liquid propane injection 2

[geofire]

Hi, I've been thinking about liquid propane injection lately using nitrous solenoids as shown here:

http://www.diy-nitrous.fsnet.co.uk/nitrous-solenoid.htm

The setup I'm contemplating is a turbo motor and I would like to add liquid propane for its high octane and how it would lower the intake temperature as it phase changes. I may try to go pure propane eventually, but for now I want to use propane assist so to speak. I'm running megasquirtnspark extra so it would be easy to trigger the solenoids only when needed, and lean the gasoline maps to compensate so it doesn't get overly rich when the propane kicks in.

in order to ensure I will be getting liquid and not vapor when the solenoid fires I was thinking of setting up a fuel pump with return to circulate the propane and purge the vapors. The question is what kind of pump do I need?

http://www.caranddriver.com/article.asp?section_id=26&article_id=2279&page_number=1

That article says a corvette fuel pump, so does that mean I can use gasoline stuff?

 

[franzh]

Where I live (Central Texas) LPG costs $1.84 per gallon then we pay an additional road use tax based on annual mileage and vehicle weight. Unleaded regular costs $2.84 and its climbing again, with all taxes calcualted at the pump.
I created a spreadsheet based on these factors plus the thermal differences between the two fuels and at these prices, there is no financial payback. Its hard to justify the investment. We need a real cost benefit to make it justifiable, more than a dollar a gallon.
Franz

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[turbocohen]

I filled my home station up for 1.51 a gallon before storm season. With gasoline at $2.69 its an ok deal.

 

[gasman32]

Walbro make a liquid LP pump. It is currently being used in the Schwan's Bi-phase liquid injection system.

 

[turbocohen]

Actually Walbro makes a gasoline pump and calls it an LP pump. I have an old chrysler canada Dakota mule with liquid fi and a dual pump module with walbro pumps. It does not compress vapor and it lacks cranking voltage flow rate and pressure required for vapor purging a liquid system during hot restart. The best pump out there was produced by Vialle. It was able to crank out enough pressure boost and adequate flow at 7 volts. The walbro supplied systems must be allowed to flow before crankup to allow vapor purge and that can take up to 25 seconds on early systems and at least 5 sec on a prototype rail was the best 20 minute hot restart I could reliably demonstrate. Try hot restarting after 10-20 minutes on a less than 80% battery that otherwise will start a gasoline vehicle and the problems get worse. The walbro also puts way too much heat into the fuel. Ideally an liquid system needs enough flow to prevent fuel vaporization and it takes either more pressure or flow. More pressure and as little flow as possible were my targets and underhood fuel components that absorb as little fuel as possible while optimizing hot restart vapor purge are the key. Heat transfer from underhood and body are hot issues npi. Been there done that filed the patents built the demonstrator systems tested in the desert and could not honestly overcome fuel injector leakage due caused by internal boil back generated deposits. Since then I have tested everyone elses systems and performed teardowns that all revealed a common problem. Hot fuel handling issues, deposits and low voltage sensitivity/failure. I have replaced my pumps twice. The original gerotor walbro pump did not have an overmolded armature and it needed one badly!

In short the Walbro pump may be ok for a back yard conversions but not an all weather oem system that can pass all of the durability validation requirements (hot cold low volts, overvolts, fuel tank empty, etc) that gasoline systems easily pass.

The Schwans system is elegant but that does not change the chemical composition or properties of propane or butane. The top feed injectors are much easier to package than the bottom feed injector my customers used but hot restart is aggravated with top feed since vapor bubbles rise.. Bottom feed self purges with less pressure required.

I wrote a guide book for liquid injectors use that was distributed to clients. The only advice worth repeating was to gain know how about refrigeration systems. Simply understanding the fundamentals of how an air conditioner works will help the hobbyist or professional engineer new to lp fuel management to understand the behavior of the fuel. My prototype fuel systems have sight glasses that are the same as the ones used with a/c systems. There are also electronic sight glasses that can be a little useful for giagnostics. The first time I sight glassed a system with the walbro pumps the problems with vapor were obvious.

Lots o luck, T

 

[patprimmer]

Great post turbochen.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[turbocohen]

Here's a potentially useless link:

http://webbook.nist.gov/chemistry/fluid/

Put thermocouples throughout any lp system, especially at the injector tip. Plug the temps into the table on the link.

Does anyone know of a link with good dispersant data?

 

[NickDV]

Interesting thread gentlemen. I see that Visteon have a brushless pump that supplies good volume at 450kPa, which might get me over the line (if my current Denso unit fails). I'm not sure about the effect of low voltage though, or whether Visteon use an overmoulded armature. Turbo I'm somewhat vexed to read your comment about hitting the wall due to deposits in the fuel system. Hopefully my latest design won't suffer the same fate here in Oz now that LPG is $0.45/litre vs petrol at $1.30/litre! One tip for the hot soak restart issue is to start the pump the instant the interior light comes on.
Cheers, Nick.

 

[franzh]

Starting the pump when the door opens is the method used by most of the liquid injection technologies. One problem faced is hot soak, where the vehicle is used in slow speed transit application (the most probably end usage), where the engine is operating at its hottest, where the undercarriage heat is the highest, and the road temperature is the greatest. This is where the fuel tank tends to absorb the most heat and the propane pressure is the highest. Again, I have seen tank pressures at the 312 psig pressure relief point numerous times.
Franz

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[rdd48856]

On bi-fuel vehicles why not starting always on gasoline and changing to LPG when all conditions are O.K. If tank temperature is very high why not stop the LPG pump and change to gasoline?

 

[NickDV]

bi-fuel systems are considered a crutch that leaves the consumer with two fuel systems to maintain, furthermore the complexities of mounting two sets of injectors (or trying to have one set deal with both fuels), having two fuel tanks etc, is way too difficult/expensive. Turbo's comments about A/C systems suggest an answer I think. Cooling the unused fuel in the return line (the condenser in an A/C system) and mounting a pressure regulator inside the tank (like a TX valve) should provide much needed cooling. Has anyone tried this technique?
Cheers
Nick de Vries

 

[turbocohen]

NickDV, I have tested several Visteon pumps including the ST-150,170 and others. It is a superb gasoline pump and a much better LPi pump than the walbro. I set up a system with two ST-170's and a pwm control to allow cranking voltage to the pump and reduced output after vapor purge. Do not use lower than 8 volts. The problems with vapor generation still persisted though and to be DOT legal the door switch trigger is a show stopper. I NEVER set up a system with a door switch activated fuel pump. The risk to service personel or rescue crews after an accident is too great. OEM's typically reach in the window and crank when evaluating aftermarket stuff so the lame door switch trick is defeated.

As for the A/C, the fuel hadling issues are easier to visualize with some A/C knowlege.

Franz, I have blown several PRD's at 450+psi tank pressure at the Arizona proving grounds. One way we accomodated the heat was to meter a little vapor to the pcv inlet from the top of the tank and that dropped temperatures moderately.

Nice posts guys, yall are thinking

 

[NickDV]

Thanks Turbo,
You're right about the service personel of course and I hadn't got around to thinking through the logic sequence to keep them safe at all times. OK. An extra button on the remote entry might be an idea worth trying instead. Now that modern diesels are starting without the customary "10 second wait till the red light goes out" routine, the consumer tolerance for waiting 10 to 20 seconds for vapour purge is just not there (ie no market comparison). I have experimented with fuel rail heat saturation to 100C (& brass injector holders!) with an A/C sight glass at the return line entry point back at the tank. At 4 litres/min I got down to 9 seconds before the vapour purged/reliquified. I'm expecting my plastic injector holders to perform much better at the 10 minute heat soak peak. If I can get the vapour purge down to <3 seconds I'll be over the moon. At the current fuel prices, an injector clean/service and additive dosage regimen is certainly worthwhile in the name of injector longevity. Hopefully the propylenes wont polymerize too badly in the guts of the injectors.
Nick de Vries

 

[turbocohen]

Nick, propylene and sulphur are equally bad except that even the best additives do not safely remove sulphur once it has formed. Injector clean up is a misnomer since the real bad stuff that deposits at the injector orifice and seat/needle cause what appear to be semi permanent failures if they are not removed before they become a form of polymerized plastic that is relatively impervious to what is nicknamed condensate fuel wash. When the system refills with fuel a small amount of the deposit is not dissolved and heat/coll/repeat cures it in place along with availability of oxygen from the intake manifold side of town.

The LPi hardware I had luck with was made from solid white glass fiber filled nylon with thermally reflective "stuff" vapor deposited inside and out to reduce heat transfer but the vapor purge was assisted by gravity. The fuel inlet on the bottom of the pod passed fuel through the injector and out the top cam bubbles (as witnessed with a/c sight glasses in the fuel lines.