Ok, how do they go about this design wise? Does rod/stroke ratio play a part? Engine RPM?
I wanna hear the science behind it!

Thanks
-Mindgame

 

Well there's always an "initial" design, in which various manufacturers have a few empirical methods. Final design will come by way of computer simulation and computational fluid dynamics modeling.

You can arrive at an "initial" concept through use of Mean gas velocities as a function of mean piston velocities.
This gives us the mean gas velocity (Vg) through the valve section:

Vg=(d1^2/d2^2)*((n*s)/30000)

Where d1 = bore diameter
d2 = valve head diameter
n = engine speed
s = stroke
[I'll use SI units since this is what I'm most accustomed to in working these formulae]

Now:

(n*s)/30000

This part of the formula is used in calculating the Vg.
This formula can, with a little more complexity, be used in calculating instantaneous gas velocity through the valve section at any calculated piston speed... instantaneous etc.

Now, the above formula used in a manipulated form (taking into account valve stem diameter intrusion if required) can be used to work out mean piston speed throughout the port/"dog-leg"/throat area.

(cylinder size (cm^3) * n)/(3000*Vg)+((valve stem area)/100)

Now what velocties to design for is a difficult one.

The port must achieve a few things:

1)Flow - must therefore acheive as low a restriction as possible

2)Impart an appropriate level of motion (be it tumble or swirl for the application) such that the engine can achieve good combustion stability (very important for emissions) and suitable burn characteristics

3)Ram tuning- Be correctly sized to achieve momentum ram effect at a designated engine performance envelope

In practice a compromise of the above needs to be achieved. The LS1 cylinder head is a prime example of this compromise.
The last point is where the port sizing comes into account-ram tuning.

I can only give examples and also say that the smallest port that gives the best outright flow will acheive points 1 and 3.

Since we are talking production "OEM", let's look at an example.
The BMW M3 3.2 litre EVO (100 BHP per litre, almost 14 bar BMEP, naturally aspirated- production engine.)
has a mean gas speed through it's economically sixed, exceptional flowing ports of 119 m/s in the dog leg at 7400 rpm. The ports are fully machined, which means they would flow about 3 % better (at valve lifts above d2/4) then if not machined, which means they can be sized 3 % smaller to achieve better ram tuning.

An AJ27 jaguar V8 engine achieve a port leg velocity of over 101 m/s at its peak power speed of 6000 rpm.(72.5 Bhp/litre, 12.7 bar BMEP).

Ive seen velocities range from as low as 80 m/s up to 120 m/s at peak power speeds.

And finally one must consider the mean inlet mach index number or Z.
As this number comes closer to 0.5 the port is starting to choke (reach sonic velocity at the minimum valve flow area). Choking substantially reduces volumetric efficiency.

Z can be ballparked with this formula:

Z= (1/2)*(bore/stroke)^2*(mean piston speed)/(mean flow coefficient*347)

Perhaps a more realistic form of the above formula takes into acount the valve lift profile also, and the above formula adapted to suit a valve lift profile is known as the gulp factor.

Do you need to know more or will this suffice?

As you can see, there are alot of variables to deal with but this is how a design is put on paper.

Hope this helps.

Chuck Riddeck
Progressive Race Engine Development

 

Maybe you can help me understand but is this important or as important in a dry intake as long as the quench is set right? I would think that velocity would increase with less turbulance before the valve?

 

Mr. Horsepower's response was very good. It's a good read.

A few more thoughts:

In OEM EVERYTHING is a compromise. Emissions, fuel economy driveability, durability and production costs are right up there in priorities with power, and often fight each other. Remember some of the stuff like emmisions is government mandated; you can't fudge on that.

Everything about the engine (and the vehicle) plays a part in the design. Yes, there is lots of computer modelling, but there are also LOTS of engines built and tested.

The LS1 design was quite a few years in design/development. The race version took a lot less time. Some of this is because there are fewer compromises, and deeper pockets for the cost of parts.

My $.02.

 

Thanks Chuck!
That's quite a bit of information to chew on and until I can digest of all of it, I won't be asking any more questions.
Nice of you to break it down to just the port. Don't believe I've ever really seen it explained like that before. I understand the issue with engine testing, prototypes, design etc. but I was more concerned with the specifics of that one aspect of a cylinder head. Otherwise the thread kinda lingers off into alot of different subjects.

BTW, you mention the LS1, why do you say it was a compromised design? Whoops! Lied about not asking any more questions, sorry.

Thanks again.

-Mindgame

 

And without getting too technical on the subject, you'd be thinking right.
Manipulating charge delivery is energy wasted. And I'll stand in saying that it is not a priority unless you're designing a diesel port or you're attempting to thwart reversion. Specific instances of such would be the port of restricted competition engines, where creating strong tumble away from the exhaust valve (about an axis perpendicular to the bore axis) is somewhat effective in minimizing reversion.
Every engine has it's own needs so it is impossible to make a blanket statement on the matter. You simply build/design everything to the engine's needs.

Yes, the port compromise was made in the name of injector targeting... ala lower emmissions, better idle quality. With this comes what we like to call, a "Zero interference port shape". The result of which was a low-flat port floor and a short side radius that is too tight for optimal performance. They go hand in hand. The LS6 head, on the other hand, was a less compromised design, built the way Ron Sperry wanted to build the head in the first place.... refining this less than optimal port condition.

Hope that answers your question?

Take care,
Chuck

 

I Didn't realize that sir!
All makes good sense though. Thanks again.

-Mindgame

 

Thank you

 

Old Post from the past!!!! And look who asked the Question!

 

Where is Chuck anyway? He made a few posts last summer but he hasn't been back since i don't believe.

Looking at his post count, it appears he used to post a LOT back in the day. Boy i'd love to be able to dig up all those old posts that have been cleaned off the server. He was an entire world of info.

 

can someone please explain the post or answer by mr horsepower a little more in depth so that I can better understand it.

 

Probably not very easily. That WAS the simple version, unfortunately.

If there is one specific point you question, perhaps someone can clarify it. If any of the terms are strange to you, research them first. Google is your friend. For example, try "bmep" definition for a Google search.