Features/simulation models

[jnnr]

This introduces a numpy-based model for solving stationary hydraulic and thermal problems. For the beginning, only networks without loops will be considered.

TODO

• [x] Define all material constants in a meaningful place.
• [x] Add a warning that hydrostatic pressure differences due to height difference are not implemented yet.
• [x] Add a warning that looped networks cannot be simulated yet.
• [x] Allow to pass a timeseries for temp_env
• [x] Use zeta values from input data
• [x] check _calculate_pipes_heat_losses - the argument is not used.
• [x] Make sure that docstrings contain the underlying formula
• [x] Fix failing test_setup_simulation
• [x] Document units of results
• [x] Fix pylint and pytest errors
• Checking
  • [x] Compare to benchmark results of the simple tree example
  • [x] Check if flow direction defined by mass flows is used in all (thermal) calculations, not the defined pipe direction, as the mass flow can go opposite to the defined pipe direction. Update: Opened issue #49.

[jnnr]

Suggestion: temperature_drop -> temp_drop update: This is a general issue. Opened #46.

[jnnr]

@joroeder: We are approaching a point were all results have been checked and have been found equal to the benchmark data. If you find the time, you are invited to have a look at the simulation example and give a review. Still missing: The localized pressure losses, global pressure losses and pump power. The distributed pressure losses in the pipes are already checked.

[joroeder]

Of course, the travis issues need to be addressed ...