Flying Model Rocket From a 3D Scale Model

Building sport scale rockets from scratch is a highly rewarding pursuit, because it involves a fair amount of engineering; scaling to a given size, stability, material choices for strength and light weight, and motor selection for a desired flight profile. Yes, it is rocket science; fortunately, there are computer programs that can do the heavy-duty math for you.

With the advent of 3D printing, this process becomes much more exciting, opening up new opportunities for detail on scale models.

When I saw a 3D model of SpaceX's "BFR" (since renamed the "Starship") on Thingiverse, I saw my next flying rocket project. This model is designed as a solid, as most models for 3D printing are, so some work must be done to:

1. Lighten it up so it'll fly
2. Scale the parts to fit standard rocket tubing
3. Make a motor and recovery system fit in it, and
4. Make it fly stable (most important!)

This will be a mostly 3D printed model, and will also use some standard size rocket tubing, plus other parts. I am assuming you have a bit of model rocket experience, as well as 3D printing experience. I recommend building a few kits before attempting a scratch-building adventure.

I take no responsibility for omitted assembly steps or bad flights. Please fly safely!

The finished file can be downloaded on Thingiverse.

There is now a barebones kit containing the 3D printed parts available on eBay:

https://www.ebay.com/itm/303604364151

• 3D printer and filament of your choice
• Estes BT-60 tubing (Inner Diameter: 1.595" Outer Diameter: 1.637"), 7.5" long
• Estes BT-50 tubing (Inner Diameter: 0.950", Outer Diameter: 0.976"), 4" long for motor mount
• 2 centering rings, .98" ID X 1.5" OD X 1/16" thick (these are smaller OD than standard BT-60 rings because they will be inside the plastic part). I used paper, plywood would also work.
• Motor hook or motor retention of your choice
• 4 ea. fins, 3 1/4" X 2" X 1/16", clear polycarbonate (for scale appearance) or plywood
• 18" parachute, shock cord
• 3/16" Launch lug with 1/8" standoff, or printed launch lug
• Epoxy. I recommend epoxy over super glue because it's easier to make structural fillets with epoxy, and epoxy is less brittle.

Files downloaded from Thingiverse are typically .stl files which, while being easy to import into 3D printing software, are not easy to edit. Fortunately, there are several excellent free tools that can help with editing such files, though they each have their strong and weak points. The two tools I use most are 123D Design (discontinued, but can still be downloaded here or here), and Tinkercad, which is browser-based.

123D Design: This program cannot edit .stl files directly; sometimes you can convert the file for editing within the program, but since it's a resource-heavy action, trying to do this frequently crashes the program. Also, if the stl file has any mesh defects, converting won't work at all. One thing 123D Design shines at, however, is scaling objects precisely by percentage, even stl files.

Because I needed to scale the parts of the rocket accurately to fit available tubing, I imported the main files into 123D Design, made a cylinder (Blue in the picture) the exact outside diameter of the tubing I intended to use, (1.637"), and mated it to the booster section of the model (the transparent gray part in the picture). Scaling by percent, in a trial-and-error fashion, I got the booster size right for the tubing. Now I knew what scale percent I needed for the upper part (yellow) and could scale it to match. I then exported the correctly sized files for the next step.

In case you're interested, this works out to about 1/200 scale.

Now I'm going to switch tools. Here's where Tinkercad shines: Though it can't do precise scaling, it can easily modify stl files. Since it's browser-based, all the hairy computing involved in editing an stl is done on a server instead of bogging down the home computer. I started a new project in Tinkercad, then uploaded (imported) my scaled parts into the workspace. (Be sure to import them without changing the scale from 100%!)

(For those new to Tinkercad, tutorials are available here and here.)

The solid parts need to be hollowed out for two reasons: To save weight and to allow the motor ejection charge to eject the nose and parachute. In Tinkercad, I combined a cylinder and a dome into one part, then converted the assembly into a hole. Clicking on any object in Tinkercad allows it to be made a hole or a solid. I then centered the hole in the upper stage of the model (using Tinkercad's alignment tool) and combined (Tinkercad calls it "Grouping") the two for a finished part.

Whenever you group a "hole" with a "solid" in Tinkercad, the part where the "hole" was disappears.

I aim for a wall of the hollowed part about .050" thick, which means I need to make the hole .100" smaller than the diameter of the part. This is a good thickness for light weight and strength. In the first picture, the part to be hollowed is made transparent to better see what we'll be removing, and the future hole is pink. Picture 2 shows the pink part made into a hole, and picture 3 shows the result after part and hole are grouped. The very front of the part is left solid, both because it's easier and because I want the nose to be stronger and somewhat heavy. We'll discuss the reason for the weight in a later step.

More Tinkercad! The booster section will be divided up into a top part and a tail part, with a section of lightweight cardboard rocket tubing connecting them. Make a copy of the booster; you'll need one for the top section and another for the bottom.

The first step is separating the top of the booster from the rest. Again Tinkercad comes to the rescue; you can delete any part of something by making an appropriate "hole," so we start by making a big square hole that deletes the rest of the booster, leaving only the top part with the detail we want to save.

Next, I make a hole of the correct shape to hollow out the part. It needs to be two different diameters because the top part is tapered. Again, I aim for a wall thickness of about .050", although I'm too lazy to make tapers, so a small cylinder on top of a larger one will suffice.

Finally, I need a shoulder that will fit the inside diameter of the tube I intend to use. This part will be the lower part of the nose cone, so the fit needs precision. This is simply made by making a cylinder the inside diameter of the tubing and making a hole down the center. I printed a couple of test parts at this point to verify the fit. Once I was happy with the fit, I copied the shoulder because I also need a shoulder for the tail part, coming up next. I then grouped the shoulder with the top section.

The tail part is made with the other copy of the booster, cutting off the top of it in the same way the bottom was cut off. The model engines also need to be removed. It's then hollowed out in the same way, although just a simple cylinder is needed this time.

The fins on the tail of the model are far too small to make a model fly stable; the reason the "big" rockets have small or no fins is they have active guidance - meaning they can steer themselves with gimbaled engines instead of relying on passive stability the way an arrow or a model rocket does. Thus, we need to make some provision for larger fins. Since the model has four small fins, the best thing to do is install the larger fins between each pair of small fins. Two rectangles, 1/16" thick, are combined at right angles to create a fin slot pattern, which is then made a hole and subtracted from the part. Finally, that extra shoulder is added to the top of the part, and the tail is complete.

The last thing to do is find the length of the complete booster and subtract the length of the new top and bottom parts to find the length of cardboard tube needed. It worked out to 7 1/2" in this case.

Back in Tinkercad, I also made a thin half-cylinder part to go the length of the cardboard tube, replacing the detail on the original part. This, in the real rocket, is called a Range Safety Package - It's an explosive charge that lets the Range Safety Officer remotely destroy the rocket in case it starts heading for Downtown Orlando instead of Outer Space! There's your space trivia for the day. (This part may also be a "cable tunnel" - A place to run wiring from top to bottom without having to run it through a fuel tank.)

Each of the four parts made in Tinkercad can, and probably should, be exported (downloaded) separately. The Export button is right next to the Import button used to put the parts there in the first place. They may need to be rotated for best printing orientation; this can be done in Tinkercad before exporting or later in your slicing software.

I printed the parts in white High-temperature PLA, although other plastics would probably be better, particularly in Arizona, in the summer... A raft is advisable because of the small bed contact area, but no supports are needed. Infill was set at 50% for strength and nose weight.

A couple of coats of automotive filler-primer help smooth out the layer lines for painting.

Before setting the design in stone - or plastic - this is a good time to do the really hairy math that will make the rocket fly straight. Fortunately, there are several excellent rocket design programs that will do most of the math for you, so you don't need to brush up on that calculus you slept through in school. The two I've used extensively, and which have sort of become the standard, are Rocksim and Openrocket. They both have advantages. Rocksim costs $123, but does a much better job on complex designs like this one. Rocksim has a 30-day free trial available here. Openrocket is free, but more limited. The .rkt file for this rocket will open in Openrocket, but will not be correct. It will be correct in Rocksim.

With a model like this, it's probably necessary to at least print the major parts in order to find out what their weight and balance parameters are. We need a virtual model of the rocket in Rocksim, then it's necessary to weigh the actual model, find it's CG (Center of Gravity, or balance point), and plug this and the weight into the virtual rocket. At this point, the model should be assembled without glue in case some weight needs to be added or fins need to be changed.

Here's why it's not a bad idea to have the nose a bit heavy: The CG must be in front of the Center of aerodynamic Pressure (CP) for the rocket to be stable. The CP is where all aerodynamic forces balance, and is automatically calculated in the program. The CP moves forward if fins move forward - like the front fins in this design - and moves aft if the rear fins are enlarged. The CG must be checked with the largest motor you plan to fly installed, and the rule of thumb is to have the CG at least one body diameter ahead of the CP with the motor loaded.

Because plastic is relatively heavy, I designed this to fly on Estes "D" motors. My model weighs 7.7 ounces, and flight simulations predict an altitude of about 475 feet on a D12-5.

I played with fin size until I had a satisfactory result and settled on fins 3" in span. The actual material is cut to 3 1/4" because the fins go through the tail and glue to the motor mount tube.

With this model, the empty CG should be no further aft than 11 3/4" from the nose tip.

If you want a kit of the 3D printed parts, find it here:

https://www.ebay.com/itm/303604364151

As with most model rockets, the motor mount is constructed first. I installed a standard motor hook on the motor mount tube. I cut two centering rings with an O.D. to fit inside the plastic tail part, and glued them to the motor mount tube, close to the ends where the rings will not block the fin slots.

I epoxied a Kevlar shock cord to the upper end of the motor mount, then epoxied the motor mount into the tail. This is a good time to dry-assemble one last time, with fins, parachute, and motor mount, and check that the CG is far enough forward. If the CG is too far aft, add some nose weight before going any further. I didn't need any nose weight, but your build might. Epoxy the BT-60 body tube to the tail.

Now the booster top can be epoxied to the nose. I epoxied a loop of Kevlar into the booster top as a shock cord anchor. If using clear fins for a scale appearance, paint the model before attaching the fins so you don't have to mask them later. I've also included properly scaled SpaceX decals for the rocket. Print at 100% or "no scaling" to get the correct size.

When attaching the fins, install a new or spent motor in the motor mount to prevent the fins from crushing the motor mount tube, because you want to press them in firmly.

Glue the "Range Safety Package" to the body tube, in line with the matching stub on the tail. Install the 3/16" launch lug on a 1/8"-3/16" standoff (or use the 3D printed lug). Make sure the launch rod will clear the fins!

Add the parachute and chute protector, and you're done!

First flight on this rocket was with a D12-3, and the flight was just about perfect. A second flight on a D12-5 was also very good.

Original design by Zastro, https://www.thingiverse.com/thing:2565361 heavily modified by me.