Because the combustion isn’t instantaneous, you have nitrogen and oxygen mixed at high temperatures and pressures.

Only if the mixture is poorly mixed, surely, as in the case of a direct injection engine (on purpose)? Otherwise the oxygen molecules will react with the petrol not the nitrogen?

Surely more air in the cylinder means more NOx not less?

Surely more air in the cylinder means more NOx not less?

NOx forms at very high temperatures. If you keep the combustion temperature low enough, it’s not a problem. There are roughly 4 nitrogen molecules per oxygen molecule in air. There are roughly 17 oxygen molecules per octane molecule in a stoichiometric mixture. There is much, much more nitrogen in there than anything else. If it gets hot enough, some of that nitrogen will form NOx before the oxygen atoms encounter a fuel particle to react with.

+1 for @thols explanation of NOx formation.

Lower temperatures drive NOx down. Excess air helps to do that.
That’s effectively what EGR does as well by the way – it reduces the peak temperatures.

NOx also only forms very locally – specifically around the flame front (not sure if it’s been covered above but the flame “progresses” through the mixture rather like watching a forest fire spread but a wee but quicker. Conditions at the flame front dominate emissions. So controlling charge motion and mixture formation helps too.

Conditions at the flame front dominate emissions. So controlling charge motion and mixture formation helps too.

How do you measure / visualise this? Or is it all CFD (etc) modelling?

Just to answer the OPs question. I have 8 diesel engines on the farm. 7 of them are pushrod OHV ones 😉

Conditions at the flame front dominate emissions.

Considering this how does the speed of the flame front affect this? Another benefit to H2 is the is has a significantly faster flame velocity than petrol which benefits efficiency and power and enables higher RPM’s, but what conditions at the flame front determine emissions? Is it just purely temperature or time or a combination of the two?

How do you measure / visualise this? Or is it all CFD (etc) modelling?

It’s a combination of:

CFD and other simulations
Theoretical hypothesising
Experimentation – both on single cylinder enginess and multi (single cylinders both cheaper and you don’t get the “averaging” effects to deal with)
Optical measurements – e.g. lasers embedded into cylinder heads etc
Occasional optical engines where cylinder walls, pistons, windows into the head can be made transparent.
There is a very cool youtube video from my old company from a project I was involved with that created a quartz (i.e. transparent) cylinder and piston with lens in the middle for optical visualisation of direct injection behaviour. Public domain so I can mention it.

Note all the the CFD needs correlating, hence the experimentation.

Oh – and reading this. It’s a standard textbook for any engine related degree course. Pretty much the bible.
Very mathematical, technical and all around combustion rather than the oily bits.

how does the speed of the flame front affect this?

It basically determines the optimal piston speed. This is one of the main factors why Diesels “run out of power” at higher RPMs.

Flame speed is a function of many things – charge motion (motion of the air in the cylinder) being a major one. By designing the ports, cylinder head and piston appropriately plus manipulating the valve events you can influence this considerably.

Np sign of the video but there’s a public domain marketing release here – apologies for the rubbish reader site.

Ricardo marketing publication – optical engine and analaysis

Brighton university marketing page talking about the optical engine

Brighton uni link

Flame speed is a function of many things – charge motion (motion of the air in the cylinder) being a major one. By designing the ports, cylinder head and piston appropriately plus manipulating the valve events you can influence this considerably.

This is what impresses me about modern engines. I was a petrolhead back in my youth, when you could pick up an old Escort or Datsun, take to the ports and combustion chamber with a dremel, throw in a cam, put on some sidedraft Webbers and a decent exhaust, and probably get double the stock horsepower. Problem is, they only ran well at a narrow rev band, so the stock engines ran well at low revs and the hot engines ran well at high revs. Carburetors didn’t help, they can get excellent peak power, but can really only be optimized for a narrow range of airflow.

I think the last time I had an engine apart was 1993 or 94, it was making a horrendous knocking noise. The noise was because the pistons were bouncing off the cylinder head. The driver was used to seeing the oil pressure warning constantly flickering due to oil surge, so just kept driving with no oil assuming that no oil pressure was normal. Opening the sump plug and nothing coming out was an early warning that something was amiss. I took that one apart, but there wasn’t anything to reassemble, it was utterly trashed. Very unsatisfying to have the final job being pulling an engine apart and throwing it in the bin instead of putting an engine together and seeing it fire up.

As much as I miss messing with engines, there’s more money in desk work than getting dirty fingernails, but modern engines amaze me. I drive a basic 1500 cc Toyota Yaris, which gets amazing fuel economy. It puts out more power than the old twin cam Lotus Escorts without all the temperamental nonsense of old engines. Engine designers today obviously understand much more about how combustion works and how to optimize it across different engine speeds. I’m betting you have bugger all chance of improving a modern engine with a dremel to the cylinder head and a cam grind.

I’m betting you have bugger all chance of improving a modern engine with a dremel to the cylinder head and a cam grind.

You might make some improvements on an NA engine with your dremel – smoothing the port and intake manifold walls out, as they’ll likely just be as-cast surfaces but not with the cams.

The exception to that is where tuners can ignore some of the restrictions that the original manufacturer had – cost and most commonly emissions, life and drivability compromises.

(Or just stick a turbo on it…. 😉 )

Or just stick a turbo on it….

Yeah, I worked with a guy who built truck racing engines. The biggest problem they had was keeping the intake system from bursting. They’d just keep cranking up the boost, but the manifolding and gaskets couldn’t handle the boost pressure so it was like the Dutch boy fingering the dykes.

Edit: I gather from watching YouTube videos that they’ve solved the intake bursting problem and now they’re dealing with the exploding crankcase problem. Ain’t gonna solve that one by doubling up on hose clips.

One more.

modern engines amaze me

And yet people still complain about them and wish for the ‘good old days’.

Now that Le Mans is running right now, I just remembered that Chevy pushrod engines have won a bunch of class victories in sportscar racing. For example, 2015 Le Mans 24 Hours, first in class ahead of a bunch of Ferraris and Porches. Not to mention the ALMS series. That big Chevy engine doesn’t seem to do so bad for an old pushrod design.