Having access to a wide variety of nozzle is a plus for research rocketry. Making your own seems like the obvious solution. But how? Buying a lathe would be a costly solution. If you don't mind having single-use nozzles, then making a mold is far cheaper solution. The nozzles shown on the right are a 38mm anchoring cement nozzle and a 98mm castable ceramic (Alumina) nozzle. This web page details the creation of the mold for that 38mm nozzle.

This information is provided as-is, use at your own risk. Measure the nozzles you create carefully (with calipers) after they dry completely to ensure the size is suited for use in a rocket motor.

While dimensions are provided, you'll need to adjust those for the nozzle being used, molding material and so on.

First, we start by making a block mold. Some of the materials used are far from ideal but will work fine for a quick mold.

This nozzle has an outside diameter of 1.32". We start by drilling a hole close to that diameter in the middle of a 2" x 2" piece of 1/4" plywood. As I didn't have an adjustable hole saw that for a hole that small, I used a 1 3/8" (1.375") drill bit instead.

Cut a second piece of 2" x 2" piece of 1/4" plywood. Do not cut a hole in this one. Mine does have a little hole in it because it was placed under the first square when that was drilled. This isn't needed, just how I happened to do it to avoid drilling into the vise.

Glue the two square plywood pieces together. Woodglue was used to do this job.

Now, we'll make the sides for the mold box. These were made from 3/16" foamboard, you can also use ply. Two of them are 2" x 1 7/8" and the other two 2 1/2 x 1 7/8". Actually, a step was skipped, so read this entire page before you make your own, to avoid the same mistake.

Take one piece of each and position them as shown. Use some tape (on the outside) to join the two pieces as shown.

Repeat the above step with the remaining mold box side pieces. This gives you two side pieces, as shown.

As the whole in the base was cut slightly larger than the nozzle diameter, it needs to be sealed for a better fit. Modeling clay can be used for this. Simply position around the edges of the hole. Take a small piece of modeling clay and wrap it in Seran wrap. Push it into the aft end of the nozzle. Do not push it into the nozzle throat, use it only to fill the entire exit cone.

Cover the nozzle with wax, or mold release, to avoid the silicone sticking to it. Be sure to cover the inside of the o-ring groove with a thin coat of wax. Also put wax on the base and the inside of the side pieces. Place the nozzle in the base, forward end pointing up. Remove any modeling clay that may be on top of the base. It should be level with the base.

Remember that step I forgot? Here it is. The mold base needs some small pieces glued in the corners. This will help connect the two mold pieces when casting. This is best done before the nozzle is placed into the base.

Position the two side pieces around the base. Use a rubber band or two to hold things in place. Here, the rubber band was double wrapped.

The silicone is now poured into the mold. This sounds easy but can be a touch tricky to get right. The idea is to avoid air bubbles, especially in the o-ring groove. Some quick tips : vacuum the silicone if possible, and use a shaker table if you have one. Pour into a corner in the mold. Pour slowly and try to pour from some distance, so you get a ribbon or thin stream of silicone.

Once the silicone is cured, open the mold box and remove the base. Remove the Seran wrap with modeling clay in the exit cone. Don't remove the nozzle yet.

Put the sides of the mold box back into place, again using some rubber bands to secure them in place. Cover the mold surface (and nozzle, including exit cone) in wax, or mold release.

You'll need to be able to inject or pour your casting material into the mold. For this purpose, a pour and air hole need to be added. Put two small rods on the nozzle. The bigger diameter the rods, the better, just make sure they don't go over the edge of the nozzle or over the exit cone. You can use a dab of modeling clay to keep them in place.

Fill the model box with mold material, silicone in this case. Be careful so you don't move the rods out of place. This will make the second half of the rocket nozzle mold.

And here is the finished mold. Also shown in this picture is the graphite insert we'll be using. This is a simple graphite disc with a hole drilled into it with a regular drill bit. The hole is equal in size to the nozzle throat diameter. This graphite disc is a slice cut from a graphite rod.

These pictures aren't very clear, unfortunately. You start by pouring casting material into the bottom of the mold. In this case, anchoring cement was used. Pour it in until it almost reaches the top of the nozzle throat area. Swirl it around so it cover the o-ring groove completely.

Next, slide the graphite insert onto the nozzle throat part of the mold. The syringe you see will be used to inject the last bit of cement into the mold. I cut a good part off the tip to make the casting material flow easier.

Put the top of the mold in place. Insert the syringe into the pour hold and pour casting material into it. As you can see, the plunger part of the syringe was removed. Once the casting material is poured in, insert the plunger part and inject it into the mold. Repeat until casting material oozes out of the other hole. Let the casting material cure and demold.

Random info and thoughts

Why the graphite insert? The nozzle throat would erode badly without it and chamber pressure would drop. Graphite is used rather than say steel washers because those can't be drilled. With the graphite insert, the nozzles can be drilled (after they are cast) to the desired nozzle throat.

After firing lots of 54mm to 98mm motors, some 38mm cement nozzles were made to test some new AP propellants. In the 38mm, I have experienced two burn-throughs. These appear to be caused by the nozzle convergence cone eroding and the flamefront coming in contact with the motor casing. Until a solution is found, these nozzles won't be used in 38mm motors, only larger ones.

The silicone used is Smooth-on OOMOO 30. It may not be the best material to use (it tears fairly easily) and isn't cheap. I have some other silicone on order to try it out. I'd suggest getting their sample size and trying it with a 38mm nozzle first.

Update : After making several mold with Mold Max 30, I can say that the Mold Max is far superior for this application than the OOMOO.

Different casting agents can be used. I started with Quickcrete anchoring cement. The ratio cement to water was 100:22. It works well but starts to harden very quickly, especially in a warm garage.

Quickcrete grout works just as well. It has a much longer working time and costs far less. This also works well with a 100:22 grout to water ratio. If it isn't pourable, add a drop or two of water at a time and stirr. All of a sudden, it will become quite liquid.

For both of these, expect quit a bit of erosion on the exit cone.

The cement and grout nozzles tend to shrink a bit after a day or two of drying in open air. This can produce a nozzle which isn't large enough in diameter and not suited for use. To work around this, I place a weight on the mold while the casting material is curing. How much weight you need takes some experimenting. Try it without the graphite inserts (to avoid wasting expensive graphite) until you get that part right.

The 54mm nozzles usually barely fit in the casing when demolded. A day or two-three later, they are just fine. If the diameter seems too large at first, give it a couple of days.

Castable ceramics (such as Cotronics Alumina or Zirconium) also work, but cost a good deal more. It also doesn't flow enough to use a syringe. You pretty much need a shaker table to cast this material.

The idea was to create re-usable nozzles. They erode quite a bit less than cement or grout, if at all. The problem was that they tended to break while removing them after firing the motor. Still, for larger motors, where you want an efficient nozzle, these materials offer a good solution.