Has anyone built a miller cycle rotary?
 

I was reading into Mazda's miller cycle engine that they use in the millenia, and I wondered if anyone had built a rotary like that. It seems like it would be way easier to try out on a rotary than a piston engine since you can grind your valve timing into the engine rather than making some sort of crazy custom camshaft.

I'm thinking about trying it with my old 12a, it seems like it could add a little power and improve my gas mileage at the same time.

Does anyone else have some thoughts on this?

http://www.mazda.com.au/articleZone....ticleZoneID=92

-Matt

"Has anyone built a miller cycle rotary?"

Yeah, Mazda has.

I couldn't find anything online about it, but I did read it in one of my books. It was a small caption and not an indepth article under a small mechanical drawing. It was a peri port using a reed valve. It had standard side intake ports, but they were way way high, nearly in the middle of the side plates. Im sure there was some sort of valve that regulated the reversion from one rotor to the next through these ports. It also had a direct injection fuel injector on the compression side above the spark plugs. Thats all I know about it. I'm sure if one searched enough they could get SAE papers along with drawings.

~Mike..............

funny you should mention that. i got ahold of an eaton m90 that i was going to use for that very idea.

Technically all the 6-port NA rotaries are a part time Miller cycle :)

They use the typical Miller cycle 30 deg delayed intake closing timing when the 6-ports are open and generate 100mm/Hg boost from the DEI effect intake.

If you want to do full time Miller cycle with a supercharger for better cruise efficiency I would think you would want the most efficient supercharger available that could make the boost at cruise and the lowest compression rotors available.

You could possibly use a VGT turbo (turbo being the most efficient conventional compressor)- though with the high p-port exhaust overlap the exhaust restriction would probably cause its own inefficiencies- save that experiment for the low overlap renisis side exhaust.

Next most efficient compressor would be the Lysholm SC like Mazda used on the Millenia S.

A very well funtioning intercooler set-up is also a must otherwise it is not much more efficient than letting the rotor compress the mixture with its compression ratio.

If you are using the early p-port exhaust you also want a way to inject fuel so it is not blown out the exhaust during the overlap phase.

This could simply be timed sequential injection of the primary injectors so they inject late enough in the intake stroke- this would work best with semi direct injection (the primary inj. moved to where the oil injectors are on the rotor housings). Mazda has done this semi direct injection location.

It would work best with direct injection into the combustion chamber or injection into a precombustion chamber that screwed into the leading spark plug hole (this would also help lower compression ratio for more efficient Miller cycle). Mazda has done these as well.

So it would actually be pretty complex to get the full benifits of Miller cycle on a p-port exhaust rotary and slightly easier with a Renisis sideport exhaust rotary.

Without addressing the low low compression conventionally used in a Miller Cycle, the loss of fuel in overlap phase and the most efficient compressor available I don't think you could possibly gain back the efficiencies lost driving the compressor.

You could address these issues most easily with a 6-port Renisis with a Lysholm SC or VGT turbo using 8.56:1 CR S4 TII rotors and a pre-combustion chamber further lowering the compression and allowing direct injection of fuel as well as a very lean localized burn. Replace your radiator with a giant IC and mount the radiator in the trunk.

That *might* work as well as Mazda's Millenia S Miller Cycle set-up.

Now think how small the gains were on that engine VS the conventional Millenia Otto Cycle engine...

Damn, it was sounding really good to me too untill I thought of that last bit :(

the miller doesn't have to have much overlap. you could always move the ports to open/close later. they don't even have to be very big to make a difference.

It seems like all you would have to do is make the intake close so late that without any pressure from the supercharger the engine can only idle. I don't see why the exhaust/intake overlap is suddenly a problem for trying the miller cycle, since 12a's are turbo'ed and blown all the time.

Also, I don't think the miller cycle requires a lower compression rotor, only at part throttle conditions the actual compression ratio is lowed, which is balanced out by not creating any vacuum. I think you would still want the highest compression rotors you could get.

Matt

so what the millenia S is the vechile in question correct?

the miller doesn't have to have much overlap. you could always move the ports to open/close later. they don't even have to be very big to make a difference.

Well, there is now way to port the exhaust and intake smaller than stock and the stock P-port exhaust rotaries have lots of overlap compared to a piston Miller Cycle engine. If you are talking about custom housings, why not just start with the low overlap Renisis?

It seems like all you would have to do is make the intake close so late that without any pressure from the supercharger the engine can only idle. I don't see why the exhaust/intake overlap is suddenly a problem for trying the miller cycle, since 12a's are turbo'ed and blown all the time.

The overlap is only a problem if you want it Miller Cycle for efficiency and not just Miller Cycle in name. The Miller Cycle cannot even idle properly without boost from the supercharger and that boost will blow your air and fuel out the exhaust port during the overlap phase thus hurting fuel efficiency.

As I noted, this can be solved by injecting fuel only once overlap phase is over, and that gets complex.


Also, I don't think the miller cycle requires a lower compression rotor, only at part throttle conditions the actual compression ratio is lowed, which is balanced out by not creating any vacuum. I think you would still want the highest compression rotors you could get.


By definition the Miller Cycle requires a lower compression piston/rotor than Otto Cycle.

The premise of the Miller Cycle is that a more efficient supercharger partially compresses the mixture so the piston/rotor doesn't have to because pistons/rotors are very inefficient at compressing (add more heat).

Yes, you can wire open your 6-ports and slap a supercharger on for a smooth idle and call it Miller Cycle, but it is so primitive a Miller Cycle you will lose efficiency from the parasitic drag of the SC.

By comparison a stock 6-port is a very elegant example of part time Miller Cycle because it uses ultra efficient pressure wave supercharging (DEI intake).

i meant you can move the pp intake port to wherever you want it. you don't hve to open it in the same location as everyone else. open it later and it's all good.

i meant you can move the pp intake port to wherever you want it. you don't hve to open it in the same location as everyone else. open it later and it's all good.

True, though you can see it would have to be very late opening indeed for a P-port intake to eliminate overlap with the p-port exhaust.

With a rotor/housing/shaft mocked up mark where the apex seal closes the exhaust port and look up clockwise to the next apex seal. This is where your intake port would have to begin open to eliminate overlap.

Look at the volume of the intake chamber at this point- not only have you not been drawing air/fuel for all that volume, you would have indeed been drawing in exhaust gasses into the intake chamber up to this point...

I therefore believe this late of an intake opening would seem to add incredible inefficiencies on its own and is in fact no part of the Miller Cycle.

The cheap easy way to me just seemed to be the Renisis housings since they already have eliminated the overlap of the traditional rotary engine.

i didn't say to get rid of the side ports, only add a pp.

working at a dealership forever, the miller engine is kind of a mixed bag. it does make more power than the 2.5 KL motor, but not a lot. and the packaging is more complex than the fd, and since the supercharger needs to come off for nearly everything, repair costs are astronomical.

so basically they run great for 60-90k miles then its $2000 (at least) to fix it, and they throw the car away

so that's different than any rx7 in what way?

I was wrong here- the Miller Cycle engine IS lower compression but the rotor/piston does not have to be lower compression than Otto Cycle, just as you stated!

The compression ratio determined by the swept volume Vs combustion chamber volume is same or higher than Otto Cycle it is the later closing valve/port that lowers the actual compression ratio. Example Millenia S has 10:1 CR that is effectively 8:1 CR due to the late closing valve.


i didn't say to get rid of the side ports, only add a pp.

Right, I was comparing the overlap of a factory sideport intake peripheral port exhaust rotary to the Miller. We get 16 deg of overlap stock and the Millenia S has 7 deg overlap and timed sequential port fuel injection.

But, after thinking about that- if you used a later S5 primary port w/ 40 deg primary port opening cutting overlap to 8 deg (while 2ndary ports are held shut) and timed sequential injection (don't know if S5 ECU did this) I guess you would actually be pretty damn close!

So, I was being overly pessimistic on this project. I say give it a go!

I would still advocate the most efficient SC (like Lysholm) since the SC will use 10-15 % engine power to turn it and you have to recover this loss with the extra efficiency the SC has compressing the mixture over the rotor/piston. I believe Lysholm is still ~10- 15% more efficient than the best Roots compressors. That and a really awsome IC set-up - I like the idea of the IC up front and radiator in the trunk :)

how about the cosmo primaries? they open the latest of any i know of.

lysholm/axial flow is out of the question right now. roots is the only thing that'll do it low enough.

people fix the rx7.... and the sunvisor isnt $400....

touche :p:

the 929 and millenia are MORE expansive than the fd's....