What is a One-Rotor-Wankel engine?

A one rotor Wankel engine is the most simple and smooth running reciprocating internal combustion engine existing today.

It only has two basic moving parts. There is no reciprocating masses, except for the combustion gases, as a result of the cycling combustion chamber volume.

Such an engine is running with next to no vibration except for torque fluctuations resulting from a reciprocating combustion.

Power is produced in this engine like in a single cylinder two-stroke engine, one combustion for every crankshaft revolution. The basic difference is the fact that each combustion CYCLE takes three crankshaft revolutions to complete. This allowing high crankshaft RPM at a comparatively slow combustion cycle, resulting in a potential power to weight ratio comparable to the lightest known two stroke engines.

With all this in mind, I decided to put more effort in realizing this engine than one would put in a normal auto engine conversion.

As it turned out, it was way more than I anticipated in the early stages of the project.

What engines are available from in-production hardware?

I found out very quickly, that there is essentially nothing except for MAZDA with the RX7 and later RX8 sports car. As stated earlier, these engines are too big and too powerful for my application. So, take one of these engine and cut it in halve, similar to what has been done to make a 1/2 VW engine some time ago. Doing this with an RX7 engine would result in about 100hp and an acceptable weight, depending how much effort I am willing to put into modifications.


Engine Design

After looking at the parts of my first disassembled 12A engine, it was clear that some major parts needed to be made of aluminum rather than using the original Mazda cast iron version.Talking in particular about the side housings.

The first issue coming to my mind on aluminum side housings is the wear surface. Nobody has come up with an after market aluminum side housing answering all the questions about a good wear surface at an acceptable cost.

In order to keep the developmental risk manageable, I decided to retain the CI(Cast Iron) surface of the side housing. Meaning a compound housing, made up of an aluminum outer part and a CI insert as the running surface for the rotor.

All aluminum casting are made of AL 356 T6. I made all the patterns and got it cast by a speciality foundry.

Subsequently having to switch to a peripheral intake port, because I did not see a possibility for integrating side ports with a  sound engineering solution. And, off course, as I learned later, the P-Port intake also has other advantages.

The prove of principle engine made its first run in ....2001. It is running presently with the final version of the compound side housing(end of 2009).

This first configuration had a peripheral port, and in addition, one side port. The side port idea was to improve low RPM idle. I immediately observed interference of the two ports at certain RPM ranges. It also turned out that on PP only, Idle around 2000 RPM was satisfactory for AC use, and I decided not to waste any time looking into the port combination any more.


Final version of compound side housing

Final motor mount design

CI section as used in first engine

AL part of sandwiched side housing

Prove-of-Principle engine as running today(2009)

First engine mounting idea

At the end of the design and development process, the only MAZDA parts left the engine, which are not modified are, the rotor, oil pump, stationary gear and some bearings.

As of today, January 20 20010, the first generation design is frozen and will be installed for flight testing in the PRAGMA air plane.

This should happen in the next couple of month.

Cutting Parts.


The first lesson I learned was: I can not afford getting the necessary machining work done by a machine shop.

So, shelving the project? Not only “nooooo...” but “ h.... no” .

As a result, a gradual building up of machining capability was essential for the success of the project.

Starting out with making an excenter shaft was like jumping in a swimming pool at 35F. But it worked out pretty well for the prove-of-principle engine.


  Finally, by the end of 2008, my machine shop was joint by a CNC milling capability, making things possible I was only dreaming off a few years ago.


I was always told, you can do it either quick or good, but not both. Wrong! Once the G-code was written and debugged, everything was quick and came out good. And that every time I made a part.

Things like measuring twenty five times, and cranking a handle the hole day, are a thing of the past.

The picture on the left is just an example what can be done in a matter of hours, including setting up, versus days with manual machine tools.


A first test run with the final configuration engine took place in early August 2010. Everything was going pretty well, except for  excessive oil exhaustion through the crank case breather.


The reason for this was the location of the breather pickup on the engine. I had it on the oil filler, which is on top of the accessory housing. Apparently, the oil is just too much in motion and foaming in this area for this solution. Hooking the breather up to a fitting in the oil pan, just below the anti-foam plate solved the problem.

The engine is now running very solid at 6000RPM and WOT. Calculated HP is 92 - 100 , depending on the BSFC No. using.


Demonstration run at Alternative Engine Gathering in Paducah, KY, 2010.

Picture by Douglas Dempsy.


In my personal view, it is already more power than I need for flying the PRAGMA.

Configuration Overview.

Engine size                          1/2 MAXDA 12A

Excenter shaft                       Modified 12A with case hardened bearings and 10:1 taper output.

Flywheel                                 Custom design with integrated balance weight.

Oil pump                                 MAZDA 12A

Water pump                           Subaru

Alternator                               30A PM.

Ignition                                   Streetfire CDI , dual coil

Carburetion                            AeroV Injector.


Moving into the area of airplane installation presented a bucket full of new challenges. Some of these are still cooking now.

For a motor mount, I choose a 3-point version, with 2 Hard points and one tension strut. A nice feature on the AVID is, it has six hard points on the fuselage for mounting the engine. Unfortunately  two of those were in the wrong location for a practical rotary engine mount. With that corrected, the motor mount turned out pretty decent.


website design software

Early fit-check of engine-mount-AC interface.

Engine instrumentation.

Looking into what to use for measuring engine parameters, I ran into several problems.

All engine monitoring systems I found on the market have capabilities enough to monitor 5 engines at the same time.

I also wanted to use a simple data aquisition system I already have. Generally used sensors for temperature and pressure are highly nonliear, and therefore require computer calculation for generating presentable physical parameters.

The way I am taking out of this is using linear sensors to begin with.

For temperature, I made my own sensor housing with standard dimensions using a LM34 series temperature sensor.

For pressure an industrial sensor from Digikey came in handy.

One disadvantage of these sensors is the fact that they need three and four wire connection. But it worked out for me better then the alternative.

With the linear sensors, I am able to use simple DVMs for indicating the actual value with a minimum of signal conditioning.


                 Pressure and temperature sensor

Finally, the engine design has been tested and debugged. A test run simulating a 20 minutes flight around the pattern, including go-around, has been conducted many times. This made me confident that the engine is ready for flight testing.

While leaving the ground test engine on the test stand for further component testing, the flight test engine is getting assembled.


As shown in the above picture, I am using studs instead of tension bolts. The reason for this is to avoid wear on the threads in the AL when assembling and disassembling the engine. From here on it is pretty much like assembling a two rotor, only some what quicker.

The side housings are already assembled with the CI insert and the stationary gear in the rear housing.

Next is bringing the rear housing into a horizontal position and inserting the rotor housing with MAZDA stock O-Rings. Insert the e-shaft and the rotor with seals, slide on the PTO-side housing,   add washers and nuts, and the block is complete.