An 800's Rebirth/The build of MechanicO

Discussion in 'Parallel Universe' started by FinTec, Dec 12, 2014.

  1. dirtysouthjacket

    dirtysouthjacket Been here awhile

    Joined:
    Mar 11, 2012
    Oddometer:
    900
    Location:
    Charlotte, NC
    Im very interested in your findings. I designed and built the first "dynamic intake runner" on a motorcycle engine (that I know of :D). Around the same time MV Augusta released their Torque Shift System in 2005-2006, which is really a two stage intake and not truly dynamic.

    Anyways, guessing by your posts your testing is all wrapped up, but if you are looking for anymore insights let me know. Good on you for really pushing the envelope with the F800!
  2. dirtysouthjacket

    dirtysouthjacket Been here awhile

    Joined:
    Mar 11, 2012
    Oddometer:
    900
    Location:
    Charlotte, NC
    70East, Diameter of the tube is a very critical parameter of the overall principle.
  3. AK650

    AK650 Long timer Supporter

    Joined:
    Jun 26, 2013
    Oddometer:
    1,943
    Location:
    Alaska
    Cant wait to see the results. Very interesting stuff. Thanks for sharing!

    Jeff
  4. arjones

    arjones Roads and Waves

    Joined:
    Feb 24, 2012
    Oddometer:
    263
    Location:
    Bahia, Brazil
    I've been following this with admiration, jaw dropped and increasing curiosity since the first post. The skills and deep knowledge required for this endeavour is amazing. I'm learning a lot and even starting to think my bike isn't that awesome thing I thought it was...
    Anyways, out of curiosity, is this http://www.advpulse.com/adv-rides/la-barstow-to-vegas-ride-on-big-bikes/ the history behind the donnor bike? Funny, because, if you look closely, it's almost the same damage reported by you on your firts posts... Long story short, these guys where doing some riding coverege thing for that website and one of them, (with a blue F800GS nonetheless) took a really bad get off and.. have a look. Maybe I have found the MechanicO's lost Family...?
    Cheers.
  5. FinTec

    FinTec Been here awhile

    Joined:
    Sep 27, 2013
    Oddometer:
    461
    Location:
    Colorado
    Answer some questions:

    Shape and diameter play a big part in resonance tuning and in general, your intake track. Not as much diameter but change in diameter is important. You need to try and incorporate a taper to the system (big at the start, smaller as you move along towards combustion chamber). For example, the stock intake tubes start out at 59mm dia, then at the throttle bodies it goes down to 50mm, then the actual bores of the throttle bodies are 46mm. They have built in taper to the entire intake track. This creates velocity and velocity is what always helps on torque. That is generally why wide open huge intakes have poor low rpm torque, their velocity is so low due to large diameters.

    Curved tubes will cause some resistance to both air flow and sound travel (our resonance affect we are tuning). But as long as you have good transitions and large radius's the negative affect is negligible. Yes, the straight tubes I mach-ed up for the test would NOT look or go well on the bike to ride. As you will see next, I have a plan to make it all fit in the air box area of the bike.

    Absolutely can do this type of mod with headers. What goes in must come out. There is some good solid science with exhaust headers as well, but I chose to just use the Remus system and be done with that side of things. Getting rid of the CAT is the biggest affect I am going to get IMHO.

    VERY cool to hear! I did find some neat information on some systems built before to allow for "dynamic" change of the intake track to move the power band around as you need it:

    http://www.emeraldm3d.com/articles/emr-adj-length-intake/
    *not dynamic but an adjustable system

    http://www.yamahapart.com/page/yamahaycci
    *neat system by Yamaha that is controlled by the computer. But two stage as well.

    I also know in F1 racing they had some systems that where dynamic in that they changed the length of the intake track on the fly. Not sure where that ended up?

    I actually thought of adding a "2 step" system to the build. Have a lever on the side of the air box with two positions: 1) makes the intake tracker longer and creates torque for off-road 2) Removes length from the track and moves power band up the rpm and works best for hwy riding. Just hard to squeeze into the space I had, so went for low end for now.

    Do you have any pictures or data from you project? Love to see what you did.

    Wow, nice find. But I think that is not MechanicO. The front wheel does not look like the one that came on my bike. And he did not have any plastic on one side. And he came with full luggage. Some day he will find out where he came from......

    Now there was something I needed to post....:getiton
  6. FinTec

    FinTec Been here awhile

    Joined:
    Sep 27, 2013
    Oddometer:
    461
    Location:
    Colorado
    So after the dyno day I went home with a thumb drive packed with data. Combed through it, graphed it, and came up with some VERY informative data in my opinion. Lets start with the main graph. This is a graph showing HP and torque for the four configurations for the test

    [​IMG]

    So where to start... lets look at Blue and Pink first. Interestingly enough these represent the two extremes of the test: fully stock vs. absolutely nothing, bare open throttle bodies. And funny enough, they are very similar. The trade off a bit through the power band, but really stay close to the same all the way up to the end. I am going to guess (and I am guessing) this is due to the Pink having no benefits from the intake track tuning, so it is only able to match stock. So a bit of a coincidence, but I think the case. But bottom line, going full "naked" does you no good.

    One of my favorite comparisons: Blue vs. Green. This tells me one glowing thing, the stock air box and filter are a hindrance to power. The data is pretty linear, removing the air box and air cleaner creates more power across the board. Obviously we need an air cleaner so you have that, but you can clean air AND have good air flow. I do think that the small air cleaner they use is NOT allowing for full potential. As far as the air box goes, I think we are looking at a product to pass EPA intake noise relations. I am no air flow engineer but when you have throttle bodies with 46mm diameters and the intake ports (front horns) on the air box are 28mm in diameter :doh, that is a major choke point. Seeing as that creates a reverse mega-phone I can see where that would help with making a quite intake. But for power, it blows...er, not suck, no wait, it does suck...anyway, you get the picture.

    Now for the big showdown: stock vs. my 10" extension (Blue vs. Red). Now first, I fully admit, not 100% fair. As we have just seen, just removing the air box and air filter has some gains. But I am still going to go with this comparison as it is the eventual goal of this build so I will be addressing the air box and air filter choke points later in the build. So to me, stock vs increased intake length would tell me where I am heading with this.

    To get a better picture of this, here is a chart of just these two tests

    [​IMG]

    This really puts it in perspective and I was ecstatic with the results. HUGE gains down low with the 10" extensions. This pretty much hit the math dead-on in my book. Two things to note:

    1) look at 4500 rpm: both torque and HP are 15% more than stock. That is a LOT of gain in that area.

    2) more importantly, the extensions filled in that entire "dip" in the stock torque band from about 3750 rpm to 4750 rpm. Every dyno chart I came across for the F800GS had this exact same dip in the same spot. Now that dip is gone and the curve is VERY linear. The HP also straightened out in the same spot.

    Like all things in life, there is a give and take, so where do I see the take. As you can see, the 10" extensions take it on chin at around 6200 rpm. There is a big dip in power, that does recover a bit, but eventually falls off sooner that stock. But, and this is just my opinion and I have no idea if others agree, who freaking cares. As you can see, I put a line up to show the rpm's I spend the vast majority of my time in. Yes, I did lower my peak HP from 72.76 to 69.74 (4.33% drop), but I increase peak torque from 53.73 to 56.81 (5.73% increase) AND moved it to the left where I actual ride the bike. I consider this a success in the test and the idea behind resonance tuning.

    Out of interest, here is the AFR graphs for the 4 runs

    [​IMG]

    Neat to see and helps visualize what is going on. Sure enough at around 4000 rpm the bike is going very lean with the 10" extension. This of course means it is getting a lot more air than it normally is in this area. The affect of resonance tuning for this area. Then you see we go super rich just after 6000 rpm with the extensions as the longer tubes are in fact now "choking" the intake and staving it of air. I suspect with some time riding the bike with the extensions, the computer would adjust and take care of these differences. Still shocking to see how lean these bikes run in stock form as well (blue line).

    So one point to make here, I do believe when I complete this part of the build I will actually see a higher gain in torque for these reasons:
    • Running a higher flowing exhaust
    • Enriching the AFR with an aftermarket system (this will be huge gains IMHO)
    • I will me be making extensions with taper and bell-mouths to give MUCH better flow than these raw pieces of no taper, no bell-mouth tubes used in the test
    • And air box design and filter with little to no restriction
    I am hoping to top 20% more torque in the 4500 rpm area and a peak torque value that is 10% over stock. These would be amazing numbers and make MechanicO incredible to ride. I swear I could see MechanicO smile during the test with the 10" extensions. He is addicted and there is no taking it away from him now.

    Just gotta build it. So next up, the custom intake tubes.......
    thirsty 1 likes this.
  7. arjones

    arjones Roads and Waves

    Joined:
    Feb 24, 2012
    Oddometer:
    263
    Location:
    Bahia, Brazil
    :clap:clap:clap:clap


    :y0!how long until the next post?:augie
  8. ERU

    ERU Been here awhile

    Joined:
    Nov 1, 2009
    Oddometer:
    101
    Location:
    Bucharest
    [ Yes, I did lower my peak HP from 72.76 to 69.74 (4.33% drop), but I increase torque from 53.73 to 56.81 (5.73% increase) AND moved it to the left where I actual ride the bike. I consider this a success in the test and the idea behind resonance tuning.


    Just gotta build it. So next up, the custom intake tubes.......[/QUOTE]

    Greetings to you and to Mecanic0.
    I'm absolutely blown away by your thread and as an F800 owner, i just feel that you did exactly what BMW should made, in the first place.

    I still have a question, you said earlier, that you lose some HP from 72.76 to 69.74, but from what i know and you said earlier in the thread, the peak power of 800 is 85 HP.

    Also from what i understand, that engine is somehow restricted by low diameters and the intake port, how about you didn't get even more HP than the stock 85?
    [I read on some other thread here on ADV, that the 800 engine is designed for about 100 hp, but put in leash in order do not cannibalize the greater 1200 GS]

    Sorry for my bad english.
  9. RubDogg

    RubDogg Adventurer

    Joined:
    Feb 21, 2015
    Oddometer:
    24
    Location:
    DE
    Very cool to read! You're really putting a ton of effort in this build. I'd love to see your face, taking MechanicO for the first ride :D
    One question, that might be slightly off topic: Do I read your dyno charts correctly that a K&N-airfilter will give me slightly more power and soften the dip around 3750 a bit? Might be worth a try...
  10. dirtysouthjacket

    dirtysouthjacket Been here awhile

    Joined:
    Mar 11, 2012
    Oddometer:
    900
    Location:
    Charlotte, NC
    I will shoot you a detailed PM when I get some more time, I don't want to detract from your thread! Just wanted to throw the offer of help out there. Great stuff!
  11. FinTec

    FinTec Been here awhile

    Joined:
    Sep 27, 2013
    Oddometer:
    461
    Location:
    Colorado
    ERU: thank you for the compliments. Cool to see interest from Bucharest! The factory rating of 85hp is "at the crank" hp. I am measuring HP at the rear wheel. Once you go through the transmission, chain, and rear wheel, you will have losses. So in my case I had about a 20% drive train loss. Also, I tested at 5300 feet of elevation so I also have some loss there as well. I know the car guys go with around 15% for drive loss, but not sure what the losses are for a typical chain drive motorcycle? Anyone know?

    Of course in my case, I was looking to measure differences between all the configurations so no super concerned on the loss as there is not much i can do for that.

    RubDogg: Well, yes, in theory, if you create less restriction with a better air cleaner, then you should see gains in the same area I show. But there might be other factors that also affect this area of the power band that the new KandN filter will not fix. What if the air box is 80% the difference we saw? Then a filter change will be a small change in comparison. What I should have done was run a 5th test of everything stock but ONLY the air filter removed. That would have been interesting and narrow it down to the issue being the air box or the filter? Or rather the percent each plays in the difference.
    RubDogg likes this.
  12. FinTec

    FinTec Been here awhile

    Joined:
    Sep 27, 2013
    Oddometer:
    461
    Location:
    Colorado
    DirtySpouthJacket: Hey, not going to steeling any of my thunder. I would love nothing more than this to become an great conversation on intake theory and how it applies to these bikes. Post away here if you like, good by me. Or PM. Just love to see what you did as well and any conclusions you cam up with.
  13. RubDogg

    RubDogg Adventurer

    Joined:
    Feb 21, 2015
    Oddometer:
    24
    Location:
    DE
    True words, didn't think of that. I might just go for one when the current one is done.
    And now I'll go back to reading and stop clogging the thread. (I'm always disappointed, when I see that the thread has new messages, open it, and it's nothing new from you Fin. ;) )
  14. banjaxxed

    banjaxxed Adventurer

    Joined:
    Nov 25, 2014
    Oddometer:
    52
    Very cool, surely variable intakes would provide optimal power at any rpm?
  15. banjaxxed

    banjaxxed Adventurer

    Joined:
    Nov 25, 2014
    Oddometer:
    52
  16. zabaleta

    zabaleta Adventurer

    Joined:
    Jul 27, 2014
    Oddometer:
    27
    Location:
    San Sebastian, Basque Country
    Incredible, to remove his hat.
    I totally agree that most of the time, we bring our bikes below 5500 rpm, and have made hundreds of thousands of kilometers in F800GS.
    Congratulations
  17. FinTec

    FinTec Been here awhile

    Joined:
    Sep 27, 2013
    Oddometer:
    461
    Location:
    Colorado
    Absolutely they would! But at a cost of complexity and available room. I would like to do this on a vehicle some day, but for MechanicO, I was happy to just move the power band where I wanted and keep it simple.

    And don't call me Shirley :lol2

    Love the videos as have not come across those before. Would also like to see the dyno graph that goes along with those pulls. Now that would telling the story.
  18. banjaxxed

    banjaxxed Adventurer

    Joined:
    Nov 25, 2014
    Oddometer:
    52
    Sorry there! Yes overkill maybe but from what you're able to do certainly within your comfort zone. I found this product sheet just for interests sake
    [​IMG]
  19. dirtysouthjacket

    dirtysouthjacket Been here awhile

    Joined:
    Mar 11, 2012
    Oddometer:
    900
    Location:
    Charlotte, NC
    So, Im in the middle of moving and getting married next month. The computer is in storage, so typing with two thumbs on my phone is agonizingly slow. :fpalm Im also going to proofread later! But sice you insist hear goes.


    Anyways yes, it was fully variable and dynamic and tied to engine rpm with a small servo and lead screw. I can go into more detail tomorrow when in front of a work computer. But this was back in 2005 as part of my time on a college Formula SAE team. So had pretty amazing resources to play with. Full machine shop, access to Panoz Motors Racing carbon fiber shop, engine dyno, a large team budget and a lot of other really bright minds. Also the car had a fully programmable Motec engine management system. It was definitely back during my mad scientist days of staying up all night building stuff and running on Mtn Dew and Red Bull. :drink

    The engine was a Honda CBR600 in basically a large go kart tube frame with a lot of carbon fiber. I cant take credit for the idea, but another guy the previous year had made a large number of different length runners and tested every single one with the engine dyno. Ill find the graph eventually, but it was pretty amazing to see all the torque curves laid out looking like a ladder as the curve shifted up the rpm curve. The theory already existed of different length runners, he was just optimizing it at the time for the engine and driving characteristics. OEM cars have already had this as well and were called "variable length intakes". Some had multiple differnt runners that would open or close depending on rpm going back several decades now. VW diesel engines have this, however it is more for emissions.

    And yes banjaxxed, because of the final end result having full torque everywhere it was like having an electric motor or cvt, no real shifting needed because of how short the race track was.

    Reading Fins post have brought back a lot of very good memories about that time.

    Anyways after someone said, "Hey lets make an intake runner that changes length" and I had the most mechatronics/automation experience, BOOM project conceived. It also helped that as we started talking about different concepts, someone mentioned that a race car at Panoz already had one. It was the Mazda RX92-P, only 3 of these cars were ever built. We were actually fairly bummed to hear such an intake so similar already existed, but it was helpful to have something to look at for ideas. It was a trombone system inside a large carbon airbox that had pulleys that pulled back and forth. I have pics of it buried somewhere as well. That car was incredible. The engine was a Mazda quad rotary!

    My system was a set of magnesium tunes with very tight slip fits that actually wore in to produce a fairly air tight seal. It was tied to the throttlebodies on the static end with a carbon fiber platen at the moving end with the trumpet shape that Finn described. The static tubes that were used in the initial testing were tapered as Finn described, but this feature was not used on the dynamic system because of the need for the tubes to slide in and out of each other. A servo motor drove a lead screw that pushed the end of the tubes up and down.

    The entire thing was housed in a carbon fiber air box. At the time FSAE rules stated that all intake air had to come through one single rule-specified orifice to increase competition and reduce engineering advantage. This produced many different creative work arounds from some teams. Like huge air boxes that worked as resevoirs to defeat the small intake inlet. So really in this application, it was a great improvement. Weight was a huge concern. Space was also a premium. Like Finn mentioned packaging a big tall expanding 12" runner is tough. Another rule was that you couldnt use your chassis more than two years and everything had to fit within the roll bars to prevent a fire. So we ended up with a slightly shorter extended length than ideal for low rpm due to to the roll bar behind the driver.

    Forgot to mention, the tubes were extended at low rpm and retracted at high rpm. This directly correlates with Finns long runners shifting the torque curve lower. I slightly disagree on the taper and air speed philosophy. I will save that for a separate post.

    Finn, I almost commented before you posted the dyno graphs during yor posting pause, that I expected to see the torque curve to drop to a peak around 7k ( I was only guessing in my head since the cc difference between the F8 and Honda 600) as well as an abrupt fall off of torque. This is fall off is very typical and we saw the same thing with initial static tube testing. However you should be able to smooth the drop some with a fuel map correction.

    I didnt mention much about the controls. It was relatevly straight forward. We fed the servo position to the Motec ECU which already knows rpm. The Motec at the time allowed for several control outputs, so I ran a simple PID vs the engine rpm and viola that sucker would just jump up and down as speed changed. It was stupid fast.

    One comment on the paper you linked earlier. It described the same similar principle (but ten years later! Sorry I had to jab at them. :D) but using a rack and pinion. This is not ideal compared to a lead screw due to speed and positional accuracy/backlash. Gear racks are also heavy and complicated to machine. Also the Arduino is unnecessary if using an engine management system.

    It was a great project, but definitely is a niche application. The rules really made it advantageous. But making a production system would be expensive, take up valuable space and have a host of reliability issues to hammer out. Especially when forced induction is so relatively cheap, powerful and reliable.

    Another team improved on the concept by making a nautilus shell style that rotated in on itself to save space. I dont know, maybe that could work better on a twin vs a 4 banger.

    Ill probably think of things to add or mention eventually. Once I get my computer out of the Shipping container late next week, I will dig up pics. This was all before everyone had cameras on everything and they went straight to the cloud, so I wish I had a bunch more.:(:

    Thanks for the trip down memory lane. But it sounds like you have a great handle on things already, im interested to see the improvement with the final tubes and packaging as well as fueling changes.

    Other crazy ideas Ive had. Slightly off topic. Ok mostly.

    Using some crazy exotic cooling fluid and a heat exchanger to supercool the incoming air. I met a guy that did this with his fuel and it worked!

    I have always kicked around a different thtottlebody design that eliminates the butterfly valve restriction.

    I also want to explore raising the redline of the engine. Microsquirt is definitely in my or maybe even another F8s future. Ive done two Megasquirt EFI projects in the past. Totally worth it!

    Finn you are absolutely correct in your assumptions and you will love the motor now. People get too hung up on total max peak hp or torque. But the area under that torque curve you have made is very exciting. Anyways, carry on! :clap
    Pressingonward likes this.
  20. dirtysouthjacket

    dirtysouthjacket Been here awhile

    Joined:
    Mar 11, 2012
    Oddometer:
    900
    Location:
    Charlotte, NC