winsteadn

I have a 12a motor and I was wondering how if possible to direct inject this motor

arghx

iTrader: (3)

No.

Kenku

Sure it is. You just have to make a belt driven high pressure direct injection pump, machine the rotor housings for direct injectors, machine a high pressure fuel rail with proper sensors, cast new rotors and machine them to have an appropriate combustion chamber shape (there's no research I'm aware of as to what that shape would look like for a rotary) and either buy an ECU that can handle direct injection strategies (a Bosch MS 6.3 should do it) or make one, and then tune it.

I mean it's completely and totally unproven ground and you're looking at 10k USD to do even a bad job of it, but it's possible.

BLUE TII

iTrader: (9)

There is data available for direct injection on the rotary.
Starting back in 1959 Daimler-Benz was developing rotary engines since Felix Wankel had ties to them from at least 1934 relating to war efforts.

Mercedes experimented heavily with direct injection on their rotary engines culminating in the C111 three and four rotor engines in the early '70s. Those engines had the injector nozzles in the intake stroke above the peripheral intake ports (near where Mazda puts it rotor housing oil injectors).

Curtis Wright started rotary engine development as well in 1959 and they used direct injection (with gasoline, diesel, and jet fuel/kerosene) with the injector located at the spark plug area (as Mazda would try with their own direct injection experimental engines in the 80s? with and without a pre-combustion chamber). Curtis Wright ended up putting a rotary engine in the RC2-60 Mustang in 1965.

If you buy the 1971 Norbye "The Wankel Engine" book you can get a lot of information on rotary engine direct injection and other developments.

Kenku

I stand corrected. Not knowing the details of their efforts though, I suspect that modern optimization techniques might produce a differently shaped combustion chamber (maybe not!) and further suspect that 12A rotors aren't shaped anywhere near optimally.

So there's ones less barrier to doing a direct injected 12A.

BLUE TII

iTrader: (9)

The direct injection stratified charge rotary engines I have seen favored Leading Deep Recess rotors- so that is available on the 12A platform.

Mazda did patent a new funky rotor tub shape for direct injection fairly recently.
Most of what Mazda does with rotaries is to meet emissions standards whereas the '60s direct injection work was mainly power and fuel economy since the rotary was already ahead of the piston engine on emissions, there were no emissions laws yet and much of the research was for aeronautics/military where there still are no emissions laws.

Emissions laws are pretty much what killed lean burn stratified charge engines as high NOx emissions result from the border of the rich pocket to the lean main body of the intake charge.

j9fd3s

iTrader: (3)

Mercedes has a long history of running direct injection, the Gullwing is direct injected, as were the aero engines, and the 30's grand prix cars. its mechanical, of course, but they were making pretty big power even now.

https://en.wikipedia.org/wiki/Daimler-Benz_DB_603

peejay

I have long thought about using a belt driven TDI injection pump on a rotary. It's an electronically controlled mechanical pump.

My main concern with the idea is if gasoline, even with 50:1 premix, has enough lubricity to keep the pump alive. With that question unanswered and possibly unanswerable with the resources available to me, I have more pressing fish to fry.

That said, I don't see why you couldn't machine the OMP bungs out a little bit and put fuel injectors there. With timed injection, that location should never see pressure, so you wouldn't really need a super high pressure fuel system. You'd just have to have injectors that could flow enough fuel at a max of 50% duty cycle or so, since you'd want the injectors to only flow between the point after the exhaust port is shut and before the intake port is shut, when compression begins. For an N/A, that is about 2000cc/min, which is doable nowadays. There might be some weirdnesses in the flow as injection vs. rotor position changes but this shouldn't be an issue if you use a speed-density algorithm instead of MAF, you'd just bake-in the varying fuel requirements and the VE table wouldn't be a true VE table but more of a fuel-demand table.

Or you could have smaller rotor housing injectors for low load/part throttle, and upstream injectors for high load use when reversion is less of an issue. But something tells me that ultra low duty cycle at part throttle would actually be a benefit here and not a loss. Maybe.

Kenku

In all seriousness, I don't see why you start with a junkyard GDI engine, lop off part of the cam, make a sealing plate for the pump that's bolted to the head, lathe down the camshaft stub that contains the GDI pump cam to accept a pulley, and so on. Then you have an electronically controlled pump intended for gasoline.

j9fd3s

iTrader: (3)

9,9,9,9,9! like direct injection, using a diesel injection pump on a gas engine has been done, every Alfa Spider from 1969 on (in the US, the rest of the world switches over a little later), runs a Spica mechanical fuel injection, which as you may have guessed is a diesel injector pump on a gas engine. apparently they are bulletproof.

item #2 in the 'everything has been done' department, Alfa Romeo also had the first variable cam timing system in 1980.

peejay

Because I don't have a machine shop available to my disposal?

And then you have to deal with the electronics. Currently, the only aftermarket controllers that will handle direct injection are SILLY expensive. DI is a whole nother ball of wax compared to port injection. Wer'e talking hundreds of volts to open the injectors, and they use the injector close ringback spike to open the next injector in sequence!

That is why I liked the TDI setup... it's self contained and it doesn't need anything exotic in the way of electronics to control it. Which is probably why VW/Bosch engineered it that way as well.

peejay

But SPICA and Kugelfischer injection pumps are a little difficult to come by... especially since so many people scrapped them in the 1970s and converted to Webers.

BLUE TII

iTrader: (9)

I think John Huijben on this forum said there was an OE box he used that took regular injector ecu output as the input and 12VDC and spat out the DI output when he built the DI rotary for his local Uni.

John put the DI injectors at the 12 O'clock position on the rotor housings like current Mazda DI rotary models.

I have always been interested in the super lean burn stratified charge rotary with precombustion chamber where you screw a combustion chamber with spark plug and injector into the Trailing spark plug hole for low load operation like Mazda did (but it gave them like 35% more NOx at low loads to deal with over regular rotary).

peejay

The current DI trend is blended port and DI injection. There are areas in the load map where DI causes excessive particulate emissions (soot), so they are running port injectors in those areas to suit. It's cheap to free at this point, since DI still needs a low (<60psi) pressure fuel system, all the added expense is a rail or two and a handful of injectors.

j9fd3s

iTrader: (3)

the SPICA pump was from a diesel, and Alfa used it for the gas car

Kenku

Well, as mentioned, there's some boxes that'll do it. I was actually thinking of doing a Duratec dyno mule to try to get port plus DI working off a megasquirt. Haven't looked too far into it, simply because I have plenty of other stuff to prove out on the dyno first... like getting the dyno itself working!

God I have too many projects...

TonyD89

iTrader: (10)

If I understand correctly, DI only causes nasty build-up on the back of the valves so they use combined port and DI to reduce the build up. Since rotaries have no valves, maybe it's a better candidate for DI. There may also be more time for fuel injection because the rotor is moving at 1/3 the speed of the eccentric shaft. Just some thoughts.

peejay

From the word of a powertrain engineer involved with DI cars, combined port/DI is to reduce soot and NOx because there are some load regions were DI is dirtier than port injection.

Unsaid by him, but inferred by me, intake coking is not a concern at the OEM level because it doesn't cause an emissions related failure. It also tends to be a problem only with certain driving styles.