CFD Research Overview Dr. Chris Roy Associate Professor Aerospace & Ocean Engineering Virginia Tech
2
Outline
• My background
• Research interests
• Research projects
• Shape optimization for tractor trailer drag reduction
• CFD of gas flow through microfibrous materials
• Potential research projects for Fall 2008
• Undergraduate teaching plans
• Graduate teaching plans
3
My Background
• BSE Mechanical Engineering, Duke Univ. 1992
• MS Aerospace Engineering, Texas A&M 1994
• PhD Aerospace Eng., North Carolina State Univ. 1998
• 1998-2003: Aerosciences Department, Sandia
National Labs (NM)
• 2003-2007: Assistant Professor, Aerospace
Engineering Department, Auburn University
• 2007-present: Associate Professor, Aerospace and
Ocean Engineering, Virginia Tech
4
Research Interests
• Computational Fluid Dynamics (CFD)
• Turbulence modeling
• Turbulence simulations
• Bluff-body aerodynamics (e.g., trucks)
• Microscale gas flows
• Hypersonic flows
• Verification (mathematical correctness) and Validation
(physical correctness) of computer simulations, or V&V
• Estimation of grid-related numerical errors
• Grid adaptation
• Validation of turbulence models
5
Truck Aerodynamics: Motivation
There are more than 2 million tractor-trailers on U.S. roads
• In 2003 each tractor-trailer averaged 62,900 miles traveled†
• On average they got 5.2 miles/gallon, resulting in a total
consumption of 26 billion gallons of diesel fuel†
• With diesel at $2.33/gallon, this corresponds to ~$60 billion
Aerodynamic drag increases with the square of the speed
†US DOE Transportation Energy Data Book: Edition 23, 2003,
http://www-cta.ornl.gov/data/
• 20% drag reduction gives ~10%
reduction in fuel consumption
• Savings of 2.6 billion gallons of
fuel/year, or ~$6 billion/year
6
Truck Aerodynamics
DES: Large-Scale Turbulent
Wake Structures Resolved
(fundamental physics:
expensive)
RANS: All Turbulent
Structures Modeled
(appropriate for design)
Two approaches to turbulence:
8
Optimization of Aerodynamic Drag
Currently developing a multi-university effort to optimize
base drag reduction devices for heavy trucks
• Advances in computing infrastructure now allow CFD to be used
within the design optimization process
• VT AOE is responsible for developing high-efficiency CFD
predictions (Roy) and conducting wind tunnel experiments
(Devenport)
• Auburn University is responsible for the Genetic Algorithm (GA)-
based optimization procedure and full-scale vehicle testing
• Currently gathering industry support and developing a 4 year,
$2.5M proposal to DOE
9
Drag Reduction via Base Flaps
Prior experiments on boattail flaps reduced drag by 18%;
however, these devices were not optimized to minimize drag
Tractor-Trailer with Base Flaps Wind-Averaged Drag Coefficient
12
H2 Reformation for PEM Fuel Cells
Proton Exchange Membrane (PEM) fuel cells are popular for transportation applications
• Protons (H+ ions) are exchanged across the electrically-insulated membrane
• Electrons are forced to travel through the electric circuit, generating electric current
• Clean technology since gaseous hydrogen and air are combined to produce water & heat
• Anode catalyst is irreversibly poisoned by ppb levels of sulfur and ppm levels of CO
• The pollutants can arise when jet fuel (JP-8), diesel, or hydrocarbons are reformed to H2
(courtesy Wikipedia)
13
Microfibrous MaterialsInterdisciplinary work with Auburn
University's Chemical Engineering Dept.
(funded by US Army TACOM)
• New structure of matter developed by
embedding catalytic particles (~100 microns) in
a matrix of microfibers (~10 microns)
• Microfibrous materials have demonstrated
remarkable properties, increasing reactivity by
nearly a factor of five
• Our goal: use CFD to identify the fundamental
mechanisms behind increased reactivity and to
help design new microfibrous materials
SEM of a Typical
Microfibrous Material
Applications include chemical processing (desulphurization of fuel
cells, ozone removal, etc.) and filtration (gas mask CO filters, etc.)
14
Geometric Modeling
Microfibrous materials are
difficult to characterize
• Initial work employs geometric
simplifications, thus allowing us to
make use of symmetry
• Particles modeled as smooth
spheres
• Fibers modeled as straight cylinders
with axes aligned in the y and z
directions (normal to the flow)
• Fluent commercial CFD code
employed in this work
Face Velocity [m/s]
Lo
gR
ed
uctio
n
0.02 0.04 0.06 0.08 0.1
10
20
30
40 Packed Bed - Simulation
Packed Bed - Experimental
Glass Fiber - Simulation
Glass Fiber - Experimental
15
Pressure Drop & Reactivity Predictions
H2S removal shows good agreement
with experiments for packed beds
and microfibrous materials
Face Velocity [m/s]
Pre
ssu
reG
rad
ien
t-
P/L
[mm
H2O
/mm
]
0.4 0.8 1.2
101
102
CFD = 0.41
Experiment = 0.41
CFD = 0.47
Experiment = 0.47
0.2
Pressure drop predictions within
15% of experiments
16
Potential Fall 2008 Research Projects
• Aerodynamic Drag Optimization for Tractor-Trailers, US
Department of Energy (DOE)
• CFD Analysis of Flow through Microfibrous Materials, US
Army Tank and Automotive Command (TACOM)
• Estimation of Grid Errors and Mesh Adaptation for CFD, NASA
Marshall Space Flight Center (Huntsville, AL)
• Verification and Validation for Power Electronics Systems,
Boeing (collaborative w/ CPES at VT)
• Verification, Validation, and Uncertainty Quantification for
Hypersonic Flows, NASA Johnson Space Center (Houston, TX)
17
Undergraduate Teaching
• Core aero/hydro courses
• Introduction to CFD (senior elective)
• Integrate CFD into core undergraduate courses
• Involve undergraduates in research
• One-on-one mentoring
• Mentoring by graduate students
18
Graduate Teaching
• New course next Spring: Verification and
Validation in Computational Simulation• Co-instructed by Dr. William Oberkampf (retiring from
Sandia Labs)
• Writing a graduate textbook by the same name
• Advanced CFD course next year
• Course on turbulence modeling (future?)
• Plan to hire 1-2 postdoctoral fellows and 5-6
doctoral students over the next 2 years
Web Site:http://www.aoe.vt.edu/~cjroy/
Questions?