Maximum Torque Per Litre

[Pumaracing]

People's claims for this never cease to amaze me. I was browsing the web the other day and came across this.

http://www.cortina-mk1classifieds.com/forum/archiv...

Down towards the bottom someone is claiming they've seen 153 ft lbs from a 1660cc Ford Crossflow on a supposedly accurate engine dyno for god's sake. That's 92 ft lbs per litre from a 2v engine with one of the worst combustion chamber shapes ever made. God knows who calibrated that dyno.

Anyway my own views on the subject are made abundantly clear on my website in various places but to restate them you're doing well to beat 80 ft lbs per litre from 2v engines and 90 ft lbs per litre from 4v ones with modified road engines on pump fuel that are built to actually last more than 5 minutes.

I'd say the normal maxima are low 80s for 2v engines and about 93 from 4v ones.

In terms of my own engines over the years I've rarely had the chance to put them on an engine dyno and rolling road figures need to be taken with a big pinch of salt but FWIW my best figures are as below.

Rolling road, 2v engines:

Peugeot 205 1.9 rally tune 11:1 cr - 78 ft lbs per litre
Slick 50 Golf Gti 1800cc, race tune, 12.5:1 cr - 80 ft lbs per litre

Engine dyno, 4v engine

Peugeot 1600cc TU5J4 200 plus bhp, 12.5:1 cr - 93 ft lbs per litre.

Now there's an engine builder of similar 4 pot motors on another forum whose torque claims make the rest of us look like eejits. His 2v engines supposedly produce more torque than everyone else's 4v ones and his 4v ones are up to 97 ft lbs per litre. He seems to derive most of these numbers from wheel figures from a local rolling road scaled up by algorithms of his own devising to get back to flywheel ones. I'll say no more.

So I'd be interested in two things.

1) Any accurate engine dyno figures for maximum torque per litre you've come across or for engines you've built yourself. Rolling road ones too if it's a known set of rollers.

2) Answers to the V8 conundrum which is as follows. It's common to see torque per litre figures from mundane American V8s that far exceed what we normally get from small 4 pot 2v engines over here. Mid to high 80s ft lbs per litre is quite common for engine dyno results for Chevy V8s and the like on pump gas.

I have two thoughts on the matter. Firstly that perhaps the crossplane V8 crank arrangement makes some difference to the pulse tuning compared to flat plane 4 pot cranks. Secondly that maybe it's a matter of cylinder size. Big cylinders are harder to fill and obtain high bhp per litre figures from but they have a much larger ratio of volume to surface area than small cylinders and therefore possibly better thermal efficiency. i.e. less heat is lost to the cooling system. Maybe this helps the peak torque somewhat.

I'd be interested in any theory on the subject.

[Pumaracing]

Nick1point9 said:

I have however noticed a slight mistake with my maths, but you have quoted me now!

(I think) it should be torque=BMEP*displacement/2

That's correct for a 4 stroke engine. For a 2 stroke it's just BMEP*displacement. There's obviously also a constant in there depending on what units are being used.

[Pumaracing]

Stan Weiss said:

Any I the only one that likes to look at this using BMEP?

Stan

Yes. Given that BMEP is directly proportional to torque per litre it serves no different purpose but is a PITA to calulate unless you can remember the constants to use off the top of your head.

[Pumaracing]

Stan Weiss said:

With most of the engines in the states listed in ci I have the same PITA using litre. Over here it is torque per ci.

Stan

Yeah well don't get me started on you bloody colonials and your reluctance to leave the 19th century and imperial measurements. It's high time we invaded you again, gave you a bloody good thrashing, annexed you back into the Queen's Commonwealth and dragged you kicking and screaming into the 21st century and stopped you being the only country in the world that can't cope with metric measurements.

While we're at it we could ban religion over there, abolish the Republican party which is near as dammit the same thing, release Sarah Palin back into the wild and then shoot the daft b1tch from a helicopter and stick a broom handle ten inches up Newt Gingrich's arse so he knows what his first two wives felt like after he shafted them.

[Pumaracing]

Stan Weiss said:

Dave,
Please do not hold back and tell me what you real think.

The highest developed N/A 2 valve engine that I know of is the 500 ci NHRA Pro Stock engine. I am not sure how old these numbers are. Torque 840 @ 8100 / HP 1432 @ 9700 - If my conversion is correct 102.5 litre

Check it out on the dyno - http://youtu.be/rKH2RWZdZhM

Stan

They are certainly impressive but they don't use pump fuel. They run on racing fuel with octanes around 120 and CRs of 17 or so. We couldn't possibly build anything like that for normal classes of racing or fast road use. They also don't have to last very long which was another of my conditions. However even allowing for all that it's a pretty staggering torque per litre.

Do you have any views on the cross plane v flat plane crank issue?

[Pumaracing]

Huff said:

The current Yamaha R1 has an inline 4 with a crossplane crank:

http://www.ashonbikes.com/cross-plane_crank

Wow! What a fantastic article. That's an avenue of engine theory I'd never thought about before.

What I would like to know is the magnitude of the variation in instantaneous crank speed during a single revolution at a nominal steady rpm. I suspect there's also a further factor involved and that's piston ring friction. With a flat plane crank when all four pistons are at TDC or BDC not only have they transferred all their kinetic energy to the crank but also the friction levels are zero because they momentarily stationary. So when they accelerate up to maximum velocity again at half stroke not only do they have to absorb kinetic energy from the crank they also have to absorb whatever force it takes to overcome the rapidly rising ring friction which will slow the crank even further.

What the magnitude of the friction element of this is in terms of crank velocity variation per cycle I also have no idea.

Brilliant food for thought though. Thanks very much for posting this.

[Pumaracing]

Stan Weiss said:

I am not sure what formulas he used to come up with his numbers. Maybe it is an American trying to use metric.

Stan

An American using metric?? Oh you so funny

Next you'll be trying to tell us there are actually Americans who get the concept of irony. In your dreams.

[Pumaracing]

It's "maths" not math. Duh!

Anyway

SAN FRANCISCO MAN BECOMES FIRST AMERICAN TO GRASP THE SIGNIFICANCE OF IRONY - Jay Fullmer, 38, yesterday became the first American to get to grips with the concept of irony. "It was weird" Fullmer said. "I was in London for the first time and like, talking to this guy and it was raining and he pulled a face and said, "great weather, eh?" and I thought "wait a minute, no way is this great weather". Fullmer then finally realised that the other man's 'mistake' was in fact deliberate.
Fullmer, who is 39 next month and married with two children, aged 8 and 3, plans to use irony himself regularly in future. "I'm like using it all the time" he said. "Last weekend I was grilling steaks on the bar-b-cue outside and I burned them to sh!t and I said "great weather, eh?".

[Pumaracing]

Any comments on the cross plane crank deal Mr Vizard or whatever other reasons V8s seem to make more torque per litre than smaller cylinder flat plane inline 4s with 2Vs per cylinder?

[Pumaracing]

David Vizard said:

Actually 'no' on this one Dave. i have given it a lot of thought as to whether there is some advantage in the exhaust system or whatever to a two plane crank instead of a single plane on a V8. Unfortunately I don't even like my own theories on this so you would probably have a field day with them.

The ProStock guys have tried them and have failed to meet the power figures they already had. Now this does not mean an absolute point is being proved here but it does give us a good pointer on which to base ideas on.

Let's start the ball rolling on this one as I would like to know - or at least have a functional idea.

Let's have your in depth thoughts on this one Dave.
DV

On the crank itself I have nothing additional to what I posted at the start of the thread. On the torque figures in general I think there may be several reasons why the V8 does so well compared to most of our 4s.

Very few inline 4s I can think of have heads that can be modified really well and even fewer have both high flowing ports and a good chamber design. I can get an awful lot of airflow out of a Ford CVH head, the most of any head I've ever modified, but the hemi chamber kills any chance of a good burn. The Pinto with filled ports is not a bad compromise but even with a dry sump and all the other fancy bits thrown at it the engine in your Pinto book only just scraped over 80 ft lbs per litre and the Chevies seem to manage that without trying hard.

The Peugeot 205 and VW Golf have ok chamber shapes but aren't downdraft enough for really decent flow figures and you find a waterway if you try to get the Golf head to flow well. The Ford Crossflow has quite nice ports but a terrible chamber.

I suspect the sheer amount of development thrown at the Chevy engine over the years and the ready availability of custom made CNC ported heads with good wet flow and nice chamber design makes it much easier to achieve good power and torque than we can over here with modified OE heads on average 4 pot engines, none of which obviously have any custom made head castings available for them.

However even with all this factored in it should still be possible to get good torque figures from a 4 pot even if the outright power per litre is limited by head flow. Given that they seem to hit a brick wall at about 80 ft lbs I'm still left thinking the crossplane crank or the absolute cylinder size must be factors in the Chevy equation.

If cylinder size is a factor then you can hopefully tell us how BB Chevies compare to SB ones. Do they produce even more torque per litre or does it go down? Maybe we can at least eliminate that variable.