Add run/clean scripts for SU2-CalculiX and restructure files

[MakisH]

The FSI/flap_perp/SU2-CalculiX was one of the first tutorial cases we got, as part of the BGCE project that resulted into http://run.precice.org/. Since then, we have many more cases that follow a few patterns, which this case currently ignores.

This PR:

• Adds a runFluid and a runSolid script, which start the two participants separately and use both the screen and Fluid.log/Solid.log for logging (using tee, which we could also apply into more tutorials). These files were previously called Su2.log and Calculix.log.
• Adds an Allrun script, similar to the OpenFOAM-CalculiX one.
• Adds an Allclean script.
• Changes the configuration of SU2 to write result files only into the Fluid/ directory (and not the root directory).
• Changes the exchange-directory from .. (flap_perp/) to the default (., i.e. SU2-CalculiX).
• Removes the duplicate configuration file euler_config_coupled_mergeSolution.cfg which was used in case of parallel simulations. This only had trivial changes, irrelevant to the parallelization (UNST_TIME, SOLUTION_FLOW_FILENAME)

Todo

• Remove the runTutorial_serial.sh and runTutorial_parallel.sh.
• Update the README.md.
• @Eder-K this will require changes in the system tests scripts, but probably not in the reference results.
• Update the wiki.

I also tried a few options to reduce the verbosity of the screen output in SU2 (e.g. not print the residuals for all the iterations), but I could not succeed.

Closes #89.

[MakisH]

This is now ready for review.

I have tested the intended functionality (starting/cleaning) and it works.

I cannot test the complete simulation, as the Solid simulation diverges in my system with ERROR: Too many cutbacks, but this is the case also with the state before this PR, as well as with preCICE v2.1 and v2.0.2. However, I don't think we need to run a complete simulation for this PR.

[MakisH]

@Eder-K please try to apply these changes to the system tests and approve when ready.

[uekerman]

Looks good.

[Eder-K]

The systests are updated, the test is now launched by executing the runFluid/runSolid scripts in the respective adapters.

However, the test does still fail during finalization due to a known double free error in the SU2 adapter: https://travis-ci.org/github/precice/systemtests/jobs/722485626#L30853
This is not exclusive to this new test structure though, as the current cron jobs on systemtest regularly fail with the same problem: https://travis-ci.org/github/precice/systemtests/jobs/722480420#L30824

[MakisH]

I also fixed the watchpoint name in the plotDisplacement.sh (somehow I had not yet committed this change).