It looks like 3 of us, at least, admit to reading Car Craft.



No need to buy the mag for the article, just point and click.

 

I'll try to explain what I mean...

You have 4 primaries coming off the heads...

Right where shorty headers usually end, you merge the pipes so you get two fatter pipes and then where long tubes end you merge those together into one collector.

Should give you a broader powerband no?

Kind of like doing step headers.

 

Grip, thanks for the explanation.

Don't know if that'd be worth anything or not. Seems that you'd get a weaker scavenging wave followed by a stronger one compared to a long-tube header which creates a single strong wave at some rpm dependent on the tube diameter and length.
Racers usually just stick to long tubes and use the collector length or different types of collectors to tune. Seems to be a much more simplistic way of doing things without adding other variables to the process. You can actually pick up quite a bit of torque in a specific rpm range with collector tuning. Not difficult to test on a dyno either because many race engine shops now use header systems built to exchange collectors quick and easy like. My engine builder uses a system like this, which takes a bit of guesswork out of having a custom header built. You decide your collector length and whether to use standard or merge type right there on the dyno. Of course in this case, we're talking time, and that's always money... but why guess.

WRT building headers,
Like I said before, the procedure I outlined is one I've seen a friend of mine, Gary Mesina, use to build his headers. He builds headers for alot of guys here locally.
Anyways, after posting my 1st response I talked to GM, trying to get a bit of info that might help blackws6 out more with his project. I sent black an e-mail on this earlier.

Gary uses a method described to him by "Headers By Ed" for laying out a set of headers... the same one I posted here. So I wouldn't be surprised to see other people using this method, HBE sells alot of header kits, flanges and the like so I'm sure they sell alot of their booklets too.
As for magazines.... bout the only car magazine I read is National Dragster, Popular Hotrodding and Super Rod. As far as "tech" is concerned, I pretty much ignore most of what the magazines have to say. Seems they are just out to sell products these days with bogus claims. I can look through a Summit sales magazine and read the "Pick up 30 hp and 40 lbs-ft of torque with the new "Bernoulli-was-Wrong" air charger! nonsense.
A bit of over-exageration on my part... but that's the way it seems these days. I do like reading guys like McFarland, Vizard, Bohacz , Scraba, Morrison, and of course Johnson. Those guys know their stuff and are much less concerned with selling a product.... they aim to educate. Otherwise I just look at the pictures... I buy Playboy for the articles.

-Mindgame

 

I'll try to explain what I mean...

Cylinder pairing, as well as primary lengths are very important in 4-2-1 (Tri-Y) tuning. See our previous posts and links.

Cylinder pairing: 5 and 7 can't be paired, but 6 and 8 are. Typical shorties have less than equal primary lengths, and from my observation, aren't in the right place for the 2-1 merging.

All of that aside, I see your point. In practice it's probably easier to fab the whole thing, or just cut the shorties where you need to, and go from there. At the price of shorties, that might be a good place to start if the primaries are big enough!

You can of course have stepped primaries like the Dynatechs in my previous link, as well as the 4-2-1 configuration. It would take some heavy duty simulation to predict what lengths and sizes to use, or some dyno time. I think there's enough here to at least work on the simulation end.

A typical street engine may need torque from 2500 to 6500 or almost 75% of the engine's total operating range of 1000-6500.
Daytona Cup engines use about 5% of the total operating range. The 4-2-1 configuration may be able to offer more average torque over the 75% operating range than does a 4 into 1.

BTW, is anyone familiar with the '63 Pontiac Super Duty cast (iron or aluminum!) headers? 4-2-1 with 3 in. or maybe 3-1/2 outlets.

Also, if I remember correctly, Smokey may have been one of the first to use 4 into 1 headers for high speed ovals like Daytona where race rpm range is very small.

I'm probably an Army of One on this 4-2-1 thing for high performance street cars, but that doesn't bother me.

My highly-opinionated $.02

 

Well, not to turn this into a Tri-Y debate but there have been quite a number of dyno tests performed over the years. Some showing mid range torque gains for the Tri and some showing no gains at all.

Most companies that build headers for competition will tell you that the Tri's hardly ever show a gain over 4-1's on a dyno but they do sometimes go faster on the track. Point... they have to be correctly designed and there are a lot more variables in the process.... the primary diameter and length, the "Y" diameter and length and the collector size/ convergence-divergence. In my opinion, that's a lot of stuff to get right. By the way, I realize this is a "street" car discussion but WJ and alot of the Pro Stock crowd are using or have used Tris over the years. Of course they have unlimited resources but it might seem to suggest that there is some power there for those with a simulation package robust enough to take advantage of it. No Desktop Dyno here.... more like "Wave".

On another note, I used a set of Doug Thorley's back in the early eighties on an extremely consistent Vega bracket car and lost 2 tenths at the track... no mph gains and even tried some tuning to maybe counter whatever the headers might have caused. Never got anywhere with that so I finally gave up and went back to my Stahls. Picked up the 2 tenths I had lost that next time out. I'll try anything once.
So I'd really need to see more positive results before I'd cast a verdict one way or the other. I know I can use a 4 into 1 and tune out a bit of power with nothing more than an adjustable collector... so I tend to stick with that.

As for SY. I remember Yunick building a 4 into 1 that paired the primaries up in one big pipe (looked like shorties but weren't) but those were just 4 into 1's with a different packaging. Never saw a Tri-Y from him though.

-Mindgame

 

I was at a buddies shop today and hanging on the wall was a set of tri-y headers for a 18 deg headed short track car. I don't doubt that the Tri-Y arangement is going to work, but like any other header combo you need to test it. A tri-Y is tricky like a stepped or merge collector, and it needs to be custom taylored too a engine just like a intake manifold.

Only bad thing I have seen with Thorley's is that they are small diameter pipes and the tuning lengths are not very long. The dynatech ones are completely different.

Bret

 

what people are missing is that collector length, size, configuration and tuning has a much bigger impact on scavenging and what power band it happens in then anything you do to the primaries (within reason, you could always make them so small or large that nothing works). The Tri-Y design allows for more variables (and sometimes fixes packaging issues) in building a properly tuned header, but if you don’t know what you’re doing you’re as likely to hurt your combination as help it.

 

Boy, ain't that the truth!

This wise statement applies to way more than header design.

 

The issue of collector tuning has been covered. Like I said before, visit a good race engine builders shop and they probably have a set of dyno headers with "exchangable" collectors. I can name three off the top of my head who I know do this.

I'm still interested in the effort of coming up with a Tri-Y design that'll work. Considering the variables, how would you go about making tests on a dyno unless you had 30 or so Tri-Y configs sitting around? Trying different length 4-into-2 primary lengths/diameters, trying different 2-into-1 primaries (secondaries).... trying different collector lengths/volumes, tapers.... lots of testing for sure.

Speaking of testing street engines with good mid range torque..... what about the Engine Masters challenge? How many of the top 10 finishers were using Tri-Y's? Seems to me that they would have been ideal for that challenge yet I don't remember seeing any of the top finishers use them.
Bret, you guys must have done alot of dyno testing to decide on an optimal header design..... I'm curious as to your findings?

-Mindgame

 

Using a simulation program which incorporates header configuration effects is the first step, just as it is in determining inlet length and configuration and cam timing. Of course simulation isn't perfect but it gives trends toward the desired end. When you get close on the simulation, build a few, or run a few cams. There was no way Bret could have run 30 or 40 cam and inlet manifold combinations on the dyno in the time available, but the many hundreds of simulations narrowed the field considerably.

In the Engine Masters competition, exhaust rules are about the most restrictive: You must use a commercially available header with a real part number with no modifications, which fits a '55 or newer US manufacturer's car inside the frame rails. No truck headers, no self-fabricated headers, etc. You can use slip on collectors, however, if they are commercially available. Also no crossovers of any kind, and two 3 inch max ID street mufflers must be used. That's the same for 2003's 470 cube big blocks as 2002's 366 cube small blocks.

As for tri-y headers commercially available with the tubing sizes to support 600 hp (small block) and perhaps 800 big block hp....nada, as far as I can find, so it's not a viable option for EM. The best Thorleys for SBC have 1-5/8 primaries and 2-1/2 collectors. That's maybe 400 hp on a good day, IMO. If you know of 2 or 2-1/4 primary, 3-1/2 or 4 collector BBC tri-ys for street cars, please send a PM.

Too bad, because they might be just what the engine likes. You are familiar with the sag in the torque curve somewhere around 3500 or so on NA V8s producing over 90 lb-ft / liter and about 100 hp/ L, right? Many drag or track engines in that hp/l range never drop below 4000, and many are never dyno'd below that. That's one of the reasons the EM test is 2500-6500, and the score is based on the average torque and hp in that range.

It's sort of a catch 22; commercial headers big enough to support the big numbers are rarely used on engines that run WOT below 4 grand, and the tri-y's are for low-po engines. Bottom line is our 4-2-1 header quest has to be for other than Bret's EM work.

More Tri-y stuff: Watching the NASCAR Cup race replay from Martinsville VA from 4/13, especially the undercar brake cam on Bobby Labonte's car, I think what I saw was part of a tri-y right hand header. Not completely positive, but almost. It goes along with my thoughts of better average torque over a broad range: telemetry showed a 4500-9000 rpm range. 4500 rpm range is right in there for a street car. Engine Masters is 4000 rpm range.

As for Smokey (earlier post), he claimed to have shown up at a NASCAR track in 1960 with a Pontiac sporting the first 4 into 1 headers on a stock car...and tech took them away from him.

He also built an 8 into 1 bundle of snakes he called 'Equal Pulse' because there was strong negative pressure pulse in the secondary every 90 degrees. Packaging was a bloody nightmare, and I don't know if he ever ran it on a racecar.

As to findings, let's say we'll discuss trends, but not necessarily hardware and data points.

 

Wow thats kinda funny. Is Ed very popular in the custom header bussiness? I have met the guy before. When I worked in retail I met him. He was wearing a jacket that said "Custom Headers By Ed" or something like that. So I got to talking to him(for some reason I just had the feeling he was the ed refered to on the jacket) and he told me about his custom header bussiness and gave me a card and whatnot. I always thought he was just a local bussiness men never thought he was an big guy in the bussiness.

 

As Old SStroker said, the rules limit that. The problem with the Engine Masters is the rules IMHO. Things like valve angle, cast manifolds (well maybe) and the like are good. Going with more rocker arm ratio, and custom fabbed headers limit alot of what is going on out there. At one point the contest will influence street engines alot and the parts made for them.

My problem is what Jon stated. Since we can't fab up a header for the contest it's something that is not going to ever be seen and applied as the ultimate technology for the street. That's why it bugs me. We really need more solid roller lobes for street cars, but PHR doesn't want solids because street guys don't use them. Why because of cost? because of adjustment? B.S. cost!!!! my god we are building $20,000+ "street" motors. Who the $%& gives about cost at that level, when in the end all it does is add cost trying to develop and run the alternative. If you ever think about the things you have to do just to make up for the fact we can't use rollers it's stupid how much it doesn't save us. BTW a roller will always increase low speed TQ because it wil require less duration to get the same lobe area. Anyways back on point. The Tri-Y's that Dynatech has we're not even legal for the contest, because they were not for a passenger car. O.k. but $2500+ CNC ported heads are? $400 Stud mounted rockers are? $1400 worth of coatings are? extremely low tension oil rings are? 1.777" rod journals are? We have to pull out racing tricks to win, and at one point they will all be outlawed. That brings about the sad fact that the "racing only" parts never make it to the street, when I apply them to it everyday.

Enough of the bad things, I wouldn't want to get into that contest if I didn't like the scope of it. Those are the rules, and they are fair across the board for all of us. The contest does make sense though because it does bring a competitive metality to solving the problems for running in a low to high rpm band, just like street cars do. I applaud the contest for that. These engines are not street race engines, but are extremely strong engines for cruisers and street rods. It's hard not to like a 1000rpm idle, instant throttle response and 600hp. All in a package you could drive to work daily in a nice climate. Do I wish 2 things were different? Yeah, because it would evolve the technology on the street, which in the end is part of the goal of the contest.

As for what I did. I needed a 30" long 1 7/8 header for my setup. I found them and plugged them on. The short duration and high lift means that the pulse is very short in time but strong in pressure so a big header was needed, if you choke that off at all it ill kill the power after peak., when you need a 500rpm peak then the header selection should be a little bit big. I usually like a big header anyways. I wish there was time to test short and longer ones and bigger and steped headers, but we needed to dial in the carb, check timing advance curves, cam advance, cam selection, intake manifolds etc....

BTW did you look at the primary lengths on those headers? A Thorley Tri-Y is very short say 10-14", but the Dyntechs are a good 26-30" long in primary length. You are looking at a double tuning pulse optimization here, that's why they have such a wide power band. (tuning length is where they first start to pair up, not when all 4 to a side come together. The Honda crowd has always had Tri -Y's and in a bolt on dyno test a Tri-Y made the most power. (I think it was the Mugen one, if any of you Honda boys are interested)

ADD ON TO THIS POST: Don't take any of this as a EM knock. The contest is very hard to enforce and alot of these rules are there to do just that. They do a really good job considering this contest is so young.


Bret

 

Most guys these days don't want to build 400 crank hp street engines. I'd say the trend is more towards 500-600 hp motors.

So whether a Tri-Y is still a realistic answer... I don't know. Not saying they're not..... just would like to see more results.

Speaking of Ed, he's been around for a looong time actually. Not as well known as Jere Stahl but he builds some really high quality stuff. No junk and not exactly cheap.... but you get what you pay for. Ed is not a big Tri-Y fan either, nor is Jere. Neither of them build tri's. Not that they couldn't.... they just choose not to. Ed also has some pretty strong ideas on headers..... what works and what doesn't. He likes equal length... which I haven't seen much gains in but he claims the gains are there. Take what you want from that. Everyone has an opinion... I can only believe what I can see though.
Seems Chevy HiPerformance did a test on headers. Ed mentions it at his site. Seems his headers did really well in that lineup of other 4 into 1's and Tri-y's. Not gonna pay for the article but I'll see what I can do in "digging" it up.
Don't ask don't tell.

-Mindgame

 

Hey, at least we're talking about tri-y's(aka 4-2-1). Of course folks are looking at 500-600 hp vs 400hp street engines, but are thay looking at torque/hp thoughout the operating range of the engine? IMO, that's what makes it go.

We might even change Mindgame's tune:
MG quote 1:

Well, not to turn this into a Tri-Y debate but there have been quite a number of dyno tests performed over the years. Some showing mid range torque gains for the Tri and some showing no gains at all.

MG quote 2:

I'm still interested in the effort of coming up with a Tri-Y design that'll work.

MG quote 3:

So whether a Tri-Y is still a realistic answer... I don't know. Not saying they're not..... just would like to see more results.

So would I!

IMO, there's a good chance race-style 4-2-1s are the way to get more average torque (and hp) in the operating range than 4 into 1s. Maybe someone who's actualy involved in high-end racing where this is being used will chime in. More than likely not.

I wish we had the resources ($ in gov-speak) to investigate this more thoroughly. I think that if we heard that the majority of short-track Cup, Busch and Truck winners were running 4-2-1s, it would become the "trick of the week" for high hp street engines, and header builders would seize the opportunity.

Magazine shootouts for things like headers tend to use what's commercially available which is perfectly understable. That doesn't mean they get what's optimum.

Hey, a little out of the box thinking gets the creative juices flowing. That's why this forum is so interesting, even with all of us prima donnas (or should that be prima dons?).

My $.02

 

Absolutely no arguements there Jon. We have to define the operating range first and of course.... everyone's def. will vary.

If for instance, I'm launching a modified car like an LT1 (something on the level that most guys here are running) at 3000-3500 rpm (that's low.. especially with a good rear), shifting at 6800-7k. Then I'm only worried about the torque curve in that rev range... more importantly at the shift points. Anything under that doesn't mean much to me. So, if a mod picked up say an average of 10 lbs-ft under 3500 rpm then it aint gonna mean much at the track where you're rpm shifts aren't even touching in that region of the curve. Of course, we're considering the fact that in this example.... it's not really contributing to the average torque in the revs we spend the most time in. Important.

Build the most avg torque at the shift points and you'll always go faster down the track. A 6-speed car for all intents and purposes can be much "peakier" than a 3-spd auto and still be plenty quick. We learned this lesson some years ago with powerglides. So, a dyno queen that makes killer average torque numbers from 2500 to 6500 rpm may not be as fast as one that makes more average just from 4000-6500 behind a car geared to work there. That seems to be a big confusion for some people. But the racers have that figures out.

You see this alot at the tracks.... guys spend so much time and money worrying about low rpm torque because the mags tell them that's what matters... then they gear the car all wrong, then wonder why someone with less torque but more hp at higher rpm (all things considered equal) is so much quicker. That, and they wonder why they can't keep the tires under the car... torque and suspension.
Torque is the great double-edged sword. Me, I'll take mine at a higher rpm and see you at the end of the track first.

Back to the street and T56'ed LT1's.......
On the street and cruising at 2500 rpm or so.... you only need a rev synched downshift to get the thing up to 3000-3500 rpm or more..... autos are that much easier. The torque made below that point just became insignificant.
So, we all agree that where the engine operates is important. The fastest cars at the track are the ones extending the revs as high as possible and gearing to work in that rev range. Different for the street but not so much different. If your shifts are taking you to no lower than say 4500 rpm then you're wasting time trying to build torque below that point if that torque doesn't play a significant part in the average in our "sweet spot"....... in this case 4500 to the shift point.

Lots of different ideas on which is best for sure.... but these are my thoughts on what works for me and so many other racers. I follow this philosophy for street cars too.

Quote:

We might even change Mindgame's tune:
MG quote 1:

Well, not to turn this into a Tri-Y debate but there have been quite a number of dyno tests performed over the years. Some showing mid range torque gains for the Tri and some showing no gains at all.

MG quote 2:

I'm still interested in the effort of coming up with a Tri-Y design that'll work.

MG quote 3:

So whether a Tri-Y is still a realistic answer... I don't know. Not saying they're not..... just would like to see more results.

So would I!

IMO, there's a good chance race-style 4-2-1s are the way to get more average torque (and hp) in the operating range than 4 into 1s. Maybe someone who's actualy involved in high-end racing where this is being used will chime in. More than likely not.

I wish we had the resources ($ in gov-speak) to investigate this more thoroughly. I think that if we heard that the majority of short-track Cup, Busch and Truck winners were running 4-2-1s, it would become the "trick of the week" for high hp street engines, and header builders would seize the opportunity.

Magazine shootouts for things like headers tend to use what's commercially available which is perfectly understable. That doesn't mean they get what's optimum.

Hey, a little out of the box thinking gets the creative juices flowing. That's why this forum is so interesting, even with all of us prima donnas (or should that be prima dons?).

My $.02

LOL,
I'm a bit of an enigma. If there's anything anyone should know about me, besides the fact that I'm bull-headed (which should be obvious), is that I never discount any idea. I just have a very hard time accepting new ones.... I'm a show-me kinda guy.
I really liked the phrase Tom Cruise used in "A Few Good Men" when he told Demi that as a lawyer it didn't matter what he "Thought".... only what he could "prove".
In much the same way.... I tend to believe what I've actually seen and know to work but I don't necessarily throw the "other" in the trash can. I just place it on my shelf-of-the-unanswered for later review. That's a rather large shelf by the way.

Ed doesn't keep that article on his server. I asked politely.
I would like to see those results though...... there are other ways.

This has been a good discussion.

Here are some links to Tri-Y's that are obviously being used on the high end.... Pro Stock, Comp Elim, etc.. Very nice work too.

http://performanceweldingheaders.com/triy.htm
http://www.burnsstainless.com/Xdesign/xdesign.html

Ahhh, if we all just had the time and money to experiment. Then again, I've had my fill of the ultra competetive stuff. The street is mucho friendlier.

-Mindgame