A lot of thought has gone into this for piston engines, and I was curious as to why I havent heard a lot about this for the rotary engine. As you all know, the fuel distribution once INSIDE the combustion chamber seems terrible. Unlike the piston engines, where there is a verticle cylinder (to maximize gravity) and port design (to increase tumble and swirl), it seems the rotary manifold promotes velocity, but hinders the mixtures motion. Am I correct to assume that since the manifold design is horizontal, that the bottem of the chamber is rich while the top is lean?
In stock trim, this may not be THAT (read: critical) bad, obviously once the boost is raised and more fuel is applied, this condition radically worsens. How well does the stock intake manifold deal with keeping mixture tumble and velocity while entering the port? How much does the port enterance effect mixture tubulance? When the rotary engine is ported, is this taken into consideration? Are there any provisions made to either smooth the transition into the combustion chamber out, or to better increase mixture atomization?
Are there currnetly any companies who make a tubular intake manifold for either the FC or the FD? It seems the smoother radiused bends and more uniform shape, as well as the ability to have the manifold tuned to a particular RPM range that there is a great potential for not only horsepower to be made, but to add a margin of safety due to better mixture distribution....

Any one out there have any ideas as to where i could find flow bench data on a stock FC or FD (perferabally both) intake manifold, and where I could find information on the custom manifolds?

Thanks for any information!

 

I would think that any changes you do with the intake manifold and port shape in an effort to produce swirl will be negated by the radically changing shape of the combustion chamber. IN any event, the super-high turbulence in the chamber is why rotaries are incredibly knock-resistant. What other production engines with 9.2-9.7:1 compression can comfortably run on 80 octane fuel?

Also note that the stock rotor housings have a pretty good "quench" area. Well, as much as can be done giventhe design limitations. When at TDC, the air/fuel is squished mostly into the bowl area of the rotor.

The stock EFI manfifolds *are* tuned-length, to a degree. The Series V's VDI setup is a multiple-length system, too.

 

"Am I correct to assume that since the manifold design is horizontal, that the bottem of the chamber is rich while the top is lean? "
If the forces upon the air / fuel were gravitational, this may be the case, but with even just an Normally Aspirated engine travelling at around 6-8,000 rpm the inlet charge is well and truly swept along by the seals.

The inlet dynamics and tuning ideals you speak of really pertain to naturally aspirated carburettured engines. EFI atomises fuel far better than any large carb will. Then it gets down to airflow - under boost conditions air is simply pumped through under force, so the niceties of swirl etc don't really count for much, it is simply pushing the air in regardless. In my country, weber style throttlebodies, or "log" inlets are common as they are relying on the turbo to push the air in. The injector in the inlet tract is liberally dousing the air with finely atomised fuel under pressure that is in excess of 60psi, so it should be spreading well through the inlet air mass.

The edge of the port, the transition into the inlet chamber is best left nice and sharp, but the air will be jammed in at better than 12psi so the delicate touches that squeeze power from a tunnel rammed small block or a high reving 4cyl vtec don't really apply.

In an NA rotary engine, tricks like variable length intakes, external injectors and 1 tb per inlet port are used to maximize power and torque. A turbo engine "just" relies on said turbo to boost the air into the inlet.

THe sort of things to pay attention to in a rotary are the cross sectional volume of the inlet and ensuring the smallest cross section is at the port itself (ie there are no other flow restrictions), the butterflyes in the tb and ensuring again there is no undue restrictions and then the placement of the injectors and their angle to the airflow.

For a darn good design of an inlet, you really shouldn't go past the latest model stock inlet manifolds and plenums (or the Jap market only designs, like the JC Cosmo). To optimise them, get an extrude hone or similar done (clean it real well afterwards). With some of the most powerful rotaries in the world running stock based inlets, these would be the obvious choice in a turbo application.

 

I have to disagree with this.

Even in N/A applications, the air is pumped into the engine under force - atmospheric pressure. 14.7psi at sea level. (ignoring pressure drop through the intet tract for the moment) At 12psi boost, that's 12 pounds above atmospheric, or 26.7psi. (also ignoring pressure drop)

I don't think anything magic happens when manifold pressure goes over atmospheric, and a turbo engine will benefit the same as a non-turbo. However, with a turbo you can simply crank up the boost to offset any deficiencies in the intake tract, while N/A folks don't really have that option so optimization is critical.

 

Peejay,
I think I am pretty much on the money or at least in the ballpark - the pressurization of a turbo does overcome minor or theoretical flaws in inlet design, and as you say, if you still lack power, up the boost! I have read somewhere that there are some "dead" airflow areas in the stock inlet tract but of course when you go looking for the reference you can never find it!

2DoritosOnAStick, the Mazda inlet is really very good, and you would be looking around for a while to find anything better... and perhaps paying a lot of money too! There must be a fundamental "ok"-ness about the design as it is basically similar in design through the 2nd gen to current crop of 13b's (both na and T). And didn't Ibarra use his to run 7's....?

"Unlike the piston engines, where there is a verticle cylinder (to maximize gravity) and port design (to increase tumble and swirl), it seems the rotary manifold promotes velocity, but hinders the mixtures motion. "

Not all piston engine cylinders are vertical, eg air cooled VW's, subaru boxers, etc. And not all inlets (carbs or throttle bodies) are vertical, eg side draft carbs, nearly every current efi v8, etc. As far as I know, the injectors spray fuel more finely distributed than a carb could ever do and thus does away with the need for promoting tumble and swirl, then hence the "focus" on getting the air in smoothly. At it's extreme, some race engines have the injectors right in the actual rotor housings (pp's) or the side plates - indeed I think one of the stock engines (was it 12a 6port or 12at?) had them there... so I don't think tumble and swirl was to much of an issue for them!

The transition into the port is very important, but again the cue should be taken from the Mazda factory... the ports are either just less than square (<90 deg) with the side plates (on the part of the port closest to the centre of the engine), or on the inlet manifold side of a ported engine at smaller angles (like 45deg), eg on a BP. But the edge is sharp, no little lips or chamfered edges or anything funky.... well, not intentionally in any engine I've seen!

 

I thought ALL factory injected engines have fuel injectors in the center housing... GSL-SE's and 12AT's do, at any rate! (12A 6P was carbureted)

What would be really cool would be a diesel-style high pressure, timed injection directly into the rotor housing. Say, put a boss directly over the trailing plug location and inject at about 40-50degBTDC, with the goal of most of the fuel going to the bowl in the rotor. You could run extremely lean overall A/Fs without misfiring since the spark plug would be firing a localized area of stoich to slightly rich mixture, and since there would be little, if any, fuel out at the ends of the chamber, you'd see a marked reduction in HC's going out the exhaust without having to resort to side exhaust ports.

I think Mazda has toyed with something similar to this at one point. You could even just use the direct injectors for low-load work (like idling and cruise) and when you need power, conventional port-mounted injectors could come on-line to ensure a more homogenous mixture.

 

hmmm, methinks either PJ has too much time on his hands or is mixing his medication again .

That insult out of the way ( ) I too recall something like this too, was it on the hydrogen powered rotary engine development?

Actually... that has me thinking... run the injector into the rotor housing where a P-port would go... and get the neat new efi/N20 injector set up from NOS... they shoot the N2O into the engine using the injector boss by slipping a fitting into the boss then puting the injector into the fitting, it pumps the N2O around the base of the fuel injector.

Sorry for the thread hijacking 2DoaS... did I make any sense back there?

 

My roomate (ASE_Joe) posted this thread under my name. I had been wondering similar things though, so I appreciate your responses and your theoretical ramblings .

I do wonder why a rotary has to run marginally richer then an "equivelant" piston engine? I've heard of some prototype head designs (designed around economy) that use advanced swirl to allow 25:1 A/F ratios! Now applied to a high performance engine this is not worth much, because outside of a certain "range", too lean will increase heat transfer, decrease chamber velocity just as too rich. But I would like to see a rotary running 12.5:1-13.5:1 A/Fs with no problems. As it is now you have to sit around 11.5-11.8 to be "safe". You say the rotary is knock resistant, which seems true. Does the fact that an apex seal can be destroyed with just one knock have anything to do with the "conservative" tuning of A/F ratios.

As of now my car is running 16+psi in third gear, and 14.5:1 A/F ratios! The car will NOT MOVE, until I cure the boosting and fuel problems. I do appreciate that it has been so forgiving so far though

 

Rotaries can and will run surprisingly well at extremely lean mixtures. (My O2 sensor doesn't even register any voltage when I'm on highway cruise) I used to have lean-out problems at the dragstrip - through the traps the voltmeter was showing .1v and fallling. No misfiring, no bad habits, just not as much power.

Now, when you add a turbo into the mix, yes going lean is a bad idea.