JBrady
Active Member
- Messages
- 1,776
- Location
- Houston Texas
The nozzle collector is similar to a merge collector with internal modification. Some basic theory and understanding will help visualize what is going on.
The exhaust event has at least 2 phases.
BLOW DOWN (BD): starts before BDC as the exhaust valve opens towards the end of the power stroke. This is the highest pressure and speed and needs to effectively reduce the residual pressure as low as possible before BDC.
PUMP OUT (PO): After BDC the piston must push against any residual pressure and flow resistance to expel the remaining exhaust gasses.
It is important to understand that small pipes and poor manifolding present little problem to BD as the pressure differential is so great that it easily overcomes moderate resistance. In fact any resistance actually acts to exert pressure on the piston during the power stroke and will actually add to the crankshaft power.
PO is where so much can be gained or lost. If there any residual pressure from BD this steals power from the crank as it must push against this pressure from BDC to TDC. If two or more exhaust pipes are connected and the manifold design is poor there is the danger of the high pressure of BD feeding into the connect pipe and cylinder that is in its PO phase. Correctly designed manifolding can actually pull a vacuum on connected pipes and actually PULL on a piston during PO.
Exhaust flow obviously increases as RPM increases. At low RPM the exhaust flow is WAY below the pipes capacity when sized for peak power support. Even pipes undersized for peak power will still have much more capacity than low RPM exhaust flow needs. This is why too large of pipe and/or poor collector hurts low end power. Slower flow allows for reversion both in the individual exhaust port pipe as well as the rest of the pipes with decreasing effect. Long pipes with small diameters are better for low RPM power while short pipes with larger diameter better for high RPM power.
Now, with that primer in mind let’s consider the merge collector. The idea is to use the directional velocity of one pipe under BD to pull a vacuum on the connected pipe(s) or at the very least not create any unnecessary turbulence. Two of many important factors in a merge collector are the angle of convergence and the volume created in the collector volume. The collector volume is the area where all the pipes join BEFORE the exit pipe. You want to have the SMALLEST total collector volume. You also want the steepest convergence angle. Problem: the steeper the angle the larger the volume.
Now, visualize that the total area of the pipes leading into the merge is more than the area of the pipe from the merge. Also, the collector volume is larger than the total area of the pipes leading into it. So, as exhaust transitions from the feed pipe into the collector the area increases, this slows the velocity and then as it enters the exiting pipe the volume reduces and the velocity increases. This is an area of inefficiency. Non-merge collectors have a much larger volume and are even worse. That said collectors have a net positive effect by combining flow which increases velocity and reduces reversion.
Nozzle Collector: I take the shape of a 2-1 merge collector and extend the center of the merge a small amount. This creates a D shape at the end of each feed pipe. This does 3 things. It moves the velocity increase point from the outlet of the collector to the end of the feed pipe. It also reduces the collector volume. It also creates a one-way anti-reversion point at the feed outlet.
Now, if the combined volume of the two D areas of each feed pipe is the same net amount as the volume of the pipe exiting the collector… what have you lost in terms of peak flow? Nothing? Below peak flow what is the restriction on the engine with the D shaped pipe collector feeds? Nothing? I see this being a good enhancement for low to mid RPM operation with negligible restriction to peak flow… maybe none.
The increased velocity and reduced collector volume should provide a strong vacuum on the attached feed pipe. Transition from feed to collector to exit may improve. For a HiPo street application this should work great. The power increase and response will probably be below typical dyno pull RPMs. Higher flatter greater overall torque curve.
One last thing to chew on. Most header designs separate connect pipes to there is no shared exhaust event. If designed properly a nozzle and even merge collector can use the BD of one cylinder to pull a vacuum on the connect cylinder during its PO phase! Some designers are actually starting to pair cyl 1&2 and 3&4 on inline 4 cylinder engines due to this effect. Traditional Tri-Y headers for a 4 cyl would pair 1&4 and 2&3.
The exhaust event has at least 2 phases.
BLOW DOWN (BD): starts before BDC as the exhaust valve opens towards the end of the power stroke. This is the highest pressure and speed and needs to effectively reduce the residual pressure as low as possible before BDC.
PUMP OUT (PO): After BDC the piston must push against any residual pressure and flow resistance to expel the remaining exhaust gasses.
It is important to understand that small pipes and poor manifolding present little problem to BD as the pressure differential is so great that it easily overcomes moderate resistance. In fact any resistance actually acts to exert pressure on the piston during the power stroke and will actually add to the crankshaft power.
PO is where so much can be gained or lost. If there any residual pressure from BD this steals power from the crank as it must push against this pressure from BDC to TDC. If two or more exhaust pipes are connected and the manifold design is poor there is the danger of the high pressure of BD feeding into the connect pipe and cylinder that is in its PO phase. Correctly designed manifolding can actually pull a vacuum on connected pipes and actually PULL on a piston during PO.
Exhaust flow obviously increases as RPM increases. At low RPM the exhaust flow is WAY below the pipes capacity when sized for peak power support. Even pipes undersized for peak power will still have much more capacity than low RPM exhaust flow needs. This is why too large of pipe and/or poor collector hurts low end power. Slower flow allows for reversion both in the individual exhaust port pipe as well as the rest of the pipes with decreasing effect. Long pipes with small diameters are better for low RPM power while short pipes with larger diameter better for high RPM power.
Now, with that primer in mind let’s consider the merge collector. The idea is to use the directional velocity of one pipe under BD to pull a vacuum on the connected pipe(s) or at the very least not create any unnecessary turbulence. Two of many important factors in a merge collector are the angle of convergence and the volume created in the collector volume. The collector volume is the area where all the pipes join BEFORE the exit pipe. You want to have the SMALLEST total collector volume. You also want the steepest convergence angle. Problem: the steeper the angle the larger the volume.
Now, visualize that the total area of the pipes leading into the merge is more than the area of the pipe from the merge. Also, the collector volume is larger than the total area of the pipes leading into it. So, as exhaust transitions from the feed pipe into the collector the area increases, this slows the velocity and then as it enters the exiting pipe the volume reduces and the velocity increases. This is an area of inefficiency. Non-merge collectors have a much larger volume and are even worse. That said collectors have a net positive effect by combining flow which increases velocity and reduces reversion.
Nozzle Collector: I take the shape of a 2-1 merge collector and extend the center of the merge a small amount. This creates a D shape at the end of each feed pipe. This does 3 things. It moves the velocity increase point from the outlet of the collector to the end of the feed pipe. It also reduces the collector volume. It also creates a one-way anti-reversion point at the feed outlet.
Now, if the combined volume of the two D areas of each feed pipe is the same net amount as the volume of the pipe exiting the collector… what have you lost in terms of peak flow? Nothing? Below peak flow what is the restriction on the engine with the D shaped pipe collector feeds? Nothing? I see this being a good enhancement for low to mid RPM operation with negligible restriction to peak flow… maybe none.
The increased velocity and reduced collector volume should provide a strong vacuum on the attached feed pipe. Transition from feed to collector to exit may improve. For a HiPo street application this should work great. The power increase and response will probably be below typical dyno pull RPMs. Higher flatter greater overall torque curve.
One last thing to chew on. Most header designs separate connect pipes to there is no shared exhaust event. If designed properly a nozzle and even merge collector can use the BD of one cylinder to pull a vacuum on the connect cylinder during its PO phase! Some designers are actually starting to pair cyl 1&2 and 3&4 on inline 4 cylinder engines due to this effect. Traditional Tri-Y headers for a 4 cyl would pair 1&4 and 2&3.
Last edited:
