2 stroke thread (with occasional F1 relevance!)

[Pinger]

Hydrogen implies an exhaust cam and a supercharger.

Pitifully low rpm limit due to valve train constraints for anything smaller than the scale of ocean going liner engines.

How could a transfer style crankcase be lubricated?

External main bearings and metered feed to big ends via crank drillings.

[Pinger]

The simple solution to much of that is DI. Injecting H2 after exhaust closure ensures the VE (air) is comparable to petrol.

Anything but simple.
Firstly no injector large enough to handle the gas flow without failing due to the heat absorbed (and lack of cooling a liquid fuel would provide) by its bulk exists.
Secondly, injecting a gas against rising cylinder pressure is a metering nightmare.

[Rodak]

Secondly, injecting a gas against rising cylinder pressure is a metering nightmare.

Yeah, I was wondering about that. What sort of hydrogen mass/volume would have to be injected into a cylinder? Obviously it would vary with cylinder size, etc. but it seems it might be difficult on a higher revving engine. It's not too difficult to meter the injection of a non-compressible liquid fuel, but a gas is another story.....

[gruntguru]

The mass required for stoichiometric operation (there is no point running any richer than stoich) is 1/34 of the mass of air. If the hydrogen is added at STP the volumetric AFR is 3.3:1 ie 30% of the charge volume is hydrogen. Power loss due to the 30% reduced air/cycle is only about half this number ie about 15% due to other advantages of hydrogen fuel. It is also possible to add the hydrogen at a very low temperature since it has to be expanded from a very high storage pressure (Toyota Mirai stores at 700 bar - 70 bar after 90% of capacity consumed). The low temperature would reduce the charge volume displaced by hydrogen.

Direct injection is not an insurmountable challenge. Again, very high pressures and low temperatures are available at the fuel rail.

Cylinder pressure does not present a metering challenge. Flow through the direct injector is choked (supersonic) and insensitive to backpressure until cylinder pressure exceeds 50% of injection pressure. At say 30 bar injection pressure, injector flow would remain constant until cylinder pressure exceeded 15 bar (which is about time for ignition).

Lots of good info here.
https://www1.eere.energy.gov/hydrogenan ... cm03r0.pdf

[Tommy Cookers]

surely no-one would now have a hydrogen engine fuelled to stoichiometric ?

for higher efficiency and low NOx it would be designed around heat dilution ie to run eg at 2 (and 3, 4 etc) Lambda
using very high boost (ie 2 stage chargers)
like current F1 but a big step beyond
(and lean running might allow fuelling in gas not liquid form)

for most applications hydrogen's big thing is more the outstanding 'mixture' range than the outstanding heat value

what I didn't say some days ago is that .....
cyclic hydrogen combustion can produce oxides other than the dihydrogen oxide better known as water
these other oxides are presumed to be short-lived in air but anyway would be far fewer with lean fuelling

[coaster]

An electronic engineer i knew took carbon rods from an dolphin torch battery and used them as electrodes in a sealed water vessel, a bridge rectifier fed dc current to the electrodes from 220vac mains.
It directly ran a 180cc side valve briggs and stratton, the motor ran but overheated.
Given hydrogen is so easy to create on demand, lets see some small capacity bench motors pushed to the limit?

[Big Tea]

An electronic engineer i knew took carbon rods from an dolphin torch battery and used them as electrodes in a sealed water vessel, a bridge rectifier fed dc current to the electrodes from 220vac mains.
It directly ran a 180cc side valve briggs and stratton, the motor ran but overheated.
Given hydrogen is so easy to create on demand, lets see some small capacity bench motors pushed to the limit?

You can create it easily with things as simple as washers or razor blades and a lemonade bottle with solar panels or battery, but once the bottle is full compressing and storing is not as easy as dealing with air.

[Pinger]

The mass required for stoichiometric operation (there is no point running any richer than stoich)

There might be given that NOx reduces slightly above stoich. Possibly acceptable/preferable for full load operation but the very high NOx region between phi > 0.8 and phi = 1.05 would have to be avoided.

Power loss due to the 30% reduced air/cycle is only about half this number ie about 15% due to other advantages of hydrogen fuel.

Why so? Is it due to the difference in specific heat ratio relative to liquid HCs?

It is also possible to add the hydrogen at a very low temperature since it has to be expanded from a very high storage pressure (Toyota Mirai stores at 700 bar - 70 bar after 90% of capacity consumed). The low temperature would reduce the charge volume displaced by hydrogen.

What kind of temperature drop do you envisage?

Direct injection is not an insurmountable challenge. Again, very high pressures and low temperatures are available at the fuel rail.

AFAIK there is no existing injector suitable, Because its bulk (due to having to pass a gas) exposes it to combustion heat it cannot withstand. Additionally, whatever is deployed as control 'pintle' will inevitably be large and thus heavy with probable limitations on opening and closing times which will limit rpm potential. At a minimum, you'd be looking at solenoid valve systems (ie, camless) for hardware more than to conventional injector technology.
If you were basing control on pressure variation (as opposed to fixed pressure and variable duration) a very fast acting pressure regulator will be required and the volumes between it and the injectors minimised.

Cylinder pressure does not present a metering challenge. Flow through the direct injector is choked (supersonic) and insensitive to backpressure until cylinder pressure exceeds 50% of injection pressure. At say 30 bar injection pressure, injector flow would remain constant until cylinder pressure exceeded 15 bar (which is about time for ignition).

Injecting all the way to ignition time will likely result in insufficient mixing time (which cannot be augmented by turbulence as that would overly increase combustion speed) and rich spots will ramp up NOx.


I'm not saying it can't be done, but what is needed straight away is a viable injector and control strategy which will end up being complex and expensive and ICE on hydrogen's key USP over a fuel cell will be its lower manufacturing cost. Fuel cell costs will likely fall so an overly complex and expensive hydrogen ICE won't gain traction. Avoiding DI would simplify and reduce cost dramatically.

[Pinger]

surely no-one would now have a hydrogen engine fuelled to stoichiometric ?

Possibly for (limited) full load operation where NOx falls again - in the same way that rich is deployed at full load with gasoline to suppress detonation.

for higher efficiency and low NOx it would be designed around heat dilution ie to run eg at 2 (and 3, 4 etc) Lambda
using very high boost (ie 2 stage chargers)
like current F1 but a big step beyond

How much boost can be utilised before NOx becomes problematic? Diesel is lean burn but is plagued with NOx due to the pressures deployed. Temperature is only half of the NOx problem. Pressure is the other.

(and lean running might allow fuelling in gas not liquid form)

Hydrogen in this context will always be gaseous.

for most applications hydrogen's big thing is more the outstanding 'mixture' range than the outstanding heat value

There's a danger of overstating it. 180:1/34:1 = 5.3. A turn down ratio of 10 is more common on existing engines. If the highest load was with twice air compared to stoich (ie, 68:1) then the available turn down ratio is 2.6. Still helpful in reducing throttling losses on a 4T but increasing pumping losses on a 2T.

what I didn't say some days ago is that .....
cyclic hydrogen combustion can produce oxides other than the dihydrogen oxide better known as water
these other oxides are presumed to be short-lived in air but anyway would be far fewer with lean fuelling

The bigger concern - with 4T at least - is the above finding its way to the crankcase. Which supercharging will further aggravate.

[Pinger]

There is - despite what I said previously - a direct injector capable of handling gas - CNG in this case here >> https://hal-ifp.archives-ouvertes.fr/ha ... 6/document
I haven't read it right through as yet but its focus is on increasing turbulence as GNG is slow burning apparently. There is some mention of late injection creating turbulence due to the fuel 'plume'.

[coaster]

[/quote]You can create it easily with things as simple as washers or razor blades and a lemonade bottle with solar panels or battery, but once the bottle is full compressing and storing is not as easy as dealing with air.
[/quote]

He was producing on demand rather than storing, the system was not self sustaining.

[gruntguru]

surely no-one would now have a hydrogen engine fuelled to stoichiometric ?

Stoichiometric operation at full load and with DI would offer 20% higher power than the equivalent gasolene fueled engine so the reduced cost of NA hardware wul be attractive. Stoich operation also allows the use of a reduction catalyst - possibly with some technique to permit lean part-load without production of NOX in the cat.

Stoich full-load also enables the 5.3 load turndown mentioned by Pinger. Cylinder deactivation could be used to extend this without throttling.

[gruntguru]

Power loss due to the 30% reduced air/cycle is only about half this number ie about 15% due to other advantages of hydrogen fuel.

Why so? Is it due to the difference in specific heat ratio relative to liquid HCs?

Primary due to the higher heating value of air when burned with hydrogen. There are other factors like higher useable CR, higher flame speed, higher diffusivity (better mixing) and smaller quenching distance (burns closer to cylinder wall) which all contribute.

It is also possible to add the hydrogen at a very low temperature since it has to be expanded from a very high storage pressure (Toyota Mirai stores at 700 bar - 70 bar after 90% of capacity consumed). The low temperature would reduce the charge volume displaced by hydrogen.

What kind of temperature drop do you envisage?

Trick question? I had to look it up - and learned that Hydrogen has a negative Joule-Thomson coefficient at room temperature.

Cylinder pressure does not present a metering challenge. Flow through the direct injector is choked (supersonic) and insensitive to backpressure until cylinder pressure exceeds 50% of injection pressure. At say 30 bar injection pressure, injector flow would remain constant until cylinder pressure exceeded 15 bar (which is about time for ignition).

Injecting all the way to ignition time will likely result in insufficient mixing time (which cannot be augmented by turbulence as that would overly increase combustion speed) and rich spots will ramp up NOx.

Not suggesting that you would. High diffusivity of H2 helps mixing.

[J.A.W.]

While on 'H2', here below is a link to a reasonably recent Kawasaki 2-stroke development
design patent which appears to cross-reference both Toyota's attempts to make a DOHC 4V
design run as a 2-cycle, & Kawasaki's own 'blown' (supercharged) 'H2' 4T engine.

But don't get too excited, its a hybrid-electric 'range-extender' tuned for hi-efficiency,
narrow-band running akin to the prototype Audi machines racing in the the 'Dakar Rally'...


https://www.visordown.com/news/new-bike ... lane-crank

[J.A.W.]

&, for any members who've missed the obsessive 'mechanical marvel' posts by a certain, ahem,
Polish member.. check this 10cyl 2-row radial 2T homebuilt out, (& at least it runs, to be fair).

https://thekneeslider.com/bistella-500- ... ek-foltis/