Airbox flow testing

[Aussie_T100]

I posted this on another bike forum several months ago and received much positive and some negative responses to it however I feel that the information is still relevant so at the risk of once again exposing myself to the naysayers I've decided to post it here.

Last year I spent much time reading various Triumph forums researching performance modifications for my 2008 carbed Bonneville and analyzing the results of modifications that

others had made to their Bonnies. One of the most common modifications is to remove the airbox and fit pod filters and this is usually done using one of several commercially

available airbox removal kits. The price of these kits seemed a little high to me considering that they are principally comprised of a piece of basic metal fabrication to

replace the airbox, a pair of pod filters and a couple of other bits and bobs. The cost of importing one of these kits from the US is not inconsiderable so I was reluctant to

go down that road. Several Bonnie owners had experimented with various modifications to the original airbox to improve airflow and had achieved positive results however there

was much debate over how effective these modifications were in comparison to just ditching the airbox altogether. Nobody had ever done a side by side comparison, either by

using a dyno or a flowbench, that demonstrated just how effective the various airbox mods were in comparison to pod filters.

Thinking about this conundrum I devised a testing method that may go some way at least toward settling this debate. Obviously modifying the airbox, tuning the bike and doing a

dyno run then fitting an airbox removal kit, re-tuning the bike and doing another dyno run is both time consuming and expensive so what was needed was some way of flow testing

both a modified airbox and a set of pods to compare their flow rates. Not having access to a flowbench I set about building a homemade device that would perform this task.

This is the first part of the device, it is basically a plenum chamber made from PVC pipe and fittings that is designed to fit up to an airbox in the following manner.

Attached to this plenum by an 8mm clear plastic hose is a homemade manometer.

The purpose of this manometer is to measure pressure drop over the various filters or devices being tested, that is, it compares the pressure inside the plenum chamber to

atmospheric pressure and registers the difference. The greater the difference in pressures the more restrictive the device being tested is. This restriction to airflow is

measured in millimeters of water lift, the higher the lift the greater the restriction.

Now all we are missing is a good source of vacume or airflow and for that job I chose two of these.

This vacuum is rated at 65 litres per second by the manufacturer and the device can be fitted with two of them so we have a potential flow rate of 130 litres per second. For

comparison a 900cc 4 stroke engine spinning at 9000rpm requires 67.5 litres per second of air at 100% volumetric efficiency so the two vacs provide more than enough airflow for

the purpose of this test.

To maximise flow the filters were removed from the vacuums.

All of this connected together for testing looks like this.

With one vacuum connected and the other vac inlet plugged or like this with both vacs connected.

Before detailing the tests I suppose I should describe the various airbox modifications that have been found by others to increase airflow:-

1. Snorkle removal.

2. Enlargement of the inlet hole, left by removing the snorkle, to the same size or larger than the filter inlet.

3. Removal of the internal baffle from within the airbox.

4. Making a second entry to the airbox on the right side and opening up the corresponding side of the filter element to match. Usually a foam Unifilter is used because it is

easier to get a seal against the newly opened end of the airbox with this type of filter.

5. The fitting of venturi or ram tubes to the carb inlets to replace the original rubber connecting hoses that connect the carbs to the airbox.

These mods apply to Bonneville, T100, Thruxton & Scrambler models as their airboxes are all identical.


The most commonly used pod filters on bikes with the airbox removed are oval K&N's (RC-2890 & RC-2900) and it is against these filters that the airbox mods will be compared.

I then procured the necessary components for testing.

My first test was to get a reading from the bare plenum by itself with no filters or airbox connected, with one vacuum connected the manometer read 6mm of water lift and with

two vacs connected the reading was 34mm of water lift.

Next I fitted the ram tubes to the plenum and recorded 6mm of water lift with one vac and 24mm with two vacs connected so at the higher flow rate the ram tubes had dropped

water lift by 10mm but at the lower flow rate they had no measurable effect.

Next I fitted the K&N pods which had been freshly cleaned and lightly oiled with K&N filter oil. The readings were 16mm with one vac and 56mm with two vacs so obviously the

pods don't restrict flow much at all.

Next I fitted the standard airbox complete with a new Triumph paper filter and the snorkle. The readings were 114mm with one vac and 360mm with two vacs. That's pretty damn

restrictive you have to admit.

Next I removed the snorkle - 46mm with one vac and 164mm with two vacs. That's a major difference just from removing the snorkle.

Then I opened up the primary intake hole in the filter cover like this.

Readings then dropped to 34mm with one vac and 130mm with two vacs.

I then removed the internal baffle from the airbox so now we have both the snorkle and baffle removed.

Readings were 16mm with one vac and 70mm with two vacs.

 

[Aussie_T100]

Airbox flow testing continued

I then made a secondary entrance to the filter on the right side and swapped to using a UNI filter with a corresponding hole made in it to match the new secondary opening.

Readings are now 16mm with one vac (no change) and 60mm with two vacs so we have reduced the airbox restriction from 360mm to 60mm at the higher of the two flow rates with the

modifications so far and we are very close to matching the pods at 56mm.

Next job was to modify the airbox by removing the rubber intakes from the front and adapt the two ram tubes to it. Firstly I made a jig to hold the airbox and plenum rigidly

with the ram tubes in place.

Once setup with a good seal between the airbox and ram tubes I ran the test and the results were:-

With one vac connected 16mm water lift - no improvement.

With two vacs connected 56mm water lift - an improvement of 4mm less water lift over the previous mods.

Test results summary:-

Results expressed in mm of water lift. Low figure is @ <65L/S, high figure is @ <130L/S aprox.

Standard airbox with Triumph Filter 114mm - 360mm
As above with snorkle removed 46mm - 164mm
As above with primary opening enlarged 34mm - 130mm
As above with baffle removed 16mm - 70mm
As above with secondary entry added 16mm - 60mm (oiled Unifilter)
As above with ram tubes added 16mm - 56mm

K&N oval pods 16mm - 56mm (oiled)


In conclusion this testing method at least would suggest that the stock airbox can in fact be modified to flow just as well as K&N pod filters.

The actual results, in seat of the pants testing, achieved by those that have fitted ram tubes to their bikes appear to be greater than that suggested in my tests (a mere 4mm

drop), it is possible that this greater increase is the result of the intake tract being closer to an ideal (tuned) length with the ram tubes fitted whereas my testing only

measured the increase in airflow in a constant (steady) flow scenario. In other words in my testing scenario the ram tubes were not able to actually work as ram tubes as they

would do in a "real world" pulsing airflow.

Another point to consider is that the K&N tests were carried out in the open, when fitted to the bike pods are quite crowded due to the proximity of the central frame tube and

the side covers so it is possible that their pressure drop would be greater in situ than shown in my testing. One other critical point is that since I did this testing other

Bonnie owners have discovered that UNI pods flow significantly better than K&N oval pods and that DNA brand pods flow even better again so even if you are going down the airbox

removal route filter choice is critical.

One final point that I would like to make is that in much of the discussion and debate on other forums regarding intake mods to our Bonnies very little consideration is given

to the actual filtering capabilities of the various options which seems a little odd to me considering that the discussion relates to road bikes in long term use not short term

use race bikes.

Enjoy your ride which ever way you chose to mod it.

 

[BlueJ]

Aussie: I have this to say about that! Nice work!

 

[outwestrider]

aussie t - thanks for esplaning zis - most excellent ... mjs

 

[klink]

My hat is off to you. You sir are a rocket scientist. Thank you for such an in-depth look at an on going controversy with an at home approach. Sticky in my book.

 

[jfenton]

That is the best information I have seen to date about airbox mods. I used to work for a company that made aerospace fans, and you test setup is very good. An finally we can compare apples to apples with all these mods.

 

[KingBear]

So what are you trying to say? ick:

 

[loxpump]

Again hats off to your work. Do you think if the air source on a moving bike would change the results? Still air compared to turbulent.

 

[mark66]

Great work! That really straightens things out a lot for me. I've been mulling this over and trying to research it for weeks. Thanks again. Ride On!@

 

[samc]

Excellent! It seems that the cheap stuff -- remove the snorkle and baffle, open up the intake hole a bit -- gets most of the way to the expensive stuff and retains the stock filter. The baffle in particular is interesting, I've long believed that its sole purpose is noise reduction, since the chamber next to the intakes is too small to be an effective dead air source, and it just kills performance at the top end, but doesn't seem to have much effect at low and mid revs where most noise testing is done. It would be interesting (but practically impossible, given your air flow source; two shop vacs make a huge amount of noise) to measure intake noise with various configurations. Some jurisdictions make a big deal about noise, so manufacturers will go after the low hanging fruit in a big way. Controlling intake noise is easier than controlling some mechanical noises.

Again, excellent! You get my vote for boy scientist of the year!

 

[Nick Morey]

Excellent research!

 

[Speed3Chris]

Aussie,
If the ship hasn't already sailed, curious if you would please re-run the experiment with snorkel in place and restrictor plate removed with stock filter and RHS airbox wall intact?
Thanks.
PS: The proof is correlating not only inches of vaccum and air flow from power vaccum to theoretical volumetric flow rate of the engine as a pump...but also correlate de-restriction to horsepower on the dyno. Pieman did some dyno testing without the resistor plate and also with pods and ARK and if he sees this thread, maybe he will post his results. If I find them before hand, I will post them.
Cheers.

 

[BlueJ]

Some jurisdictions make a big deal about noise...

So you must not have any HD's on the road near you! Lucky! Even an airbox-free Bonnie with Preds is quiet compared to some of the HD's around these parts.

 

[Aussie_T100]

Thanks to everybody for all of the positive comments, it's good to see that my efforts are appreciated.

Again hats off to your work. Do you think if the air source on a moving bike would change the results? Still air compared to turbulent.

Quite possibly yes, but that is not taken into consideration when people are doing dyno comparisons either is it.

Aussie,
If the ship hasn't already sailed, curious if you would please re-run the experiment with snorkel in place and restrictor plate removed with stock filter and RHS airbox wall intact?
Thanks.

Too late for the airbox I used for testing as it now has the right side modified but I do have another spare stock airbox that I could use to run that test. I suspect that the snorkle will just cork it back up again despite the removal of the baffle but a test would soon prove or disprove that theory. I'll see if I can find the time in my busy schedule to do that for you.

 

[Sal Paradise]

Aussie,

This is a fantastic bit of work. I would say its very comparable to what Ventura did on the other forum.Your method seems very sound and scientific and with you and Ventura both using seperate methods but coming to nearly identical conclusions, for me the matter is settled. I see very clearly how the two compare. By the way I have the double entry modified airbox and it seems to have really helped in both sound and power.

Thanks for your work and thanks for sharing. Great post!!

 

[Sal Paradise]

PS: The proof is correlating not only inches of vaccum and air flow from power vaccum to theoretical volumetric flow rate of the engine as a pump...but also correlate de-restriction to horsepower on the dyno. - Chris

Chris you always go down this road, where " ..wise men fear to tread". Aussie wisely limited his work to volume and pressure and made no claims to analyzing engines. This make his conclusions and information more valid and clear.

 

[Speed3Chris]

PS: The proof is correlating not only inches of vaccum and air flow from power vaccum to theoretical volumetric flow rate of the engine as a pump...but also correlate de-restriction to horsepower on the dyno. - Chris

Chris you always go down this road, where " ..wise men fear to tread". Aussie wisely limited his work to volume and pressure and made no claims to analyzing engines. This make his conclusions and information more valid and clear.

Actually Sal, more like I go down the road that dull men fail to understand and what keeps me largely from further discourse. At some point I will explain and correlate Aussie's testing to horsepower which does btw correlate to not only Aussie's results but Ventura's AF analysis as well. I think many but perhaps not yourself will enjoy this discussion. Most remove their airbox because they perceive it will increase hp. The question is how much hp and at what RPM any benefit is realized. Pieman has done this testing and the results are available. It always comed down to interpretation of data which includes Aussie's scientific testing and if possible how it relates to hp.

 

[Speed3Chris]

Too late for the airbox I used for testing as it now has the right side modified but I do have another spare stock airbox that I could use to run that test. I suspect that the snorkle will just cork it back up again despite the removal of the baffle but a test would soon prove or disprove that theory. I'll see if I can find the time in my busy schedule to do that for you.

I see. A thermodynamic analog is restriction is additive. What this means is, for each restriction, they are additive like resistors in a circuit.
Therefore a snorkel has a resistance value...the restrictor plate has a resistance value, removing the RHS wall has a value and opening up the snorkel hole has a value. The reason for testing with the snorkel and with the restriction plate removed (how my bike is set up) is because one of the predominant reasons the airbox is on the bike is to mitigate sound...other two reasons being dirt and water ingression. Pull the snorkel and intake noise increases considerably. Removing the restrictor plate as you know is a major contributor to de-resistricting the box. So I was interested in your result of leaving the snorkel and removing the restrictor plate. If too much trouble to swap in the RHS wall of the box and re-perform the test, I fully understand.
Thanks Aussie.

 

[Sal Paradise]

Chris ,

Simply stated - Aussies work is a proof in itself. He set out to measure the pressure ( or lack of) and he acheived that and presented it well.

That other "proofs" could be acheived with full engines is obviously true.but the sheer number of variables ( jetting, pistons, carbs, exhausts, ) is mind numbing and you are getting into criteria as well such as where do you want you max torque and do you go for max hp or max torque, etc. You could have dozens or more likely hundreds of combinations of components all producing different results.

Aussie wisely tested one single component and let each owner use the info as he could best apply it.

 

[Speed3Chris]

Sal, forgive me but Aussie's test is at maximum flowrate and 20% above the flowrate of a 865cc Hinkley motor near redline. Flow and restriction is not linear as it relates to horsepower. Horsepower is why most strip off the airbox.
You may not find the following of any interest but it will resonate with others.

First, let’s dissect Aussie’s flow testing:

Based upon his calculation of a 900cc 4 cycle engine spinning at 9K rpm = 67.5L of air/sec. Applying this more precisely:

A 865cc motor spinning at 7K RPM…the max most of us spin these motors = 50.4L of air/sec. This is an important boundary condition and why it needs to be acknowledged. Now relate this value to the volumetric flowrate of the power vacuums used:
1 vacuum pump = 65L of air/sec or a full 22% more air than the 865 motor flows up near redline. What this does is dismiss the use of the second vacuum. Is it interesting to note that over two times the volumetric flow rate has a non linear response to vacuum? Yes but this is known widely in thermodynamics. Flow and restriction aren’t linear in the turbulent range of flow in particular and unless running close to a 2.0 L engine, not relevant.

So…further distillation of Aussie’s values with 1 vacuum pump which again flows a full 20%+ more than a 865 motor at redline:

Standard airbox with Triumph Filter :114mm
As above with snorkle removed : 46mm
As above with primary opening enlarged: 34mm
As above with baffle removed: 16mm
As above with secondary entry added : 16mm
As above with ram tubes added: 16mm
ARK with K&N oval pods: 16mm

Because the one vacuum pump Aussie used in his experiment actually flows 20% more air than a Hinkley motor near redline, the values above are even closer together than posted. This is because vacuum grows exponentially with respect to flowrate and restriction as proven by Aussie’s testing with two pumps which is also based in scientific fact.

Again, the test findings are interesting but they only relate to differences at redline and real world difference in restriction is actually less than the data posted because most ride these bikes between 4k-7k rpm and don’t live at redline.

Summary: Aussie’s findings in terms of restriction show:
1. There is a monumental decrease in restriction between a stock airbox and combined removing the snorkel and restrictor plate.

2. There is very little difference between 1. above and gutting or removing the airbox and installing pods unless riding at redline but let’s examine that further.


Let’s bring in Pieman’s excellent dyno comparison to supplement Aussie’s testing and see if the two agree. I went ahead an made annotations on Mike’s chart to illustrate some points about RPM as it relates to horsepower:

OK, what do we come away with?
Here are some observations:
1. At 4.5K RPM it doesn’t matter if you have a stock airbox, restrictor plate removed or remove the airbox and add pods.

2. Beginning at 5K RPM things start to separate a stock airbox from one that is modified. The restrictor plate really starts to restrict flow on the order of 3 horsepower or so.

3. 6500 RPM is a good reference point to discuss changes to the airbox as follows:
- At 6500 RPM or just shy of redline, an airbox with restrictor plate removed flows as good as NO airbox. Be sure to read that again.
This btw correlates closely to Aussie’s results in terms of magnitude of vacuum. Also there is only a 5 hp difference at 6500 RPM between a stock airbox and one completely removed.

4. At 7000 RPM things start to separate which isn’t shown in Aussie’s test.
Pods start to flow better than the restrictor plate removed from the airbox. At 8000 RPM there is about a 5 hp difference between airbox with no restrictor plate and no airbox with pods.



What does the above mean? You are best served to relate it to how you ride your bike. If you don’t spin your motor much over redline..say shift up around 7K RPM which is where I shift when on it, then there really is no reason to remove the airbox unless you want more noise. When I am riding aggressively through the twisties I stay in the zone between 5-7K RPM where there is little difference between removing the airbox or the airbox intact with restrictor plate removed. If you raise the rev limiter and like foray’s up to 9K RPM or just under, than ARK + pods maybe what you want. The thing to keep in mind is an airbox is a muffler for the intake. Just as with exhaust if you tune the motor right, you will extract maximum horsepower with mufflers removed. Same with the intake. Remove the airbox and at 7.5K RPM or so it will start to matter in terms of maximum horsepower but if you take the airbox off it will matter in sound every time you open the butterflys. So it really comes to the same tradeoff between maximum flowing pipes and stripping the airbox and associated intake roar. All what you want. The thing that Pieman and others have taught us however is and perhaps the most important point…Bernoulli withstanding for who understand his principle, there is no downside to removing the restrictor plate in the airbox…only just a fractional sound increase which to me sounds as it should. I have learned this first hand. It is all good just as Pieman’s dyno chart shows…all the way from mid range on up and I highly recommend it. There is only a fractional increase in sound to my ear by taking out the restrictor plate and with a sectioned snorkel, the intake makes sweet music. We each have a different ear however and some will prefer to remove the airbox and hear the intake in all its glory. For those that want almost the same advantage without the sound penalty I highly recommend removing the restrictor plate and keeping the airbox unless you spend a lot of time above the factory redline. In summary, Aussie’s results on flow and restriction are confirmed by Pieman’s dyno testing and the slight distinction up top is likely due to the flow difference between K&N and DNA pods...the latter being known to flow the best.