Engine and Transmission Tech

I'm seriously considering "merge collectors", for the 290" Huff GT 1 motor. I suspect there are gains over conventional 4 into 1, but I would like to see some science. Thanks, roverman.

I found this interesting:

[www.southernskies.net]

Carl, that's a pretty darn nice site you've found; perhaps if something similar could be found for the four stroke engine it would help us better understand the difference between the two.

It also reminds me how EFI intake runners are tuned to length in order to create a resonance/pulse bounce that helps push in the air charge into the cylinders. IMHO, this thread has been highly educational for lots of us; between Dan and Jim I'm starting to feel a little more confident in my exhaust design.

Jacques

Thanks Dan, I have made a note and will check at the local library next time I go there, perhaps they can get it. Sounds like an intriguing read.

Carl, that is a real find on the scavenger pipe. The author obviously knows his stuff. What I picked up on was the function of the reverse cone in ramming the escaped intake charge back into the cylinder, that was really the piece of the puzzle that I was missing, and very effectively and simply explains why the pipe is both extremely effective and very sharp in cut-off (peaky).

This makes it child's play to explain why the scavenger pipe would not be effective on a 4 stroke engine.

"This pulse is timed to reach the exhaust port after the transfer ports close, but before the exhaust port closes. The returning compression wave pushes the mixture drawn into the header by the negative pressure wave back into the cylinder, thus supercharging (a bigger charge than normal) the engine. The straight section of pipe between the two cones exists to ensure that the positive waves reaches the exhaust port at the correct time"

In this version of the 2-stroke the transfer ports take the place of the intake port. Obviously in a 4 stroke the intake does not close before the exhaust valve so this effect could only be used with a custom ground cam with an extra bump on the exhaust lobe. I don't know if anyone has ever tried that but it would be pretty impractical for a V8 due to the size of all the individual scavenger chambers, especially given the larger displacement of each cylinder. It might be made to work with something like a Briggs & Stratton racing class possibly. However, it still does not explain what is going on with the small diameter "stinger", (closed ended pipe, I got that part) and how that compares to the typical V8 system. It is clear enough that a pressure pulse after each firing ejects a quantity of exhaust from the stinger, and this is clear to see when the bike sits and idles and the oil begins to build up smoke in the exhaust. But I think there is more to this story, some of which we might make good use of. In particular, I don't think we've given enough attention to the temperature decreases from end to end, nor to the smoothing of high and low pressure pulses later in the system. If you look at pressure curves for gasses it's pretty clear that the needed volume for a given pressure increases substantially with temperature. In a typical exhaust system we're dropping *hundreds* of degrees in temperature from one end to the other, in most cases well over five hundred degrees. I think we need to have a look at exactly what that means to us in terms of system pressure, and therefore flow requirements.

JB

Why not the best of both worlds, The" on demand 2/4 cycle engine". This was/is, designed/tested by Ricardo Research Foundation. International Patents. Impressive performance and economy. roverman.



Edited 1 time(s). Last edit at 11/13/2010 10:20PM by roverman.

Jim, I observed the same thing off of Carl's link; this type of tuned scavenger muffler made a huge difference on our snowmobiles, however, it had to be built to the exact (specific) engine application.

I believe that this is what Dan was explaining all along, a well designed four cycle exhaust at its best can create the necessary low pressure trail (vacuum like) that finishes the cylinders charge filling. There's a muffler that I'm interested in that has its inlet designed to compliment the scavenging effect; surprisingly, it's only sold with the same seize outlet.

Jim If I understood you right, one comment that you've made is that whatever horsepower gain or lost resulted from the exhaust was no big compromised on a 400 H.P. engine, I agree, but, the only worthwhile result should be at the rear wheels, hence, the well built exhaust dictates that rear wheel result. I'm sure that it also helps the engine rev a lot quicker/straight up (torque?).

It's hell to know just enough to get yourself in design trouble but not enough to work yourself out of it without the help of knowledgeable comrades like you and Dan, and brother, I thank you for that help.

Cheers,

Jacques

This thread doesn't have enough photos, so I thought I'd stir the pot by posting photos of something I don't think has been talked about yet: anti-reversion devices for four-stroke headers. (Photos from an upcoming article...)





ref: [www.hytechexhaust.com]

Quote:

REVERSION is the secondary pressure wave that travels back up the primary pipes and enters into the cylinder on valve overlap. As this pressure wave travels back up the pipe, it brings with it all the residual gases still left in the pipe. This is what contaminates the fresh intake charge. Enter in stepped headers and ANTI-REVERSION chambers, placed at strategic locations in the primary pipes. These methods are employed to tune the arrival of the exhaust wave and to diminish the effects of the high pressure in the pipes. The results are higher volumetric efficiency and more power.

Thanks Curtis. I notice it uses a "merge collector". Can we now have some science, to go with these power inducing thingies ?roverman.

Unfortunately, I don't know much about the thought or testing behind HyTech's headers. I'm keen to hear any comments or insight here! (I've photographed two Formula Fords that use them. Interestingly, the two car owners both told me the headers came about a foot longer than shown above. They trimmed the tail end for practical/safety reasons.) Am I correct to think the primaries are paired oddly compared to regular tri-y headers?

Curtis, I think tri-y headers are refering to V8 applications ? "Bank" firing order, is usually different from L4's,(except for "Freds"miraculous, flat plane, 180 firing, 32 valve, RV 8) ! Go gettum Fred ! roverman.

Vizard covers anti-reversionary headers in "Performance with Economy".
His testing was on a V8 with bank separated 4-into-1 headers with the AR
cones built into the flange. The cones act as a diode, allowing flow
out of the exhaust port but not back in and are less sensitive to larger
primary sizes (though you should still pick the best diameter for your
operating range). His conclusion was that AR headers work well, providing
better fuel economy and power but a cross-over is mandatory. Also, the
carb will need to be retuned as the AR headers provide a stronger signal.
The extra power comes primarily in the low and mid-range if the primary
pipe diameter is held constant. One example cited was a 305 SBC powered
Malibu that went from 17 to 21 MPG with no other changes other than fitting
AR headers. Another example was an engine that wouldn't run smoothly below
4000 RPM due to reversion was able to run at 3000 RPM. Blackjack and Cyclone
were the holders of the patents but have since been bought out and, AKAIK,
the AR line was dropped. You may be able to buy pre-made flanges from a
company like Headers by Ed. The Japanese import crowd have been using AR
chambers in the primaries similar to the exhaust shown.

Dan Jones

Here's a link that explains exhaust almost to perfection:

[www.popularhotrodding.com]

Not much said there about the resonator cones. It occurred to me that I'm in a perfect position to add those to my system as I have to shorten the pipes about an inch due to added engine width. Somehow I keep removing simplicity. Now the question is the best method.

One point I'd like to add is that the recommendation in using the primary sizing chart was to concentrate on the "RPM range of interest" in determining primary size. Note that the closest thing to rpm given was the amount of flow, so if your primary range of interest is daily driving the flow rating will be a mere fraction of what it would be for maximum performance, and better to err in a downward direction. So then, if your heads flow 150-200 cfm but you only put your foot in it on rare occasions, maybe those 1-5/8" primaries aren't needed.

Another point, relating equal length primaries to race engines where a <3000 rpm band makes such unnecessary, in street driving, and especially spirited mountainous driving where the desired usable RPM range can be in excess of 4000 rpm easily, equal length primaries make a whole lot more sense.

YMMV

JB

Jim, are you expecting to have "reversionary flow" problems, with that bad-boy blower strapped on there ? Going to dyno this build ? Measure "pressure" in the exhaust ? If so, I suspect you'll find 1 3/8" primaries, overly restrictive ? Good Luck, roverman.

Doubt they'll be any more restrictive than the exhaust valves and ports. Remember Art, this is not a max HP project. Despite the blower, the goal is not ultimate top end performance, but all around driveability and good economy. Adequate power will be there. There is nothing to choke power down to levels below what I was producing with the 215, which was in the 300hp range and the headers were quite adequate for that application. If you'll recall, Vizard's recommendation was to err downwards in tube size, and I also had the opportunity when they were built to pick the brains of Ed Henneman ("Headers by ED"), one of the most knowledgeable exhaust system guru's on the planet at the time. (mid 80's) The tube size was optimized for a 300 cu.in. engine so yes, for a normally aspirated 350 at full tilt-n-boogie it'd probably be a little undersized and a few ponies might be lost at the top. But for a blower motor that is going to spend 99% of it's life under 4 grand? Even if it does see 7 on occasion I think these tubes are going to turn out just fine. Remember, I consider 300hp in an MG to be adequate and the goals of durability and low maintenance are equally important. I could put a cast iron manifold single exhaust on the car and meet that goal sitting in the shade, so how are those 1-3/8" primaries going to hurt?

JB

On the dyno question,.... maybe. I'm not sure what real purpose it'd serve. I'm not aware of a shop that has one around here so it's an investment in time and money that can be better spent other places.

> One point I'd like to add is that the recommendation in using the primary
> sizing chart was to concentrate on the "RPM range of interest" in determining
> primary size. Note that the closest thing to rpm given was the amount of flow,

The amount of flow is a function of the valve lift for a given cylinder
head, not an indicator of RPM. If you click on the graph, the primary
sizing curves are explained:

"Fig 4 This chart applies to normally aspirated engines. For street headers,
where low-speed torque is of prime importance (especially with a stock
converter and high rear end gears), use the lower line to select the
appropriate primary size. For hot street machines having reasonably big cams
and decent compression, use the middle line to size the primary. For race
engines, use the top line. If nitrous is involved, check out the nitrous header
section."

The lower curve is best applied to heavier vehicles with automatic
transmissions and/or tall final drive ratios and would be conservative on a
lightweight British car with a big American V8, a 5 speed, good cylinder
heads and a decent cam. AFR 165's flow nearly 190 CFM on the exhaust side
(184 CFM @ 0.500", 188 CFM @ 0.0600", 189 CFM @ 0.700"). Assuming a cam
with modest 0.5" lift on the exhaust side, even the lower curve calls for
1 5/8". Furthermore, a 5.0L HO with the stock E7TE heads only flows around
110 CFM at the same lift (grinding out the Thermactor bumps will help that
some) and Ford's stock tubular exhaust manifolds were sized at 1 1/2".

I'm working on a SBF with Floo Tek heads where I may use 1 1/2" primaries
for clearance reasons. The Floo Tek heads are cast in Australia, assembled
in Indy and are essentially copies of the FRPP GT40 turbo swirl heads. They
flow around 160 CFM @ 0.500" lift on the exhaust side. They will be used in
a 1971 Mustang convertible with 2.79:1 final drive ratio and an automatic
overdrive transmission with 2500 RPM stall speed. With an HO servo, the
transmission up shifts around 5200 RPM and there is only room for 2 1/4"
intermediate pipes through the convertible chassis bracing (though I'll
likely step up to 2 1/2" for the axle hop and Magnaflow mufflers). If the
same engine were in a lightweight sports car with 5 speed manual transmission,
I'd for sure use 1 5/8", especially with heads like the AFR 165 with their
excellent exhaust port. That assumes there is space permitting. On the
convertible Mustang, I'll be running simulations to compare primary diameters
and collectors to help me make up my mind but I think the 1 1/2" primaries
will be okay if I optimize the collectors.

> where the desired usable RPM range can be in excess of 4000 rpm easily,
> equal length primaries make a whole lot more sense.

The paragraph following that discussion explains that, even with equal lengths,
much of the potential benefit is lost in V8's that use a dual plane crankshaft.
Things would likely be different with a flat plane crankshaft. Also be aware
the presence of bends changes the apparent primary length. 180 degree headers
are much more sensitive to equal lengths and one builder of such headers I know
sizes the lengths acoustically so that each pipe tunes in on the same note.
You'll usually find better payoff by concentrating on the collectors. It's
relatively easy to change collector length and styles. One of the SBF drag
racers I know has had very good results with the merge style collectors on
evrything from a relatively stock fuel injected 5.0L to a high winding 289.

Dan Jones

It's also interesting to note what some of the SBB guys are doing with ported stock exhaust manifolds. Getting into the 12's with full sized cars running the 350 small block and cast iron manifolds, they claim about a 5-10 hp loss over tubular headers. That's 2nd hand, but those boys are fairly serious about their drag racing so even if it's padded a little it's still sort of impressive. The Buick manifolds are a pretty free flowing design.

JB

> It's also interesting to note what some of the SBB guys are doing with
> ported stock exhaust manifolds. Getting into the 12's with full sized
> cars running the 350 small block and cast iron manifolds, they claim
> about a 5-10 hp loss over tubular headers. That's 2nd hand, but those
> boys are fairly serious about their drag racing so even if it's padded
> a little it's still sort of impressive.

The biggest difference between exhaust manifolds and long tube headers
tends to come at lower RPM (see my dyno data at the bottom). Also, the
can be 20+ HP difference among different long tube designs. I'd wager
those Buick headers aren't optimized and/or there is a much larger
difference at lower RPM. The best of the SBF iron manioflds were the
K-code 289 hipo manifolds. A better alternative would be the 5.0L HO
Mustang tubular steel "shorty" manifolds. In addition to the OEM, there
are a wide variety of aftermarket versions available in 1 1/2", 1 5/8",
1 3/4" and 1 1/2" to 1 5/8" primary diameters, equal and unequal lengths
in mild or stainless steel. Long tube headers still out-perform them but
they are useful where stock location catalytic converters must be retained
for emissions purposes or where long tubes won't fit.

Here's a magaxzine test of stock, K-code and long tube headers on a
dynojet. Note it's a fairly mild 302 engine, through the mufflers and
the power is at the rear wheels.

[www.mustangandfords.com]

There's another test which recorded a much bigger differnce on a 347 SBF
stroker:

[www.mustangandfords.com]

They claim 104.5 hp and 96.6 lb-ft of torque difference between Hooker
Super Comp headers and K-code manifolds. though it appears they also
installed a K&N air filter at the same time.

> The Buick manifolds are a pretty free flowing design.

The Ford 351C-4V engine has larger valves (OEM size is 2.19" diameter
intake and 1.71" exhaust) and their iron exhaust manifolds have
ports that are much larger ports than a Buick 350. Here are
the results from dyno testing I conducted on a fairly mild
Ford 351C. At 4200 RPM, where we started the pull, the iron
manifolds are down 59.1 HP and 72.6 ft-lbs of torque. It
would be even worse down at 2500 RPM. This is because of the
short length of exhaust manifolds. The long tubes are able to
tune in at a lower RPM. The gap narrows at higher RPM but is
still 26.5 HP at 6100 RPM. Note that the long tube headers
are no where nearly optimized for this combination. Still,
if you plot this data, you'll see the difference in area under
the curve is huge (more than you'd likely get from an expensive
set of aftermarket heads or port job). Couple this with the
50 HP difference in mufflers we saw and you can turn a 400 HP
engine into a 300 HP with a restrictive exhaust.

Ford 351C-4V Dyno Project Engine
hydraulic roller street cam
flat top pistons
unported iron 351C-4V heads
12 inch extension and Magnaflow mufflers

Iron 351C-4V Exhaust Manifolds Hooker 1 3/4" primary headers
RPM HP Torque (ft-lb) RPM HP Torque (ft-lb)
4200 259.7 324.9 4200 318.8 397.5
4300 274.3 334.7 4300 325.0 397.7
4400 284.8 340.1 4400 331.6 395.9
4500 294.8 344.0 4500 339.4 396.1
4600 305.4 348.7 4600 349.9 399.2
4700 315.1 352.2 4700 360.6 402.9
4800 330.1 360.6 4800 368.0 402.8
4900 343.2 368.8 4900 372.4 399.5
5000 345.5 363.1 5000 378.0 397.0
5100 354.0 364.2 5100 388.6 399.7
5200 364.8 368.7 5200 397.1 401.1
5300 368.6 365.4 5300 401.3 398.2
5400 377.5 367.0 5400 404.9 393.8
5500 382.5 365.5 5500 409.9 390.9
5600 383.0 359.2 5600 412.1 386.7
5700 384.7 354.4 5700 407.9 375.8
5800 383.8 347.0 5800 401.4 363.8
5900 376.6 335.6 5900 398.9 355.1
6000 369.3 323.2 6000 401.3 351.2
6100 380.5 328.4 6100 407.0 350.2

Dan Jones

For the record, are we discussing ID. or OD. on tubing sizes ? Any solid data on merge collectors ? Stepped primary headers, properly built, function similar to anti reversion cones? Port mismatch at port can be "a" on flooor, "b" on sides, and "c"never on roof ?. Why not use oval tubes to fit available spaces ? Is it not the "area" inside the tube and bend radii ?Thanks, roverman.



Edited 1 time(s). Last edit at 11/18/2010 01:33PM by roverman.

Good info Dan, about the hp/rpm losses and the area under the curve. I have long been a believer in chasing that volume vs. absolute hp peak so to me this is very revealing (and to be honest, 5-10 hp loss sounded like pie-in-the-sky). I'm also a firm believer in long tube headers for the street. Art, I can't speak to others' comments but my measurements were inside diameter.

About the anti-reversion cones, I looked at some test results (Harley) and the graphs I saw showed no gain and an upper end loss due to restriction. These were slip-in cones that are simply inserted into the end of the header pipe. Starting to sound like snake oil, and I'm suspecting that any advantage is going to be in a very specific application under ideal conditions (racing perhaps). Without more than that they would hardly be worth the trouble. But having said that, a Harley exhaust is hardly an ideal tuned system. Also it seems they were primarily interested in eliminating a dead spot in the power curve where reversion was thought to be interfering and causing a power loss. All of this leads me to suspect that a properly designed set of headers is not going to have much of an issue with reversion and therefore the cones will be of little benefit.

JB