Airbox design factors vol.1 : resonance


I have been reading about airbox design and resonance from Kevin Cameron for a while now, and in particular within the last week.   He maintains that the resonance effect in an airbox can be quite powerful-  up to 10-15% extra cylinder filling when the airbox is in 'resonance'  and that resonant airbox tune is of more significance than any ram air effect.

I finally tried to get serious about this stuff and have been doing a bunch of math (which is giving me a headache)  trying to model various  resonances that one might apply to a v-twin,  which has such an irregular intake beat.

Also, I searched out and bought several excellent SAE papers on the topic of airbox resonance-  several by Robert Fleck of QUB who you may recall published many of the cool 2-stroke tech SAE papers.

At any rate, they modeled various airboxes in F1 applications and built a test box around an R6 engine to predict and then measure pressure activity inside the airbox.

They modeled the airbox in several ways , from very simplistic (assuming uniform pressure thru the entire volume)   to very complex, modeling the airbox in its entirety using CFD (computational fluid dynamics) and also testing with various pressure transducers.

Everyone has been assuming the simplistic behavior "well, it pulses at such and such a frequency"  but in reality the inside of the airbox is a pretty busy place.   And during all this pumping and wave action, quite often adjacent cylinders make more or less power because of wave action or localized changes in density because of what is happening from the cylinder next to them.    Even connecting a *single* R6 cylinder (to avoid all the noise caused by 3 other cylinders in there trying to pump) to a box that resonates at 75 Hz, then running the engine at the exact coinciding  RPM,  the pressure inside the airbox was not fluctuating at +/- 10-15% above atmospheric pressure, which is the oft-quoted figure.

in SAE  2002-01-3318 Fleck is at it again, with exhaustive analysis of the inside of an F1 airbox.  what is interesting is 2006-01-3653.pdf  where they study effect of airbox resonance specifically.   On the testbed R6, running at the resonant freq of the test airbox,  8800 rpm, the measured pressure fluctuations were pretty mild.

+/- 1 KPa



The test above has one cylinder connected to the airbox, and driven at resonant freq, to eliminate noise from the other cylinders' influence in the airbox


That's +/-  10 millibars, which is the kind of ram pressure you can get at 80 mph on the ZX9R- but the ram pressure is valid at any rpm range.    Airbox resonance ONLY works at that set RPM and to a lesser extent, at 2nd and 3rd harmonics.


I dunno if I have the instrumentation or smarts to test this on my own- my dyno is a crude instrument, but I plan to test various airbox inlet ducts-    so I'm leaning on the SAE stuff!




Short story:  I am beginning to think the "tuned airbox" business is akin to a boost bottle-  effective to fix a dip or hole, but not a huge power producing device.   I think the main airbox benefits are


1) noise reduction

2) good home for an air filter

3) provides benefit of cold air intake (1% power change for every 10 degrees F intake air change)

4) allows ram air which may not be all that huge (compared to atmospheric)  but is substantially better than a non-ramair box, even by 60 mph.


That's where we get the real benefit.   I do not see where an airbox that resonates at a single frequency is going to boost power anywhere except at one RPM zone.


Anyhow, now you know how I spend my winter!   Now I have to go see if this TLR swingarm will bolt onto my TLS.  the new airbox is almost done-    and I shortened the throttle bodies... we shall see,  we shall see...


And, yes!   pressure fluctuations-   on a v-twin, the airbox is a whole different animal.   It's crazy in there.   my RS250R has a 12-liter airbox-  yowee.   F1 cars typically have a 50 liter airbox atop a 3 liter engine.


Here's the deal.  With the intakes 270 degrees apart, the 2nd cylinder in a v-twin takes a breath from an airbox full of less-dense air, than the first cylinder did.   To simplify a bit...


Assume that the airbox is NOT replenished-  you just have to suck air out of a sealed airbox with no inlet hole (we'll look at one revolution, of course)


a 12.5 liter airbox = ducati 999

at 1 atmosphere pressure-    suck 500cc into your intake, leaves ~ 0.96 atm in the box.    pull another 500cc intake, and leaves  ~ 0.92 atm in the box.   So even after the engine draws a full 1000cc breath from the airbox, it still contains 0.92 atmosphere of pressure.


How about a 10.5 liter box?   RC51


A 10.5 liter airbox  at 1 atmosphere pressure-    suck 500cc into your intake, leaves ~ 0.95 atm in the box.    pull another 500cc intake, and leaves  ~ 0.90 atm in the box.   After the engine draws a full breath from the airbox, it still contains 0.90 atmosphere of pressure.



8.5 liter airbox?  (TL1000S)


an 8.5 liter airbox  at 1 atmosphere pressure-    suck 500cc into your intake, leaves ~ 0.94 atm in the box.    pull another 500cc intake, and leaves  ~ 0.88 atm in the box.   After the engine draws a full breath from the airbox, it contains 0.88 atmosphere of pressure (there are some errors here because the engine is drawing ever-thinner air from the box, but I'm keeping it simple) .



This is super-simplified, with the engine pulling from a completely sealed-off (no inlet) airbox.



 I just do this to get a feel for what's happening.   Looks like in that example  the 12.5 liter airbox delivers 4-5% more air density at the end of the event, than the 8.5 liter airbox.


To add a layer of reality (and complexity) , the airbox is getting replenished via an inlet port (or ports).   The flow into the airbox, in the absence of resonance,  is controlled by the size/restriction of the inlet hole and by the pressure drop across the hole (airbox pressure vs external pressure)


Get this! A smaller airbox can get along with a small inlet port, much more easily than a large airbox- because there is a larger pressure drop in the small airbox.   In just the same way that an engine with high vacuum can work with a smaller carburetor, than an engine with low vacuum.  The engine with low vacuum MUST have a larger carburetor, just to have equal flow through the carb.


In a steady-state model- where the engine sucks air smoothly from the airbox in a continuous draw, and the airbox is replenished smoothly with constant flow from the airbox inlet, the airbox will drop to a certain level of vacuum and stay there.  That level of vacuum is decided solely by the restriction of the inlet port to the airbox- having nothing to so with volume.


So a big or small airbox (say, on that ZRX1100) would both fall to the same amount of pressure.  Again, twins are goofy, because the intake is so asymmetric- cylinder #2 gets really robbed if the airbox has been momentarily sucked down to low pressure by cylinder #1.  So, bigger is always better.   Or maybe another way to say it, is a 270 degree twin suffers more from a small airbox, than an i4 or a 360 degree twin.


Kevin Cameron says, what is cool, is that at certain RPM-  the resonant RPM of the airbox-  determined by box volume and inlet area/ length  -   the intake restriction acts like IT IS NOT THERE.    For a certain span of RPM, you get- pretty much-  free airbox filling.   We can choose the RPM at which we get this free filling-  where would you want it?  Streetbikes usually have it tuned to reduce a dip in the torque  curve in the middle RPMs, before the engine comes "on the cam".   If I were a racing sort I'd tune it to resonate at an RPM where I'd be racing it.


However, the SAE papers I have been reading do not seem to bear out any large improvement in cylinder filling at a particular RPM due to airbox resonance.   Although, the resonance is there, to be sure.   However, sometimes it helps some cylinders and sometimes fights against others.


V-twins huff and puff pretty hard.   They need big airboxes.



BUT, KC is persuasive and still feels that airbox resonance is a significant player.


I hope to dyno test this on mine, by playing with various inlet tube lengths.   In fact, the TL airbox , stock, really is by the book for a resonator.   It is 8.5 liters volume,  and has a pre-chamber before the airbox that the ram ducts feed into.  The ram ducts do not go directly into the airbox.    The pre-chamber and the airbox are connected by a small duct,  6 cm in length and with 31 sq cm of area. Exactly by the book, for a resonator.  It is resonant at ~ 95 Hz.   with a 360 resonance (excited by a single cylinder- which oddly enough was what the SAE paper from QUB was saying) this works out to 5700 rpm-  and KC says that the box is usually tuned to iron out a flat spot or dip in the powerband as the bike comes onto the cam.


360 resonance in a twin, unfortunately, looks like this- the red humps align with a twin's exhaust- or intake pulses.  so cylinder #2 is really getting screwed.





720 degree resonance in my bike would look like this-   seems much more reasonable.


Anyhow, what am I doing??


I found an article from some guy, who knew some tuner, who was shortening TL1000S throttle bodies and getting 130+ hp with slipons and a fuel map tweak.   Now, that sounds dandy but the guy is not online anymore, can't find the tuner, etc .  Still, it makes me curious.


The TL-S has rather long throttle bodies-  I called up my RC51 buddy and mine are at least 30mm taller than his.   So I bought a spare set of TBs and chopped them down an inch.



While I'm at it, I'm making the airbox volume larger.   I can't resist.  So I hacked the tock box on half, and am sealing it to the underside of the fuel tank instead, and out to the sides of the frame rails.   I'll measure when done, but I should have 11+ liters.   I know, I'm making multiple changes at once, but I can't go test each little thing.   Just no time, and my dyno guy is hours away.


Well, still there has to be a duct into the airbox.   So, I'm thinking, why not fool with the inlet duct thingy.


The stock TL-S makes good torque till 8000 and then nosedives after that.    I'm thinking- what if I re-tune the airbox to "help out"  at something higher, like 8500 rpm?  Will I get a big hole at 5700 now?   Or a little hole? Can I make a dyno pull, slip in a longer inlet pipe,  and re-dyno, and see a change in the power output?  Will I get a bump here or there?


A corollary to all this is,  right now I am using a modified stock airbox.   I've gutted the flapper assembly (like every other TL owner)  and  cut that fancy little 6-cm inlet duct right out of it.  I just gave it a bigger hole.  and, you know what, I have more power everywhere and no dip at 5700 like one might expect, after I got rid of that resonator duct.  nice smooth dyno chart.


So... is it all BS?   Because I certainly didn't notice a speck of difference after getting rid of the "resonant"  thing.   More air gave me better power.


Guys have dynoed their TLS with the airbox lid off, and seen more power everywhere-  HOWEVER on a dyno with no ram air supply, they are missing half of the benefit of the airbox.


Now, the variable-length inlet stacks on the R1, very cool.   Like a poor man's F1 intake.   So why don't they have variable-geometry airbox inlets?



Lotta things to consider when fooling with an airbox.