Improving Exhaust Performance with Crossover Pipes in 4 Cylinder Headers

[tsslaporte]

Currently this engine uses a 4-2-1 Setup with a crossover a few inches away from the exhaust port on the header primaries, only on cylinders 2 and 3(which is well before the first collector)

This Engines Firing Order is 1-2-4-3The Crossover linking Cylinders 2 and 3 confuse me since are not paired. Since Cyl 2 Fires then next to fire is Cyl 4, so there isn't any scavenging from these two (Cyl 2 and Cyl 3). I believe this is done to alter the sound of the exhaust, is my assumption correct?I have an idea to create a dual exhaust system which

(Cyl 1 and Cyl 2)
and
(Cyl 4 and Cyl 3)

Are paired

So the exhaust is a Dual 2-1 system. But, now two cylinders will not receive scavenging (Cyl 1 and Cyl 4)

To Remedy this I propose to put crossover pipes on the header primaries between

(Cyl 4 and Cyl 2)
and
(Cyl 1 and Cyl 3)

So now all cylinders can receive scavenging.
What are your opinions on the system? Is there any literature on crossover pipes in headers?

 

[jack action]

The Crossover linking Cylinders 2 and 3 confuse me since are not paired. Since Cyl 2 Fires then next to fire is Cyl 4, so there isn't any scavenging from these two (Cyl 2 and Cyl 3). I believe this is done to alter the sound of the exhaust, is my assumption correct?

First, cyl 2 and 3 are paired. To find which cyl is paired with which one, you take the firing order (ex.: 1-8-4-3-6-5-7-2) and separate it in 2 groups:
1-8-4-3
6-5-7-2

The pairs are the following: 1 & 6, 8 & 5, 4 & 7 and 3 & 2.

So in your case: 1-2-4-3 gives 1 & 4 and 2 & 3.

Second, even if they are paired, what you see is probably not a «desired» crossover performance-wise but something for functionality. A picture would be nice. It may be a place to put the O₂ sensor similar to this header:

Finally, you seem to think that cylinders must be connected together to help scavenging, and this is not really how it works.

The basics are that you can help scavenging of a single cylinder by choosing an appropriate diameter and length of the exhaust pipe which is discharging directly into the atmosphere. Simple as that; you can't get more scavenging than that. What is actually helping the scavenging is the fact that at the pipe outlet there is an infinite area change («diameter» of the atmosphere «pipe» is infinite) and that sends a strong pressure wave back to the engine which - if timed correctly - will help scavenge the cylinder on the next cycle.

Two problems arise with such a simple system:

First, rules, available space, practicality or cost forces you to merge pipes together.

Second, it is timed for one particular rpm only. That tends to make the highest hp gain you can have, but in a very small rpm range. If you use your engine in a wider rpm range, it is the average hp in that range that counts, not just how high it can go.

To resolve this problem, this is where stepped tube headers (primary pipe diameter varies along its length), crossover tube or pairing of primaries (such as 4-2-1 designs) come into play. What all these designs do is to increase gradually the area change such that instead of having one big pressure wave returned at the end of the pipe, you have many small ones returned along the pipe length at the expense of a weakened one at the end of the pipe. This provides a smaller hp gain but on a wider rpm range. When you joined 2 pipes together, choosing the right ones such that one will help the other is also a concern. But it does not create a bigger pressure wave than what you could've had with both pipes separated; You can only, at most, recover what was lost.

 

[tsslaporte]

First, cyl 2 and 3 are paired. To find which cyl is paired with which one, you take the firing order (ex.: 1-8-4-3-6-5-7-2) and separate it in 2 groups:
1-8-4-3
6-5-7-2

The pairs are the following: 1 & 6, 8 & 5, 4 & 7 and 3 & 2.

So in your case: 1-2-4-3 gives 1 & 4 and 2 & 3.

Second, even if they are paired, what you see is probably not a «desired» crossover performance-wise but something for functionality. A picture would be nice. It may be a place to put the O₂ sensor similar to this header:


Finally, you seem to think that cylinders must be connected together to help scavenging, and this is not really how it works.

The basics are that you can help scavenging of a single cylinder by choosing an appropriate diameter and length of the exhaust pipe which is discharging directly into the atmosphere. Simple as that; you can't get more scavenging than that. What is actually helping the scavenging is the fact that at the pipe outlet there is an infinite area change («diameter» of the atmosphere «pipe» is infinite) and that sends a strong pressure wave back to the engine which - if timed correctly - will help scavenge the cylinder on the next cycle.

Thank you for your reply.

The one with the crossover on 2 and 3 is the smaller displacement 600cc engine, the headers without it is the 1000cc engine.

Why would they remove scavenging on the bigger engine? There is no 02 sensor there.

Yes I understand that at the end of an Opened pipe is the best for scavenging since the gas rapidly expands outside of the pipe leaving a depression inside. The reason you connect cylinders is so that the other cylinder can see this effect during valve overlap.

I also learned that my setup would only see 540deg scavenging from one cylinder at a time, where as what you mentioned 1-4, 2-3 would see scavenging every 360 deg.

So I understand that now.

But I am not understanding the pairing the stock headers (pictured) use and that cross over? I don't seem to properly understand pairing I guess.

 

[jack action]

More interesting with a picture.

Keeping in mind that it is practically impossible to determine if an exhaust system design is good or not by simply looking at it (only testing or a good computer simulation can) I'll try a guess, assuming people who designed these exhaust systems knew what they were doing. Also, the definition of «good design» depends on what you want to do with the engine.

We already said that 2 & 3 and 1 & 4 should be paired, which is not the case in these exhaust systems (obviously it is a constraint due to the fact that it has to fit a motorcycle). I guess that this crossover will help restore that pairing (at least for one pair). The fact that it is so close to the valves means that the effect only appears at high rpm (it simulates short primary pipes), and that would favour high peaky hp. The 600cc has usually a higher rpm than a 1000cc.

Why does the 1000cc doesn't have a crossover lower on down the pipe? Maybe it was preferred to promote hp at lower rpm because it was felt that there was enough power at high rpm due to higher displacement?

 

[tsslaporte]

More interesting with a picture.

Keeping in mind that it is practically impossible to determine if an exhaust system design is good or not by simply looking at it (only testing or a good computer simulation can) I'll try a guess, assuming people who designed these exhaust systems knew what they were doing. Also, the definition of «good design» depends on what you want to do with the engine.

We already said that 2 & 3 and 1 & 4 should be paired, which is not the case in these exhaust systems (obviously it is a constraint due to the fact that it has to fit a motorcycle). I guess that this crossover will help restore that pairing (at least for one pair). The fact that it is so close to the valves means that the effect only appears at high rpm (it simulates short primary pipes), and that would favour high peaky hp. The 600cc has usually a higher rpm than a 1000cc.

Why does the 1000cc doesn't have a crossover lower on down the pipe? Maybe it was preferred to promote hp at lower rpm because it was felt that there was enough power at high rpm due to higher displacement?

Thanks,

For crossovers is their any info on them? Like a paper or book?

 

[jack action]

You will not find info on crossover alone. Exhaust system performance is almost impossible to predict just by eye-balling it or with simple formulae. The pressure waves travel at the speed of sound, which in turn varies with temperature; and exhaust gas temperature can vary greatly from one end of the pipe to the other. To make matters worst, pressure waves bounce back from one end to the other and crosses each other. Each time, they modify each other's characteristics; variations that only a full computer simulation can predict. One good book to learn about the subject (although not an easy read) is http://books.sae.org/book-r-186/. There is a similar book by the same author for http://books.sae.org/book-r-161/.

By the way, I just taught of another reason why the 1000cc wouldn't need a crossover. Is it possible that the 1000cc does not have the same firing order as the 600cc?

 

[tsslaporte]

Yep Same Firing order.

I was eying that book before, Ill buy it soon. Even aftermarket exhaust for this engine still use the same type of layout its odd I guess it could be for packaging. Youd think in this competitive market that they would pair it right but I guess they need to keep upgrading it for future years. I looked at another 4 cyl (S2000) and one of the aftermarket headers were paired right.