A 4-Rotor discussion.
 

First, this is a discussion on the methods and logic behind building a custom four rotor engine. And to make moderators happy, i put it here becuase i want to use S4 parts and an FC chassis.

Rules:
1. if you dont have anything smart to say, dont.
2. this could be a good brainstorming session and denouncer of myths.

Here is why im starting this.

In order to finish college and get a degree, you must have a senior year project. For MCE they call it a senior design project. Basically a year long research project utilizing what ive learned over the years of schooling. Granted, im presently a few years away from having to start, but i would like to start formulating ideas and concepts that could be applied. and at the same time give a resource for such information, seeing that most of this is "secret" from many people who have succeeded.

After sitting down and thinking about what the biggest obsticle is, the eccentric shaft, i came up with three methods that could be used in order to produce a functioning four rotor engine. For each of use i will give them three letter acroymns becuase i dont want to type them out that much.

Notation i will be using:

| FR1 | FR2 | RR1 | RR2| <--showing the location of the rotors. FR meaning Front Rotor, and RR meaning Rear Rotor with the number 1 or 2 meaning which rotor in the series.. the hookey picture above shows the layout of the rotors in discussion from front cover to bellhousing.


METHOD 1: Staggard Firing Rotors. (SFR) Version 1

Essentially, this is what the mazda lemans car does. There is a three piece eccentric shaft custom made for the application. Due to the length of the shaft and the assembly of a rotary engine, it has to be three pieces. otherwise, there would be no way for it to be assembled. The center nods are the primary piece. these are the FR2 and RR1 rotors. Then the FR1 and RR2 nods are added on to each end. basically two "tailshafts". Also, the rotors are staggard at 90* apart. This allows for a more continous firing order and also helps with the balancing of the motor. basically the firing order is | (1) | (3) | (2) | (4)|

The fact that this is a custom piece leaves alot of room for errors and alot of need for machining experience. The location of the bearings must also be considered in this design.

I feel that for those will out a high paid design team and a machine shop with the latest in Computer aided maching, this more than likely is NOT a fessable method.


METHOD 2: Dual Firing Rotors (DFR)

This is the "granny speed shop" way of making a four rotor, and more likely a better route to take than the SFR v1 method. Mostly because of the cost differences. In short, you are stacking engines. FF1 and RR1 will fire at the same time while FF2 and RR2 do the same. So basically you are doubling the power strokes that are being put out. The trick is the adapter plate that would connect the two engines togther. the rear rotor must be connected to the from rotors through a coupling of the eccentric shafts. Also, oil and coolant must be able to reach these parts. This can be resolved in several ways including using the stock oil pump in the front section AND rear section. With the rear section being accomadted into the design of the adapater plate. This goes the same with the coolant.

Ducting for coolant and be ran through the plate with an electric pump providing the coolant.

METHOD 3: Staggard Firing Rotors (SFR) Version 2

This is the a combination of Method 1 and 2. The effect of the staggered rotors, with the easier production of coupling engines.Basically the two sections are coupled with the rear section being 90 degrees out of phase. Then with some proper timing you can have the fire order so that each rotor fires individually. and balanced..

What would be the power and revoultion differences between dual firing the rotors and have a continous firing setup?

that and what can everyone add to this? let the ideas flow!

Here is some usefull facts about the mazda rotaries http://www.thecarricos.com/ACRE/ , especially the 4-rotor chapters.

How about welding two rotors and make a BIG 2-rotor 26B? That would be awsome, and you can use a one piece shaft and still be able to assemble it.

That might run, but it wouldnt be very practical. A test was done a few years back to find out wether or not the number of cylinders mattered. Heres the results:
3.0l 1 cylinder- 20hp
3.0l 14cylinder- 400hp

although the effect wouldnt be as drastic on a rotary since there is no "dead" points and the rotor never stops, the test shows that multiple lightweight pistons/rotors can improve power exponentially. This is why formula 1 cars use 3.2l v12's and not huge v8s.

Personally I like the first option, as two 2 rotor engines strapped together are still two 2 rotor engines, not a true 4 rotor.

I can't see why a single bespoke/custom made eccentric can't be made.
It would be considerably more stiff than multiple pieces.

You'd use one end plate, 2 centre housings and one front plate.
Oil pump should be high volume possibly racing unit, though super high pressure would not be required.

Ignition should be direct fire coils run off an ECU.
Firing sequence may not strictly matter.
The standard 2 rotor pattern is similar to a 6 cyl piston engine
So you have a choice of making a very smooth 12-cyl or a double banger 6 cyl.

IanC

Formula 1 has an argument between V8 and V12, with V10 as the compromise largely based on valve area against friction and complexity.
V8's have the lowest cylinder friction, but V12's have greater valve area.


Ianc

Sorry to be off topic but F1 is going to a V8 next year http://www.autocar.co.uk/popups/video.asp?ar=216402 It is a 2.4L V8 turning over 20,000rpms yes that's correct! And they are doing this to try to slow them down and cut cost. I wish F1 still had turbo's allowed oh well wishful thinking.

problem is, you cant assemble the engine with it being one peice.

What other ideas have you thought of , any about other pieces that would have to be fabricated or changed to make it happen?

well i would need to make some sort of coulping to connect each eccentric. Then from here comes the needs for an adapter plate to go between the two sections to house the coupling and to provide the feed for coolant. more than likely it will be made out of a large alumnium piece. I have access to a CNC machine but not a fancy one. so making some insance eccentric is out of the question. but drilling and forming an adapter block is possible. Also, i will be using a TH400 trans becuase i dont think the TII trans will take this much power and not die.

I kinda like the idea of 2 13b's put together @ 90 degrees. The tension bolts would be shorter & may help with flex. Even though it's not a true 4 rotor I think it would still sound like one, be affordable & achievable.

What are your ideas on the coupling.

ive seen how they couple big blocks together and its not that hard. bolting the eccentrics will be less than desireable as the whole of the power would relay on a bolt but if i connect it will a chain in a circle it would hold. Its hard to explain but think of the main pulley being a bike gear and the output shaft from the front section being connected with another eqaully sized gear. now, if they are put together, a thick chain can be put around the gear-a-majig, thus give us coupling of the two motors. its how they do those crazy death 2000hp tractor pulls things. so im sure it can work for this application.

Sounds strong, & it would help absorb some of the shaft flex.

What about end play on the rear engine?

only thing is i might have to make the adapter piece in two parts. becuase in order to get the chain connected and clamped, i would have to get the rear section on while trying to get them to line up, so if i left a bottom accessto get to some of the bottom of the pulley i could clamp the connection. then attach the bottom piece.

the adapter combinded with the coulper should give more than enough rigity to the setup. im going to balance the whole damn assembly before i would do the full go ahead assembly also.

Parts wise for stock things thinking i would need AT LEAST: these are all S4.

NA block. for everything except the intermediate plate.
sell na intermediate plate
TII block. <--- for two extra housing and the intermediate plate. oil pump and whatnot
sell TII rotors
get two addition 9.4 comp rotors with the same weight stamp as the first ones. eccentric from the motor those came from.
another TII intermediate plate.
no OMP. premix.
S4 NA CAS.

also,

Standalone. period. no way around that.
electric water pump.


(going to be a PP setup)

I was looking at doing this, Building up both shafts & turning them back down to 1.690 the dia of the main facing them off to the correct length cutting a key slot & using a clamp type collar for a coupler.

essentailly where is alot of mixing and matching. the production of an adapter plate will the most difficult as the coolant feeds have to line up exactly as does the thickness of the plate in relation to how the couplers meet up. must a VERY accurate. otherwise, play will start to develop.

I think PP is the only way to go , the intake would be much easier to build. I also think a 4 rotor would be more streetable with a pp than a 13b

it should be a VERY potent engine. im going with the TH400 for that reason. and im sure i wont stop at the torque levels ill be getting. and seeing i would have to move the trans anyway, might as well upgrade while i do it

Did you look at the way Granny's did their coolant feeds, they ran them into the top of the center housings. seems like it would cool the rear rotors better.

Especially if you boost it

http://www.kiwi-re.com/index.php You might want to get in contact with this person, Alec Bell, he makes custom 3 & 4 rotor engines. He has custom eccentric shafts and uses the center intermediates to make the engines shorter and help in controlling flex. Engines are PPort motors look around his site and you will see pictures of the 3 rotor engine I didn't find any for a four rotor but there is a guy with one name is Bryce McEwan he has a 4 rotor FD. Car is from New Zealand.

Thanks.. I saw a vidio of that car , It's pretty much what sold me on a 4 rotor

Whats your idea on it ?

looks like they modied the intermediate plates to run the coolant to the top of the plate for the in and out. It said they blocked off thefront cover componets shortened down the assembly. which makes sense. that area needs to be as simple as possible,

Do you think it would help keep the rear rotor cooler?

It would be a tight space.

its more direct so i would presume so. you would almost have to inorder for it to cool anything. run the lines so that the rear gets coolant which is not fresh from the hot front section is smart. seeing by the time it affected the rear, little if any cooling would take place.

What also interests me is that by running it "stacked" so that the Fr1 and RR1 fires at the same time but 180 degrees offset meant the need for two counterweights is eliminated. that might be the smarter way to go. Seeing that space and simplicity will become a important factor. Also, with the cooling, you could then have the adapter plate be much more rigid if there are no cooling ducts. its obviously carbed with the crappy points igniton and it still made 550ish Hp. meaning that using more state of the art methods for the igniton, i would be easy enough to break that barrier. FI is best.

so lets say that Method 2 is altered. so the shafts are 180 out of phase AND fire together to act as natural counterbalances.

Seems to me that the counterweight is used to offset the weight of the rotor some, witch in turn would put less stress on the main bearing. So to me I think it would be a good idea to keep it.

there is a page now about making the coupler. there is a link at the bottom of the page. its interesting. but it seems like the draw bolt wont not allow for the oil jets to function

you retain outside counters but the inside counter would require no counterweights seeing that the rotors themselves act as counterweights.

Yea I saw that, they drilled the center of the shaft 10mm so oil could pass.

Lynn e. Hanover said " If you make the junction ridged, you would go nuts trying to keep the rear crank in it".

I'd like to do a 4 rotor engine some day. I'd prefer one rotor face every 90° so it will sound like a 4 rotor engine. The problem with two rotor faces every 180° is that it would sound like two 13Bs reving up together instead of the awesome 4 rotor sound.

i have to get to class.

They would but only if you used a sollid type coupler like granny's did.

Nothing sounds better.

I am actually 50 % done on a 4 rotor 12a build up, the coupler we are getting machined right now is a 35 spline unit, and then we had the eccentric shafts built up and 35 splined, its actually the same spline as a ford 9 inch rear end 35 spline axle. The couple actualy bolts onto the back engine so you can adjust the end play of each engine seperatly before putting them togethere. we are running the engines in sync for hopes of some good torque multiplication. we are also using an external oil pump and electric water pump, our center connecting plate is 3/4 aluminium. We are only going to run a bridge port but the engine is going to be twin turbo'd at 18 psi, we are hoping to make between 800-900 rwhp on pump gas, which should be possible, and should make the rx3 its going into a fun street/strip car. we hope to have it finished by summer of 2006.

We looked at doing it the way with the chain and gear like a fellow has done before in that tractor puller, but the engine would have to be a great deal longer, and the system of doing it seemed very mickey mouse.

Thanks for the info, the coupler sounds like a genius design. How does it work? do you use some sort of bearing? Iv'e been wondering about end play becouse it's so important. With this setup are you using the counterbalancer?


Thanks again for the info.


Bob

i have a very detailed drawing of it i will try and find a scanner. I only use you front counter weight of the front engine and the rear of the rear engine. i just re-read what i said and may have been a litle missleading we only welded up the rear of the front engines eccentric shaft and cut it down too the 35 spline the front of the rear eccentric shaft is being having a 24 spline of the original diameter, then a bolt is slid through on the large 35 spline and and screws into the front of the rear eccentric shaft you you can still have all the stock bearings in play, you just need to get the front end cut down to accomidate for the lack of the oil pump and counter weight. We are trying to build this as easily and cheaply as possible so we will be able to sell kits

So you are using a draw bolt all the way through the front shaft?

im having a hard time understanding this 35 spline to 24 spline coupling method.

That is a really cool idea. Are you feeding oil to both engines through the front housings? & what about cooling routs?

Hey low impedence, if granny's made 550hp w/carbs , your Pport should be awesome even n/a . Are you thinking about running any boost?

more than likely im going to run this NA until ive worked out my quarks. Make some nice tubular headers. Then more tuning than i care to think about. Looking into individual throttle bodies too.

And your going to put it a 2nd gen, I mesured mine it's going to be a tight fit. I cut out all the radiator mounts & will have to move the trans location back 8'' to 10''.

yeah i think im going 8 inches back into the firewall and the rest upfront

Alot of work but worth it.