| |
By Sandeep Sovani and Ashok Khondge, Fluent Inc.; Sunil Jain, International Truck and Engine Corp., Fort Wayne, IN
View the pdf of this article
Two views of the AISDS truck model
Large tractor trailer trucks are a major mode of transporting goods all
over the world. From an aerodynamic point of view, large trucks are
bluff bodies in a high-speed flow subject to enormous drag forces. It
is estimated that a modern truck with an average drag coefficient of 0.6
driving at 65 mph spends 50% of its fuel on overcoming aerodynamic drag.
Many aerodynamicists believe that these drag coefficients can be reduced
by as much as 25%, which could translate to billions of dollars of fuel
savings
annually. A number of devices, such as boat-tails and pneumatic suction/
blowing equipment are currently being conceived and evaluated to
reduce truck drag. To understand and optimize these devices, it is important
to have a sound understanding of the flow field around a truck and
in its wake with and without the devices in place.
It is well known that the flow around a truck and in its wake is highly
turbulent and unsteady. While turbulent flows can be modeled accurately
with direct numerical simulation (DNS) or large eddy simulation (LES),
these approaches need impractically large computational resources for
solving high Reynolds number flows. The detached eddy simulation (DES)
model, new in FLUENT 6.1, is a modified version of the one-equation Spalart-Allmaras
equation is solved.
Time variation of the drag coefficient; the average drag is within 0.8%
of the measured value
Vorticity iso-surface around the vehicle, colored by velocity magnitude,
shows the transient flow field in the wake region
A 1/8th scale replica of a Class 8 tractor trailer truck (referred to
as an
AISDS model, courtesy of International Truck and Engine Corporation) has
recently been solved using the DES turbulence model in FLUENT. A constant
flow velocity of 85.9 m/s, corresponding to a vehicle-width-based
Reynolds number of 1.88 million, was imposed at 0° yaw at the upstream
flow boundary. This flow condition was chosen to match the experimental
data. A new hex-predominant strategy was used to create the mesh. In
it the geometrically complex parts of the domain (such as the underbody
and regions near the vehicle surfaces) were meshed with tetrahedral elements,
while the rest of the domain was filled with hexahedra. A transient
CFD simulation was performed. The inherently unsteady nature of the turbulent
flow around the truck and in its wake was captured accurately in
the simulations. The boundary layer on the trailer separates a short distance
downstream of the trailer leading edges, shedding vortices occasionally.
After
convecting downstream, these vortices play a role in determining the structure
of the truck’s wake, which extends several vehicle lengths behind
the
truck. The wake structure is also strongly influenced by the flow emerging
behind the vehicle from underneath the trailer. The time-averaged drag
coefficient
taken from the CFD simulation was found to be within 0.8% of the
corresponding experimentally measured value.
|
|
|
