Engine and Transmission Tech

Any experienced exhaust Guru out there?

I have been reading, and working (above my head) with all sort of calculations, to formulate the perfect exhaust system for my MGB powered by a Ford small block (331) stroker. Having said that, the reality is that having the wrong mufflers or pipe size can defeat the perfect induction system.

Combustion chamber filling doesn’t only depend on the intake side for maximizing fresh intake mixture volume; in fact, the low pressure developed by a strong scavenging exhaust system is the main contributor to pulling in the next batch of fresh oxygen rich air charge.

Some racers find it more difficult to fine tune the exhaust extraction to maintain a lengthy low pressure scavenging trail than it is to formulate the induction system. BTW, maintaining back pressure turns out to be a myth, it has nothing to do with fine tuning exhaust formulas.

So far, I’m thinking that the section immediately following my headers (through fender well) is the critical area to good scavenging. The problem is that on the MGB, there’s a sharp turn following the collectors, so I’m using as long a radius as possible with a 2-1/2” pipe (all the way to the back)

I’m trying to generate ideas from members that have been there (FSB) and have some trial and error experience with the MGB short wheel base and the exhaust. How about an “X” pipe? Its benefit are credible, however, there’s not much room to accommodate it; what mufflers can excel on flow attributes while maintaining a civil rumble?

I know that some of you may think that I’m over thinking this one, but, after seeing some dyno numbers based on exhaust change (gain or lost), it got me convinced that your engine’s performance (and MPG) relies on a strong partnership between the intake and exhaust systems. Please let me know your thoughts on it.

Jacques

Quote:

I know that some of you may think that I’m over thinking this one

Jacques,

There's no such thing as over thinking. In the end, you may wind up going the same tried and true route as others, but the time spent learning is NOT time misspent.

Jacques, I think you are right about what you said, but have left out one critical element, namely the camshaft. It has a radical effect on both intake and exhaust and the systems that are ideal for one cam will be all wrong for another. So the very first thing that you have to do is to define the performance parameters of the car and determine the rpm range that the car will spend most of it's time in, then also how wide that band is and how "peaky". For instance a highway cruiser might spend 90% of the time (or even more) in a very narrow band centered around or somewhat below 2000 rpm. That use band would be extremely peaky. A car driven only in the mountains might be constantly going between 1500 rpm and 5000 rpm so it's use band would be much flatter. Most are somewhere in between.

Once that is determined a suitable cam for the car can be selected with lift, duration, and overlap that are calculated to enhance the desired operating conditions with emphasis on performance or economy or driveability, or torque within the use band, etc. Only then do you really have enough information to properly design the induction and exhaust system.

Where in the induction system much attention is paid to mass flow, in the exhaust it helps greatly to think in terms of pulsed flow. All extractor type exhaust systems, from a dirt bike to the most complex headers rely on this dynamic, primarily because the exhaust pulse is vented from a pressurized chamber. So how you handle this slug of exhaust gas will determine what happens when the intake valve opens, as well as how much the conditions will vary with engine speed.

Back to the dirt bike, these engines are renowned for a peaky powerband, primarily because the pipe is tuned to provide the greatest possible suction volume at the exhaust port at a particular rpm, known as being "on the pipe" and they create tremendous power at that engine speed, but can be absolute dogs below that.This would be bad in a street driven car. By contrast, the old and poorly appreciated cast iron log manifold, precisely because it is incapable of being tuned to a precise range, gives a very flat performance curve which, it turns out is just what you usually want in an average passenger car. The curve is low and about all you can do to raise it is increase the size of the passages, remove obstructions, and smooth corners to increase flow but you are making almost no use of scavenging whatsoever, with the result that the engine will produce a flatter torque band. For something like, oh I don't know, maybe a really big engine in a really small car like the MG-Roadmaster that was going to be driven in the mountains a lot, this would be an excellent idea. But unless you really do have excess displacement most people will opt for a more performance oriented solution. But note, "performance" in this sense almost always equals a more peaky powerband in some degree. It's just one of those compromises we always have to make. Note that the type of exhaust system used will have a big impact on how much overlap is used in the cam, where a very peaky exhaust can be moderated somewhat by less overlap, etc.

So leaving header/manifold design for later, you next have a series of pulses moving towards the muffler. Usually they are not evenly spaced, and can be thought of as areas of greater density separated by areas of lesser density, some closer together and some farther apart and getting larger or smaller based on engine speed and load. By the time they have passed through the muffler they have been partially merged as well as cooled a bit which is why it is common to see a smaller tailpipe than the header pipe. Muffler design can be complex but basically the size of the box determines the frequency band of attenuation, which is why a large under body muffler and a smaller resonator is a popular combination. The crossover pipe seeks to even out the exhaust pulses, with the X pipe being the most effective means of doing so. However it does require equal length header primary tubes for maximum effectiveness. Bear in mind, the firing sequence causes problems with these pulses which is why the crossover is desirable at all and if you look carefully at the firing sequence you will see why this is so.

Next, pipe size is directly related to engine size and operating speed. The modern trend is towards big pipes but it is clear that this is a trend which favors high speed, heavily loaded applications. From the '70s muscle cars it was very clear that upwards of 300 hp was easily supported by a header down pipe size of 2-1/4" and a 2" tailpipe, and in fact this was almost the standard. The point being that bigger pipes can be used if desired and may help produce more peak power, but could potentially narrow the torque band a bit. At the same time, the exhaust system as a whole will have one resonant frequency where it is most efficient. Larger pipes move this point up the rpm band which is good for power but not necessarily good for economy. For that the exhaust is often tuned to the peak of the use band.

So that is maybe a useful starting point.

JB

Jim, I'm with you. The system needs to be thought of as a whole.
Too many "builders" buy into the hot part of the moment mentality and end up disappointed.
We also have the fuel injected and carbed camps. What works for one does not work for the other. Sewer pipe sized exhaust on an FI car works OK, it's a disaster on a carbed car.
You are clearly a "been there done that" guy and I think your 2-1/4 recommendation is right on.
Jacques you are absolutely on the right path by thinking the whole engine through from air cleaner to tailpipe. But judging from your post you haven't decided what you want your car to do. You mention racing tuning, street manners and mpg. It's like the old saying, you can have your car "good" "fast" and "cheap" pick any two. But tuning your air management system is even more restricted, pick any one and a quarter.
If your tuning for racing then it all boils down to header and collector dynamics. Which become pretty straight forward when you're aiming for specific hp/torque/rpm targets. Exhaust pipes and mufflers don't enter the picture.
Tuning for mpg is a very interesting subject and while optimizing the drivetrains efficiencies is the path to greater hp I don't think you built a 331 to be a gas miser.
That leaves us with the street. In the real world,not the magazine Disneyland, a really good two valve 331 might get 350 hp on a good day. (We really should be talking torque but hp is so much cooler)
Dual 2" pipes will flow enough air to support 400 hp or so. With decent mufflers attached this will probably drop to around the 350 hp range. Add in a few bends and we're looking at 300/325hp. This would be a good choice if mpg trumps ultimate hp. Exhaust temps at the tail pipes will still be nice and high which indicates good velocity and aids longevity.
Dual 2-1/4 pipes will cover 450-500 hp. Add some bends and a couple of good mufflers and you can pass 375hp. This should cover your engine nicely and not be so big as to reduce valuable ground clearance and slow down the exhaust into the problematic range. As a side note a single 3" will flow as much as the dual 2-1/4s due to the reduced surface area despite the greater volume of the duals. However a 3" exhaust would require your mg to become an SUV.
Unless your using 180 degree headers, rate your mufflers on CFM not pulse tuning. I don't want to cause trouble but, Flowmasters flow like crap. Dynomax turbos flow very well and Magnaflows sound awesome.
Crossovers will help a compromised system and make the engine sound more rev happy. An H pipe seems to be better power wise. Think about the straight through nature of it vs the 90' bend required for each side of an X pipe. But the X pipe sounds smoother.
As far as tailpipes go, a smaller pipe ie. a 2" pipe on a 2-1/4 system will give the exhaust that old timey RRRap sound. A bigger pipe IE. a 3" on a 2-1/4 system will give it that boom sound that makes the engine sound bigger than it is.
In the end you can throw numbers at it all day long but what matters is
"does it work or not"
Anyway that's my take on it.

Cheers
Fred

Side pipes anyone ? Better ground clearance. Properly designed to reduce air getting under car,(crude skirt).Less heat transfer onto floor boards. Lastly, looks ? roverman.

I've done that. So has Joe Scheaffer. We agree it'd be hard to go back, but there are real disadvantages too. Aside from being noisy, getting fumes in the car is a real pain, especially sitting at a light with just the right crosswind. I'll be taking a real hard look at routing a tailpipe along the same path used in the MG-Roadmaster. On mine I think it can be done, but it has fully enclosed sidepipes. On that system heat is an issue that must be dealt with.

JB

JB and Fred,

I appreciated your posts.

I'm building a 'Nasty-Boy' 1958 Austin Healey 100-6 two seat roadster. Also employing a 331 SBF like Jacques. Have also procured a narrowed Currie built 8 inch Ford rear-end with TrueTrac & 3.55 gears, and a Tremec TKO-600 5 speed trans with the shorter legged OD 5th gear. Rear tires are 225/60/15s on 15x7 Dayton Wire wheels.

On the dyno, my engine peaked at 400 bhp @ 5800 rpm / 416 lb-ft @ 4200 rpm. Report starts at 2800 where it was 375 lb-ft and 200 bhp, and closed out at 6000 with 350 lb-ft and 398 bhp. Has a forged steel crank, forged pistons, Edelbrock aluminum heads and intake, a tuned 'Stage 2' Holley 650, a 'mild' Comp Cams Hydraulic Roller Cam kit, and all the other standard goodies to hold it together.

The combination employed is meant to deliver a nice smooth, strong torque curve, and I think this goal has been accomplished to date. I want to keep it that way!

I too am researching exhaust options... And given the age of the car, cats are not an issue even in California (yeah!), where I grew up and hope to drive the car when it is done (presently living in exile in a suburb of Pittsburgh, PA - "the Paris of Western Pennsylvania"). Because I expect you'll ask or assume - yep, the engine is going to mostly live between 1500 rpm and 3000 rpm with only the occasional run up to 6000 rpm when embarrassing random German and Italian steel - but I want to be sure the engine will get to well past 5500 rpm long before it starts to choke itself off.

Trying to find a nice set of short tight center dump headers to clear mounts and such... Engine is bolted low and flat parallel to the ground to a custom Jule Enterprises frame (rails 17 inches apart and 23 inches on the outside - engine mounts are customized very stout units meant for a biscuit type combination, but mated to a pair of C5 Corvette hydraulic mounts that bolt directly into the frame rails). From what I've read in notes by David Vizard (a guru), header tuning isn't that important on a street bound V8 (where this will live - hell this thing is going to have AC...) - his reference engine for what I read was the Chevy 350. Any header recommendations..?

Also trying to design a complete system that will not rob too much from the 'go-fast' side of the equation while keeping the noise to a real minimum. I've read about really quiet 'three-muffler' systems used on big Turbo-Buick 10 second dragsters that are so quiet you can barely tell the engine is running (see: [www.turbobuick.com])! That is my 'goal' - super quiet performance.

Thinking a pair of DynoMax Super Turbo mufflers with an 'H' pipe - maybe a DynoMax 'X' if I can make it fit... I understand that the implementation of either an 'H' or an 'X' will both increase bhp and quiet down the exhaust if properly employed.

Thinking of starting with 2 1/4 pipe from the headers back and running 2 inches out from the mufflers to some sort of shiny tips...

But what about resonators - placement and dimensions?

And how far back should they be from the headers?

I understand the 'H' pipe (or 'X' pipe if I choose that route) should be as close to the headers as possible (and in the case of an 'X' pipe should be sure to be made to look to the gasses like an extension of the tuned-length from the headers). Does this hold true if resonators are employed?
Any recommendations on type of resonators (the Turbo-Buick guys used some sort of 'plane-jane' Summit Racing stuff)?

I've also read the trick is to use the resonators as the end of the tuned-length section from the headers, but the dump into the resonator has to be properly designed for the gas in the pipe to not think the resonator is just an extension of the tuning pipe... I can figure that part out from what I've read, but haven't seen a good formula for calculating the length of the tuning-pipe from the headers... Got one?

Getting going off the line is not my greatest concern, but pulling away with smooth power delivery (not peaky-spikey) is something I want to keep and enjoy....

Any suggestions or guidance would be very appreciated!

Again, enjoyed your thoughtful posts gentlemen!

Regards,

Mark

WOW! Thanks Jim and Fred, I feel like I’m revisiting my exhaust tech article all over again; well said. I’ve originally started reading this tech info on Sanderson’s website:
[www.sandersonheaders.com]

Then I went on searching various tech articles and publications. I’ve failed to mention that my goal targets mostly street cruising and an occasional street light stand off for fun.

Fred, I’m curious as to why you would assume that I would seek a gas miser out of a 331? This light weight fun little car with hidden muscles that carries modern fuel management will do a fine job at managing fuel consumption under light throttle; that’s the beauty of modern EFI horse power, it has thought us that fuel efficiency through proper computing, is horse power. My friend has a 2004 Corvette with the base line 400 H.P. “LS engine” that cruises at 28 M.P.G., lots more cubic inches than my 331.

Jim, all you said, concerning the camshaft choice is dead on target. I’m using the old reliable Ford Racing E-303 bump stick in EFI form. It is older technology, but well proven for my needs, it also seems that a lot of existing data revolves around it.

A couple of areas not mentioned that can enhance/tune the given specs of a camshaft are the rocker ratio and the crankshaft small timing gear (slotted to advance or retarded power curve). I wish that there was a way to install a camshaft phaser (variable valve timing) so the choice of power band and R.P.M. would no longer be a problem.

I’ve installed a full roller rocker with a 1.72 ratio, hence, the duration is now a bit longer and will actually (IMHO) enhance cylinder fill if it exist in harmony with the exhaust resonance (again IMHO), I could be wrong on my calculation/assumption.

It is very true that pipe size are in direct relation to engine cubic inches, however, you should mostly size the diameter based on the desired horse power (see The Sanderson chart) and just like an intake runner that is sized too big will kill (poor air velocity) the low end torque, an exhaust pipe sized too big will also kill the velocity necessary to produce scavenging (trailing) low pressure. Again, with assumptions based on engine improvements, 2-1/2” piping is within the rules of given specs. Again, take a look at Sanderson’s pipe size chart; it matches all responses given on most respectable Mustang web forums.

Lets not get into forced induction as it changes a lot of rules, but, mufflers designed around turbochargers, hence “Turbo muffler”, seems to be a decent choice because of its lower flow restriction, however, it may not suit everyone’s taste for that perfect sound.

I’ve spoken with a few avid racers and I was told to brush on a stripe of paint from the collectors up to the muffler inlets then look where bubbling starts and ends; supposedly, it give a general idea of the effective thermal area to work your components in, sounds reasonable to me. Since physics dictates that hot flows towards cold by nature, it makes sense to use that data and attach it to thermal velocity and scavenging low pressure.

Anyway, I wasn’t looking to start a debate on who is the smartest on exhaust science, but, I was trying to evaluate some questions specific to the MGB layout such as:

1. How does the unique “short wheel base” of the MGB affects the rules of the design? It should be great help in scavenging exhaust and creating low pressure evacuation/filler aid?!?!
2. Are oval pipes helpful to under carriage room restriction?
3. Would an “X” or “H” pipe be affected by the short wheel base?
4. Would the stripe painting technique help create some basic blue print?
5. What advantages would result in mounting mufflers on each side of the fuel tank versus underneath the seat floor pans?
6. Has anyone experienced using different system on the MG and able to speak about noticeable trial and errors (do and don’t)?

Well thought out fabrication is well spent money. I don’t believe that the system has to be complex to deliver maximum performance, just well thought out. Guys, thanks a million for your input, it helps tremendously in asserting a plan.

Ground clearance will always be a problem with any MG set up for fast cornering and hard driving. The mufflers and pipes will always hit first so anything that can be done to increase clearance there will help. Also, shorter pipes do not need to be as large in diameter as longer ones so for instance, using 2-1/4" instead of 2-1/2" pipe buys you extra clearance, as does tucking the pipes in closely. Remember, even the 2-1/4" is significantly larger than original and leaves less ground clearance. This was the stimulus behind the exhaust system on my roadster, which never extends below the body and therefore gives maximum ground clearance. I simply couldn't keep an exhaust system on the car before going to that system and was constantly grounding the pipes. However, most cars don't sit that low so it isn't quite as much of a problem. It is still a concern with speed bumps and such though. Oval pipes are fine for a straight run but I think you'll run into trouble with bends. For mufflers I'd try turbo mufflers under the car and smaller glass packs beside the tank. I think that will give a good balance with minimal back pressure. Use the thinnest turbo mufflers available underneath and center the tank. Then if you don't care for the sound it's fairly easy to modify it. We went with rear glass packs only (or they might be baffled bullet mufflers) on the MG-Roadmaster and it sounds good and that does give extra ground clearance, but might get a bit loud over time. It's a low compression engine though and a high compression engine will have a sharper and louder sound. An "H" pipe is worth doing but I'm not sure I've ever seen one on an MGB. Again, crossing under the frame rails cuts ground clearance and this is at a spot that is likely to ground anyway. I've considered a crossover but never have designed one for just that reason.

JB

Thanks Jim, after reading your postings for 2-3 years now, I sense that no one puts as much time in R & D as you do; that alone should crown you as "The Go To Guy".

Where do you find the time?

Jacques

I'm an insomniac ;-)

JB

@ California Kid

Take a look at SpinTech's "Cruiser" mufflers. That's what I'm using on my Healey. Only 2 1/4" tall and supposed to be relatively quiet. I was looking for a system that would give me the most ground clearance for the Healey and the only way I could think of something that might give me more than the Cruisers were chambered pipes like on Dan Master's MGB-GT, but I think they would be louder than the SpinTechs. Haven't got the exhaust on yet, so don't know what the ultimate sound would be like. Sanderson makes some block huggers for SBF's that might work for you, but I've read some people aren't too happy with their quality: [www.sandersonheaders.com]

HealeyRick,

Thanks for the tip. The SpinTech product's "oval" tubing is an interesting option.

I didn't see any detail on the make up of the SpinTech mufflers (will be going back to look again)... I really like what I've heard about the DynoMax Super Turbo's design and function - David Vizard (http://www.amazon.com/David-Vizard/e/B001JOWD4M) was instrumental in the original design - apparently often copied by others.

I have looked at the Sanderson headers. I have not heard anything about the quality, but the gasketless design gives me concern.

There are some other headers that are not exactly center-dump that might work, and alot of the center-dump headers look like they would dump right into the frame rails... Might have to go 'custom' to properly fit...

Thanks again for the post.

Mark

Hey Jacques,
I was just using the gas miser example to show a tuning route that some might be interested in. I think if you re read my post you'll realize that I discounted that option in your situation. Mpg in your case is going to end up where ever it ends up, which should be pretty good with the correct gearing.
Your friends Vette aught to get excellent hwy. mileage. It's engine management and engine design are light years ahead of the SBF stuff. Add to that an overdrive ratio that puts the engine just off idle at cruising speed and you have a winner.
Use caution with your rocker ratios. Everything changes with the ratios and not always the way you would expect. In general a ratio change is used when the cam choice is limited. The duration of your cam stays the same with a ratio change. The lift and valve acceleration change and often overstress the valve springs. Modern cams are designed around specific rocker ratios and tend to under perform when this is changed. If you are looking for a certain performance level then chose a cam that meets it and you'll be much better off. Changing your cam timing doesn't get you any more power ( unless it was grossly out of time in the first place). What it does do is allow you to move the torque peak around. Advancing the cam pulls the torque peak back some to a lower rpm level and retarding it moves it up. Once again care needs to be exersized here.Like the ratio change it alters the piston to valve clearance and if overdone can send your exhaust temp through the roof.
I still think that the 2-1/4" system is your best bet for several reasons. #1 is ground clearance #2 is it will support your HP level well.#3 is exhaust velocity and #4 is ground clearance.
A 2-1/2" system has bragging rights but it won't make your car any faster.
Take the "Turbo muffler" thing with a grain of salt. Some are very good and some are not. Legend has it that the turbo muffler started out on the old Corvair turbo. GM found that the turbo muffled the car so much that a full blown muffler was no longer required. Street racers found that this new muffler with fewer internals sounded good and went better than the mufflers of the day. Thus the "Turbo Muffler" moniker. Reality was that they were terrible by todays standards but the name lives on. Choose your muffler by flow level and sound quality not marketing hype.
The paint stripe on the pipe is a time honored tuning tool and it works well. Build your system without mufflers or crossovers and paint it with the cheapest stuff you can find. Check for cops and take it out for a hot lap around the block. You're going to find the paint burnt within a foot or two of the collector. Your crossover goes here. Install the crossover then drive it like you stole it again. The hot spot after the cross over is where the resonators go. Use a resonator with a greater volume than your pipes and the system will think it ends there. After that put your mufflers wherever you wish and your done.
The reason for the hot spots is due to harmonics and resonance. The pipes get hotter wherever two exh pulses collide. The out going pulse meets a returning one. This effectively becomes a restriction and you want to provide a release for that pressure. As you can see wheelbase doesn't really enter the picture.
I think the spin tech pipes are a great. Use them where the clearance is minimal then use the adapters to round them up for better muffler choices.
Thats all I've got for now.
Cheers
Fred

"I don't know everything that works......but I know an awfull lot about what doesn't".

Fred,

Your guidance to Jacques on the paint trick is superb! Makes complete sense! You also confirmed most of my understandings to date - your rationale for 'H' or 'X' pipe location give better justification than I had found elsewhere.

Following is a good explanation of some of the peculiarities of V8 exhaust design I found: "Experience indicates inline four-cylinder engines are more sensitive to primary pipe length, but a two-plane cranked V-8 is not two inline fours lumped together. It is two V-4s and, as such, does not have even exhaust pulses along each bank. With a conventional, as opposed to a 180-degree header, exhaust pulses are spaced 90, 180, 270, 180, 90 and so on. The two cylinders discharging only 90 degrees apart are seen, by the collector, as one larger cylinder and accounts for the typical rumble a V-8 is known for. This means the primaries act like they do on a four-cylinder engine, but the collector acts as if it were on a 3-cylinder engine having different sized cylinders turning at less revs. (Doesn't life get complicated?) This, plus the varied spacing between the pulses appears to be the cause of the system's reduced sensitivity to primary length..." courtesy of David Vizard from an article published in February 2009 in Super Chevy titled "Exhaust Science Demystified."

[www.superchevy.com]

Re. validation of using an 'H' or 'X' pipe, because of the unbalanced pressure caused by the peculiarities of V8s noted above (i.e. the "3-cylinder engine having different sized cylinders"), "Virtually all V-8 exhaust systems can be refined by the addition of a balance or X-pipe. These have two potential attributes: increased power and reduced noise. Extensive dyno testing on both of these factors has indicated balance and X-pipes are 100 percent successful at reducing noise. The reductions amount to a minimum of 1 dB to a maximum of 3 dB with 2 dB being common. As far as power is concerned, things are a little less certain. With engines between about 325 to 550 hp, experience indicates that in about 60 percent of the cases (mostly with balance pipes), the engine can deliver as much as 12 additional hp, with 5-8 being the most common. The other remaining 40 percent tested showed virtually no change in output either up or down. Based on such results, we can conclude that a balance or X-pipe is always a positive asset and never a negative..." again courtesy of David Vizard's article.

The choice between a balance ('H') pipe and an 'X' pipe is a little more personal, but some analysis has been done in less than scientific manners that lean twoards 'X' pipes being both quieter and more productive.

See: [www.boyleworks.com]

Also note that the system design suggested by David Vizard in the referenced article incorporates a large resonator box that is also basicly just a large 'X' pipe...

You may not have given me the math I was hoping for, but you certainly gave me a map to follow...!

Thanks!

Cheers,

Mark

> I have been reading, and working (above my head) with all sort of
> calculations,

The formulas are a starting place but the wave dynamics are complex enough
that dynamic simulation is required. David Vizard wrote an excellent exhaust
system article "Exhaust Science Demystified" that was published in the May
2005 issue of Popular Hot Rodding and another good article "No Loss Exhaust"
that was published in Drag Race Monthly. In these articles, Vizard gives some
design criteria for header and muffler sizing. The equations are based what
the exhaust port flows at your maximum exhaust valve lift and apply to bank
separated 4-into-1 headers on a V8. I use his equations as starting points
for Dynomation iterative simulations so I wrote a little routine to calculate
the parameters. It's in Fortran and I've cut out the relevant section below.
For a SBF, treat the calculated collector length as a minimum value for best
performance. The SBF typically responds to long collector length (helps
increase velocity in the short exhaust port), regardless of cam or RPM. Long
in this context is up to 24 inches. The Flowmaster 4-2-1 collectors work well.
The optimum collector length is best determined by testing but there will
typically be a noticeable burn pattern inside the collector which will indicate
the length the engine wants. The primary diameter should be based upon what
the exhaust port flows and is typically a bit larger than the exhaust valve
diameter.

C Calculate exhaust parameters (initial values for Dynomation)
C
C OID = optimal internal diamter (inches)
C Ex_CFM = exhaust port flow (CFM) at 28" H20 (peak within max valve lift)
C FD = flow density (around 80 CFM/square inch @ 28" H20)
C CDiam = collector diameter (inches)
C
C Based upon Vizard articles:
C "No Loss Exhaust", Drag Race Monthly (date?)
C "Exhaust Science Demystified", May 2005, Popular Hot Rodding
C
C Primary lengths should fall between 24 to 36 inches (for bank separated
C 4-into-1 headers on V8's with dual plane crankshafts)
C Collector length between 12 and 20 inches suitable for up to 8000 RPM
C Mufflers should flow 2.2+ CFM per HP (the point at which there is a
C 1% difference between mufflers and open pipes), e.g. 500 HP requires
C two 550 CFM mufflers

ELSEIF (OPT .EQ. 11.0 ) THEN
WRITE(*,*) 'Horsepower Goal:'
READ(*,*) HP
WRITE(*,*) 'Peak Exh Flow in CFM at 28" Hg within max cam lift:'
READ(*,*) Ex_CFM

FD = 80.0
MFLOW = 2.2 * HP
OID = SQRT (Ex_CFM * 1.27/FD)
CDiam = 1.75 * OID

WRITE(*,*) 'HP Goal =', HP
WRITE(*,*) 'Exhaust Flow =', Ex_CFM
WRITE(*,*) 'Required Muffler Flow =', MFlow
WRITE(*,*) 'Primary ID (inches) =', OID
WRITE(*,*) 'Collector ID (inches) =', CDiam
WRITE(*,*) 'Primary lengths between 24 to 36 inches'
WRITE(*,*) 'for bank separated 4-into-1 headers on'
WRITE(*,*) 'V8s with dual plane crankshafts)'
WRITE(*,*) 'Collector length 12 and 20 inches good to 8000 RPM'

> to formulate the perfect exhaust system for my MGB powered by a
> Ford small block (331) stroker.

I'm working on a couple of small block Ford (SBF) engines at the moment.
One's an EFI 5.0L with AFR 165 cylinder heads, the other a carburated 331
with Floo Tek heads (Australian castings, assembled in Indy).

> Having said that, the reality is that having the wrong mufflers or pipe
> size can defeat the perfect induction system.

Stock exhaust systems can be very restrictive and mufflers are absolutely
critical. I recently dyno tested three different ANSA mufflers on a 400
to 525+ HP 351C Ford V8's, along with stock exhaust manifolds and several
sets of heads. The ANSA mufflers lost over 50 HP compared to Magnaflow
stainless steel mufflers. The Magnaflow mufflers were within a couple of
HP of open pipes. We gutted the ANSA mufflers and they still lost 38 HP.

> By contrast, the old and poorly appreciated cast iron log manifold,
> precisely because it is incapable of being tuned to a precise range, gives
> a very flat performance curve which, it turns out is just what you usually
> want in an average passenger car.

I've got to disagree here. Headers are among the best bang-for-the-buck
modifications for a SBF, increasing power across the rev range. Even on
an engine like the 351C-4V which has cast iron manifolds with large passages,
the effect is quite dramatic. When we tested cast iron manifolds against
long tube headers, the headers were better across the rev range. At the
peaks, the OEM cast iron exhaust manifolds were down 34 HP and over 38 ft-lbs
of torque compared to the headers. In the 4000 RPM range, the manifolds cost
50 HP and the idle was noticeably worse. Bottom line is you can lose a bunch
of power on a poor mufflers and cast iron manifolds. The best mufflers for
performance are the straight through designs. The better ones use stainless
steel packing that doesn't burn out like the fiberglass stuff. Some of the
muffler companies are flow rating their mufflers now. To be non-restrictive,
mufflers should flow 2.2+ CFM per HP (the point at which there is a 1%
difference between mufflers and open pipes). For instance, 300 HP requires
two 330 CFM mufflers. Cross over pipes near the collectors can help with both
power and noise. Cams with narrow lobe centers are more sensitive to back-
pressure and require good mufflers to work properly.

> BTW, maintaining back pressure turns out to be a myth, it has nothing to do
> with fine tuning exhaust formulas.

Yes, that myth dies hard. While racers do add header collector extensions,
it's not to increase backpressure. They do it to alter the acoustic tuning.

> Back to the dirt bike, these engines are renowned for a peaky powerband, primarily
> because the pipe is tuned to provide the greatest possible suction volume at the
> exhaust port at a particular rpm, known as being "on the pipe" and they create
> tremendous power at that engine speed, but can be absolute dogs below that.
> This would be bad in a street driven car.

This pronounced effect is specific to 2 stroke engines and does not necessarily
transfer to 4 stroke engines like a SBF.

> The crossover pipe seeks to even out the exhaust pulses

On an engine with back pressure, the primary effect of a cross-over pipe is to
permit sharing of muffler flow, increasing the effective flow rate of the
mufflers.

> From the '70s muscle cars it was very clear that upwards of 300 hp was easily
> supported by a header down pipe size of 2-1/4" and a 2" tailpipe, and in fact
> this was almost the standard.

Going from a larger down-pipe to a smaller tail pipe is a power killer.
It's nearly always better to step up to a larger muffler and tail pipe size.

> The point being that bigger pipes can be used if desired and may help produce
> more peak power, but could potentially narrow the torque band a bit.

The collector outlet diameter is key here. For a wide power band, don't use
too large a collector.

> At the same time, the exhaust system as a whole will have one resonant frequency
> where it is most efficient.

But there will be subharmonics as well so the effect is not only at a single RPM.

> You mention racing tuning, street manners and mpg. It's like the old saying, you
> can have your car "good" "fast" and "cheap" pick any two.

A well coordinated engine can provide a large increase in power with only a small
decrease in cruise fuel economy. I recently designed an engine for a friend's
Pantera. His poorly coordinated prior engine made We went from under 300 HP to
526 HP and he now gets better gas mileage (at least when he keeps his foot out of
it). Though it's an oldie, David Vizard's Performance with Economy book is still
recommended.

> Tuning for mpg is a very interesting subject and while optimizing the drivetrains
> efficiencies is the path to greater hp I don't think you built a 331 to be a gas
> miser.

Getting good fuel economy on the highway is about low drag, minimum cam overlap,
high (dynamic) compression and tall gearing. Running the engine well below peak
torque at cruise (via tall gearing) helps fuel economy because you need a larger
throttle opening for the same power which reduces throttling losses which, in
turn, means lower BSFC. Ideally, you'd run at near WOT at cruise (with any sort
of power enrichment disabled and timing optimized) but acceleration would be very
sluggish and you'd need to change gears just to accelerate or run up hills. Lower
RPM also leaves more time for complete combustion.

The new Corvette Z06 is an example of what can be done these days. It makes 505
net HP yet still meets emissions goals and is rated at 26 MPG highway. The Z06
has relatively low drag, a short duration hydraulic roller cam (211/230 degrees
duration, 0.591" lift), high compression (11:1) and very tall gearing (in sixth
gear). The result is good highway fuel economy even though the displacement is
427 cubic inches (7 liters). The large displacement and minimal overlap means
lots of torque which allows the engine to cruise at low RPM yet still have plenty
of power to pass without down-shifting. The lightweight valvetrain, high cam
lift, and excellent cylinder heads (360 CFM intake at max lift) allow it to make
excellent power. The 400 HP LS2 in the base model Vette is rated at 28 MPG
highway and many owners report pulling 30 MPG on the highway.

> In the real world,not the magazine Disneyland, a really good two valve 331 might
> get 350 hp on a good day.

While I agree that magazine numbers often inflated (David Vizard and Rick Anderson
are two builders that publish honest numbers), a good 331 can make considerably
more than 350 HP but it requires really good cylinder heads.

> (We really should be talking torque but hp is so much cooler)

A discussion for another post. For now, allow me to suggest what you want to
design for is the maximum average power between shift points.

> I don't want to cause trouble but, Flowmasters flow like @#$%&.

and are sensitive to distance from the collector.

> Dynomax turbos flow very well and Magnaflows sound awesome.

Magnaflows work very well.

> Also employing a 331 SBF like Jacques...
> On the dyno, my engine peaked at 400 bhp @ 5800 rpm / 416 lb-ft @ 4200 rpm.
> Report starts at 2800 where it was 375 lb-ft and 200 bhp, and closed out at
> 6000 with 350 lb-ft and 398 bhp. Has a forged steel crank, forged pistons,
> Edelbrock aluminum heads and intake, a tuned 'Stage 2' Holley 650, a 'mild'
> Comp Cams Hydraulic Roller Cam kit

Which version of the Edelbrock heads?

> From what I've read in notes by David Vizard (a guru), header tuning isn't that
> important on a street bound V8 (where this will live - hell this thing is going
> to have AC...) - his reference engine for what I read was the Chevy 350.

What? Vizard is a strong proponent of headers on street 350's with plenty of
dyno testing to back it up. Perhaps you're confusing equal primary lengths.
Vizard has shown that bank separated 4-into-1 headers are insensitive to
relatively large primary length variations. 180 degree headers are very sensitive
to equal primary lengths. On a 90 degree V8 with 2 plane crankshaft, 4-into-1 bank
separated headers are much less sensitive and are able to tolerate large variations
in primary length. Much more important than equal length is the radius of the bends.
On a SBF, the collector is more important than the primary length. The mufflers are
as important as the headers.

> I've read about really quiet 'three-muffler' systems used on big Turbo-Buick
> 10 second dragsters that are so quiet you can barely tell the engine is running

Turbos tend to quiet down exhaust all by themsleves.

A Helmholz resonator can be used to tune specific frequencies but absorptive
media works over a wider range. Lots of small holes in the tubes, surrounded
by the absorptive media. More volume makes it easier. More cylinders raises
the frequency and that diminishes the size of the elements required. Large
volume can smooth out pressure waves. The pressure waves can excite vibrations
in the shell of the muffler and cause it to radiate noise. That's why shells
are nearly always round or oval and are often double wrapped for damping.
For Panteras which have dual tipped muflers, we've made removeable inserts
to quiet them down on trips.

> Thinking of starting with 2 1/4 pipe from the headers back and running 2 inches out
> from the mufflers to some sort of shiny tips...

I would not reduce the diameter at the tail pipe. Stay the same or larger.
However, we have made removeable inserts to quiet down loud exhaust on trips.

> I’m using the old reliable Ford Racing E-303 bump stick in EFI form.
> I’ve installed a full roller rocker with a 1.72 ratio

For nearly all aftermarket heads, the E-303 will work better if you run 1.72:1
rockers on the intake side and 1.6:1 on the exhaust. I've personally tested
this and the effect is even stronger on dual pattern cams. What EFI intake
manifold are you using? If you are running one of the long runner EFI intakes,
that will drive much of the tuning considerations. Make sure your springs are
strong enough to handle the E-303 and 1.72:1 rocker ratio.

> Again, with assumptions based on engine improvements, 2-1/2" piping is within
> the rules of given specs.

2 1/2" duals all the way back will work fine as far as the engine is concerned.
I have no experience with what fits in the MGB but you want a long collector.
Sacrifice the primary length if you must to get the desired collector length.
The larger the pipe size, the harder it is to make the bend over the rear axle
and still clear everything. It's okay to increase pipe size as you go aft but
you do not want to have larger pipes dump into smaller ones as you go aft.

What cylinder heads are you using?

> The duration of your cam stays the same with a ratio change.

While the duration of the lifter stays constant, the duration at given valve lift
(and the period of overlap) changes. The effect is relatively small, however.

Dan Jones

Excellent reply Dan,

I was heading to my library to get Mark some recommendations for his "math". But you beat me to it.
Like you, I strongly recommend anything that Vizard has written as mandatory reading.
Another good reference is Scientific Design of Exhaust and Intake Systems by Phillip Smith and John Morrison.
The book is a little dated (at least mine is) but physics don't really change over time.

Cheers
Fred

Wow, lots of Vizard fans here. Used his Mini books on the A Series BMC engine as a bible when building my bugeye. David does a weekend seminar for $550. I'd bet this would be a huge bargain for pro motor builders:



[www.davidvizardseminars.com]



Edited 1 time(s). Last edit at 11/03/2010 08:44PM by HealeyRick.

Dan,

Wow....

First things first, my apologies - I over simplified my statement re. headers. You were absolutely correct in assuming I was referring to equality of primary lengths - or rather the 'lack-of-the-necessity-of' equality for my application / non-180-degree 4-into-1 split system...

I've noted that secondary length is important (seen 18" to 24" as good rule of thumb)... but your note is the first I've seen keying in on collector dimension importance for the SBF. Am I making a mistake in nomenclature - in this discussion are you using collectors and secondaries interchangeably? If not, what guidance can you give on collector dimensions?

The Edelbrock heads I'm using are the Performer RPM - 1.6 inch exhaust valve / recommended 1500 to 6500 rpm range:
[www.edelbrock.com]

Having a hard time finding commercial headers for the SBF that have primaries of larger diameter than 1.6" and will fit my car (short / center dump / non-block hugger - need to clear 23" outside width of frame rails / 1958 AH 100-6 on a custom Jule Enterprises frame).

I'm assuming I should be going with headers with 1 5/8" ID primaries because of the heads I'm using... I expect the primaries should be ~18" long and lead to 4-into-1 collectors with conventional 2.5" ID terminations that will be plumbed down to ~24" long 2 1/4" ID secondaries leading to a 3" ID resonator (maybe a Dynomax 'X' pipe / killing two birds with one stone / expansion to 3" used to tune length of secondaries) before plumbing to a pair of Dynomax Super Turbo mufflers (or comparable) that flow ~440 cfm each. I will likely follow your recommendatin and not skinny down the exhaust to 2" after the mufflers as was previously considered.

Afraid I might be going custom on yet another aspect of this project for the headers... In that event, you noted that bend radiuses are important - any specific guidance assuming 1 5/8" ID tube?

And Dan, thanks for your thorough and thoughtful post... You obviously know your stuff...

Cheers,

Mark

Thanks for the very informative post Dan. It's good to get some of the mysteries of exhaust tuning cleared up a bit, and we all know there is plenty of mystery about it. So how about another question? Could you elaborate on that business about not stepping the pipe size down after the muffler? That's a trick the manufacturers have used pretty much forever and are still doing so there must be some pretty good reasons for it. Seems like if it was that much of a power killer it would have disappeared a long time ago. (Of course we might say that manufacturers aren't interested very much in performance but I don't really believe that. Maybe it's because I'm not trying to find that razor's edge of maximum performance either.)

Another thing worth mentioning is that extremely few of the headers used in LBCs are equal length headers, which means a great deal of the effectiveness of having headers is lost right there. In fact, as bad as cast exhaust manifolds are, some well executed designs are out there which perform as well as some of the minimalist shorty tube headers. So it's really more a question of degree and of quality of execution I think. And here is where the advantage lies in building your own headers. As an example, I built headers for my 215/MGB roadster 25 years ago using guidance and parts from Ed (Headers by ED) and built true equal length primaries +- 1/8". The performance of those headers has been exceptional, but fitting them into an MGB was a challenge. The good news is that they are good enough that I'll still be able to use them on the 340 that is going in the car now. At 1-3/8" x 34" on the primaries the tubes are only a smidgen larger than the exhaust valves and at 2-3/8" the collector could be a bit larger without hurting anything but it will work just fine. The set was sized for a 5L displacement so it was a bit big for the 215, and a bit small for the 340 but close enough. And it keeps me from having to build another set. Considering that I had 80 hours of construction time invested, I consider that a very good compromise.

There are probably more compromises in a car's exhaust system than in any other part of the vehicle, at least for a passenger car and especially a LBC. Header primary size certainly can be in that category. Fitting all those tubes in is difficult enough at 1-3/8", if they had been 1-5/8" I don't think it would have been possible without very major redesign of the car's monocoque structure, or else seriously compromising on primary tube length. Choosing between equal length tubes and a step down in tube size, I think the advantage goes to the smaller equal length tubes. Maybe Dan can prove me wrong on that one.
Finally, although I don't really see the relevance of the information about 180* headers it's good to know and interesting from an engineering perspective. Since it is such a plumbing nightmare even in an open chassis race car though, not much of that applies to LBCs. In a practical sense we're pretty limited on where and how the pieces can be fit together and often performance considerations have to give way to practicality.

JB