Combustion System

[Brian Coat]

Thanks for posting.

The two injector version of TJI was filed as a patent as long ago as 2011 so there has been plenty of time for other race engine teams to try to make it work with one injector ... and for engineers to move about a bit.

There was a Cosworth simulation paper in RET in 2013 where they spoke of a novel knock mitigation method but this was undisclosed and of course they never landed a contract.

[J.A.W.]

An example of commercially available combustion testing equipment, http://www.tfxengine.com/index.html

[gruntguru]

Article came out today regarding TJI and the DI sitting in the main chamber and pre chamber fuelled by compression stroke.
http://www.motorsportmagazine.com/opini ... ine-update
Interesting that the link says ideas percolated from a core group of Cosworth engineers to Mahle and Mercedes AMG HPP. All 3 companies are roughly the same area close to each other in Northampton.

This article confirms what I suggested in the V6 thread:

"I recall at least one paper by Attard et al where the pre-combustion chamber did not contain a supplementary injector - operating instead with the same mixture as the main chamber (stoichiometric in the paper (a Rotax engine IIRC)). The system still offers the benefit of rapid combustion although not the ultra-lean AFR.

Such a system could conceivably operate at say lambda 1.2 - 1.4 in the pre-chamber, with DI stratification producing a much leaner mixture in the regions further from the pre-chamber. So - only one injector per cylinder."

Mixture strength in the pre-chamber would be heavily dependent on spray direction and controlled to some extent by the timing of the injection pulse.

[NL_Fer]

Isn't the injectorless pre chamber older than the version with?

Like the pre chamber sparkplugs?

[Brian Coat]

GG: You are right, the Rotax test paper is SAE 2013-01-1629 and Rotax 914 normally has no injector at all ...?

This MSU test paper shows some nice photos of TJI combustion, albeit with a different arrangement to that claimed to be used in F1.

http://www.egr.msu.edu/zhug/Conference% ... zation.pdf

[Brian Coat]

Isn't the injectorless pre chamber older than the version with?

Like the pre chamber sparkplugs?

Yes, some of the stuff from the likes of Hoerbiger is older but was the jet concept properly established in these? And they appear to have been aimed at lower speed CNG engines?

[henry]

If they are using a high geometric compression ratio with late inlet valve closing might some of the mixing for the very lean element of the stratified charge be taking place outside the cylinder in the intake plenum?

[J.A.W.]

Article came out today regarding TJI and the DI sitting in the main chamber and pre chamber fuelled by compression stroke.
http://www.motorsportmagazine.com/opini ... ine-update
Interesting that the link says ideas percolated from a core group of Cosworth engineers to Mahle and Mercedes AMG HPP. All 3 companies are roughly the same area close to each other in Northampton.

This article confirms what I suggested in the V6 thread:

"I recall at least one paper by Attard et al where the pre-combustion chamber did not contain a supplementary injector - operating instead with the same mixture as the main chamber (stoichiometric in the paper (a Rotax engine IIRC)). The system still offers the benefit of rapid combustion although not the ultra-lean AFR.

Such a system could conceivably operate at say lambda 1.2 - 1.4 in the pre-chamber, with DI stratification producing a much leaner mixture in the regions further from the pre-chamber. So - only one injector per cylinder."

Mixture strength in the pre-chamber would be heavily dependent on spray direction and controlled to some extent by the timing of the injection pulse.

& Rotax does build the SI 2T ETEC engines which run the needfully rapid response 'voice coil' type DI (& HCCI )..

[J.A.W.]

If they are using a high geometric compression ratio with late inlet valve closing might some of the mixing for the very lean element of the stratified charge be taking place outside the cylinder in the intake plenum?

Given the even firing V6 interval, common plenum, & turbo back-pressure modulation?

( Well, either that or an M-P "African swallows are non-migratory" analogy.)

[gruntguru]

I would be surprised if LIVC is operated as a strategy at full-power/high-rpm. The reduction in cylinder filling would increase the boost and compressor power requirement. The uptake of variable intake trumpets confirms that the designers are using wave tuning so what would be the point of pushing some charge back out?

[R_GoWin]

This article confirms what I suggested in the V6 thread:

"I recall at least one paper by Attard et al where the pre-combustion chamber did not contain a supplementary injector - operating instead with the same mixture as the main chamber (stoichiometric in the paper (a Rotax engine IIRC)). The system still offers the benefit of rapid combustion although not the ultra-lean AFR.

Such a system could conceivably operate at say lambda 1.2 - 1.4 in the pre-chamber, with DI stratification producing a much leaner mixture in the regions further from the pre-chamber. So - only one injector per cylinder."

Mixture strength in the pre-chamber would be heavily dependent on spray direction and controlled to some extent by the timing of the injection pulse.

With the spark plug inside the pre-chamber, this must have been a challenge to avoid misfires at very lean operating points. (and would be accentuated if purging of the prechamber does not happen easily, and traps burnt species)

I can get my head around how the TJI as described in the Mark Hughes article would work in homogenous mode. And like you say about the rotax engine with stoichiometric mixture. But for part load conditions or other operating points where the main chamber by itself is lean, it could become a challenge to achieve suitable AFR for ignition to initiate inside the pre-chamber. So in that way, does this architecture of TJI end up limiting how lean you can go?

Agree that its heavily dependent on spray direction. I guess also maybe the shape of the piston too - to direct the vortex from tumble flow to feed the pre-chamber.

[NL_Fer]

It is the whole mistery of current f1 engines. How to get a lean main combustion chamber, richer pre chamber and all with a single injector.

But injection timing, steering and swirl are most likely to do the trick.

Also interresting, are they reaching auto ignition under certain certain load levels?

[Frank_]

if the supposed perforated thimble/pre-chamber hangs down into a piston bowl (smaller than this) then its positioning in relation to the inlet valves might ensure enough swirl and purging ?

[gruntguru]

This article confirms what I suggested in the V6 thread:

"I recall at least one paper by Attard et al where the pre-combustion chamber did not contain a supplementary injector - operating instead with the same mixture as the main chamber (stoichiometric in the paper (a Rotax engine IIRC)). The system still offers the benefit of rapid combustion although not the ultra-lean AFR.

Such a system could conceivably operate at say lambda 1.2 - 1.4 in the pre-chamber, with DI stratification producing a much leaner mixture in the regions further from the pre-chamber. So - only one injector per cylinder."

Mixture strength in the pre-chamber would be heavily dependent on spray direction and controlled to some extent by the timing of the injection pulse.

With the spark plug inside the pre-chamber, this must have been a challenge to avoid misfires at very lean operating points. (and would be accentuated if purging of the prechamber does not happen easily, and traps burnt species)

I can get my head around how the TJI as described in the Mark Hughes article would work in homogenous mode. And like you say about the rotax engine with stoichiometric mixture. But for part load conditions or other operating points where the main chamber by itself is lean, it could become a challenge to achieve suitable AFR for ignition to initiate inside the pre-chamber. So in that way, does this architecture of TJI end up limiting how lean you can go?

Agree that its heavily dependent on spray direction. I guess also maybe the shape of the piston too - to direct the vortex from tumble flow to feed the pre-chamber.

We know it is possible to produce a stratified charge using direct injection. No doubt the PC is located in the centre of the "rich" zone - ie where most of the fuel spray is directed. Early injection will tend to mix throughout the chamber to create the lean, main charge. A late injection pulse, timed to arrive at the PC when the gas flow into the PC (pushed through the orifices during compression stroke) is at its maximum (perhaps about 30-40 deg BTDC) will add extra fuel into the PC and the surrounding rich zone.

[R_GoWin]

Its interesting that Attard has moved on from MAHLE has been working for the Fiat Chrysler group since 2014. I wonder how much of technical interaction the chaps at Maranello have within the FCA group. AMG HPP and the Brackley guys have made it known a few times about leaning in on the expertise within Daimler.