Exhaust Collector Discussion

[TGP37]

But enough about him... back to the interesting technical discussion.

Here are a few teaser diagrams of my new v6 header collectors. When I decide to weld up the turbo headers, currently stuck with the stock logs.









Made in Studio Max. The idea is no open turbulant area in the center and each pipe directs its exhaust pulse at a slight offset. This means the previous pulse will be less of a restriction and a vortex should be created.

[SappySE107]

I don't know if that is what will happen though unless you have equal length primaries, and even then you want the smoothest transition possible for each pulse to retain its strength. The speed of the exhaust vs the length of the primaries will determine when if ever that vortex could be created.

[bob442]

Here are a few teaser diagrams of my new v6 header collectors. When I decide to weld up the turbo headers, currently stuck with the stock logs.

[ATTACH=CONFIG]6039[/ATTACH]

[ATTACH=CONFIG]6040[/ATTACH]

[ATTACH=CONFIG]6041[/ATTACH]

[ATTACH=CONFIG]6042[/ATTACH]

Made in Studio Max. The idea is no open turbulant area in the center and each pipe directs its exhaust pulse at a slight offset. This means the previous pulse will be less of a restriction and a vortex should be created.

i'll take one PM me

[ForcedFirebird]

True merge 3-1 will naturally create a "vortex". I don't know if over-engineering would benefit here, there's a reason Burns collectors are consistently winning races.

[Superdave]

I'd do some small scale testing on that design well before trying to cut that out of some steel tubing, building a test rig to accurately see what's going on will be a major undertaking though.

[HOYS]

Cutting those pipes would be awful hard.

[pocket-rocket]

I bet you could print out some paper guides and use a die grinder to rough in those curves.

[TGP37]

It does take some work, cutting the steel. But I did make a molding around the bend which can be cut with razor. I'll transfer the template on screen to the paper molds (I have an unwrap printed out but it is messy). Once the paper molds, which are pretty strong, are cut to shape just right, then they slide on the steel nice and tight. A quick blast with spray paint to mark the cut-out area hit it with the cutters.

I have a rotary cutter that can slice this steel fairly well.

Weld it up and then just clean up the insides like any port job.



Part of the offset in the design is for a reason. 1-3-5 or 5-3-1 will change how the next primary interacts with the previous primary. So I'm not sure which way yet, CW or CCW. But one way places the exhaust pulse much closer to the previous pulse in a tighter spiral. The other way places the next pulse further away in a more open spiral. Not sure which way would be better.

Also, there 'should' be a negative pressure wave created in the other two pipes as the blast is pulsed out the one.

But then it dawned on me, for velocity and other special dynamics, the blending of all 3 should meet in a spot of similar diameter to flow 1 pipe, not 3. The negative pressure wave created in the other 2 would be stronger.

So I need to flow test two different ideas, but wow, flow testing a pulsing exhaust is going to be rough. Can't just push air through all 3 at once.

next idea.....

[TGP37]

If all works out well, and I can replicate my final choice, I would definately make a few more for the cost of materials and a little work.

We'll cross that bridge once I get the first pair made up. The first pair will take the longest, charting cut lines, etc.

[Superdave]

A few years ago i was testing some exhaust stuff for the 60v6 and built a pulse rig using 3 120VAC air solenoids hooked up to the air compressor. I was using a variable speed DC motor with a cam on the end to actuate some microswitches to pulse the solenoids and push air through the primaries. It actually worked pretty well.

too bad i dismantled it and turned it into a air gun... but it'll shoot a 3" screw through 2 layers of drywall @ 120 PSI

[AaronGTR]

Well, then you get into the differences in composition and temperature of exhaust gas versus compressed air, and that might affect the results of the testing. What it probably comes down to is that it would be difficult to get test results that would accurately predict anything without actually testing them on a running engine in a dyno cell, or without a powerful computer running a CFD program. Something that basically only OEM's and high level race teams have the resources for.

But hey, that's just my hypothesis. I known far less about this stuff than some of you guys, and I'm learning a lot just by reading this thread.

[TGP37]

Well, then you get into the differences in composition and temperature of exhaust gas versus compressed air, and that might affect the results of the testing. What it probably comes down to is that it would be difficult to get test results that would accurately predict anything without actually testing them on a running engine in a dyno cell, or without a powerful computer running a CFD program. Something that basically only OEM's and high level race teams have the resources for.

But hey, that's just my hypothesis. I known far less about this stuff than some of you guys, and I'm learning a lot just by reading this thread.

I am learning a lot as well. This turbo I built was my first major project on a car/engine. I had dismantled moped engines at least several dozen times from 14-16 yrs old.

I think Superdave was onto something pulsing 120psi like that. It may not take every detail in consideration, like carbon deposits and unburnt fuel, but it seems it would still help find which setup had the best negative pressure wave travelling up the other two primaries.

I got a good ability to visualize things clearly in my head. I can see how a smoother transition would allow a stronger negative wave reversion, if that is what it is called.

I am seriously considering building a shop to built v6 header collectors. And to design the absolute best possible....I have the mental character and the love of squeezing every hp possible out of the engine. So designing a flow bench that accurately measures exhaust flow would be critical. I am thinking a single head (3 exhaust ports) and bolt the head to a feed of compressed air. Using MAP sensors at key points to measure pressure waves.

Maybe I could start a small company building these things. I would definately need a plasma cutter and a clean professional welder, not just a flux core. With lots of porting sand rolls. Handcrafted header collectors.....I like it.

[TGP37]

A few years ago i was testing some exhaust stuff for the 60v6 and built a pulse rig using 3 120VAC air solenoids hooked up to the air compressor. I was using a variable speed DC motor with a cam on the end to actuate some microswitches to pulse the solenoids and push air through the primaries. It actually worked pretty well.

too bad i dismantled it and turned it into a air gun... but it'll shoot a 3" screw through 2 layers of drywall @ 120 PSI

Ha ha, I see you like to build stuff. I am like that, if I can't find a tool or something I need, I make it. In example, exhaust flow metering 3x.

[AaronGTR]

hehe, yeah... that's the other thing I was thinking. "man, a plasma cutter sure would be nice for cutting curves in pipe like that!"

[TGP37]

hehe, yeah... that's the other thing I was thinking. "man, a plasma cutter sure would be nice for cutting curves in pipe like that!"

Oh yeah, I used a dremel to cut the steel pipe for my turbo manifold, then used a sand roll to make the ID just right. Merging 2 pipes smoothly is much easier then 3. It took some time but it was worth it.


The turbo manifold wasn't fully welded in these pics, but it shows how anal I am about the boundary where the two meet, I spent a while grinding smooth and polished it.

I determined few things before I started to build this engine up, optimize intake and exhaust as best as possible was one of them. The exhaust post-turbo is done and the entire length has zero interference along the way (within reason). The exhaust is wide open and the only resistance is the dual flowmaster 40's.

The y-pipe that splits the exhaust into dual (fake dual exhaust, lol) was also cut just right and equally divides the exhaust flow.