Food for thought; does having a single WG on an elephant trunk negate twin scroll?

[Tippy]

Been pondering this for awhile without any hard data. Does having an elephant trunk WG circuit using a single WG negate the benefits of a twin-scroll setup (using a divided header of course)?

I'm leaning towards it doesn't 100% completely negate the advantages of a twin-scroll setup nor 100% allow the system to work properly.

My thought has been that the exhaust pulses would have to make a ~160-170 degree turn in the WG circuit where the 2 pipes merge to cause energy cancellation.

That seems like it wouldn't really happen.

Reason why IMO?

The elephant trunk WG circuit has proved to be a restriction not allowing the exhaust to be dumped as it is as seen by many here with boost creep. This tells me that once the flow gets to heading in the direction it is going, it doesn't like to change direction.

Thoughts?

[jsveb]

I am not an exhaust/header expert, but I do understand airflow pretty well.

It does negate the effect of a true twin volute system ( I believe it is called )

You need to separate the two banks if you want the benefit of tuned headers. The individual exhaust port pulsation will travel across to the other bank if it is allowed. This negates the Venturi effect that otherwise gained by a tuned port system.

Not a 100% correct, but the answer to your questions is yes.

[Tippy]

I'm struggling with "tuned" and "venturi". Are divided headers/twin scroll turbines "tuned"? I don't think they are? When think of venturis, I think of a convergent/divergent nozzle. Where do these play into twin scroll systems? Trying to understand since you understand airflow well.

The more I think about it, it seems it would negate advantages but wanted data if someone had it.

Rarly8?

[jsveb]

Tippy, my knowledge is from aviation, though the theory is the same, some time when it comes to cross disciplinary application there are other factors to consider than what I am used to.

Imagine a closed system, meaning no energy etc can influence what we are about to discuss. This means everything happens on an exchange basis, so to speak.

If we are standing in a big pipe with atmospheric pressure. We have got a static pressure of about 1 bar ( 1.01325 to be precise or 29.92") in standard atmospheric conditions. This means, if you put your hand in front of you the pressure you feel is the same above and below, left and right and so forth.
Now, if you move that hand from left to right, the leading side will feel a slight increase in pressure. It is more pronounced when doing the same thing in our car at speed.
Remember I said there could be no added "energy" this means the pressure we felt when holding the hand still 1(bar) has got to be lower now, since we can feel an increase by moving it. Guess what, it is.
The stand still pressure was called static pressure. When we move our hand, we call that pressure, dynamic pressure. The sum of the two, we call total pressure
Ptot=Pdyn+Pstat
Once again, recalling this is a "zero sum game". This means when holding our hand still Pstat is 1bar, hence Pdyn must be zero.

I know this is long winded. Once you understand these basics you can begin to understand aerodynamics and even engine performance as affected by environmental conditions.

Back to the Venturi thing:
We can conclude whenever we create some dynamic pressure the static pressure drops. Guess how airlines fly; they create a pressure differential between the upper end the lower side of the aero planes wing. This pressure differential is called lift and sucks the airplane up into the air.

I am cutting it a bit short here, but I hope you follow. The Venturi effect is basically the suction that takes place when an air masse moves; creating a dynamic pressure along with a corresponding lower static pressure, suction.

We can use this to our benefit in an exhaust system. We can imagine it actually takes time for an exhaust pulse to travel through the whole exhaust system.
Let us say #1 fires and creates a pulse from the left bank. If this is not isolated this pulse can travel and create a higher pressure for the #4 cyl to counter in its way through the exhaust system.
Also the #1 cyl plus might aid the #2 cyl by sucking out the exhaust plus from that port.

This is why the header design is very important, or one of the reasons.

I hope this is not completely nonsense

End of 1st lesson on airflow fundamentals.

[Unobtainium]

Jsveb,
Brilliant description.

[RarlyL8]

This is a tuned system. Each primary is the same length, each secondary is the same length. The plenum is split. In theory each pulse will hit the collector and then tubine at the same individual time/rate as the firing order. There is disturbance at the waste gate braches but it is not detectable in a practical sense.





The twin waste gate setup is a twin volute system which can fully utilize the twin scroll turbos design.

[Tippy]

jsveb - I never thought of a header as a Venturi due to the shape. I understand what you're explaining since a Venturi creates low pressure to increase scavenge and headers create scavenge (to a degree).

Good analogy on the dynamic principles. I'm an A&P BTW.

Rarly8 - the WG branch is where all the theory I was debating in my head is at.

Can the 160-170 degree turn at the merge where the 2 branches meet have that much influence on contaminating the pulse energy?

[proffighter]

The Venturi effect is already given by the cylinder head/exhaust duct into headers which helps transverse flow scavenging (correct english term?). Not sure about another Venturi at headers.

BTW To understand that clear (my lack of english, sorry):

You mean a Venturi effect where the single tubes of the cylinders lead into one (or two at RarlyL8's headers) So the gas stream of lets say cylinder 1 causes scavenge in tube 2 and 3, right?

[Tippy]

Looking at the header Brian posted (looking at WG branch), that one CLEARLY negates the twin scroll advantages as it looks like it would easily cause pulse contamination.

On my header though, the WG branches start out 90 degrees from secondaries, the curve another 90, then travel about 6 inches or so then merge to a single tube.

Now, where they merge, would require the pulse energy to make about a 160-170 degree turn.

Beings this collision is so far from exhaust valve, would it really contaminate the energy 100%?

Trying to figure out.

Sorry if being dense......

[xbmwguy]

hey tippy my mode headers are a twin scroll set up using one wastegate. i asked bob when he was building them last year if at my power goals a twin wastegate system would be better he assured me no. i am able to keep boost exactly at 11.5 lbs (.8 bar) and at 17.5 lbs (1.2 bar) with zero boost creep.

[Tippy]

Wow, Bob uses single. Ok, I know people consider Mode to be great headers. Interesting. Wonder why he uses single? I mean, if I were designing from scratch, I'd no doubt put in twin WG's. I was just trying to figure with my old school headers that I converted to divided, is the single WG killing the effect?

I'm wasn't as concerned with boost creep (I have that due to the elephant trunk WG circuit and the basis for my (flawed?) theory) as canceling/negating/nulling the advantages of the divided scrolls.

[Tippy]

BTW, do you have a pic of your headers xbmwguy?

[xbmwguy]

here is a couple

[RarlyL8]

If you design the waste gate circuit to flow well there is no noticeable performance difference on most engines, even with twin scroll turbos. The big HP engines can require exact control from each bank which the twins will insure. The plumbing is (obviously) a bit less complicated as well.

[Dai]

Mine is also made by Mode, and Bob gave the same answer here. The header and turbo is twin scroll design but with one wastegate. Here is mine.

[Tippy]

First off, thank you xbmwguy and Dai for posting pics!

I'm really leaning towards the Mode's negating the effects of twin-scroll. I mean, I can clearly see how cross contamination (exhaust pulse energy collision) can happen easily in the WG circuit. Not trying to offend xbmwguy and Dai as your headers are stunning. I would love your headers! There might be some improvement doing as Rarly8 does and use twin WG's?

Anyways, the theory of twin-scroll is to have good scavenge effect at low RPM to decrease spool time and backpressure reduction at high-RPM to increase HP.

Many outside of the Porsche world claim this is very advantageous. Maybe we don't see the enhanced results of twin-scroll due to long tube headers pre-turbo (loss of exhaust energy due to heat loss) and moderately lower boost?

Chris Carroll, are you there?

[jsveb]

I believe the reason people might not see any significant effects using dual WG, is because there is some other and more significant restriction in the equation - somewhere.

[proffighter]

If I understand correct, the advantage of twin WG's is better response etc... which is while WG is still not open. If so, the WG pipe does not see any flow, which will not affect the twin scroll much then, due it's angle (On my RarlyL8's) even less.

[Raceboy]

Using a single wg on twins scroll system requires to assemble the wg pipes correctly. Most photos above that picture exhaust headers are made wrong having wastegate pipes connecting right after exiting from bank. You need to separate pulses right to the wg valve, then you can easily use single wastegate and still get the benefits of twin scroll. And since the split-pulse is important on boost treshold and transient response (less so on reversing effect at full boost), it all that matters.
Like this (made it for my 924 turbo "Gulf" race-car, mild steel, sandblasted and ceramic coated with Ceracote C105 Titanium):



And just for reference, BMW used also twin-scroll with single wastegate in their F1 turbo motor. Guess it works then:

[RarlyL8]

930's use much lower boost levels than water cooled racecars which is why the waste gate circuit design merge is not so critical. I tried a split WG plenum on a 1bar engine and noticed no difference. Part of the control of dual waste gates involves the extremely short tubes and short or non existant tailpipes. It is theoretically very precise, but again only noticeable on high HP engines running non-standard boost levels. The design of the waste gate itself is even more important in these applications as cracking pressure/response and crispness of control allow much quicker spool out of a corner.