Jump to content
Sign in to follow this  

Airbox Modification and Simulation Testing - Conclusion

Recommended Posts

So this is a final thread listing conclusions from a long-term analysis I have been doing on the airbox and intake system of the FZ07. For those not familiar with that thread, it’s https://fz07.org/thread/7684/intake-simulation-adding-velocity-stack I’ve been updating it as the simulation and analysis has been completed over the last month or two. This thread will add the hard data and show some conclusions from the work.
It’s been very interesting, and there were a couple of surprises as well. I looked at stock, removing the lower half of the snorkel, removing the total snorkel, removing the entire filter and filter cover, installing the MWR filter and cover, and I played around with the tract length and shape as well. I also created my own external filter design and ran that as well. You can see more data on that here: https://fz07.org/thread/7851/prototype-intake-filter-external-airbox  
I need to re-emphasize that this is a very limited test and removes a significant number of variables from the equation that defines the performance of this engine. It is very simplified and only looks at the static flow capabilities of the airbox along with some of its components. I used a pressure differential of 3 inches of water but also ran a few tests at 1.5 inches, which didn’t add anything to the results because of the limitations noted above. That which flowed best at 3”, still flowed best at the lower pressure, so no need to delve deeper there.
First I’ll list the results. I’ll list the actual flow for the stock system but all others will be compared to it as a percentage, so consider it the baseline. The actual number is not important, just the relative change of the number. Units were in cubic meters per second.
1. STOCK: 0.0362 m^3/sec. Very, very roughly an engine speed of 5000rpm flows .03 m^3/sec, but, this number doesn’t include all sorts of dynamic variables, so it’s very rough, but gets you a sense of where the tests were at which should be considered as in “the middle of the engine range, ” not necessarily 5000 rpm. This is only the baseline from which all other tests were compared.
2. INTERNAL NOSE OF SNORKEL REMOVED: -0.8%. That’s right, the flow was DECREASED be removing the internal part of the snorkel. I explained why on the original post – see there for more detail. Notable in that 2WDW also showed a slight decrease with this setup, supporting this approach and results.
3. ENTIRE SNORKEL REMOVED: +16%. A very nice gain can be had here if you jet accordingly to the increased flow. This is why all the tuners out there who post on this forum all recommend to remove the snorkel. I do too. Remove, optimize the jetting and you can get a few ponies.
4. STOCK FILTER AND FILTER COVER REMOVED: +80%. Yeah, no surprise. Take the lid and the filter off and it flows like a burst dam. Interesting to note on this is that I believe that this is the configuration that Hord did to create his dyno run that had the very, very fat midrange gains of ~5 HP that are shown on the first page of the post. But, no filter, so not a good idea.
5. VARIOUS SHORTENED TRACTS: No change. That’s right, zip. I tried this with stock, with the snorkel removed, and with the filter and cover removed and basically got no change, at least nothing equitable to noise. But there was a small DECREASE for the fully-open configuration at 1.5” shorter, but no change with 4” shorter. This is kinda counter-intuitive, isn’t it? Open up the box, shorten the runner lengths to decrease the flow restrictions and it should run better, right? Nope, there appears to be more magic here than we originally thought. I did the runs at stock length minus 1.5 inches and a very short one that was about 4 inches shorter than stock. Now it DID alter the flow characteristics, but not the flowrate, which is what I’m looking at. This was interesting and not what I expected.
6. MWR COVER AND FILTER: +70%. Youzza! This should be it. Damn near the same as running it wide open, but something is amiss here. I write about it in the previous posting but this potential is not necessarily attained. I’ve seen dyno runs that were noted as back-to-back, same day, same bike but with MWR vs. DNA and the MWR was squeaking out some gains. BUT I’ve heard from two different tuners offline that they are not seeing it in their work. This is where things have to be kicked back to the “heavily-simplified” clause on this simulation. There obviously is some dynamics going on here that is causing the filter to not work to its max potential. For kicks, I added a nicely shaped external velocity stack to this config and re-ran the tests and got a slight increase of 1.3% showing that there’s always room for improvement, but it’s a lot of work to gain a small number.
7. EXHAUST TRACT REDESIGNED AS VELOCITY STACKS: +13% MORE THAN WIDE OPEN. In short, taking the small diameter intact tracts and making very clean ellipse-based velocity stacks added a 13% gain OVER the previous gains. So vs. stock, adding the largest of the velocity stack designs I looked at (they were designed straight-back, no curves) and removing the cover and filter flowed more than twice as much, 105% increase. This is some serious sh!t if it applies to the dynamic characteristics as well. BUT it would necessitate making these stacks and getting them in there, and that’s not something I’m up for for a, what, 1HP gain or something. It can be done with 3D printing but, again, lots of work for an unknown gain.
8. MY EXTERNAL DESIGN WITH STOCK TRACTS: +46%. Hmmm, was hoping for a better result as the MWR was showing pretty sweetly. But then I looked at both this design (I’ll refer to it as the A.C. design) and the MWR with the large velocity stacks, and my design actually flowed BETTER than the MWR, so there’s some magical stuff going on here that is just out of sight. And again, without all the actual flow dynamics added, we really don’t know what’s actually happening without dyno tests.
• Duh: larger openings generate more flow. With the cover and filter removed, it flows the best.
• With stock tracts the MWR/cover is flowing very, very well. It came in second in the simulation with standard length tracts, but was tied with my design when I altered the intake tracts. There was talk of increased performance due to increased surface area of the bottom of the filter, it’s not see in my simulation because it introduces a negative flow characteristic external to the filter and decreases the pressure in the filter, erasing the potential gain.
• The filtering element is not making that much difference relative to all the other factors. That is, changing a filter because it flows slightly better is not doing anything of note because the filter is not having that much effect on overall flow. This also is not intuitive – yes, it’s having an effect, but when you look at the overall flow path and variables, I could completely take out the filter and have it barely have an effect on overall flow, so minor changes in filter flow are not going to have any effect. This kind of buttresses to K&Ns info that you don’t need to rejet with their stock replacement element, right?
• Moving and playing with the length of the stock exhaust tracts (exhausting the airbox, that is) had no meaningful effect on flow. Again, not intuitive and was one of the things that started making me see that the system of all these components is more important than any given feature. Meaning that airbox volume and shape, tract length and size and location, filter location, intake design, etc. are all working together to have a synergistic effect. If designed well, they work together and optimize the flow even with inherent restrictions.
• The last comment works here as well – the airbox shape is very critical to shaping and directing the flow. The narrow part where the tracts open, is actually narrow to force the flow into the openings and not let it slide down the wall. When I pulled back the tracts just a smidge, you could see the dynamic at work. At stated before, there wasn’t a loss due to the change, but you could see that it was affecting the dynamics of the flow. This was real important when starting to play with tract length and opening location. I’ll touch on this a few times – the stock design actually does pretty well – the engineers were doing some nice work while having to account for noise and emissions issues as well. Things are shaped, oriented, and located as they need to be, and it’s pretty easy to add a bit of chaos to that dance if you play with it without knowing what the effect will be.
• Remove the snorkel and get your bike rejetted by one of the know techniques. Yes, louder, but it makes a difference. I’m seeing ~2 HP in midrange potential.
• Do not cut the snorkel. Remove it completely.
• Properly designed, large velocity stacks DO have a nice increase in flow. I think we get away from the tuned aspect of the parts to a “bigger is better” mode. It may not be as tuned, but by virtue of being wide open, it flows like a sombitch.
• Interesting to note, a slight gain (+1.3%) was had by adding a significant velocity stack to the outside of the airbox. Notable because the minor radiused or square edged openings, while theoretically different, aren’t really different in practice, as this big, bad mother was only able to muster a negligible gain.
• One you start playing around with the tract length, you realize they are also located to equalize flow in the two chambers. Once I started playing around with length, even though they were in the same general area and layout as before, the flow relative to the two tracts started to diverge and become different. Again, pointing to the tuned and optimized design already in place.
So that’s the results from most of the tests, and it should give you an idea of what I found, objectively. Now on to the subjective results – what I saw and how I interpreted them.
One of the components here that is causing issues is the speed of the flow. On the smaller orifices (stock, snorkel removed), the flow is coming down very fast and causing some disruptive flow. It’s having a hard time changing direction and maneuvering cleanly to the ports. One of the findings here is that having a bottom on the filters (not the MWR) was actually HELPING the flow even though you lost some filter surface area. The flow is screaming in, hitting the bottom of the filter, creating a high-pressure zone and forcing the flow out laterally towards the ports. I would not have thought of that. I tried a run without the bottom to the filters and had it as additional flow area, and the flow was DECREASED slightly! Again, this is because the flow is being shaped and directed by the pressurized filter. Without the bottom, the flow is hitting the bottom of the airbox and blowing past the openings and dancing around quite a bit before it circles back to exit. It’s not as strong on the MWR, but makes me wonder about its dynamics while actually running, so I went and looked at my bike (MWR filter and cover) and rev’d it while looking at the filter, and it’s jiggling at certain engine speeds. A vibration sets in then stops, dependent upon the engine speed. So if this is happening regularly, I can see why there is a flow impediment that’s not being visualized with this analysis. And that would be very subjective to each bike and where its natural frequencies and harmonics are depending upon quite a few factors. That COULD explain why some of us are seeing increases and others not – bike specific. This is definitely speculative, but possible, esp since I’m seeing some aspect of it on my bike.
I don’t have any data on the DNA to support this, but I think the DNA and the MWR should be about the same in performance, unless there is a real dynamic problem with the MWR that I’ve alluded to. On a bench, all else being equal, the MWR will likely flow a little more, but in real life, in this application, you should be able to consider them equal. I wonder if the MWR filters (particle size and pressure drop per a given flowrate) as well as the DNA? I haven’t taken the time to look it up, but I wonder.
So WTF is going on with MWR? It’s flowing fantastically here and should be the bomb, but the feedback from the tuners is not so much. It’s possible that we’re so far over the threshold of limitations in the filter element that changes are not having much effect overall. Possible, but there is also the possibility that there is a dynamic movement that is taking place, and the filter is quite flexible, so it’s possible. I think it would be very easy to set something up to settle it down, but there’s no data to support this supposition at the moment except the vibration I saw and the fact that some of the tuners are not showing a difference.
A while ago Hord commented that he thought the internal filter was “shrouding” the ports and inhibiting flow, so an external filter was a likely place to find an increase in power, IF this shrouding was taking place. Well, I’m not seeing that dynamic in this simulation, but I’m not using the actual filter mechanical properties, so maybe they are actually there to some effect, but I don’t think so from the simulation. I can see the flow pattern and how the filters are affecting it, and I don’t think a shrouding is happening, which is unfortunate, because the whole point of the external filter design was that this shrouding WAS taking place. So only the dyno knows for sure, so . . .
Next . . . is a trip to the dyno. I want to run snorkel removed (as a baseline), the MWR setup, and the A.C. design, back to back, same day, same bike. What do I expect to see? Not a whole lot, a few ponies over the baseline, but A.C. relative to MWR one will be slightly better than the other, but by a negligible amount. The thing to keep in mind here is that an engines power is a complex dynamic bit of magic and all the little facets matter - but only to a degree, up to a threshold. Once you go over a certain threshold, bigger local gains may do nothing to power because you’re not changing a meaningful factor anymore. Below this threshold, it can make a large difference because there really is a restriction in that local part of the chain, like removing the snorkel. Consider that Hord took nearly the entire airbox/intake flow restrictions out of the equation and had some nice gains at the top-end, but if it really was as simple as removing the restrictions, he should have had comparable gains throughout the rpm range! Think about that for a second – it’s not restrictions necessarily because he took them all out and didn’t get midrange increases. This is an important understanding, I think. It’s not reasonable to think he would have had increases everywhere because he was tuning for max HP at high RPM, and we know there are dynamics at work here that are quite complex and flow restrictions can actually be tuned in; that is, the bike works better at some engine speeds WITH the restrictions as they alter tune, most notably when tuning for midrange vs. high RPM. So if we’re not removing a meaningful restriction, and we’re already over the threshold for making meaningful changes in the local flow path, we’re not going to see anything overall either.
One of the forum members (@gregjet ) here commented about the tuned aspect of the intake system. I’m not sure what his background is, but his perspective rang true with me after running all this simulation and looking at the dyno runs strewn about this forum: “The most important thing on the intake side is the length of the intake and the harmonic used to calculate it. Longer isn't better, shorter isn't better. The right length is best. There is more than one length depending on the harmonic number you use ( a base harmonic is seriously long , so that's out). That lengths depend on the RPM for which you wish to tune the motor. Next is the SHAPE of the inlet "tube" and opening. This has a huge effect on the flow. ( "Funnel" shape) On top of that is the airbox resonance and filter restriction. And the actual airbox inlet shape and restriction. It isn't simple. The 07 is very restrictive at the outside end, so any help there will be useful.”
Another forum members perspective: @rhb : “Believe the tuner who says intake mods are not the place to look for great HP increases, especially between 2 different "high flow" brands, lid or no lid. These difference while they may be visible on today's dyno run (vs, tomorrows or another bike etc), are largely academic.”
Another: @hordboy : “This is interesting stuff, but static flow models are only one small part of the equation as I'm sure you know. The tuning of the intact tract has a far bigger impact. Runner length, volume, etc. The test I did removing the filter and snorkel assembly completely, proved the stock runner length is ok for midrange, but the volume of the airbox is too small, and the runners are badly shrouded by the filter. I think the solution there is a flat panel filter over the airbox opening, or an "upside down" filter on the outside of the box. Room is limited for that.”
For additional info check out this thread. I’ve alluded to it, but here’s the actual info showing his results that MWR was better than DNA by a smidge. https://fz07.org/thread/6237/mwr-dna-filter-dyno-tested
But, all in all, I’m hoping to see that the external A.C. design will mimic Hord’s midrange boosted dyno run. I’ll post again after the dyno runs.
  • Like 6

Share this post

Link to post
Share on other sites
You can removed the snorkel, snorkel plate, and airfilter then take a oiled MX filter like this stuff https://m.summitracing.com/parts/ufi-bf-2?seid=srese1&cm_mmc=pla-google-_-shopping-_-srese1-_-uni-filter&gclid=CjwKEAjw_6XIBRCisIGIrJeQ93oSJAA2cNtMCAzN73i14rS7EXIZB22PMfBA-bCQLkEA_gucG31ZhBoC9Tvw_wcB
Then cover the opened that once contained the airfilter overlapping some of the airbox and (I used some long ass zip ties) secure the filter material so that it covers that opening. I used an ejk to add the fuel I needed and got great results! $20 bucks and some redneck engineering...I used 2 layers of different density (the red and grey materials )

Share this post

Link to post
Share on other sites
Very balanced view of the whole concept, outstanding work, Case in point was my ECU flash producing decel popping, not a bad thing in itself, but surprising. Why was it the case? many possibilities from not enough air put into the flash based on a dyno run of a similar but not identical set up or even the fuel left on during closed throttle, or a combination of these and other factors. The FTECU active tune dialed it in precisely.
The take away for me is after a point we reach diminishing returns in terms of cost per HP, and real performance increases for daily street riding. But it sure is fun trying to figure it out :)
Thanks for your work.

Share this post

Link to post
Share on other sites
I figured out the MWR vs. DNA issue regarding the other thread. Take a look above and see the edit. Kinda interesting.

Share this post

Link to post
Share on other sites
I figured out the MWR vs. DNA issue regarding the other thread. Take a look above and see the edit. Kinda interesting.
@sansnombre ,
The MWR vs DNA thread with the dyno graphs was from my bike. It's me, Paulie, same Paul as from BellissiMoto (for those of you who didn't already know). I did that testing because of course my co-workers wanted me to use the MWR filter because that's what we sell.
However, ya'll know me by now, I won't run anything I don't actually believe in...so if I was gonna change from DNA to MWR I needed to see for myself. And as you pointed out, the MWR DID make gains over the DNA.
I had no idea that the Akra Ti was a catalyst for those gains, but I've also heard reports from customers that the MWR made similar gains from the filter they switched from (not sure what exhaust they were using though).
In my original testing, with the DNA filter and DNA cover, and STOCK exhaust, that combo was worth about 2-2.5 hp, with nice gains in the midrange.
My take away from your post (purely from a which filter and cover to buy standpoint), is that the MWR is the most COST EFFECTIVE way to get these 2-2.5 hp gains, and slightly more gain with using the Akra Ti.
The cost of the MWR combo is $70 CHEAPER than the competition.
I applaud you for all this testing.  Fantastic work my friend.
- Paul 

Share this post

Link to post
Share on other sites
Thanks heaps for the work sansnombre. Good quality hands on physics/engineering testing.
My background is simply passionately interested amateur ( in gas flow stuff in this case). A lot comes from interest in 2 stroke tuning from race mod days. In 2 strokes power mods come primarily from gas flow chnages so it kinda makes you concentrate on those issues.
Interesting your notes on elliptical crossection stacks tends to confirm Gordon Blair's work in that area.
Besides dynamic flow, the resonances of airbox/inlet tract and, including and NOT including the cyl volume are probably going to influence the results.
I agree about the high pressure "cone" in the filter helping direct the flow out to the filter material. I thought of making an actual solid one to help remove any turbulance that may form especially as the revs change.
I guessed that removing the bottom of the snorkel may screw up the flow, but the whole thing is so restrictive I didn't worry about it much. BTW we didn't get any statistical significant difference removing or leaving the snorkel on the dyno. That I suspect has to do with the plate/hole turbulence of the stock filter top. I made a crude ( very) shaped inlet top to the DNA filter just to help smooth the inflow ( not tested just seat of the pants added. There is a picture on the forum https://fz07.org/thread/3667/derestricting-lams-mt07-australia ). Been meaning to make a better shaped integrated one, but I will revisit your work when I go to do it now.
Even though I eventually found out the the firing order was 270deg ( not 90 as I originally was told by Yamaha) , without the actual inlet cam angles , I still suspect flow interference between the two inlets at any overlap. If one cyl is still closing, whilst the other is opening, than the two inlets side by side will interfere with each others flow. If they don't actually directly interfere then there may be some residual dynamic flow interference in the air box. eg. The left cyl intake occurs and the flow will be clockwise from the top, second intake opens and has to reverse the circulation. Theory only, not tested, but worth a look in the quest for ultimate power ( mine isn't a race bike so I haven't pursued the edge gains).
Interestingly I have a Subaru ( normally aspirated) which has the most insanely long inlet system, so I am having a really good look at it at present in an attempt to clean it up.
Great work either way and thanks.

Go forth and modify my son...go forth and modify...

Share this post

Link to post
Share on other sites
Wow, great work, thanks for the write up!

'16 Yamaha FZ-07, '15 Yamaha FZ-09

Share this post

Link to post
Share on other sites
Nice, thanks for sharing.

Share this post

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  


Important Information

By using this site, you agree to our Terms of Use.