gemellocattivo.com

Have you ever thought of what would be required to add DI to an existing engine?

There are plenty of resources and technical papers which analyze the spray patterns, injector requirements, benefits (power, emissions, economy, stratified charge, ...), calculations and calibration methods, etc. There are also plenty of people in the aftermarket who upgrade DI engines with different injectors etc., and even a few standalone controllers capable of controlling them.

Finding an engine with similar power/cylinder should have a reasonably well matched fuel injector, and an HP pump from an engine with similar power would be a good starting point but the lobe profile (3 or 4 lobes and pump cam lift) can be changed a bit to adjust it.

However, the fabrication work isn't really trivial, especially since the DI injector would probably have to pass through the water jacket.

Anyone have thoughts on the mechanical aspects of a conversion?

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke

You probably want to look at hp/cylinder when selecting an injector.

Fabbing it is a matter of finding a path that is chamber >water jacket > outside.

if that's true all the welds are on surface you can get to easily and its just drill, drop in the bung, weld.In the chamber I'd leave the bung short and fill from the end of the bung out into the chamber with weld so you clean the chamber as was and still know you haven't ground all you weld away.

I'm a computer engineer, I didn't think of that.

I was trying to think of how to seal a bung on the back of the chamber casting where there wouldn't be room to weld. Drilling it all the way through and welding it on the chamber side would be much easier.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke

I'ts not an easy job for anyone which is part of why you don't see it being done. This forum is one of the few places you'll find welding heads to be considered normal The other part is that other than improving emissions it doesn't add much and is generally considered to REDUCE hp if emissions aren't a consideration so racers don't mess with it.

You could probably do it without welding.....thread the bung into the chamber and seal with a pipe thread or o-ring then another o-ring and nut on the outside. For me it's usually faster to just weld but welding used castings is not so easy because they have all kinds of junk sucked into the pores that makes it want to bubble so you have to work the out and they suck up a lot of heat so you need to pre-heat and use a big welder.....mine in good to 400A, but for a motorcycle size head 250-300A should be enough.

I think welding could be done suitably. The head isn't very used and I can get a new one if necessary. I think our larger welder goes a bit above 300a. I'll ask.

Not really concerned with power loss either.

Theoretically, it can improve fuel economy considerably (by running stratified instead of throttled). Emissions are actually worse steady-state when running stratified - HC goes up when stratified if the mixture formation isn't perfect and the exhaust composition isn't suitable for a 3-way catalyst. As far as I can tell, no production cars run stratified yet, they all run homogenous stoichiometric and the only DI advantage to that is in-cylinder cooling for improved knock. But we're already getting into the lean emissions and really low cat efficiency with our PFI engines so it's not a huge loss or change.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke

There's more to it than that for production cars. being able ti time the fuel to start after the exhaust valve closes allows them more freedom on cam timing and lobe design and that translated to about a 10% hp bump for them. and then they get to run higher compression and that improves mileage. So there are benefits....bit it still seems to be down on hp compared to what can be done with port injection I thing due to better charge cooling with well designed port systems.

I haven't really seen any utilize more aggressive cam timing. Its possible I haven't looked at the right ones. I do know that MultiAir2 has extremely aggressive valve events at times, but that's PFI.

Higher CR comes from better knock limit, so that's a good thing too.

PFI has much better homogeneity of the mixture. DI systems are quite bad at this. I assume that would hurt power.

Most companies don't want to deal with the expense of more complex catalyst systems required for any emissions other than strict stoichiometric (which can use a relatively simple, cheap, and common three-way cat).

Stratified combustion produces Diesel-like efficiency and fuel economy (and it's basically calibrated/controlled like a Diesel with spark), but the emissions aren't matched to a 3-way cat so OEMs are reluctant to actually use it due to the emissions penalty.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke

Step 1: Pick your engine well.

Probably the biggest factor is simply whether or not there's spacer for the injector bung under the intake port. If your target head doesn't have that, you're done; pick a different engine. Also, 4V chambers have better geometry for DI than 5V chambers, as OEM's building 5V engines (Audi, Ferrari) went back to 4V chambers for DI. Of course GM's new DI V8's use 2 valve chambers and seem to be pretty impressive at making power.



I don't know what the limitation is. If timing the injection right guarantees that no raw fuel goes out the exhaust port, then there shouldn't be an emissions limit to how high a state of tune a manufacturer can build, unlike with port injection. DI + throttle per cylinder should allow some pretty impressive states of tune without emissions compromises. Of course all the OEM's were in such a hurry to add turbos to tiny DI engines to meet fuel economy requirements that naturally aspirated DI engines remain relatively rare and we haven't seen much serious development of them.

I can tell from experience that the 2014 Corvette is a BEAST.

The racing application is a little different.

I haven't heard of injector swaps or aftermarket controllers which can run DI. I'd like to learn more about that.

The engine we currently use with PFI is a 4-valve DOHC engine. There appears to be enough space in the combustion chamber for a DI injector recess as well as enough space outside of the head, but there is a water jacket in between them. There is a bridge in the water jacket but I don't think it's thick enough to enclose the injector, in the vertical direction. A bung would be necessary.

Second issue is high pressure pump mounting and cam lobe. There is space on the camshaft for another lobe (Between the two cylinders, on either camshaft, they are identical). Some sort of mount block would probably have to be made from scratch and maybe bolt on top of the cam bearings on either side? It looks like most DI pumps have followers which move up/down with a slip fit in their bore. We would probably need oil flow to this area as well.

Third issue is fuel lines. This isn't huge, it's like making any other machined fuel rail but with more precision and higher pressures.

Electronics are relatively simple. Solenoid injectors need a peak and hold supply with high-voltage peak (~65v). There are ICs which can do this as well as injector driver boxes which work with a PFI ECU. The HP pump needs a 12v peak and hold circuit and needs to be commanded at specific angles, we can probably use a standard 12v peak and hold injector circuit for this and drive it similarly to a spark output.

Software would be similar to PFI for homogeneous mixtures. PFI fuel software would probably be adequate to start with, plus HP pump control.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke