Water injection, need ultimate response, what design?

[BLSTIC]

I'm about to build a very octane sensitive engine. Factory released figures show a 7% increase in torque in the same engine, between regular and premium fuels (91RON to 98RON), with that total being 10.5% higher than the otherwise damn near identical earlier variant that was tuned for 91 and had lower compression. Naturally I want that torque. There are higher compression again engines, but those are designed for stupidly high octane LPG, so I'm not sure I should try running 12:1 on a 3.6x3.9" bore/stroke short revving engine.

Because of the apparently octane sensitive nature of the engine design I want to run water injection to make the most of it. Because it's a naturally aspirated engine with bulk low-rpm torque (420Nm from a 4.0L engine, or 310ft-lb from 243ci) it needs to react really fast. Fast enough that I'll probably use the rate of TPS increase as an earlier trigger than solely manifold pressure.

So what systems have you guys got that are suitable for naturally aspirated engines, have good atomisation, and have close to zero response time?

[teoman]

Commercial or DIY?

Fastest response time would be with a gasoline injector.

[gone-ot]

+1 on high-pressure injection of H20+methanol.

[BLSTIC]

DIY.

I hadn't thought of a fuel injector (recently) but previous musings have lead to injector and pump durability concerns.

Unless those soluble oils for machining don't encourage detonation when mixed in a bare minimum ratio with water?

[cRiPpLe_rOoStEr]

Those soluble oils seem more suitable for cooling in CNC mills, but I'm noto sure if that would be safe to inject to an engine. But anyway, what engine are you going to build? Is it based in the Ford Falcon "Barra" engine?

[BLSTIC]

That guess was scarily accurate. Are you... THE MAN? Should I wear my tin-foil hat lest you steal my dastardly ideas for fuel economy?

But seriously, what gave it away?

[BLSTIC]

Tractor guys use it in their mechanical water injection pumps, so I assume it's usable. Not sure how detonation resistance goes but the oil mass will only be about 0.3% of the fuel charge so even if the oil was 0 octane it couldn't drag the total down by much

[RustyLugNut]

Use ethanol rated solenoid fuel injectors. They work well for water injectors. Use a water pump that is rated at 80 psi and pressurize a working chamber that has an elastic buffer (spring or gas). Use a basic Arduino to deliver your needed flow between 55-80 psi through a hexfet with a capacitor bridge using pulse width modulation. Easy peasy. The pump does not need to run continuously and the Arduino can be programmed to cut in the pump below 55 psi to replenish your reservoir.

[cRiPpLe_rOoStEr]

Originally Posted by BLSTIC

That guess was scarily accurate.

When you said it's a 4.0L engine, the first thing that came across my mind was the Falcon. Later I checked Wikipedia, only to find out the displacement and bore x stroke matched. Well, when I remind the Brazilian F-250 had been available for some years with the Canadian Essex 4.2L V6 as its base engine (not the regionally-sourced MWM straight-6 turbodiesel), it does surprise me somewhat that the Barra had never even been offered as an option for the F-250 back there.

Back to topic: are you considering to add only one injector right above the plenum or individual injectors for each cylinder? When it comes to corrosibility, injectors rated to operate with ethanol might fare OK with water. Trust me, I'm Brazilian and familiarized with ethanol-powered cars.

[BLSTIC]

Single injector mounted upstream of the throttle is the plan at the moment, providing I can get one with a fine enough spray pattern.

The original plan was an ultra-high pressure coffee maker pump with a single nozzle, with pump switching. This allowed for 200+psi at 150cc/min. I don't know how fast it would react, but atmisation is almost perfect at full flow