Exhaust... stuck in a way of thinking..?

[Jon]

Last time I attended a RWYB I had the pleasure of listening to a highly tuned 911, with dual Stingers... I have always loved that look and seeing it racing late at night was really special. When I stumbled over this pic it hit me a second time...



why did Porsche do this kind of thing back in the day? And why didn't the VW racers copy them?
I have always "known" that the pipes in he collector should be arranged after firing order... to induce a nice screw motion in the exhaust... where I got that from I have no idea... I suspect i read it someplace. With this "knowledge" in the bottom I have never really thought of a alternative to the standard VW header design.
My little engine wants short pipes.. real short pipes, and that is hard with just one collector. The big boys run the pipes over the engine to get it short enough.
If I did make a couple of these headers I could have made them with two different collectors to tune the engine after whatever use I wanted.
The only real disadvantage would be the posible "trike" sound.
Any ideas why Porsche did this, and why it would be a big mistake on a VW?

[nicolas]

very cool find. i would like to know more about it as well

good sponsor for that car BTW

[alex d]

Any ideas why Porsche did this, and why it would be a big mistake on a VW?

hmmmmm, the fact that they have 6 cylinders and we have just 4 may have something to do with it for sure

[Jon]

Yes, they have three sylinders firing into one pipe, and we would have two. To me it looks like the most important things about heathers are primary pipe lenght and Primary pipe size. In this article for intance, the firing order isn't even mentioned http://www.carcraft.com/techarticles/header_basics/index.html, I would think they would if it was a key factor?
I have looked at the firing order in the 911 to see if there can be a clue there, but it doesn't look like it matches in any particular way..??


I mean if it was good enough for racing Porches, why isn't it good enough for a bug? 

[Soddli]

Have to aggre with you JHU.
I have not reflected much about it, but just look at the exhaust at the 904...2-1 headers with megaphones there also.
And on 911's it is only 3-1 that is used in racing...look:
http://www.paragon-products.com/product_p/p911-ba.htm

JUST DO IT!

[Soddli]

Here is a picture of the 904 race-system:

[John Maher]

I mean if it was good enough for racing Porches, why isn't it good enough for a bug? 

The answer is in the diagram you posted. Each side of the Porsche flat six fires its cylinders with equal spacing between each power stroke. Therfore you can treat each side of the engine as a separate entity as far as running two tuned exhaust sytems is concerned. For maximum efficiency you want each exhaust pulse entering the collector in a regularly spaced sequence.

Now take the VW flat four...
The full 4 stroke cycle takes two crankshaft revolutions i.e. 720°
One exhaust event occurs every 180° of crank rotation (one quarter of the cycle).
If you build an exhaust system along the lines you propose, the two exhaust events per head, per 4 stroke cycle, will be 180° apart, followed by a 360° gap, while the same thing takes place on the other side of the engine (but during the 360° gap on the other side).

Not the best approach for designing a well tuned extractor. End result... loss of power.

Ever wondered why those dual cannon systems you see on buggies are a POS? 

[Jon]

Thanks for a good response, I sorta knew that the Buggy way wasn't the way to do it, but didn't know exactly why.
How does evenly spaced pulses improve performance? Is there a simple way of explaining it?
If I get you right, joining 1 with 3 in one collector, and 4 with 2 in the other collector would be the way to do it?
But would the benefits of a short primary weigh up for the positive effect of having all fours coming together in the collector?

I know JPM have been talking about testing a 4-2-1 system...

[alex d]

hmmmmm, maybe joining 1 with 3 and 4 with 2 would make more sense...

[Jon]

hmmmmm, maybe joining 1 with 3 and 4 with 2 would make more sense...

Thanks for the tip!

[John Maher]

How does evenly spaced pulses improve performance? Is there a simple way of explaining it?
If I get you right, joining 1 with 3 in one collector, and 4 with 2 in the other collector would be the way to do it?
But would the benefits of a short primary weigh up for the positive effect of having all fours coming together in the collector?

When each pulse exits the primary pipe into the collector, a negative wave returns back up the pipe. If it reaches the heads just as an exhaust valve is about to open, it will help scavenge the cylinder of gasses more effectively than the pulse returning while the exhaust valve is closed. This improved extraction in turn can help pull more intake into the cylinder during the overlap period. Merging the 4 pipes together improves scavenging and ensures each cylinder gets the same hit. Merging pipes 1-2 and 3-4 will have little or no scavenge effect and power will suffer as a result.

Depending on primary pipe length, the rpm at which peak scavenging occurs will vary. Tuning the pipe length to suit a particular rpm will lift torque beyond that possible with a non-tuned header.  The longer the pipes, the lower the rpm the effect. Shorter pipes are used when trying to bump up torque at higher rpm. There are formulae and computer programmes to help predict and design tuned length systems

A 4-2-1 system is common on many inline 4 cylinder engines. It's not the ultimate for peak bhp but tends to boost mid-range torque. Depending on application it may offer an overall performance advantage. Highly unlikely you'll see a 4-2-1 header on a drag car but can be useful for circuit racers where an increase in mid-range is more beneficial than maximising peak power higher up the rpm range. Packaging one on a flat four is tricky - more difficult to manufacture than a typical 4 into 1 and probably the main reason for lack of availability.

Many years ago I built a 1914cc engine for a Spyder kit car. The customer opted to have a custom stainless steel header made specifically to suit the car - a professionally made 4-2-1 design. Compared to previous engine builds of similar specification it was down on peak power but torque in the mid-range was significantly better than I'd seen before for the specification.

Generally speaking, what you gain in one part of the rpm range, you'll lose somewhere else. If you're only interested in one specific part of the rpm range, stuff like this can pay dividends.

[Jon]

There are formulae and computer programmes to help predict and design tuned length systems

Hehe thats how I ended up trying to find a way to build a super short exhaust... it's hard getting under 800mm primarys with a conventional VW header.
I have that "scientific design of inatke an exhaust systems" book, so I aware of the negative pulses, but I must have missed the savaging effet you mention, back to the book I guess.

What do you think about having the primary's arranged 1-4-3-2 in the collector? I guess the scavenging effect works like a ejector, if so it wouldn't matter to much?
Thanks for great replys by the way!