Quote:
Originally Posted by Nigel_S
I still think this is wrong. I've just had a look at the calculations I did when I changed my header and the result is 31.49 inches which agrees with your figures, however I'm using 16" primaries so why does practice not coincide with the calculations? The headers that are on sale tend to be either 16" primaries & 16" secondaries or 24" & 8" both designs giving 32" total, I've not come across 32" primaries for sale.
|
I see exactly where you're coming from, but consider this: if you were looking at 4-1 header instead of a 4-2-1 header, the primaries
would be ~32" long, and hit the merge collector at that point.
Of course, in that case you're also getting cross-reflection from twice as many cylinders at the 32" mark with a 4-1 design... I'm not sure how much of an effect that is
. The idea of the looong primaries here is definitely unknown territory design-wise (i.e. not covered by any of the convenient formulae), I do know this, but as someone's signature around here says, "If we knew what we were doing, it wouldn't be called
research now, would it?"
Again, though, take a look at the modern "hybrid" style that have been apparently been making waves in certain segments of the performance world recently for their wider torque band - they have primaries of similar length to a 4-1, and fairly long secondaries (more than just a "hybrid collector", where the 3 Y merges are stacked as tight as possible). The extra length means they're generally cat-less (i.e. you remove the cat to fit the extra length of the collector).
Other features of note with the 'hybrids':
* The primary:secondary ratio appears to be about 2:1, though it seems to vary a bit.
* Most seem to have an anti-reversion step roughly half way down the primary... or alternately near the 15" mark. Hard to tell - some seem to have multiple steps in the primary.
* The first set of collectors seem to be paired 1, 3 and 2, 4 or 1, 2 and 3, 4 - what the heck?
* All the designs are proprietary and design data / formulae / reasoning behind it is hard to come by (or at least I haven't found any on the net).
Quote:
Originally Posted by Nigel_S
I think the reason is that almost all of the formulae documented on the web are for V8 engines while we have straight 4 engines which use a different firing order. The V8 calculations time the pulses for a cylinder which fires 360? degrees later but we need to time them for the other cylinder in the pair which fires 180 degrees later thus we want primaries half the length. Or something like that, I'm not too sure about the V8 firing orders and angles.
|
Actually, in a flat-crank V8, each bank is essentially a 4-cylinder engine in itself, with exactly the same exhaust pulse to RPM ratio (though potentially different firing order). Or another way of looking at it, a flat-crank V8 is a pair of 4-bangers welded together at right angles, sharing the same crankshaft. Tri-y headers are designed for each bank using exactly the same parameters as for a 4-cylinder, then potentially joined further back by an H pipe or X pipe to balance the banks.
In a cross-plane crank, the firing order is all messed up, and cylinders firing 360 degrees (crank) apart are on opposite banks, so if you want 'equal length' tri-y headers, they're generally routed to the back of the engine before the first merge, producing an appearance referred to as "bundle of snakes". Burns Stainless also has an interesting article with another possible solution which allows shorter tubes, given that the back of the engine can be a long way away - routing some of the pipes under the engine.
Burns Stainless LLC - Bundles of Snakes and 180 Degree Headers
