The 13B NA Power Thread.

[NoDOHC]

rotor housings are cheap.

I want to do a semi-pp, leave the primary ports alone (they close pretty early) and fill the secondary ports, then connect a throttle plate for the two primary runners (I was going to use part of a TII manifold). Each of the peri-ports gets an ITB that feeds the port directly. A progressive throttle linkage will allow the engine to only run on the primary ports until about 30% throttle, at which time the P-ports will start to open.

I think it will work very well. If you don't try it, I intend to. I have the utmost confidence that you can do the project justice.

[RotaryProphet]

I had an idea for a setup where I would connect the primary and secondary ports together, and run each set to two of the ports on a four barrel throttle body, then run a pair of peripheral ports to the other two ports.

With a pair of reasonably small peripheral ports that open late to help keep overlap down, and a staged throttle linkage, it'd probably be reasonably drivable.

[PercentSevenC]

It's been done before. I think it was Jaytech that made a square-bore 4-barrel semi-PP intake manifold.

[NoDOHC]

The trouble with the late intake close P-port is that you might as well go with a side port. All the advantage of the P-port is lost if the intake port closes after pushing 40% of the engine displacement back out the port.

I think that the progressive throttle is the answer, keep the P-port throttle plates as close to the housing as you can get them and run the primary ports normally. The primary ports will not be open after the P-port has closed, so there is no loss of charge air. If you leave the secondary ports functional, your intake charge that just came in the P-ports will leak out the secondary side ports as the compression stroke begins, negating a large part of the P-port benefit (which would be a full compression stroke allowing better torque output).

Overlap is useless under vacuum, but does not hurt that much at WOT. In fact, with a well tuned exhaust and intake, the engine will actually flow unused air from intake to exhaust during the overlap, slightly cooling the combustion chamber and increasing net thermodynamic work from the next air/fuel charge. This will increase Volumetric efficiency and torque output.

If the primary ports (with no overlap) are used for idling and cruising, and the peripheral ports kick in only when the gas is floored, the car should have good driveability. I was going to use 6-460cc/min injectors, 2 in each p-port and 1 per primary port (in the center iron). This should give a decent transition, especially if the engine was tuned on TPS.

[Whizbang]

i am going to try straight Peripheral with the side ports eliminated.

but regarding the semi-pp, i wonder how an rx8 engine would work out with the whole side exhaust thing....hmm

[Whizbang]

well i know that the ITBs i have will flow 400 cfm per runner. What i need to figure out is what diameter runner would flow approx. 400cfm so i can size the port but i cant think of a way without knowing the velocity of the air moved.

[NoDOHC]

400 cfm is all?

The stock throttle bodies flow more than that (I saw a write-up with them tested at 950 cfm when ported).

Originally Posted by Gravity Fed

i cant think of a way without knowing the velocity of the air moved.

Velocity is not the point here, but i will explain how to calculate it later. The point is air flow. Each rotor requires 0.65L of air per revolution, if you crunch some numbers, you can see what the engine will require.

I would put a peripheral ported rotary at 110% VE Average (based on port timing and overlap). This means that each revolution of the engine will require 0.71L of air.
If you figure your rev limiter at 9500 rpm, that gives you 9500 rev/m * 0.71 = 6745 L/min which gives 238 cfm. According to this math, your ITBs are plenty large.

Unfortunately, this math is not exactly accurate, as it assumes sufficient plenum volume to allow constant flow through the throttle plates. This is not so, as you will be mounting your ITBs as close to your P-ports as possible.

The air flow into an engine can be approximated as offset-sinusoidal (we will make this assumption for simplicity).

Therefore, the minimum air flow is 0 and the maximum air flow is 2X the average air flow. This means that the port will require approximately 2 * 238 or about 480 cfm at peak flow. Basically, you will have some pressure drop across your throttle plate during the intake stroke, but not a lot.

The runner should be sized with a velocity stack for the inlet, gradually reaching the same diameter as the throttle body, then with a constant cross-sectional area all the way from the throttle plate to the rotor housing. A slight decrease in section height while passing through the housing should give good fuel mixing characteristics.

To compute the velocity is easy, however it has no real bearing on anything unless you care about Helmholtz tuning. Never the less, you simply take the peak air flow (480 cfm) divide by the cross-sectional area of the runner in ft2 and divide by 60 to give ft/sec. this can then be converted to mach easily by guessing Mach 1 at about 1070 ft/s. You want 0.3 Mach (about 300ft/s) peak intake velocity for best resonance tuning effectiveness.

Tuning for 9,000 rpm, throttle body/runner diameter should be about 2.144" (about 2 - 1/8" or 54mm) Runner length should be around 9 inches from port beginning (cross-sectional area stabilizes at bottom of velocity stack) to port end (the combustion chamber)

These are all rules of thumb, but they are good ones. I think that you will find that your intuition is a good as a calculator on this.

[Whizbang]

from efihardware:

Each throat flows 399.6 CFM @ 20.4 inches of Vacuum.
On the 55/53/50 tapered bore model with a 75mm long Pro-Series Ram tube fitted.

without running both setups on the same equipment with some standards enforced, it might be difficult to compare them.

im assuming the velocity stack going into the throttle body will be sufficient, but if the throat is 50mm (ill have to measure it one day soon) i would not want the runner to be a larger diameter than the throttle body for a variety of reasons as far as i can reason, plus its only going to flow as well as the smallest part of the track.

Then again i will be also limited by what is available for piping. For 50mm, a 2" od piping with a 1.902 ID would be the best to try and maintain consistency and easy to purchase from speedymetals.com. Ill just have to verify if its not too thinned wall for welding or bending. If that is the case, then ill step up the OD.

[Whizbang]

also interesting note regarding mach numbers, the mach number increases with altitude, the difference might not be worthwhile given the lack on fine adjustment on other areas of the design, but i figured id throw that random fact out there.

[NoDOHC]

The speed of sound decreases with increasing altitude, so the mach number of a given velocity would be higher, yes. I forgot that you were tuning for high altitude. You will be moving less air at higher altitudes, so I don't know if that will offset the compressibility effects or not.

You are right, the throttle body should be the same size as the runner. Any change in cross-sectional area will not only adversely effect your Helmholtz resonance tuning, but will also increase pressure loss in the runner (hurt flow).

You are going to keep the two ports on the center iron, right? They will open later and close sooner than the P-port, which means that they will have no adverse effect on the P-port power, but they will allow the engine to idle and run successfully at partial throttle (all the p-ports I have seen have serious issues at partial throttle operation, they liked WOT only).

I really like this idea, it is what I wanted to do for a long time.

Keep up the good work!

[Whizbang]

it might be able to concoct a manifold with the center ports intake (and streetported) since i planned on making the whole thing away. Might a little time consuming, but time is free. Just lots of measuring and tweaking before a final weld. But would there have to be some form of additional throttle to limit the peripheral ports interference at partial throttle.

The first engine will probably but a standard PP, but for future use, its certainly interesting.

[NoDOHC]

Actually, it would take less time to leave the center iron side ports open and just cut the outside runners of a turboII manifold, leaving the primary runners only and a single throttle plate at the end. You would have to leave enough of the flange to catch the two studs in the housings and that single bolt in the center plate, but you should have room for that.

It takes a lot of time to fill ports in (although I would DEFINITELY fill the secondary ports, as leaving them will hurt your power a lot). The primary ports will not effect the p-port at all, as the open later and close sooner.

If you set the throttle up so that 50% on the pedal is WOT on the primaries and 0% on the secondaries (peripheral ports) and the 100% is WOT everywhere, I think you will have something. (I actually wanted the primaries to continue past open until they had closed going the other direction - 170 degrees of throttle plate rotation - but either way will work.)

The slider is a very good means of doing this (has been used on transmission kickdowns, mechanical secondaries for carburetors, etc.)

As I said before, this is my idea of a 6PI system that would actually make more power (although it is technically a 4PI).

EDIT: Oh yeah, don't worry about porting the primaries, just clean the manifold up a little. This is not where your engine will be getting the air to make big power, just the air to idle and cruise, which is better if turbulent (smaller port is better).

[Whizbang]

hmm...its an interesting thought. Im still set on building the first engine as a regular P Port, but the next engine i have to always out do myself. i could probably use two different diameter throttle rotors. have one for the throttle cable to connect to, then attach a second rotor of another size (pending what is being driven) and have that rotate the other throttle cable.

[RX200013B]

If you set the throttle up so that 50% on the pedal is WOT on the primaries and 0% on the secondaries (peripheral ports) and the 100% is WOT everywhere, I think you will have something.

this is what i am tring to build the winter, just a little different
t2 center SP, 6pi ends w/ bridge on the lower ports, itb's
with the 5,6 ports only open at WOT

should bring out some good torque!? i hope

[Whizbang]

i might actually just be opting for a good four port streetport for the first engine in the rally car. I really doubt i will be able to put much more to the ground. Really depends on what mood im in when i get to building the engine. Making the Peripheral just seems like a better way to go. Id rather have more power on hand than less.