Frank D.B. wrote:

A question for Alegzandar Kovatsch:
>
> http://www.jestartech.com/jestar_TTW/Jestar.html

The pictures at this website do not show any steering linkage,
but there are kingpins. Is that correct?  It actually looks like
the driveshafts control the steering. How does the steering
work?

That question is ok.
Those pictures are made just to persuade some professors to let me do my own project for my diploma thesis, because I was a student of department for structural and dynamic analysis and optimization, and I want to do my diploma thesis with department for vehicles, where I haven't passed a single exam.

The diagram explains a lot about the steering system.

The diagram represents the dependency between the angle of turn of front wheels and roll angle (fi - o with vertic line) on a 2F1T TTW.
The theta (O with horizontal line) angle is for the outer, and psi (half circle with vertic. line) is for the inner wheel in a curve of the radius of 20 m.
What was varied is the angle alpha (from -15 (front of axis down) to 20 deg (front up)) - that is the angle that pivoting axis closes with the ground.

Horizontal axis of the diagram represents how much the vehicle rotated around the roll (pivoting) axis - in general case that is not the actual roll angle (angle between the vehicle and the ground), but it is always pretty close to the actual roll angle.

On diagram you can see that needed turn of front wheels (in a constant radius curve) increases as the vehicle leans, if the alpha is negative (theta and psi for alfa = -15 deg) the values for theta and psi go up as the "roll" increases. That is because the rear wheel gives the "parasite" steering that demands additional steer input. That parasite steering of the rear wheel increases as the roll increases (as Mr.Foale said in his overview - I am just giving you the exact numbers for a certain case).

What is even worse, the theta and psi don't climb equally. Theta - psi difference changes as the fi changes.

Theta - psi is constant only for alpha = 0. That is also the case when theta and psi are constant during the entire range of fi.

That means that the required turn angle of inner and outer wheel in one curve does not depend on the roll angle only if the roll axis is horizontal.

That is why every patent application of 2F1T demands pivot axis horizontal. That is the only case when you can use the Jeantaud trapeze or principles of Ackerman Steering for calculus of geometry of the steering mechanism. I couldn't use them, so I didn't put that system on the images.
I knew all that before having this diagram, because I knew that the position of axis of the rear wheel will not be predictable without any calculus.

If the pivot axis, (alpha>0) is front up, than, when the vehicle leans, the rear wheel goes inside, steering the rear outside, (just like 4WS cars in the "parking lot" mode) which means that you must steer less if you tilt more, and this more/less relation also depends on the angle of inclination of the pivot axis.
Furthermore, the psi and theta curves intersect at 0 value - that means, at a certain roll angle, for a certain construction (alpha value) you will have to steer the wheels left when you are turning in a right curve because the rear wheel gives too much steering - and as the diagram shows, that will be some serious counter steering at the extreme tilting angles (aprox 40 deg tilt).

Further problems relate to the drop of the rear wheel downwards as it leans in curves.
That will lean the pivoting axis back end downwards, and even if we design pivoting axis horizontal in zero tilt, it will change as the tilt increases.
So the vehicle will not have perfect (predicted by calculus) steering when I most need it.
I have composed the equations that will correct the values in the diagram that I'm giving you, and I can tell you that the corrections are in order of size of a degree or two, or even less, and that is all a lot actually for a high performance vehicle (entire setup range is for some things not more than that).

The vehicle should be actually built to have certain behavior under certain circumstances. I believe that the angle of pivoting axis should be incorporated in zero tilt so that it gets "perfect" when the vehicle assumes position where it has to work "perfect"...
So after all this talk about the angle of pivoting axis (and we could talk about it forever), I can tell you something that I'm sure you already know, even though I don't have an idea of who you are - a perfect angle of the pivoting axis, as well as probably the steering mechanism (and all of that together with a bunch of parameters not mentioned here - even though they all effect each other), will be one for Jerez, and some other for Suzuka, and some third setup will be the best for some third racetrack.
This means that a certain TTW can be "built for curves". Here I talk about the variable tilting axis angle (wich I have all figured out), just like the fork angle is adjustable on BMW telelever front fork, or just like shock absorbers are adjustable and so on. That is of course only if we speak of the competitive use of TTWs.
For "civilian" use just one setup could be, obviously, good enough.

On the other hand, it is possible that some high performance design of front geometry of the 2F1T TTW could be very hard to steer and control for "civilians" every day. That is not unusual, a friend of mine has a Ducati 748 motorcycle - it's radiator boils after 15 to 20 minutes of city ride from cafe to cafe. Does it mean that Ducati is a bad motorcycle?
No, it just wasn't built for that purpose.
The same goes, will go actually, for TTWs when they take part in covering the streets.

In simple words, some layouts of steering system should take in consideration the roll angle, and some shouldn't.
I have a couple of solved examples of steering systems, and what is more important to me is the principles that were used to design the mechanism. Anyway, I will use Pro Engineer software to design my mechanism, and I don't have to have anything else but "just" a good idea. I just don't believe I'll have the enough of time to "test drive" my TTW in Pro Mechanica, since I have deadlines, and my work done so far is already better than some degree projects I've seen.

Steering system could be designed as an all mechanical system, or it could be some processor that controls hydraulics that controls wheels and takes the driver's steering input as one of the parameters when making a decision about what to do with the wheels.
This may sound too complicated, but in fact, it is nothing but just another black plastic box under the seat, together with a bunch of other black plastic boxes. I see it that way, since it is not my job (reliability of electronics in steering is a serious business, ARK (active rear kinematics), is the only system that I know of, developed by BMW, and it was difficult to them, all I know about such system is that it must get stuck with only driver's input in case of breakdown of electronics, and that is the idea to start with).

Drive shafts on Jestar are a different story. As it is shown on the site, and as my teaching assistant said, Jestar eats standard CV joints for breakfast, lunch and every time he is bored.
As a matter of fact, CV joints are built for some known uses, they stand maximum angles when the load is minimal. You turn the steering wheel to the extreme only during the parking lot maneuvers, when you don't press the pedal to the floor (if you like your CV joints that is). I have contacted a small company in NY, and they make custom cv's for 700 HP machines, but even their cv joint steel balls wear out (although they don't know of titanium spraying).
In angles that TTW demands from CV joints, they would wear out even faster because, we twist the accelerator on the way out the curve, so when the cv's are under "bad" angles, we put them under the worst loads.
All this time I am talking about CV joints on the way out of the diff case, and tripod joints (which are on the diff side in the cars) on the wheel side of the drive shaft, (I am talking about my TTW layout). That is because I don't desire the turning angles that car wheels have, since I have 1300 to 1400 mm wheelbase (like GP motorcycles), on a 1200 mm wide vehicle, so I don't need maximum values of steering angles that cars need.

Second solution (there are several that I though of) for transmitting power from a tilting engine to the non tilting wheels would be a solution with diff case in the back, and a pair of cone sprockets coming out of it, sending two drive shafts (one on each side) to the front pair of cone sprockets and than via second pair of drive shafts to the wheels, in that way the vehicle could tilt 55 deg to each side, with all CV joints working within 15 to 17 degrees to each side, and that is ok, except for the fact that:
a cone pair to near 90 deg turn
a cv
a drive shaft
a cv
a cone pair to near 90 deg turn
a cv
a drive shaft
a cv
on each side of the vehicle, is what will bring extra cost and losses, so my advisor just said: "Delete that!"...
I still believe that two longitudinal driving shafts bringing torque to the wheels have something to do with the right solution. The point of bringing the torque from the tilting engine to the non tilting wheels is in not disturbing the tilting. We don't want to twist the accelerator and hit ground with the shoulder one nanosecond later, right...

I am not thinking of a high performance TTW just because I'm young and stuff...
High performance is the area where problems faster come out on surface.
Just like chain forces on motorcycles were irrelevant until somebody raced a motorcycle with over 100 HP...

For example, imagine that steering and tilting are somehow connected, over an active tilting system...
It sometimes happens with motorcycles (not if one rides like a grandmother) that rear end slips sideways, and you have to input some additional steer to the opposite side to maintain on wheels (lean and steer opposed to usual, like on bike speedway racetracks).
You'd have to have some kind of a button to switch off the tilting system for a "short amount of time", or should your processor "recognize" the situation, and how sure could it be...
This will not happen to a TTW that goes to bring the groceries... Usually...
How would the just mentioned TTW behave?

And I'd like to say something to Mich:

Imagine a system, a mechanical system, where a valve opens when some regualtor (a pendulum) detects lateral acceleration or something like that, and oil pump starts pumping oil into a cylinder and makes the vehicle tilt, keeps it there while a certain "forces situation" is present, and gets it back or pushes it further in case of change. With balance established valves get closed, tilting and non-tilting parts get "frozen" in the position...
No artificial input data but what was given as construction parameters (valve diameters, strokes and so on) is present.
A natural mechanism, no driver or processor input.

Would you call that an assisted or natural lean?

In this mail I was talking about 2F1T TTW layout.
Everyone who might see this letter is welcome to comment it, there is a lot more to talk about the steering of TTWs and pivot axis inclination!

After studying the breathtaking nice pictures of the Jestar TT of
Alegzandar I have some doubts concerning the drive shafts and couplings.
I think they have to bend much to far to be reliable. This could be
better if the centerline of engine outputs of the shafts comes in the
same sphere as the centerlines of the front wheels Depending on the
amount of lean this could work  but I still have my doubts. If the
engine/clutch/gearbox/differential combination is based on an actual
existing combination I like to hear which combination that might be.
 
 

Engine on the pictures is "my own" design and you will see it later, when I update my site.
Every composition with longitudinal engine, T case in front of it, and diff in front of the T case is OK, such combination is Fiat 750, which I was took a look at just a couple of days ago, but it was low performance (32 HP) and heavy (about 130kg)...
That is usually the combination from a vehicle that has rear wheel drive and engine, and then you have to replace the half shafts with those that come from a vehicle that had a front wheel drive. Of course, it is lot easier to say it than to do it. Further, sometimes, like in case of that Fiat, it is possible that the manufacturer has an engine that was supposed to be elsewhere, in the front, while you can actually put them together.
Fiat has that gearbox from a rear wheel and engine 750, that you can combine with a 1300 engine that was in the another model's front.
That would be what I like to call "pocket money prototype"...
That is the layout of old VW Bettle, or Porsche 911, with changed driving shafts. Renault 4 had similar layout, if we talk about the "pocket money prototype"...
Do you think that TTW with Porsche 911 machine could impress the world?
I think it could , I just don't have any money to tear apart an old 911...
I believe that there was a car manufactured in Soviet Union, with engine in the back, gears and diff in front of it, air cooled V4, although, I'd rather tear apart a 911 Turbo then pay for fuel for a soviet vehicle...