The ads above are Google-sponsored.
Clicking on them at every visit helps support this website!
Clicking on something inside an advertisement helps even more!!


Tuning:  Intake and Exhaust Systems.
(ramblings on exhaust and intake system function & modifications)
PLUS:  Premium fuels versus mileage & combustion temperatures.
R. Fleischer

This article contains a discussion of factors that go into the size, length, width, opening area, ETC., of the intake and exhaust system.  It is not a how-to article, but offers a fair amount of not easily found information.

Firstly, just to show you something to whet your appetite, take a look at this calculator, this is the updated URL for Mez Porting's calculator.   This, with appropriate input from you, will give you a lot of information.  Even if you do not plan on doing the calculations or using this page, take a good look at it....the things that it asks for; and, the bottom comments, will prove interesting.   This calculator is for singles, twins and four cylinder engines.
(at Snowbum's last check, unfortunately that link was NG)
I suggest you try, and ask where that article is now located....I could not find it on the site!
Please let me know!

Due to VERY complex effects having mostly to do with camshaft timing, size & length of intake & exhaust systems, & shape & size of combustion chamber items... such as port and valve sizes, angles, flow, eddies, etc....;  modifications for power or torque is a mix of science & art.  Even with modern high speed computers a large amount of experimentation is often required. Many seem to think that if the exhaust system is 'opened up', that in itself will give more power, more performance.  They also seem to think that bigger carburetors will do the same.  Same for 'freer flowing' air-cleaners, or individual air-cleaner elements at the carburetor mouths (and eliminating the airbox, etc.).  Many commonly held beliefs are WRONG!

changes really will, almost always, give improvements across the rpm range
Colder air to the air cleaner, even as little as just insulating the air cleaner interior. A good 3 angle valve job, higher compression ratio within gasoline limits; gearing to better match engine characteristics for performance only.

Many changes will often REDUCE performance. Of course, one has to define performance.   Top speed? Quick acceleration to some particular speed? Tractability? Flatter torque curve? (so as to not needing excessive shifting).

The only accurate way to 'prove' your "improvements" to horsepower and torque, top end performance, acceleration, etc., is either known sections of road with test instruments or on the drag strip; or, better yet, on a calibrated, or at least repeatable-readings, type of dynamometer.  Certainly you can road test for characteristics like throttle response and smoothness. Dynamometer time can be expensive.  One could get together with some friends and build one, it need not be calibrated, just consistent. NOT expensive. A several hundred dollars and a lot of labor would get you a dyno.  Purchasing one is likely to cost a minimum of $4000.00; a really good setup will cost much more.

It is counter-productive to 'hop up' an engine if the results are such that the lower & mid range torque are severely reduced....UNLESS you intend the engine purpose to be for mostly or only racing; or, where you might be able to 'stay on the cam or pipes', that is, you keep the rpm well up, all or most of the time. 
A peaky engine makes for a really bad engine for touring, even can be unpleasant for sport touring. What you may end up with is an engine that can produce more top rpm horsepower, but requires a lot of shifting of the transmission at, perhaps, anything under large amounts of throttle and rpm. 

HOWEVER, it IS possible to use a 'hotter' camshaft, which normally might make the engine 'peaky', to which you add compression, exhaust & intake 'tuning', perhaps a modified ignition, etc., such that the resultant horsepower output is increased and the resultant torque curve is reasonably flat enough & the engine not too peaky. This, over-all, is likely the goal of many modifiers for street usage, whether they understand, or not. Doesn't hurt racing performance either!

Only the most serious street rodders would put up with a truly quite peaky engine. 

The following discussions will describe some of the details of engine design and tuning, and give some pointers.

Intake system:

Increasing the size of the carburetors is not always the best thing to do.  As carburetor throat (venturi) diameter is increased, it becomes harder and harder to obtain enough velocity through the carburetors (for that, one usually needs increased rpm and/or displacement) to enable them to atomize fuel correctly. In some cases, as some owners of BMW R80 and R100 machines well know, the 32 mm carburetors provide better throttle feel and even acceleration into midrange & sometimes a fair amount higher, than the 40 mm carburetors.  With carburetor sizes, & this also applies to fuel injection throat size, smaller throats give better low to mid-range throttle response, sometimes even upper mid-range performance. That is because high air velocity through the carburetors or FI is needed to allow them to do their job. At JUST the extreme top end, let us say at or near wide open throttle & fairly high rpm to very high rpm, the largest carburetors that will cause decent atomization, are usually the best for JUST power output. A manufacturer usually strikes a balance in sizes.  Unless one can stay in this high throttle/rpm range (with a more peaky cam you'd also want close ratio gears), OVERALL acceleration may suffer; & the suffering can be a considerable amount.   The Bing CV carburetors TEND to act like a rather modest variable venturi, but do not provide the almost instantaneous 'snap' of a directly controlled slide or butterfly carburetor.  Many have installed Mikuni carburetors to get that SNAP effect.  What is not often realized is that over-all acceleration over a distance may be the same.

It is more important than often considered by those after more and more power, that in the mid-range of rpm, good response is needed on a touring or sports-touring bike. If the carburetor is too large, all sorts of strange effects will occur, including the engine simply refusing to accelerate smoothly when the throttle is suddenly opened a large amount. The CV carburetor tends to greatly minimize that effect, but in some instances it will still be evident. Still, it is the combination of enough quantity of mixture, AND the throat velocity, to ENABLE an atomized mixture to get into the cylinder that is important.  If the velocity is too low, the next intake stroke of the piston will not allow much mixture into the cylinder. It is the total sum quantity of atomizedmixture that gets into the cylinder, at the right time, that is the important thing (this is a simplification).

In ALL cases, the best power is available (assuming jetting can be modified if need-be) from the coolest possible temperature of the incoming air (within reason), since the vast portion of the burning mixture by weight, and volume, is the oxygen in the incoming air.  Cooler air has FAR more oxygen, allowing more power, if the gasoline quantity/mixture is adjusted to match. This is a BIG effect.  A definite increase in power is available from getting cool air to the carburetor intakes. However, if this air is not warm enough, one can freeze up the carburetor (venturi icing); especially if the air contains a lot of moisture. Contrary to popular belief, MOISTURE in the air (high humidity) means LESS power, as the air has LESS usable oxygen.   Generally cold air is not a problem once the engine is warmed up....or not with air intake systems that are warming the incoming air. When cold, the engine may not run well as the fuel is not atomizing well, condenses on cylinder walls and other parts, and is no longer a vapor, especially with cold parts; this is all especially so with carburetors, and much less so with a good fuel injection system.  That is why carbureted vehicles have 'chokes', particularly for cold weather operation, allowing lots more richness, really a brute force method. Fuel injection systems can have far better control, as they can automatically compensate for air density; and most compensate for oxygen in the exhaust, air pressure, temperature of cooling water, ETC.   BMW Airheads, ALL models, have warmish to warm+ air coming to the carburetors, due to the central aircleaner, and how that air cleaner gets its air. It is quite different, for this warmish-warm+ air for the ClamShell models versus the twin snorkle rectangular air cleaner models, but the effect IS THE SAME.

You WILL get more power, perhaps lots more, if you find a way to get cooling air to...and through... the airbox.  It is often difficult to do it neatly
, and to avoid ingestion of rain, leaves, etc. A NASA type duct, including the angle separation system used on turbine motors (airplanes) works well...sometimes.

Ram air: Sometimes folks ask questions about ram air possibilities for "no-cost supercharged horsepower".  Forget about it unless you are planning to ride over 150mph.  Below 130 mph, the effects are EXTREMELY small, and effects even at 150 are just noticeable. The effects above 150 are noticeable and worthwhile. And, no,... a larger or extremely large scoop does NOT mean you will see improvement. You could make a scoop/funnel the width of the motorcycle, and all it would do is likely add a lot of drag. Yes, that sounds wrong, 'common sense' is incorrect here.  There is some information on intake areas velocity in my formulas article:

Dynamic air pressure by or on a moving object is treated by engineers as a certain size of flat plate moving through the air at some speed.   Dynamic air pressure is proportional to the SQUARE of speed, but the horsepower to attain that speed goes up as the CUBE of that speed change.   Read that again!!

Consider your Airhead or any other motorcycle, moving at 68 statute miles per hour.  That is easily converted to feet per second, and the value is 99.7, let us just say 100 feet per second.   Forgetting atmospheric standard pressure on everything....there is about 12 pounds of pressure per square foot of surface area.   At 200 feet per second (about 136 miles per hour) the pressure is 48 pounds per square foot.  (see, I did the calculations for you!!).  The drag is 4 times higher for a doubling of speed (I put this information here so you'd know why gasoline usage increases so fast with increasing speed).   If one was to divide by 144, you would have the pressure per square inch, in this case 0.33 PSIThat is basically almost nothing (2%) compared to atmospheric pressure forcing itself into the cylinders (about 15.0 psi at sea level).  Thus ram-air supercharging effect is very small until speed gets VERY high.  It starts to make some reasonably usable difference around 150 mph.  Thus, ram air pressure does NOT help at ordinary to quite high road speeds.  

Intake tract length:  This is a particularly difficult idea to get across to many folks, and the same effects, in reverse, are in the exhaust system. This effect is very noticeable at all throttle settings, but the effect varies greatly with those settings:

When the intake valve is in its 'open, at least somewhat' phase, & in conjunction with a time just after the piston reaches bottom, & the valve is still not yet closed (we are not going to get into cam timing theory here), the air coming through the intake system is NOT a steady flow.  In fact, at any time, intake valve open or not, there is still some flow into the intake system, as this flow pressurizes against itself, readying itself for the next opening of the intake valve, so to speak. This flow is varying in velocity & pressure, depending on the valve opening & more complicated things. Pressure here means absolute pressure, or, if you will, referenced to atmospheric.   To picture this simplified, let us start at the beginning of the intake stroke.   The valve is opening, the throttle is open a bit or more. As the piston lowers, it reduces the atmospheric pressure in the cylinder, allowing the outside air pressure to push air/fuel mixture into the cylinder. View it as sucking the outside air/mixture inwards if you must.  The piston eventually stops lowering, and in modern engines, the valve closes a bit later. Since this happens at a fast rate, even at idle, the intake flow is in PULSES. The incoming air slows when that valve closes, and this slowdown occurs VERY suddenly. View this as a slug of air slamming against a closed valve, if you want to think of it that way. These pulses are described by engineers in a type of complex interacting mathematics dealing with 'waveforms'. You can think of it as the air piling up on itself.  Because of these pulses, some complex things happen in the intake system, and the one that I want to discuss is a fairly complicated one, the reflected pulse.

The intake system will act like any one specific pulse is partially reflected backwards. This is not inertia
, and if you cannot picture it yet, just accept it as fact!  I will discuss this from a different viewpoint later, in the exhaust, which might be easier to understand. This reversion/reflection effect occurs at the point of intake, which in the later airheads, as an example, is approximately at the tips of the snorkels!   If, at any one given and constant throttle and constant rpm, one could adjust the LENGTH of the effective intake system, one could find a correct length so that the REFLECTED pulse hits the intake valve at exactly a particular moment of time, IN RELATION TO an existing normal INcoming pulse of mixture. This will have the effect of boosting power, as more fuel/air mixture will be forced into the NO cost whatsoever in loss of energy, etc. Of course, you will use more fuel!  FREE POWER!!!...except for the gasoline usage!  I once actually made up a VERY crude sliding tube system for the intake, one of these replaced both snorkels, and on a dyno, moved the tube-inside-a-tube, noting very distinct changes.  I made a second one up, guess-tuned it INITIALLY for 5000 rpm, and varied it while on the dyno.  Believe it or not, this crude device was made from the inner cardboard 'roll' of a roll of kitchen counter wipes.    The effect on rpm/power had to be seen to be believed.   It was necessary to jockey the throttle and tube length to keep the rpm more or less constant.  A change of as much as 8 horsepower was noted.   I want to CAUTION here that I did this only crudely, and a truly more vigorous AND PROPER analysis/testing would have REQUIRED changing the carburetor jetting and an exhaust analysis at each try, and I did not do this, due to lack of time that day.

Engines have been built with variable length intact tracts, controlled by throttle position, rpm, and other things. 
In the past, most of these things were not very practical.  In the last 15 or 20+ years, some car and motorcycle manufacturer's have managed to incorporate MANY of these things.   There have been some clever ways to
dynamically change the effective length and size of the intake tract.  There have even been engines with multiple intake valves that had a valve closed at times other than normal, and especially notable is the various methods of variable camshaft timing and multiple intake butterflies.   Manufacturer's also found similar ways to INcrease compression ratios far higher than in the long-ago past.   NEAR RACE tuning on camshafts has become relatively common, yet the engines maintained nice torque curves and were very tractable, that is, smooth throttle response, lots of low and mid-range power, and plenty of top end power.

Many engines are using what in the past would be regarded as modestly radical camshaft timing and would have low torque and lousy idling, all in the lower mid-range rpm area, except that the various cam timing changes and intake tract tuning GREATLY enhanced performance by improving those lower and mid areas.  These engines managed to keep high rpm performance; all this on regular grade fuel, with some engines having even better performance on 91 octane, yet could run on 87 octane, due to sensors and automatic adjustments.  A few of these engines made NEARLY THREE HORSEPOWER PER CUBIC INCH!

A few manufacturer's added turbochargers (a very few used superchargers) and managed to keep reasonably high CR.

Modern metallurgy and casting techniques allowed the engines to have greatly extended life.


The Airheads, and some generalities:

BMW has Service Information bulletins, particularly on some R80 models, for drilling holes near the snorkel opening, to deal with a reflected pulse velocity problem (poor fuel mileage) (at least I think it was the reflected velocity pulse) that was not so perfect in the mid-range (of course, the technical terms details were left out of the bulletin).   BMW was certainly not trying to restrict power!  Just because it works to improve fuel mileage on some R80 models (or some form was on R45/R65 models) does NOT mean that it works on other models!!!   ...and, in fact, it does not work on most of the R100 models, without some other changes, the R100's being mostly tuned correctly right from the factory.  The article about the intake on the R80 models is on this is article #9 of the Technical Articles List (in the fuel system section):

Late models of the BMW Airheads have intake 'snorkels' to the airbox that restricts the intake opening size, to a smallish taper.  There is a bell-mouth associated with that.   The section with a taper seems to go in the wrong direction!  For some strange reason (I leave YOU to think about this), the snorkels are NOT the same cross-section near the bell-mouth.  LOOK AT THEM!

All this might APPEAR to greatly restrict performance & if openings were small enough, this would certainly be true. However, IF BMW designed the taper & the bell mouth to keep laminar (smooth) airflow & JUST the correct the almost supersonic velocities in the system (which are normal), then BMW was VERY clever.... & is my belief that they did just exactly that, allowing only a relatively small amount of restriction.   Laminar flow means that the incoming airflow (the reversion pulse too, but let's ignore that here) flows smoothly, with little if any burbling, eddy currents & swirling, etc., which would cause losses, during its way down the snorkels. 


Laminar air flow?  Reversion pulses?   Sounds like I'm getting techy, and, yes, I am...but will try to keep things relatively simple.

Some folks may be interested in the effects of changing the length of the intake system. This is OFTEN seen with those 'racer boys' (real or otherwise) who remove airbox stuff & let each carburetor breath through a screen or a filter, right at the carburetor inlet itself.  I thought some of you might like to know this information.  The distance from the intake valve to the intake opening (to the outside air), in inches, is equal to APPROXIMATELY 79200 divided by the rpm, for the maximum torque peak. One of the terms I managed to remove (from the formulas involved in getting that simplification) has a term in it that is a SQUARED function, meaning that things happen FAST with SMALL changes in some items in the formula. The above simplified version of the formula includes certain assumptions I made about the speed of sound inside the intake system, effective pressure, valve timing, & temperature but it is pretty close to the real world.

IF you remove your airbox-to-carburetor items & put stubs on individual carb intakes, you certainly do have a tuning effect, but the effect is now at a very much higher RPM, with the shortened intake length.  NOTE that this mathematical effect is exactly what testing shows, and it is exactly what the result is from using the individual cone-shaped filter units that are popular on Racer-Boy carburetor intakes.  There can be other effects, from the weight causing fuel frothing, and the need for jetting changes.....but I will stay with the torque curve effect for the immediate discussion.

Keep in mind, again, that the length of the intact tract, overall, is from the beginning of the air flow into the system, through the filter, any tubing and connections, through the carburetor, through the head adaptor, and to the inlet valve.

This intake system length ALWAYS affects the rpm for torque peak.

About 7 inches from valve to carburetor inlet to the intake valve was what I used, decades ago, for Airheads, with the stock air cleaner and tubing removed, for calculations for dyno work.  Let's be generous, assume 8 inches there, and 5 inches for a large inlet filter, shaped to act as a pipe. That gives 13 inches of tract length.  That is 6092 rpm for the torque peak (& it is likely much higher).  I think 9 inches is generous enough, for 8800 rpm. Seems about right, from what I remember. The engine will likely be peaky, unless other things are done...such as tuning the exhaust for a different rpm!! (amongst MANY other 'things').   One of my racing engines had such an intake, with specially formed bell-mouths, & a rather restricted exhaust, & smallish carbs, but very light valve gear, & a special camshaft, & a very light flywheel & clutch. We had no problem with WELL over normal redline, to say the least! Not a good engine to go touring on! It had good acceleration above 5300 or so, and above about 6200 it was fierce. 

The bottom line is that if you shorten the intake length, the torque peak will rise, & once it gets high enough, the 'tractable' torque at the lower rpm area is AWFUL.  
With a wilder camshaft on an Airhead, the bottom end torque can be so bad that you must slip the clutch a LOT, taking off from a stop.

The Exhaust System:

At this point you might be thinking that all of the intake effects discussed ALSO happen in the exhaust system, due to complex relationships with the exhaust valve, length of system, intake opening, etc.  YES...the same sort of formulas & things generally really do apply. Please keep in mind that the things that happen in the exhaust system, also happen in the intake system!   I'll try to explain this in a laypersons viewpoint (I hope), in a gross simplification:  

The effect of exhaust system SIZE is generally small, & what is usually most important is the effective cross section area from cylinder head port to the outlet...or at least the muffler intake. A smaller diameter exhaust header pipe (meaning all of the piping, up to the muffler) can HELP midrange torque, but may or may not hinder top end. It is important, GENERALLY, that the muffler interior volume be substantial, the reasoning for this is quite complex.   In general, a noisy small diameter straight-through muffler does NOT perform well!  Many have replaced the BMW mufflers on the Twin-Shock Airhead models & found that loud exhaust does NOT equate with more horsepower, in fact, performance may suffer....a LOT.  BMW did their homework on the old twin rear shock model's mufflers.   The BMW Airhead mufflers that are under the frame, do NOT perform as good as the old BMW twin mufflers, so BMW made other changes, to help compensate for this.  That help was not perfect, as seen by the hp and torque figures for those bikes....but, it is decent.

Let's get into this a bit deeper:

The exhaust valve begins to open, & the mostly burned gases begin to exit, & the process continues, until a little bit AFTER the exhaust valve closes, due to inertia of the gases. Each power stroke results in a pulse of gases, of uneven shape. Multiple firings of the cylinder means a 'train' of unevenly shaped pulses going down the exhaust pipe. At higher & higher RPM, the pulses get shorter & shorter, but more often. YES, exactly the same type of thing as the incoming air for the intake system. Now, speaking of any single pulse...when some portion of this pulse...for our purposes, let us say the very beginning,,,, reaches the exit mouth of the muffler tip (or unmuffled pipe...etc), the gases, which have accommodated themselves to the pipe system before this, now suddenly see vastly different conditions at that very exit point. The gases see atmospheric pressure, atmospheric temperatures, increased effective diameter of the pipe (NO pipe!!)...ETC.   The gases are confused, & think they have run up against a brick wall (REALLY, a solid wall!!), which then gives way, modestly smoothly, & allows the rest of the pulse to go to outside air. It does not 'give way' instantaneously, it seems so, but is not so. It gives way in proportion to the intensity of the gas pulse, basically suddenly strong, then fades slowly, all this happening in an seemingly instantaneous manner, unless you have instruments to measure it.  At the "gas confusion point", the gases are reflected back up the pipe, all the way to the exhaust valve & even into the cylinder if the valve is open.   The reflection can be full, partial, partial distance, or a combination.    One of the complicating factors is that ANY inconsistencies in smooth pipe interiors will promote a point of reflection, which may be quite minor, or could be major.   This includes the change in effective pipe diameter as the header pipe enters the muffler, or before, crossover pipes with poor smoothness of the curve that should really exist where they meet the main header pipe, ETC.  Every little jog or anything not smooth to the flow creates a change in the traveling exhaust, perhaps a detrimental reversion wave, small or large.  This is just ONE of the reasons that exhaust system design is part art & part engineering.   You might think that 'old-fashioned' exhaust mufflers are for 'show', but a LOT of engineering went into such as REAL fishtail & other 'weird' looking mufflers. 

You can think of the exhaust gases as a series of soft rubber slugs if you wish, traveling in the exhaust system, with a bit of separation between them. They hit an inconsistency...or, especially, the outside air, and SOME rubber bounces back, pushing each previous one ahead of it...and so on. So, what we have is a reversal/reversion of SOME of the exhaust energy. This wave is at low pressure, so you might think it has little real-world effect and that it would or could be a bad thing, except that the pipe length can be made such that a pulse can arrive at the valve at any portion of that pulse, or in-between pulses at any given RPM area.   On a PRACTICAL basis, it would require a much TOO LONG exhaust system to take full advantage of this effect for lower rpm, so the manufacturer must deal with that in other ways.  
Yes, it is true, if the exhaust system was extended, perhaps by a FEW feet behind the motorcycle, the results would be very nice for torque output.

NOTE!  The careful reader may well say, what about using a LONG intake, and a SHORT exhaust (or, vice-versa)?  ...well, that would be workable, but the intake, if long, will have more problems staying cool, and have many other problems, including the FACT that the exhaust must exit someplace not near the rider's knees!  Changes have repercussions. On a practical basis, due to gas expansion, mechanical location problems, and MANY other things, the exhaust system is always longer than the intake system, and the exhaust tuning on 4 stroke engines is usually done mostly by pipe diameter, and one or more crossover pipes, and some muffler innards that are often VASTLY more clever...or, at least, have vastly more real engineering in them, than just for sound reduction. 

In some systems, the exhaust flow starts out with a small diameter and is a constantly expanding taper...this can be made to work very well indeed....but is expensive to fabricate.  It was very commonly done with 2-stroke engines, with many welded-in sections, called Expansion Chambers.  The outlet of such an exhaust system can be a very small diameter, and OFTEN IS!    There is an amazing amount of power produced by the small old 2 stroke engines for their displacement, and the exhaust exit is quite small.
How does THAT fit in with your pre-conceived ideas about muffler and pipes size and 'open pipes' theories?

I have NOT, so far, except in reports of experiments, seen anyone with a truly wide-range dynamically variable tuned exhaust system...although some single bypass systems are in use, and there are now appearing some mechanically controlled ones with more than a fixed one-two change.   If the mechanical problems could be sorted out, it might be possible to combine a variable camshaft, separate pair of intake valves for differing cam lobes, variable intake length and cross-section; together with a variable exhaust, all electronically controlled, and obtain an almost unbelievably flat torque curve.

As you have read, the length and size of intake and exhaust systems have big effects.   In quite a few ways, some of these things use the same formulas and are similar to tuning a pipe organ, at least in some areas.  Engineers use the same formulas for open and closed pipes in organs for some of the needed calculations.

If this pulse (or lack of pulse; that is, in-between pulses) of reverse energy meets another oncoming (from and out of cylinder) pulse at exactly the right moment, it CAN reduce the pressure (or increase it) in the pipe at that particular point, and THIS effect then travels back down the pipe. With a lowered pressure, the exhaust is
extracted from the cylinder far better....and when this occurs with the intake tract still flowing, as it is with valve overlap timing, MORE intake charge goes into the cylinder as well.   Almost like a supercharger.   This means that a tuned exhaust can extract more exhaust AND help suck in more mixture to be burned!  This seems like the best of both worlds, and CAN be, but the effects vary with just about every other thing, including, ESPECIALLY the RPM, but also throttle position, entire intake system design, the camshaft timing for the valves, shape and sizes of things, and dozens of other parameters.   Many decades ago, before computers helped the engineers, all this was done by first cutting and trying (with some engineering formulas).   Nowadays with high speed computers and special instruments, all able to analyze and correlate hundreds of changing characteristics at the SAME time...there is far less 'cut and try'.   However, overall, exhaust and intake design is extremely complicated.    Even today, computers and lots of measurements with sophisticated equipment do not usually tell the whole story, just give one some directions/hints, granted rather good ones usually.

In the INtake system, it is desirable to fashion the system so that the reflected pulse enhances the flow by creating even more of a vacuum than normally would exist; or a longer length of the gases slug.  In the exhaust system, the reflected pulse, from the exhaust outlet tip, is returned to the exhaust valve area, and if timed correctly, will REDUCE exhaust system pressure, and thereby, in a complex manner, allow MORE INtake of fuel mixture to occur, due to the valve overlap period, including the effects of the intake charge pushing the remnants of the diluted exhaust out the exhaust valve...AND reducing 'back pressure', allowing more flow in the first place.  If you are getting the idea that back pressure IN ITSELF is NOT the most important are absolutely correct. The problem is in getting it all to work together, as you change ANY one thing, it seems to change many more things.    That is why well-experienced master tuners/modifiers are in such high demand for commercial racing.   Today's engineers and some master tech's use very sophisticated computerized tools for analysis.  The factories have very sophisticated dynamometers; and methods of changing almost any part of the intake, exhaust, or engine internals, at will.  YOU will not have such a sophisticated multi-million dollar setup.  The factories can change various things until they get the particular performance they want...>>and for touring, that means a high torque output, with a relatively flat curve.

If you consider what you have been reading, then you SHOULD now understand that "Back Pressure" in the exhaust system is LESS important than the pulse timing; back pressure has LITTLE effect until it is really high!  MORE on this a bit further down this page.

Design of TRUE performance enhancing exhaust systems, especially the mufflers, used to be part art, as well as part science.  Today, vehicle manufacturer's have LOTS of pressure transducers....and visual transducers!.... plumbed into prospective 'mufflers' and other exhaust components, and results analyzed on very high speed computers.  

You might be surprised to find out that antique 'fishtail' exhausts (not the much later ones for appearance sake only) can be a good enhancer of performance, as are several other designs, IF properly done for performance, and not for just appearance. 
Generally speaking, it is desirable to have the exhaust system gases constantly expanding, even if ever so slightly, in cross-section (that means larger and larger inside diameter of the round pipe...or, its equivalent in other shapes), from exhaust valve to outlet.  In some designs the exhaust tip outlet is then restricted, changing the velocity, and reflection pulse. Does this begin to sound a bit like an intake system too?

Megaphones of the tapered larger outlet types are not practical on the street, unless internally muffled, which tends to defeat their performance.  A proper megaphone for performance on a stock Airhead would be quite long, with a SLOW taper, and it PROBABLY would stick WAY out and beyond the end of the motorcycle.  SOME types of megaphones have a reverse tapered outlet, which enhance the performance, whilst others are sales gimmicks.  Many folks find that the BMW stock muffler is better, performance-wise, than they thought, after spending lots of $$$ on aftermarket exhaust systems.   SOME of the BMW airhead models with mufflers/collectors located under the transmission can be modified, usually the sound increases some, and the torque MIGHT or might not increase/decrease.  In other instances, you can make some decent, if not huge improvements.   

A general rule is that IF putting other 'mufflers' on your bike vastly improves performance, you likely have a poorly tuned motor to start with.

Most folks seem to have the idea that reducing back pressure in the exhaust system is the holy grail of design. This is NOT so. What is REALLY desired is a prominent and properly timed extraction effect at the exhaust valve....due to the effect this will have on the incoming fuel-air mixture. Typical exhaust back pressure has a rather small effect on power,,,,within limits of course. Perhaps a few more words will help in this area:

Shortly after the spark plug ignites the fuel-air mixture, the flame begun by this spreads rapidly, and the very rapid expansion of gases produces a very high pressure, which is very effective in moving the piston downwards. As the piston moves downwards this pressure reduces by a huge amount. By the time the exhaust valve begins to open, the pressure is fairly low. The pressure then is probably under 100 pounds per square inch; still enough to do some work.  The pressure in the header and exhaust pipe, due to expansion into that area,
is very much lower, especially at low rpm, but still fairly low even at high rpm. It is only a FEW pounds per sq. inch. If you do not believe this, have someone put the throttle up for 5000 rpm, and cover both exhausts with your hands, try to stop the engine, while your friend turns the throttle wide open. There is no way that any internal pressure, without your hands, would in itself reduce engine power greatly, compared to the very high combustion pressures. A horsepower is 550 foot pounds per you REALLY think that some sort of pressure on your hands is relative here? (ok, so that is a bad analogy, so what). It is the effect of the waves, in a properly designed system, that reduces this pressure somewhat, but reduces it AT the valve area, in synchronization with the incoming charge (remember: valve overlap...the intake valve is opening to allow incoming charge while the exhaust valve is still partially open), thereby allowing MORE mixture to be packed into the cylinders.  THAT 'packing of mixture' is where the largest portion of the power increase is noted, with a good exhaust system, from the improvement in the force on the piston during the early portion of the power stroke......AND HOPEFULLY CONTINUING DOWN THAT STROKE A LONGER DISTANCE THAN PREVIOUSLY.  Just a very small amount of lengthening of the amount of the stroke that usable pressure is on the piston makes a BIG power output difference....but there are practical limits.

Many decades ago, 'boxes' were put where we now think of 'muffler', and this box, which was NOT a TRUE expansion chamber, collected the exhaust mixture, quieted the exhaust, and had not much restriction.   Early engines mostly used that system.  An improvement on this was the Brooklands 'muffler', which developed into the full-blown type called a 'fishtail'. This design used a slot (unknown to many who have seen fishtail mufflers...likely because those in the U.S. are GARBAGE designs, with design mostly just for appearance-sake and are all without the needed slot) in the outlet, the resultant of which was that the AREA of the outlet was somewhat LESS than the input pipe area...which reduced the sound level.   (hmmm...whatcha think about the smaller inlet size of snorkels now??)

Due to very complex waveform development inside the design, the fishtail muffler had a greatly reduced pipe pressure, compared to an OPEN PIPE!  In fact, it was roughly 50% better! YES..this means MEASURED RESTRICTION meant LESS restriction in OPERATION!  If the fishtail was removed, and only the 'box' used, the results were in-between an open pipe and with the fishtail.

The very next step in early design was the adding of smallish holes in internal baffles.  This caused many complications, which were eventually solved.   In fact, it was these complications that needed fixing that brought about a large amount of knowledge about muffler, collector, and pipes design.   It was now possible to have a exhaust system that muffled AND performed. I know this sounds absolutely wrong, common sense seems to tell us that restrictions ALWAYS mean lower power, but it is true.

In the intake system, the reflected pulse can be timed, within reason, to boost effective amounts of actual cylinder filling fuel/air mixture. This effect is why RAM TUBES are seen on so many vehicles these days. They lengthen the intake system, adding more power robbing drag (from wall effects, and hopefully keeping laminar flows), but boosting far greater the reflected pulse effect. Well-designed ram tubes can work well, whether between carbs/FI and head, or carb/FI inlets and intake of aircleaner.  Ram tubes generally allow more radical cam timing, without the engine becoming too 'peaky'.

BMW factory-sold 'sport camshafts' are rather versatile, performing fairly well without any necessity for radical camshaft timing, (radical camshaft timing usually results in a very rough loping idle and VERY poor low end performance), and there is no dangerous super high valve lifts either.  Use of the factory sport cam with a compression increase is helpful.  It is also helpful to reduce the clutch/flywheel weight, to allow the engine to accelerate a bit faster, although this has some drawbacks such as more vibration.   Some sporty camshafts, the BMW mild one included, EFFECTIVELY raise the dynamic-effective compression ratio DURING high rpm operation. I will NOT get into why...there are several effects.  You can NOT just change the camshaft in your Airhead, without other work.  If you do install such as the BMW sport camshaft, be prepared to do LOTS more, or your lower RPM performance will SUFFER.   The BMW sport camshaft will considerably detract from performance if you do not.  The sport cam will STILL only START to "perform" above 5000 rpm....typically the improvement is mostly notable from 6000 or so.  I can NOT recommend the BMW sport cam (often called 'the 336") for any sort of street Airhead for pleasurable touring characteristics.  For a Cafe Bike, OK. For racing, yes.

Other, hard to quantify effects, come from combining the intakes and/or exhausts. In our BMW bikes, the intake filter housing feeds BOTH carburetors. Hence there is now a complicating factor (VERY complicating) of the unequal 'sucking' strokes/pulses of the two out of phase cylinders, and the effect THIS has on the incoming wave charge. AND, that effect varies with throttle opening and rpm, ETC.   However, properly designed, the effects can IMPROVE THE TORQUE CURVE.

A sales department telling engineering management that 'a two into one exhaust is needed for appearance-sake', for an upcoming model means hundreds of test and design hours, on intake and exhaust!   The sales department was probably responsible for the rear disc brake on the older RT and RS models; the brake is not very good compared to the drum version.

What all this means, on a practical basis, is that you really need a dyno to REALLY hop up a bike....OR, someone who already has done exactly this, or has been very lucky, to give you all the exacting and exciting details!!!

Now, specific recommendations, that I feel I am safe in making:

Before we get into things, a few flat-out statements.  These apply to stock, relatively stock, somewhat modified Airheads, and fairly well modified Airheads.  In every instance I mean for the street....NOT for full-out racing.   It makes no difference the engine size, compression ratio, piston changes, milling the heads; nor, making carburetion changes, nor even installing the BMW sport camshaft; these comments apply to all:

Use of a two-into-one exhaust system will likely REDUCE performance.  The Super-Trapp type of exhaust is tricky, hard to get it to perform correctly, and is VERY likely, even with large numbers of discs, to perform MUCH less well than the stock BMW exhaust.  Use of K & N air filters is NOT going to improve performance, and WILL decrease engine life. Use of individual air cleaners, screens or filters, of any type, at each carburetor throat, instead of using the stock air cleaner system, will CAUSE PROBLEMS.  The engine will be less tractable, due to the bad effect on the flatness of the torque curve.  In some cases the weight will cause vibrations that will cause the fuel to FOAM.   To get around this, the proper type of rubber intake hoses (especially durometer rating),  & mechanical support for the carburetor & perhaps the individual air cleaners/screens is needed.  More power IS available from individual air cleaners, but the torque peak is moved upwards in rpm, making for a MUCH more peaky engine. are some discussion points:

1. Camshaft:
A BMW sport camshaft can be used. The BMW Sport Cam has a deleterious effect at lower rpm. This is a timing cam, not a lift cam, and use of higher lift, by incorporating high ratio rocker arms, is probably possible, but one must be VERY careful.  Pay attention to the rocker arms, pushrods, piston interference, etc.  The performance improvement will be best above 5700 rpm.    This means that without a LOT of other changes, you will LOSE performance lower down. There are companies in Germany making a variety of camshafts for the Airhead.  DO consider them.

Compression ratio: the higher the better, within reason for fuel, etc. On a practical basis, probably about 9.5 is the highest with a stock camshaft, and maybe 10.5 with a sport camshaft. This includes dual-plugging.  With a really clean combustion chamber with no sharp places, and everything tuned to the Nth, even 11.0 might be possible on 98 octane or better, with dual plugs. The problem is mostly the octane of available gasoline. If you have the time and do not mind experimenting, the squish area, as in the 1977 airhead, can be played with, with some improvement available if it is reduced, or re-incorporated. One problem in any hopping up, is that if you are using a R100 engine, it is already running quite hot. If you are willing to do valve jobs more often, you can go pretty far, approaching early Oilhead output.   The larger valves of the early models, with the larger carburetors, are going to be needed with higher compression and a sport camshaft.    A good 3 angle valve job, ...careful valve guide work, ... are all necessary, except for the most extreme motors, where polishing, reducing weights, shaving guides, modifying lifters and pushrods, etc., are helpful.   I do not think going to a larger bore via special expensive aftermarket kits is a good idea.  BUT...if you already need cylinders, maybe going to 1050 cc or so is OK, but do NOT increase the bore on the will probably be UNreliable.  Keep in mind that going over 9:5 in CR, or going to large bores, will, or can, cause reliability problems.   Very careful attention to the valves, valve springs, etc., is going to be NECESSARY.

Flywheel, clutch, transmission: The early heavy flywheel/clutch assembly is fine for most uses, but not for best acceleration and fast shifting, where lighter items are better, but give more vibration, which can be somewhat (ONLY somewhat) reduced by balancing.  Even the later 'flywheel' (carrier) and clutch is too heavy for the most spirited sport riding and racing. Some of the later clutch parts are well known to us wrenches as going out of balance, from warping (like the plate ring), so balancing is relatively important.  Vibration robs power, not just annoying to you physically.   SOME may want to go into the transmission and modify it for easier/faster shifting (particularly at highest rpm).  There is an article on this website on how to PROPERLY lighten the flywheel:

4. Intake system:
you should consider modifying for cold air TO AND THROUGH the airbox.  Except for a near race or actual race motor, OR A CAFE BIKE, I suggest you NOT eliminate the stock airbox entirely....that is, I suggest you do NOT use individual screens or filters right at the carburetor mouths (thereby not using the airbox at all).   If you do go to individual filters, AND ARE AFTER PERFORMANCE, NOT JUST LOOKS, BE SURE TO DO some work with tufting or smoke, to ensure a good airflow to those filters, and that airflow coming from the forward area is not being curved excessively, and creating a light vacuum at the filters.  IT IS VERY IMPORTANT THAT THE ADDITIONAL WEIGHT OF THE AFTERMARKET CONE OR OTHER TYPE OF INTAKE FILTERS (MOUNTED AT THE CABURETOR INTAKE) NOT VIBRATE, WHICH WILL FROTH UP THE FUEL IN THE FLOAT BOWL AND CAUSE YOU ENDLESS PROBLEMS BECAUSE YOU DON'T KNOW ABOUT IT AND TRY EVERYTHING ELSE AS A 'FIX'.

With regards to cold-air:  Avoid effectively lengthening the intake system, unless you WANT to lower the rpm for torque peak.   If the cross section areas are too small you will run into non-laminar flow effects, so avoid this. You want to avoid supersonic velocities. You MUST run tests to be sure that the area of cold air pickup really is in an area that is not affected by the front end/fender/etc...and surrounding parts...that can actually cause loss of effective air pressure there. AGAIN, generally unless engine is heavily massaged, use of the stock airbox, modified for smooth flow and cooler air, is recommended. I have reservations, however, about using the clamshell box...and think that it should be eliminated.   Get rid of ALL smog items under the air filter on the later models. Rework that airbox for a SMOOOOOTH flow of air internally. Yes, you want turbulence, but you want it AFTER the carburetor...INSIDE the cylinder! Insulate as required with cambric, etc., so that as little as possible hot metal comes in contact with the incoming air. HINT: Sometimes you will get a bigger improvement from making the intake system smaller, IF it remains cool! If you keep the intake system from being excessively heated by the engine you will gain a LARGE amount of power. The power gain from cold air modifications is REALLY significant. This was one of my personal secrets.

Snorkels: They can be cut back, a bit at a time, or, removed entirely, and you will see the BAD effect any shortening usually has on a stock machine.  They can also be opened, but not much, unless you are going to very high rpm consistently. Shortening them will move the torque peak upwards.  Generally, if the carburetor size is stock, you can do minor mods on the snorkels on both the R80 and R100 machines, but watch for loss of midrange torque, which you will want to make up elsewhere if you go too far. If your engine has two sizes of snorkles, try using the larger one for both. Try the reverse too.  You will need jetting changes for each change.  Gas mileage can suffer if the rpm position for the snorkel effect is off, BMW found that out with the R80, and R65, and then modified the snorkels with holes...and later, UNequal length snorkels. For machines WITH a fair number of modifications, you likely will want to chop off enough of the snorkel to increase the cross section area of the intake, but do not go too far.   As you chop the snorkel, you will greatly affect low mid-range.  I have found that putting some holes in the inside length of the snorkels works well even with the stock lengths...but while this is true, some turbulence is generated, and cutting the snorkel a inch and a half or so, and bell mouthing the remainder, works very well, and far better than the holes.  If you are truly anal, you might try modifying the two snorkels to one larger one, and making it adjustable in length...and somewhat in cross-section.  This is NOT all that difficult!

5.  Ignition: Dual plugs.  For racing, or very spirited riding at VERY HIGH rpm, single plugs MIGHT be better, this is NOT cut and dried; you may run into octane problems if your CR is high, but this problem tends to fade at the highest rpm to being almost non-existent.   The stock BMW ignition systems are fully capable of good performance. The NON-canister points models are not so good at 6000+ rpm.   The points canister models are probably OK to 7300.   Install Porsche points in the canister and you may go even to 8000 or so, but the latest stock BMW numbered points for the canister are EXCELLENT for high rpm.  The BMW stock electronic ignition as far as Hall device, module, and coil(s), are quite adequate for 8000 rpm. Only the most modified engines require tighter controlled and more powerful ignitions. The spark plug resistance caps can be eliminated on the NON electronic models, although I think you will find that 1000 ohm caps are nice for spark plug gap longevity, and I think the spark waveform is better, than with zero ohm caps, so I would NOT eliminate the 1000 caps.   You can NOT use 1000 ohm caps on the electronic ignition models, without some danger of frying the ignition components.  I think it can be done with some shielding and filtering, but have NOT tried it, nor experimented with it, which I once planned to do with my oscilloscope...the idea being to prevent any voltage spikes getting to the Hall elements.

The ignition curve should be modified to match the rest of the engine, particularly if the engine is heavily modified. On a mild performance enhanced bike, this can be as simple as bending the ears in or out a bit, and/or using a slightly stiffer spring on one weight....on some early bikes it will be simply using a different model of advance unit. Restricting advance can be done with shims too.  The combination of points and all the various automatic advance models and springs that have been used on the /5/6/7 models, allow quite a few versions, and might be adequate to make up a matching advance, or you can modify the weights.  One can certainly modify the springs and weights to get just about any curve one wants. Frankly, I think that at this point some dyno time will allow one to make a custom advance curve.   One or two springs and sometimes reshaping of the weights are all that is really usually necessary.   For a RACE bike, all this is moot, as you are nearly always at full advance; and, probably have a performance ignition driven off the crankshaft for best stability, and near perfect ignition as the timing chain and sprockets, tensioner guides, etc., wear.

One final word about the automatic advance:   The best power and performance, IN GENERAL, assuming that the octane of the gasoline is adequate for the compression ratio, etc., will be obtained with a fairly fast advance curve.   Granted that very little power is on hand below 2500, still, a fast advance will help acceleration from the low rpm point.   If possible, the use of the stock early /5 advance unit, or springs (both for pre-1979 cams), or simply reducing the weight of the stock ignition advance unit weights (sometimes just one weight or spring), allowing a faster (lower rpm for maximum advance) advance, will be of some help. These things do not help but a rather small amount.

With a VERY hotted-up bike I simply suggest purchase of one of the CRANKSHAFT-triggered aftermarket ignitions.

Addendum #1:
Modifying the stock early mufflers (twin shock models) for SOUND & RATHER SMALL power increase:

1.  On the NEARLY FLAT REAR FACE of the stock muffler, drill 1/4" holes, TWO places, one at top, one at bottom, centered in the flat area.

2.  Obtain an old drill bit meant for metal, not wood, of about 5/8 to 3/4" diameter size.  Heat soften the shank end for an inch or three, and BRAZE or WELD this bit to a long piece of material. Be SURE the brazing or welding is secure.  If you have a bolt-together drill bit extended used by some electricians in house wiring, you can use that, but be SURE the bit can NOT become separated from the is then difficult to remove from the muffler innards! Drill from the muffler OUTLET end (nothing to do with the holes in 1.), into the way far deep inside center, drill it through. (if you were clever, you could do this from the inlet side, easier...(not always!...the mufflers VARY in construction at that end)...but IF you CAN drill it there, try it....the problem is usually that you see a dome, hard to make mark or drill a dome when inside something.

3.  At the rear outlet, at both 9 O'clock and 3 O'clock, use a punch to punch-prick a starting tit, INSIDE the outlet, about 3/4" IN from the flat face....and drill on an ANGLE, leaving the drilled part moving the electric drill motor sideways, after the hole is first drilled through.  This makes the final inside hole sort of elliptical, by means of the side flutes cutting the metal of the muffler. Do NOT overdo this.  The drill suggested for this elliptical hole is 1/4" to start with....up to 3/8" maximum.

4.  Make no jetting change unless you are already on the very edge of being lean.  If the leanness is for wide-open-throttle, go up '5'  on the main jet.  If the leanness is in the mid-range, go to one step larger on the needle JET (NOT needle position).  Check for recommendations for the UK version of your bike (assuming SAME size carburetors, 32 or 40 mm).  If not loud enough, add TWO more holes, 90 degrees to the others at the outlet flat face, but NO MORE elliptical holes.  If still not loud enough, change those 4 flat area holes to about 3/8".  This mod helps power SLIGHTLY, moves the torque peak up about 300 rpm,  is definitely louder....all with stock appearing mufflers to "most inspection folks".

5.  For those that want a slightly throatier sound (this is for a street bike, not racer), do the 5/8" way-deep hole, and NO elliptical holes, and TWO each 1/4" holes...holes at top and bottom, not sides.   

An extra advantage is that condensation inside the muffler is GREATLY decreased, and the muffler will last for many decades.

Addendum #2: 
Krauser 4 valve heads:
I do not want to get deeply into these here. I will just make some quickie generalizations.  They are tricky to get to work properly because they were never fully machined and completed.  Once the work is done, they can work well, but may need more fiddling with such as valve clearances, etc.   They DO offer increased intake area, and WILL boost horsepower output a fair amount.   Here is my article on them:

Addendum #3:
Premium fuels versus mileage and combustion temperatures.
There is an old controversy over possible increases in combustion chamber temperatures, ETC., when using premium gasoline's in lower compression BMW airhead engines, where lower octane is supposedly not needed.  Gasoline burns at about the same rate under normal, that is, not detonating, etc., conditions. The output (BTU) per gallon of Premium gasoline is potentially....or even be a tad LOWER than for Regular. Therefore, I think it is likely that SOME premium gasoline's WILL give LOWER gas mileage than a regular gasoline will.....assuming here that the engine will run properly on Regular grade gasoline in the first place.   Temper this with the fact that gasoline formulas change, and also change between summer and winter grades of gasoline's. Winter gasoline tends to have rather volatile things like butane or propane in them...etc.   In the West, California in particular, oxygenates were added to most fuels for decades.   These additions GENERALLY cause ~5-10% POORER gas mileage.  They are NOT good for your engine, carburetors, hoses, etc. Nowadays, ethanol is added to nearly all road fuels.  Without jetting changes, they make an engine run leaner.  The fuel mileage decreases even more, perhaps another 3 to 5% when you rejet for performance with such fuels.   When dealing with this subject on a practical basis, one of the first things to check is the EURO (UK) jetting and needle position, for the same size of carburetor.  The UK jetting is almost always better, richer, for performance.

Addendum #4: 
The old 'bible' book on tuning.  Still quite good, even now:
"Scientific Design of Exhaust & Intake Systems"; Philip H. Smith and John C. Morrison.

08/04/2004:  add Addendum #1
12/15/2004:  clarify and fix typos.  NOTHING substantially changed.
12/17/2005:  Add addendum #2
08/04/2006:  minor editing
06/16/2010:  recheck article
06/05/2012:  recheck article, clean up a couple of typos, layout changed slightly. NOTHING substantial.  Links not checked.
10/06/2012:  Add QR code; add language button; update Google Ad-Sense code; update article
11/28/2012:  Minor updating and cleanup.
10/02/2014:  Add Addendum #3, which used to be elsewhere's on this site.
02/19/2015:  Add Addendum #4, for the book.
Spring, 2015:  Began an update, intending to add more testing information.  Decided that I will never get to it, so updated what I had, 09/18/2015, due to my responses in a Facebook thread.
11/18/2015:  Change font for clarity, clear up a few details.  Recheck for the MEZ article, still no find.  Fix scripting.
08/03/2016:  Recheck MEZ.  Update metacodes, scripts, layout, clarifications. 
09/02/2016:  Fix two typographical errors in Addendum 3, and one grammar error in same.

R. Fleischer

Return to Technical Articles List Page

Return to HomePage

Last check/edit: Friday, September 02, 2016