Extra Power in a diesel with LPG

[minormatt]

I must admit that is more what I was thinking initially...

[SimonHobson]

Yes, that would work, but I suspect it would be "quite variable" in it's gas delivery.

Firstly, the cylinder pressure would vary with temperature - so variable pressure at your metering nozzle.

Secondly, with under-bonnet temperature it would start to vapourise. At the start line, you'd have only vapour in the line, when you hit the switch, liquid will flood into a warm hose - so you'll initially get a slug of warm vapour followed by a mixed phase of vapour and liquid. The warmer the hoses etc, the more vapour you'll get out.
Whether you could work round these issues with a pressure regulating valve, and some suitably applied ice* would be an exercise for development work. You'd want to cool the liquid after it's left the cylinder - it'll keep it;s pressure, but you'll introduce a margin by which it would need to warm up again before it starts to vapourise.

You might be able to avoid some of the problems by using the tank & pump from a liquid gas injection setup. There you have an in-tank pump which boosts the pressure, a pressure regulator under the bonnet, and a return line that allows for a constant flow of liquid round the system. It'll add expense though.

Either way, introducing liquid propane into the intake will also cool the intake charge. At atmospheric pressure, it boils off at around -40˚, at 1 bar it's around around -20˚C(according to the graph at http://encyclopedia.airliquide.com/Ency ... p?GasID=53)

* As an aside, the JCB diesel land speed team used only ice for their coolant. The vehicle had a big ice box in the front through which they ran water which was used for both intake charge cooling and engine cooling.

[minormatt]

I've had a silly thought today...

If I were to get a disposable propane cylinder (i.e proper 1" threaded plumbers one) and inverted it I could mount it upside down (so it forced liquid out) I could use this as the source of the LPG? It would mean I didn't have to cut any holes in the bodyshell and would be able to easily remove the set up for use on the road... If the cylinder only lasted 1/4 mile it wouldn't be a problem even if there was only a little bit left I could use in the blowlamp!

Any thoughts?

[Gilbertd]

It should work. I'm not sure you need to inject liquid as all you are trying to do is add a little LPG to the air to make the diesel burn better. I've got one of these http://www.bernzomatic.com/item.html?id=30 that uses either propane or MAPP gas for fuel which is liquid so lasts forever but the Oxygen cylinder just contains Oxygen vapour under pressure so lasts for about 10 minutes if you are lucky. That gets interesting though if the cylinder falls over when you've got it lit and liquid gets to the nozzle. Consequently it gets used more as a source to directable propane as an alternative to Easystart, for lighting bonfires and bleeding brakes (drill a small hole in a spare master cylinder cap, jam the nozzle into the hole, turn the gas on and it forces brake fluid through the system so you can simply open the bleed nipples to let the air out, saves getting the missus to sit in the car going, up, down, up , down).

[minormatt]

I think its worth a play...

If I have time I might knock something up over the next week.

I would think MAPP is going too far though

[classicswede]

As you only want it on the drag strip. and then only for 10-12 seconds at a time when you already have a full turbo boost of 1 atm.

Why not feed liquid lpg direct into a drilling in the manifold .
Your gas bottle should provide the required pressure.
If you were to use something like a gas cooker jet you would have a large choice of jet sizes.
A simple electronic timer would open a standard cut-off valve for the required time.
manually operated when you needed the additional boost
you would only need a minute bottle and everything could be removed before gong back to roadwork.
High pressure braided SS hose should satisfy the scrutineers.

just a thought.
No doubt attract derision for the health and safety brigade.

It is going to go BANG

[bxman]

your 20 second run is not going to use that much LPG the experts could probably work it out I suspect it would probably be less than 10cc.
IMO The vapor pressure in the tank/bottle will maintain the lines up to the solenoid valve in a liquid state. I believe it is about 100psi at 50 deg F.
The manifold should not be that hot and with a short 1mm bore hose between solenoid valve and jet there should not be to much time for the gas component to be purged.

you can try it before drilling the manifold or even fitting it on the van.

keep us posted so many people forget to give feed back.
good luck.

[rossko]

If it's purely for the dragstrip you can simplify enormously.

Starters, use a full-throttle switch for turning LPG on - you don't need it any other time. Still smart to have a crash shutoff, RPM or oil pressure sensitive maybe. I'd be suprised if the scrutineers didn't insist.

Detachable plumbing is a great idea for avoiding road use issues.
You may not need to upend a bottle for liquid supply, just use a more generous hose size for vapour - this is only a supplemental fuel sized like running a 500cc engine.

I too think you should do some sums to estimate how much gas you need. Full power diesel is what, around 12:1 AFR (by mass)? Work out how much fuel you're eating per cycle, and you then estimate for say 15% LPG. From the rev range you're going to cover, you can work out the rate for LPG. Guesstimate in the 1g/s league?

Can't see how you'd meter the LPG against rpm without a mixer; but you might simplify and compromise with a fixed steady dribble sized for best effect in the mid range.
Presuming you are going in after the turbo, we've already mentioned referencing pressure to manifold.

Lorry style systems use an LPG injector to modulate flow, but I'm guessing you don't want to get into full blown systems development.

I'd still be a bit worried about overrevving, unless you're confident it cannot. The diesel might cut out to prevent in-cylinder ignition ... but that leaves you pumping hot gas/air mix through a hot engine and exhaust. It'd be likely to blow a street exhaust to bits, but I guess you disconnect that?

Fraid I feel it'll end in a bang without onerous, cautious development

[SimonHobson]

IMO The vapor pressure in the tank/bottle will maintain the lines up to the solenoid valve in a liquid state. I believe it is about 100psi at 50 deg F.

You'd be wrong. In the cylinder is saturated vapour, the pressure will be exactly that required so that the rate at which vapour molecules move from liquid to vapour exactly matches the rate at which they go the other way. If you lower the pressure at all, then the equilibrium is changed, and more molecules will come out of the liquid than are going back in. This will continue until the liquid has cooled sufficiently (by the refrigeration effect of the liquid evaporating) for equilibrium to be re-established. The same applied if you raise the temperature at all.

To put it in perspective, the effect is sensitive enough that it will drive gas between tanks while driving. On my Disco, I had three tanks, and the rear tank would always cross-fill from the cill tanks - even though it was slightly higher up. Quite literally, I could watch the fuel gauge go up while driving - it would take 50-100 miles to transfer about 10l of gas.
I've also observed the effect when filling from red bottles. A bottle that was full would cause far less problems than one that was only part full - even though the pump itself was below the liquid level.

So, you have a cylinder and draw off liquid. That flow goes through pipework and fittings which will all cause a slight pressure drop. It ends up in pipework under the bonnet which will be warmer.
By the time it reaches the engine, it will be a mixed phase - unless there's enough heat input to completely vapourise it.
This is one of the problems faced by the liquid gas injection people - the underbonnet temperatures tend to make the liquid vapourise. That's why they use an in-tank pump - both to increase the pressure above the saturated vapour pressure, and to circulate fuel back to the tank to minimise it's opportunity to vapourise.

[minormatt]

I will put in a generic rev limiter to stop the over-revving.

I do like the oil pressure switch cut out idea. And a simple full throttle switch will be easy!

I also have spare engine which can go in it if I blow the existing lump up (its not the original one anyway, its from my other van - the one that now has a V6 in)

[minormatt]

Speaking to Dai last night I think I have a suitable way forward for this. Its going to be quite a simple system which is easily removed (well most of the electrics will be permenantly mounted), Installing/removing the gas system should be doable in a matter of minutes, with some jubilee clips and a couple of bolts.

I will be reporting back with pictures

[dannyf]

I done some testing with Gas in the mk4 1.9tdi and at 160bhp it would smoke like a good one
we used some LPG in the inlet to clean up the burn an make a bit more power

[tommo2009]

I used to use a bank of four injectors into each port in the manifold (similar to a petrol set up). then just switch the injectors on or off with the idle switch as mentioned earlier. The bigas injection reducers are ideal as you can take the lid off and cut the spring inside to give you full control over gas output ie you can wind it down to zero output.
From the vapouriser back is the same as any standard set up with both front and rear lock off solenoids.

First venture into this was an escort van, still makes me smile to this day. but be warned if you put too much gas in you WILL destroy the engine.

To set mine up i usually get someone to hold the revs at aroun 2500 then turn the system on. gently wind the pressure up on the vapouriser until an increase in revs is heard. about a 10% increase is sufficient