29 replies to this topic

[Spa65]

As a long-in-the-tooth follower of Grand Prix racing I thought I'd ask a couple of questions in the forum as some of you guys are certain to know the answers. These date from the 60's.

 

First of all, how did they manage to downsize the Ford DFV engine down from 3 litres to 2.5 for the Tasman series? Was it a simple case of putting very thick liners inside the existing cylinders and adding comparably smaller diameter pistons? Anything else done e.g. to improve the fluid flow in and out of the cylinders? Or was something even more drastic done?

 

Secondly, how did they do the opposite, e.g Coventry Climax making a 1.5 litre unit into a 2 litre (or thereabouts) unit? I believe BRM and Ferrari did something similar as well. I can't imagine these engines being capable of being bored out to such an extent, and a larger throw crankshaft would seem impracticable.

 

Or maybe they realy did do one-off specially redesigned engines rather than modifying existing units 

 

I've only been curious about this for about 45 years. Put me out of my misery before I pop me clogs.

[Greg Locock]

If you've got to reduce the swept volume of an engine the easy peasy way is a short throw crank and longer con rods. If it is an interference engine that might cause problems, but other than that the resulting engine should be a bit of a sweety in comparison with the original one. You wouldn't design it that way as it is heavy and has a higher deck height than necessary. Ford Australia mucked about with a 3.2 derivative of the 3.9 several times, with a turbo it would have been economical and torquey enough. It revved nicely and was smooth.

[gruntguru]

The DFV spawned a number of derivations. In 1968; Cosworth created the DFV's first derivation, a 2,500 cubic centimetres (152.6 cu in) version for the Tasman Series, the DFW. DFV to DFW conversion simply involved substitution of a short-stroke crank and longer connecting rods. 

(Wikipedia)

[Catalina Park]

With a motor like the DFV if you tried to reduce the bore you would run into problems with amount of space required by the valves.
So reducing the stroke is pretty much the only option.

[Kelpiecross]

The 4-cylinder Climax engine was as big as 2.7 litres - this was really stretching the engine and they used tungsten balance weights on the crankshaft - tungsten having a density over 19.

[Lee Nicolle]

If you've got to reduce the swept volume of an engine the easy peasy way is a short throw crank and longer con rods. If it is an interference engine that might cause problems, but other than that the resulting engine should be a bit of a sweety in comparison with the original one. You wouldn't design it that way as it is heavy and has a higher deck height than necessary. Ford Australia mucked about with a 3.2 derivative of the 3.9 several times, with a turbo it would have been economical and torquey enough. It revved nicely and was smooth.

I have owned dozens of EA, EB Falcons and have never seen a 3.2. I think that engine was a myth!!
I was playing with speedway modified sedan engines in the past. I felt a 3.3 crossflow and turn it harder may have been a good thing. But nobody wanted to spend the money!

With the Cosworth what did they do when they made it bigger? There was a 3.9 from memory. I knew the 2.5 and 3 litre scenario. As a 3.9 it would have quite short rods and the pin very high in the piston. Neither ideal but it did seem ok in Sports Cars.

Edited by Lee Nicolle, 30 September 2013 - 10:21.

[Paolo]

If you've got to reduce the swept volume of an engine the easy peasy way is a short throw crank and longer con rods. If it is an interference engine that might cause problems, but other than that the resulting engine should be a bit of a sweety in comparison with the original one. You wouldn't design it that way as it is heavy and has a higher deck height than necessary. Ford Australia mucked about with a 3.2 derivative of the 3.9 several times, with a turbo it would have been economical and torquey enough. It revved nicely and was smooth.

 

So, in reverse, the easy way to increase swept volume would be to increase  stroke.

Of course this will cause increased mechanical stress, but the real question is: will it actually increase power?

From what I read, power depends largely on bore more than on stroke; a marginal increase in combustion quality can be expected , but is there anything else to be gained?

Would an hypothetical Cosworth with doubled stroke be substantially more powerful than a standard one, supposing the same max rpm can be met?

And, just supposing that mechanical stresses are not relevant here, would valve flow be a limiting factor in obtaining the same max rpm as the standard unit? 

Edited by Paolo, 01 October 2013 - 14:56.

[Lee Nicolle]

So, in reverse, the easy way to increase swept volume would be to increase  stroke.
Of course this will cause increased mechanical solicitation, but the real question is: will it actually increase power?
From what I read, power depends largely on bore more than on stroke; a marginal increase in combustion quality can be expected , but is there anything else to be gained?
Would an hypothetical Cosworth with doubled stroke be substantially more powerful than a standard one, supposing the same max rpm can be met?
And, just supposing that mechanical stresses are not relevant here, would valve flow be a limiting factor in obtaining the same max rpm as the standard unit?

[Canuck]

Ha!  It's like one can't get away from the topic!

 

Output of the engine is enhanced by an increase of displacement, all other things held the same so if you increase the displacement via bore or stroke, you will obtain more power.

Doubling the stroke and keeping the RPM the same would suggest that the original design was grossly under-stressed and over-built I would think.  At any rate, if this were possible, you would increase your energy output more or less linearly with the displacement increase, ignoring increased friction and similar losses, this assuming you managed to overcome the intake-valve's inherint restriction in trying to feed a large increase in displacement at the same RPM.

Increasing the bore allows you to increase the displacement (and valve size) without impacting piston speeds - more displacement, same piston speed, more horsepower specifically (or PS or kilowatts as opposed to a more generic interpretation of "power" when we're perhaps meaning energy).

 

An engine is nothing more than an air pump so anything that allows you to move more air in a given time, allows you to burn more fuel and release more energy.  If you double the displacement but leave the valvetrain - on the assumption that the valvetrain was optimized for the original displacement - you will most certainly run into a flow restriction that caps your output.

[Rasputin]

I wonder what Cosworth did exacly to first turn the 3.0 DFV into a 2.65 DFX and later to a 3.5 DFZ? 

[malbear]

Ha!  It's like one can't get away from the topic!

 

Output of the engine is enhanced by an increase of displacement, all other things held the same so if you increase the displacement via bore or stroke, you will obtain more power.

Doubling the stroke and keeping the RPM the same would suggest that the original design was grossly under-stressed and over-built I would think.  At any rate, if this were possible, you would increase your energy output more or less linearly with the displacement increase, ignoring increased friction and similar losses, this assuming you managed to overcome the intake-valve's inherint restriction in trying to feed a large increase in displacement at the same RPM.

Increasing the bore allows you to increase the displacement (and valve size) without impacting piston speeds - more displacement, same piston speed, more horsepower specifically (or PS or kilowatts as opposed to a more generic interpretation of "power" when we're perhaps meaning energy).

 

An engine is nothing more than an air pump so anything that allows you to move more air in a given time, allows you to burn more fuel and release more energy.  If you double the displacement but leave the valvetrain - on the assumption that the valvetrain was optimized for the original displacement - you will most certainly run into a flow restriction that caps your output.

The trouble that is eventually run into by increasing the bore as well as the valve size ( assuming that the displacement is constant) is combustion chamber shape. It virtually just becomes the cutouts in the piston for valve clearance and hence the surface to volume ratio becomes rediculously large( an orange peel with a lot of distance to the edge from the central spark plug) , so heat losses to the piston become large as well as combustion in the end gass slow or prone to detonation below 12000 rpm. efficiency and power losses become very significant. The best efficiency is about a cylinder volume of 400cc and a bore and stroke around the 80mm mark. best power with a bore around 90mm + with a stroke oround 50mm

Edited by malbear, 30 September 2013 - 20:30.

[Canuck]

No argument from me.  I started writing more and more but realized the rabbit hole could go on for some time so trimmed it up and left it awfully basic.  I do have a job to attend to after all ;-)

[Greg Locock]

I have owned dozens of EA, EB Falcons and have never seen a 3.2. I think that engine was a myth!!

 

I'm not too sure whether it ever actually got into production, My impression is that it was actually sold at one point, but nobody bought it. We used to build one every now and again as demo of what a more modern engine in the Falcon would feel like. The trouble is nobody except NVH was that keen on it.

[Kelpiecross]

I'm not too sure whether it ever actually got into production, My impression is that it was actually sold at one point, but nobody bought it. We used to build one every now and again as demo of what a more modern engine in the Falcon would feel like. The trouble is nobody except NVH was that keen on it.

The 3.2 Falcon engine certainly did get into production - I had an engine I bought from the wreckers. When I bought parts from the Ford dealer they initially told me there was no such thing as a 3.2l.
I like big straight sixes (not V6's) - the 3.2 is a great engine but the situation is the same as with the Jaguar XJ40 3.2 and 4.0 engines - the AJ 3.2 is a great engine but is the best part of a couple of seconds slower to 60mph - and the mileage is almost identical - so why would you buy one? Second hand 3.2l XJ40's are almost impossible to sell except for a pittance.

[gruntguru]

Ha!  It's like one can't get away from the topic!

 

Output of the engine is enhanced by an increase of displacement, all other things held the same so if you increase the displacement via bore or stroke, you will obtain more power.

Doubling the stroke and keeping the RPM the same would suggest that the original design was grossly under-stressed and over-built I would think.  At any rate, if this were possible, you would increase your energy output more or less linearly with the displacement increase, ignoring increased friction and similar losses, this assuming you managed to overcome the intake-valve's inherint restriction in trying to feed a large increase in displacement at the same RPM.

Increasing the bore allows you to increase the displacement (and valve size) without impacting piston speeds - more displacement, same piston speed, more horsepower specifically (or PS or kilowatts as opposed to a more generic interpretation of "power" when we're perhaps meaning energy).

 

An engine is nothing more than an air pump so anything that allows you to move more air in a given time, allows you to burn more fuel and release more energy.  If you double the displacement but leave the valvetrain - on the assumption that the valvetrain was optimized for the original displacement - you will most certainly run into a flow restriction that caps your output.

I can't believe you said all that without once uttering the "T" word!

[Lee Nicolle]

I'm not too sure whether it ever actually got into production, My impression is that it was actually sold at one point, but nobody bought it. We used to build one every now and again as demo of what a more modern engine in the Falcon would feel like. The trouble is nobody except NVH was that keen on it.

The 3.2 is listed in the owners handbooks up to EB2. I am told it was a throttle body engine like base 3.9s in EA, just shorter stroke with different crank and rods as usual with Fords.

[Paolo]

Ha!  It's like one can't get away from the topic!

 

Output of the engine is enhanced by an increase of displacement, all other things held the same so if you increase the displacement via bore or stroke, you will obtain more power.

Doubling the stroke and keeping the RPM the same would suggest that the original design was grossly under-stressed and over-built I would think.  At any rate, if this were possible, you would increase your energy output more or less linearly with the displacement increase, ignoring increased friction and similar losses, this assuming you managed to overcome the intake-valve's inherint restriction in trying to feed a large increase in displacement at the same RPM.

Increasing the bore allows you to increase the displacement (and valve size) without impacting piston speeds - more displacement, same piston speed, more horsepower specifically (or PS or kilowatts as opposed to a more generic interpretation of "power" when we're perhaps meaning energy).

 

An engine is nothing more than an air pump so anything that allows you to move more air in a given time, allows you to burn more fuel and release more energy.  If you double the displacement but leave the valvetrain - on the assumption that the valvetrain was optimized for the original displacement - you will most certainly run into a flow restriction that caps your output.

 

 

The trouble that is eventually run into by increasing the bore as well as the valve size ( assuming that the displacement is constant) is combustion chamber shape. It virtually just becomes the cutouts in the piston for valve clearance and hence the surface to volume ratio becomes rediculously large( an orange peel with a lot of distance to the edge from the central spark plug) , so heat losses to the piston become large as well as combustion in the end gass slow or prone to detonation below 12000 rpm. efficiency and power losses become very significant. The best efficiency is about a cylinder volume of 400cc and a bore and stroke around the 80mm mark. best power with a bore around 90mm + with a stroke oround 50mm

 

Thanks, I now see the error of my ways. The idea that an engine output is roughly proportional to the top surface of cylinders is justified by a little math I went to re-read to understand it better.

Increasing swept volume increases power for the same rpm, but if one does so by increasing bore he will face structural and fluidodynamic challenges, forcing to decrease max rpm.

In the end it can be shown that for a given technological level power is proportional to top surface, at least until bore becomes excessive. A lot clearer now.

I'll even utter the T-word: for given max power and max rpm,  better combustion in long bore engines should improve low rev-torque, shouldn't it? 

Edited by Paolo, 01 October 2013 - 10:09.

[Wuzak]

I'm not too sure whether it ever actually got into production, My impression is that it was actually sold at one point, but nobody bought it. We used to build one every now and again as demo of what a more modern engine in the Falcon would feel like. The trouble is nobody except NVH was that keen on it.

 

Wasn't the 3.2 used as a concept by Ford USA called T-drive?

 

The engine was mounted east-west, and drove the rear wheels. 

 

Edit:

http://www.drivingen...ive/default.htm

 

Apparently the T-drive concept could be 4, 6 or 8 cylinder in-lin engines, front, rear or all-wheel drive.

 

The straight 6 was, IIRC, the 3.2l I6 from Ford Australia, modfied for the job and with a cast aluminium block.

Edited by Wuzak, 01 October 2013 - 10:24.

[Spa65]

 

Secondly, how did they do the opposite, e.g Coventry Climax making a 1.5 litre unit into a 2 litre (or thereabouts) unit? I believe BRM and Ferrari did something similar as well. I can't imagine these engines being capable of being bored out to such an extent, and a larger throw crankshaft would seem impracticable.

 

Thanks for all the replies. It now seems pretty easy how they made the DFV smaller by changing the crankshaft throw and using different length con rods.

 

However I'm still confused about the situation regarding my second question above. I would have thought that the original 1.5 litre V8 engines had the cylinders too close to each other to allow the substantial reboring necessary. Otherwise why make the cylinder block so big in the first place? Not to mention additional cooling problems with less metal between the cylinders. Similarly using a larger throw crankshaft - why have that extra space that wasn't needed for the original crankshaft? Unless deliberately large crankcases gave less resistance to the rotating crankshaft with less oil getting splashed about in that direction. I'm making wild guesses - my expertise is restricted to rebuilding a Honda 250 4-stroke many times.

 

So how did they make them bigger? Maybe sneaking in both a slightly bigger bore and a slightly bigger crankshaft, but that would be almost a complete engine redesign. I'm as puzzled as ever.

 

Thanks for your replies in anticipation.

[Greg Locock]

Small increases in volume can be achieved by making the bores a bit bigger, siamesing them, and so on. But there isn't likely to be more than 8% or so in that in a reasonably modern engine.

 

So they must change the stroke.

[Canuck]

I can't believe you said all that without once uttering the "T" word!

Concerted effort...

[desmo]

There's an ad currently running on American TV for some huge pickup truck with Dennis Leary reading ad copy on the voice over that says "Torque is power".  

[gruntguru]

There's an ad currently running on American TV for some huge pickup truck with Dennis Leary reading ad copy on the voice over that says "Torque is power".  

If only somebody had told us that earlier - could have saved 1000 posts or more.

[Canuck]

Familiar...it wasn't Dodge was it?

[desmo]

No Ford, must have the same ad agency.

[Charlieman]

Thanks for all the replies. It now seems pretty easy how they made the DFV smaller by changing the crankshaft throw and using different length con rods.

 

However I'm still confused about the situation regarding my second question above. I would have thought that the original 1.5 litre V8 engines had the cylinders too close to each other to allow the substantial reboring necessary. Otherwise why make the cylinder block so big in the first place?

 

I wouldn't say that Cosworth had an "easy" job. They designed the DFV and DFW at the same time, providing the necessary clearances for the short stroke version. Additionally, there would have implications for lubrication and piston speed/acceleration to consider. Note that BRM did a similar exercise for the V12.

 

When stretching Climax engines for the 3 litre F1 regs, remember that technology had moved on a bit. Improved metallurgy, machining processes and casting techniques might have permitted a modest bore increase. However the Climax FWMV received a half inch stroke increase. Does anyone know what Paul Emery did to the Climax FPE to stretch it to 3 litres?

[bigleagueslider]

The Cosworth DFV was a 3.0L V8 they designed from scratch, so they could have used any bore/stroke combination they wanted to. The DFX was a 2.65L short-stroke version of the DFV designed for Indy Car, and the DFL was a 4.0L long stroke version designed for sports cars.

 

The long stroke DFL sports car engine was probably the most difficult design challenge for Cosworth.  Increasing the stroke of a flat-plane crank V8 resulted in serious dynamic vibration issues.

[Charlieman]

The DFX was a 2.65L short-stroke version of the DFV designed for Indy Car...

 

...developed, initially, by the VPJ/Parnelli Jones Indy car team. Few people in the UK at the time had considered how to pump such a volume of fuel through a racing engine.

[Peter Morley]

When stretching Climax engines for the 3 litre F1 regs, remember that technology had moved on a bit. Improved metallurgy, machining processes and casting techniques might have permitted a modest bore increase. However the Climax FWMV received a half inch stroke increase. Does anyone know what Paul Emery did to the Climax FPE to stretch it to 3 litres?

 

They fitted Jaguar pistons to the Climax FPE to make it 3 litres e.g. they increased the bore.

I used to have drawings and part numbers used by them at the time, but they all went with the Shannon and its FPE to Austria.

 

FPE used liners which made the conversion easier.

Given that the later FPF was effectively half an FPE and ran from 1.5 to 2.7 litres (admittedly larger engines had an improved block), Climax had obviously been pretty generous with the wall/liner thicknesses.

[Charlieman]

Thanks, Peter.