Exhaust Manifold

[Christ]

Originally Posted by Shawn D.

Well, how "low" are you and your passenger going to be? That's the frontal area issue, not the engine. If you put the engine upright behind you and your passenger, it will have a negligible effect on frontal area.

Even if your scheme was practical, there'd be no "pumping gains" to be had, really. Taking a few bends out of the exhaust is going to have a negligible effect. You'd be better off to fabricate a system with mandrel-bent curves instead -- gains from that are noticeable and well-documented.

You just contradicted yourself... if you're taking the bends out, flow will have increased velocity and lower pressure than if there are (even smooth) bends.

You say that there are no measurable gains from "taking a few bends out of the exhaust", yet you claim that there are measurable gains from "mandrel-bent curves" in the exhaust.

So, given the OP's question, you're essentially telling him that more power can be had using a mandrel-bent, curvy exhaust, than a straight, bend-free, less restrictive exhaust, correct?

In the OP's representation, the "few bends" that are being removed are normally critical pressure points... They're the areas that commonly are re-worked for higher performance to begin with. Straight flow in that area would help prevent flow-reversion characteristics and the overall pressure drop compared to a "normal" head design might yield interesting effects.

I wouldn't be so quick to shoot it down, is all I'm saying.

Unfortunately, with respect to the angle of the valves in the head, a truly flat, straight exhaust would require the engine be mounted at an angle, as well. I believe it's approx 15*, but don't quote me on that.

So that would require the engine be mounted at least 15* from horizontal... which is great for draining oil away from the cylinders, as well. Dry sump the system, and there is no pooling issue.

Given the money and milling apparatus, it could be pulled off, I say.

I'm with Frank on this one though. Fun to discuss, but unlikely to occur.

[theunchosen]

Originally Posted by Christ

You just contradicted yourself... if you're taking the bends out, flow will have increased velocity and lower pressure than if there are (even smooth) bends.

You say that there are no measurable gains from "taking a few bends out of the exhaust", yet you claim that there are measurable gains from "mandrel-bent curves" in the exhaust.

So, given the OP's question, you're essentially telling him that more power can be had using a mandrel-bent, curvy exhaust, than a straight, bend-free, less restrictive exhaust, correct?

^_^ Its fun to wait and see who corrects something when I sit on my hands(thread).

So Christ I was kinda waiting on you to show up here, I'm not aware of a block that has the exhaust exiting the block in a straight line from valve to tip so I don't have any way to speculate on the difference.

I would assume its pretty significant considering manufacturers always use smooth curves and make it travel in as much of a straight line as possible. Also too many curves are the problem with turbos and high performance race kits try and get the air to the turbine with as few twists and turns as possible.

[Christ]

Well, lets picture something for a second, then you can tell me why you don't think manufacturers have created heads with straight exhaust profiles yet.

View the attached images:


See the shape of the valves?

Now see how they sit in a cross-sectional view:


The second image is representative of a typical valved cylinder head.

Given the design you're attempting to achieve, your valves would have to operate in a completely different manner, so as to prevent the loss of gasses back into the head/oil containing areas, and to prevent the presence of mechanical parts in the (very hot) exhaust stream.

[Christ]

Given the second image, several things become evidently "bad" about current valve design and flow restriction, which, if addressed, would create significant gains on their own. Unfortunately, with technology advances occurring the way they do, the oft-quoted phrase comes to mind: if it's not broke, don't fix it.

It would cost a variable fortune to redesign completely such a small aspect of a much larger picture. It would essentially be like recoloring the faded sun in a painting of a magnitude reminiscent of the Milky Way.

IOW - there is easier to pick fruit within arm's reach, so why go to the shed to get the ladder?

[Frank Lee]

First things first: how much loss is there from a few turns?

And: When is a straight exhaust path beneficial- if at all- i.e. is this thing going to be run at WOT at high rpms a high percentage of the time?

Can you find an example of a perfectly straight exhaust system... anywhere on anything?

[Frank Lee]

Originally Posted by Christ

Given the second image, several things become evidently "bad" about current valve design and flow restriction, which, if addressed, would create significant gains on their own. Unfortunately, with technology advances occurring the way they do, the oft-quoted phrase comes to mind: if it's not broke, don't fix it.

It would cost a variable fortune to redesign completely such a small aspect of a much larger picture. It would essentially be like recoloring the faded sun in a painting of a magnitude reminiscent of the Milky Way.

IOW - there is easier to pick fruit within arm's reach, so why go to the shed to get the ladder?

If poppet valves are that bad maybe sleeve valves will take over any day now?

[theunchosen]

Ah, thanks I had forgotten to consider the valves themselves. I knew something was awry when I mentioned valves but I couldn't quite grasp it.

Ok so in that case there would be a very slight bend so that the exhaust pipe could miss the cams or whatever you have that drives the valves.

Doesn't have to miss it by much just enough room to put a thin layer of something to shield them. So instead of what looks like 45 degrees of overall bend in the above diagram to maybe 20 degrees to miss the cams. Compared to the final exhaust system(in my car) it shifts 90 degrees straight out 90 down 90 towards the rear of the car(in sloped curves) opposed to the proposed 20-30 degrees total?

[Christ]

Originally Posted by Frank Lee

First things first: how much loss is there from a few turns?

And: When is a straight exhaust path beneficial- if at all- i.e. is this thing going to be run at WOT at high rpms a high percentage of the time?

Can you find an example of a perfectly straight exhaust system... anywhere on anything?

1 - The gains surely wouldn't justify the expenses of labor and research/development. There is obviously "lower hanging fruit".

2 - A straight exhaust path would be beneficial if properly designed, even under low RPM circumstances. A straight path could be designed to be non-restrictive and still provide optimum flow for the exhaust gasses at a given engine speed/volume output. The length and stepped sizing of the exhaust could be tuned accordingly to create the harmonic scavenging of exhaust flows, creating a sort of "super efficient" area in the engine's operating range. Obviously, that area should be the most often used speed.

Inversely, scavenging can be had simply by combining the exhaust pulses, which a true straight exhaust simply cannot do.

3 - Jet engines?

[Frank Lee]

You have knowledge of straight through systems and how much the gains are?

[Christ]

Frank - I'm not saying poppet valves are a bad thing, I'm saying exactly the opposite. There are obviously gains to be had from modifying the flow tract available with poppet style valves, but there are also other redesigns that would please the wallet and engineering minds much more.

theunchosen - The total exhaust package on my Caravan actually looks something like a teardrop when viewed on a single axis, even given that the engine sits relatively high in the engine bay, and the initial exhaust pulse happens perpendicular to the ground.

For lack of imagery, consider how an upright inline engine is normally oriented, then imagine that the normal exhaust manifold exhausts fumes at about -20* to the flange surface of the head, where exhaust continues to flow mostly in a straight line, until there is a 20* bend (-20 + 20 = straight) which continues almost the length of the van.