Tuning gasoline engine running diesel

[decipha]

I know this is quite odd so ill give specifics

For the past few years I've been programming a universal self learning engine control unit and the primary objective of this unit is that it is completely self independent with no user input required.

with that said, I'm in the process of writing of updating my flex fuel strategy and the thought popped in my head that if someone was to accidentally put diesel in a gasoline engine I want to be able to recognize this directly and adjust accordingly

Presently, I have it setup so that if there's still enough gasoline diluted in the diesel that the plugs can still ignite the mix it will simply go through the basic flex fuel strategy and adjust ignition timing and air/fuel ratio to get the engine running as efficiently as possible, this functions fine and there's no issues here.

The problem is that if there's too much diesel that the plugs can't ignite the mix I have to be able to calculate at which point to stop trying to fire the injectors since its hopeless in an effort to prevent washing down the cylinders and damaging the engine. Now this gets even more involved because one of the most common uses for this ecu is for methanol fueling which requires a significant amount of fuel as compared to gasoline.

I would do some trial and error testing but I hate to destroy an engine for a 'what-if' scenario that's unlikely to occur anyhow

If anyone has any insight or have done any testing like this I would greatly appreciate any thoughts, thanks in advance

 

[pmrobert]

Perhaps you could keep a running average misfire counter and when that hits a predetermined number either send an alarm, shut down or both. You could have a smaller number for an alarm init and if things are looking bleak then fire the shutdown algorithm.

 

[DanEE]

Flex Fuel Vehicle?
Misfire counter sounds like a good idea, especially if this is a fuel injected vehicle application. Fuel injected typically means remotely mounted (tank installed) electrically driven, high pressure fuel pump...

In the case of an accident, that high pressure pump needs to be shut down quickly otherwise a ruptured fuel line, or damaged engine fuel rail even to only one cylinder, etc would rapidly lead to a fuel dump on a hot engine and vehicle fire..

Many different techniques done to do this, but I know on some Toyota vehicles, six detected, successive misfires causes power to be cut to the remote fuel pump and of course engine shutdown.

Might this same criteria be used to protect your engine?

 

[Lou Scannon]

Not sure how extensive your control system is, but if it comprehends catalyst in and out temperatures there's sure to be some shenanigans going on there that you could recognize as soon as igniting the diesel in the cylinder becomes problematic.
I know this is all mostly hypothetical but while running successfully on a diesel mix, knock sensitivity may also be an issue you should take into account. I suppose the existing knock protection should cover this eventuality as far as protecting the engine is concerned, but the knock subsystem might provide additional data for your algorithm.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[decipha]

i have a misfire counter already in use for setting obd2 fault codes and adjusting coil dwell
as well as for calculating crossfires (this is a not so common problem I have already addressed due
to some odd exhaust header manifold designs) and a bunch other calculations

when excessive misfires are detected on a single cylinder i simply disable that one cylinder
if its global then the fueling is adjusted to try to tame the misfire, if it can't tame the misfires
then the engine is leaned out as much as possible and ignition timing is retarded slightly to
keep the engine running as safely as possible, the main objective is to keep the engine operating
though so if the ecu can keep the engine running it will unless a fail-safe is actived (such as loss of
oil pressure, etc..)

the ecu has an inertia switch on the case same as most vehicles have either in the trunk or in the kick panel
this works in conjunction with an accelerometer since alot of the vehicles that use these ecu's are drag
racing engines that can get triggered faulty when coming off the trans-brake during hard launches

i really dont want to kill the engine due to multiple misfires if it can still operate the engine successfully
i do have a fuel pressure sender that gives fuel pressure and temperature readings to the ecu, so a
ruptured fuel line isn't a concern of mine

i do have two egt sensor inputs but most of the vehicles using this ecu do not use it.

i'll give it some more thought, i really appreciate everyones input

 

[GregLocock]

Fascinating project, although fairly high up on my wtf scale. Anyaway I don't want to teach you your own business but "the engine is leaned out as much as possible and ignition timing is retarded slightly to keep the engine running as safely as possible," sounds like a recipe for burning exhaust valves to me.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[MikeHalloran]

Given an engine that's not too cold, I'd expect the Diesel to ignite without waiting for the spark, so retarding the spark may not be an effective strategy.

Given the structural damage that continuous preignition (can you detect that?) is alleged to do, I'd want to shut down the engine immediately, unless it's running a fire pump.




Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

What CR do you run?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[decipha]

It takes a significant amount of heat to split an exhaust valve, that's not even a concern for any engine operating at less than half of its pumping capacity, really isn't a concern once you factor in the cooling effect of the misfiring as well

retarding the ignition timing will put more of the combustion in the exhaust, this is done to prevent excess heat on the spark plug since the diesel will require a different heat range

Preignition is already calculated and if detected the engine will shut down, that's priority #1 in the failsafe strategy, that's already implemented and functions correctly

I've done a tad more research and it appears that ignition timing and fuel flow (when diesel isn't at high compression) is relatively similar between diesel and gasoline

Most of the engines operated by this ecu are factory production engines running gasoline and e85 between 8:1 and 11:1 static compression

 

[pmrobert]

The problem is that if there's too much diesel that the plugs can't ignite the mix I have to be able to calculate at which point to stop trying to fire the injectors since its hopeless in an effort to prevent washing down the cylinders and damaging the engine. Now this gets even more involved because one of the most common uses for this ecu is for methanol fueling which requires a significant amount of fuel as compared to gasoline.

How long can it run if there's too much diesel that the plugs can't ignite the mix? If this is operating a methanol fueled drag racer I would hope that there would be almost no chance that the end user would mistake gasoline/petrol or diesel for methanol. Sometimes you reach a point of diminishing returns trying to cover every single possibility - of which there are literally an infinite number of.

 

[ornerynorsk]

A bit out of my league, but is there a sensor that could be situated in the fuel tank that might detect the different make-up of fuel should diesel or kero be introduced?

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[decipha]

there is an input for ethanol content sensors, typically these sensors are mounted in the fuel feed line on factory production vehicles so equipped, however many manufacturers have stopped using them due to their high failure rate and cost, they have a hefty price tag of several hundred dollars (U.S. currency)

however, most factory ECU's infer ethanol content from air/fuel ratio makeup and knock detection to calculate an approximate ethanol content, this is fairly straight forward and simple to implement in a production vehicle with known variables

 

[Tekton]

A short story regarding a gasoline engine which happily ran on diesel - - for a while

A decade ago an acquaintance was driving a 1987 VW Golf 4 door, GF engine code, 1781 cc, 8.4 : 1 compression, Bosch K Jetronic injection - - in decent mechanical condition

Departing from Vancouver BC in December, he refilled the tank part way to Winnipeg, shortly thereafter noticing a slight decline in performance which he attributed to steadily dropping temperatures

Within 150k of filling, the engine low oil pressure warning light began to glow; he stopped and checked the dipstick, found the level was higher than normal, and proceeded to the next gas station where a mechanic advised the oil be changed and, having done so, sent him on his way

By the time he reached Winnipeg the oil pressure light was again glowing; he concluded the oil pump was at fault and duly replaced it

Family gathering over, he was heading back to Vancouver when the dreaded light lit up once more

He again checked the dipstick and, now it was daylight and noticing the level was high again, smelled it

Turned out the gas station attendant had pumped approximately 45 litres of diesel into the 60 litre tank

Extraordinary the engine ran satisfactorily (he was a "steady" rather than enthusiastic driver) and started, particularly given the sub zero temperatures

Granted that Northern Winter Grade diesel has a SG of around 0.82, but even so - -

How to guard against such with today's 10.0 : 1 and higher compression motors?

Decipha reports that fuel quality sensors are both expensive and problematic, and my own immediate thought of a specific gravity sensor won't work either as the 3 diesel variants span 0.81 to 0.96 @ 60F, the lower of which is very close to gasoline

Turboed gasoline engine egt's in particular register wide temperature variations, and it's unlikely a low percentage of inadvertently added diesel would cause the temp to register above or below accepted gasoline fluctuations unless the vehicle was being pushed hard

Definitely a tricky challenge


Tekton

 

[decipha]

thanks for that it helps me out greatly

at this point i'll just leave it be and let the ecu learn as it would any other fuel change, as long as the vehicle can operate and it isn't dead on the side of the road, it will be satisfactory, i highly doubt anyone will fill up with diesel since the majority of the engines operated by this ecu are highly modified performance oriented engines, however it only takes that one time the wife decides to drive the vehicle