MFI on a 3.6 or 3.8?

[724doorE]

Anyone done this? While I realize before I even type this that many will simply say, "why would you do that???" But a 3.8 RS in an early car with mfi would be pretty neat! and if one had access to a core 3.6 for cheap enough.....

Well, again, anyone done it?

[pjv911]

Go with slide valves and efi . Forget about MFI . Oh yea you may need to mortgage the house first.

Kurt Williams

[Az911]

How does slide valve injection work?

[Jack Olsen]

I believe it's been done. In my opinion, it's cost-prohibitive... but cool.

[beepbeep]

Hmm...my understanding is that MFI space-cams are ground and thus "hardware-mapped" to fit engine that system sits on. It would probably be helluva work to "map" it by grinding new cams, installing them and testing car on dyno. Also, there are probably some other limiting factors as you would need roughly double amount of fuel that 2.7 needs.

I wonder if this played any roll when Porsche used similar sized P&C's for 917. It must have been lot's of work to "map" those things right.

MFI is/was cool in it's time but i fail to see the fascination. There are perfectly functioning EFI systems that will work witg TWB's and give you as instant response as MFI.

[Bill Verburg]

I agree w/ Kurt & Beep. MFI was ok in its day but that day is long gone. EFI is far more flexible and will reward you with smooth, reliable power.

For street and light track Motronic w/ mild cams is fine. For more serious track use Motec or DTA or similar management w/ TWM or similar throttle bodies and lumpier cams is the way to go.

[jluetjen]

I'm sure it could be done by someone with the appropriate ingenuity, but the three major issues (inorder of increasing complexity) that I could think of would be...

1) You would need to get a set of appropriate throttle bodies.

2) The space cam, depending on the engine configuration, porting and cam, a special space cam may need to be created.

2) Do any of the MFI pumps have cylinders with enough capacity to squirt enough fuel??? Keep in mind that on a 2.7 or 2.8, each mfi cylinder needs to squirt enough fuel for about a ~466 cc cylinder or nominally 34cc's of gas per cylinder per engine cycle. The 3 liter SCRS MFI would have been pumping nominally 35cc's of gas per cylinder per cycle. A 3.8 has a cylinder capacity of 633 cc's and thus would nominally require 46 cc's of gas per cylinder per cycle. That is a 41% increase in fuel flow over the fairly rare 2.8 MFI set-up. To do this I could imagine that you'd need to re-engineer the MFI pump with the cylinders having larger bores (with matching pistons), or updating the cam shaft that runs through the pump that acts like a crankshaft does in regular motor, or both. I could picture that being a fairly significant undertaking.

Keep in mind that from a fuel injection perspective, a 917's motor was just 2 RSR motors on a common crank with an MFI pump which has twice as many cylinders. But each MFI cylinder has no more capacity then in an RSR application. Now maybe a 935 pump could be modified, but what will it cost to get one?

(PS: In a moment of "free thinking" it dawned on me that a 2.0 MFI pump could be used and run at twice engine speed. This would double the fuel flow and as we have already determined, when the fuel gets injected doesn't really matter. I just wonder if the thing could hold together? Alternatively, why not hook up 2 x 2.0 pumps and injectors in parallel??? Hmmmm )

[724doorE]

John, How many cc's were the mfi 935's pumping??RSR turbo 2.1 at 540hp??

Designing my own patterns for throttle bodies is of interest to me anyway. Having them cast and machined is no problem either.

An alternative would be MFi style TB's with EFI or motrinic or something. That would give the "look" with the useability of the modern electronic systems.

Depending on the design of the pump piston, one might be able to simply bore the cylinders to an appropriate size and manufacture larger pistons. Another annoying thing about the MFI pump, its freakin heavy being cast iron!

[Bill Verburg]

TWM throttle bodies/DTA engine management 380hp

[jluetjen]

Dennis;
Those numbers were just "back of an envelope" calculations. Basically engine capacity / 6 = cylinder capacity. At 100% VE, the amount of fuel required by any cylinder will be 1/14 of the cylinder's air capacity. That should be close enough for sizing requirements.

I don't have a clue about a 935's MFI pump capacity, but if at wide open throttle if they were running 1 bar (aka: 1 atmosphere) of boost, the fuel requirements will be roughly double what the engine would require at ambiant pressure.

Thinking about my hair-brained idea further, I like the idea of "ganged" pumps. One could drive a set of injectors low in the ports with a "T" 3D cam, these would be the "low RPM" injectors. The second pump could drive a set of injectors which are very high in the stacks which would be the "high RPM" injectors. This pump would use a n S or RSR 3D cam which would run considerably richer at high RPMs. While both sets would be firing all of the time, the higher ones would get richer at high RPM when the intake velocity is much higher.

Nobody seems to like the idea?

[Early_S_Man]

You're on a roll, John, so keep at 'em ...

How about the Bosch MFI pump from a Mercedes 6.3 V-8??? It certainly had the fuel capacity per cylinder!

[jluetjen]

Or better yet the pump from an ME109 and just use every other nozzle.

[Matt Holcomb]

My late mechanic wanted to build a 3.5 litre MFI race motor after completing my engine.

His solution: one MFI pump per bank of cylinders.