CIV E 632-COMPUTATIONAL HYDRAULICS AND HYDROLOGY
SPRING 2007
FINAL EXAM
WEDNESDAY, MAY 16, 2004, 1530-1730
Name: ______________________ S.S. No. ________________ Grade: _______
|
Instructions: Closed book, closed notes. Use engineering paper.
When you are finished, staple your work in sequence,
and return this sheet with your work. Answer all 11 questions (one bonus question).
|
-
Why does the diffusion wave model of catchment dynamics
provide grid independence for a wide range of grid resolutions, while the kinematic wave
model (the Li method) does not?
- Compare the distributed method of runoff generation (overland flow) with the lumped method (unit hydrograph)
in terms of their advantages and disadvantages, in theory and in practice.
-
What two kinds of errors arise in numerical modeling?
Explain which one is related to stability, and which one to convergence.
Why is numerical modeling a compromise between stability and convergence?
-
Why is a second-order numerical scheme not necessarily better than a first-order scheme?
What artifice is usually resorted to in order to provide stability at the expense of convergence?
- Why is it often necessary to artificially extend the channel reach in dynamic wave modeling?
What would happen if a unique rating curve (a kinematic boundary condition) is specified at the downstream boundary? Why?
-
Why does the Muskingum-Cunge method work best at Courant number equal to 1? What happens for Courant greater than 1?
For Courant less than 1?
- Compare the tradeoffs between explicit and implicit schemes. Which one is better?
Why?
- Do sedimentation bed transients (channel bottom perturbations) in alluvial rivers travel upstream or downstream?
Explain.
-
Why do dam-breach flood wave peaks eventually reach the same value at a certain distance downstream?
-
What two types of boundary conditions arise in two-dimensional groundwater modeling?
Explain.
- Should two-dimensional flow to a side river embayment convect (go straight into) or circulate (turn around)? Explain.
|