20 replies to this topic

[blueprint2002]

As everyone in this forum knows, Renault introduced the turbocharged engine to Formula 1 in the late seventies, with a 1500cc version of their Le Mans-winning 2-litre V6, soon showing that the turbo was the way to go. Apart from BMW, Hart and Alfa Romeo, most of the competition followed suit with their own turbo V6s, and very successful many of them were. BMW and Hart had sound, but different, reasons for using turbo L4s; only Alfa Romeo built a turbo V8, this last without success.

Concurrently, the Cosworth DFX Indianapolis turbo V8 was developed, based on the legendary DFV, and was just as successful as its parent. In fact, the Indianapolis brigade had a ten-year start with successful turbo engines, both the Offenhauser L4 and the “four-cam” Ford V8 dating back to the late sixties.

So it is interesting that Ferrari, Honda, TAG and Ford (Cosworth) all opted to build their F1 engines as V6s, following Renault’s lead. (Even though Renault’s original choice of the V6 for their competition engine in the early seventies does not seem to have been with turbocharging in mind).

And, also interestingly, since 2014, the FIA has mandated turbocharged 1600cc V6 engines for F1.

This raises the question: Is the V6 particularly suitable for turbocharging when the total displacement is 1500 to 1600cc?

Or has it been chosen simply because Renault happened to turbocharge an existing V6 engine, whose performance they needed to enhance, so as to win at Le Mans? And met with resounding success. (The straight-6 may reasonably be dismissed, as it is none too suitable, architecturally speaking, for a modern racing car, and for the majority of modern road cars too).

Those Indy turbo L4s and V8s were initially 2800cc, later somewhat reduced if I remember rightly. Could the difference in displacement account for the difference in approach? (Seems unlikely, as here, too, existing “aspirated” engines were adapted and developed with turbochargers).

[lustigson]

This raises the question: Is the V6 particularly suitable for turbocharging when the total displacement is 1500 to 1600cc?

 

The original hybrid-turbo plans were for inline 4 cylinder engines, but I seem to remember that Ferrari were opposed, because they have no 4-pots in their road car line-up. Hence the move to the V6 configuration.

[Charlieman]

The Honda V6 had its origins with the F2 2 litre engine. The TAG, Ferrari and significantly later Cosworth were clean sheet V6 designs. John Barnard's autobiography is worth reading if the TAG engine interests you. Barnard wanted an engine with the cleanest profile and packaging for a ground effects car with side tunnel venturi. TAG/McLaren commissioned Porsche to build an engine according to Barnard's criteria, and photos show that it was the slimmest and tidiest of the V6 engines. Unfortunately for McLaren, the engine was too late to ever run in a first generation ground effects car but Barnard was able to use the profile successfully in his designs.

 

The first Renault engine had a single turbocharger and the layout was asymmetric. Later engines had a separate turbo and intercooler for each cylinder bank (ie symmetric), with the turbos, exhausts and intercoolers positioned away from the venturi exits. Twin turbos were individually smaller than a single turbo, theoretically spinning up more quickly and reducing boost lag. Overall, the V6 twin turbo was better for venturi tunnel packaging and twin turbo layout.

 

Barnard had previously been involved with the Cosworth DFX project which started at the Vel's Parnelli Jones team. Cosworth were not particularly interested in the Indy/USAC turbo concept and supplied kits of parts to specialist engine builders. The engine builder implemented fuel system, turbos and ignition, unlike the DFV for which you could buy a complete "standard" engine from Cosworth.

[kikiturbo2]

the original 1500 cc engines were mostly v6 bacause v6 is compact and stiff and can be used as a stressed member, plus leaves space for underbody tunels.. IIRC the i4 had to be canted on one side and have external srtucture to carry the gearbox/rear suspension load

[PhantomRaspberryBlower]

I wonder if the Motori-Moderni V6 was half an Alfa V12, given the involvement of Chiti?

[ceesvdelst]

I think the BMW was perhaps the most compact and best engine overall, but was after 83/4 in the wrong car.

 

The TAG was a super engine, in a great car that seemed to last for about 3 or 4 years!

 

Honda, well what can you say.

 

I have often wondered how good the Cosworth V6 would have been in 88, rumour was it would have been superb in lower boost economy style racing, certainly up there with Ferrari and better than the sold out Megatron motors. 

 

Also would have been fascinating to see what might have happened had Ford not come unstuck with the 4 cylinder engine in 83/84. 

 

But yes, I think size, width, maybe other reason mean the V6 simply ruled. 

[Wuzak]

As everyone in this forum knows, Renault introduced the turbocharged engine to Formula 1 in the late seventies, with a 1500cc version of their Le Mans-winning 2-litre V6, soon showing that the turbo was the way to go. Apart from BMW, Hart and Alfa Romeo, most of the competition followed suit with their own turbo V6s, and very successful many of them were. BMW and Hart had sound, but different, reasons for using turbo L4s; only Alfa Romeo built a turbo V8, this last without success.

Concurrently, the Cosworth DFX Indianapolis turbo V8 was developed, based on the legendary DFV, and was just as successful as its parent. In fact, the Indianapolis brigade had a ten-year start with successful turbo engines, both the Offenhauser L4 and the “four-cam” Ford V8 dating back to the late sixties.

So it is interesting that Ferrari, Honda, TAG and Ford (Cosworth) all opted to build their F1 engines as V6s, following Renault’s lead. (Even though Renault’s original choice of the V6 for their competition engine in the early seventies does not seem to have been with turbocharging in mind).

And, also interestingly, since 2014, the FIA has mandated turbocharged 1600cc V6 engines for F1.

This raises the question: Is the V6 particularly suitable for turbocharging when the total displacement is 1500 to 1600cc?

Or has it been chosen simply because Renault happened to turbocharge an existing V6 engine, whose performance they needed to enhance, so as to win at Le Mans? And met with resounding success. (The straight-6 may reasonably be dismissed, as it is none too suitable, architecturally speaking, for a modern racing car, and for the majority of modern road cars too).

Those Indy turbo L4s and V8s were initially 2800cc, later somewhat reduced if I remember rightly. Could the difference in displacement account for the difference in approach? (Seems unlikely, as here, too, existing “aspirated” engines were adapted and developed with turbochargers).

 

The adoption of turbos in Le Mans and Indy/CART racing may have occurred earlier due to equivalence factors.

 

For F1 the equivalence factor was 2. For Indy it was 1.5 (4.2 N/A vs 2.8 turbo)?

[blueprint2002]

The original hybrid-turbo plans were for inline 4 cylinder engines, but I seem to remember that Ferrari were opposed, because they have no 4-pots in their road car line-up. Hence the move to the V6 configuration.

 

The Honda V6 had its origins with the F2 2 litre engine. The TAG, Ferrari and significantly later Cosworth were clean sheet V6 designs. John Barnard's autobiography is worth reading if the TAG engine interests you. Barnard wanted an engine with the cleanest profile and packaging for a ground effects car with side tunnel venturi. TAG/McLaren commissioned Porsche to build an engine according to Barnard's criteria, and photos show that it was the slimmest and tidiest of the V6 engines. Unfortunately for McLaren, the engine was too late to ever run in a first generation ground effects car but Barnard was able to use the profile successfully in his designs.

 

The first Renault engine had a single turbocharger and the layout was asymmetric. Later engines had a separate turbo and intercooler for each cylinder bank (ie symmetric), with the turbos, exhausts and intercoolers positioned away from the venturi exits. Twin turbos were individually smaller than a single turbo, theoretically spinning up more quickly and reducing boost lag. Overall, the V6 twin turbo was better for venturi tunnel packaging and twin turbo layout.

 

Barnard had previously been involved with the Cosworth DFX project which started at the Vel's Parnelli Jones team. Cosworth were not particularly interested in the Indy/USAC turbo concept and supplied kits of parts to specialist engine builders. The engine builder implemented fuel system, turbos and ignition, unlike the DFV for which you could buy a complete "standard" engine from Cosworth.

 

 

the original 1500 cc engines were mostly v6 bacause v6 is compact and stiff and can be used as a stressed member, plus leaves space for underbody tunels.. IIRC the i4 had to be canted on one side and have external srtucture to carry the gearbox/rear suspension load

 

 

I wonder if the Motori-Moderni V6 was half an Alfa V12, given the involvement of Chiti?

 

 

I think the BMW was perhaps the most compact and best engine overall, but was after 83/4 in the wrong car.

 

The TAG was a super engine, in a great car that seemed to last for about 3 or 4 years!

 

Honda, well what can you say.

 

I have often wondered how good the Cosworth V6 would have been in 88, rumour was it would have been superb in lower boost economy style racing, certainly up there with Ferrari and better than the sold out Megatron motors. 

 

Also would have been fascinating to see what might have happened had Ford not come unstuck with the 4 cylinder engine in 83/84. 

 

But yes, I think size, width, maybe other reason mean the V6 simply ruled. 

 

 

The adoption of turbos in Le Mans and Indy/CART racing may have occurred earlier due to equivalence factors.

 

For F1 the equivalence factor was 2. For Indy it was 1.5 (4.2 N/A vs 2.8 turbo)?

 

 

Thanks, guys, for filling in various facets of the picture and/or for your views on the matter.

With regard to Ferrari and TAG (Porsche) F1 V6s, the “Autocourse History of the Grand Prix Car - 1966 to 85” mentions that both considered the V8 as well, but opted for the V6. However, there is no explanation why this was so.

The V6, no matter what the angle between the cylinder banks, is inherently less well balanced than the V8. When the angle is 120 deg, the vibration can be rather less than at the more usual angles of 60 to 90 deg. Additionally, a 3-throw crankshaft provides even firing and thus the smoothest torque. With angles of 60, 80 or 90 deg, even firing is possible only with 6-throw or split-pin crankshafts, these last not being the best for long, trouble-free life. And either makes the whole engine bulkier and heavier. Or uneven firing must be accepted, which affects the torsional vibration characteristics, this may again demand palliative measures.

The 90-degree V8, on the other hand, is perfectly balanced when the crankshaft is of the cross-plane type, not so good with the single-plane crankshaft which is usual with racing engines, but still better than the V6. And it fires evenly, with more and smaller torque impulses in each cycle, so the output is that much smoother.

Since vibration (of any sort) is one of the principal sources of machinery failure, it is a major cause of unreliability and short life; clearly something to avoid as far as possible. Technology has developed to the point where the effects or results of vibration can be overcome (suppressed may be a better word), but as always, the price has to be paid, one way or another.

Perhaps architectural (or “packaging”) considerations took precedence over mechanical realities, though the differences here too would seem to be marginal.

[Charlieman]

Thanks, guys, for filling in various facets of the picture and/or for your views on the matter.

With regard to Ferrari and TAG (Porsche) F1 V6s, the “Autocourse History of the Grand Prix Car - 1966 to 85” mentions that both considered the V8 as well, but opted for the V6. However, there is no explanation why this was so.

The V6, no matter what the angle between the cylinder banks, is inherently less well balanced than the V8. 

 

I do not recall any explanations from John Barnard for the V6 versus V8 selection. Remember that the flat crankshaft DFV V8 was known to be a lumpy engine but F1 teams were able to cope with it for two hour races. The DFV was used in a few sports cars for endurance racing -- it took a long time before the detuned DFV worked reliably. My assumption is that F1 teams considered the V6 -- smaller reciprocating mass for a 1.5 litre -- as an acceptable compromise. 

 

Regarding engine capacity equivalence, Porsche were sceptical about the Renault V6 project when it was announced. Porsche and BMW were the big name developers of turbo technology for road racing but they had done it with a 1.4 equivalence factor. They did most of the hard work, almost certainly with smaller budgets than Renault. I'd say that Porsche underestimated how much they had achieved.

[gruntguru]

There is always an optimum displacement per cylinder for any given design. For most racing engines it is somewhere around 300cc.

[blueprint2002]

There is always an optimum displacement per cylinder for any given design. For most racing engines it is somewhere around 300cc.

 

I believe this is sometimes attributed to Keith Duckworth. If you know of any reference which I could consult on this subject, please do let me know. Thanks.

What I'd like to know is how "optimum displacement" is defined. From the view point of power output? Breathing ability? Mean effective pressure? To name but a few possibilities.

And, what, if any assumptions are operative. Stroke-Bore ratio? Valve-area to piston-area ratio? Materials of cylinder, head, and piston? Again, others are also possible.

The date when this judgement was made could also be significant, because some of the assumptions could be implicit in the technology of the time. 

[gruntguru]

"Optimum" with consideration of a number of factors but with greatest weighting applied to power and breadth of power band. Clearly fuel efficiency, durability, packaging etc are also important in a race engine.

 

Tony Rudd was one (pre-Duckworth) adherent and refers to the concept a number of times in "It Was Fun". Interesting that he was involved with two of the most spectacular failed attempts to produce engines with tiny cylinders - the V16 and H16 BRMs.

 

Obviously changing any of the assumptions you listed would change the optimum cylinder size so you would have to assume that each of these is optimised for race engine requirements. No doubt advances in design and materials has shifted what is "optimal" over time but it is interesting to note that even during the march towards 20,000 rpm, the last 3 litre formula seemed to converge on a 300cc cylinder size.

[blueprint2002]

I do not recall any explanations from John Barnard for the V6 versus V8 selection. Remember that the flat crankshaft DFV V8 was known to be a lumpy engine but F1 teams were able to cope with it for two hour races. The DFV was used in a few sports cars for endurance racing -- it took a long time before the detuned DFV worked reliably. My assumption is that F1 teams considered the V6 -- smaller reciprocating mass for a 1.5 litre -- as an acceptable compromise. 

Very likely you are right.

And yet, the “lessons of history” are not always straightforward.

Ferrari’s V6 Dino was supreme in the final years of the 2.5 litre formula, replacing the L4s, L6s and V8s that preceded it, and defeated by the Cooper in spite of the Climax FPF, rather than because of it.

Only a couple of years later, the 1.5 litre Dino was outmatched by the V8s of BRM, Climax and Ferrari themselves.    

[blueprint2002]

"Optimum" with consideration of a number of factors but with greatest weighting applied to power and breadth of power band. Clearly fuel efficiency, durability, packaging etc are also important in a race engine.

 

Tony Rudd was one (pre-Duckworth) adherent and refers to the concept a number of times in "It Was Fun". Interesting that he was involved with two of the most spectacular failed attempts to produce engines with tiny cylinders - the V16 and H16 BRMs.

 

Obviously changing any of the assumptions you listed would change the optimum cylinder size so you would have to assume that each of these is optimised for race engine requirements. No doubt advances in design and materials has shifted what is "optimal" over time but it is interesting to note that even during the march towards 20,000 rpm, the last 3 litre formula seemed to converge on a 300cc cylinder size.

 

Thanks GG. That is another rather puzzling fact! 

Edited by blueprint2002, 09 October 2019 - 01:14.

[Greg Locock]

Isn't one of the limiting factors Mach choking of the intake valves at high rpms? So you could, I think, work out some sort of cylinder size vs rpm vs bore relationship for a 4 valve head?

[Henri Greuter]

The adoption of turbos in Le Mans and Indy/CART racing may have occurred earlier due to equivalence factors.
 
For F1 the equivalence factor was 2. For Indy it was 1.5 (4.2 N/A vs 2.8 turbo)?

And in Endurance, at least in the 70's when Porsche introduced the turbocharged Carrera RSR Turbo the handicap factor was 1.4 (2.1 liter turbo vs 3 liter) for the Gp6 open sportscars.

Mindy you, as unfavorable as the factor 2 appears, it has been worse. The first postwar F1 engine formula was for 4.5 liter unblown vs 1.5 liter supercharged, thus factor 3! Despite that, the Prewar Alfa Romeo 158's reigned supreme for a long time. But none of the 4.5 liters that went up against the Alfettas were multicylinder DOHC 4 valvers so that certainly helped the Alfettas.

Edited by Henri Greuter, 09 October 2019 - 13:20.

[Henri Greuter]

"Optimum" with consideration of a number of factors but with greatest weighting applied to power and breadth of power band. Clearly fuel efficiency, durability, packaging etc are also important in a race engine.
 
Tony Rudd was one (pre-Duckworth) adherent and refers to the concept a number of times in "It Was Fun". Interesting that he was involved with two of the most spectacular failed attempts to produce engines with tiny cylinders - the V16 and H16 BRMs.
 
Obviously changing any of the assumptions you listed would change the optimum cylinder size so you would have to assume that each of these is optimised for race engine requirements. No doubt advances in design and materials has shifted what is "optimal" over time but it is interesting to note that even during the march towards 20,000 rpm, the last 3 litre formula seemed to converge on a 300cc cylinder size.

A little known case op optimum design....

the Heidegger team developed a straight6 1.5 turbo with power takeoff between the cylinders 3&4. And the story is that they showed it to the Mclaren team and John Barnard. The latter supposedly went ballistic, proclaiming that no matter how good the engine itself could be, that it was a nightmare and impossible to put within an F1 car and make that work and be fast. And that the days of designers building a car around the engine were over and done and that it was by now the other way around.


I think that one benefit of the V6 is also overlooked compared with the in-Line 4's. Dus to the symmetry, it was possible to divide the work for a single turbo over two smaller ones which then had less stress to cope with. From what I have understood, the single big turbo's on the Bimmers (& Hart's) were under a tremendous stress and more vulnerable as a result.

[Henri Greuter]

As everyone in this forum knows, Renault introduced the turbocharged engine to Formula 1 in the late seventies, with a 1500cc version of their Le Mans-winning 2-litre V6, soon showing that the turbo was the way to go. Apart from BMW, Hart and Alfa Romeo, most of the competition followed suit with their own turbo V6s, and very successful many of them were. BMW and Hart had sound, but different, reasons for using turbo L4s; only Alfa Romeo built a turbo V8, this last without success.[/size]
Concurrently, the Cosworth DFX Indianapolis turbo V8 was developed, based on the legendary DFV, and was just as successful as its parent. In fact, the Indianapolis brigade had a ten-year start with successful turbo engines, both the Offenhauser L4 and the “four-cam” Ford V8 dating back to the late sixties.[/size]
So it is interesting that Ferrari, Honda, TAG and Ford (Cosworth) all opted to build their F1 engines as V6s, following Renault’s lead. (Even though Renault’s original choice of the V6 for their competition engine in the early seventies does not seem to have been with turbocharging in mind).[/size]
And, also interestingly, since 2014, the FIA has mandated turbocharged 1600cc V6 engines for F1.[/size]
This raises the question: Is the V6 particularly suitable for turbocharging when the total displacement is 1500 to 1600cc? [/size]
Or has it been chosen simply because Renault happened to turbocharge an existing V6 engine, whose performance they needed to enhance, so as to win at Le Mans? And met with resounding success. (The straight-6 may reasonably be dismissed, as it is none too suitable, architecturally speaking, for a modern racing car, and for the majority of modern road cars too).[/size]



[/size][/font]

You mentioned the Offenhauser. One of the fatures of that engine since its inception, including most of the Miller predecessors it was derived from and/or found its roots into, The Offy had an integral cylinderhead, thus no removable cylinderhead for valve maintanance etc. But also: no head gasket. Because of that, as well as running on alcohol fuels, it could use quite a high compression ratio.But it also openend an option for blowing the reduced shortstroke versions debuted in 1966. And that worked like a charm.

Edit: Brian Hart eventually went to a similar construction for his engines, these were called the monoblocs end edit


It is interesting to point out but when we are ravelling about the 917/30 CanAm with its turbocharged 5.4 liter flat12 putting out some 1200 hp in 1973; the best Offies could do the same with half the capacity and number of turbos and 1/3th the number of cylinders.....

Edited by Henri Greuter, 09 October 2019 - 13:35.

[Charlieman]

The Offy had an integral cylinderhead, thus no removable cylinderhead for valve maintanance etc. But also: no head gasket. 

In the 935/936 era, Porsche built some engines with electron beam welded cylinder heads. At the time it would have been bleeding edge technology application for something so big.

[BRG]

It is interesting to point out but when we are ravelling about the 917/30 CanAm with its turbocharged 5.4 liter flat12 putting out some 1200 hp in 1973; the best Offies could do the same with half the capacity and number of turbos and 1/3th the number of cylinders.....

True, but the Porsche Can-Am engine was anew development and proved unbeatable so there was no need for Porsche to continue pushing it forward even if the Can-Am hadn't collapsed.  I suspect that they could have extracted a LOT more laugh if they had continued to work on it.

 

Whereas Offy development was on-going for far longer and probably undertaken by several engine shops competing with each other.  Hence they would certainly have progressed a lot further.  Another point might be that the Porsche motor was for road courses and the Offys for ovals, requiring very different driveabiity modes.  

[Henri Greuter]

True, but the Porsche Can-Am engine was anew development and proved unbeatable so there was no need for Porsche to continue pushing it forward even if the Can-Am hadn't collapsed.  I suspect that they could have extracted a LOT more laugh if they had continued to work on it.

 

Whereas Offy development was on-going for far longer and probably undertaken by several engine shops competing with each other.  Hence they would certainly have progressed a lot further.  Another point might be that the Porsche motor was for road courses and the Offys for ovals, requiring very different driveabiity modes.  

 

Almost all correct and true.

 

The only thing I don't think is correct is much more progress being made with the Offy. Its strongpoint compared with the Quadcam Ford was literally its strength to cope with for the time rediculous boost levels. But after the 1973 season, that strongpoint was taken away with the boost levels maximized on a level that was no problem at all for an Offy to cope with but against a V8 at the same boost gave away 50 to 75 hp by default. And that was pretty much the end for the Offy since that opened the door for batter V8s than the Quadcam Ford: The Cosworth.

 

BTW, the 1000+ hp Offy of 1973 was on the very edge already. At Indy they dropped out of the race as flies. That race was flagged mercifully at 332.5 miles but given the number of cars still left at that moment, I wonder how many would have made it to the end had the race gone the full distance and how much they had been detuned to make it to begin with in the final stage of the race..