I know we have the 'calculator' for the normal 4-1 headers to get the relationship right between tube length, diameter, engine size, and RPM, but how would I do this with a Tri-Y setup? Also, what tube size would you guys go with. I was thinking maybe the normal 1 3/4" primaries, then 2" secondaries, and into a 3" collector. Am I close? I wanted to start foolin around with my own header design cause it looks like fun. Any input would be great

 

You could try calculating the cross sectional area of the primaries, secondaries and collectors. That'll give you some info to work with.

For instance, the CIA tri-y headers have a primary that's about 80% of the collector cross section, with the secondaries in the middle, and those headers seem to work very well.

Just a thought for a starting point.

 

I like your ideas about trying 4-2-1's. They are used on some high-end engines, but were first used maybe 80 years or so ago on 4-cylinder engines. Still are on some ricers, and even production engines. Vortec 4200 I6 uses a short 6-2-1 cast header.

It's probably more difficult to determine the "best" diameters and lengths for the primaries, maybe steps, secondaries and collectors than it is for a 4 into 1. A good start is the book Scientific Design of Exhaust & Intake Systems; Third Edition by Philip H. Smith and John C. Morrison. It's available online or sometimes at Barnes & Noble. There were a few threads that discussed it here in Advanced tech.

Like with any engine, the header and exhaust design has to compliment the rest of the engine design. A Nextel Cup Daytona/Talladega "plate" engine which runs in a couple hundred rpm band has pipe sizes and lenghts somewhat different from a Bristol or road course engine. Some Cup teams have software which can help determine header configuration, but there is a lot of "cut and dyno" header work done. Of course they are looking for those last few horses.

That being said, even Cup cars have to make compromises on lengths and bend radii to get them in the car. The header fabricators stay very busy making those Inconel headers which probably cost nearly 6 grand a pair.

IMO, you can come close with diameters and lengths using Smith's book and some of the better engine simulators. I'd use primaries sized for the power output of the engine; 1-3/4 sounds good until you get into the high 500s (fw) or low 600s where 1-7/8 might be better. Again, secondary size and collector size is important as is total system length. 2 or 2-1/4 secondaries and 3 in collectors are probably close. Something like 2-3/4 might be better, but you pretty much have to stay with what's available in mandrel bends.

Fabricating a good 4-2-1 which fits in a real car is a big challenge. If you hire it done in stainless, $2000-2500 seems to be the cost. I think I'd build my "development" set from plain steel because of all the recut and refitting and scrap I'd end up making. If they worked well, I'd either coat them or duplicate them in stainless.

I think a well designed and built 4-2-1 for LS-1 F-bodies in the 400++ rwhp range would be a useful commercial venture, but I don't think enough people would pay nearly twice what the high-volume 4 into 1 headers cost today to make it viable for the major header manufacturers. Even Thorley only has a limited number of Tri-Ys and none with the right pipe sizes for strong small blocks.

I'm not trying to discourage you; just the opposite. Go to it, but read the book first.

My $.02

 

Thanks for the great replys. Deffinently stuff I'll keep in my head, and I will be checking that book out very soon. I was also wondering how much a decent madrel bender would run me. I'd be looking at somethin kinda small to fit in a packed garage and somewhat cheap. Thanks again

 

Mandrel benders are NOT cheap. Every size tube requires a different mandrel as well as outside "shoes" like a crush bender for each bend radius.

There are many pre-bent steel and stainless mandrel bends available in various tube diameters, bend radii and wall thickness if you look around. Most folks building a custom system piece together sections of bends and straight. Even F1 engines usually go this way. They don't bend an entire tube. That's more what happens when you are doing a production run of headers for volume sales.

Good luck.

 

i was told mandrel bending machines run in the $50k area.......is that true?

 

Can be much more. Of course it depends on the feature set. I have seen ones that cost over $250K. I don't know how much a basic bender would cost. The basic OEM stuff costs over 100K to start. Plus tools.

 

Here is a Mandrel bender for around $3k. Tube Shark

 

i dont think thats mandrel.

also, the bends dont look like they would be tight enough for things like header primaries.

 

After doing some reading in the book mentioned by OldSStroker, which I also recommend getting, I would recommend against a 4-2-1 header. They don't really do what they are suppose to do. If you can fit a 4-1 header in there, I would recommend it.

 

There are a few Cup teams that might disagree, along with some Pro drag racers and high-end road racers. What leads you to believe 4-2-1s don't work?

 

mastdrver may want to check out this article on Jack Burns

http://www.popularhotrodding.com/tech/0310phr_burns/

Some good info on tri-y's, who's using them, why they're using them and how they work. Not bad for a magazine article and one of the reasons my tight-butt forked out a few bucks for a subscription to Pop Hotrodding.

-Mindgame

 

Thanks. I couldn't find that.

Scott Parkhurst is an interesting guy. Bret and I have spent some time with him at various gatherings. He has a lot of technical knowledge, but isn't afraid to get folks like Jack Burns to contribute their expertise. IOW, he doesn't let his ego get in the way. Scott is a big reason we subscribe to PHR.

Of course he does have a thing for Buicks, and not necessarily the turbo GN ones, but I don't hold that against him. BTW, he's in California, of course, but is originally from Buffalo NY area, I believe. I think you two would get along, 'Game.

 

Yes... California. Can't help but find it ironic that so many gearheads live in Cali.

I would like to meet Parkhurst and thank him for raising the bar for hot rod magazines. Not to mention, he has McFarland write an article time and again.
I have to put that on my list of things to do before I fertilize the good earth. Well, that and attending PRI.

Will actually make it out to LV for SEMA 2005, something I have wanted to do for a long time.

-Mindgame

 

http://www.popularhotrodding.com/tech/0310phr_burns/

great article !


with the Tri-Y system, you are trying to keep the exhaust gas velocity and scavenging (negative wave) higher, but the primary pipe lengths become more critical..you only have a few inches to play with. You want a negative wave at the "Y" ..if you get the primary pipe length wrong, you won't see full benefit of Tri-Y.

the Tri-Y design can scavenge more...so you might need lobe centers spread and more towards 65 Exh/Int Flow Ratios like ProStock Cars, when take to extreme.

Tri-Y design can use smaller OD pipe diameters, and engines with very high CR (will have lower EGTs ) , along with RaceCars that run a lot of MPH (header pipe air cooling effects)...will like Tri-Y headers that have smaller pipe ODs than normal 4-1 or 4-2-1