16X, and direct injection
 

I was drinking with a very qualified machinist friend tonight who knows Wankel's pretty well.

We where thinking of attempting to install injector bosses for direct injection in a set of spare housings.

What both of us noticed in the promo pictures of the 16X is that the injectors inject into the top of the rotor housing, after the housing passes the intake port but before the rotor starts compressing the mixture.

So in other words, we think that injector bosses in this location would not require ultra hight pressure injectors like those used by diesels or GDI vehicles such as the mazdaspeed 3 and could use standard injectors.

His thinking is to install 2 injector bosses in each housing, and run injectors large enough and fuel pressure high enough such that the injectors can stay fully sequential.

He is going to lay out some parts and figure out how many degrees of rotation would be available and where exactly the compression stroke starts to see if this is even possible.

He built a set of housings in the early 90's with an extra set of spark plugs, so he' says adding injector bosses to an area with no compression would be easy.

Anyone ever try this, know someone who has or have any links that would help.

There is no commercial intention in this endeavor and anything learned will be shared freely with all.

i have no knowledge of this area, but it is something that i thought of before

if it works, you should make a documentation of it, so others can try it out.

good luck!

While I appreciate, as I'm sure the rest of the community does, the lack of commercial intention, I'd approach the project with mass-production/recreation in mind.

A friend of mine and I have been considering doing this for few years now.... Haven't yet but I think it is a very good idea!

Though I'm very curious as to why Mazda didn't put this into a production rotary by now? They tested this over 20 years ago with pretty good results.

Those engineers now what they are doing, so it just makes me think they found something "wrong" with this system on a rotary?

I still think we'll end up trying it though!

-J

sounds very cool. if you ever get it working, i'd love to stop in and see it.

What they did in those early tests was different, they where injecting into the compression stroke like normal direct injection, and for most of the tests using no throttle body, but using fuel to control engine speed.

Their main issues where cost of components and smog. The direct injection that they tested resulted in exhaust temperatures to low to heat the cat converts up to light off at idle and low load.

huhh (shrug) gotta love mazda, found somthing that worked and then got there heads stuck in there butts again and backed out of it.

That's is real direct injection but I like the idea Mazda uses in the 16x. In all reality this could also be done on a piston engine (injecting at BDC) so there must be a reason they are pursuing the other method.

Like you said, it's pollution control. The advantage of injecting after compression is no fuel get trapped in the corners of the combustion chamber. More complete burning of the fuel.

I wonder if atomization of a normal injector would be good enough so close to the compression/ignition event?

I imagine they are injecting right into the pocket of the rotor. It probably minimizes the amount of fuel collected twords the trailing tip of the rotor.

On a 13B with peripheral exhaust ports, direct injection with enough fuel would basically eliminate all fuel lost to port overlap.

You could in theory make a PP engine or big BP get better gas mileage than a stock port, with nearly as good low speed performance.

this is some pretty neat stuff. i'd love to see the results of something like this in practice.


subscribed. :)

How come everybody thinks that because the OEM didn't mass produce a design that they found a critical design flaw with it??? !!! Car manufacturers are interested in making MONEY. Piston engine manufacturers have toyed with direct injection in spark ignition engines for the last 70 years or more all with good results. So why isn't every piston engine on the planet running with it? The main reason is it is much more expensive to produce even in a rotary.

I have to cut this short, time to go to work.

However, unlike piston engines, I think that Mazda KNOWS that the future viability of the rotary engine hinges on the improved fuel economy and lower emissions that this technology will provide. Although direct injection helps piston power, it's not nearly as critical to the go-forward viability of the technology as it is in the rotary engine (IMO).

made some progress here. We are trying to determine where to place the injectors, so as to begin and end injecting into the face of the rotor just before compression begins.

To provide the maximum advantage it has to be as high up as possible without getting into the compression stroke.

If the injector has to inject into the compression stroke, basically at all then precise fuel control is going to be hard with standard pressure injectors, because its not possible to predict the pressure on the other side of the injector.

As it is this is going to be really hard to accomplish because, there is such a limited time to inject the fuel, maybe at best the injectors will be limited to 25% duty cycle.

On one of the better ECU's that can control big injectors and provide a way to tune in limits to the primaries it should work.

To figure this out we have mocked up some parts and will be trying a few things out.

Subscribed.
I love new things and this one should be pretty easy to DIY. :bigthumb:

Ok, I just wrote this big long post of other things you are going to have to take into account and since I wrote it all from an engineering point I don't think it meant dick, so I'll try to write it a different way in which it will be useful. And it's been a while since I've talked to anybody who has actual experience with direct injection so some of the specific details might be fuzzy.

1. Larger injectors require more pressure to properly atomize the fuel.
2. In an intake runner high velocity helps to shear the fuel once it is injected which also helps to atomize it.
3. There will be a threshold RPM at which the system you design won't have enough time to inject all the fuel needed in this "uncompressed" area.

I'm not sure what kind of shearing forces inside the chamber would be during the intake stroke in a rotary but I would guess they would be less than in the intake manifold until you start hitting the compression stroke. Also faster moving air drops pressure on a surface (think airplane wings) so the faster air inside the manifold helps reduce the amount of fuel pressure needed. This will mean right off the bat you will probably need more pressure than what the injector would need inside the intake manifold.

For the problem with injecting enough fuel at higher RPMs if you solve that with larger injectors you will need more pressure, If you solve it by adding another injector you will need to increase you're volume, if you can't get either of these to work then you will either have to start injecting the fuel earlier and/or later. As you inject the fuel earlier you will loose, or lessen some of the benefits of having it direct injected. As you inject the fuel later, you will increase the benefits but will require much higher fuel pressure to do the job.

right, the duration available to inject fuel will be fairly limited to about 25-30 deg (maybe less), and can't go beyond that, the rotor won't be there to prevent fuel from flowing in the way it blocks the side port and does just that in rotary with port injection.

Any fuel injected early or late will flow out of the exhaust, not simply lose the benefit of direct injection.

Also why large injectors are required.

The fuel will have to be controlled with excellent precision within this window, and limited to this window.

interesting...

Very interesting, I love watching ideas like this unfold!
What engine setup are you planning to use for the prototype?

I was thinking that running a leading and trailing injector might be more efficient then one large one. It would be similar to the spark setup, but would run a 0* split. I think the increased spray area would provide better atomization.

The hardest part still seems like it would be engine management. The stock CAS should be precise enough as long as you run a high resolution EMS (Megasquirt II, etc.), but additional software needs to be written.

I don't think megasquirt II is a sequential injection controller.