| |
In the past, the only way for manufacturers to evaluate a proposed automotive
HVAC system design was to build an expensive prototype and test it in
the laboratory. Several years ago, engineers at Visteon began experimenting
with CFD with the goal of evaluating automotive air handling system designs
without the need to build a prototype. The complex geometries used in
air handling systems present major obstacles in computer simulation. They
typically include ducts that expand and contract, change from round to
square cross-sections, go through complex curves throughout their length,
and have many branches and internal walls.
The CFD codes used in the past at Visteon involved a structured mesh
approach that required that the block structure be defined by hand before
a volume mesh was generated. It took weeks or months to produce a grid
for analysis of complex air handling system components using this approach
- as long as building a prototype. Recently, Visteon began using TGrid
3 to generate an unstructured tetrahedral volume mesh from a triangular
surface mesh. While tetrahedral meshes are useful for complex geometries,
prism elements are more suitable for resolving boundary layers. The prism
elements, extruded from surface triangles, allow a more accurate solution
with fewer elements. TGrid creates a hybrid mesh consisting of prism layers
in near-wall regions and tetrahedral cells in the remainder of the domain.
It also layers prism elements at the walls. The result is better accuracy
without the time-consuming task of building an all-hex mesh.
Evaluation of windshield defroster/demister performance.
Flow distribution in an automotive air handling system.
By switching to FLUENT and taking advantage of parallel processing,
Visteon made dramatic improvements in solver technology. The latest FLUENT
release partitions the flow domain for multiple processors and runs an
instance of the solver on each processor. Visteon uses both a 16-node,
shared-memory Silicon Graphics compute server and clusters of workstations
for parallel processing of CFD models. On the SGI compute server, the
FLUENT solver achieves about 90% processing efficiency – 14.4 times the
speed of a single process. This makes it possible to solve many typical
models overnight.
As an example of the design improvements that can be achieved by applying
these new capabilities, consider the design specification that performance
of the HVAC system be linear with respect to the temperature dial on the
instrument panel. In other words, moving the dial from position one to
position two should have the same impact on temperature as moving from
position two to position three. In the past, the linearity of the temperature
dial couldn’t even be estimated until full-vehicle prototypes were constructed.
At that point, changes were costly and the testing data provided little
or no input on what types of changes were required. Using FLUENT, Visteon
engineers can determine the linearity of a proposed design as soon as
the solid model has been created – in a matter of days. The flow patterns
revealed in the analysis help them to quickly iterate to an optimized
design.
|
|
|