ENH: Individual Fins

[MateusStano]

Checklist

• Tests for the changes have been added (if needed)

Description

Individual Fins model have been added:

• Usage can be seen in the documentation files. All equations and modelling are also described in the docs file.

• A lot of structural changes were made to all fin classes. Please review this new structure with care.

• If you run a simulation with the standard Fins model and another with individual fins, the results will be a little different. This happens because the individual fin model calculates forces for each fin separately, using the airflow at each fin’s exact position. The standard model (Barrowman) treats all the fins as one and uses a single point along the rocket’s center. In theory, this means that individual fins are more accurate then the standard model

• There are a couple of TODOs relating to cant angle control. The center of pressure position relative to CDM is precalculated in the Rocket class and then accessed in the Flight. This causes a problem if the fin’s cant angle changes during the simulation because that changes the center of pressure position, but right now there’s no easy way to update these values in the Rocket while the simulation is running.

Breaking change

• Yes
• No

[Copilot]

Pull Request Overview

This PR introduces an Individual Fins model to improve the accuracy of fin force calculations by computing forces per fin rather than using a single aggregated point. Key changes include:

• Adding new fin classes and updating fin imports in multiple modules.
• Enhancements for printing and plotting fin data, including new fin print and plot classes.
• Updates to documentation reflecting the new individual fin model.

Reviewed Changes

Copilot reviewed 40 out of 40 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
rocketpy/rocket/aero_surface/fins/init.py	Added new fin exports (Fin, EllipticalFin, FreeFormFin, and TrapezoidalFin)
rocketpy/rocket/aero_surface/aero_surface.py	Introduced a rotation matrix for surface-to-body conversions and minor cleanup in force computations
rocketpy/rocket/init.py	Updated imports to include new fin classes
rocketpy/prints/aero_surface_prints.py	Added new printing methods/classes for individual fin types
rocketpy/plots/rocket_plots.py	Updated aerodynamic surface validation and added code for drawing individual fin geometry
rocketpy/plots/aero_surface_plots.py	Added multiple new plotting classes and methods to visualize individual fin shapes and parameters
Documentation files	Updated indices and references to include the new individual fin model

Comments suppressed due to low confidence (1)

rocketpy/plots/rocket_plots.py:361

• [nitpick] Consider vectorizing the fin rotation operation rather than iterating over points to improve performance.

for i, p in enumerate(points): points[i] = surface._rotation_fin_to_body @ Vector(p)

[codecov]

Codecov Report

Attention: Patch coverage is 61.25166% with 291 lines in your changes missing coverage. Please review.

Project coverage is 78.53%. Comparing base (4df0b38) to head (45cebc4).
Report is 22 commits behind head on develop.

Files with missing lines	Patch %	Lines
rocketpy/plots/aero_surface_plots.py	29.47%	67 Missing ⚠️
rocketpy/rocket/aero_surface/fins/fin.py	31.25%	66 Missing ⚠️
rocketpy/prints/aero_surface_prints.py	33.33%	28 Missing ⚠️
rocketpy/rocket/aero_surface/fins/_base_fin.py	73.07%	28 Missing ⚠️
rocketpy/plots/rocket_plots.py	38.63%	27 Missing ⚠️
...ocketpy/rocket/aero_surface/fins/elliptical_fin.py	37.03%	17 Missing ⚠️
rocketpy/rocket/aero_surface/fins/free_form_fin.py	39.13%	14 Missing ⚠️
...ketpy/rocket/aero_surface/fins/_free_form_mixin.py	90.08%	12 Missing ⚠️
...etpy/rocket/aero_surface/fins/_elliptical_mixin.py	73.80%	11 Missing ⚠️
...cketpy/rocket/aero_surface/fins/trapezoidal_fin.py	45.00%	11 Missing ⚠️
... and 2 more

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #818      +/-   ##
===========================================
- Coverage    79.11%   78.53%   -0.59%     
===========================================
  Files           96      106      +10     
  Lines        11575    12445     +870     
===========================================
+ Hits          9158     9774     +616     
- Misses        2417     2671     +254     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Gui-FernandesBR]

Suggested change

	try:
	print(f"Tip chord: {self.aero_surface.tip_chord:.3f} m")
	except AttributeError:
	pass # it isn't a trapezoidal fin, just don't worry about tip chord
	if isinstance(self.aero_surface, What you want):
	print(f"Tip chord: {self.aero_surface.tip_chord:.3f} m")

[phmbressan]

I agree with the suggestion. hasattr could also work.

[Gui-FernandesBR]

Suggested change

	name="Fins",
	name="Trapezoidal Fin",

[Gui-FernandesBR]

Suggested change

	"""Calculates and returns the center of pressure of the fin set in local
	"""Calculates and returns the center of pressure of the fin in local

[Gui-FernandesBR]

Suggested change

	Name of fin set.
	The trapezoidal fin's name.

[Gui-FernandesBR]

TrapezoidalFin.d

vs

TrapezoidalFin.diameter

[Gui-FernandesBR]

We have only one fin in the set, no?

Suggested change

	Aspect ratio of each fin in the set
	Aspect ratio of each fin in the set

[Gui-FernandesBR]

isnt the evaluate_center_of_pressure common for both TrapFin and TrapFins?

[Gui-FernandesBR]

This way you skip the "skip line" entry

Suggested change

	1. By use of the :class:`rocketpy.GenericSurface` or :class:`rocketpy.LinearGenericSurface` classes,
	1. By use of the :class:`rocketpy.GenericSurface` or :class:`rocketpy.LinearGenericSurface` classes, \

[Gui-FernandesBR]

do you really need the return?

[Gui-FernandesBR]

maybe what you want to do is to create a setter method to the cant_angle parameter?

[Gui-FernandesBR]

Suggested change

	surfaces : list, AeroSurface, NoseCone, TrapezoidalFins, EllipticalFins, Tail, RailButtons
	surfaces : list[AeroSurface], AeroSurface

[Gui-FernandesBR]

(minor) these checks could be moved to an another method

[Gui-FernandesBR]

As both Fins and TrapezoidalFinsMixin have the to_dict method, which method will the TrapezoidalFins class use?

[Gui-FernandesBR]

Alternatively, one could define the to_dict here as an abstractmethod, so the children classes must define the method.

[Gui-FernandesBR]

Should be faster (copilot suggestion), and by far more readable:

        # Replace infinities and handle invalid values in chord_lead and chord_trail
        mask_invalid = (
            np.isinf(chord_lead) | np.isinf(chord_trail) | np.isnan(chord_lead) | np.isnan(chord_trail)
        )
        chord_lead[mask_invalid] = 0
        chord_trail[mask_invalid] = 0

        chord_length = np.where((chord_length < 0) | np.isnan(chord_length), 0, chord_length)
        max_chord = chord_trail - chord_lead
        chord_length = np.minimum(chord_length, max_chord)

[Gui-FernandesBR]

The evaluate_geometrical_parameters method is too long.
If possible, please consider refactoring it to split the same code into smaller pieces.

[Gui-FernandesBR]

separate method

[Gui-FernandesBR]

0.288?

Where does it come from?
Any reference?

[Gui-FernandesBR]

nice!

But who is super here? the mixin or the fin?

[phmbressan]

A docstring would help here. This method is useful for both Fin and the current Fins classes, right? Otherwise, it should be only in the child.

[phmbressan]

Why using complex_step here, is performance that crucial? I think it is a bit less stable in the Function class, since it was mainly tested only for lambda Functions. I have no idea if it works properly in the case of spline or polynomial interpolations.

[phmbressan]

I agree with the suggestion. hasattr could also work.

[phmbressan]

Do we need these default child classes? I understand their value for scalability (we can add things to them if needed), but couldn't we use the base _FinPrints?

[phmbressan]

I understand the use of mixins so that you provide custom geometrical methods for each fin type. My question is why not simply use a bit of composition here (e.g. more like a strategy pattern): have a geometry attribute that handles all the specifics of geometry (a class _FinGeometry that has essentially the same methods as the mixin).

My question arises, since using mixins here is alright from a functional standpoint (it works), but it is a bit unconventional given that it is more used to provide more I/O related methods or "secondary" behavior to a class in general.

My brainstorming is something like:

class _BaseFin(ABC):
     def __init__(self, geometry, any common non geometric attributes (if any)):
        self.geometry = geometry
        ...
    # Ideally most of the code here should be concrete
    # and delegate things to the geometry (e.g. `return self.geometry.lift_coefficients`)

class TrapezoidalFins(_BaseFin):
     def __init__(self, same init as today):
         geometry = _TrapezoidalSetGeometry(geometry args)
         super().__init__(geometry)

class _FinGeometry(ABC):
    # Basically all abstractmethods to define structure

class _FinSetGeometry(_FinGeometry):
    # Any common specific logic of finsets
    # if no big specific logic, this class is not needed

class _TrapezoidalSetGeometry(_FinSetGeometry):
   ...

class _IndividualFinGeometry(_FinGeometry):
    # Any common specific logic of finsets (e.g. frame rotations)

class _TrapezoidalFinGeometry(_IndividualFinGeometry):
   ...

This is simply a brainstorm of mine I just had while reading the code. I believe this captures semantically a bit more the structure than mixins, while keeping flexibility.

On the many setters that update the _BaseFin geometry (e.g. change on span), I would simply create a new geometry and set the self.geometry. If performance is a concert, then make setters inside the geometry that are called.

[phmbressan]

Good change.

[phmbressan]

Why this method should not mutate the instance?