Compare several methods for cylinder2D problem

[wacyyang]

In this PR, we are going to collect and compare several methods to solve NSE for the benchmark problem Test 2D-2. The first step is to use exact same setting, compute drag/lift force and compare results with other literature. Those methods include:

• conforming mesh with rans2p method
• conforming mesh with rans3p method
• immerse boundary method
• shift boundary method

[codecov]

Codecov Report

Merging #683 into master will increase coverage by 3.23%.
The diff coverage is 83.88%.

@@            Coverage Diff             @@
##           master     #683      +/-   ##
==========================================
+ Coverage   37.64%   40.87%   +3.23%     
==========================================
  Files         207      264      +57     
  Lines       73703    77424    +3721     
==========================================
+ Hits        27745    31647    +3902     
+ Misses      45958    45777     -181

Impacted Files	Coverage Δ
proteus/mprans/MCorr3P.py	0% <ø> (ø)	⬆️
proteus/mprans/RANS3PF.py	77.83% <ø> (+77.83%)	⬆️
proteus/mprans/RANS2P.py	77.78% <ø> (+2.81%)	⬆️
...s/tests/cylinder2D/conforming_rans3p/save_batch.py	0% <0%> (ø)
proteus/tests/ProjScheme_with_EV/NS_convergence.py	0% <0%> (ø)
proteus/mprans/VOF3P.py	0% <0%> (ø)	⬆️
proteus/tests/cylinder2D/ibm_method/cylinder_so.py	69.04% <100%> (ø)
...ylinder2D/conforming_rans3p/pressureincrement_p.py	100% <100%> (ø)
.../cylinder2D/conforming_rans3p/pressureInitial_p.py	100% <100%> (ø)
proteus/mprans/PresInc.py	77.32% <100%> (+77.32%)	⬆️
... and 101 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9e4add7...8f722ae. Read the comment docs.

[wacyyang]

@cekees I am not sure about this line and line36. Are they correct?

[cekees]

@wacyyang, could you look over these changes and make sure we don't need to keep any of them before we delete this branch? I went ahead and closed the PR.

[wacyyang]

@cekees I just removed some debug lines, tested rans2p and rans3p methods again and got right drag/lift coefficients. So I think it is safe to merge it into master branch now.

[cekees]

Couple of ideas:

• What about putting one main README file in the cylinder2D directory that describes the problem and just short ones explaining the variations in the subdirectories?
• I'll push up the bit of extra code to add so that these will run on travis (see the changes to setup.py for conforming_rans3p test). You'll need to do the analagous thing for the other test directories.
• Remove the TODO's from the README and make them github issues or put them in a python docstring in the code with the sphinx .. todo: directive http://www.sphinx-doc.org/en/stable/ext/todo.html
• Spell and grammar check your README files.

[cekees]

The tests aren't running for me locally. You should see them run when you do make test in the proteus directory. @manuel-quezada has some experience with these kinds of issues.

[cekees]

Looks like a path issue in setup.py for the new data files.
https://travis-ci.org/erdc/proteus/builds/342120672#L3954