Cylinder head question: Port cross sectional area over different conditions

Hi all,

This is my first post to the forum, and given the huge knowledge base here I think this is the best place to get my question answered.

My general understanding of port cross sectional area is that is should be primarily dictated by rpm and cylinder displacement demands. It has also been my understanding that forced induction methods are a density modifier and not necessarily a velocity modifier in the context of our engine. However, typically air moves quicker from a point of higher density to a point of lower density, so in theory higher boost pressure would lead to greater air velocity going through a port. Considering the airspeed limit of around 200 m/s, in theory with enough boost we would need to alter the CSA to accommodate the increased velocity.

In real life situations the boost increase also heats the air which would have a counter effect on the velocity due to the reduction in density. I have read so many different perspectives on this topic by all kinds of experts in the industry that my head is starting to spin.

I also have questions on valve sizes relating to this topic, but I think that discussion may require its own thread.

Curious to hear your perspectives.

Demetri

That's a very good first post, and excellent questions! I'd add a couple of other parts, too - because the pressure gradient (differential) between the manifold and in the cylinder, what affect does the increased pressure of the cylinder 'depression' on the intake stroke, from back pressure in the exhaust manifold(s) have on the filling, and does the increased port pressure alter the ~200m/s port velocity, as the trans-sonic velocity will change with the pressure?

There have been several tests, most available on-line, with different port and head variations being tested on N/A engine, mostly V8s', but I don't recall anyone doing a similar direct comparison with forced induction, whether supercharged or turbo-charged. The small block/LS engine has MANY such options, and would seem to make for an excellent 'proof of concept' test bed.

Different engines, but if you haven't done so, I'd recommend checking out Gale Banks' channel, as he's done a LOT of reseach into turbo'ed engines in his very long career - mostly diesel of late, but same principle applies - more air-mass in = more power!

https://www.youtube.com/@bankspower/featured and especially https://www.youtube.com/watch?v=iBoyW2e0ra0&list=PLwtmrqcWzLtQsQGxE-x2i8GaPgEuItUkn

David Vizard is another VERY smart chap, but his presentation style can be rather 'challenging'.

Well, according to Vizard what works on NA engine will work on forced induction engine too. That's why he (together with another guy) made a special program for head porters to measure 4 most important characteristics that will tell you how efficient the porting is. As always it's a combination of air mass and velocity wich results in air charge kinetic energy (as we all know it's half of mass times velocity square Ek= ½ mv2) that has huge factor (but not limited to) in cylinder fillling. Providing the back pressure factor can be kept about to 1:1 the rules for forced induction would be the same as for NA engine.

From my experience with forced induction it's always better to stay on the tight side rather than go to bigger sizes. For instance, the smaller diameter throttle body gives better response and more power (to a limit) due to higher air velocity that results in higher kinetic energy since velocity is square in formula. The same with oversized valves. Increasing valve size 1mm on 4g63 isn't beneficial unless you have camshafts with more than 11.5 mm valve lift whilst increasing valve size 1mm on Evo X is absolutely useless- you will make more power with smaller valves.

As it appears to me air velocity plays huge role in power outcome. However it should be a balance - if you want to make 2000 hp instead of 280hp that the engine was making from factory you definitely need to make some calculations on air velocity and make adjustments if required...

So much to say! Thank you for your replies, I am very familiar with both Gale Banks and David Vizard. Gale's videos taught me about air mass, David taught me... well a lot of things, but I have found myself becoming less enthralled with his work as of late because I have found he may be a bit rigid in his philosophies and not really explaining the "why" of how things work (I have a lot of his books).

I digress

Gord: In theory there is 'optimal' cross sectional area for every given combination (regardless of hp goals) that can be mathematically calculated (I've lost track of the equation) which combines the maximum air mass with optimal velocity. I have done a lot of research on the subject, but by far the most digestible resource on this subject on Youtube is a guy named Bain Racing. I have linked the channel below. My question remains the way you put it: does the increased port pressure alter the ~200m/s port velocity, as the trans-sonic velocity will change with the pressure?

I hope someone can chime in on this.

Georg: I am a big supporter if the 1:1 ratio due to compromised cylinder filling capabilities with un-evacuated exhaust in the chamber with higher back pressure ratios (as I'm sure you know). Regarding valve sizes, this is one of the mysteries I'm trying to solve regarding engine design at the moment. The old method of "lifting the valve X% of valve diameter" (a theory vizard espouses in his books) has not held up in modern engine design especially if you look at even street-oriented motorbike engines. As stated above, the 'optimal cross-section' theory states that the cross section for optimal VE at 300 hp, is the same at 2000 hp, assuming rpm range and displacement remain constant. In practice that is usually not the case with a 2000 hp (due to displacement/rpm increases) engine but for the sake of the theory this should be the case.

Sorry for the long answer, any further insights are greatly appreciated.

Bain Racing: https://www.youtube.com/@bainracing/videos

Since you are very familiar with David's work you should be aware that it's not only flow numbers, air mass and velocity are in the picture. Things like port energy and port density are as equally important. So as I said before - it is always a combination of shape and geometry on both sides that gives you the most efficient port. To me this is where the knowledge and experience kick in over just pure math that's only good in theory. For instance, I remember Vizard's story where he talked about old times when he was working on some small displacement engine. He was trying to install most efficient carburettor that would allow him to atomise fuel droplets as small as possible but ended up making much more power with carburettor that did exactly opposite- atomised fuel very poorly with big fuel droplets. The results was achieved against all the existing righ formulas and rules and to these days he doesn't have any explanation why it happened... My point is that knowing the theory not necessarily will always get you to the best achievements but combination of knowledge and practice will.

A quick check suggests the pressure will make negligible difference, with the higher temperature possibly increasing the transonic velocity slightly.

So, if you consider it as "x" volume of fluid (air) passing through a CSA of "y", in "z" fractions of a second, you can swap around the values to get an estimated average velocity, which is pretty much the same for N/A and forced induction.

However, theory doesn't always match empirical results, as has been observed with the aforementioned testing.

With the droplet size, I would surmise the evaporation of the fuel, reducing a small percentage of the mass of air, had more negative affect than the possibly poorer fuel burn with a little more oxygen available. Some engines being rather more sensitive to the latter than others.

So to summarize thus far (feel free to correct me): Theory would indicate that CSA (optimized for correct velocity and mass flow capacity) remains relatively consistent largely irrelevant of boost pressure, but empirical testing and evidence may suggest that there may variables that are unaccounted for (that is at least how I would put it). At this time I am in no position to conduct testing to put these theories to the test, but your responses have given prompted some things to think about.

Thank you both for your insights this has been insightful and what I was hoping for from the forum. Curious to hear from other members as well.

By the way, forgot to mention another important factor for efficient port- discharge coefficient. The throat under the valve seat should be big enough but not to exceed certain percentage of valve size. If done properly it will provide best exhaust gases velocity which is crucial for performance engine. Too small- choking at high revs, too big- velocity is too low. In both cases you're loosing power to be had.