Tilting Quad Velomobile


Fair to say this is the most technology we have ever seen on this forum !

Blown away Paul
Thanks Paul.
It' been 14 years of try "this" or try "that" as I started this project with minimal theoretical knowledge. Much of the "technology" is the result of practical failure -vs-success with the failures far outnumbering the successful elements.
While developing the caster wheel & hub, concurrent with the tilt mechanisms, I was also internet searching and experimenting to find the ideal steering linkage to compliment the caster steering.
The relationship between F to C and the steering linkage proved to be bothersome. On the one hand, some form of linkage was needed to provide Ackerman geometry as well as to keep the front wheels aligned and to steer when maneuvering backwards. But, on the other hand, the direct linkage must not interfere with the F to C action of the front wheels when moving forward, and yet, must somehow work in concert with the tilting and suspension actions of the front wheels.
Well, according to my wife, the prolonged pursuit for the best-practice linkage somehow crossed the threshold from perseverance into stupid stubbornness. I believe her exact words were "blah, blah, double stupid"
After numerous failures, some so so results, I came up with the linkage shown in this photo.

The linkage consists of the following parts observable in the photo above.
--two bicycle brake levers to be mounted to the tilt control lever handles (not shown) in a manner which allows me to push the brake lever with my thumb while grasping the tilt lever handle.
--two cables attached to the thumb levers and linked together by a short bicycle chain which is housed within a cassette consisting of two small idler sprockets.
--a steering shaft with a nine tooth sprocket fixed at the cable cassette end ( or posterior end ) and a steering knuckle fixed to the anterior end. The knuckle pivotally connects two short rods to blade- like constructs, which will be pivotally attached beneath the upper front swingarms. These blades can move synchronously ( up and down ) with the swingarms as the vehicle tilts, but, they can also pivot laterally to move the tie rods to alter the steer angle.
Hopefully, when you see these parts in there proper places and moving back and forth in the following videos, this will make more sense.
For the next two videos I "butchered" ( couldn't resist using Paul's fitting expression ) the upper right swingarm of a previous vehicle and inserted a thin piece of aluminium to reduce the bulk of the swingarm. This skinny substitute allows us to see the lateral movements of the pivot blade and tie rod as the right wheel is turned to the right, a little to the left, and back to straight ahead.

The frontal view of the next video provides an opportunity to see the action of of the steering knuckle and the linkage to the tie rods. The steering shaft and affixed knuckle are located behind the square tube of the mainframe. The hole in the square tube provides access to a bolt which holds the shaft and knuckle in place. Rotating the shaft and knuckle clockwise/ anti-clockwise impacts the linkage to the tie rods.

Next time I will describe the rear wheel drive as it is today and what it might eventually look like.
Thanks to the members who cautioned me last year about the legal considerations of installing a Series Hybrid Drive system. I subsequently decided to incorporate a jack shaft to drive the right wheel and later this year I plan to install a hub motor to drive the left wheel. Almost the mirror image to the drive-train Hugh has put into his revamped delta.
The chain, brake, and aero-covers will be the focus of this post.
This is what it looks like now.

And this is what it looks like from above the wheel

The brake is tucked away within the concave wheel. After I install the front brakes, all 4 wheels will be fitted with non rotating aero-covers. The following videos show the chain drive operating without a cover, with an aero-cover and finally with a chain cover.
Operating with the non rotating cover in this next video
I will be poking holes in the cover to cool the brake. But first, I'll test the effects of the regen-braking from the hub motor on the delta model before I poke too many holes. The next video shows the rear drive operating with both an aero and a chain cover.
The covers shown in the videos are proof-of-concept quality. I plan to test different materials, sizes, and designs because I don't want to "butcher" the shell's tail section any more than is necessary to accommodate the tilting action of the rear swingarm at the chain-line.
My next post will focus on how the drive, tilt, and steering work together as I ride the quad along the straight and narrow and around curves.
Using blue cardboard to simulate a bicycle path, I will attempt to show how the drive and tilt impact the steering.
For the majority of my riding time, while pedaling along straight sections of a pathway, I simply hold the tilt levers to keep the the vehicle in a substantially upright position. No manual ( thumb ) steering input is required. The tilt mechanisms function like a variable lean-lock device which I can adjust to accommodate variations in road surface and slope.
Although It's not observable to me while I'm pedaling forward, I'm aware that the forward rotating front wheels are constantly and freely adjusting ( castering ) to my tilt control inputs which keep the vehicle steered along the straight and narrow.
I try to illustrate this in the following two videos,First I pedal slowly forward in a straight line while holding the tilt levers stationary. In the second video I put both feet down and push to coast backwards in a straight line. No steer inputs are made in either direction.
Well, I actually put my feet down and push twice in the next video.
When entering a bend in the pathway or a sharp corner I increase the tilt just enough to free-to-caster around the bend or corner at my current speed. Again, manual ( thumb ) steer inputs are not required.
In the next video I try to demonstrate the F to C steering action while slowly negotiating a curve to my right. I start by first tilting the vehicle from side to side to indicate that controlled tilting (without manual steer input) will be used.
When I tilt the stationary vehicle there is no change in the steer angle, but as I pedal forward and actuate the seat and levers to tilt the vehicle, the forward rotating front wheels change their steer angle in response to the controlled tilting of their steering axis to follow the curve in the pathway. Although F to C operates at this slow speed, it works much better when accelerating from a stop into higher speeds where it performs smoothly and predictably (always the correct steer angle for the tilt and speed of the vehicle). A bonus is, that actuating the tilt control mechanisms becomes easier with increased speeds as the dynamic forces kick in.
Riding this prototype is quite basic. Unlike the engage/disengage requirement of most steer-tilt- control vehicles, this prototype does not require transitioning from manual to caster and back to manual steering at a certain speed threshold, and there is no obligation to think about the technique of steer/counter steer. I simply control vehicle tilt at all times and rely on free-to-caster steering to get me around the bends, corners and straight sections as F to C operates from full stop to full speed and back to full stop.
Since assembling this prototype and riding it within my neighbourhood, I have found that that manual steering is rarely needed. And when it is used, I just hold the vehicle in a substantially upright position and engage the thumb levers for steering.
In the next video I demonstrate this approach on a curve to my left. I begin by reaching for the tilt levers and press the thumb levers to show what I do to alter the steer angle of the front wheels when riding. I then pedal slowly forward, holding the tilt levers stationary to maintain the vehicle's upright posture and apply steer inputs with the thumb levers. I then maneuver backwards using thumb steering
Operationally, this vehicle is a lot simpler than my previous ones. As shown in the above video, when I'm thumb steering, I'm not tilting. And, as shown in the one before, when I'm tilting, I don't have to steer.
Although I have not yet tested this prototype in all riding situations, I'm satisfied with how it performs so far. But, I'm a little disappointed with its weight. I had hoped that removing the hand cranking drive train and building a lighter two-part mainframe would have reduced the original weight of 36 Kg by more than the 3.7 Kg I succeeded in shedding.
And, although I have constantly looked for ways, I haven't been able to significantly reduce the number of components or their complexity, except for the delta model, which has one less front wheel and spindle, three fewer swingarms, a simplified steering knuckle, and a less intricate nosecone.
I look forward to your replies- especially any suggestions on how I might be able to simplify any of the components.
Four photos illustrate the form and the materials I used to make the nose cone and cabin section.
First photo is of the male mould for the right side of the nose cone.

I made separate moulds for each side with the intention that I would convert them into left and right side female moulds which could be bolted together for the fiberglass layup and then separated to release a complete nose cone with a smooth exterior finish. That didn't happen.
Instead, I settled on pulling off separate F. G. sections from both male moulds and then I epoxied the two halves together, Since laying up the fiber glass onto the outside of the male moulds was challenging enough, I decided not to attempt laying up the material inside of a two part female mould.
As you can see, the nose cone ended up looking a bit odd with its multiple curves and indents. The central indent is for a sectional cutout for the front swingarms and tie rods to move freely when tilting and steering. The deeper indent, which looks like an elephant's ear from from an oblique angle, is to allow the front wheels to steer sharper angles. The outer edges of the "ears" portion is where the cabin's frame settles onto. And the stickie-outie top and back of the nose cone permits the front of the cabin section to tilt forward (clam door opening) without bumping into the front of the nose cone on the quad or the front wheel of the delta.
Admittedly, the quad's nose cone resembles the skull of some prehistoric creature as seen in the next three photos, especially with the skeleton-like 12mm square tube aluminium frame of the cabin section nestled behind it.
The three photos show the nose cone and cabin frame together as viewed from the side,

from above,


and from the front
Sorry for the lying down frontal view. And the "top view" also looks oddly inserted. Well, at least all three photos got inserted.
The oval shape in the mid-top of the cabin frame is where the windshield will be placed. The hole in the front of the nosecone is for inserting a headlight, and the smaller side holes are where the front swingarms and tie rods will stick out to slide the inserted pieces of coroplast up and down as the vehicle is tilted from side to side.
The four photos were taken some time ago, but my inability to recall how to insert photos on this site has kept me from posting till now. Fortunately, a much younger "assistant" is helping me with this post and I'm taking notes.
Since the cabin section is almost completely finished, it shouldn't be too long before I can post the next part of the build.
My next post will focus on how the nose cone and cabin section are attached to the quad's frame, and how the ventilation and clam door opening of the cabin are designed to work.
Wow you don't take any prisoners do you ?

That looks very very good , your fiberglass finish looks flawless (y):)

Can't wait to see it balanced on the chassis

Thanks Paul. It must be the lighting, as the flaws in the fiber glass nose cone show up in the next photos.

VELO'S CABIN --Dimensions and Access
In the photo below it appears that the track of the outbound wheels is wider than the velo's shell. But the cabin's widest part is actually the same dimension as the distance between the outsides of the fiber wheels.

The cabin's width-height relative to the compact and sub-compact is readily observable.
At 75 cm wide the velo is narrow enough to fit through the back door of the garage, while the cabin's height of 140 cm provides a 10 cm space between the top of my head and the roof. The seat height of 45 cm is a combination of "the ease" of getting "into and out of" at age 77 and my desire to "see and be seen" in traffic.


The fiberglass nose cone is colour matched with the yellow coroplast side panels. The roof is clear plexi painted yellow and the small yellow visor-vent located below the windshield is made from aluminium sheet bent to match the contours of the blue plexi extending from the windshield to the nose cone. The blue central section and the clear plexi to each side is actually one piece of plexi which wraps around from one side to the other and reaches back towards the side windows. The windows slide up and down inside of plastic moulding inserted within slots that I cut into the the square tube frame.
My plan is to post a video to demonstrate how the sliding windows and visor-vent are actuated. Assuming all goes well with video inserting.o_O

The next photo shows the cabin lifted and tilted forward to rest on the front swingarms between the nose cone and the front wheels.

Somehow I ran out of space to complete the previous post. Still learning the ins and outs of posting.
In the above photo there are two short steel rods with screw-pins inserted at the tops to keep the cabin from tilting further forward. (One of the rods is vaguely visible between the right foot pedal and the nose cone). And these rods also help to guide the front of the cabin back down to its correct position on the nose cone, which is also when the back of the cabin settles down onto the tubing at the seat area. This tubing will serve as the front of the frame for the tail section of the shell.
The clam door cabin design provides easy access and exiting as well as a "quick removal" feature.
In my next post I hope to include videos of the cabin's ventilation (sliding windows & visor vent) and its quick removal feature.
Looks great. A quality build.
Thanks Popshot.
I'm pleased with the "rounded" form of the cabin compared to my previous mostly coroplast shell. But, I'm beginning to think that plexi is not the ideal material for this. I was supprised that I was able to bend the plexi around the curves of the tube frame and insert it into the yellow plastic moulding which is attached to the square tubing. Initially, my plan was to make the roof from fiber glass, but this "no-mould" quick-fix plexi approach became difficult to resist.
Although it all holds together, there is tension, and I'm concerned about developing major cracks in the plexi. I already have some small stress cracks around the screw holes where the plexi is fastened directly to the frame.
I've been "baking" the cabin in the hot sun, hoping to relieve some of the sprung tension. But it's difficult to know if it's helping without removing a section of plexi to check out its formation.
I'm hoping some AZ members have experience with this and can inform me on the best approach.
BUILDING the SHELL continued.
Cabin Ventilation
As I sit in the cabin, I slide the window down and up as I open and close the visor-vent. For this video I stuck a strip of white hockey tape to the top of the window so that its movement would be more visible. Each window can hold any position along its up and down travel, but the visor-vent remains totally open or closed.

Cabin Removal
Prior to recording the next video, I unscrewed the pins from the tops of the short steel rods to allow the cabin to be lifted all the way off. I discovered that the cabin's weight and its wedged position between the nose cone and the tube frame of the tail section keep it snugly in place when the cabin is closed.

Well, I did it again. I ran out of space below the inserted video before I completed typing in the text to end the post, or rather, to introduce my next post which will focus on VISIBILITY and RIDE-FEEL with the cabin attached
One word ...WOW. Nice job hopefully you will have many km's of enjoyable rides. If my experience is any indication be prepared for a lot of attention from the public.
One word ...WOW. Nice job hopefully you will have many km's of enjoyable rides. If my experience is any indication be prepared for a lot of attention from the public.
Hi Hugh.
Thanks for your (+) response. I must admit, after 15 years of tinkering, I'm looking forward to more riding. And I'm really looking forward to one day getting it to the e-assist status of your delta.
The photo below was taken in a nearby park along a fenced-in pathway. This is my view from inside the cabin.

I tried to make the cabin feel "roomy" and hopefully enhance the ride-feel by having the clear plexi extend from the side windows down to the nose cone. This gives me a clear view of the front wheels and along each side.
The photo shows that I have a clear view of the front wheels and the pathway, but I'm beginning to think that the amount of plexi exposed to the sun may result in a need for greater ventilation-that the sliding windows and visor-vent may not be enough, especially after the tail section is attached. Lots to think about. And the tail section will take sometime to complete.
In the meantime, I plan to test-ride the quad with just the nose cone and cabin attached to get a better feel for it. A few years back, when I rode my proof of concept models, I became aware of how the shell not only blocked my view of the front wheels, but also, how the shell's weight distribution affected the ride-feel, especially when tilting at slower speeds.
So my short term plan is to check-out the ride-feel with and without the cabin (quick removal) to compare the quad's performance when cornering at slower speeds, when encountering side winds, when traversing a side-slope, plus riding over speed bumps and potholes etc.
I guess my next post will focus on a specific test-ride result.
Man, I see here a hud projecting additional information like airspeed, target aquisition, temperatures and personal satisfaction. I see also the gazilions of buttons, levers and switches activating " Gerwalk".
Click for DIY Plans!