Case Studies
West Lafayette, IN - February 2002
Gaseous fuel mixing with air.
David Glase from Purdue University develops models for simulating turbulent
combustion. Improved accuracy for this type of simulation could potentially
benefit internal combustion engines in cars, gas turbine engines in airplanes,
power plants, and other turbulent combustion processes.
His goal is to improve efficiency while reducing pollution. The best way
to solve problems of this type is with advanced computer modeling of combustion
processes.
His simulation code runs on an IBM SP supercomputer. The code runs in
parallel on 16 processors, using a total of about 6GB of RAM and about
10,000 hours of CPU time. Visualising the mountains of data from simulations
like this once posed a huge challenge.
2-D slice displaying mixture composition of gaseous
fuel mixing with air.
Today Tecplot can smoothly tackle data of this magnitude.
Version 9.0 enables quick rotation, sooming and translation of 2-D and
3-D data sets that consist of millions of points. New Plot Approximation
Mode options vastly reduce rendering time, and increase responsiveness
for viewing even the largest data sets.
By exploring, slicing and adjusting plot attributes, David's data yields
a better understanding of underlying physics and insight into the simulation.
Axial and transverse slices through the jet can be
useful for spotting problems in the simulation. For example, implicit
numerical dispersion errors show up as jagged edges on contours that should
be smooth.
Iso-surfaces are a great way to quickly assess 3-D
data. The level of turbulent mixing can be seen in the plot at the top
of this page, large vortical turbulent structures can be isolated with
plots of pressure (above), and the small-scale turbulent structure can
be easily seen with a vorticity plot (below).
Streamtraces illustrate the nature of the vector field
flow within the volume. Transparent iso-surfaces enable us to view the
shape as well as internal phenomena.
A 2-D slice along the axis of a 3-D simulation of a
fully-turbulent jet flow . The grayscale contours represent mixture fraction.
This can be thought of as a smoke tracer added to the flow. The lower
half of the plot shows velocity vectors, colored and scaled by their relative
magnitude. The duration of the movie represents about 70ms of time in
a 26mm gas jet operating at a Reynolds number of 21,000.
Download
the AVI movie (11,668 KB).
David presents most of his time-averaged data using
XY-plots. His current research focuses on duplicating experimental data
that is measured from actual jets. For comparison, he plots the axial
velocity decay along the centerline of the jet, as well as radial velocity
profiles at several axial locations. These, along with similar plots of
turbulence intensity and Reynolds stresses allow him to judge the effectiveness
of the turbulence models and numerical method chosen for his code.
XY-plots enable comparison of simulation data with
experimental data.
A 2-D axial slice with a single iso-surface through
time-averaged data.
Transverse slices with a transparent iso-surface through
time-averaged data
These images display a variety of exploration and presentation
options Tecplot offers with its unique combination of technical plotting
and powerful 3-D visualisation. Chris Wingard from Oregon State University
summed it up best when he said "With Tecplot, the only limitation to what
I can do is my imagination." David Glase noted "I can do anything I dream
up."
|
RSS Feeds
System Requirements
