The Road to the 50% Thermally Efficient F1 Internal Combustion Engine

[johnny comelately]

In the centre and last photos, the round capped vertical standing part with three lines looks like a pressure regulator with a return to tank line, is that your view?
The first image with the line I can see why it could be considered fuel particularly as there is no evident fuel lines near the coils.....
And the single wires are for separate knock sensors as opposed to measuring ionic differences at the spark plug??

lets assume this a split unit method:
If the injectors were separate and postioned where you say they could be aimed upwards at the inlet of the prechamber, thus allowing the rich charge to be pushed by the piston effect into the prechamber.
Aiming an injector stream into airflow is an advantage for atomisation from aerodynamics and heat rise effect
Maybe you have hit on something here.

[gruntguru]

Do you think a relatively lean a:f mix in the prechamber (compared to cylinder charge) is feasible or even unavoidable? Residual combustion products must play some role in this, there won't be much evacuation of the prechamber during the exhaust stroke beyond pressure attempting to equalize.

No the pre-chamber needs to be richer than the main.

Too much fuss is made of scavenging combustion products out of the pre-chamber. In round terms the 18:1 CR will push 18 parts fresh charge into the pre-chamber to mix with one part residuals.

[vorticism]

In the centre and last photos, the round capped vertical standing part with three lines looks like a pressure regulator with a return to tank line, is that your view?

That's the high pressure pump. Looks like only two lines to me. The braided hose on top is the low pressure feed, the steel tubing is the output.

As for the fuel rail, look below the exhaust port covers on this 2017/18 Ferrari:



Compare to road car assemblies.

No the pre-chamber needs to be richer than the main.

Not sure. As long as it's not too lean to combust, it should still be able to provide jets due to combustion occurring in a fixed volume.

[johnny comelately]

Well, I think you have hit upon a good answer Vorticism.
A keen eye on those photos I think has solved the question of how the TJI works.
It was always a compromise with the Mahle original, having the injector huddled up in the prechamber with the spark plug.

This postulation provides a far, far better atomisation and still supplies an adequate mixture into the prechamber for ignition.

The jig saw puzzle continues

Note to editors : we need a thumbs up emoticon

[gruntguru]

Note to editors : we need a thumbs up emoticon

There is a rating button, top right corner of each post.

[gruntguru]

No the pre-chamber needs to be richer than the main.

Not sure. As long as it's not too lean to combust, it should still be able to provide jets due to combustion occurring in a fixed volume.

Main chamber at 1.4, pre-chamber leaner than this?
- Ignition reliability (spark in PC) would certainly suffer.
- Burn time in the PC would be variable.
- The papers on TJI mention the jets contain radicals which initiate combustion as they proceed, so it is not just heat and the jets are not "flames". I am not a chemist but I assume the PC mixture should be at least stoichiometric to maximise this effect.

[vorticism]

No the pre-chamber needs to be richer than the main.

Not sure. As long as it's not too lean to combust, it should still be able to provide jets due to combustion occurring in a fixed volume.

Main chamber at 1.4, pre-chamber leaner than this?
- Ignition reliability (spark in PC) would certainly suffer.
- Burn time in the PC would be variable.
- The papers on TJI mention the jets contain radicals which initiate combustion as they proceed, so it is not just heat and the jets are not "flames". I am not a chemist but I assume the PC mixture should be at least stoichiometric to maximise this effect.

As you say maybe intermittent aiming of the plume at the prechamber face is sufficient and I'm not giving enough credence to the possibility of the plume to stagnate vs disperse into the main charge before being entrained/squashed into the PC. Maybe they even have a recess or scalloped face in the direction of the injector to ensure trapping the plume.

This does make me think they could be using a moving prechamber door/face. The regs would not disallowing as it would be neither an intake nor exhaust valve. Operated by spring force, differential pressure, or camshaft. Open mode would permit a large opening for fuel entrainment, shut mode would be a hard stop wherein combustion could only escape through the TJI holes.

On some of the earlier engines they had the injectors and the spark plugs alongside each other. Begs the question of whether or not those were also TJI, only with the injector in prechamber to ensure a rich mixture. The main charge fuel then pulled through the TJI orifices during the intake stroke.

Yet on later engines we see this horizontal cross-cylinder injector arrangement, while still presumably TJI. Was it found that running the limit on prechamber richness was ideal? Perhaps a leaned prechamber and a richer main had the best effect.

[gruntguru]

As you say maybe intermittent aiming of the plume at the prechamber face is sufficient and I'm not giving enough credence to the possibility of the plume to stagnate vs disperse into the main charge before being entrained/squashed into the PC.

The plume doesn't need to be entrained. During injection I can easily see AFR in the centre of the main chamber at 0.5 or even richer. If that injection is timed to coincide with the period of maximum flow into the pre-chamber (90* BTDC - TDC compression) the AFR in the PC can be manipulated to be as rich as desired.

[vorticism]

I think at some point someone realized that method (what you're postulating) works better than injector in prechamber (when limited to one injector). Precombustion chambers (and by default some sort of ignition jet) have been around for decades; one example being high dilution gaseous fuel stationary engines. That being the case the smart guys in the engine room would have known this and thus prechambers were probably in use starting in 2014. However some started with putting the injector in the PC (image below), others, or perhaps later, they switched to injector outside the PC. The limit thus may have been getting sufficient fuel into the main charge through the prechamber orifices vs tuning injection profiles and CC geometry to permit a relatively lean PC ratio while optimizing the pre-ignition mix of the main charge. Or rather, it may have always been exterior to the prechamber, and it was only positioning that mattered, with horizontal winning out over vertical.

maybe some acronyms are in order

Outwardly fueled prechamber (OFPC)
Inwardly fueled prechamber (IFPC)

or

single injector outside prechamber (SIOP)
single injector inside prechamber (SIIP)

f.e. as it relates to above paragraph, 2014 Ferrari engine with fuel lines atop cam cover, distinct from Honda and Merc exhaust side rails. Whether or not this implies SIIP or SIOP I can't say. It does tell us that at the least horizontal placement won out over vertical placement.

[johnny comelately]

I think at some point someone realized that method (what you're postulating) works better than injector in prechamber (when limited to one injector). Precombustion chambers (and by default some sort of ignition jet) have been around for decades; one example being high dilution gaseous fuel stationary engines. That being the case the smart guys in the engine room would have known this and thus prechambers were probably in use starting in 2014. However some started with putting the injector in the PC, others, or perhaps later, they switched to injector outside the PC. The limit thus may have been getting sufficient fuel into the main charge through the prechamber orifices vs tuning injection profiles and CC geometry to permit a relatively lean PC ratio while optimizing the pre-ignition mix of the main charge. Or rather, it may have always been exterior to prechamber, and it was only positioning that mattered, with horizontal winning out over vertical.

maybe some acronyms are in order

Outwardly fueled prechamber (OFPC)
Inwardly fueled prechamber (IFPC)

or

single injector outside prechamber (SIOP)
single injector inside prechamber (SIIP)

The ignition point has to be adquately rich, particularly with these enormous pressures (being one of the main problems of top fuel engines and often what blows them up - failure to ignite at the right time)
Try starting any engine at 1.4 lambda and what do you get , at best a backfire.

Knowing that these units are indexed I imagine that would facilitate a peculiar hole placement in the prechamber.
If there was a fill hole that took the injector spray from the exhaust placed injector it could be tangentally placed and indexed to the spray thus facilitating the fill with rich mixture.
and that same hole would provide a jet flame that would create a circular flame around the combustion chamber along with the more centred holes (a picture says a thousand words if I could draw, which i might later)
Problem solved....somewhat

[gruntguru]

I think at some point someone realized that method (what you're postulating) works better than injector in prechamber (when limited to one injector). Precombustion chambers (and by default some sort of ignition jet) have been around for decades; one example being high dilution gaseous fuel stationary engines. That being the case the smart guys in the engine room would have known this and thus prechambers were probably in use starting in 2014. However some started with putting the injector in the PC (image below), others, or perhaps later, they switched to injector outside the PC.

The 2 injector approach as used in Mahle research is more accurate and shows good results up to lambda 2.0

Interestingly, in the early days of this research, a gaseous fuel (I think Methane) was used in the pre-chamber. This may have been due to difficulty obtaining an injector that would meter minute quantities of liquid fuel and also create a homogeneous mixture in the PC. Regardless there are a few papers showing results from both techniques (liquid and gas) and the gas in PC versions usually perform a little better.

[vorticism]

I think at some point someone realized that method (what you're postulating) works better than injector in prechamber (when limited to one injector). Precombustion chambers (and by default some sort of ignition jet) have been around for decades; one example being high dilution gaseous fuel stationary engines. That being the case the smart guys in the engine room would have known this and thus prechambers were probably in use starting in 2014. However some started with putting the injector in the PC (image below), others, or perhaps later, they switched to injector outside the PC.

The 2 injector approach as used in Mahle research is more accurate and shows good results up to lambda 2.0

Interestingly, in the early days of this research, a gaseous fuel (I think Methane) was used in the pre-chamber. This may have been due to difficulty obtaining an injector that would meter minute quantities of liquid fuel and also create a homogeneous mixture in the PC. Regardless there are a few papers showing results from both techniques (liquid and gas) and the gas in PC versions usually perform a little better.

I'll have to look into those. Modern piezoelectronic injectors may have opened new opportunities.

After all my pontificating I realized I left out one more option. The injector, even if the rail is on the exhaust side of cylinder bank, might still connect to a prechamber by routing it through the exhaust port web. Now, it's probably not ideal to overly thicken the web there to accommodate this, so you're probably still on the right trick; I thought it might be be worth mentioning though.

[johnny comelately]

Can you use this link please, can't quite get the upload image thing to work atm
https://www.deviantart.com/sherlockedsh ... 914218191/

I put the bore positioned injector in just to help the discussion

[johnny comelately]

This is the drawing for the insert version of what we did use
https://www.deviantart.com/sherlockedsh ... -732335564

[johnny comelately]