I'm in the middle of making a small-scale aesthetic prototype, which I'll post about later, but in the midst I made this little video showing how the double parallelogram tilting front suspension works. The prototype is made out of black foam core with hockey tape in the joints. Bone simple, very cheap and easy to modify. More on that later.
They were apparently absolute murder to ride, but nothing quite says vintage transportation like the Velocipede. The thing I've always found fascinating about these is their huge front wheels. The physics of a bike like this must feel very odd compared to bicycles of today. The force of angular momentum of a wheel (the force that makes it want to stay upright and that leans it when you top turn it) is directly proportional to the torque moment of that wheel. Or, said in non physics babble, the bigger and/or heavier a wheel is, the more stable it is when its rolling. It's kind of an amazing force, when you think about it. The two tiny rollerblade-like wheels of a little razor scooter generate enough angular momentum to hold an adult upright. In larger, powered two-wheelers like scooters and motorcycles, wheel size makes a huge difference in the character of the machine. Scooters have traditionally had much smaller wheels. Most vintage small-frame Vespas had only 8" rims, while their large frame brothers weren't much larger at 10". My modern Vespa has 12" wheels, but still feels very much like a scooter — agile, and a tad butt-heavy. Motorcycles tend to have anywhere from 14" to 22" rims depending on the style. When I test rode a Triumph Bonneville in 2008, what I noticed immediately was that it took about half the speed I was used to in order to get rolling stability. Once moving, the larger wheels and rolling mass of the Bonneville felt so solid — like it was on rails. What would wheel size mean for the Streetliner?
Bigger front wheels
While playing with the proportions and broad strokes design details of the Streetliner this week, I wondered what would happen if I swapped out Burgman-size front wheels for larger wheels from another bike in the Suzuki fleet. The size of the rear wheel is likely fixed because of the shape of the engine/transmission casings. Big wheels up front would definitely give it a stronger visual connection to the old race car designs that have inspired the project. Especially if I grabbed the spoked wheels off of something like the Boulevard S40. But beyond aesthetics, it would make some significant differences to handling comfort, lean feel, and possibly even make the front suspension easier to build. It looks the business though, doesn't it?
That's an 18" rim up front as opposed to the 14" stock Bergman rim on the rear. It definitely calls back some nostalgic racing mojo. Larger diameter wheels would mean that the front end would take bumps better and have greater rolling stability. It'll also mean more room inside that wheel for all the tilting mechanicals, brakes, and steering attachments. What it would also mean is that the lower and upper swing arms of the tilting suspension could be further apart — letting me have not only better ground clearance, but deeper leaning on the same distance between the wheels.
If you go back to my initial lean study, I explored the relationship between body height and leaning. With the larger wheel size up front, a couple of good things happen. First, I'm able to spread the swing arms apart, meaning that at the same lean angle, there's less interference. This is important at both the wheel attachment points and where the dampening assembly will go (the shocks and their connecting structure, which isn't shown). The more room there is to work in those assemblies, the better. Not only will it be easier to build, but it'll mean more opportunity for building fine tuning and adjustment points right into the structure.
The biggest difference may indeed be felt in leaning stability. The greater mass of those larger wheels would mean nearly instant leaning stability. Whenever I see a Honda Goldwing slowly lean its way through an intersection turn, it amazes me that the 900 lbs or so of the Goldwing are able to be held up at such low speeds. The Goldwing has an 18" wheel up front and a 16" wheel in the rear, and that's enough to keep it stable, even at low speeds. With two 18" wheels up front, the leaning stability ought to be massive. But more than that, having the swing arms further apart vertically, means that the side-to-side forces generated by the wheels when they precess (based on steering input) will not only be greater, but have more leverage on the body of the vehicle.
The one thing I wonder about, however, is what it will do to the driving feel of the vehicle? Greater stability means that it will take more input force on the steering to induce the lean — and that force will grow with speed. That's a good thing, as you wouldn't want the vehicle to be all twitchy at 70 mph. I'm not worried about it, I just wonder what the feel will be. I imagine it wouldn't be much different than a big cruiser motorcycle with the Tilting Motor Works kit installed.
From the beginning, I've envisioned the Streetliner with wheel pants for added aerodynamics and efficiency. This inspired by both '30s era aircraft like the Gee Bee but also similar projects like the Aptera. Exposed wheels simply aren't very aerodynamic. Putting a streamlined shroud around the wheels will add effeciency (and look fantastic). However, I'm not exactly sure just how much efficiency is gained. Adding wheel pants will definitely add time and complexity to the build, and at some point I'm going to have to do an aerodynamics study to see how much more efficient the shrouded wheel is than a naked wheel. One thing I have to keep in mind though, is that unless I can put a decent tail on the pants, there won't be much advantage in the wind. However, that tail will quickly limit how far I can turn the wheels (because the tail of the pants will hit the body). That means a large turning radius, and for a vehicle intended for mostly city use, that's not a good thing. So if I have to sacrifice wheel pants for the sake of turning ability, that's a compromise I can't not make. But I do just love the look of the pants. I hope I get to keep them.
With the broadest strokes of the exterior design penned as far as is practical at this point, it's time to do some real work. Up to now, I've pretty much just eye-balled it. With the pieces starting to come together, it's time to get more specific and more precise. So today I explored the relationship between body ground clearance and tilt geometry. There are a lot of factors effected by this relationship. Some I anticipated, some I didn't.
The major factors are these:
- Lean angle
- Ground clearance
- Center-of-gravity shift
- Eye height of the driver
Tilting Study A
I started with a low profile body height that would keep the suspension swing arms essentially level. From studying the Brudeli Leanster, it looks like so long as the swing arms are parallel, they need not actually be level. This makes sense as I start to lean the elements visually. One of the things I was most curious about was the relationship between body height and maximum practical lean angle. Tilting things over in Illustrator, it looked like 35º was about all I was going to get before bodywork was likely to start scraping. My first surprise was how far the rear wheel slid right, which made instant sense, but was unexpected. It's exciting to see just how far over that body swings from centerline — pushing the center of gravity deep into the apex of the turn. Given that it's nearly impossible to low-side a tilting trike, moving that much mass to the center of the turn has me pretty jazzed thinking about the cornering capabilities of a vehicle like this. My one hesitation is in line-of-sight. Sitting too low is a major safety concern.
Tilting Study B
So what happens if the body gets picked up quite a bit higher than the wheels? Using the same length swing arms, this obviously brought the front wheels in tighter to the body. The end track of the vehicle is yet to be determined, and it's not as though the swing arms can't be lengthened if needed. I was again surprised at just how far the rear wheel swings from side to side. A deeper lean angle of 45º seems likely in this arrangement. 45º is considered the standard of a deep lean, as this is the depth to which sport bikes are able to lean — their riders laying padded knees into the asphalt. In essence, this is what having two wheels up front does for the Streetliner. There is an issue with this kind of arrangement, however. A closer examination of the swing arm attachment makes it look pretty obvious that unless I bend the ends of the swing arms, they're going to interfere with the outside wheel. Also, if the body is too high, then I'll need a really wide track to meet one of my minor design criteria — that leaned fully over, the vehicle will not actually tip over and fall if not moving. With no ability to put my feet down, this little detail is crucial. On the positive, it'd mean a nice high line-of-sight. It would seem prudent to find something in between.
Tilting Study C
In splitting the difference, I found a nifty elegance. By aligning the bottom of the body with essentially the axle height, I think I've found the sweet spot. 40º seemed easy with only the wheel pant in danger of interference. I bet 45º would be possible depending on the engine and transmission casings. With further study and measuring, this configuration yields an eye height of approximately 47". This puts it almost exactly where my line-of-sight is in my MINI, which is actually very good. But more exciting is as much as 9" of ground clearance. That should make for both very clean aerodynamics and easy road going over fairly substantial road hazards.
Obviously, this is all just estimation, but it's been a fascinating thought experiment. It's also yielded a deeper understanding of seating ergonomics, but that's for another post.
Thanks to the Yahoo! tilting vehicles group, I just discovered the Brudelli Leanster. The front end geometry is functionally the same as the Aprilia Magnet and the Tilting Motor Works design. One thing it does differently is that it has a tilt limit mechanism that keeps the bike from leaning over too far. As the photos show, the bike is able to lean over full tilt and not fall over. That's what I want for Project Streetliner. It's very nice to see that it's possible. I'm also encouraged by the structural details of the frame and suspension components. They seem to be the kind of things that would be straightforward to fabricate. Very cool.
For more, check out their website: Brudelitech.com
I love LEGO! The next best thing to working this out by hand with a big pile of LEGOs is using their software to render a mechanical prototype of Project Streetliner. Obviously, the proportions aren't any kind of exact, but seeing the basic shape and structure in 3D has me pretty stoked. Also, having to work through the basic mechanical details (what stays solid, what's allowed to pivot, etc.) has given me a much better understanding of the mechanics involved. All in all, it's not that complicated. It's also really interesting to start thinking about the structural considerations of the safety cage / chassis.
Now that I have this so much more planted in my mind, I'm actually tempted to skip the functional model all together and go straight for the Pathliner prototype. That's probably not a great idea, but it's a temptation all the same. I just want to keep moving — keep making progress.
LEGOs are a great way to prototype something mechanical. But what if you don't have enough of the right kind of LEGOs? Well that's where the LEGO Digital Designer software comes in very handy. It's basically CAD for block heads. Here I've rendered how my front end could work were it built out of LEGOs. I really think like I'm getting very close to where I can proceed to the next stage of prototyping: a functional small-scale model.
In the past few days I've had a terrific email exchange with Bob from Tilting Motor Works. His company has developed a fascinating kit to convert essentially any standard motorcycle into a leaning tadpole trike. But more importantly, he's done over 10,000 miles of real world road testing on his tilting suspension design. I asked him if directly linking the shocks to each other had any adverse effect on handling. Specifically, I wondered if hitting bumps while leaning or in other tricky positions would give the vehicle any proclivity to change course suddenly or alter its tilt. This was his response:
Yes, the force can be transfered directly to the other wheel but my experience shows me that it is no issue even while cornering or going over railroad tracks. Actually all suspensions ultimately transfer the force to the other wheels, just not as directly. I tested my suspension by laying 2x4's all around my cul-de-sac and running over them at different speeds and angles. The bike performed great.
This is such great news! A front suspension setup similar to what Bob's developed seems the perfect match for Project Streetliner, but I didn't know how capable that kind of dampening really was. I also asked Bob about comfort and he was happy to report that his front end is far more comfortable than stock and that his VMax prototype is always his first choice in a stable of fun vehicles. He's graciously offered his time and expertise and I'll be giving him a call very soon. Thanks, Bob!
This submitted to Tumblr by lamidesign:
I wonder if there could be a more mechanically simple solution than tilting the wheels. The wheel tilt is all about moving the CG - the contact patch is the same, and the tires really don't care if they vertical or tilting.
Can the entire sprung chassis pivot, or rock - I'm thinking like the inverse of a hammock between front and rear supports. If the body of the car leans into the turn then the wheels do not really have to.
That's a good thought. I had to think about this for a moment and based on my research, here's what I've concluded.
You are correct about contact patch — the tires won't care one way or the other and shifting the CG is indeed what adds to the dynamic stability of the vehicle. But there's one key reason why the front wheels need to tilt with the body: conservation of angular momentum. Utilizing the physics already induced by the spinning wheels to lean them over (and the body with them) is already the simplest in terms of mechanism (even if it's a little harder to wrap your head around). I can let the tilt mechanism free float and lean the vehicle just with the steering inputs — like I would on a motorcycle (via counter-steering). If I keep those wheels straight up and fixed but want to lean the body, I'd have to use a powered mechanism to force the body to lean into the turn. Were it just a hammock, like you've described, the natural inclination of the body would actually be to move the CG to the outside of the turn (because of inertia) which would quickly flip the trike.
The same gyroscopic force that holds a two-wheeler up is also the force that allows you to lean that vehicle via counter-steering. If I were to use a delta trike (two wheels in the back) instead of a tadpole trike, then I could use the angular momentum of the steerable front wheel to enter and exit the lean like you would on a motorcycle — all while leaving the rear wheels fixed (there are a number of bicycles that work this way). Unfortunately though, on a vehicle the size and weight of Project Streetliner, I'm afraid that the single wheel might not actually provide enough leaning force unless it were a pretty massive wheel. Furthermore, the complexity of driving two rear wheels instead of just one becomes just way more trouble than it's worth.
In 1983, Mother Earth News built a three wheeled leaning vehicle called the 3VG. It used hydraulics to lean the body of the vehicle into the turns. They seem to have incorrectly assumed that a 3-wheeler could not tilt on counter-steering alone:
Although we tried a number of different mechanisms in our efforts to find a system to induce camber in the car (because it has more than two wheels, it can't lean naturally like a motorcycle), only a handful proved successful. The "inclination" is at present controlled through a combination of mechanical, hydraulic, and electronic components ... all of which are available as "off the shelf" equipment, and some of which are so basic that they're commonly used in many toys and pocket calculators.
Both the Piaggio MP3 and the Tilting Motor Works prototypes prove that natural leaning in a tadpole trike works just like it does for a conventional 2-wheeler. They later describe the supposed problem of the force it would take to lean a vehicle of the size of their prototype:
It'd be ideal, for example, if the lean system were a stone-simple mechanical affair with a minimum of moving parts, but research showed us that such a design has some real drawbacks . . . one of which is that it takes almost superhuman force to initially lift the vehicle from the full-bank position. That's why we use hydraulics.
As I read that again, it sounds like they were again operating under the assumption that tilting the front wheels and using counter-steering simply wouldn't work. They don't really say if they ever actually tried. It sounds like they were trying to lean the body independent of the wheels, which would certainly require amazing force to overcome the weight and inertia of the vehicle body — especially one as large as what they built. Given what I've seen from both the MP3 and the TMW bike, I think that the size and weight of the vehicle could indeed have a significant impact on its ability to utilize counter-steering. I can only conjecture that the Mother Earth News boys either never tried, or that the significantly larger size and weight of their vehicle was more than the angular momentum of the wheels could easily overcome.
My design spec is to keep the vehicle weight under 600 lbs. The Piaggio MP3 500 weighs in at 530 lbs, so if I'm able to land anywhere in that weight range, then I should be just fine. That does bring home the point, however, why this vehicle isn't going to be something that anyone could just get in and drive. If you're not already a competent scooter or motorcycle rider, the coordination necessary to pilot a vehicle such as this would be completely lost on you, as you're essentially turning the "wheel" right in order to go left. That's hardly intuitive, even though we've all done it on bicycles since we were kids. That's also why it won't have a steering wheel at all. Ha!
Thanks so much for your submission!
I got an email a few days ago from Bob over at Tilting Motor Works. His tilting trike system is designed to be bolted onto the front of virtually any motorcycle. It appears to be using the same basic geometry as the German Tripendo recumbent bicycles and the Aprilia Magnet concept trike. His 10,000 miles of road testing seem to be a pretty sound endorsement for this kind of geometry. Good stuff. $10k is definitely beyond my budget for Project Streetliner, but I appreciate Bob reaching out and hope to still have him involved somehow.
Some may wonder why I'm insisting on a tilting 3-wheel design for Project Streetliner. Watch about :30 of this video and it's pretty obvious. The trouble with non-tilting trike is well explained by Wikipedia:
The disadvantage of a rear drive, non-tilting three wheel configuration is instability - the car will tip over in a turn before it will slide, unless the centre of mass is much closer to the ground or the wheelbase is much wider than a similar four wheel vehicle.
So unless the wheels tilt, you've got to essentially crawl around corners. I'm not planning to race this vehicle, but I do want to be able to take evasive action in traffic. So tilting it is! A vehicle like the Morgan was low enough and wide enough that it doesn't really have this issue. But if I want to really benefit from a smaller frontal aerodynamic cross section, I'm going to need to keep things narrow. The advantages of a Piaggio MP3-style front end keep piling higher and higher.