yeah many new turbos, and a tremendous amount of different housing options. pretty cool stuff!

 

I agree there must be a tradeoff at some point, and the very same qualities that give a divided housing superior low rpm characteristics probably become less efficient than a single inlet as rpm increases past some point. This probably explains why the divided setup seems to be a darling of street setups and not all-out race cars (not that I recall seeing anyway).

Of course, it's very difficult to empiricalize this from behind the comfort of a keyboard (like you said, testing is needed to uncover the nitty gritty details). But one can almost certainly guarantee that as always, you give up something to gain something else.

 

Care to elaborate more on what you are saying here Geoff? If the collector is symetrical across both axis and runner length is equal across the board it does not matter which hole the runner goes to. If this was a divided set-up you'd be dead right, but all four runners are merging together negating any effect of cylinder pairing.

Edit: The Airwerks stuff looks to be promising. Some sort of concoction will be going on a bike soon, hopefully the results are as good as everyone is hoping for.

 

in a 4-1 why is pairing important?

and can you elaborate on how a long tube is more succeptible to reversion than a short tube...

 

I'm of the same opinion as Shearer and Trinydex, in that if every primary runner is absolutely equal, I don't see where pairing is relevant.

As for reversion, it occurs only at lower engine speeds when a large exhaust system and long overlap cam set is used, and is essentially a non-issue here. Why? Because even cams such as the HKS 272 set deliver only 214 deg duration at 1mm and give very decent low speed characteristics. For the sake of discussion however, a collector is in effect an anti-reversionary device. Once exhaust pulse exits into a larger area (collector) and expands, the slight negative pressure of reversion is insufficient to allow it to suddenly reform and be sucked back into the cylinder. Just have a look at going back through the exit of a collector and one can see the return path is virtually impossible in an aerodynamic sense. This being said, the shorter the distance between the exhaust port and the collector (essentially the lower the primary runner volume), the lesser the volume of intact exhaust pulse than can be sucked back into the cylinder. This is why tubular manifolds that employ anti-reversion cones position them immediately after the exhaust port exit.

 

hmm... i see what you're saying... but in the case of a long tube runner isn't it also true that there is usually a following pulse that can't/won't get sucked back after it enters the collector simply because of the fact that there is another pulse coming out where in a sufficiently short runner (say the stock cast mani if there was no twin scroll) a pulse can actually get pushed into another runner for lack of collector and because there's no followup pulse "cleaning out" the runner?

and i don't know if this would be more pronounced in pulse timed headers or equal length. but in equal length when you hit the collector you have the power of two pulses to push gas back up other tubes, whereas in pulse timed headers you always get a scavenging effect, but i suppose that's most pronounced after the turbo tho....

so another question, why aren't pulse timed headers more popular and powerful especially for turbo applications? or are they and all this el stuff is just hype.

i would say equal length is good in the sense that it doens't vary back pressure across the motor's different cylinders... that is one good thing.

 

Excellent book on the subject ; )


http://www.amazon.com/exec/obidos/tg...glance&s=books

 

are ya going to let us know how the kit is?

 

Triny, you have it wrong. No two cylinders fire at the same time. It takes two rotations of the crank to complete the firing order, so a 4 cylinder has 2 combustion events per rev. Cylinders 1&4 and 2&3 fire 180 degrees apart from each other. So with a true equal length manifold only one pulse will reach the turbo at a time.

 

Well i didnt order a full kitt only mani+wg,the rest will be custom made so i guess you have to wait for someone who goes for a full Evo 8 kitt,if i had an Evo 8 (i have Evo 6TME) this kitt would be my first choice,hands down.

 

this is gonna be a long one!

sorry for the confusion. To be honest there really is no clear answer on how the pairing of cylinders affects flows in a 4-1 system. Some engineers agree with you, others strongly disagree. What i have noticed is from one particularly noteworthy header designer from the west coast stresses the importance of proper cyl pairing on 4-2-1 and 4-1 setups. He feels that the proper cyl pairing creates a stale boundary layer between the cylinders sort of like those thermal air barriers at supermarkets which keep the heat in (for lack of a better example right now)

There are 2 points i guess i should mention

1) the cylinder pairing is such that if you wanted to run a divided housing, all you/we would have to do is seperate cyls 1+4 and 2+3, not have to completely redesign the manifold or be stuck pairing the wrong cyls together

2) the best book i have made a number of points in one of the early chapters regarding turbocharging manifold design, so i figured ill just copy them verbatim with my comments in parenthesis

- it is difficult to obtain even steady state characteristics over the full potential operating range of the turbine under representative conditions

- partial admission turbines (basically anything with a wastegate) add the further complexity of unbalanced mass flows even under steady state conditions

- the turbocharger turbine never operates under steady flow conditions, even in the so-called constant pressure system (something like a generator that operates at a constant boost and rpm level) and the effects of unsteady flow should be considered (can be calculated by a quasi-steady technique in which the size of the turbocharger is defined by its steady state characteristics, and the measurements are taken at the inlet flange)

- the effect of out-of-phase pulsating flows in partial admission turbines is not fully understood (*** this is the one which applies most ***) However, good general practice is to pair cylinders according to the engine's respective firing order, keeping flows in-phase.


With regards to the long tube being more susceptible to reversion than a short tube, i strongly disagree. I really dont follow how you came to that conclusion? It really goes against traditional exhaust theory.

i strongly disagree that reversion only occurs at lower engine speeds. I do agree that a true merge collector is also a bit of an anti-reversion device.



last note, equal length means nothing. The defining point is equal pressure. Putting a measuring tape inside of a 180 degree bend and reading the length will only tell you the circumference, not the actual friction and pressure drop, which are what really matter.

 

I think you've got two different concepts confused. Pulse tuning is a class of manifold design, equal length is a design enhancement that may or may not be present in a pulse manifold.

Pulse tuning is basically what you call *any* manifold used in a turbo app whereby all the primary runners collect at the turbo housing inlet. Such a manifold may or may not have equal length runners. Equal length runners are desirable because they evenly space the exhaust pulses.

Once a pulse gets through the collector, it's not coming back. When reversion takes place, it is more prevalent at low engine speeds and involves only the pulse that just left the exhaust port, not the following one. This a problem with large primary runners and long overlap cams. Unless you find yourself with really crappy low speed performance due to long overlap cams, or go to a *smaller* hotside, you shouldn't have to be concerned much about reversion where your EVO is concerned.

 

The possibility for reversion to be significant at higher engine speeds is possible in suspect applications, being far more likely to occur in a heavily cammed N.A. engine with drag pipes as opposed to the typical EVO given the most popular cam sets and its turbo size.

I agree with you that the term 'equal length' can be misleading, especially in an application (e.g. EVO) where bends are necessary due to spatial constraints, as disparity between them can certainly affect pulse timing.

The 'thermal barrier' concept is a novel one. A 4-2-1 manifold will certainly benefit from cylinder pairing, but as for 4-1... ???


Actually, primary length doesn't really factor into the susceptibility to reversion at all in the real world, which is I had to conjure an exaggerated, purely hypothetical scenario in an attempt to illustrate the point.

 

I'm a turbo geek and this is the first I've heard of this (well Geoff brought it up first)-- more info on new turbos please !

 

I have to disagree with you Ted. If you look at pressure ratios between pressure in the intake manifold and pressure in the exhaust manifold under boost, you will definately see why reversion would be more prevelent at higher RPM, full load situation on a boosted car. The ratios are typically 1:2 at full load, for a good spooling street car. When I say 1:2 I mean that for 20psi of intake manifold pressure there is 40psi in the exhaust manifold. Whenever there is a period of overlap the exhaust gas is pushed back in the combustion chamber due to this unbalance of pressure, thus diluting the charge and hurting power. This is why you typically see a "turbo" cams to have low duration and high lift with a wide 112-114deg lobe separation angle. On race set ups, with large turbos, the turbine wheel can flow well enough that the pressure ratios can be around 1:1. This would allow the use of a much higher duration cam, similar to what an N/A car would use, and this would yield substantial gains over a typicall "turbo" cam