Flow coefficient of a "head-valve" system

[Robert]

Hello Guy,

I have a question about the flow coefficient of a "head-valve" system.

If you get the curve Flow=f(Lift) and you want to normalize the flow in % of the "ideal" flow to obtain a flow coefficient

Cflow = Flow effective/Flow ideal

What is the value of Flow ideal we have to consider ?
Is it the flow obtained without valve ? Is it a theoretical flow obtained by calculation ?

I found that this flow coefficient (its average or mean value) is useful for the sizing of the valve.

Many thanks in advance,

Robert

[Guy Croft]

I am not sure I would ever want to take a mathematical approach to this. The coefficient you refer to is not a constant because the aerodynamic behaviour of the medium (air-fuel/gas) varies significantly with the relative position of the valve. All I can tell you is the smaller the valve the better its discharge coefficient. Two small inlet valves, for example, with area equivalent to one large one will flow more. One reason among many that mulitvalve heads give superior power. The area of the valve head should generally be such that the ex valve has min 70% or ideally 75% - 80% of the inlet and much depends on the engine capacity and port sizing as to what is best.

To continue the discussion you would need to present your findings.

G

[TomLouwrier]

Just a quick thought: if you take "the integral of (flow as function of (the lift curve combined with the flow found for the corresponding lift heights))" you should be making some sense.
Probably not explaining myself very clearly, but it's been a long day and my math's on the blink a bit.

Then again: real gas flow in a running engine is much different from the values seen on a flow bench. The relation (correction) between these could be found by logging air flow on a running engine, at different speeds. You'll deal with pressure, gas speed, Reynolds number, turbulent flow, aerodynamic properties of the ports...
All in all so many correction factors involved you would likely be better off sampling air flows as mentioned, on a lot of engines and see if you find a pattern between valve sizes and cam profiles. Empirical approach.
I'm quite sure there is an answer in there, but the amount of data you need in order to go do some meaningful mining is huge. This would be an effort on Ricardo / Lotus / OEM level.
(Don't make me drool now...)

Multiple small valves flow better than one large one because of the relation between throat area and curtain area. One changes with D^2, the other with D. Set it up in a spread sheet and experiment a bit with different sizes. You'll see.
Honda once built a race engine that was limited by regulations to a maximum of 4 combustion chambers. So they gave it oval cylinders and put in 8 valves per chamber. Massive flow, corresponding with a V8 of similar capacity. The NR500.

regards
Tom

[Robert]

Hello,

I have the answer in your book Guy (the first one). Page 48.
If you take the integral as explained by Tom (Excell sheet with trapezium method) of the flow A of the page 49 (bare 130TC head) and normalized it with the Qth = 21.8*d²sqrt10=138 cfm (with d= 36mm for a 130TC, value given p45) you can find a Cdm=47.6%.

Sorry but now it is quite difficult to me to explain what I mean because the(my) English and because I don't really master all the theory found elsewhere. In fact the velocity relationship mentionned p52 certainly leads to what I see concerning the Mach Index.
Because it is writen that IM must be less than 0.6 to get a good flow rate. I can verify that using 43.5mm valve @ 6500 rpm the MI (with Cdm=47.6%) is about 0.44.
Or that the max revs is 8700 rpm to get MI=0.59. So there is a big margin between the OE spec and a max preped spec...

Robert (Am I clear ??)