Well that took a while but I finally managed to make a picture of a few simulations. Never mind the ridiculous names (RP , fear not, it is not). Power figures are halfway realistic but should not be taken too seriously as I have been fiddling with burn models and it is based on an actual race engine and we don’t want to give too much away, do we . If you want it to say 200 bhp it can be arranged..
Two power curves, the only change is the exhaust. The black line has a purpose build 3-1 race header. It looks more impressive that it is due to the scaling of the graph, but it helps readability.
Before anyone starts emailing angry : yes it starts at 4000 rpm and goes to 9000 rpm.
Is 9000 rpm a good idea with a 3.2 inch stroke? Basically yes if you are interested in only power, but there is a large BUT, and that is that the piston speed is 24.4m/s. At 9250 it is 25 m/s and at 9800 27 m/s and that means that it then has a higher piston speed than a world super-bike, 20.000 rpm old skool v10 F1 … Basically the only things that have an even higher piston speed are top fuel and pro stock dragsters (ca 30m/s) but they only have to last for a few seconds.
Is >9000+ rpm possible for standard stroke an A-series engine, principally yes, but it will not last long (as in road engine long) . It will also need some very fancy parts to actually make more power. Starting with a Killer 5 port head or an 8 port or better a KAD or BMW-K 4v with porting. Then a nice crank,rods, light pistons so it can stay together and in the end it will be very very expensive. They do tend to go Co Coo on you as well, and it involves oil, blood, sweat and tears and an overheating credit card.
But any-ways, spinning a bit faster in a race setting is not such a bad idea if you wallet can stand the stress and it actually produces a faster car ( I’m not even saying more power, because you can have more peak power and go slower). However the 5 port head does not breathe all that well at higher rpm’s, so it does make sense to not rev it where it has trouble breathing and try to make as much power as possible is the rpm band where it does breathe ok.
A more fundamental question is why does it not want to do it. The head flows about 130 CFM and the often used rule of thumb says 2 bhp per cfm (which is BS I know, but still), hmm maybe try 130 x1.67 ( which a more conservative rule of thumb number coming from Superflow) is still 217 bhp..
So why do we only get 123 ish simulated BHP? Well apparently the rule of thumb does not work for an A series, but it does for a lot of other engines otherwise it would not be a rule of thumb ( a classic case of wrong but not unfounded).
Breathing is not CFM flow capacity on a bench then. Well that is bad because I know exactly how to get a lot of flow.. use a very big hole. What makes power are not high cfm numbers but high trapped mass numbers. It is chemical power (there is a very good NASCAR 101 lecture on this on youtube) in the end that makes cylinder pressure, which turns the crank, which makes torque.
TORQUE * REVOLUTIONS = KW
And KiloWatts are a measure of power per second (1000 j/sec)
So to make lot of KW you need TRQ and RPM.
For a high trapped mass you need a efficiently working induction tract , meaning intake and exhaust working in unison with the cam timing to cram the absolute maximum amount of mass into the cylinder, then close the door quickly and hopefully the mix is in a state that will burn in a manner that produces the right amount of cylinder pressure at the right part of the crank rotation.
Ok via this enormous segway we come back to the root problem : half of the engine has a very bad bmep at high rpm. The other half is actually not all that bad
As I have stated before : In an A series the inner cylinders do have the problem that the exhaust is shared, but to make matters worse it is basically tuned wrong for the rpm the outers are happy at. This results in a lower trapped mass and less stuff to burn, causing less cylinder pressure
The P/V Loops that illustrate the difference in power (here surface area of the upper loop) for individual cylinders.
The red and green lines shown in the figure show that the 2 & 3 cylinders basically keel over at 6500 on the engine with the needlessly complicated name, but the” rerun “lines do flatten out but they don’t fall off. The outer cylinders are basically unaffected (keep in mind that this sim is based on a race engine that pretty well sorted and it could be a lot worse). In reality the power will fall off a bit at the very top but as I said before I am still futzing with the burn model.
So there you have it, loose a little pipe, add a little pipe. Below power per cylinder compared between a 3-1 ( which is better than a LCB in this application) and a improved tuning version. Note that the green and orange lines gain quite a bit (30 vs 26 isch bhp ergo a gain of ~8 bhp on the inner cylinders).
This idea has had a fair few iterations before it landed here. The original idea I started out with was suggested by a guy who has a long and prolific background in killer 2 stroke engines btw. His suggestion initially was to lengthen the outer until they were just right then force down the frequency using basically a very long LCB
The beauty of simulations is not that they can tell you what to do, but that you can go trough a lot of ideas relatively fast without cutting and fabricating a zillion exhaust manifolds. You do have to have a fairly reasonable grasp of the matter at hand though, otherwise you will iterate yourself into oblivion.
If all is well you end up with a thing worth building and testing in real life on a dyno.There is still no guarantee that it works , but if the model is good enough, it should be quite close.
We will build it (Meaning I wont but it will be build).
If you really want to build one drop me a line.