Turboflow

- Centralize testing structure. To manage which configuratoin files that should be included in the different tests, and for which configuration files that there should be made new regression data for is now managed through
- `tests_manager.py`

- Critcial exit flow angle is now calculated with deviation model
- Add beta_crit_out as variable and another residual equation
- Gauging angle is the arccos(area_throat/area_exit)
- Aungier deviation model interpolated between mach = 0.5 and mach_crit_throat should be used

- Added configuration file validation functionality
- Added validation schema to documentation (updates automatically)
- Updated documentation for configuration validation
- Updated the regression tests YAML file with the new configuration options

- Refractor the "utilities" import in `cascade_series.py` and `performance_analsys.py`
- Add documentation and validation for the keys expected in the results dictionary
- Moved `compute_overall_performance()` calculations to a dedicated function
- Refractored and renamed `compute_stage_parameters()` to `compute_stage_performance()`.
- Added a new generic function: `validate_keys()` that should be used for all validations of dictionary keys.
- Added the new `validate_keys()` generic function to `geometry.py`
- Simplified the argument list of `compute_efficiency_breakdown()`
- Renamed `cascade_series.py` to `flow_model.py` and replace the abbreviation "cs" by "flow"

- Add additional problem evaluation after root finding completion (this ensures the solver.problem object has the converged solution)
- Add elapsed time to root finding solution report
- Add function `print_simulation_summary()` to print report after running performance analysis.

- Changed the storage for plane and cascade data. The KEYS used to initialize was no longer needed as the dataframe could be initialized as an empty dataframe. What properties that are stored is now controlled by the output of the evalauete plane functions and choked model functions. This makes it more easier to control and more flexible imo.
- in fluid_properties, more specifically FluidState class, I commented out "fluid_name", "converged" and "identifier". I had troubles storing these values in the pandas dataframe, and they dont seem to be stored anywhere else either way.
- in loss_model, I removed the 'loss_coefficient' from the loss_dict. I had troubles storing these values in the pandas dataframe, and the value is already specified in the config file.
- Changed method for giving constraints to the design_optimization code. Now we give one list of constraints, where the type is specified within each constraint.