Anti reversion cones on turbo application?

[Wiz_kid53]

Has anyone done this? Given the inherently high exhaust pressures in a turbo application I would assume they would benefit greatly from anti reversion but I’ve never seen it done. Anyone have any experience or know why people don’t do it?

[gunt]

It a kinda is done , from exhaust flange always keeps a step , some use a stepped manifold and twin scroll to help scavenging , which all aid in exhaust flowing one direction , but the aim is the turbo sizing and efficiency , its actually more the expansion speeds of the gas rather than pressure although that is how we measure it

[Wiz_kid53]

The port to primary step I figured was a given but I haven’t seen an instance where an engine did not benefit from an anti reversion device at some point down stream. I understand how to size a turbo but my thought is that the high exhaust pressures make it easier for there to be backwards flow into the cylinder during overlap which is why people for so long have reduced overlap on turbo cams. So if you can do the same thing without changing the cam you should be able to broaden the torque curve pretty significantly.

Might be worth mentioning that this instance I am applying this question to is for a Harley, so 2 cylinders and I was thinking about doing an AR device on each cylinder and possibly at the collector as well

[ptuomov]

Has a benefit from “anti-reversion” devices documented in any well designed system? Turbo, NA, whatever?

[Wiz_kid53]

I believe the point you’re trying to make is “if you design an optimum system it won’t be necessary” however we all know that designing a perfect system that also fits packaging constraints is next to impossible. I agree that it’s a bandaid fix but it’s an effective one. And my thinking is that a turbo magnifies the downsides of any exh system in that regard so the AR devices should bring back some area under the curve especially at lower boost/rpm levels.

[dannobee]

Good question that I've often asked myself.

But if we look at state of the art turbo stuff like F1 turbo cars or Indy/Champ turbo cars, you won't find any. I'd suspect that, given their unlimited budgets and access to practically any material, if it worked, they'd be on their cars. I can only suspect why they don't run them. One (of many, I suppose) likely reason is that a properly sized modern turbo setup will actually have more boost than backpressure. The negative pressure gradient would likely obviate the need, and could in fact, pull more fuel and air out the exhaust during overlap. Not necessarily a bad thing for power, but the fuel mileage would suffer.

[ptuomov]

What kind of engine is this for? A cross plane V8 has completely different turbo exhaust problems and needs compared to say an in-line four.

[dannobee]

What kind of engine is this for? A cross plane V8 has completely different turbo exhaust problems and needs compared to say an in-line four.

He said earlier that it's a Harley-Davidson. 2 cylinder.

[ptuomov]

What kind of engine is this for? A cross plane V8 has completely different turbo exhaust problems and needs compared to say an in-line four.

He said earlier that it's a Harley-Davidson. 2 cylinder.

How many degrees are the exhaust pulses separated in that engine?

[dannobee]

Fires one at 315 degrees, then the other cylinder 405 degrees later.

[Wiz_kid53]

Good question that I've often asked myself.

But if we look at state of the art turbo stuff like F1 turbo cars or Indy/Champ turbo cars, you won't find any. I'd suspect that, given their unlimited budgets and access to practically any material, if it worked, they'd be on their cars. I can only suspect why they don't run them. One (of many, I suppose) likely reason is that a properly sized modern turbo setup will actually have more boost than backpressure. The negative pressure gradient would likely obviate the need, and could in fact, pull more fuel and air out the exhaust during overlap. Not necessarily a bad thing for power, but the fuel mileage would suffer.

So my thought process with your pressure gradient point is to run a 2” primary with ar cones, this should keep exhaust bp down without allowing tons of reversion at low rpm/boost and keep it free flowing on the big end. Unfortunately I don’t have easy access to a dyno to sit and watch exh bp for good data, I do the majority of my tuning on the street. That being said it will be important to get that data as soon as possible so that my intake to exhaust pressure ratio is not theoretical, I’ll have real numbers. Luckily I do instrumentation for a living so I can get the appropriate equipment to get the data I need. It’s just not something I can do at this exact point in time.

[ptuomov]

Fires one at 315 degrees, then the other cylinder 405 degrees later.

So there’s no need for any tricks to separate pulses. Isn’t this a case where turbo inlet should be the size of the exhaust port and the exhaust primary and the “secondary” to the leading to the turbo should also be the size of the exhaust port? The pipe length ideally such that the wave travels it four times for two round trips from the exhaust valve opening to the valve overlap period?

[Wiz_kid53]

Are you saying one primary twice the length of the other?

[BLSTIC]

I'm interpreting ptuomov as saying there's no combined gas flow, so the manifold can all be sized as a primary and you can set the length so that you get a favorable negative pressure return wave at the rpm you want.

Cylinder to cylinder interference should be negligible considering it's mainly the blowdown pulse on EVO that wreaks havoc on other cylinders.

The other thing to consider is that it's backpressure on overlap that is your main concern now that blowdown isn't a thing. Would this benefit from having as short (low volume) a manifold as possible regardless of pulse tuning lengths?

As for your original question... How big are your ports compared to what is ideal? You might find that anti reversion devices at the port flange are just correcting a design flaw. Also i can't find it now, but I recall someone saying the 883 (or whatever the more modern smaller one is) Harley head was the best for overall power production simply because the others were just too big.

[Wiz_kid53]

I’m not sure about the 883 heads, the sportster stuff doesn’t apply here because the engines are completely different. I have a twin cam. As for port size I’m not sure but I’d say no bigger than 1.8” for an educated guess