Greendomize

67King has made some good points and is very knowledgeable, but it can be quite difficult for others to understand since his descriptions are very quantitative. I'm a Mechanical Engineering major at Georgia Tech and I have recently taken an entire course specific to internal combustion engines, so I figured I would share what I know. I'm also a product validation engineer for Cummins Inc.

First the goal of any intake system is to increase volumetric efficiency, which is a way of quantifying how effectively an engine is taking advantage of its displacement volume. It takes into count terms such as the number of revolutions per cycle of the engine, mass flow rate of air, air density, displacement volume, and engine speed. Obviously you want a larger mass flow rate of air, so as the mass flow rate goes up your volumetric efficiency will increase. As air enters the intake and makes its way through the intake manifold and the intake valve/valves there is a resulting pressure drop (you're essentially choking the air flow), causing a reduction in the mass of air entering the cylinders, decreasing volumetric efficiency. Since this is a discussion specific to cylinder heads I won't talk about intake manifold design, but the largest restriction in the intake system is at the intake port. Which is especially true at high engine speed because the incoming air is much more turbulent. This is why you will always notice that modern engines have a larger diameter intake valve compared to the exhaust valve. Another reason behind the intake valve being bigger is because hot exhaust gases carry much energy than a relatively cool intake air charge (therefore flows faster) and can exit the combustion chamber much quicker than the intake air can enter it. Intake valves need to be larger to maximize volumetric efficiency becuase losses in volumetric efficiency in the intake have a much greater effect on engine performance vs the minimal flow restriction caused by having a slightly smaller exhaust valve.

There is actually an equation that we use to determine the minimum valve intake area necessary for a modern engine, but I won't go into detail on that unless you guys really want to know. Regarding valve design, there is often not enough wall space in a cylinder to fit spark plug and exhaust valve/valves and still have enough space to fit the minimum valve intake area. For this reason, most engines are now built with more than one intake valve per cylinder because you can more effectively make use of a given cylinder bore by breaking one larger area into two or three smaller areas. Having multiple intake valves gives more flow area (less volumetric efficiency loss) and less flow resistance than having one larger valve. Also, when two or more valves are used instead of one, the valves will be smaller and lighter, which can be opened and closed much faster than a larger valve (allowing higher engine speeds). Smaller valves also allow for the use of lighter springs and reduces the forces in the mechanical linkage.

Like someone mentioned earlier, having multiple valves does add cost and complexity to the design because you need a second camshaft and additional mechincal linkages. In general, the greater volumetric efficiency of multiple valves overshadows the added cost of manufacturing and the added complexity.

Hope that helps!

-Ben

67King

Ben, I'm a Ramblin' Wreck, too. Undergrad in 97, Masters in 2000. You might appreciate the livery of my race car. Great school, obviously. I hope you are doing FSAE if you have any intesrest in staying in the industry.

BTW.....you ever run across a PhD from Tech at Cummins named Dinesh Bansal? He's a tribologist, not an airflow guy, though.

I assume the equation you are talking about is related to Z-Factor. I mentioned it briefly, but it makes some assumptions, such as a flow coefficient based on lift/valve diameter. That's a great first cut. When I had actual cam profiles and flow coefficient information available, I used that. One of the issues you run into is that you can only run so much lift, and it is easier to get larger L/D from smaller valves. One interesting thing I found out was that many of the big 2V's in production, like the Hemi and some of the LS engines, never get into the Zone III region of flow (I is seat area limited, II is seat shape limited, III is port area liminted), because they can't lift the valve enough to get out of Zone II.

LS1Porch

I worked for a very knowledgeable head porter for a few years and we put some of my old stuff on his flowbench and i read through all his Reher-Morrison training manuals.

I didn't read through this whole thread, and i doubt i could say anything that hasn't already been said, but i'll say this: Your heads are pretty decent. If you were talking about them for an n/a application they'd be pretty lackluster for a number of reasons already mentioned. But you guys are talking about forced induction and for that purpose i doubt they'll be a limiting factor.

Greendomize

Awesome car! I'm not in GTMS, but I am a member of Wreck Racing. I chose that over GTMS because I'm more interested in building production cars for the track than I am pure racecars. Although, they are still very interesting. I have aspirations of turning my 951 into a racecar once I graduate and buy another daily driver. I'm in the process of a thorough engine rebuild at the moment of which I will be posting on the forum in a few weeks once everything is complete.

The equation is was referring to was: A = 1.3*B^2*[(Up)max/c] = (pi/4)d^2
where A is the total inlet valve area, B is the bore, (Up)max is the average piston speed at the maximum engine speed, c is the speed of sound in the intake, and d is the diameter of the valve. i.e. A = (pi*d^2)/4 for 1 valve and A = (2*pi*d^2)/4 for two valves.

Regarding what LS1Porch said, the turbocharger does not actually eliminate volumetric efficiency losses. Yes, on a naturally aspirated engine your volumetric efficiency will always be less than 100% because no intake system if perfect and turbocharging the engine can raise the volumetric efficiency to levels above 200%, but there is always improvements to be made in the intake system as the efficiency could be even higher than what it already is in the 951.

333pg333

Bump...you guys should watch this if you haven't. VERY interesting!! Mr K is a very open guy about his products.

67King

THis is actually not true. Tuning can get it well over 100%. The new 5.0L I worked on needed to hit 108% to meet its initial target (400 horsepower). It exceeded its target, and is closer to 111%. And that is just at peak power, I don't have it for peak torque, which will be higher. Gross numbers (no air induction system or exhaust system) for several engines are well over 120%.

The BMW engine Michael Mount and I were talking about had VE's north of 130. Hard to come up with an exact number, as I had to infer FMEP. I had pretty solid FMEP numers to start wtih when I was doing the 5.0L, and tweaked them to account for the changes from the baseline engine (4.6L 4V). I had estimated that we'd pick up about 30 horespower with the changes I recommended, we actually picked up over 50. I still have the target setting worksheet I came up with, and was showing 120% VE at peak power. But again, it exceeded my predictions by a pretty fair amount - due primarily, probably, because I didn't account for who was building it and doing the blending with the 34mm valves.

Again, that is ALL about tuning, though. If you take a look at BMEP number for various engine architectures, you'll see that V6's tune harder than anyting else (well, I6's, too). I4's are not far behind. V8's don't tune worth a crap. That is because the cruciformed crank makes the pulses in both the intake and exhaust overlap, and pull down VE to some cylinders more than it should. THat's why Ferrari's use flat cranks - it lets them tune like I4's. I've heard that the E92 M3 also uses a flat crank, but I haven't verified it.

All that said, this still does have some level of importance on our engines. If you look at the Helmholz equations, you'll see that pressure is not a variable. The caveat here is that longer runners will help you tune more earlier, which will get the turbo spooling a little faster. But outside of that, I still like to engineer the intake, because I'd rather do it with tuning, rather than boost, because the air charge temperatures will generally be lower.

Greendomize

Patrick, I saw that earlier today. Very cool idea to see in action, but the idea itself has actually been around for quite a while now. I always figured the valves would be electrically actuated as opposed to being pneumatic, but ultimately either way is a significant improvement.

Harry, you are right I forgot to account for intake tuning. Although, I thought that this was only for a specific RPM range. At what RPM's is this effective on the 5.0L? By the way those new 5.0's are very appealing to me. That might be the car I choose to buy once I graduate.

URG8RB8

Green:

I think you would be amazed at just how many of us on this forum are engineers. There is a thread about this some time ago. I have been an engineer longer than you have been alive and still find Harry's posts quite exceptional! Now that I see he also has a master's from GT, it is of no surprise! At least we (UF) have a better football team!

Harry:

My colleague here in Thailand has a masters in Tribology, you simply don't hear that term very often.

67King

Finally had a chance to watch this. Unfortunatley, this concept is probably at least 2 decades old. I've known about it for 15. It has been tried with electronics and hydraulics so far. This one appears to be pneumatic. Now, IF he can get it to work, that will be absolutely phenomenal. it holds an incredible amount of potential. But it is a tough nut to crack. Despite that, though, I'd say it has a greater chance of making it somewhere with a small, nimble, and focused company like Koenigsegg. Man I would have loved to have worked there!

Here's a 12 year old article outlining some of the same stuff. Just did a search for 42V valve actuation and found it: http://www.edn.com/design/automotive...ce-engine-cams

Duke

Apparantly they converted the head of their testmule, a Saab 9-5, to run the pneumatic valves and have had it running for 60000 kilometers for everyday use with cold starts in -20C etc.

michaelmount123

Who's gonna get this thread back on track?

V2Rocket

So with some of the above-mentioned details on the cylinder head, is there anything to be done fairly easily to get a little more flow efficiency out of it? King has said that the ports are pretty good due to location and shape, and in the past people have said that they are already large enough for much larger amounts of airflow. Someone said the bigger restriction is the shape of the intake which can be solved with a variety of options. But as someone else said, the intake valve is very close to the chamber wall.

Could that be alleviated somewhat by simply beveling the chamber wall on the sides where it is close to the valve heads, for engines with larger than stock bores? The chamber would still seal up to a larger bore but the extra space might allow for opening up the "shrouded" area.

67King

I'll try to do a little modeling, and see what the "ideal" valve size would be. When I did this with a Ford 2.3 that was stroked to 2.5L, the answer I got was 47mm for peak power at 6500 RPM, but that engine had a VERY stout bottom end. That also comes with the caveat that I had to factor in what size aluminum tubing I could get since I was also doing a custom intake. It is a system, not a collection of components.

I don't know how shrouded the valve is, I don't think it is very shrouded.

If you want to make peak power at 6500 RPM, you'll need shorter and ibgger intake runners. I know that there are a lot of things that I don't like about the intake, but I'd also caution folks against looking at an intake like a cylinder head. Its primary job should'nt be considered to flow well, but rather to tune well. It needs to do both, but if you ignore the tuning and just try to make it flow, you'll get disappointing results. Or at teh very least suboptomized ones.

I'll see what I can throw together maybe this weekend. This is all in theory, though, which is only half of the equation.

Thom

Agreed - my (good) aftermarket intake, featuring larger and much shorter runners, has freed up the breathing ability of my engine between 5500 and 6500 rpm, which is where the stock intake was strangling it - I was getting compressor surge from the middle of the rev counter up to 7000 rpm with the stock intake, which is not the case anymore.

TurboTommy

What would the internal flow area of each individual intake runner be, as compared to the individual intake ports.
Wouldn't that be a major factor?