How to fight exhaust reversion when the primaries are way too short

[Calypso]

Apparently resonator at manifold is not exactly new idea

1F9DD8FB-C630-4B30-BD4D-B5110695E624.jpeg

[ptuomov]

That recent photograph of a fully functional high performance engine may be indicative of how this superior design took over the world by storm...

[Calypso]

That recent photograph of a fully functional high performance engine may be indicative of how this superior design took over the world by storm...

Maybe they had room for headers instead at the time?

[Calypso]

One in Red Bull F1. Maybe they didn’t have space for muffler.

72639A04-7C4F-4A18-8F4A-79FE0A1C3E8A.jpeg

[ptuomov]

One in Red Bull F1. Maybe they didn’t have space for muffler.

72639A04-7C4F-4A18-8F4A-79FE0A1C3E8A.jpeg

I don't think anyone really knows what's the purpose of that pipe. They use exhaust a lot to create aerodynamic cheats, is that an open pipe or closed? If it's closed, are they trying to take out pulses from the exhaust air flow that they use to guide the air flow to the rear wing?

[modok]

-My mind's been locked on to understand Bernoulli's Principal more and more lately,

-It's not about 'hairsplitting', but it's about 'speaking the correct language'.

-Exhaust heat energy transfer is not applicable to this topic?


-From another perspective, consider the definition of 'sophistry':
sophistry is evidenced by using 'superficial' terminology.

You said many things
Unified terminology is difficult. I've commented on that before, however I don't see things so black and white. There are different "models" of exhaust design. There is no unified theory that is sufficiently functional at this time. If the model functions, then it's useful. It may not truthful to what's actually happening. I can give MANY examples, but fact is if you want to understand what they are thinking, then learn their model. If third harmonic means fourth harmonic, and over scavenging means reversion, then just have to learn that, or teach them, and you are not immune to this.
"Bernoulli's observation" is useful, but it is not a law of nature nor necessary building block to understand what is happening. it is actually just a rule of thumb observation which is true most of the time.......making it a rule and building upon it may actually hinder further progress.
So.....it steams of irony.
I'm not making fun I'm just having a good time

[Cubic_Cleveland]

Given that I'm running on pump gas and with low compression, I really want to empty that combustion chamber from hot exhaust gasses.

I tend to agree on the overlap. If the gas flow goes in the wrong direction, shut the valve! Filling the intake runner with hot exhaust gas isn't going to help make things better. It's going to make things worse.

In terms of the shorter exhaust duration, in my opinion, more complicated. Definitely helps at low rpms to reduce the 180-degree blowdown interference. Now so sure at high rpms when the problem is 90-degree exhaust blowdown interference. I guess it depends on whether the exhaust manifold design strategy is to separate the 90-degree pulses (like in the LS7 manifold) or to lay them on top of each other (as on one side of the Coyote 5.0L manifold).

Why does the LS7 manifold work so well? is it because the four-in-a-row collector aims all the short primaries in the same direction? That's the sort of effect that the 1D software (Vannik's) I use isn't really accurate with, you'd need some sort of 3D software to model that. Why isn't the LS7 manifold killed in the top end by the 90-degree exhaust blowdown interference?

This is another "too short" cross-plane V8 exhaust manifold that works really well on pump gas, yet simple 1D simulation would say it would be really hurt by the 90-degree exhaust blowdown interference:



It has the same feature as the LS7 manifold, namely that the primaries are pointed in the same direction before the paths merge. Maybe that's a clue?

Is that off the newish Koenigsegg? Are they flat plane crank or bent crank? I remember thinking when I read an article that those bends seem tight and restrictive, and maybe a simpler manifold would work just as good. But then again, they do make massive power and their engineers seem to know a fair bit more than me

[ptuomov]

Is that off the newish Koenigsegg? Are they flat plane crank or bent crank? I remember thinking when I read an article that those bends seem tight and restrictive, and maybe a simpler manifold would work just as good. But then again, they do make massive power and their engineers seem to know a fair bit more than me

Yes, that's one of the Koenigsegg engines. It has a cross-plane crankshaft, which makes sense with the long stroke.

http://www.koenigsegg.com/build128-the- ... gg-engine/

Our crankshaft is a 90-degree design that has very small and light counterweights to suit Koenigsegg’s very light pistons and connecting rods. The lightness of the rotating assembly together with the small area intake plenum and refined software calibrations make for a very responsive engine.

We’ve been asked on several occasions why we don’t switch to a 180-degree crankshaft design, which theoretically would allow for more power due to even more optimal exhaust pulses. In many aspects this is a very simple thing to do as the connecting rods and pistons can stay the same. The crankshaft, camshaft and some of the parameters in the software would have to change due to a different firing order and some difference in exhaust gas re-circulation. But that is pretty much all that it would take.

We have experimented with 180 degree crankshafts over the years but have decided against using one for the time being. The reasons for this decision? Well, we’re not exactly short on power as things stand right now and the 90-degree design gives less vibration and smoother engine characteristics. This is very important in a car where the engine is bolted straight to the carbon monocoque, without any cushioning, as it is with the Agera. We also find the 90 degree V8 rumble in combination with the turbo whine and fast response make for an evocative, powerful and unique sound.

Somehow that exhaust manifold with unequal length short primaries works very well even though the pulses are all wrong. I'm guessing that it's something related to the primaries exiting parallel at the collector, like in the LS7 "four-in-a-row" collector.

[amcenthusiast]

I think discussion of engineering concepts that apply to 'reversion' should include:

Physics: -sound wave theory, such as constructive and destructive interference, how heat adds energy to increase velocity, how cubic square area affects velocity, resonance/resonant frequency
(what else?)

Thermodynamics: -internal vs kinetic vs potential energy of a gas
-elastic potential energy of a gas, heat transfer/heat dissipation, thermal efficiency -effect of ceramic coating
(what else?)

Physics: -Acoustics, reflection, cancellation, how sound travels around a corner
(what else?)

How do established principals such as Hooke's Law, Bernoulli's Principal for gas though a venturi, Heimoltz Resonator (what else/other laws/principals apply?)

For me the fun part is learning the engineering laws and principals that transcend time, whereas the mainstream auto literature may promote styles and trends only to stay in business.

Could a switching magnet be used to manipulate exhaust gas molecular content?

[ptuomov]

I think discussion of engineering concepts that apply to 'reversion' should include:
Physics: -sound wave theory, such as constructive and destructive interference, how heat adds energy to increase velocity, how cubic square area affects velocity, resonance/resonant frequency
(what else?)
Thermodynamics: -internal vs kinetic vs potential energy of a gas
-elastic potential energy of a gas, heat transfer/heat dissipation, thermal efficiency -effect of ceramic coating
(what else?)
Physics: -Acoustics, reflection, cancellation, how sound travels around a corner
(what else?)
How do established principals such as Hooke's Law, Bernoulli's Principal for gas though a venturi, Heimoltz Resonator (what else/other laws/principals apply?)
For me the fun part is learning the engineering laws and principals that transcend time, whereas the mainstream auto literature may promote styles and trends only to stay in business.

Almost all of these are concepts that are incorporated into a 1D simulation software. What would you like to know from those simulations?

Here's an example of how those physics concepts result in mass flow in the wrong direction for one of the cylinders in the passenger side cylinder bank. This is at 6000 rpm.

RightBank6000.jpg

During the overlap, the exhaust port doesn't ever really flow in the wrong direction for any of the cylinders. However, both the exhaust port and the combustion chamber have such a high pressure in cylinder #1 during its overlap that when the intake valve opens, burnt gas flows into the intake port. Then later this burn hot gas returns to the cylinder and causes cylinder #1 to knock.

What's dawning on me is that the shapes of the exhaust manifold are really important and that even the best 1D simulation is not going to be very accurate in explaining how the exhaust manifold shapes impact issues like that.

There's a paper that combines the 1D overall simulation with 3D exhaust manifold simulation:

Turbocharger Matching Method for Reducing Residual Concentration in a Turbocharged Gasoline Engine
2015-01-1278, Published 04/14/2015:

Abstract
In a turbocharged engine, preserving the maximum amount of exhaust pulse energy for turbine operation will result in improved low end torque and engine transient response. However, the exhaust flow entering the turbine is highly unsteady, and the presence of the turbine as a restriction in the exhaust flow results in a higher pressure at the cylinder exhaust ports and consequently poor scavenging. This leads to an increase in the amount of residual gas in the combustion chamber, compared to the naturally-aspirated equivalent, thereby increasing the tendency for engine knock. If the level of residual gas can be reduced and controlled, it should enable the engine to operate at a higher compression ratio, improving its thermal efficiency. This paper presents a method of turbocharger matching for reducing residual gas content in a turbocharged engine. The turbine is first scaled to a larger size as a preliminary step towards reducing back pressure and thus the residual gas concentration in-cylinder. However a larger turbine causes a torque deficit at low engine speeds. So in a following step, pulse separation is used. In optimal pulse separation, the gas exchange process in one cylinder is completely unimpeded by pressure pulses emanating from other cylinders, thereby preserving the exhaust pulse energy entering the turbine. A pulse-divided exhaust manifold enables this by isolating the manifold runners emanating from certain cylinder groups, even as far as the junction with the turbine housing.

This combination of appropriate turbine sizing and pulse-divided exhaust manifold design is applied to a Proton 1.6-litre CamPro CFE turbocharged gasoline engine model. The use of a pulse-divided exhaust manifold allows the turbine to be increased in size by 2.5 times (on a mass flow rate basis) while maintaining the same torque and power performance. As a consequence, lower back pressure and improved scavenging reduces the residual concentration by up to 43%, while the brake specific fuel consumption improves by approx. 1%, before any modification to the compression ratio is made.

What this paper says when translated into plain English is that (according to their 3D simulation) running a twin-scroll divided exhaust manifold _with a single scroll turbo_ gets you a lot of the benefits of a twin-scroll turbo. Somehow keeping the interfering pulses separate in the exhaust manifold until the runners all point in the same direction (into the turbine housing inlet, in this case) directs the pulses forward to the turbine and not back up the other runners.

I have no access (or expertise to use) software that can accomplish that. So I'm kind of flying blind extrapolating from other people's findings and from my observations of what is working for LS7 and Koenigsegg. That is, keep the interfering pulses separated until the runner all point in the same direction and then run a collector that doesn't "venturi merge" at all but instead just allows the cross-sectional area to increase and that prioritizes the flow from all interfering runners being parallel at when they join.

Could a switching magnet be used to manipulate exhaust gas molecular content?

No, I think I'll just use my money more productively to upgrade to billet muffler bearings:

[cjperformance]

Is that off the newish Koenigsegg? Are they flat plane crank or bent crank? I remember thinking when I read an article that those bends seem tight and restrictive, and maybe a simpler manifold would work just as good. But then again, they do make massive power and their engineers seem to know a fair bit more than me

Yes, that's one of the Koenigsegg engines. It has a cross-plane crankshaft, which makes sense with the long stroke.

http://www.koenigsegg.com/build128-the- ... gg-engine/

Our crankshaft is a 90-degree design that has very small and light counterweights to suit Koenigsegg’s very light pistons and connecting rods. The lightness of the rotating assembly together with the small area intake plenum and refined software calibrations make for a very responsive engine.

We’ve been asked on several occasions why we don’t switch to a 180-degree crankshaft design, which theoretically would allow for more power due to even more optimal exhaust pulses. In many aspects this is a very simple thing to do as the connecting rods and pistons can stay the same. The crankshaft, camshaft and some of the parameters in the software would have to change due to a different firing order and some difference in exhaust gas re-circulation. But that is pretty much all that it would take.

We have experimented with 180 degree crankshafts over the years but have decided against using one for the time being. The reasons for this decision? Well, we’re not exactly short on power as things stand right now and the 90-degree design gives less vibration and smoother engine characteristics. This is very important in a car where the engine is bolted straight to the carbon monocoque, without any cushioning, as it is with the Agera. We also find the 90 degree V8 rumble in combination with the turbo whine and fast response make for an evocative, powerful and unique sound.

Somehow that exhaust manifold with unequal length short primaries works very well even though the pulses are all wrong. I'm guessing that it's something related to the primaries exiting parallel at the collector, like in the LS7 "four-in-a-row" collector.

I took this from the Koenigsegg link you put up "
Quote
""We’re the only high-powered, turbocharged production engine in the world that has less back-pressure in the exhaust manifold than the boost pressure in the intake manifold. Let’s say, for example, that we have 1.5 bar of boost on the Agera RS in the intake manifold. For that engine, we will have around 1.3 bar of back-pressure in the exhaust manifold, which is unheard of. Normally, if you have 1.5 bar in the intake manifold then you’ll have something like 2.5 bar in the exhaust manifold before the turbo. With that high pressure, you would have all the associated hot EGR and you’d have to reduce boost to prevent detonation, which means you’d have to back off your timing.

That means power loss. Big power loss.""end quote

This pressure ratio/differential is the real key here. This is negating a lot of the 90* issue in that the higher intake pressure is allowing exhaust gas to be evacuated/pushed out of the cylinder on overlap by the higher intake pressure. Sure it is still not perfect but I think that is their gold star and coupled with a short but uni directional manifold/header is probably as good as its going to get unless twin turbos per bank were used, of course adding cost and reducing space.

[ptuomov]

I took this from the Koenigsegg link you put up:

We’re the only high-powered, turbocharged production engine in the world that has less back-pressure in the exhaust manifold than the boost pressure in the intake manifold. Let’s say, for example, that we have 1.5 bar of boost on the Agera RS in the intake manifold. For that engine, we will have around 1.3 bar of back-pressure in the exhaust manifold, which is unheard of. Normally, if you have 1.5 bar in the intake manifold then you’ll have something like 2.5 bar in the exhaust manifold before the turbo. With that high pressure, you would have all the associated hot EGR and you’d have to reduce boost to prevent detonation, which means you’d have to back off your timing. That means power loss. Big power loss.

This pressure ratio/differential is the real key here. This is negating a lot of the 90* issue in that the higher intake pressure is allowing exhaust gas to be evacuated/pushed out of the cylinder on overlap by the higher intake pressure. Sure it is still not perfect but I think that is their gold star and coupled with a short but uni directional manifold/header is probably as good as its going to get unless twin turbos per bank were used, of course adding cost and reducing space.

Of course the pressure differential (or ratio) between intake and exhaust is the issue. However, I am convinced that the cycle average pressure ratio is not the issue. We'll soon be at about exactly absolute pressure ratio of one at 6000rpm between intake and exhaust, if you use the cycle average numbers. That used to be considered good.

However, if you take the 1D simulation of intake and exhaust pressures (forgive me, sensors are not going to be hooked up to this engine already in the car), the exhaust pressure at the beginning of the overlap is 2.5x in cylinder 1 compared to cylinder #2! The exhaust to intake absolute pressure ratio for the worst cylinder #1 is 2.5x and for the best cylinder #2 it is 1x. There's no plausible reduction of the cycle average exhaust back pressure that would fix that.

RightBank6000.jpg

Right?

I don't think there's anything that one can do that moves the needle except changing the exhaust manifold geometry. And that's the challenge here.

[modok]

I can't read which trace is which, but the big green one must be the culprit, oh yeah it's RUINING the scavenging.... but look how fast the pressure FALLS.
Open that exhaust valve 20 degrees sooner see what happens

[ptuomov]

I can't read which trace is which, but the big green one must be the second cylinder exhausting, oh yeah it's RUINING the scavenging.... but look how fast the pressure FALLS.
Open that exhaust valve 20 degrees sooner see what happens

If I advance the cams by 20 degrees, everything gets a lot worse. If I retard them by 10 degrees, 6000rpm gets a little better. I speculate that with this exhaust manifold geometry, retarding #1 lobes and advancing #3 lobes might help a little but not fix the issue.

[modok]

yeah man, not the whole cam just open that valve sooner.
Is that not possible using your software?