Homemade Tin Can Turbine With 3d-printed Compressor

In this instructable we will learn how to make a really working gas turbine engine at home. There are many tries to make a homemade gas turbine, described over the net. There are even some instructables, e.g. that one:

• Turbine Engine;
• How to Build a Jet Engine!;
• The Recycled Jet Engine;
• ...

But only a few of them are truly functional.

A typical report on a successfull homemade turbine begins with a bunch of photos of machine tools and stages of material processing. One usually needs at least a lathe, a milling machine and an argon welding tool and this makes the gas turbine engine much less affordable than it could be.

The most complicated thing to manufacture and the most critical for the turbine to work is its compressor stage. Usually one needs CNC or manual driven precise machining tool to build it successfully. Luckily to us, the compressor works at low temperature and can be 3D-printed. Also luckily to us 3D printers are affordable nowadays and most DIY-ers have them at home.

I must note that I'm not the first who got an idea to build the sufficient parts of gas turbine by using 3D printer. AFAIK the first one was Axel Borg from pulse-jets.com forum. You can take a look at his site: amazingdiyprojects.com. I must also note that the turbine, being described here, is totally independent elaboration from the one presented at amazingdiyprojects.com. I must state clearly that neither me, nor people known to me, have ever purchased the mentioned project, and hereby I am in no way can be imputed in disclosing any proprietary information in this instructable.
People skilled in the art can note different design, different technology approach and even different blade angles. The only contribution of Axel Borg was that he had shown that 3D-printed version of compressor will work if necessary efforts were applied. It appeared to be enough for me to apply the necessary efforts and design my own version. I'm gratefull to Axel for this.

Another thing, being usually extremely difficult to reproduce at home is so called "nozzle guide vane" or simply NGV. Luckily by trials and errors I've found a way how to make it without using an argon welding machine or other exotic tools.

You will need:

• 1) 3D printer, able to work with PLA filament. If You have an expensive one, like Ultimaker - good for You, but cheaper things, like Prusa Anet, will suit too;
• 2) Obviously You should have some slicing software, properly tuned to work with Your printer;
• 3) Even more obviously, You should have a proper amount of PLA filament to make all the printed parts. ABS won't suit for this project, since it is too soft. You probably may use PETG, but it is untested by me, so do it on the risk of Your own;
• 4) A tin can of proper size (100 mm diameter, 145 mm length). Preferrably the can should have a removable lid. Many sorts of tea or candies are distributed in such cans. At the least case You can take a common can (from say pineapple slices) but then You will need to make a metal lid for it all by Yourselves;
• 5) A sheet of galvanized iron. 0.5 mm thickness is optimal. You may choose other thickness, but You may face difficulties with bending or grinding so be prepared. In any case You will need at least a short ribbon of 0.5 mm thick galvanized iron to make a turbine shroud spacer. 2 pcs of 200 x 30 mm will suit;
• 6) A (smaller) sheet of stainless steel to make the turbine wheel, the NGV wheel and the turbine shroud. Again 0.5 mm thickness is optimal. You may use a chimney steel or even to sacrifice a stainless steel dinner dish for this purposes. A wide spread 555 "kitchen steel" works extremely well here;
• 7) A hard steel rod to make the turbine shaft. Beware: mild steel simply does not work here. You will need at least some carbon steel. Some hard alloy will be even better. Here we will assume that the shaft has its diameter of 6 mm. You may choose another base diameter, but You will then need to find proper materials to make a hub;
• 8) 2pcs 6x19 ballraces 626zz (11/14/21: earlier it was written 6x22 - it was a typo, thx to Yossis2 for note);
• 9) 1/2" pipe nipple 150 mm long And two end fittings;
• 10) A drill and some rig to secure it;
• 11) A grinder. In the worst case some whetstone will suit;
• 12) A dremel (or other hand engraving tool) with engraving discs;
• 13) A general purpose metal working tool set including hacksaw, pliers, screwdriver, M6 thread die, scissors, caliper and so on;
• 14) A piece of copper or stainless steel pipe to make a fuel spray;
• 15) A set of screws, nuts, clamps, vynil pipes and other common DIY stuff;
• 16) Propane or butane torch

If You want to run the engine You will also need:

• 17) A tank of propane. There exist petrol or kerosene engines, but it is a bit tricky to get them to work on these fuels. Better to start with propane and decide later, wheteher You want to transit to the liquid fuels or You are already happy with gas ones;
• 18) A manometer, capable to measure pressures of several inches of water column. Alternatively You can use a jar of water, connected to a transparent pipe attached to a ruler;
• 19) A digital tachometer is also desirable. One can certainly estimate the rpm by the sound of the turbine, but it requires a lot of experience;
• 20) A starter. To start the jet engine one can use:
  • a fan (100W or more rated, and better centrifugal one)
  • an electric motor (100W or more rated, 15 krpm capable; Yes You can use Your dremel here).

The hub will be made of:

• one 1/2" pipe nipple 150 mm long;
• two 1/2" female hose barb connector
• and two 6x15 ballraces 626zz.

For some reason the ballraces fit correctly to the nut parts of the fittings, so all You need here is to remove the dead weight: Use Your hacksaw to cut off the pipes from the fittings, and use Your drill to enlarge the remaining holes. You will end up with two cup nuts. Put the ballraces in the nuts and screw the nuts onto the waterpipe connector. Now You will be able to insert the shaft into the newly made hub and rotate it freely.

Theory (and experience to some extent) says that there's no difference whether You make the shaft of mild steel, hard steel or stainless steel. So choose one, that is more affordable to You.

If You expect to get a decent thrust from the turbine I'd advise You to use a steel rod having 10 mm (or larger) diameter. However when preparing this instructable I had only 6 mm one, so we will take it as an example.

Use Your M6 thread die to thread one end of the rod on 35mm of its length. You next task will be thread the other end of the rod in such a way, that when the rod is inserted into the hub (it is assumed that the ballraces are put onto the end of the pipe nipple and tightened with the hose barb nuts) and when nuts are screwed to the end of thread at both ends, there would stay a tiny bit spacing between the nuts and ballraces. This is a very complicated procedure. If the thread is too short and the longitudinal play is too large You can thread the rod further on. But if the thread appears to be too long (and there is no longitudinal play at all) there is no way to recover it and You will need to start up with a next rod sample. Better to have an excess of rods here.

You may also want to grind the shaft with a sandpaper to fit the ballraces more precisely.

For this sample there exists one prospective material - shafts from a laser printer. Their stainless steel looks hard enough and they are exactly 6 mm diameter. Their drawback is that 20-25 krpm is their limit. However the same can be said about any 6 mm steel rod. If You want higher rpm - use thicker rods,

Here are the STL files for the matrices for the NGV wheel:

• NGV_matrice_down.STL
• NGV_matrice_up.STL

Download them and 3Dprint

And here are the STL files for the matrices for the turbine wheel:

• turbine_matrice_1pass_down.STL
• turbine_matrice_1pass_up.STL
• turbine_matrice_2pass_down.STL
• turbine_matrice_2pass_up.STL

Download them and 3Dprint too.

It appears that the blade shape becomes more smooth if one presses the vane not into the final shape during one step (pass) but into some intermediate shape (1st pass) and only then - into the final shape (2nd pass). Therefore I provide STL's for both type of press matrices. For the 1st pass and for the second one.

Yes in this design the NGV has the shape of the wheel. Just like the turbine, but with mounting holes in it.

This design uses 2 kinds of steel wheels. Namely: the turbine wheel and the NGV wheel. To make them use a stainless steel. If made of mild or galvanized one
their lifetime would be barely enough to show You the engine working.

You could cut the wheels from a sheet of metal and then drill a hole at the center, but most probably Your drill bit will miss the center. Another approach is to drill a hole in the sheet of metal, and then to glue a paper template, so that the hole in the metal and the place for the hole in the paper template become coincident. Finally You can cut the disc with scissors.

You may find and download the templates below:

Turbine wheel template:
turbine_wheel_template.doc or turbine_wheel_template.pdf

NGV wheel template:
NGV_wheel_template.doc or NGV_wheel_template.pdf

For Your convinience the images are already in the properly scaled "doc" and "pdf" files. Load them into a text editor, capable to operate with doc-files (One can use Microsoft Office, Libra Office or Open Office for example).

Download the files, open with Your office software and print on a common paper using a common (not 3D) printer.

Now You can cut the wheels from a sheet of metal and drill the auxiliary holes. (Note that the central holes should be drilled already. Also note that the turbine wheel has only the central hole.)

It's also a good idea to leave some allowance when cutting the metal, and then grind the circular border of the wheels using a drill and some grinding device. However if You trust Your abilities to cut a good circle with scissors only, You may omit this procedure.

It might be better to make a number of reserve wheels at this step. Further on it will be clear why.

Freshly cut wheels are hard to be placed into the pressing matrices. To do the job use pliers to twist the vanes a bit. The wheels with pre-twisted vanes match the matrices much more easily. Clamp the wheel between halves of the press matrix and compress in vise. It really helps if the matrices have been lubricated with machine oil beforehand.

Vise is rather poor press machine, so, most probably, You will need to hit the assembly with a hammer to compress it further. Use some wooden pads in order for not to break the plastic matrices.

One may try to save some efforts by shaping the vanes using only the 2-nd pass press-matrices. However the aerodynamics of the blades will be a bit worse this case. The exhaust temperature will then be a bit higher.

Two stage shaping (using 1st pass matrices first and 2nd pass matrices to finalize the shape) gives definitely better results.

The doc file with template for the support tripod is here:
support_tripod.doc or support_tripod.pdf

Cut the part from a sheet of stainless steel, drill the necessary holes and bend the part as shown on the photos.

If You had a lathe, You could make the whole part from some bulk material. However not many people have a lathe at home. The DIY way to do it is to cut a number of planar discs from a sheet of metal put them one onto another and bolt them tightly to obtain a volumetic part.

These parts stay comparatively cold, so there is no need to make them of the expensive stainless steel. Use 1 mm thick sheet of mild (or galvanized) steel here.

The doc files with templates for the spacers are here:

• spacer_big.doc or spacer_big.pdf - for the bigger one
• spacer_small.doc or spacer_small.pdf - for the smaller one

You will need 2 smaller discs and 12 bigger ones. The numbers are given for 1 mm thick sheet of metal. If You use thinner or thicker one, You'll need to adjust the amount of discs to obtain the correct total thickness.

Cut the discs and drill the holes. It's also a good idea to bolt the same discs together, to clamp the bolt in Your drill and to grind them round. In such a way You can get a set of discs with equal diameter.

As the support collar keeps the whole NGV assy, You should use some thick material here. You can use some suitable steel spacer, or some sheet of (black) steel not thinner than 2 mm.
The doc file with template for the support collar is here:

• NGV_support_collar.doc or NGV_support_collar.pdf

Now You have all the parts to assemble the NGV inner. Install them on the hub as it is shown on photos.

As all predecessors we use a reaction type turbine. The reaction turbine needs some pressure to operate properly. And to keep the hot gases from expanding freely we need so called "turbine shroud". Otherwise the gases will loose the pressure immediately after coming through the NGV. To function properly the shroud has to match the turbine with a small clearance. Since we have the turbine wheel and the NGV wheel having exactly the same diameter we need something to provide the necessary clearance. This 'something' s the turbine shroud spacer. It is just a stripe of metal that would be rolled over the NGV wheel. The thickness of this sheet determines the clearance value. Use 0.5 mm here.

I think there's no need to provide a template here. Just cut a stripe 10 mm wide and 214 mm long from a sheet of any steel 0.5 mm thick. Note that You do not really need stainless here.

The turbine shroud itself will again be a piece of metal to be rolled over the NGV wheel. Or, better, a pair of pieces. Here You have more freedom to choose the thickness. The shroud is not simply a stripe since it has attachement ears.

The doc file with template for the turbine shroud is here:

• shroud_template.doc or shroud_template.pdf

Roll the shroud spacer over the NGV blades. Fasten with some steel wire. Find a way to fix the spacer, to keep it from motion when the wire will be removed. You may use soldering or brazing here.

Then remove the wire and roll the turbine shroud over the spacer. Again use some wire to make the wrap tight.

Just do as shown on the photos. The sole connection between the NGV and the hub is those three M3 screws. It limits the heat flow from the hot NGV to the cold hub and keeps the ballraces (rather) safe.

At this step You can Use Your shaft with the turbine wheel to check if the turbine can spin freely. If not - make the alignment of the NGV shroud by changing the position of fixing nuts on those three M3 screws. Vary the tilt of the NGV until the turbine can spin freely.

The doc file with template for the main part of the liner is here:

• liner_main.doc or liner_main.pdf

Glue this paper over a sheet of steel. Drill the holes and cut the shape. It is not necessary to use stainless here. It appears that liner works good even if made of galvanized (roof cover) steel. Roll the cut into a cone. Nothe that to keep it from unrolling we use curling here. If the steel is soft enough You may curl even with pliers.

The front part of the liner is here:

• liner_front.doc or liner_front.pdf

Again use this template to make a cone. Use a chisel to make the wedge slits and then roll the thing into a cone too. Fix the cone with curling. Both parts are kept together only by friction in the complete engine. So no need to think how to attach them at this step.

The impeller consists of two parts:

• impeller_base.STL - this one is the disc with blades
• impeller_up.STL - this one is the covering (shroud)

This is Kurt Schreckling's impeller, having been modified heavily by me in order to be more tolerant to the longitudinal displacements. Note the labyrith, preventing the air from coming back due to the backpressure.

3D print both parts and glue the covering onto the disc with blades. The best results can be obtained using an acrylic epoxy here. But You may also try other glues.

The thing has very complicated shape. And when the other parts can be (at least in theory) produced in DIY conditions without the use of precise machinery, this one can not. To make the things worse this part is responsible to the efficiency of the compressor to the most extent. It means that the fact whether the whole engine will be operational or not depends strongly to the quality and precision of the diffuser. That's why don't even try to make it manually.
Entrust Your robotic printer with the job.

For the sake of 3D printing convinience, the compressor stator has been divided into several parts. Here are the STL files for the:

• fire_cone_mount.STL
• case.STL
• connector.STL
• intrake.STL
• stator_base1.STL
• stator_base2.STL

3D print and assemble as shown on the photos. Note that a nut with 1/2" pipe thread is to be attached to the central body of the compressor stator. It is used to keep the hub in place. In this example the nut is attached by 3 pcs M3 screws. Here is the template, where to drill the holes in the nut:

compressor_nut.doc or compressor_nut.pdf

Alternatively You may find another way to keep the nut inplace. You may try to epoxy glue it if Your epoxy is good enough.

Also note a heat screening aluminum foil cone. It is used to prevent PLA parts from softening due to heat radiation from the combustion liner. As a source of the aluminum foil one can use any pare beer can here.

You will need a tin can 145 mm long and 100 mm in diameter. Better if You can use a can with some lid. Otherwise You will need to install NGV with the hub into the bottom of the tin can itself, and have further problems with reassembling the engine for service.

Cut off one bottoms of the tin can. In the other bottom (or better in the lid) cut a 52 mm circular hole. Then cut its border into the leafs as shown on the photos.

Insert the NGV assembly into the hole with leafed border. Wrap the leafs with a steel wire tightly.

Make a copper tube (6mm outer dia, 3.7mm inner dia) ring. Or better You may use a stainless steel tubing. The fuel ring should fit tightly into Your tin can internals. Solder (or weld) a fuel support pipe to the ring.

Drill fuel spray nozzles. These are just 16 pcs of 0.5 mm holes evenly spread over the ring. The direction of the holes should be orthogonal to the air flow. I.e. You should drill the holes on the inner side of the ring.

Please note that the presence of so called "hot spots" in the exhaust of the engine depend almost solely to the quality of the fuel ring. Dirty holes, or uneven holes and You will end up with the engine that brings a pair of NGV vanes to white glow, while all the other gases are still cold. Such an engine will just destroy itself in bare tries to start it up. The presence of the hot spots depends much less to the liner quality than others try to say. But the fuel spray ring is essential.

Check the quality of the fuel spray by the test fire. The flamelets should be equal to each other.

Once completed, install the fuel spray ring into the tin can body.

All You need to do at this step is to put all the parts together. If things went good, ther will be no problems in performing this. Otherwise You will need to fit the parts manually.

Lute the tin can lid with some kiln sealant. If You have no access to the kiln sealant, You can use a silicate glue (Yes, that nasty office glue) with some heat resistant filling agent. One may use graphite dust, steel powder and so on.

After the engine has been assembled check if its rotor spins freely. If it does, make a preliminary fire test. Use some reasonably powerfull fan to blow into the intake or just rotate the shaft with dremel. Turn on the fuel slightly and ignite the stream at the back end of the engine. Adjust the spinning to let the flame into combustion chamber.

PLEASE NOTE: at this step You are not trying to start the engine! The sole purpose of the fire test is to heat it up and to look whether it behaves good or not. At this step You can use a cylinder of butane, being commonly used for handheld torches. No need for propane yet. If the fortune is on Your side, You can proceed to the next step. However it is better to seal the engine with a kiln sealant (or with silicate glue, filled with some heat resistant powder).

You can start the engine using any common procedure. Either by blowing air into its intake or by spinning its shaft with some starter motor.
Be prepared to burn up several NGV (and maybe turbine) wheels while trying to start. (That's why at step 4 it was recommended to make a few reserve ones.) Once You become convenient with the engine, You will be able to start it flawlessly at any time You want.

Usually You have to use propane as a fuel at this step. I suspect that if manufactured and assembled with extremely care and accuracy this engine may self sustain on butane gas (for handheld lighters and torches) but commonly it can not.

Please note, that at the present state the engine can serve mostly educational and recreational purposes. But this is a fully functional turbo jet engine, able to spin up to any desired rpm (including the self destructive ones). Feel free to improve and modify the design to fulfill your (RC-flight) goals. First of all You will need a thicker shaft to reach higher rpm and therefore thrust. The second thing to try is to wrap the outer surface of the engine with metal pipe - fuel line and to use it as vaporizer for liquid fuel. This is where the engine's design with hot outer wall comes handy. Yet another thing to think about is a lubrication system. In the simplest case it may take a form of little bottle with some oil and two pipes - one pipe to take pressure from the compressor and direct it to the bottle and another pipe is to direct the oil from the pressurized bottle and to direct it to the rear ballrace. Without the lubrication the engine can only work for 1 to 5 minutes dependently to its NGV temperature (the hotter NGV the shorter runtime). After that You need to oil the bearings all by Yourselves. And with the lubrication system added the engine can run for a long time.
CREDITS:

Thanks to Kurt Schreckling and Thomas Kamps for their books.

Thanks to Axel Borg for the idea of 3D printed compressor and to Mike Everman for his excellent forum (pulse-jets.com), where this idea was published.

SPECIAL THANKS to: bbewamit on Youtube for his video of a simple homemade turbine. From that is seen on the video the turbine is doubtfully self-sustain but still the video is very encouraging.

Also note that I am NOT planning to sell and send the kits. Just get a 3D-printer and follow the instructable. Thanks for understanding.

Note that in response to public demand I've added the cross sectional view of the engine with parts labeled.