Transcript Adam Koenig Final Presentation

Adam Koenig, Wichita State University
Mentors:
Dr. Ron Riggs, University of Hawai’i, Manoa
Dr. Sungsu Lee, Chungbuk National University
Krystian Paczkowski, University of Hawai’i, Manoa
HARP REU Program, August 3, 2011
Overview
 Background and Motivation
 Benefits of CFD Approach
 Description of Utilized CFD Tools
 Simulation Description
 Results
 Conclusions and Recommendations
Background and Motivation
 Experiments have collected data on tsunami bore
formation1,2
 Tsunami blockage and funneling is less studied
 Such data would be useful for establishing design
parameters for structures in and around streets that
would be affected by this channeled flow, especially if
flow is accelerated
Goals
 Numerically simulate tsunami bore channeled
through a city street
 Identify effects and phenomena caused by buildings
obstructing bore flow
 Quantify relationships between tsunami bore
parameters and flow properties in the street
Benefits of CFD approach
 Much more inexpensive than experimental tests
 No scaling problems
 Greater flexibility in test parameters
Software/Models
 This study uses OpenFOAM v 1.7.1, a free, open-source CFD
software package for a wide range of fluid problems
 Solver: InterFoam, a solver for two incompressible,
immiscible fluids that uses a VOF method to generate a
volume where the sharp interface between phases would
exist
 Turbulence: k-ε model, a RANS based turbulence model
with transport equations for turbulent kinetic energy and
turbulent dissipation
Hardware
 JAWS system at Hawaii Open Supercomputing Center
 320 Dell PowerEdge 1955 blades with four 3.0 GHz
processors per blade
 Cisco SDR infiniband (10Gbit/sec) interconnect
Domain Description
 The domain of this test consists of a 120×290×30 ft
rectangular prism with two half-buildings obstructing
the end
 The half-buildings are each 45 feet wide and 90 feet
long
Domain Description
 The inlet consists of a 3.6 ft high patch spanning the
back wall of the domain
 The inlet speed was controlled by setting a constant
velocity condition across the surface of the inlet. Tests
showed that there was no difference in the channel
flow of a total pressure inlet was used.
 The inlet speed was adjust to give the desired Froude
number of the bore. The study focused on bore
Froude numbers between 2 and 3 from experimental
data1,2.
The Mesh
 The mesh consists of two groups of hexahedral cells
stacked in the domain
 Mesh density is 1.25 ft/cell in horizontal directions
 Vertical density is 0.9 ft/cell up to the height of the
inlet, and 1.65 ft/cell from top of the inlet to the top of
the domain
 This meshing allows for acceptable resolution
throughout the domain with improved resolution in
majority of flow area
The Mesh
Limitations/Difficulties
 Zero velocity boundary condition in wall above inlet
 Data is only valid until reflected bore strikes back wall

Reason for long domain
 Open boundary resulted in flow anomalies and crashed
simulations
 Gap Aspect Ratio
 Dimensions chosen to fit regular two-way street and
building size based on Empire State Building
 Actual aspect ratios would vary considerably
Simulation Example
Observations
 Water pools in front of obstructing buildings at height
significantly greater than bore height
 Original bore reflected back out to sea as a hydraulic
jump
 Remaining water cascades between buildings into the
street
Results
Pool Height vs. Bore Froude Number
4
3.5
Pool Height (m)
3
2.5
2
1.5
1
0.5
0
0
0.5
1
1.5
2
Bore Froude Number
2.5
3
3.5
4
Results
Outlet Height vs. Bore Froude Number
1.6
1.4
Outlet Height (m)
1.2
1
0.8
0.6
0.4
0.2
0
0
0.5
1
1.5
2
Bore Froude Number
2.5
3
3.5
4
Results
Pool to Outlet Height Ratio vs. Bore Froude Number
3
Pool to Outlet Height Ratio
2.5
2
1.5
1
0.5
0
0
0.5
1
1.5
2
Bore Froude Number
2.5
3
3.5
4
Results
Outlet Velocity vs. Bore Froude Number
9
8
Outlet Velocity (m/s)
7
6
5
Outlet Velocity
Bore Velocity
4
3
2
1
0
0
0.5
1
1.5
2
Bore Froude Number
2.5
3
3.5
4
Results
Reflected Froude Number vs. Bore Froude Number
0.7
Reflected Froude Number
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0.5
1
1.5
2
Bore Froude Number
2.5
3
3.5
4
Conclusions
 Pooling height, outlet height, and outlet velocity all
positively correlated to bore Froude number.
 Height ratio independent of bore Froude number
 Outlet velocity never exceeds bore velocity, but
numbers are very close at low Froude numbers
 Possibly a consequence of inlet height and sheet flow in
bore
 Reflected bore relatively constant for tested range, but
possible negative correlation
Recommendations for Future Work
 Study effect of gap aspect ratio on flow property
relationships
 Determine whether inlet height affects funneling
behavior and whether inlet height affects bore shape
Acknowledgements
 Krystian Paczkowski, for his insight into the inner
workings of OpenFOAM software
 Dr. Susan Brown, for continuous assistance with data
storage issues
 Dr. Ron Riggs and Dr. Sungsu Lee, for their guidance and
insight into fluid behavior problems
This material is based upon work supported by the National
Science Foundation under Grant No. 0852082. Any opinions,
findings, and conclusions or recommendations expressed in
this material are those of the author(s) and do not necessarily
reflect the views of the National Science Foundation.
Works Cited
 1Robertson, I. N., H. R. Riggs, and A. Mohamed.
"Experimental Results of Tsunami Bore Forces on
Structures." Proceedings of the 27th International
Conference on Offshore Mechanics and Arctic
Engineering. Estoril, Portugal. Print.
 2Robertson, I. N., H. R. Riggs, K. Paczkowski, and A.
Mohamed. "Tsunami Bore Forces On Walls."
Proceedings of the ASME 2011 30th International
Conference on Ocean, Offshore, and Arctic
Engineering. Rotterdam, The Netherlands. Print.
Questions?