head port size

[brashear]

Hello,
I have found plenty of information on valve port diameter based on a Mach index, but haven't found anything on head port (the port that connects to the intake manifold) size. I know that having a smaller port will increase swirl and lead to better power at low RPMs, but is there a point of having the mach index on the head port greater than the valve mach index? I am using throttle body injection, if that makes a difference. Thanks all.

 

[Lou Scannon]

Hmmmn... your OP is the equivalent of dipping your big toe into the Pacific ocean. How deep do you really want to go?

 

[patprimmer]

I am sure port CSA directly influences velocity, but I don't quite see how it impacts on swirl.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[evelrod]

I agree with Hemi....pretty open ended. Even the "greats" of head design have had some 'stinkers', e.g. some of the early Gurney-Weslake designs based on faulty Ford port dimensions.
I read a study by a Japanese engineer that had some 'different' ideas about port size/shape v swirl...Sure wish I could give you more than a whole country to search...Also, I recall a study by Suzuki on combustion chamber 'swirl' from the early 80's (I think).
For sure, a smaller port will generally have a higher exit velocity at the valve, promoting more combustion chamber turbulence.
A good example of a cylinder head with ports and valves too big would be our current Lotus twincam...putting the biggest valves possible and the largest ports possible results in a loss of hp, a substantial loss. Changing to smaller intake ports, larger exhaust ports, smaller valves resulted in a gain in hp and tq !

Rod

 

[patprimmer]

Rod

I can certainly see where increased velocity past the valve head increases general turbulence and in a 2 valve design might increase swirl in the chamber.

Certainly to a point velocity is important to power, especially after the point where the fuel is added.

Certainly air speed should be subsonic.

Certainly inertia in the column is important to filling while the valve is open after BDC.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[thundair]

Another port to big for its own good was a 4V 351 Cleavland Ford.

I read some where maybe on Speed talk that a max velocity was about 520 fps.
Here is a calculator

http://www.speed-wiz.com/calculations/engine/port-velocity.htm

Also most ports are about 80% to 85% of the valve diameter.



I don't know anything but the people that do.

 

[evelrod]

A little background on our "twincam" efforts over the years.
Currently we are getting 206 hp from our Lotus 1594cc, reliably. No drama with this one (fingers crossed).
Efforts in the past to get max from an 1600 cc Alfa was not effective until switching to a "GTA" twin sparkplug head.
Efforts early on with a Fiat 124 cyl head were pretty much a waste of time. About 175/185 was tops, even after extensive valve angle changes and port changes.
So, my point, just saying that even though all three of these designs are very similar, each exhibits different results when modified for max performance.
Yes, a bit of "swirl" can be introduced in a two valve setup and, possibly we may be benefiting from that...I'm not convinced that it is not simply turbulence, tumble, whatever...swirl if you want, that is causing the 60's Cosworth design to be more efficient than the Italian designs.
This is not "news" as all these engines are <50 years old and the power figures that we are getting today are pretty much the same as was common in the distant past...pretty much!

In fairness, I should point out that the Cosworth design was a race specific effort while the Alfa and Fiat engines were street engines modded for race. Makes a difference!

Rod

 

[evelrod]

Thundair:
Pretty close there on the port size for our latest...
1.625" intake valve and a 34mm port....
Actually these dimensions were on a 1972 Brian Hart/Vegantune rally head I had in 1980...Like I said, ancient history as modern four valve designs have long since surpassed any thing out ancient Lotus twincam can produce....Then again, our efforts are for VINTAGE RACING and must adhere to the 1972 rule book (We try).

Rod

 

[adrag]

Not to hijack but I have a question about that port velocity calculator on speedtalk. I ran the numbers they input and got a much lower port velocity. Isn't it just s=Q/A? So:

Port Dia=1.500"
Bore=4.00
Stroke=3.500
Rpm=5000

Port csa is (1.500^2)*PI/4 =1.767 in^2
Cyl volume= (PI/4)*(4^2)*3.5=43.98 in^3
Volumetric flowrate= (5000rev/min)*(.5 intake/rev)*(43.98in^3/intake)=109,950 in^3/min

s=Q/A=109,950/1.767 = 62,224in/min or 5185ft/min
86ft/s
58.9 mi/h

Much lower than their numbers. This is assuming 100% VE at that rpm of course.

 

[brashear]

Now that I think about it, I don't think making the velocity higher at the head port than at the valve will be advantageous, as far as velocity is concerned. The final velocity of the mixture entering the chamber should be determined by the valves. A greater restriction to the flow prior to the valves will, I think, just cause the flow the speed up, then slow down again at the valves, and achieves nothing except pressure losses.

adrag:
Why did you use 0.5 intake/rev? If each head port is responsible for only one cylinder, and neglecting any buffer and inertial effects, the port is only "used" during the intake stroke.
Here's how I calculated it:

Average piston speed:
(5000rev/min)*(2strokes/rev)*(3.5in/stroke)=35000in/min=48.6ft/s

since the area of the bore and port are proportional to the square of their diameters, so too must their flow speeds

Average port speed=48.6ft/s*(4/1.5)^2=345.6ft/s

Couldn't find your speedtalk calculator, but

http://www.wallaceracing.com/piston-speed-velocity.php

gave the same result as my calculations.

Cheers

 

[adrag]

I used .5intake/rev because assuming it's a four stroke there is one intake event every two revolutions so intake/2rev or .5intake/rev. I see what you're doing though by calculating avg piston velocity and then using the ratio of bore to port csa. Is this the typical method to calculate port velocity? Why does the method I used give a number that is so far off? I remember calculating fluid velocity from volumetric flow rate and csa all the time in my fluids class many moons ago. Also, that wallace calculator gives a higher reading (407.24 ft/s) compared to your calcs.

 

[brashear]

Right, there is one intake event every two revolutions, but since there is no buffer between the head port and valve, the head port has only air flowing through it during the intake stroke. By using 0.5intake/rev, you are assuming that the intake port has a constant flow through it during all exhaust/intake/compression/expansion strokes, which is not true.

Think of it like this: you have a short straight pipe with two restrictions, the first being the head port, the second being the intake valve, followed by a piston. A piston moves at a certain speed and is part of a slider crank. What speed? Say the crank rotates 5000 revolutions a minute, and for each rotation, the piston moves up by 3.5in, then down by 3.5in. The average piston speed must then be: 5000 revolutions a minute, times 3.5in * 2 (down and up).

The square of ratio of diameters is just a shortcut I used. You can get the same results with your calculations by replacing 0.5intake/rev with 2strokes/rev.

For the Wallace calculator, make sure you put the port area (1.767 in^2) and not the port diameter to get 346ft/s

 

[Marquis]

When I design gasoline engines for performance - if the port is cast- I usually size for a mean gas velocity at peak power engine speed at the cylinder head entry of about 80-90 m/s. A fully machined port like used on some BMW- perhaps an entry velocity of 95 m/s.

These are guidelines and based on years of empirical benchmark data of good engnies-nothing is cast in stone.

http://www.auto-scape.com

 

[TrackRat]

In my experience you would not normally want the port area to be small enough to have a higher mach index than the valve seat I.D. as you are essentially restricting system flow at peak rpm and gaining nothing at low speed because the valve controls the flow not the port area. The port area is used to adjust the gulp factor and peak velocity for high rpm use.

Higher velocity may or may not increase swirl or tumble which primarily impacts low speed and part throttle power vs. mid and high rpm power. Tumble and swirl are of limited use unless your primary goal is low BSFC and emissions goals. Squish and chamber shape for fast burn are more productive across the entire operating band, for best power in my experience.

Port sizing can be done in several ways. What has delivered the best results for me is to match the port area at the head flange to the peak flow area of the port + 5%-10% depending on how high an rpm you plan to use. For the intake manifold I typically use a 2-5 degree taper depending on the runner length and intended peak rpm.

There are numerous ways to achieve a desired goal. Some are better than others. These are what work well for me on many different engines. YMMV.