Why Not Cross Tri-Y's (180 deg header alternative)?
06-05-2013, 05:38 PM
#32
Rennlist Member
Could be? I'd always thought of exhaust pairing in termers of engine cycle, not crank cycle but I could be wrong in my perception.
So, what then would you call the typical tri-Y on an inline 4, as far a s degree seperation, and for that mater what cylinders would you pair with the traditional firinig order of 1, 2, 4, 3?
So, what then would you call the typical tri-Y on an inline 4, as far a s degree seperation, and for that mater what cylinders would you pair with the traditional firinig order of 1, 2, 4, 3?
06-05-2013, 06:01 PM
#34
Rennlist Member
06-05-2013, 06:09 PM
#35
Rennlist Member
06-05-2013, 09:14 PM
#36
Official Bay Area Patriot
Fuse 24 Assassin
Rennlist Member
Fuse 24 Assassin
Rennlist Member
Let's be realistic here; if 180 degree headers are re-made, can we place have something sectioned so that it's easier to install on these cars? I'm talking about collectors where pipes off each bank merge, then those merges meet up with another collector downstream. It shouldn't be that hard to do. Yes, I do know the more collectors and flanges there are, the more leaks may occur.
06-05-2013, 09:44 PM
#37
Archive Gatekeeper
Rennlist Member
Rennlist Member
FWIW, here are the Tri Y's that are (were) on the Zombie.
Creatively wrapped around the steering shaft:
Creatively wrapped around the steering shaft:
06-06-2013, 05:45 AM
#38
Nordschleife Master
When exhaust pulses are considered to be 180 it is usually with reference to one cycle of the ignition system sequence so in effect it would be 360 degrees of the crank on a 4 stroke engine. The easiest way to visualize this is in reference to the distributor (provided a single distributor system, dual distributors complicate this visualization).
The reason the Tri-Y works well on the inline 4 is that it is a flat plane crank. Your pairing would work well on a flat plane V8 as well, but with a cross plane crank the 180 degree firing cylinders are on opposite banks. For a flat plane crank the firing order would be something like 1, 7, 2, 5, 4, 6, 3, 8. So from this you can now see that you would get 180 firing cylinders paired using the Tri-Y exhaust. On a cross plane
To truly get 180 pairing on our engines you would have to pair (1, 6), (3, 5), (7, 4) and (2, 8).
The reason the Tri-Y works well on the inline 4 is that it is a flat plane crank. Your pairing would work well on a flat plane V8 as well, but with a cross plane crank the 180 degree firing cylinders are on opposite banks. For a flat plane crank the firing order would be something like 1, 7, 2, 5, 4, 6, 3, 8. So from this you can now see that you would get 180 firing cylinders paired using the Tri-Y exhaust. On a cross plane
To truly get 180 pairing on our engines you would have to pair (1, 6), (3, 5), (7, 4) and (2, 8).
OTOH my back of envelope doodle aimed for best primary flow pairing 1/2, 3/4, 5/7, 6/8 in which have least flow overlap, then tuned pairing of 1/2 + 6/8 at the secondary merge, but same question.
06-06-2013, 12:33 PM
#39
Rennlist Member
Thread Starter
OK, you guys made me put together another diagram! 
FJ, I did not diagram what I call the 360 degree setup, because I believe it is too difficult to accomplish for the 928 configuration!
I don't believe there is anything bad about the 1/7 4/6 merging. They are 180 degree separations. Re-posted my original diagram (first one, below) - to me, this would work very well.
Agree with you and ptuomov that your suggested pairings of the primaries create less flow overlap in the secondary pipes. However, look what happens (second diagram, below) if you then Y-merge these secondaries into tertiary pipes. You get 90 degree bunched pulses again - this is what I thought needs to be avoided. Also, if you went this route, it would seem to me that getting the lengths/flows would become quite critical, and earlier in the piping.
This problem in the tertiaries, for the least-overlap primary pairings you and ptuomov are suggesting, is what ptuomov was pointing out, and why (I believe) he suggested doing a 4-to-1 merge to the secondaries and skipping the tertiary pipes altogether. That could be an interesting set-up (and sound), but unless you get one huge cat, or split it back out afterwards and place cats further downstream (making them much less effective, I believe), it could not be done on a car requiring cats.
FJ, I did not diagram what I call the 360 degree setup, because I believe it is too difficult to accomplish for the 928 configuration!
This problem in the tertiaries, for the least-overlap primary pairings you and ptuomov are suggesting, is what ptuomov was pointing out, and why (I believe) he suggested doing a 4-to-1 merge to the secondaries and skipping the tertiary pipes altogether. That could be an interesting set-up (and sound), but unless you get one huge cat, or split it back out afterwards and place cats further downstream (making them much less effective, I believe), it could not be done on a car requiring cats.
Last edited by hernanca; 06-06-2013 at 12:39 PM. Reason: Cropped those dang diagrams!
06-06-2013, 01:42 PM
#41
Nordschleife Master
How about you pair 1/2, 3/4, 5/7, 6/8 but make the 2, 3, 5, and 8 primary runners longer? Longer by enough to make the pulses equally spaced at the y's at the tuned rpms. I haven't worked out the math to see if the length required for this would be absurdly long. Quick back of the envelope math says that it may be too long of a difference to be practical: 1800 ft/s *(1/(5500 rpm * 60 sec/m * 4)) = 4.9 ft. Ok, back to the drawing board...
06-06-2013, 03:01 PM
#42
Rennlist Member
Generally the intent of exhaust design is to control the exhaust pulse reflected wave effect for scavenging between cylinders. You will have multiple reflections traveling in any given section of the exhaust. These reflected pulses will be determined by several different effects, i.e. length, diameter, diameter step change, collector/tube merger…
The design of the tri-Y system typically includes equal length primarys and equal length secondarys. The intent is to achieve the same reflected wave profile for each cylinder so that all cylinders perform equally and at the same tune. This is achieved by pairing cylinders that are 180 degrees out of phase (relative to engine cycle. This will give evenly spaced primary pulses as you will have a pulse occur every 180 engine cycles in each given primary tube. The primaries will in turn generate pulses in the secondarys at the same 180 degree separation. The reflected wave in the secondarys will be influenced by the merger of the secondarys into the collector for the single exhaust where the collector will see evenly spaced pulses every 90 engine cycle degrees. The exhaust will end up seeing the reflected wave impulse that is the added sum of all the reflected waves. The wave pattern will be determined by the length and diameter of the given sections, but provided the system is symmetrical all cylinders will see a similar pattern.
When you pair cylinders that are not evenly spaced you can get reflected wave patterns that are different between cylinders, this can be combated to some extent by varying the runner length to create even pulses at the collector, however this will only be tunable at a given RPM. Anything off the tuned rpm will have different cylinders seeing different pulse reflection patterns. This difference in tuning can be compensated for to some extent on sequential injection systems by using individual cylinder trimming for different conditions. This is not to say that you can’t build an effective header with unevenly spaced exhaust pulsing, as that is essentially what almost all 928s are currently running and we all know that there are some very well performing 928s out there. I was merely wanted to expand on the discussion, with regard to optimizing the exhaust design in theory more so than in practice. I’m not saying it is even practical to implement a 180 cycle degree exhaust system on the 928. I was in no way saying that your design doesn’t have merit, just that if you were trying to mimic the effect of duplicating 2 four cylinder tri-Y systems that the proposed pairing was not quite as effective.
The reality is that everything is a compromise to some extent and who know it may not give the same peak that “properly paired” header gives it may work in favor of increasing the overall area under the curve which would actually be more beneficial.
The design of the tri-Y system typically includes equal length primarys and equal length secondarys. The intent is to achieve the same reflected wave profile for each cylinder so that all cylinders perform equally and at the same tune. This is achieved by pairing cylinders that are 180 degrees out of phase (relative to engine cycle. This will give evenly spaced primary pulses as you will have a pulse occur every 180 engine cycles in each given primary tube. The primaries will in turn generate pulses in the secondarys at the same 180 degree separation. The reflected wave in the secondarys will be influenced by the merger of the secondarys into the collector for the single exhaust where the collector will see evenly spaced pulses every 90 engine cycle degrees. The exhaust will end up seeing the reflected wave impulse that is the added sum of all the reflected waves. The wave pattern will be determined by the length and diameter of the given sections, but provided the system is symmetrical all cylinders will see a similar pattern.
When you pair cylinders that are not evenly spaced you can get reflected wave patterns that are different between cylinders, this can be combated to some extent by varying the runner length to create even pulses at the collector, however this will only be tunable at a given RPM. Anything off the tuned rpm will have different cylinders seeing different pulse reflection patterns. This difference in tuning can be compensated for to some extent on sequential injection systems by using individual cylinder trimming for different conditions. This is not to say that you can’t build an effective header with unevenly spaced exhaust pulsing, as that is essentially what almost all 928s are currently running and we all know that there are some very well performing 928s out there. I was merely wanted to expand on the discussion, with regard to optimizing the exhaust design in theory more so than in practice. I’m not saying it is even practical to implement a 180 cycle degree exhaust system on the 928. I was in no way saying that your design doesn’t have merit, just that if you were trying to mimic the effect of duplicating 2 four cylinder tri-Y systems that the proposed pairing was not quite as effective.
The reality is that everything is a compromise to some extent and who know it may not give the same peak that “properly paired” header gives it may work in favor of increasing the overall area under the curve which would actually be more beneficial.
06-06-2013, 05:32 PM
#43
Rennlist Member
Thread Starter
How about you pair 1/2, 3/4, 5/7, 6/8 but make the 2, 3, 5, and 8 primary runners longer? Longer by enough to make the pulses equally spaced at the y's at the tuned rpms. I haven't worked out the math to see if the length required for this would be absurdly long. Quick back of the envelope math says that it may be too long of a difference to be practical: 1800 ft/s *(1/(5500 rpm * 60 sec/m * 4)) = 4.9 ft. Ok, back to the drawing board...
Generally the intent of exhaust design is to control the exhaust pulse reflected wave effect for scavenging between cylinders. You will have multiple reflections traveling in any given section of the exhaust. These reflected pulses will be determined by several different effects, i.e. length, diameter, diameter step change, collector/tube merger…
The design of the tri-Y system typically includes equal length primarys and equal length secondarys. The intent is to achieve the same reflected wave profile for each cylinder so that all cylinders perform equally and at the same tune. This is achieved by pairing cylinders that are 180 degrees out of phase (relative to engine cycle. This will give evenly spaced primary pulses as you will have a pulse occur every 180 engine cycles in each given primary tube. The primaries will in turn generate pulses in the secondarys at the same 180 degree separation. The reflected wave in the secondarys will be influenced by the merger of the secondarys into the collector for the single exhaust where the collector will see evenly spaced pulses every 90 engine cycle degrees. The exhaust will end up seeing the reflected wave impulse that is the added sum of all the reflected waves. The wave pattern will be determined by the length and diameter of the given sections, but provided the system is symmetrical all cylinders will see a similar pattern.
When you pair cylinders that are not evenly spaced you can get reflected wave patterns that are different between cylinders, this can be combated to some extent by varying the runner length to create even pulses at the collector, however this will only be tunable at a given RPM. Anything off the tuned rpm will have different cylinders seeing different pulse reflection patterns. This difference in tuning can be compensated for to some extent on sequential injection systems by using individual cylinder trimming for different conditions. This is not to say that you can’t build an effective header with unevenly spaced exhaust pulsing, as that is essentially what almost all 928s are currently running and we all know that there are some very well performing 928s out there. I was merely wanted to expand on the discussion, with regard to optimizing the exhaust design in theory more so than in practice. I’m not saying it is even practical to implement a 180 cycle degree exhaust system on the 928. I was in no way saying that your design doesn’t have merit, just that if you were trying to mimic the effect of duplicating 2 four cylinder tri-Y systems that the proposed pairing was not quite as effective.
The reality is that everything is a compromise to some extent and who know it may not give the same peak that “properly paired” header gives it may work in favor of increasing the overall area under the curve which would actually be more beneficial.
The design of the tri-Y system typically includes equal length primarys and equal length secondarys. The intent is to achieve the same reflected wave profile for each cylinder so that all cylinders perform equally and at the same tune. This is achieved by pairing cylinders that are 180 degrees out of phase (relative to engine cycle. This will give evenly spaced primary pulses as you will have a pulse occur every 180 engine cycles in each given primary tube. The primaries will in turn generate pulses in the secondarys at the same 180 degree separation. The reflected wave in the secondarys will be influenced by the merger of the secondarys into the collector for the single exhaust where the collector will see evenly spaced pulses every 90 engine cycle degrees. The exhaust will end up seeing the reflected wave impulse that is the added sum of all the reflected waves. The wave pattern will be determined by the length and diameter of the given sections, but provided the system is symmetrical all cylinders will see a similar pattern.
When you pair cylinders that are not evenly spaced you can get reflected wave patterns that are different between cylinders, this can be combated to some extent by varying the runner length to create even pulses at the collector, however this will only be tunable at a given RPM. Anything off the tuned rpm will have different cylinders seeing different pulse reflection patterns. This difference in tuning can be compensated for to some extent on sequential injection systems by using individual cylinder trimming for different conditions. This is not to say that you can’t build an effective header with unevenly spaced exhaust pulsing, as that is essentially what almost all 928s are currently running and we all know that there are some very well performing 928s out there. I was merely wanted to expand on the discussion, with regard to optimizing the exhaust design in theory more so than in practice. I’m not saying it is even practical to implement a 180 cycle degree exhaust system on the 928. I was in no way saying that your design doesn’t have merit, just that if you were trying to mimic the effect of duplicating 2 four cylinder tri-Y systems that the proposed pairing was not quite as effective.
The reality is that everything is a compromise to some extent and who know it may not give the same peak that “properly paired” header gives it may work in favor of increasing the overall area under the curve which would actually be more beneficial.
06-06-2013, 05:51 PM
#44
Rennlist Member
I still have these headers and exhaust system. I hope to get it cleaned up and go for a test fit, as I have never tried them myself. May be helpful to the thread though, most compact tri-y I have seen for the 928.