The BIG ONE: Aerospace

I know I haven't posted much about my model rocket project for like the last month, but that was all according to plan. I do not have nearly the amount of time to spend on this project compared to my CPU project as this is not a school related project. However, I feel like this post should make up for my drought of information.

Start: OpenRocket

I started this project in OpenRocket, building an initial model of the rocket using an Estes D motor and a Featherweight altimeter. At this stage, the goal was not precision, but how I was going to proceed from here. OpenRocket is well suited for high-level questions like overall dimensions, mass distribution, and expected flight behavior, including stability margin, altitude, and acceleration trends.

This model was rough, however, it allowed me to sanity-check proportions and get a first-order sense of how the rocket might behave in flight before committing to detailed geometry or fabrication decisions.

Translation into 3d model through CAD

The next step was translating that conceptual design into FreeCAD. This model marked the transition from “does this make sense” to “can this actually be built and analyzed.” Unlike OpenRocket, CAD forced me to confront geometry explicitly: wall thicknesses, transitions, fin profiles, and how components physically connect.

Because I intended to use this model for CFD, I also had to decide what details to preserve and what to simplify. OpenRocket abstracts geometry heavily, but CFD does not. Every fillet, edge, and surface affects meshing and solver behavior. The resulting CAD model is complete enough to serve as a baseline, with the understanding that future iterations will evolve based on both simulation results and real flight data.

The beginnings of CFD

This version of the model represents my first serious attempt at CFD. I used the CFDOF OpenFOAM plugin in FreeCAD to convert the solid geometry into a solvable case, but beyond that, most of the setup came down to experimentation. Defining inlet and outlet conditions was particularly sensitive; small mistakes often caused the computed forces to blow up to unphysical values.

The ParaView visualization shows the magnitude of forces acting on the rocket body, currently on the order of ~10 N. There are still streaks of unusually high force that I believe are numerical artifacts rather than real flow features. At this stage, accuracy is secondary. The purpose of this work is to understand how setup choices influence results, so that later refinements are meaningful.

Once the mesh was built and the case compiled, I ran the solver while monitoring the residuals. The early noise reflects the flow field spinning up from initial conditions, followed by steady decay as the velocity fields settle. A spike partway through corresponds to a pressure–velocity correction step, after which the solution smooths out again. Pressure converges more slowly than velocity, but the overall trend indicates the simulation is approaching a stable, believable solution.