Designing a header for fuel economy (for a Metro G10 1.0L)

[freebeard]

From Permalink #90


Did we decide you'd get better scavenging with a radius between the pipes? Instead of V-notching and circularizing the three pipes, maybe cut them with a V-tab and bend it out to for 1/3 of the transition. The cut the collector with three V-tabs and bend them in to close.

Edit: I think there would be less moving metal around, unless you triangularize the collector piece. Or maybe even if...

[Shawn D.]

Originally Posted by freebeard

From Permalink #90


Did we decide you'd get better scavenging with a radius between the pipes? Instead of V-notching and circularizing the three pipes, maybe cut them with a V-tab and bend it out to for 1/3 of the transition. The cut the collector with three V-tabs and bend them in to close.

Nope, you decided that. This isn't about blood flow. If a radius like that were better, that configuration would be used in high-$$$$ systems by top race teams, but it isn't. To the contrary, to prevent reversion it's considered preferable to have a sharp step instead of a smooth bend like that at a transition.

[freebeard]

Fair enough. But it wasn't falsified (I think). There was a post between #90 and #91 that I replied to, that has since been deleted.

It might broaden the rpm range it's effective in or weaken the overall effect. Hot rodders would test this on a flow bench.

[Daox]

Oh to have a flow bench and hours upon hours of time to test fun things...

[freebeard]

AutoSpeed - Ultimate DIY Automotive Modification Tool-Kit, Part 3

With the Dwyer Magnehelic manometer you'd be well on the way. They cost $30 or whatever, but I'm not sure what you could do short of a test engine on a bench for the exhaust pulses.

[Shawn D.]

Originally Posted by freebeard

Fair enough. But it wasn't falsified (I think). There was a post between #90 and #91 that I replied to, that has since been deleted.

It might broaden the rpm range it's effective in or weaken the overall effect. Hot rodders would test this on a flow bench.

A flow bench is a constant-state test. For exhaust systems, the tests are done on a dyno on a running engine, not a flow bench, as a dyno is dynamic. In an exhaust system, the key word is "system" and the effects must be tested against multiple dynamic branches, not an isolated component in constant flow. You cannot test scavenging, reversion, wave travel, or effects on other cylinders by using a flow bench.

While I think your point about the relative volume of branches is valid, I do not believe your point about the smooth radii is. The real reason the radii are smooth between blood vessel branches is structural. Any time you have a "notch" between two components with a sharp transition, that's a high-stress area that can lead to failure. It's true for real "branches" on trees as well -- any tree (like a Bradford pear) with sharp notches between a trunk and a branch is much weaker and are more likely to fail with high wind or snow load. This is also the reason for structural fillets that are added for no performance-related reason, as well as the purpose of "stop drilling" a crack to stop its propagation.

[freebeard]

I was already backing away from the flow bench. What sensor would you say are used on the dyno. A high-speed recording manometer? Microphones?




I did this to illustrate the end-cuts I proposed. The collector would have three tabs. Convergent angle is not represented. This is thirds of a Christmas wrapping core, not what you were thinking.

What is the relationship between scavenging and anti-reversion. It seems they are at cross purposes.

[RobertISaar]

scavenging can cause anti-reversion if there is a low-pressure area at the end of the primary when the exhaust valve opens for a given cylinder.

[oil pan 4]

What if its too efficient and vacuums out the cylinder causing more air and fuel to be pulled in?

This would be ideal for an N/A diesel in all conditions, because more air is always better for them.
For a gasoline engine this is only good if you want to burn lots of fuel and go fast.

[RobertISaar]

Originally Posted by oil pan 4

What if its too efficient and vacuums out the cylinder causing more air and fuel to be pulled in?

possibility, generally one would design a cam around it with a tiny amount of overlap