12 replies to this topic

[DOHCPower]

What is the currently accepted CFM at max cam lift on current F1 cylinder heads?

Also, when questioning a recent 'expert', he revealed to me header design is primarily based on valve size. How you found this to be true, or does it have more to do with swept volume of a single cylinder or cfm?

[hydra]

The state of the art for cylinder head flow is around 105-110cfm/square inch @ 28" depression at a lift/diameter ratio of 0.36, which means that for an F1 head with 40mm intake valves, we're looking at around 420cfm @ 0.550-0.575"

And IMO, header design should be a function of rpm, cylinder displacement, and valve size - in that order

[DOHCPower]

Thank you very much for the response.

When you say that the state of the art heads flow at these values, are you stating all F1 teams, or only the elite?

[hydra]

Dohcpower,
I was talking about the state of the engine art in general, based on a chart I once saw published. I really doubt the difference in specific head flow (aka flow coefficient) is more than 1% from best to worst...

[DOHCPower]

Interesting. So in what area's do teams gain horsepower on other teams if not in cylinder head flow? Friction "maintenance" ?

[hydra]

Remember, this is for SPECIFIC head flow - e.g. cfm/square inch of valve area. The primary way teams gain horsepower is by raising rpms and tuning the rest of the engine to suit, while of course minimizing frictional losses. This brings up a rather interesting point, one that I've been thinking of for a while now...

Every year (especially back in they heydey of engine development - around 5-10 years ago) manufacturers release a new engine, with a 1mm larger bore, and 1mm smaller stroke (for example), and another 500rpm, and ~25 more horsepower. Makes you wonder why they had to go with such incremental changes... Why couldn't they have gone from a 91mm bore (Peugeot/Asiatech) to 97-98mm (Ferrari of similar vintage) straight away, gaining 1000rpm, 50bhp, and saving several years of evolutionary "development" in the process?

[desmo]

It seems like in the late '90s there was a bifurcation of b/s strategies, with one camp led by Ferrari-Supertec going for a ~95mm bore, and another camp- Ford-Ilmor-Honda- at around 93mm. I assume the smaller bore guys were aiming at lower max piston accelerations, higher BMEPs and lower frictional losses. Perhaps the smaller bore engines made it possible to make an attractive compromise between valve lifts and compression ratio as well.

A good place to look at upsides/downsides of some of the b/s compromises is SAE Paper 2002-01-3315 by Ferrari man Boretti, here's a bit from the end:

The proposed engine solutions are a very satisfactory
basis for engine development. Despite in previous
engine projects the maximum power engine speed and
the bore/stroke ratio was significantly smaller, the model
reliability may be considered good. These solutions
obviously need to be further refined by investigating
small variations of all the model parameters.
As the bore is increased and the engine is tuned for
higher engine speeds, curves of charging efficiency are
shifted towards higher engine speeds with a slight
increase in magnitude. Conversely, curves of indicated
mean effective pressure not only shift, but also have a
reduction in magnitude. Due to the behavior of indicated
mean effective pressure and to the inevitable increase of
friction mean effective pressure running higher engine
speeds, curves of brake mean effective pressure are
shifted as the bore increases while magnitude is
significantly reduced.
By increasing the bore size, both power output and
engine speed for maximum power operation increase.
Conversely, for maximum torque operation only engine
speed increases while torque output reduces.
Engines having bore B=94, 96, 98, 100 and 102 mm
supply maximum brake mean effective pressure mepb
≅15.2, 14.9, 14.6, 14.3 and 13.9 bar at engine speed
NM≅14,800, 15,400, 16,100, 16,700 and 17,400 Rpm
respectively.
These engines also provide maximum power Pb=893,
907, 918, 925 and 932 HP at engine speed NM=17,200,
17,950, 18,700, 18,900 and 19,650 Rpm respectively.
The maximum power mean piston speed is 24.8 m/s in
the engines with bore B=94, 96 and 98 mm, while it
reduces to 24.0 m/s in the engines with bore B=100 and
102 mm.
The best engine has been selected by using the sharpest
car acceleration criteria. The best engine available is not
the engine that provides the best power curve vs. engine
speed, but the engine that provides the best power curve
vs. car velocity. This is the engine having bore B=98
mm.
Moving from bore 94 mm to bore 96 mm, then to bore 98
mm clearly improves the distribution of power output vs.
car velocity. Conversely, moving to bore 100 mm and
bore 102 mm does not improve to any further extent the
distribution of power output vs. car velocity.
The paper does not cover all these aspects pertinent to
general car performances.

[J. Edlund]

Originally posted by DOHCPower
Interesting. So in what area's do teams gain horsepower on other teams if not in cylinder head flow? Friction "maintenance" ?

These flows are measured under static conditions. The actual performance in an operating engine can differ quite a lot. The head with the highest flow may not be the head that gives the most power.

What the head flow can give you is a measurement of the port flow coeffiecient.

[DOHCPower]

Interesting point. But would you agree that the head that flows the most is the head with the greatest potential?

[J. Edlund]

Originally posted by DOHCPower
Interesting point. But would you agree that the head that flows the most is the head with the greatest potential?

Not necessarily, given equal conditions a higher flowing head suggests a port that has a lower flow restriction which is a good thing, but may not be what is most important for the power output.

I think port flow is much more an issue with production based heads where the ports may be very restrictive compared with the heads that are designed for racing use.

Originally posted by DOHCPower
Also, when questioning a recent 'expert', he revealed to me header design is primarily based on valve size. How you found this to be true, or does it have more to do with swept volume of a single cylinder or cfm?

The behavior of the pressure pulses will basicly depend on the ratio between the port and manifold area. Port area, the port to manifold ratio and combustion pressures used hence have a very large effect on both intake and exhaust pipe area. Length is more dependant on engine speed and the local speed of sound. Since the local speed of sound in exhaust is much higher than in the intake air, exhaust pipes must be much longer than intakes; something like three times longer.

[DOHCPower]

Very interesting on both points. I would assume that an engine revving to 20,000rpm would need all the flow it could get. Can you explain to me a design where a head that flowed 300cfm would outperform a head that flowed 400cfm, both engines same displacement, compression ratio, rev limit. I am thinking what youre saying here is that sometimes its the QUALITY of the air flowing into the cylinder that leads to power, not just the QUANTITY.

2nd point regardring the exhaust pipes needing to be something like 3 times as long, this is obviously because the exhaust gas/air is moving faster than the inlet air due to increased temperature, correct?

[J. Edlund]

Originally posted by DOHCPower
Very interesting on both points. I would assume that an engine revving to 20,000rpm would need all the flow it could get. Can you explain to me a design where a head that flowed 300cfm would outperform a head that flowed 400cfm, both engines same displacement, compression ratio, rev limit. I am thinking what youre saying here is that sometimes its the QUALITY of the air flowing into the cylinder that leads to power, not just the QUANTITY.

2nd point regardring the exhaust pipes needing to be something like 3 times as long, this is obviously because the exhaust gas/air is moving faster than the inlet air due to increased temperature, correct?

In order to get the outputs of a F1 engine you need a volumetric efficiency of something like 120% at maximum power speed. This means that you must take advantage of several effects to fill the cylinders, intertia of incoming air and pressure pulses. These factors may not benefit from the highest flow. The highest flowing head may not be the highest flowing during lower lifts, this must also be considered. Then we have air motion, squish areas and several other factors that can have a positive effect on power even though the flow may be reduced.

Pressure pulses travel with the speed of sound. The speed of sound is higher at higher temperatures (but is also affected by other gas properties) hence the requirement of longer pipes in the exhaust.

[DOHCPower]

Bringing this thread back from hiatus, i got to thinking some in class today... (imagine that)

Lets just say at 7000rpm you have a cylinder head that flows 250cfm through 2 30mm exhaust valves. At this particular engine speed, with this particularly flow capability through these size valves, there is a calculatable volume of space being pushed out the exhaust port. By properly fitting tubing diameter and length, is it possible to create a negative pressure wave that travels down the exhaust pipe the exact moment the exhaust valve opens and the piston begins to push the spent gasses out? The negative pressure wave would "pull" the exhaust gasses out, correct? Not unlike the pressure wave the pulls intake air into the cylinder bore on the intake stroke.....

Am i on the right track?