Mechanical Fuel Injection, Because I Hate Electrons

tuna55 wrote: OK guys, I am very glad to see that each of you can pee very very far, very impressive. Now. EFI is off the table. It just is. I'd tell you all why, but it is one of those "don't want to get your hopes up" kind of things. It's to satisfy a rule book. Figure 2 liter engine at 36-40 psi on race gas. Looking for something like 600 hp. Streetable on occasion would be a great goal. The ability to start with a key and idle for a few minutes without fouling plugs or overheating would be a huge benefit. We have looked at these options put forth thus far and I see $$$$ -> science experiment, but nothing really in between. Right now the best we can figure is a diesel V8 pickup setup with two injectors per cylinder. It's not a challenge budget, but it's not "How can I spend 15K for a box of parts??" either.

That's a lot of flow for any system. More so, for one that needs to be flexible.

one big issue for the diesel system (outside of the fact that it will have problems lasting any decent amount of time) is timing- a diesel will be sequential, whereas on this 4 cyl it won't be. I'm sure you'll manage to make ti work, but....

Lucas and Hilborn will be the best bets. Making a diesel work will require a lot of engineering to make the input position relate to air flow in some kind of way- mechanical ones are mostly just a throttle input, since there's no requirement to be within an a/f range for it to burn.

Well, let me correct that. The PUMP will probably work. The logic part is where you'll need to be super extra creative.

Can you remind why a carb won't be good enough? It will be a lot better for probably 3-5 years of mechanical pump development- if it's just a small team, one engine, do real dyno development, access to a good machine shop, and someone familiar with mechanical adding machines.

Is the lack of lubricity of gasoline a concern for trying to use diesel injection bits?

In reply to ransom:

(it would be a concern of mine, but Marvel Mystery Oil can be used to help)

Oh, one more diesel concern- injeciton pressure- not sure how well the pump runs when you run say 150psi (what SPICA's are) injectors vs. the well over 1000psi diesel pump is used to running.... may make calibrating more challenging.

(edit- I have found out that Bosch menchanical injector run at 39psi, BTW, if that helps)

http://www.kinsler.com/page--Home-Page--13.html

You can get a Lucas or a constant flow system. may be easier to start with.

alfadriver wrote: Can you remind why a carb won't be good enough? It will be a lot better for probably 3-5 years of mechanical pump development- if it's just a small team, one engine, do real dyno development, access to a good machine shop, and someone familiar with mechanical adding machines.

I have not explored this, but I have always understood that carbs were boost limited and unable to be intercooled. I haven;t ever really understood why, however, and if you can shed some light on this, that would be great.

In reply to tuna55:

Pre-turbo carbs will have an issue of the fuel condensing in the intercooler- just like it can condense water.

Post turbo carbs need to be sealed to the intake pressure- not an easy thing to do, but you can build a box around them so that they reference manifold pressure vs. just sealing them up. someone posted a picture of that before.

Not sure about a boost limit, but WWII planes used to run lots of boost even with carbs, so it is possible. (oh, and some pre and post was GP cars did as well- iirc, the 158 and 159 Alfettas used to run a LOT of boost with 1500cc)

I'm not a carb guy- two of my cars should be carburated, but they are in pieces, two others are SPICA, and all the rest are modern. (and I know a thing or two about EFI, but I don't want to cross streams... ) I'm not the best person to answer carb questions.

But the pressures you guys intend to run- outside of drag racing and USAC, that kind of boost really didn't catch on until there was some kind of electrical injection available. I don't even remember the Offy's running that much boost- but they ran a constant flow system that was a MAJOR PITA to get restarted.

Basically, if you run binary, you can get a mechanical injection work. it's all that stuff inbetween that's hard.

The only way you can intercool a carb, and thus run more than about 1/2 bar through the turbo, is if you put the carb in a sealed box and use filled floats, so the positive pressure from running the blow-through turbo won't crush them.

The sealed box makes the whole carb "see" whatever pressure you're blowing at it with the turbo, so it can function properly. As you know, carbs only work with vacuum, or at least a referenced vacuum relative to everything else.

If you try to run a suck-through system with a carb, and then intercool it, you will end up with gas condensing on the inside of the intercooler. Obviously, this is bad, and very very very very very unsafe.

mguar wrote: In reply to alfadriver: The reason for the hard start on high boost Offy's had everything to do with making ultimate power at a minimum weight. (I'll be glad to go into the details at depth if you wish)..

Uh, that would be my point. To make a fuel system flexible enough to run at a high boost (which is the intent of the project), AND do anything else is very hard.

your example of the D type is bad- it's not boosted. Boost requires mechanical logic that is not easy, IF you wish to be capable of running basically anywhere else. You could also quote the Mercedes SL, or the Vette, or a number of mechanical injected, but not boosted, cars as examples- and they all miss the detail of boost, let alone intercooling on top of that.

I do understand fuel injection. I also undersand boost, and the fueling logic that is required to run boost.

Latest issue of Car Craft has a Maverick powered by a 5.3 liter LS-type GM truck engine with a carb and a turbo and a hat, not a pressure sealed box. I know Car Craft has written about this sort of setup before and the cars seem streetable and they can put down serious amounts of power.

I'm not sure why the ducting from the turbo to a carb couldn't go through an air to air intercooler first...

Hillborn = WOT and nothing else CIS = Everything but WOT

Pick one.

Since you plan to run boost, does it really need to be port or sequential?

Why not start simple with a single ( or more if needed for flow rate) injector set up in a TBI style installation. It would solve the pressure problem and eliminate the timing issues.

I would think a throttle body drilled to accept some poppit style injectors from a GM SFI system with a flow valve linked to the throttle linkage could be a starting point, and a pressure regulator to help with drivability.

Once you get the simple system running then you can expand it if needed to get better mixture balance on all cylinders.

This:

and a lot of index finger exercise.

(I have nothing to add except stupidity. )

tuna55 wrote: EFI is off the table. It just is. Figure 2 liter engine at 36-40 psi on race gas. Looking for something like 600 hp.

Sounds like a straightline terminal velocity ruleset/build. I would run it with whatever comes along that is cheapest and most fool proof (read carb with hat) at low boost and get some time on the track(s) before upping boost. No one breaks records first time out.

Even if it's not a Vmax event you're likely to have have way more problems with traction than power if a 2L 600hp motor is the end goal. Good chassis with bulletproof 300hp is always better to start than bad chassis with twice as much hp on the brink of molten chunks.

mguar wrote: Finally Carbs cannot be Tuned to richen at certain Boost pressures like Injection can.

Wrong.

Sorry but I've got a factory turbocharged Pontiac in the garage that will argue with you. Same with the carbed turbo Buicks and all of Holley's carburetors that are made with a boost-referenced power valve.

The enrichment circuit is plumbed externally in these carbs so that it can see manifold pressure. Rochester and Holley both made it work.

mguar wrote: In reply to Paul_VR6: Not true.. You need to spend some time with Kinslers catalog..

.. and there are still compromises over efi. The only nice thing on mech is potentially better atomization.

Carb with turbo, I would draw through.

I fiddle with a '57 Fuelie Corvette at work for one of our regular customers.

It's not too bad of a system once you wrap your head around it.

Once it's warm, it runs great but the cold start and warmup system can be a bit fiddly.

It -loves- WOT and runs ok at idle and cruise.

It's a very rudimentary system but it gets the job done. In it's day it was the cat's pyjamas.

I'd bet that if you put it on a dyno against an indentical engine that runs EFI there wouldn't be a significant difference.

Interesting page on the Rochester mechanical FI, along with pages on Kugelfischer etc:

http://arrc.ebscohost.com/ebsco_static/repair-tips/8852CH15_ROCHESTER_MECHANICAL_FUEL_INJEC.htm

My dad had a '58 FI 'Vette that no one knew how to work on, this was ~1977. He sold it pretty cheap for that reason.

Lucas mechanical FI information:

http://www.lucasinjection.com/

Pretty interesting. The 'History' page has some cool stuff on it.

mguar wrote: In reply to Trans_Maro: That's very interesting.. One of the great things about this site is the stuff I learn I didn't know.. I think I'd better go to the Holley and Rochester web sites for further information..

As near as I can tell, none of the people who know how to make a carb work with a turbo still work for Holley. At least they didn't when I contacted them back when I was trying to get a blow through turbo working on my Dart, so I gave up and went EFI.

There's some guys out there who can mod Holleys to work with boost. Personally, I'd rather try using a carb with metering rods and use something that makes the rods move out under boost, but I've never seen such a setup offered for sale and didn't care to get the tools to make my own.