twin scroll turbine & 5-cylinder engine

[Artsi]

Hello

My wife's son is building an Audi 5-cylinder 10-valve (petrol) turbo engine, 2144cc. The cylinder head will most likely retain stock size valves, but will receive a mild porting job by a professional (in order not to kill off-boost performance, porting job will not be done to support astronomical power levels). Targeted power level for the first evolution version is about 400-450hp. This first version will most likely have a low budget Garrett GT32 single entry turbocharger.

Second evolution version most likely shall have a ball bearing turbo by Garrett (perhaps GT3071 or 3076).
I have been playing with an idea of making use of the twin scroll turbine option. However 5-cylinder engine presents a challenge, to which I am unable to come up with answers.
I don't know whether it's feasible to make odd cylinder engine to sing well with twin scroll entry turbine.

Two crank rotations makes one complete four stroke trip. Divide that with 5-cylinders, I gather we get a spark every 144 crank degrees.
Firing order is 1-2-4-5-3.
I guess that would make, say, cylinders 1-4-3 go together, and remaining 2-5 go together.

I'd like to see this engine to have a broad operational rev range, which is what twin scroll turbine is supposed to support.

I do welcome constructive commentary over this issue, and also comments on what to take into account in making a tubular exhaust manifold (long tubes create large volume versus short tubes, is it good or bad, pros and cons, equal lenght runners, diameters of the runners, etc..)

-Arto L.

 

[BrianPetersen]

In theory ... I don't think a twin-scroll is the best choice for this application. No matter what cylinders you split up, you're always going to have two in sequence in the same group. The blowdown pulse from the second of that two-in-a-row will interfere with the end of the exhaust stroke and potentially the overlap period of the first of the pair and it's only going to affect that one cylinder (basically increasing the amount of exhaust residual in that one cylinder only). Bear in mind that with your proposed cylinder grouping 3 and 1 fire in sequence in that order. 1 might be at the start of the firing order but 3 is at the end just before the next start

If the cam timing is really mild so that there's little overlap, perhaps it won't matter too much. But that kinda defeats the purpose of using a twin-scroll.

If you run all 5 cylinders together in one manifold and split that into operating both scrolls then at least all 5 cylinders will be treated (almost) equally but then there's no point using a twin-scroll.

I can't really comment on the performance of a log versus long-tube exhaust manifold other than noting that OEM turbo applications are almost always log manifolds or very close to it, and aftermarket can go either way. OEM has some considerations that the aftermarket might not care about, like keeping catalysts happy and lighting them up after start-up as quickly as possible. Space is almost always an OEM consideration, too, and tubular manifolds take up a lot of it. If controlling heat in the engine compartment is an issue then that's in favor of a log manifold, too, because there's less area of hot surfaces.

 

[GregLocock]

There's very little (but not necessarily nothing) to be had from long exhaust runners in a turbo application. The slight gain in efficiency from preserving the pressure peaks is more than offset typically by the longer pipe run and hence thermal losses. Packaging wise you are heating up the entire engine bay with the long runners, never a good thing.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[EdStainless]

Exhaust headers are actually very forgiving.
Look up Engine masters on the R&T youtube channel.
They did some testing on a dyno changing headers and then crushing tubes on headers.
Almost nothing mattered at all. It would make a bigger issue at higher power density, but not in low to mid-range.
Restrictions on the intake side matter more.
In almost every case with a turbo you want things as close and compact as possible.

I don't see twin scroll helping.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Lou Scannon]

The conventional wisdom on pulse turbocharging, via joining of exhaust ports into discrete turbine inlet channels, is that it can work well when done properly with 240 degree spaced pulses; 360 degree space pulses can also be made to work, but suffer to some extent from the gaps between pulses. Spacings of 180 degrees or less, or any grossly uneven spacings are really not worth the bother, as there is not enough pulse to be captured (although the benefit of small volume and short distance always exists).
Any turbine inlet conditions that consist of significantly uneven flow (the definition of a pulse system) will result in less than optimum turbine efficiency, since the flow angle onto the blades is continually oscillating well above and well below the design point. A good pulse system overcomes this disadvantage due to the capture of the kinetic energy of the blowdown pulses that is otherwise wasted.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz