As a high quality supplier of
HVAC systems to the automotive industry, Behr has developed
an accurate methodology for predicting thermal comfort of
passengers. This article describes how Behr uses STAR-CD
as the central tool for prediction of thermal comfort in
the early stages of a vehicle design.
The parametric cabin modeler
Mesh generation for a passenger compartment based on CAD geometry is a time-consuming
task. The many CAD models that describe the cabin need to be collected, cleaned-up
and matched together before meshing can commence. Depending whether a structured
or unstructured meshing strategy is adopted, this whole process takes from
three to eight weeks. In most cases, the CAD models of the car are not available
to a supplier at the early phase of a development process.
 |
Fig. 1: Different types of vehicles created with the parametric cabin
modeler: A - sedan, B - truck, C - hatchback |
To avoid these problems, and to speed-up the process, Behr has developed
a parametric cabin modeler in collaboration with ICEM CFD. The geometry of
the model is based on more than 200 parameters that can be specified from
the new car design (if available) or similar existing models. The combination
of parametric geometry and real CAD geometry is also possible, ensuring that
design critical geometrical details are adequately captured (Figure 1). One
major advantage of the tool is that it is possible to position vents at any
surface in order to quickly test different ventilation strategies. A mesh
of tetrahedral surrounded by near wall prisms is created automatically by
the tool. In order to reduce the computational overhead, most of the tetrahedrons
are converted to hexahedrons within the meshing tool. Applying so-called "density
volumes" in jet regions to refine the mesh locally provides a high-quality
mesh for the simulation process. The quality of the mesh and the calculation
model have been confirmed in extensive validation studies where the CFD results
showed good agreement with experimental data [1].
Using this tool, the set-up of time for a typical cabin model has been reduced
to less than three days.
Solar and thermal radiation
In order to evaluate the thermal comfort of a proposed car interior, it is
necessary to predict the impact of both solar and thermal radiation on all
surfaces inside the cabin using STAR-CD’s discrete beam approach.
The computed solar incident radiation distribution for a real case is shown
in Fig. 2. This case was set up with a solar incidence of 45° altitude
and -45° azimuth. A direct solar radiation of 500 W/m2 was specified.
 |
Fig. 2: Incident solar radiation (top) and surface temperature distribution
(bottom) |
The perspective of the picture corresponds to the inclination of the sun.
The windows are represented by red edges. Only a few surfaces of the cabin
interior receive direct solar radiation. The A-pillar, the roof, and the
side door shadow a major part of the driver. Figure 2 shows the effect of
the solar incident radiation upon the surface temperature distribution.
Virtual thermal dummy
The prediction of the thermal comfort in the passenger compartment by means
of CFD is based on an experimental in-house method named MARCO (Method to
Assess theRmal COmfort).
MARCO is a thermal dummy equipped with 21 heated sensors that measure the
resultant surface temperature (RST) as a function of the air temperature,
velocity (convection) and radiation. The activity of the person is represented
by heat sources in the sensors. The RST has been experimentally correlated
with the human perception resulting in an LMV value (Local Mean Vote). From
these results, comfort zones for each of the 21 parts of the body have been
defined. [2].
In order to calculate the RST from the CFD model, a numerical thermal dummy
is included in the cabin model containing 21 temperature-measuring sensors.
Since the CFD solver takes into account the radiation and the convection
to calculate the surface temperature, the numerical RST is equivalent to
the one obtained experimentally and can therefore be calibrated with the
human perception LMV.
 |
Fig. 3: Distribution of incident solarradiation and of an example of
use |
Example of use
One major benefit of investigating thermal comfort by cabin flow simulation
is the possibility of comparing different ventilation concepts at an early
stage of the design without prototypes. In the following example, the influence
of the rear center and B-pillar ventilation on the rear passenger comfort
is assessed. Figure 3 shows the inclination of the solar radiation.
The airflow in the cabin with both ventilation concepts is illustrated in
Fig. 4 by streamlines. The streamlines marking the rear center and B-pillar
ventilation jets are colored in red. With the rear center and B-pillar ventilation,
the rear passengers are passed by more cool air. In the system without rear
center and B-pillar ventilation, the upper part of the body, in particular
breast and belly, is too warm (Figure 5). The very low LMV value of the right
hand of the passenger in the case of rear ventilation is caused by the direct
contact of the hand with the rear center vent jet.
 |
Fig. 4: Streamlines in a vehicle without (left) and with rear center
and B-pillar ventilation (right) |
 |
Fig. 5: Comfort values of the rear left passenger in a vehicle without
and with rear center and Bpillar
ventilation |
Conclusion
CFD cabin flow analysis is a valuable enhancement of the simulation tools
used in the product development process. The application of the parametric
cabin modeler enables a fast generation of CFD meshes of passenger compartments
allowing CFD analysis to be applied at the earliest phase of development.
References
[1] W. Kühnel, F. Guilbaud, C. Proksch, T. Heckenberger, D. Heinle,
CFD Cabin Flow Analysis as Part of
the Development Process, Vehicle Thermal Management Systems Conference Proceedings
(VTMS 6),
SAE Paper No. C599/053/2003, pp. 243-253.
[2] C. Bureau, H. Kampf, B. Taxis-Reischl, A. Traebert, E. Mayer, and R.
Schwab, MARCO – BEHRs
Method to assess thermal comfort, Vehicle Thermal Management Systems Conference
Proceedings (VTMS 6),
SAE Paper No. C599/005/2003, pp. 223-233.
|
Your route to this page:
