Will the Flexi fuel vehicles able to run with methanol 3

[Azmio]

Hey guys,

I know that FFV can run easily with ethanol, will it be possible to run FFV with M85 or M100 too?

 

[dgallup]

Not for long.

 

[BrianPetersen]

Methanol doesn't like staying mixed with gasoline.

 

[patprimmer]

You would need considerably more fuel supply capacity and the corrosion of metals and chemical attack on seals will be more severe. If the E85 engines have the right materials in the fuel system, which is quite likely, then so long as the fuel system will supply enough for peak demand, then no problem.

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.

 

[Azmio]

back in the mid 90s, the Big 3 came up with many M85 compatible cars for California. I dont know what happen next but the program just simply died.

AS for the E85, whichever in contact with fuel will not be made out of plastic or aluminum but preferably steel. The question still remains, will the current FFV built for ethanol capable of running with methanol?

 

[patprimmer]

I would think so, but I don't know so without seeing the materials used.

Even mild steel will corrode in contact with methanol or ethanol, but it corrodes faster in methanol as the methanol absorbs more water more quickly.

If rust resistant steel parts are used in contact with the fuel, it should be OK

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.

 

[drwebb]

Are you telling me that California- the state that led the caniption fit over MTBE in gasoline- considered methanol as a practical fuel???? Sounds pretty cryptic to me.

 

[Azmio]

webb,

well, unless it is possible to delete history, they did have fleets of methanol powered cars back in the 93s. It was a big thing during that time, try search the internet and you will many reports from the programs.

 

[dgallup]

Methanol is dead as a flex fuel these days. Modern seals, plastics & steels used in fuel pumps, pressure regulators and injectors are NOT designed for methanol capability.

 

[TDIMeister]

Even if the fuel system materials were fully compatible with methanol, existing Flex-fuel vehicles may not operate properly or optimally with methanol content in the fuel. The reason may not be what you think it is.

The current method of detecting the quantity of ethanol in the fuel is via a sensor placed in the fuel line that measures the impedance or dielectric constant of the sampled fuel.

http://www.google.com/url?sa=t&ct=r...G0b2EGty1hTbqxOjw&sig2=6uLFjPf1_ctkz9uIZSThzg

Gasoline has a range of dielectric constants due to its mixed composition, but it is in a fairly narrow range. Ethanol has a completely different dielectric constant, and by measuring this dielectric constant of the blended fuel and the known flow rate, it is possible to determine the mass fractions of ethanol and gasoline.

A somewhat simpler and less sophisticated method is to measure the change in stoichiometry at the oxygen sensor. Gasoline has a known stoichiometric AFR (~14.5-14.7:1); ethanol 9:1. If you know the fuel mass flow rate and the stoichiometric AFR, you can determine the mass fractions of the blended fuel.

The problem with methanol is that it has both a different dielectric constant AND stoichiometric AFR than ethanol. Remapping the respective values can remedy this, BUT you have maps holding ONE tranfer function, not more, so if you remap for methanol, you lose the transfer function for ethanol.

In a true Flex-fuel car to be able to also work with methanol in addition to ethanol, you have to now consider TERTIARY blends of gasoline, ethanol and methanol, and this cannot be addressed by remapping alone.

One solution would be to sense BOTH AFR and dielectric constant, and instead of 3-dimensional maps (fuel rate; dielectric constant OR stoichiometric AFR; mass fraction for two fuels), you have a 4-D map (fuel rate; dielectric constant AND stoichiometric AFR; mass fraction for 3 fuels).

Obviously, this is not a plug-and-play change...

This has further implications of truly optimizing engine operating parameters for the benefits of ALL the fuels used in their respective percentages over the entire operating range, but this adds yet another variable to be considered and optimized.

 

[Azmio]

Orbital published an interesting paper of how they simply picked any passenger car in the market and run it with ethanol. From the report, at idle and part load, the ECU automatically adjusts the AFR ratio using the oxygen sensor.

The cars worked well in most conditions other than full load and engine start up. During these two conditions the ECU switched to open loop causing the ECU to look into the mapping table rather than to receiver feedback from the oxygen sensor

 

[Lou Scannon]

I was involved with methanol in California in the '90s. The flex-fuel vehicles started appearing in the late 80's. The state government provided incentives for methanol fuelled vehicles. Methanol fuel was seen at the time as a path to significant reductions in ozone precursors from engine exhaust, and as an economically viable fuel source assuming mature technology and infrastructure. When the original economic and infrastructure projections proved overly optimistic, and aldehyde emissions came under scrutiny, the appeal of methanol fuel faded quickly.

"Methanol doesn't like staying mixed with gasoline."

Pure methanol and gasoline are quite miscible. The problem arises when water is included in the blend. Water can and will enter the fuel through normal weathering processes, and will go into solution due to the co-affinity between water and methanol. However, when the water content goes above a certain level, the solution will separate into 2 or three different phases. Engine operation will tend to deteriorate when this occurs.

 

[TDIMeister]

The O2 sensor by itself tells the ECU nothing about the actual AFR but the lambda value, or rather simply a binary yes/no whether the air-fuel mixture is at lambda 1 or not, because the zirconia piezo-electric element in the O2 sensor has a very strong step-function response within a very narrow window of lambda around 1 (wide-spectrum O2 sensors are a different breed).

Getting the stoichiometry correct only tells you (the ECU) that you have the fueling quantity correct, but as you said, is useless at full-load when the mixture is enriched beyond stoichiometric or in lean-burn engines operating at part load.

To determine the actual AFR at closed-loop lambda=1, you need to know also know the instantaneous mass flow rate of the fuel AND air.
mass air = mass fuel * lambda * stoichiometric AFR

Actually, you can see that we have two unknowns, requiring two equations, but you can use use MAF sensor readings or time histories to come up with the other unknown.

Getting the amount of fuel correct is only one part; knowing the precise composition of the blended fuel has other optimizing implications. For example, if you know precisely how much gasoline, EtOH and MeOH you have in your fuel, you can also optimize the ignition timing maps, which you cannot do with only a knowledge of stoichiometry alone.

Additionally, you might put some intelligence into your ECU as well, for example knowing how different fuels react in cold start, and optimizing the measures accordingly. Or by knowing precisely the composition of the fuel, you can adjust the fuel pressure in the accumulator rail to maintain a constant injector duty cycle, etc., etc.

 

[mattsooty]

TDIMeister

In actual fact what you have described is a binary lambda sensor as opposed to the more modern wideband sensor.

Wideband lambda sensors have an integral O2 pumping circuit which allows, within reason, a meaurement of actual lambda. As opposed to the rich/lean output of a binary sensor.

One of the problems is that of fuelling adaptation. Whereby the ECU will implement a longterm trim factor to ensure that Lambda=1 is always achieved, useful in the prescence of air leaks etc. By using flexfuel or fuels of differeing stoich afrs these adaptions can quite easily reach the threshold of failure and the MIL will be illuminated.

MS


MS

 

[TDIMeister]

"In actual fact what you have described is a binary lambda sensor as opposed to the more modern wideband sensor."

Yes, I know, and I specifically distinguished both in my second post above (although I called it "wide-spectrum" instead of wideband).

I don't know that wideband sensors have now displaced the original sensors I described. I thought they are mostly only used in lean-burn gasoline engines (which don't exist in North America) and new Diesels with PM and NOx traps.

 

[mattsooty]

For recent and future emissions regulation compliance wideband sensors are pretty much de rigeur for feedgas afr measurement, however, cheaper binary sensor are still sufficient for post main & starter catalyst O2 storage and monitoring.

MS

 

[Azmio]

There is still no clear answer to my question. If an FFV car can run with E85, roughly if we blend 56% of methanol with gasoline, it will give the same AF ratio as E85.

With the FFV fuel rail and tank being designed and developed to withstand gasohol corrosive nature, I just wondering why it is not possible for the current FFV to run with M60?

 

[TDIMeister]

The answer seems quite clear to me. Even you get the stoichiometry correct (the lambda sensor would guarantee that at the very least, regardless of what fuel blend you run), the engine will not run optimally with respect to ignition timing, etc. for methanol. The reason is as far as the engine is concerned it's running stoichiometric, fuel be damned, and everything is hunky dory. I don't think it was ever said by anybody that methanol WILL NOT work in an existing unmodified ethanol FFV, rather that it won't be optimal.

If your FFV uses a dielectric constant ethanol sensor, it will get false readings for alcohol content in the blended fuel. The ECU could see what's going on with signals coming from the MAF-, fuel quantity- and Lambda sensors, and see that something doesn't jive in the stored maps with the signal from the ethanol sensor, and throw an out-of-range or implausible MIL fault. It might be fine on a SPECIFIC blend of methanol (e.g. M60 as you say), where the stoichiometry and other properties happen to coincide with an mapped ethanol/gasoline blend (it might not be specifically E85).

 

[patprimmer]

Azmio

Your question is not quite clear possibly due to fine meanings of words and inference. That is why some are saying what could possibly be done and others are saying what is done.

Do you mean can you run existinf FFV without modification on M50

Do you mean can you run an existing FFV with minimal tuning changes on M50.

Do you mean can you build an FFV with hardware components that could easily be produced if the demand existed to justify manufacture of the components suitable for M50 or M85.

Do you mean can the existing or new designs be made flexible enough to run petrol, E85 and M50 reliably and efficiently.

Different answers seem to have presumed some of these different options.





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.

 

[Azmio]

TDImeister,

"If your FFV uses a dielectric constant ethanol sensor, it will get false readings for alcohol content in the blended fuel"

Is it the alcohol or the oxygen content that the sensor is sensing and feeding it back to the ECU?

" the engine will not run optimally with respect to ignition timing, etc. for methanol"

Normally, ignition timing, load and rpm goes hand in hand in the mapping table. The knock sensor on the other hand will push it further to the brink of mild knocking to really push the ignition timing to the limit.

The current FFV available are intelligent enough to be runned with gasoline, E50 and E85. Can anyone explain how it is being done?