Jump to content

Diesel Motor LPG Misting... bit of a vapour project


Recommended Posts

Me and a mate have been bouncing round an idea to use a petrol efi setup to run and tune an LPG fumigation set up

For those that haven't heard of this the LPG gas is bleed into the air intake the small amount of it acts like an accelerant improving combustion.... this is supposed to give improvements in economy and power....

It really does work but.. I find some of the claims from marketer selling these setups... lol yea well... as for the crowd that promise a an 80hp gain... with no qualifiers as to what motor its fitted too or size, style, turbo or not

But as I said there are enough gains that I'm going to give it a go

So done a bit of research on this, LPG has a couple of issues with its use,

BIggest is too much fuel will melt pistons lol

Im bleeding into the intake a flammable gas... I dont want it there when its not going to be used

Also Im not certain on how a cold cylinder would like this

The idea is to set this up first off with an over ride switch to turn it on

then going to setup a water temp thermal cut out for too hot or too cold

Tap the injector (injectors if we cant get the gas flow) into the intake motor side of the of the turbo

Fit a MAF air flow meter on the intake, setup a crank position sensor, air temp, throttle position sensor and the O2 sensor in the exhust

with these there we should be able to use a petrol engine ecu to "map" the LPG gas delivery in a nice controled way

to adjust flow we can alter injectors (size and number), pressure of gas (rough estimates calculate out to 6-10psi)

If the map proves to be wrong for the amount of gas we can either crack it or run a piggy back that we can programe

Any thing we are missing ?

Link to comment
Share on other sites

A couple of points you may not have considered.

I believe that electrically powered petrol injectors rely on the petrol to lubricate them.

I believe LPG (vapour or liquid) doesn't have this lubrication property, so petrol injectors passing LPG have a short life.

You are putting more fuel into the cylinder so, IF you can burn ALL the fuel, you will get more power, irrespective of any 'accelerant' effect.

To burn ALL the fuel, from both sources, you need more air. Think about how you are going to get more air into the cylinder.

The underlying assumption behind that comment is that (as you appear to want as much power as possible) the basic diesel fuelling is already set up to give maximum fuelling at maximum power, such that more diesel fuelling produces a socially unacceptable amount of black smoke (unburnt diesel). If the diesel fuelling is set up like this, adding more fuel, in the form of LPG, will exceed the boundaries of what can be burnt, so giving black smoke.

In these circumstances your LPG fuelling control may need to incorporate an LPG cut when the diesel fuel delivery is at or near it's maximum.

Also bear in mind that your air delivery capacity is reduced, because you are 'blocking' the air with LPG vapour.

Effectively you can be in a situation where you are hit twice, at the same time, by adding more fuel AND reducing the air intake capacity.

Optimise your diesel delivery as already described (are the diesel injectors mechanical or electrical?) and any LPG advantages will be at part throttle where you will burn additional fuel (without exceeding the air supply) and (possibly) benefit from the 'accelerant' effect of the LPG vapour ensuring more complete combustion of the diesel fuel droplets.

Ironically, there is more scope for this improved diesel combustion if you are dealing with a mechanically injected engine with notoriously bad combustion (low power and high fuel consumption compared to it's peer group).

Start with a more modern diesel (electronic (possibly multi-pulse) control of the injectors, higher fuel line pressure, and better injectors, combining to give better vaporisation (smaller fuel droplets, better distributed over the combustion space and combustion time slot) and the scope for 'improved combustion' due to vapour LPG is minimised.

If the diesel fuel delivery is electronically controlled then any LPG vapour advantage centres around decreasing the diesel delivery at the same time the LPG delivery is increased. Any economic advantage depends on the relative costs (taxation?) of the two fuels in your part of the world. If LPG is noticeably cheaper then your fuel costs per 100km may reduce. Not enough to give you economic payback over 5 years, but this is a technical challenge, right?

HTH.

Link to comment
Share on other sites

Also bear in mind that your air delivery capacity is reduced, because you are 'blocking' the air with LPG vapour.

Effectively you can be in a situation where you are hit twice, at the same time, by adding more fuel AND reducing the air intake capacity.

It can work the other way round: if you inject the LPG into the intake stream as a liquid, as it warms and expands it will absorb the latent-heat-of-vapourisation from the intake air, cooling it, causing it to increase in density, and so potentially getting you _more_ air into the cylinder.

Link to comment
Share on other sites

A diesel tends to burn with circa 40% excess air anyway. Even with thick black clouds of smoke out of the back you still have excess air.

The way I've seen this set up on an automotive engine before is by using a trigger on the throttle pedal. Once a certain throttle position is reached, LPG is injected.

I'd also be wary of introducing it at the turbo intake - that's a lot of potentially combustable air/fuel and if it ignited then you might have some excitement. I'd work out a way of introducing it into the inlet manifold.

Large marine diesels are also burning LPG these days, both in a marine environment (LPG tankers burning cargo boil off) and in CHP installations in large buildings. SmasherWebbs knows more about this, but it's reasonably common.

Link to comment
Share on other sites

At the moment this is just Vapour... If we work out a system that is tunable and isnt going to cost too much, I have a project coming up a Sll 109 that I'm fitting a Nissan TD27turbo motor and trans... at the moment it is just a bit of fun both of us like to solve problems lol

Yes I know about LPG in petrols, I used to have a RRC setup on dual fuel LPG/petrol and the funny part I not long ago sold the LPG kit from it.... the "mixer" part that converts liquid to gas is what most of the kits on the market use

Now just to save this degenerating into an "argue with the snake oil salesman" thread..... we dont care lol there is enough studies on the net that back it from a reduction in emissions stand point... also it has been used in the trucking scene that I have found reference back to the 70's as a power boost with throttle linkage systems that pritty much just opened a BBQ bottle into the air intake, catch was you got to greedy you melted your pistons lol

What we are looking at is the new ECU systems that bleed a metered amount all the time dependent on rev's, load, ambient air temp etc they are marketing them as giving an increase in hp and economy.... now as dubious as this sounds there is enough evidence that the Australian govt is offering a subsidy to the installers for fitting them!

What we are playing round with is trying to use readily available stuff so we could make our own "open source" system that will do the same as the Aussy setups

This is the idea of using a petrol car EFI setup, if we can work out a way of tuning it to work to the diesel system, ideally with out re-mapping it, so either simple electronic piggy back pulse modulators or even simpler mechanically.... at the moment we are working off a honda accord 2L motor (also sold as the Rover 600) it runs a Mass Air Flow sensor so will work on the diesels intake so we have a measure of load, the distributor could be pulled apart to give the bits to make a crank position sensor so it has a measure of revs, adapt the throttle position sensor onto the throttle lever or to the diesel pump linkage, ambient air temp sensor is easy and the O2 sensor to help tune it... I don't know about this.... my concern is how similar are the exhaust gases O2 levels.... my mate is thinking of using it as a control by running a piggy back on it to change the signal so it can richen or lean the injection

The injector or injectors as the delivery device into the plenum chamber because they are an easily monitored setup all you need is a duty cycle multimeter then you can calculate then you calculate that plus orifice size and psi you are able to calculate how much gas

The bit Im working on at the moment is the delivery, the injectors

Tanuki you are dead right the conversion of LPG from liquid to gas is very endothermic this would give the bonus of giving a cooler air charge - more air molecules in the cylinder charge..... there is a couple of issues with this; first to take advantage of this you would need some way to increase the amount of diesel to match... the flip side of this is if we wind up the diesel pump to take advantage of this we must run LPG or we over fuel with black clouds if we dont or worse damage the motor

The second issue is at higher duty cycles and flow the injectors will ice up, the liquid to gas conversion is so endothermic it will turn any moisture in the air to ice at the nearest contact point; the tip of the injector.... with all the problems with tuning

There are a couple of bonuses to using liquid injection, first calculating calorie / BTU figures this is as easy as it gets, the other bonus is this system uses liquid if it gets gas in the lines it is alot safer as this drops the calorific value.... where as a gas system that gets a bit of liquid is in simplified terms getting 250 times what its used too!

The gas "injection" way, this also has a couple of issues other than getting liquid in the lines...first as David said gas doesn't lubricate very well if at all lol

The second is the injector is designed to have 30 to 50psi from a liquid pushing them open.... I have a feeling that 10psi of gas isnt going to open them fully this makes calculating calorie / BTU figures impossible

An industrial regulator will bring the pressure up to make them work but still leaves the lube issue

But I'm more inclined to go the gas route as this to me looks like the simpler to sort

Link to comment
Share on other sites

If you want to run a defined map, then a megasquirt ECU would do it quite happily. It won't care if it's LPG on a diesel - you'll need to feed in throttle position, and engine speed. Connecting the various temperature sensors up isn't technically necessary, but you might like to so that you can have an accurate temperature gauge. You'll need to make sure you choose 'Alpha-N' control algorithm. Then set up the fuel map for the right fuelling.

It'll need some fiddling, but perfectly doable.

edit: Actually, look at Microsquirt. Smaller, cheaper.

Link to comment
Share on other sites

what we are trying to do is to increase decrease flow to match demand from the motor.... just relying on throttle position and revs will melt your motor when you load up eg climbing a long hill, or towing a load where you have your foot down

We are using the MAF sensor to tie in with the throttle position sensor and crank position sensor so the computer has an indication of load vs acceleration...

I was also thinking of a pyrometer with a high temp cut out on the gas 1250F/680C, also water temp cutout as oil temp and water are pretty consistent and water is easier to tap into, My mate was thinking the knock sensor could work as well being the diesel knocking is a different frequency to detonation knock the catch is we need to reverse the ECUs response so it limits the gas flow.... if it and the O2 could be tuned it would make for a self tuning system

The rough theory is the car ECU map as its a 3D curve that takes in these things... so its just a case of giving it the right inputs.... if it comes down to the map is just not usable then we have a mate who is a auto electrical genius, who can remap em... the idea is to do this with minimal special gear or expensive bits

Link to comment
Share on other sites

Throttle position and engine RPM can be used to calculate engine load - it'll be pretty accurate on a diesel as throttle position is directly proportional to torque demand. Mass air flow is nice to have, but not essential - particularly in your application. You don't need to be particularly accurate, just accurate enough not to melt pistons.

Megasquirt should work more or less out of the box using the Alpha-N algorithm. No need to fiddle with code. Or if you do (to add your high temperature cut off etc) it'll only be code tweaks rather than having to write the whole thing from scratch.

Because of the excess air in a diesel, O2 adjustment is unlikely to work.

Link to comment
Share on other sites

Sounds like a regular sequential LPG ECU is what you need, on top of an electronically controlled diesel.

The diesel injector opening times should give a good indication of load, as they do on a petrol engine. Only other thing needed there is an rpm signal.

If you can inject before the MAF, the diesel ECU should take the added air into account, but not sure if that's a good idea.

Link to comment
Share on other sites

I wonder if you could use the alternator tachometer output for the engine speed for MS...? Just saves one more thing to hook up :)

It sounds like it is less critical in this instance than on a petrol engine, as you already have one form of metered fuel going in.

Ideally though, if it's an electronic diesel, then picking up on the injector timings should work very well, and no need for TPS or RPM signal.

Link to comment
Share on other sites

The diesel injector opening times should give a good indication of load, as they do on a petrol engine. Only other thing needed there is an rpm signal.

Yes, but the opening pulse is in the order of a few hundred volts. And the injector inductance is used to generate the next few hundred volts for the next injector opening. You'd have to do some careful electronics to turn this into something useable.

What engine is this being fitted to? The electronic version of the 300TDi engine uses a PWM driven fuel rack - you could potentially use that signal as a direct indication of engine load.

Edit: Now I've thought about it - you could use turbocharger RPM, or boost pressure. The latter will of course be limited when the wastegate opens, but seeing as you want to limit the gas at high engine loads anyway it might well work in your favour.

Of course Megasquirt comes with an air pressure sensor so again, it would work right out the box - no need for Alpha-N, you could keep the default speed-density algorithm.

Now I'm thinking about it on my land cruiser. I've got an MS ECU kicking around and I think it'll be pretty trivial to set up.

Link to comment
Share on other sites

Ford was running a high pressure lpg system on their transit engines and in the ldv these had high pressure gas injectors. might be worth having a look around the scrap yards as the biggest fault with the system was the ecu or pump in the gas tank.

Link to comment
Share on other sites

Because of the excess air in a diesel, O2 adjustment is unlikely to work.

Thats my view on it to....but my mate thinks he can still use it with some modification of signal, another option would be using a wideband o2 sensor.... this would then need modification of signal so the ECU doesn't take it as extremely lean... but this takes us away from our goal... commonly available parts that are cheaply found

Sounds like a regular sequential LPG ECU is what you need, on top of an electronically controlled diesel.

The diesel injector opening times should give a good indication of load, as they do on a petrol engine. Only other thing needed there is an rpm signal.

If you can inject before the MAF, the diesel ECU should take the added air into account, but not sure if that's a good idea.

What is different between a petrol and LPG ECU, I imagine there maps would be slightly different, from my experience with the dual fuel rover the gas flow on the map will be higher as its was the weaker of the two when used through the carb.... also unless you live in Australia where there are alot of LPG cars made by the manufactures finding a second hand ECU is way out of budget

lol we have no intent of injecting pre turbo... this is an explosive gas and will be treated with caution (if we decide to fit this idea to something)

I wonder if you could use the alternator tachometer output for the engine speed for MS...? Just saves one more thing to hook up :)

It sounds like it is less critical in this instance than on a petrol engine, as you already have one form of metered fuel going in.

Ideally though, if it's an electronic diesel, then picking up on the injector timings should work very well, and no need for TPS or RPM signal.

Yes, yes and no lol we aren't critical as this isn't the primary fuel this is just an additive

Given the vintage nature of New Zealands fleet of diesel motors think mech pump lol

Ford was running a high pressure lpg system on their transit engines and in the ldv these had high pressure gas injectors. might be worth having a look around the scrap yards as the biggest fault with the system was the ecu or pump in the gas tank.

Again same issue of commonly available parts... if I wanted to go this route I'd be looking at the new Aussy LPG Fords and Holdens that are direct injection liquid LPG they have solved the icing issue.... but we come back to price and availability again lol

it can be done - one of the guys in the DIscovery Owners club runs his 300tdi with an lpg system. Think it uses some kind of vacuum system to regulate when the gas enters the manifold

Sounds like one of the older styles of LPG fumigation.... they had a venturi tube that was fitted into the air intake

dsl-i61.gif

From what Ive read in the forums they work but with a delay and there are comments about damage when you go from hard throttle to a stop

Our idea is to inject directly into the plenum chamber we dont get the same build up of gas, there is some issue with the turbo boost and having to change the psi of the gas to compensate for the boost pressure but this isnt a hard one to sort given that the gas bottle pressure dependent on air temp will be from 50-80psi

Link to comment
Share on other sites

From what Ive read in the forums they work but with a delay and there are comments about damage when you go from hard throttle to a stop

Our idea is to inject directly into the plenum chamber we dont get the same build up of gas, there is some issue with the turbo boost and having to change the psi of the gas to compensate for the boost pressure but this isnt a hard one to sort given that the gas bottle pressure dependent on air temp will be from 50-80psi

Cough. Megasquirt. Am I selling this yet?

Link to comment
Share on other sites

What is different between a petrol and LPG ECU, I imagine there maps would be slightly different, from my experience with the dual fuel rover the gas flow on the map will be higher as its was the weaker of the two when used through the carb.... also unless you live in Australia where there are alot of LPG cars made by the manufactures finding a second hand ECU is way out of budget

lol we have no intent of injecting pre turbo... this is an explosive gas and will be treated with caution (if we decide to fit this idea to something)

The difference is that the LPG ECU is specifically made to work with another ECU's injector timings. The map in an LPG ECU is a percentage of the original timings, which is pretty much what you need.

But if the diesel injector pulses are indeed hundreds of volts (I wonder why?), it's indeed a bit more difficult.

Link to comment
Share on other sites

Cough. Megasquirt. Am I selling this yet?

LOL so are you telling me Megasquirt can compensate a gas line pressure outlet for boost pressure.... wouldn't it be easier with a diaphragm attached to a regulator and a adjuster nut

The difference is that the LPG ECU is specifically made to work with another ECU's injector timings. The map in an LPG ECU is a percentage of the original timings, which is pretty much what you need.

But if the diesel injector pulses are indeed hundreds of volts (I wonder why?), it's indeed a bit more difficult.

LOL still doesnt work because of the cost.... rare things cost money!..... and just for reckless "a megasquirt" would be cheaper lol

Cost wise with the right ECU I can get it flashed with a new map for $50 thats less than 20 pound for the englishman amongst us

Link to comment
Share on other sites

Fair enough, the ECUs are expensive, but they do sound like the solution you need. If you want it purely mechanical, I'm sure that's perfectly possible (venturi + adjustable vapouriser), but I'm just proposing an easily tunable solution for the electronic engines :)

Link to comment
Share on other sites

LOL Im not saying your wrong, but out of all of my mates only 1 drives an efi diesel.... alright I did hear somewhere that New Zealand has one of the oldest vehicle fleets think the average age is 20-25yrs all of the guys that drive diesels me included are mechanical pump indirect injection... oddly enough that seems to be from what Ive read the best motor to do this too

Talking with diesel specialists and from what Ive read on the net if you have a newer direct injection efi you are better off with more boost than this... the boost will give you better returns and safer than LPG

Link to comment
Share on other sites

LOL so are you telling me Megasquirt can compensate a gas line pressure outlet for boost pressure.... wouldn't it be easier with a diaphragm attached to a regulator and a adjuster nut

Yes, yes it will. Why include regulator adjuster when it can be done inherently on the fuel map?

Seriously, everything you have described so far can be done by Megasquirt without writing any additional code. You get full control over the fuel map and all parameters and a pre-developed tuning and logging platform. The standard fuelling algorithm is based around manifold air pressure - which in your case is exactly what you need - although it'll be pressure, not vacuum (MS won't care, standard sensor is good to 2.5 bar I believe). Hell, it'll even compensate for air temperature.

Using a petrol vehicle ECU sounds like an enormous challenge, and you'll end up with something bodged.

But if the diesel injector pulses are indeed hundreds of volts (I wonder why?), it's indeed a bit more difficult.

It's to ensure the injectors open as quickly as possible. Opening time is a big factor in fuel delivery calculations on a diesel given the relatively short time for injection. Once open, they're held open by the usual +12V.

I forgot to mention in the earlier post, but there also tends to be multiple squirts per cycle. As many as 4 or 5 in some cases I believe [citation required]. It makes for a smoother running engine and reduced emissions. Although at high power/rpm they all blend into one squirt.

Link to comment
Share on other sites

So reckless you are talking about a larger injector and map it for an increased duty cycle to compensate for the decreased gas pressure into the manifold due to the rising boost.... hmmm that has some merit

Alright then since your the man for MS, can MS take O2 sensor imput and adjust, would prob have to be wideband due to the lean nature of the extra air How about a knock sensor... this is where Im going to make you think lol ideally can we code the knock to drop back say 200rpm on the map with a delay then advance, so say under high load that sustained, a big hill, heat build up in the cyclinder will start to pre-ignite the gas... knock - the map rocks back revs (less gas) driver still holding foot into it knock again another drop on the map etc end of delay each knock movement will advance

If you cant do something with that how about using the dwell angle to reduce duty cycle, or the line presure... basicly I need knock to drop gas flow....

If it is do-able I dont mind learning another language lol that and hardware were my things when I did computing.... Im guessing there is a compiler for MS

Link to comment
Share on other sites

Yes, O2 sensors are supported out of the box. Both narrow band and wideband.

Knock sensors are set up with petrol engines in mind, so they'll adjust the spark (which you obviously won't be using). You have to tweak the PCB to enable the input. There's no reason why you can't tweak the code to make it do what you want.

It wouldn't adjust the revs on the fuel map - it'd take the figure from the map and then adjust it by a certain %age. The same theory as O2 adjustment.

I do question how well a knock sensor or an O2 sensor would work on a diesel though and wonder if you're perhaps making it a bit overcomplicated.

Aside from the knock sensor, you shouldn't need to touch the code. Everything is set up through a nice UI on a laptop. If you do come to tweak the code, it's written in ASM and comes with a command line compiler.

There's also the option of using MSII - which is a daughterboard that fits into the CPU socket on a MS PCB. It runs C which you might find more user friendly.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

We use cookies to ensure you get the best experience. By using our website you agree to our Cookie Policy