Gunning for high gas mileage

Discussion in 'Introduce Yourself' started by DougPan, Dec 26, 2008.

  1. DougPan

    DougPan New Member

    I've been lurking in these forums for a while, and suppose it's time to finally dive right in. I'm in the super mileage vehicle team in our university, and the whole point of our club is to build a vehicle that gets the best possible gas mileage.

    My club has spent a whole bunch of time with composites, and we can make some pretty nifty stuff (fully enclosed aerodynamic fairing, carbon fiber "skateboard" chassis). We've reduced the total weight of the car to ~40 kg and have an estimated drag coefficient of .1 .

    However, in our rush to get the space age stuff done, we've neglected the most important part: the engine and drive train. We plan to use a Honda Gx25, and a double reduction gearing to drive a Nu Vinci hub. The gx25 has a custom made EFI setup and will be optimized (hopefully :whistling:) later.

    The competition that we enter is requires that the "car" (it's really a motorized recumbent bicycle) travel an average of 15 mph for 9 miles. This means that we use a burn and coast strategy where we accelerate till around 30 and coast until we're almost stopped and then we accelerate again.

    The major complication is that I have no idea how to properly setup a double reduction gearing. Many of the projects I've seen on the site use the station gearbox. However, in our current mounting, there is little room for the box.

    More info on our EFI setup is available here (

    Thanks for this wonderful resource!

    PS I plan to motorize this old beach cruiser I have laying around the house. Any suggestions for a good frame type setup?

    Attached Files:

  2. skyl4rk

    skyl4rk Guest

    The GX25 is a good choice. You will need to work out how to freewheel with the motor off, most kits cannot do this.

    Pulse and glide is definitely the way to go.

    I have heard the NuVinci has a lot of drag. If you really need a geared system, I would suggest using a chain drive to an internal gear hub. If you don't have a lot of money, find an old Sturmey Archer and rebuild it.

    If you don't use an internal gear hub, you might be able to use one of these freewheel kits and pull start (?) the motor every time you need to pulse.

    If I were doing it cheaply, I would build a jackshaft similar to the Dax Titan and use a pull start motor with a freewheel hub. The trick is to get the right rpm range to pull you up to 30mph hopefully within the optimal torque curve rpm range for the engine.

    As for the frame motor setup for a cruiser, good luck, I am also looking and have been for years. The best option right now is the Grubee kit:
  3. mabman

    mabman Member

    Can you pedal?

    Gas only. Answered my own question.
    Last edited: Dec 26, 2008
  4. skyl4rk

    skyl4rk Guest

  5. mabman

    mabman Member

    Because their engine is modified via the fuel injection and whatever else they have done to it, a dyno test would have to be re-run. Those graphs are for the stock motor.

    I think the Nuvinci, in that it is an internal hub, with their new developer kit View attachment DevKit_Datasheet_V4Web.pdf that is due out any day would be an excellent choice for this project. The ability to program your gearing via software is about as exact as it can get. Contact the factory direct.

    Any drag from the hub, and all hubs create drag, could be reduced greatly by having a freewheel between it and the other power sources. Something like this, P1010048.JPG

    You know of course that the driver of that rig is going to want to turn the Ipod up to 10 and be wearing noise cancelling head gear as running an engine in that shell is going to be LOUD.
  6. DougPan

    DougPan New Member

    Thanks for the replies, guys. It's been pretty helpful.

    To skyl4rk:
    Thanks for the many links, I'll take the time to slowly go through them and learn what I can. I am pretty new to a lot of bike terminology and that Sheldon Brown website is pretty fantastic.

    To mabman:
    The vehicle's motive force can only come from the engine. We currently have a set up where the engine will start reliably. However, the big problem, as you've stated, is to generate an accurate torque curve ( or even better, a brake specific fuel consumption curve). I've looked around this forum for a couple of ideas, and am trying to decide between two setups: an inertia dyno and one that drives a large electric motor. I think that the one with the electric motor would allow for longer tests thus letting us play with injection times and what not. But the inertial one seems much easier to make.

    Secondly, thanks for the developer kit attachment. As a team, we were thinking of using either a linear actuator ($$$) or a tiny dc motor with an encoder that's been geared up (down?) so that it's able to pull the shifter. That way, we can link the motor rpm, and wheel rpm in order to have the nu vinci shift juuuuust right.

    Unforunately for the driver, we haven't taken her comfort too much into consideration yet. :sweatdrop: And the suggestion of noise canceling headphones is a good one. Another thing we were considering was an insulated enclosure for the engine. Since we pulse and glide, the heat generated when the engine is running will keep the engine warm and help avoid combustion in the transient regime. IE. when the motor is cold, it will not be as efficient as when the motor is hot.

    Thanks for all the help guys. I really appreciate it. Is there any interest in the making of the aerodynamic fairing? If there is, I'll put it together. It's Christmas break, and I might as well do something useful. :eek:

    Edit: I started studying the attached picture, and am not too sure what benefit there is in having the freewheel as opposed to a live axle would be. It seems that for a bicycle setup that has to juggle between pedaling and motor input, the freewheel is there so that the rider can rest his legs on the pedals as the motor revs up.

    Edit 2: That said, I would really appreciate it if you had pictures of that set up from different angles. Other examples of that would be great too! Maybe I'm just not "getting" it.
    Last edited: Dec 27, 2008
  7. mabman

    mabman Member

    I have some better pics around here somewhere of that setup and another currently under development and will try and dig them up later. But for our needs we want to be able to pedal, motor and coast independently while minimizing drive train drag especially for the pedaling and coasting part, yet also pedal and motor concurrently and as pedaling is not part of your parameters perhaps that setup would not be valid. The idea is to insulate each drive with a freewheeling mechanism, in this case simple bmx type freewheels through a jackshaft.

    If you are solely dependent on the motor for your drive system and want to coast efficiently then you still need to reduce drive train drag to a minimum as you know, while keeping weight to a minimum also. As I said all internal hubs have drag, they also are heavy. The Nuvinci setup with the servo will give optimum shifting/rpm but maybe the weight and complexity is a detriment and you would be better off with a plain old derailleur type system and a well trained pilot? A small tach could be enlisted and optimum rpms/load determined easily on a small dyno to achieve this? As old as it is the old style chain/freewheel system is actually still the lightest and most effective drivetrain. It has less of a co-efficient of friction than the new belt drives actually. But the belt drives do not require maintainence of course.

    Correct me if I'm wrong but the way I see it your greatest fuel expenditure will be from a dead stop? So the gearing you have to get off the line is pretty important in that it will do so effectively and that means a pretty low gear ratio to start. Because these motors all require gear reduction I think that the simple 5:1 box that is available through the scooter parts places like the one in the photo previously supplied would be a good place to start. It is relatively light in weight and compact. From there you can design your additional reduction necessary via sprockets probably? You just need a top gear that gets you to 30 which can be calculated easily using your wheel size and sprocket teeth front and final. I would say off the cuff that a 120" gear would do the job for 26" wheels and more for you as your wheels are 20"?

    It just occurred to me that WWU here is probably in on this competition and I should see what they are up to. I talked to the head of the dept. there at one point and he wasn't at all interested in what we are up to trying to get MAB's recognized as LSV's.
    Last edited: Dec 27, 2008
  8. mabman

    mabman Member

    And as long as your offering I will take one of these:
    oct 091.jpg

    I like this design because it looks like it would really easy to insulate the engine from the cockpit via a simple insulated bulkhead.

    And btw I thoroughly enjoyed your google pages, at least what I could understand and think that fuel injection system is bad. Now I want to make one!

    As far as the competition goes I realize that there have to be parameters set and that 9 miles and 15 mph average is it. But I am more interested in the real world environment and i think that FI could help in that realm most of all that I saw on there. The current setups flirt with 200 mpg but as you load things down the mileage goes down also. Also for us it is a matter of how much you are willing to pedal which can be a big gas saver. My personal goal is 300 mpg with a bike capable of carrying 250 lbs. of cargo, while carrying said cargo. A fraction of 2800 certainly but I want to average 30 mph over 900 miles at that number.
  9. skyl4rk

    skyl4rk Guest

    The NuVinci is a great system if you have power to burn. If you are in a mpg competition you cannot afford the internal losses due to friction. Note that you will be spending about two thirds (or more) of your time freewheeling with the motor off. Low drag during this period is critical go overall mileage per gallon.
  10. ZnsaneRyder

    ZnsaneRyder Member

    Since it's a small engine, how about using a centrifugal clutch with bicycle gears and shifter to get your MPG? Any type of CVT will consume more fuel. A clutch on the other hand, will fully lock at high RPM and you don't waste revs.

    I'm getting 100+MPG with heavy 196cc engine and only one ratio, no shifting, so I know that with the right gearings, your 25cc should get 200MPG.
  11. mabman

    mabman Member

    It is going to take more like 2850mpg to win that competition. Last years winner had 2800.
  12. skyl4rk

    skyl4rk Guest

    It depends how the track is set up. A lot of the mpg runs have a rolling start and they just keep rolling until it is done. So there is never a dead stop during the contest.

    The questions to answer for the pilot are:

    1. What is the "start motor" speed
    2. How much throttle to apply
    3. Which gears to use while accelerating, and what are the start and stop speeds for each gear
    3. What is the "stop motor" speed (30mph)

    As long as the pilot has a speedometer and knows which gear they are in, they can do the actual driving by feel. An rpm gauge is useful but not necessary.
  13. DougPan

    DougPan New Member

    We currently have the engine set up so that we have a centrifugal clutch and one way bearing so that we can accelerate accelerate the wheel, cut off the engine, and let the wheel coast without moving the chain. A video of our intended setup can be seen here (

    In the video, the wheel is running with minimal load and probably spinning way fast. Therefore we need some way to gearing the engine output down (up?) so that the chain speed that the wheel "sees" is in the vicinity of a bicycle's chain speed. That way, as mabman pointed out, we have flexibility in terms of which gearing type we use.

    Hmmm. You're right. I was going off the specs provided by Nu Vinci in terms of drive train efficiency. According to "Bicycling Science" by David Wilson, chain and sprocket systems have efficiencies in the high 90's (95 -98%). I had thought about how heavy the wheel would be after sticking the nu vinci hub in, but did not realize that the car would also suffer a ~10% drive train efficiency drop.

    I guess I'll be paying for the hub out of pocket and keeping it for myself then. No matter, the hub is cool enough to use for my own projects.

    The car starts from a dead stop, so the initial acceleration to top speed would consume the most fuel for that specific pulse and glide cycle. However, the track that we will race on is notorious for strong winds, so depending on the wind gods (lateral winds please!), there may be an entire section of the track where we're just burning gas just to maintain speed.

    For an ideal case of a flat track, no wind, we still need to determine the actual drag coefficient of our fairing and actual wheel friction numbers. Based on that, we think our minimum speed of the line will be the point where the losses from the wheel are equal to the losses from wind drag. And the maximum speed will be the lowest possible speed to achieve the 15mph average. Could you elaborate on the 5:1 gear box reduction? I tried a simple google search for "scooter gearbox" and didn't get too many pertinent hits.

    Regarding the EFI, the benefits of EFI increase the more time the engine is running, but right now, the guys who did the project can't tell me if it's more or less fuel efficient than a properly calibrated carb. :sweatdrop: The big problem is that the oxygen sensor that we use is meant for use in cars so it needs ~6 seconds to warm up and after that we aren't even sure if there is enough exhausted gas to properly activate the sensor. Delphi makes a mini-switching 02 sensor, but I haven't had any luck in finding someone who actually sells it.

    If you end up competing at the shell eco-marathon (31 days till end of registration!), maybe I'll see you there. It's a fun time, even though it's in the middle of the semester. Bring the bike too, it sounds pretty neat.
  14. mabman

    mabman Member

    I don't have any designs on that competition, it is not real world enough for me. As I said I would be happy with a fully loaded cargo bike with motor assist getting 300mpg on surface streets while able to keep up with the flow of traffic. Going to France would be ok I guess.

    Here is one source for the gearbox. It is cheap and it reverses the rotation of the crankshaft also please note. But it is the smallest and lightest gear reducer that I know of and can be gone through I think to optimize it's function.

    There are some better shots of the blue bikes drivetrain here taken by the bikes maker, Jim G. That is the same type of gear reducer on that setup.

    As you can see the bike is pretty burly and not a wind cheating machine at all. But form follows function and you can see how it does here in town at this vid:

    This is another prototype that we are working on: P1010779.JPG
    We are trying various types of off the shelf drives in order to shave down the cost of the final product but haven't hit the lottery yet there. We are focusing on load capacity at this time because it works and we had to choose between that and about 5 other concepts for monetary reasons.

    Thanks for opening up this aspect to the board. It really makes one think!
  15. DougPan

    DougPan New Member

    Thanks for the links. Googling "scooter transmission" yields a lot more useful results.

    The vid was also pretty eye opening. I didn't realize these motorized bikes went quite so fast. I had initially wanted to make a recumbent bike frame to act as the super mileage vehicle's chassis, but welding takes a long time to learn how to do it well.

    I'm curious as to how you plan to improve the load capacity. I thought that with more load, the only thing that could be done is to lower the gear ratio. And that decrease in gas consumption mileage is unavoidable. What ideas are you thinking of?

    Even though it may be "wind cheating", a front fairing/windshield should greatly reduce the drag. One of super mileage vehicle's teams predecessors was the human powered vehicle teams. They were around seven or eight years ago, set a string of world records, then disappeared. Here is the design report for their 1999 tandem cycle. Even though fairing manufacture is a REALLY labor intensive project, it's pretty neat.

    One other thing I was thinking of (which may not be feasible) is to distribute the weight of the cargo to both the front and the back even though that may compromise the ability to steer well. Having all that weight in the back, can't be good for the rolling resistance of the rear wheel.
  16. mabman

    mabman Member

    A big part of our high mileage plan includes actually pedaling along with the motor:sweatdrop:. If you don't want to pedal you get worse gas mileage. Pretty simple really.

    Based on our experiences riding tandem bikes over the years we are not that worried about trying to distribute the load evenly front to rear. These bikes are about the same wheel base as a tandem and so with the motor/drivetrain/fuel load weighing in around 16 kilos that leaves us with 90 kilos of capacity safely. Plus a trailer can be towed. But for us pedaling efficiency is about as high a priority as engine.

    The blue bike has a 20" rear wheel which is inherently strong but weak when it comes to gearing. The current gear ratios allow for good torque off the line but the pedaling stops at 25 mph and the motor will pull it to 35 on level ground. If you can pedal at that speed you can "get on top" of the motor, similar to riding with a strong stoker on a tandem, except there are no humans that can put out the wattage a 1.5 hp engine can:D

    The next bikes will have the ability to run larger diameter wheels up to 29"es which will help alot on top end gearing. But we are thinking that you could also run 24" diameter wheels if you need a lower COG and are shorter in stature. The bb height will be lower but for on road use that shouldn't be an issue. As you can see on the blue bike it is really high and the bike handles well enough but could be better. Especially with some fatter tires because the roads tend to be rougher than you think.

    Distilling alcohol is next on the agenda after the bikes are somewhat dialed in. This will open up many possibilities for the future of this type of LSV and others.
  17. skyl4rk

    skyl4rk Guest

  18. graucho

    graucho Active Member

    A lot of the computer programs you use and fuel intake are over my head. But, I do under stand the EFIE portion
    (due to hydrogen fuel cell installations in vehicles)

    Doug, would your competiton allow any perpetual motion in your design? A few years ago I made a perpetual motion
    wheel enclosed in one of my hydrogen bike-experimental builds. It would add weight, but in a good way. We machined
    a perfectly balanced disk, 1"center shaft, 18" wide, 1" thick, complete with magnetic assist. (28.4 lbs)

    When we had it mounted on the axle (with no load) at 35 mph it spun for over an hour before we stopped it.
    Between the weight and magnetic assist It didnt want to stop. When the bike was completed I would pull in the
    clutch between 25-30 mph and I could roll for a good mile.

    I wanted to get 100 miles from 1/2 litre of hydrogen out of a 350cc engine. Im more of a backyard inventor and
    no where close to you guys. Your goal of 2850 mi WOW! makes my head spin. I just wanted to know if these
    principals could relate or help in your cause? Or maybe you have to go strictly from engine and intake mods. I wish I
    was younger and lived in CA. so I could hang out and learn from you guys. Good luck on your project. graucho

    Attached Files:

  19. mabman

    mabman Member

    Hey Graucho, we don't see nearly enough of your work around these parts. You include little dribs and drabs but how about some content!
  20. John-Forrest

    John-Forrest Member

    Getting max gas mileage...

    It should be obvious that most of us got into this way of getting around because we're trying to cut costs and save gas. We have reached peak oil production and are going to have to start rationing what's left while we figure out other ways to get energy. Biodiesal fuels would be ideal for these kind of bikes since we use so little compared to cars. I've heard of using algae in ocean to make fuels.

    The chart is very educational and answers a question of my own. What is the best way to save gas while riding? Good maintenance is a start - if you have a loose sprocket or spark plug is coming out, you are not going to get very good gas mileage.

    I shut off engine whenever I can. I shut it off at stop lights and enjoy a few minutes of quiet, then pedal to get going again. (If a car starts honking behind me, to **** with them, I'm setting an example.)

    Going down steep slopes is an ideal place to start coasting. Pull in the clutch handle, lock it, and fly! You can actually go faster since engine tends to get in way. Then just as you start climbing the next hill, pedal it going again. Occasionally, I just like peace and quiet, so I pedal on flat ground and get some exercise. If you have a light bike, you can just use engine for climbing.

    Going up steep slopes I really think it helps to LEAN towards hill - just like you would on a bicycle. If engine starts working hard, I also help it along by pedaling.

    I recently got an 80 cc engine so I could climb hills better and keep up better downtown. (I think it's really safer to be in middle of lane, since people getting out of parking lots can't see you very well.) I thought I'd be going a lot faster, and I can get up to 35 mph but I don't like it since engine starts working too hard and vibration loosens nuts.

    I discovered engine is most "comfortable" at 20 mph and there I save most gas. It's quieter and better for engine. I actually get about the same gas mileage as a 49 cc. As pointed above, even on a 150-250 cc motorcycle you get best gas mileage at 45-50 (around 70) and can get similar gas mileage to a moped riding top speed.

    Also, you can take time to enjoy ride, look at the scenery, and feel a LOT safer. Who needs to go fast?

    Along that line, I notice that downtown when some car impatiently swerves around me, I catch up with them at the next stop light and the first to go. I've wound up going over a couple of miles and get to my destination the same time that car did, and have less time parking, so I'm in the store before that car driver!