Printable Version of Topic

Posted by: SirAndy Nov 2 2004, 04:47 PM

anyone here know if i can use the MegaSquirt to run just a single injector at a very high pulse, like let's say 32 times per revolution of the crankshaft ...

possible or do i need to build/program my own FI-Brain ?
Andy

Posted by: Mueller Nov 2 2004, 04:53 PM

i think the limiting factor will be how long you need the injector open for each "squirt"

Posted by: SirAndy Nov 2 2004, 04:57 PM

well, let's see, 32 times per revolution, it's probably going to spin up to 10k soooo, i would need 320000 squirts per minute ....

any injectors out there that can handle this ???
Andy

Posted by: Mark Henry Nov 2 2004, 04:59 PM

Yep if you did trick it and you were trying to fire it 32 times per revolution of the crankshaft, on a running engine the injector couldn't keep up.
The injector still has mechanical limitations.

Posted by: SirAndy Nov 2 2004, 05:01 PM

i just though about something ...

if i run 4 injectors on the same intake i could reduce the squirts per injector to 80000 per minute and get the same effect ...

like this ...


Attached image(s)

Posted by: SirAndy Nov 2 2004, 05:02 PM

so, what's the max. squirts per minute for a good injector ????

anyone any ideas?
Andy

Posted by: Mueller Nov 2 2004, 05:03 PM

is this for the engine design you and discussed a few years ago?? I was actually thinking about it lst week while sitting in traffic...I still don't have a decent combustion chamber idea

Posted by: SirAndy Nov 2 2004, 05:04 PM

yes. i think i might be onto something ...

i would love to get a prototype running!
Andy

Posted by: lapuwali Nov 2 2004, 05:28 PM

Most injectors have a minimum cycle type (fully closed to fully open to fully closed again) in the 2-3 millisecond range. So, you're looking at 300-500/sec as a maximum. This also doesn't include the time it has to deliver the fuel, which can be lengthy at most common fuel pressures (several more milliseconds). A 10K rpm four-stroke twin (say, a racing bike) is about the worst case for these things, and that's only 40 injection events per second. Even such engines often use two injectors per cylinder just to get enough fuel in the thing in the alloted time.

Figure 50/sec is pushing the envelope, and design from there.

Posted by: SirAndy Nov 2 2004, 05:32 PM

ok, so i found injectors that can be run up to 50k rpm!

that means if i run 8 injectors on the intake i could have it spin up to 10k rpm and run only 40k squirts per minute per injector!


can MegaSquirt handle 8 injectors ???
Andy

Posted by: SirAndy Nov 2 2004, 05:35 PM

eewh, that wouldn't work. i need *MUCH* more than that ...

Andy

hmmmm, i might be able to run them at 100% (wide open all the time) to get a constant flow as i would have 32 power "strokes" per revolution ...

Posted by: Dr Evil Nov 2 2004, 05:47 PM

MS handles 8 injectors at up to 10K RPM, IIRC. Why not just do a K-jet type of approach? CIS?
HTH

Posted by: Andyrew Nov 2 2004, 06:05 PM

What would this be for andy?

I dont get it....

Posted by: SirAndy Nov 2 2004, 06:09 PM

i'am designing my own "rotary" engine that has absolutly nothing in common with traditional rotary engines. completely different approach.

single intake, 32 power "strokes" per revolution, high compression, high rpm (probably in the 10k range) ...

need to figure out how to feed the beast ...
Andy

Posted by: Aaron Cox Nov 2 2004, 06:15 PM

crazy german!

im intrigued!

Posted by: physician Nov 2 2004, 06:39 PM

how many rotors? intake ports?

injectors work with duty cycle. and you may have problems with injecting the right quantity of fuel in the time allowed. i add the fact that in that time and speed the fuel may not atomise corectly to burn right. in bikes, they put injectors before the throttle body and sometimes, injectors shoot away from the valve( to the throttle body) so it got enough time to atomise corectly. you may consider thism don't know if it will help you

i want more info on your project, look interesting to me... it's enternaing until i finally put my gsxr 954 itb and tune them on the race civic... but that's another story here...

Posted by: Jeff Bowlsby Nov 2 2004, 06:40 PM

It sounds like a full time duty cycle. The PW would approach zero...so why not just run an injector full time...(CIS?) and vary pressure.

Posted by: SirAndy Nov 2 2004, 06:44 PM

2 rotors, but they both share the "power strokes" (don't know how else to call them, they ain't strokes, but ...), in my example each rotor get's 16 power strokes per revolution, so both rotors deliver equal power.

one intake port, one exhaust port ...

hmmmm, maybe i could just run a weber carb for the prototype ...
Andy

Posted by: lapuwali Nov 2 2004, 06:49 PM

I'm in agreement. You can regulate pressure very accurately. Pulsed injection doesn't sound like a good fit with this application, unless you inject the fuel well away from the ports, and don't bother to time it (thus, providing something like continuous injection, but varying fuel flow using pulses instead of pressure).

I'd be interested to see a basic sketch of the design. If it's really not been done before, I'd salute you. Nearly everything imaginable has been tried over the past 100 years, 90% of it in the first 30 years.

Posted by: Andyrew Nov 2 2004, 07:12 PM

what about running 2 or 4 megasquirt systems?

just a thought...

Posted by: SirAndy Nov 2 2004, 07:20 PM

the only thing it has in common with a wankel engine is the fact that something is rotating ...

no valves, no pistons, no crank, no camshaft, no pushrods

the force put on the rotors goes straight to the transmission ...
obviously, at this point, i'm not willing to share the design, altough Mueller & Brad have seen it. i'd like to get a prototype running first ...

Andy

PS: yes, a constant flow of air/fuel seems the way to go, especially because i will have 32 compress/spark/detonate/exhaust per revolution.
it should even be possible to "force" the air/fuel in there. think Kompressor!

Posted by: TimT Nov 2 2004, 07:24 PM

Andy, you want to run injectors at a max of 85% duty cycle. You may have to double,triple or more the number of injectors. Since there will be so many firing events in the engine, batch fire is suitable.

You could just make a few more injector drivers, make them in seperate boxes if need be. ie piggy back the exisitng drivers

Posted by: physician Nov 2 2004, 07:25 PM

i can't wait to learn more about it! keep on the good work and plz keep us update!!

to your ingeniosity!

Posted by: Andyrew Nov 2 2004, 08:24 PM

Giant turbo with big fins where the combustion pushes the turbine wheel to spin thus you need many firing per second which is what your concerned about....

Am I close?

Posted by: Mueller Nov 2 2004, 08:52 PM

you've described a very basic turbine, jet engine.....

if you have any extra turbochargers, you can build one yourself

Posted by: Andyrew Nov 2 2004, 09:34 PM

Mike, I know I have.. lol

I've got a few turbo's... ya...

Hmmm

Posted by: Aaron Cox Nov 2 2004, 09:41 PM

my unc has a turbine race car.... insane power!

Posted by: lapuwali Nov 2 2004, 09:42 PM

Yup. The "turboprop" engine, used on small airliners and on many helicopters, uses a turbine to directly drive a shaft to turn a prop/rotor. Turbines have been tried in cars, with indifferent success. Part of the problem is horrible throttle lag. On the Rover/BRM turbine car raced at Le Mans in '62 or thereabouts, throttle lag was such that there were several seconds between pushing on the throttle and getting a response from the engine. You had to start "accelerating" while still braking for the turn. On a fast course like Le Mans, this was doable. On a tight course, forget it. A turbine was also raced at Indy at least once, where it was much better suited.

A much better approach is to use a turbine to drive an on-board generator so you have a gas-powered electric. Volvo made a concept car driven this way a few years ago.




Well, if you won't let me do it, do a careful search before you get all interested in doing this in some way that will make money. Prior art does not have to have been patented for it to invalidate a patent application. So many wacky ideas have been tried, that there's an excellent chance yours has also been tried. No disrespect, but there's not much point in keeping it secret if you can't make a buck on the idea, and you can't reasonably make a buck on it without a patent.

Posted by: SirAndy Nov 2 2004, 10:12 PM

at this point it's not about the money. i just want to proof to myself that my idea is going to work.

we'll talk money and patents after i blew up someone elses garage ...
Andy

Posted by: ajracer Nov 3 2004, 09:56 PM

Sir Andy Interesting Post:

I can't say for sure that I really understand what you are
trying to design other than an engine with 32 ? power
strokes per revolution ? and something that will rev to
10,000 rpm. (I like the 10K part).

I do know of a guy in Toronto area that has fabricated
a custom intake/valve setup than he runs on a rotary
engine. It is very unique and extremely powerful setup.
I understand he has also been involved behind the scenes
with providing experimental info and designs the Mazda
of his intake system. (I believe he uses Electromotive
FI system that is fully programmable). If you want more
info please PM for info. All the Best, I must get back to
work on my LT1 Chev V8 conversion as winter is starting
to come to Toronto.

Allan