-
-
Notifications
You must be signed in to change notification settings - Fork 13
Future Design Refinements
Many design parameters will ended up being rounded for construction purposes. For example, the length of a plate settler should probably be rounded to the nearest centimeter. In order to achieve this values should be rounded or tolerances should be used in order to ensure that construction will work correctly with the design code's outputs.
Different countries have access to different materials. For example, in the Mathcad code there is a country pipe database which specifies which pipe sizes are available in Honduras and which are available in the United States. In the current version of the Python code there is only a single universal pipe database.
Update the docstrings for every function to comply with the numpy standard. Ensure that doctests are added to the docstrings as well.
The former "expert input" values can be used in a optional inputs drop down in GUI for customization or so people can play with different plant parameters and see how the plant design changes.
Remove unit wrappers and any unnecessary .magnitude conversions within the function calculation. Add .to to every function return.
Remove anything that is not a universal constant. Universal constants include gravity, vena contracta coefficients, K minor values, and other things that are physically dominated. All other things should be in a YAML dictionary.
Convert the dictionaries to classes so that each unit process has the attributes of the dictionary and the methods which are currently functions in the Python script.
The length of the entrance tank is defined in optional inputs as the maximum length possible (the length needed for a 60 L/s plant). An improved design would use the plant flow rate and other hydraulic parameters to determine a variable rather than constant length for the entrance tank.
The head loss of the LFOM is currently an optional input. In the old design engine the head loss was calculated based on plant flow rate, so recreating this calculation would improve the design over the constant of 20 cm that is currently used.
Currently this function uses an approximation in order to avoid integration, but using the SymPy package the more complete method used in Mathcad which employs integration could be implemented.
The current AutoCAD drawings do not reflect the current design for the plate settler support structure. The current system uses a 3-inch longitudinal structure supported by 2-inch cross supports. See this photo from Honduras of the Fall 2017 most updated version. Future design code and template work should reflect this change.
Our current failure mode to determine head loss through the backwash layer depends on an assumed angle of repose for wet sand of 45 degrees. The actual angle is 15 degrees, so it would be useful to have the ability to change this angle in the design process. That should create smaller trunks and larger orifices (both of which would make fabrication easier).