Pushrod engines in F1

[LHutton]

Could they provide an advantage now that revs have come down to a more manageable level?

[Kchrpm]

*quick Google research*

It looks like some Cup-related circle track pushrod engines have been able to have 10k-11k redlines and last entire races, but that was naturally aspirated at around 6 liters, and I don't know what the differences in g-loading will do (or if they matter at all).

I assume the development costs would be rough, though, considering F1 teams haven't worked with pushrods in decades.

[tigeraid]

The #1 advantage to pushrods is that the engine is generally a little lighter, and more compact, dimensionally. And considering how well F1 teams have manipulated engine location and size management already, I don't think it would matter much to them. The weight difference would be minimal considering the exotic materials they already use. The main reason to use pushrods is to fit under a Corvette's low hood, basically. ;)

[LHutton]

*quick Google research*

It looks like some Cup-related circle track pushrod engines have been able to have 10k-11k redlines and last entire races, but that was naturally aspirated at around 6 liters, and I don't know what the differences in g-loading will do (or if they matter at all).

I assume the development costs would be rough, though, considering F1 teams haven't worked with pushrods in decades.
Longer strokes equal higher piston speeds. Those 6 litre engines pushing 10krpm have similar piston speeds to 2.4 V8 F1 engines at 19,000rpm.

I was actually thinking of this though:

The chief advantages would be CoG and drag related. I guess valvetrain friction and MPG might also play a part.

http://www.fanmercedesbenz.com/1994-penske-mercedes-pc23-indycar/
http://8w.forix.com/penske-mercedes-pc23-car.html


http://forix.autosport.com/8w/pc23-500i/500i-cutaway.png

http://www.fanmercedesbenz.com/wp-content/uploads/2014/12/4879-10.jpg

[Kchrpm]

I was under the impression that pushrod valve tech limited RPM capability, no matter the stroke length.

I did look up the Ilmor, but that was just used for one race, and was competitive mostly because it was given higher limits for boost and displacement limits, not because of any intrinsic benefits from being a pushrod design.

I guess I just don't know what the intended benefit would be of going to a pushrod engine in F1. Unless there is no point, and it's just a thought experiment, then yeah, you could pull it off.

[LHutton]

It does limit rpm but longer stroke engines are naturally more limited rpm-wise anyway.

Unless we're talking about a massive loss in power with pushrods below 11,500rpm, that's about the limit of the rpm most teams are using because of fuel savings and reliability. The advantages are obvious. Lower engine height = Lower CoG = Lower bodywork = Less drag = Better handling + Less fuel consumption + Better straight-line speed. Less valvetrain friction = Cooler running + More net power.

Ilmor demonstrated that pushrods could run with some reliability at 10,500rpm back in the '90s and NASCARs run near 10,000rpm all day long. Is it possible, or is there some major reduction in torque at higher rpm with pushrod designs that I'm overlooking?

[Godson]

I'd wager, depending on the OHC setup, that a pushrod would have MORE valve train friction.

A big issue with a pushrod engine design is flow. OHC engines only need to worry about water jackets, cooling jackets, and studs for the flow of fuel/air into and out. You add pushrods to the mix, you then need to move stuff around on the heads to allow room for the pushrods, which can be a big sacrifice at time. I'd also wager with the high cylinder pressures of f1 engines, getting a pushrod engine to seal at 12k rpm with a good layout for power would be difficult while still allowing good packaging.

[Kchrpm]

We're past my depth of knowledge of the subject. I would build something in Automation, but they tend to have valve float come in very early with no real way to fix it.

[LHutton]

Interesting answers that somewhat confront my conventional understanding but it was just an idea and an interesting thread. Cheers guys.

[Dicknose]

I think they all run pneumatic valves (for closing).
Wonder when they will go with the same for opening and have it electronically controlled.
No cams at all.
No losses, no moving, rotating metal parts.
Automatic variable timing.

Reliability is probably what's holding them back.
But once someone goes and finds it works and has a benefit then it will probably quickly be the new standard
(Just as valve springs got replaced with pneumatic)

[Godson]

I think LMP cars and f1 have all been pneumatic since the mid 2000s.

[Random]

I think they all run pneumatic valves (for closing).
Wonder when they will go with the same for opening and have it electronically controlled.
No cams at all.
No losses, no moving, rotating metal parts.
Automatic variable timing.

Reliability is probably what's holding them back.
But once someone goes and finds it works and has a benefit then it will probably quickly be the new standard
(Just as valve springs got replaced with pneumatic)

Rules are what's holding them back: no variable valve timing or lift.



5.9.2 Variable valve timing and variable valve lift profile systems are not permitted.

[Freude am Fahren]

Doesn't Koenigsegg have a camless engine?

Or at least a prototype

[Godson]

Variable valve timing rule is fucking dumb.

[Rare White Ape]

I thought that by now the rule makers would embrace electronic valves, the pinnacle in variable valve timing, and not have to worry about any of this.

It took a while for MotoGP to even allow pneumatic valves.

[LHutton]

If you varied valve duration, would that count as timing or lift?

[Godson]

Timing.

[LHutton]

Ah okay, so timing is based on start, middle and end of lift rather than just one point during lobe action wrt the regs.

[Godson]

http://lmgtfy.com/?q=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcAl w9RpTZ4I

[Kchrpm]

Ha, that is perhaps the stupidest use of LMGTFY ever, putting in a direct link to a YouTube video.

[Freude am Fahren]

:lol:

[LHutton]

http://lmgtfy.com/?q=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcAl w9RpTZ4I
Err I know how cams work, my question pertained to how 'valve timing' is defined in the F1 regulations. If it was based on a single point in lobe action, i.e. start, middle or end, then variable duration could theoretically be used without contravening the regulations on variable valve timing.

[Random]

It isn't defined, that I recall, but changing anything about how and when the valve moves is de facto changing the timing and lift, no? Some of the variable valve timing schemes were just cam phasers, after all.

[LHutton]

It isn't defined, that I recall, but changing anything about how and when the valve moves is de facto changing the timing and lift, no? Some of the variable valve timing schemes were just cam phasers, after all.
If they've left the definition of 'variable valve timing' open-ended then yes that probably would be the default. The second part was what I was getting at - did the rules just prohibit camshaft position relative to the crank? I'm guessing the answer is 'no'.

The original pushrod idea stemmed from the fact that power output is effectively cock-blocked by fuel flow restrictions anyway and nobody is really going over 11,500rpm despite the official 15,000rpm limit, so if the you could still pull in enough air to use that fuel flow with a pushrod ohv engine, have a sufficiently wide powerband and maintain reliability, the weight, CoG and packaging/drag benefits may be worthwhile. DOHC was originally adopted because fuel flow and rpm was entirely open-ended but capacity was limited, hence DOHC engines revving to 20,000rpm were the best solution because of the massive power benefits, but now that that's effectively all changed, how does the pushrod vs DOHC solution weigh up?

[Freude am Fahren]

Space is probably used about as well as can be right now. Taking away the cams up top may lower the cg, but I wonder how much larger the engines might need to be in plan view to give room for the cam(s?)/pushrods? If NASCAR can manage 10k with them though, I don't doubt that F1 could add another 1500rpm in short time.

Are NASCAR engines limited in RPM by reliability or rules?

[LHutton]

Space is probably used about as well as can be right now. Taking away the cams up top may lower the cg, but I wonder how much larger the engines might need to be in plan view to give room for the cam(s?)/pushrods? If NASCAR can manage 10k with them though, I don't doubt that F1 could add another 1500rpm in short time.

Are NASCAR engines limited in RPM by reliability or rules?
Rules. These days there is no official hard and fast rpm limit but final drive ratios are restricted on a per track basis, which means they don't get much past 10,000rpm. Without that they would reach >11,000rpm as they do in Pro Stock.

http://www.epi-eng.com/piston_engine_technology/comparison_of_cup_to_f1.htm

I guess it depends whether the space between the V is currently being used. These are the best pictures I can find. Difficult to tell really. In the Ferrari, Mercedes and Cosworth designs, the lower height of the engine could lead to a lower overall height, for the Renault it's less clear. I think the main issue would be reliability though. 1 race - fine, several races - ???

http://i1.ytimg.com/vi/4vIjJg0lXgc/maxresdefault.jpg

https://fbcdn-sphotos-g-a.akamaihd.net/hphotos-ak-prn2/1049192_678634435484359_785563494_o.jpg

http://www.rssportscars.com/photos/cars/2014-mercedes-amg-w05-formula-1-car/mercedes-w05-formula1-08-engine-m.jpg

http://4.bp.blogspot.com/-1MLw3WjQ9nE/TtAlkr9jEsI/AAAAAAAF6VE/L_3Qt-Xz7vI/s800/c0602008H.jpg

http://www.racecar-engineering.com/wp-content/uploads/2014/01/upCosworth1.jpg

http://ysfdesign.net/wp-content/uploads/2014/06/2014-Cosworth-V6-Engine-Render-05.jpg

http://media.theformula1.com/blog/wp-content/uploads/2013/05/FerrariEngineV82012.jpg

[Freude am Fahren]

Yeah, that space between houses the MGU-H for all of them I think, except maybe Mercedes, which may have it externally, but has something there still to attach the turbine to the compressor. MB may be the only one with room in there.

[Godson]

Ha, that is perhaps the stupidest use of LMGTFY ever, putting in a direct link to a YouTube video.

Stupid response was due to a stupid question. I was merely mocking him. It was a poor choice of action on my part.


Err I know how cams work, my question pertained to how 'valve timing' is defined in the F1 regulations. If it was based on a single point in lobe action, i.e. start, middle or end, then variable duration could theoretically be used without contravening the regulations on variable valve timing.

I am going to try and keep things simple for the sake of not getting into the super in-depth engineering design theories, etc.
--------
If you read the rule book you’d have your answer. Nothing about the valve timing can change. Not the intake opening/closing, not the lobe separation, not the camshaft lobe centerline, not the exhaust cam opening/closing. Nothing. Period.

Also you DON'T understand how a cam works when you talk of varying the duration and not adjusting the valve timing. Changing the duration is doing exactly that. Duration is the exact measurement in degrees of the valves action on the camshaft. You change the duration, you are changing the timing of the valve opening/closing. Hence, you are varying the valve timing, and breaking the rules.

For example, BMWs VANOS system does exactly as you are talking about, by advancing and/or retarding the lobe centerline of the camshafts IN RELATION TO the timing of the crankshaft aka, the engine.
Honda does variable valve timing by using an ADDITIONAL set of lobes on the same camshaft. It is essentially 2 camshaft cuts, one for economy and one for performance, on a single shaft.

Both of these designs do a similar thing, vary the timing of the valves. VANOS allows variable timing. VTEC allows variable timing AND variable lift.


If they've left the definition of 'variable valve timing' open-ended then yes that probably would be the default. The second part was what I was getting at - did the rules just prohibit camshaft position relative to the crank? I'm guessing the answer is 'no'.

The original pushrod idea stemmed from the fact that power output is effectively cock-blocked by fuel flow restrictions anyway and nobody is really going over 11,500rpm despite the official 15,000rpm limit, so if the you could still pull in enough air to use that fuel flow with a pushrod ohv engine, have a sufficiently wide powerband and maintain reliability, the weight, CoG and packaging/drag benefits may be worthwhile. DOHC was originally adopted because fuel flow and rpm was entirely open-ended but capacity was limited, hence DOHC engines revving to 20,000rpm were the best solution because of the massive power benefits, but now that that's effectively all changed, how does the pushrod vs DOHC solution weigh up?

Changing the relation of the camshaft centerline also known as lobe centerline in relation to the camshaft is still varying the valve timing. The valves are timed off of 0 degrees TDC of whichever piston is deemed to be timed off of. This is usually Cylinder number 1.

Power is limited by 3 major factors. Fuel amount, fuel delivery, and how it is burned. DOHC engines do a better job at running at high RPM than pushrods, part of this is due to the wider intake and exhaust runner designs. 4 valve engines (predominantly used by OHC designs) also have a prime advantage over the 2 valve designs (predominantly used by pushrod engines) in which they are naturally a hemispherical design and allow a better burn of the fuel/air mixture. But I digress. We have discussed this previously in this thread. Narrow intake/exhaust runners create a higher velocity of fuel delivery, and in turn allow a faster closing of intake and exhaust valves. This in turn allows greater variability in the ignition of said fuel (this is very rough explanation), which can create more torque. Not power. Good example of this is the Gen II Small blocks versus the Gen III/IV engines using the same cam and intake designs. The Gen II are going to create more torque in their curve relative to the Gen III. Gen II designs also run out of revs sooner than their Gen III counterparts. As I said, this is very, very, very simplified.



To add another angle at the valve timing debate. If you change the length of the rocker arm, known as the rocker arm ratio, you are also changing not just valve timing, but the valve lift. So technically, a variable ratio rocker arm (while stupid in theory do to the complexity of the design needed to withstand the forces when simpler and more effective designs are present) would still be variable valve timing.

[Kchrpm]

Stupid response was due to a stupid question. I was merely mocking him. It was a poor choice of action on my part.
I can understand if you wanted to mock him, but how was he supposed to Google a specific address that he didn't already know?

Proper lmgtfy etiquette would be to use the topic to be searched, not the intended result. It has been used on me several times, so I feel that I am an expert.

[LHutton]

Stupid response was due to a stupid question. I was merely mocking him. It was a poor choice of action on my part.

I am going to try and keep things simple for the sake of not getting into the super in-depth engineering design theories, etc.
--------
If you read the rule book you’d have your answer. Nothing about the valve timing can change. Not the intake opening/closing, not the lobe separation, not the camshaft lobe centerline, not the exhaust cam opening/closing. Nothing. Period.

Also you DON'T understand how a cam works when you talk of varying the duration and not adjusting the valve timing.
FFS dude, I'm a bloody engineer.:rolleyes: I was simply probing to find out how detailed the rules are on 'timing'. There are several parameters that determine valve operation, and 'timing' on its own leaves it a bit fuzzy. The timing of the opening, the peak or the closing? For the last time I DO understand how a cam works, either that or I got real lucky when I rebuilt two engines. If you don't like my terminology, fine, that's a different issue.


Changing the duration is doing exactly that. Duration is the exact measurement in degrees of the valves action on the camshaft. You change the duration, you are changing the timing of the valve opening/closing.
You're changing one but not necessarily both. I could open the inlet valves at the same time but close them later and I have increased duration and likely overlap too, if the exhaust valve opening time is unchanged and its duration is also increased. I could theoretically even have the same fully open timing point if I've completely changed the lobe or rocker profile.



Power is limited by 3 major factors. Fuel amount, fuel delivery, and how it is burned. DOHC engines do a better job at running at high RPM than pushrods, part of this is due to the wider intake and exhaust runner designs. 4 valve engines (predominantly used by OHC designs) also have a prime advantage over the 2 valve designs (predominantly used by pushrod engines) in which they are naturally a hemispherical design and allow a better burn of the fuel/air mixture. But I digress. We have discussed this previously in this thread. Narrow intake/exhaust runners create a higher velocity of fuel delivery, and in turn allow a faster closing of intake and exhaust valves. This in turn allows greater variability in the ignition of said fuel (this is very rough explanation), which can create more torque. Not power. Good example of this is the Gen II Small blocks versus the Gen III/IV engines using the same cam and intake designs. The Gen II are going to create more torque in their curve relative to the Gen III. Gen II designs also run out of revs sooner than their Gen III counterparts. As I said, this is very, very, very simplified.
Hemispherical designs can be achieved with ohv engines too, i.e. the Mopar Gen II and Gen III hemi engines. 2014 F1 engines employ direct injection, so fuel delivery itself need not be affected by head design, as fuel enters the cylinder directly, not in an air-fuel mix via the valves as with port injection. Highly efficient mixing and atomisation can be achieved by using swirl injection methods. The action of the air itself as it enters the cylinder is the sole remaining factor and whether the cylinder displacement to valve area ratio can pull in the enough air to burn the restricted fuel supply. It's a matter of weighing up any lost power, if any, against potential drag and handling benefits, which is probably a job for a supercomputer. Maybe someone has already done it, maybe they haven't.

Aside from that reliability is the big one. DOHC definitely has a reliability advantage at those rpms.