure 22, when it is increasing the GLDR from zero to one, the drag coefficient is increasing. At GLDR as one,
the drag coefficient is improved by about 8.36% from previous case.
Fig. 22. Drag coefficient of a standard 3D Truck at FFRR=0.8, MFRR=0.2, TFRR=0.2 and CTAR =0.861 with variation of GLDR.
3.6. Effect of profiles and devices with best ratio on cabin of truck
Then a standard 3D truck with FFRR=0.8, MFRR=0.2, TFRR=0.2, CTAR=0.861 and GLDR=1, is investigated, as shown in
fig. 23.
Fig. 23. A standard heavy truck model with FFRR=0.8, MFRR=0.2, TFRR =0.2, CTAR=0.861 and GLDR=1.
The following figure 24(A) shows Contours of turbulence kinetic energy (J/kg) around the standard 3D truck without
aerodynamic profiles and devices at velocity 20 m/s. It is observed that, at the front, above cabin, gab between cabin and
container, above container and at rear of container there is formation of high turbulence and formations of turbulent eddies
because of adverse pressure gradient and the flow separation occur at edges. From figure 24(B), the flow around the truck body
is smoothing and the turbulence and wake formation are decreases due to use all aerodynamic profiles and devices.
(A) (B)
Fig. 24. Turbulence Kinetic Energy (A) A standard 3D truck model without aerodynamic profiles and devices (B) A standard 3D truck model with all
aerodynamic profiles and devices studied.
The following figure 25(A) shows the 3D stream lines around the standard 3D truck without aerodynamic profiles and devices
at velocity 20 m/s. These stream lines show the big recirculation and the air flow is random and non-uniform around the truck
volume. From figure 25(B), the recirculation is very small and the air flow is more uniform and aligned to the surface around
the standard 3D truck with all aerodynamic profiles and devices, that’s decreasing drag coefficient. With installing all
aerodynamic profiles and devices at their best ratios and positions on cabin of a standard 3D truck, about 36.03 % drag reduction
is enhanced compared to the standard 3D truck.
A B
Fig. 25. 3D stream lines (A) A standard 3D truck model without aerodynamic profiles and devices, (B) A standard 3D truck model with all aerodynamic
profiles and devices studied.
NOMENCLATURE
Symbol and Abbreviations Meaning and units
A Project area of truck (m2)
CFD computational fluid dynamic
CD Drag coefficient
CTAR cab truck angle ratio
D Drag force (N)
EWT education wind tunnel
FAR flap area ratio
FFAR flat flap angle ratio
FFRR front fillet radius ratio of cabin
FLR flap length Ratio
GLDR Gap length device Ratio
MFRR mid fillet radius ratio
UFRR under fillet radius Ratio
TFRR top fillet radius ratio of cabin