Sub minimum-diameter rockets

SU-1 #

On September 7 2024, I launched my first rocket that achieved supersonic speeds. It also shattered my previous record of 700m, raising the bar to almost 1800m. The rocket was around the same length as a bottle, and weighed 0.5kg.

Here’s how I achieved such performance in a tiny size.

Minimum diameter rockets #

Typically a model rocket will have a component called the motor mount. This is a part that transfers loads from the motor to the whole structure, and retains it during the coasting phase of the flight.

Minimum diameter rockets don’t have a motor mount. Instead, the airframe internal diameter is the same as the motor casing external diameter, with a perfect fit. There are retaining and load-transferring structures present. The fins are most often bonded to the airframe, but that’s not the only way to do this.

The advantages here are a lower airframe diameter which reduces drag, and a lower mass thanks to the absence of a motor mount and a reduced airframe diameter.

Subminimum diameter rocket / submin #

The submin is a step further down the optimization path. Here, the airframe is the casing of the rocket motor. This offers the greatest weight reduction, especially with a composite case and casebonded propellant. The downside here is that if the motor is single use, the whole booster section, which may include the fins, becomes a single-use component.

In built a CNC filament winder last year (still gotta do a writeup on that one), so the composite single-use submin was the obvious way to go.

Building a submin #

The goals were clear:

• Break my record of 700m
• Break the sound barrier
• Recover

I started doing some preliminary simulations in openrocket to get an idea of what I had to do. With a propellant that can provide 120s of ISP, I determined that a low H class rocket motor with around 200Ns should do the job with some safety margin. A 28mm ID and 31mm OD was chosen for the dimensions, and the vehicle ended up having a length of just over 500mm.

Like with my previous projects, simplicity matters above all. A simple design will mean higher reliability, lower cost and less work at the desert with launch prep.

I decided to build the rocket as an assembly of 3 components:

• Avbay with the electronics, nosecone, and coupler tube
• Fincan that slides onto the motor casing and can therefore be reused
• Motor, with an extended casing so the avbay coupler can be inserted

Avbay #

It’s typical for these kinds of builds to have a tricky avbay design. This was no exception. The avionics I selected were my Tracker-PTR, and the legendary UWS board, which is a fully analog system that measures the earth’s magnetic field to detect the pitchover at apogee and deploy the chutes. This was important to me, because it makes the whole system unaffected by the effects of supersonic flight.

I took pictures of the UWS board, and then imported and calibrated them in CAD to get an accurate 3d model. Next, I threw in my model of the UWS and designed a simple frame to hold everything together. The frame is sandwiched between the 2 boards, and has holes for securing it inside a piece of airframe section to form the avionics bay. On the other side I glued a coupler made from some phenolic tube. This also serve’s as the cannon that deploys the cute and separates the front section from the booster.

The fiberglass tube gets capped off with a PETG nosecone, which covers the tracker antenna and arming switch. This is the last point on the checklist, which involves securing the nosecone with 2 screws after arming the deployment board.

Fincan #

Perhaps a little under-engineered compared with other parts of this project. The fincan is made from tip-to-tip fiberglass composite over a printed PLA core. It was very quick and easy to make, looks great, is really stiff, and I didn’t need a mandrel for the tube. The fins + tube are all one piece, which gets laminated after printing and a quick prep. The part then slides onto the motor before tieing the main line, and rests on the thrust ring. For future builds, I might just bond pieces of G10 directly to the casing.

Motor #

As mentioned before, the motor is a single-use, composite-cased motor. The case has 2 layers of tow, wound at 55 degrees for a total of 1.5mm case wall thickness. It’s so thick because the fiberglass tow I’ve got isn’t great, so the layers come out thick. Carbon fiber would need much more layers to achieve this thickness. The nozzle is from my SUGARCOAT mix, which performed well in this flight. Foreword closure was just an opacified epoxy plug, where I also submerged a loop for the main line while the resin was liquid. Given the 1s burn time and geometry, I didn’t bother with a liner. The thing is single-use anyway. The case extends a few cm past the closure, and that’s where that avionics coupler sits during flight.

Performance #

The rocket was designed using OpenRocket and RASAero. Simple trapezoidal fins and a Von Karman nosecone. There was no barometric altimeter or accelerometer on board, so all data is taken from the GPS, which isn’t accurate. According to the data from the tracker, Sonic Boomer reached an apogee of 1785m, which is within 100m of the value given by the RASAero sim. I’ve never seen anything leave the pad so fast, the launch was really something. The vehicle was recovered in perfect condition, sans the paint on the fincan which got some mach rash.

Planned work for the future #

I really want to fly this as a second stage of a larger vehicle. Flying on a K class booster the thing should hit mach 2.

I’m also going to scale this project up. I like the idea of performance brainrotted designs, and with my winder these builds are super fun. I have parts for 38mm and that’s probably what I’m going to do next.