arghx

iTrader: (3)

Here's an interesting little gem that I found in the SAE paper from 20 years ago on the series 5 engines. Mazda did back-to-back testing using the same basic turbo and OEM electronic boost controller. The only difference was the turbine housing and manifold--single scroll manifold and housing (like the s3 12A turbo) vs fully divided (like the s5 13BT). I presume the sort of "in between" system of the s4 was not part of the test.





I suppose this just confirms commonly held belief, but it's interesting to see actual figures on a rotary engine from guys with lab coats.

haydenw

Thats a good find !!

StavFC

fantastic.
that ends the 'so it makes a difference, but HOW much difference' question then.

would love to see twin scroll vs one of those SP quickspool valves, but thats a lot of effort for pure testings sake.

arghx

iTrader: (3)

Stav,

Mazda had a factory quickspool valve on the series 4. They went to the s5 fully divided twin scroll for a reason. I think it's an inferior design over true twin scroll, although it probably offers advantages over regular single scroll.

Here are Mazda's single turbo manifolds through the years:

Series 3 (12A Turbo):



Series 4 13BT with factory quickspool valve (twin scroll actuator):



Series 5 13BT with independent twin scroll:

arghx

iTrader: (3)

Here are Mazda's factory turbine housing designs:

HT18S (12A turbo):



It looks a lot like a standard T25 inlet. Now here are the 13B HT18S-2S turbine housings. Series 4 is on the right, and is intended to be used with the factory quickspool valve. Series 5 is on the left and is intended to be used with the factory divided manifold.

StavFC

Yeah indeed, but I'd still like to see back to back tests considering the way Supra and big diesel owners bang on about how good they are despite having access to twin scroll setups

Trots*88TII-AE*

iTrader: (11)

The rotary engine has characteristics that those others don't though, the twin rotor with independent twin scroll design. It has to do with pressure waves and other black magic I don't understand

StavFC

I also wonder, as the testing was clearly done on small turbos, how things are affected as things go bigger.
Wonder if the effect is more, less, or similar. Hmm.

arghx

iTrader: (3)

One of the big advantages of a twin scroll design is that if you design the manifold right, you can take advantage of the firing order of the engine in order to reduce back pressure and maintain exhaust velocity. I bet it makes the biggest difference on a rotary and inline 4 cylinder piston engines, especially from the literature I've seen.

So on a peripheral exhaust rotary you have two exhaust pulses that alternate: front and rear. With a single scroll manifold you will have residual exhaust in the manifold increasing backpressure. The front rotor fires and the exhaust hits the turbine blades. Then the rear rotor fires and has to push through left over exhaust from the front rotor's exhaust pulse. This is because there is no separation between the exhaust from each rotor.

With an independent twin scroll setup, the exhaust pulses from each rotor never directly interfere with each other. The front rotor fires, and as the rear rotor fires the front exhaust manifold runner has enough time to dissipate the residual exhaust. I didn't understand this until I read an article (not by Mazda) about the same type of thing in inline 4 cylinder engines:





and you can see in this diagram the reduction in backpressure:



and the accompanying increase in spool on this 4 cylinder motor:



On a 4 cylinder you may have a firing order of 1-3-4-2. So if you put 1 and 4 to one scroll then 2 and 3 to another, you are always alternating which scroll receives exhaust. You pulse the left scroll, then the right scroll, then the left scroll, then the right scroll. The exhaust never interferes. And that's the problem with these "quick spool" valves and the factory twin scroll valve Mazda was using. You can't take advantage of this backpressure reduction (if the engine configuration would even allow it). If anything, with the quick spool valve closed you are increasing backpressure. It may spool the turbo faster compared to a normal single scroll manifold but the engine is still working against the residual exhaust.

Now it may not work exactly the same on a rotary, especially because there are no valves, but it has to be a similar type of thing going on (high exhaust velocity and backpressure reduction). I would think that on a single turbo V6 or V8 with a twin scroll setup you have more cylinders and so much more plumbing that there is less advantage to a twin scroll design. It's got to be one of the reasons why twin turbos are usually used on V configurations.

Trots*88TII-AE*

iTrader: (11)

You always have the coolest articles.

Black Magic --- 0

Proven Theory --- 1

Narfle

iTrader: (16)

Neat.

BurntOrangeT2

iTrader: (30)

good read

ThirdGenFTW

+100 internets to you sir!

BDC

Great post, man.

B

arghx

iTrader: (3)

Here are pics of the Evo 9 exhaust manifold and turbo:





and here's a pic of Subaru's factory stainless equal length divided turbo manifold on the 2010+ EJ255 boxer engines



The manifold uses a double-layer structure to retain heat, like the Ford 3.5L Evoboost twin turbo engine and the Rx-8 Renesis engine. To make the short equal-length design fit, it has a low-mount location with a separate oil scavenge system:

TRWeiss1

iTrader: (20)

Very interesting indeed! Thanks for posting this.

Trots*88TII-AE*

iTrader: (11)

WHAT?!?!? They didn't use a Quick-Spool Valve? What a bunch of idiots at Subaru...

scrubolio

iTrader: (8)

very interesting!