[python] Re: New Jetrike Rev B Plans

  • From: Henry Thomas <whpthomas@xxxxxxxxx>
  • To: python@xxxxxxxxxxxxx
  • Date: Wed, 24 Jan 2007 20:42:49 +1000

Sorry, I meant to send this reply a while back, but saved it as a draft and forgot about it until now...

25hz wrote:
I think lots of builders and riders of more "unique" HPVs think certain
design aspects magically cure all handling ills.  Some others I have talked
with who have also been considering a tilting rear end delta are no

Yes, while I hope I don't fall into this category, I still may. I am not expecting the trike to have any better high speed stability than the python, but I do expect it to perform better than a no-tilting trike.

The reason's for building a trike are numerous, and include: better low speed stability on steep hill climbs, keeping your feet clipped in when stopped, better load carrying ability and a wider birth -- forcing cars and trucks to slow down and wait for a passing opportunity.

Even though the tilting and steering mechanisms aren't directly
connected and one, in theory, would be able to function without the other
with a little rider finesse, they will still interact.  If a bump is hit
with the left rear tire (for example), regardless of the linkage design
(swingarms, parallelogram, etc), the left wheel will rise and through the
connecting linkage(s) try to force the right wheel down.  If there is no
depression on the right side to coincide with the bump on the left, then the
only method of reacting to the bump is to raise the pivot point up and to
the right in a small arc.  On any tilting delta I think, this is going to
result in a steering input.  Unfortunately homebuilders usually don't have
access to exotic light materials or the technology needed to eliminate bump
steer without making a bent that weighs a ton.  I've ridden trikes with
pretty bad bump and brake steer, but it's amazing what you can get used to
if you really want to.

My expectation is that the tilting mechanism on this will halve the distance that the frame is deflected compared to the tire. Given that the tilting is independent of the steering (like the front wheels of a tractor), I would also expect that the deflection will be parallel to the front wheel, aside from the (hopefully) small forces induced by the self-centering effect, I hope the amount of lateral movement will be minimal -- but I guess this is where I hope "certain design aspects magically cure all handling ills".

I think that method of tilt limiting is an excellent way of doing it so you
don't have to add separate stops and plates.  If I read it correctly, are
you saying that your tilt limit so far is about 30 degrees from vertical due
to the CoG concerns?  With your design, item 35 is almost wide enough to put
three holes in it so likely close to 25 degrees of adjustment can be added
by simply adding the holes.  This of course will change the swingarm height
unless the tie rods can be adjusted sufficiently, putting extra adjustment
holes in the end of the rocker arm wouldn't help with this because that
would change the rocker arm angle, but more on that a bit later.  As for the
cam, that's way too detailed for me :)

One of the things you should be mindful of with this design (having simulated many variations) is that you can end up with geometry that has the opposite effect, causing the seat to move towards the pivot point as it leans. This in effect accelerates the lean once it has begun, it would feel a bit like wheel flop -- this was what actually cause most of the wheel flop in my first prototype. And as you did with the python, I used the parameter driven CAD to attempt to simulate it and get a better idea of what was going on and what worked.

When you have started working on your version of this design, please send me your geometry and I can at run it through my simulation and give you some indication of what you might expect from it. The model runs from a design table, so it is very easy to add a different configuration to it. Once I build my trike and feel confident in my understanding of how it works, I will post some geometry recommendations for different wheel sizes and tracks.

This is exactly right, the geometry is all about getting the
inside swing arm that you are leaning on to move down by a
smaller amount than the arm on the outside that is lifting
you up. I have not included the main steering pivot raising
in any of my assumptions, I have actually treated them as
independent, as they probably should be. In my experience the
tilting mechanism should function independently from the steering.

Cool.  That's an intersting way of using that and somethign I will
definitely keep in mind for future tilting/steering designs.  On a normally
steered FWD bike or trike, you could separate the steering and tilt no
problem.  On a python though, I can't imagine the amount of strength and
body english required to be able to ride in a straight line with the python
tilted to the side :)  Maybe some of the European clan has done something
like this with their python deltas?
I guess this was the other reason I though an independent tilting mechanism was important for a python front end. On a sideways lean -- like a banked track -- it would be very difficult to steer a non-tilting python trike, because of that massive negative gate, and all the weight of your body then pushing on the steering. On my trike I hope and would expect that the rider would be vertical. This does raise and interesting point. That is where to place the handle bar. I am going to but binders on both the drive train just in front of the pivot and on the rocker arm. I don't at this point know what will be more important, controlling the tilt or controlling the steering.
So, 190 degrees is the optimum angle to produce the best amount of lift, or
the best amount of lift with angle B?
My new geometry page http://jetrike.com/geometry.html explains this a bit better now, but angel B is the important one, Angle C has a small additional effect.

Gotcha, I didn't think of that.  With the first tilting delta, while wanting
to keep it as short and narrow as practically possible, the rear wheel could
and did hit the back of my arm sometimes on hard corners.  This wasn't fun
with the screws in the tires for ice racing :)
Because the tilt locks, I simulated that in the design, and made sure that the tire is about 50mm behind the seat at full tilt. The swing arms also move on a plane parallel with the center line, so you don't actually turn into them when it leans, so contact isn't going to be an issue.
When I  first tried my python, I believed it was
unrideable, yet, my friend got on it and was riding it within 5 minutes.
While I understand your assertion that changing parameters will have a
detrimental effect on the self-righting aspect, the python is an amazing
little beastie and it does all kinds of things you wouldn't think a bike
could do and still be rideable.  Adding a tilting, lifting, CoG moving delta
rear end to it is just a bunch more variables to toss into the mixing bowl.
I must confess a little anxiety about this, all the simulation in the world isn't going to tell you how it is to ride.
I'll do up a 3D version with a few of my mods and your dimensions but put
extra holes in the tie rod flanges and make sure the tie-rods have as much
adjustability as possible.  After changing and testing as many things as I
can on it, if it doesn't work, then it can always be reset back to your
original specs.  Did you experiment with item 35 being a little longer in
conjuction with longer or different swingarm angles?

Item 35 can be longer, as long as the tie rods can also be lengthened. The only thing I was concerned about was them scraping the ground at full tilt over uneven ground.

That's too bad that your steel selection is somewhat limited, but it's cool
that you're using high strength steel.  I only use mild steel because 4130
is about 17 times more expensive than mild.  1" x .063 box tube is about
$1.10/ft while CroMo is over $18/ft.  I also prefer square over round stock
because it is much easier for me to work with and align, so that sort of
limits my choices too.  I found some engineering formulae to calculate the
bending moduli of square and round profiles and I found that mild steel is
more than strong enough for what I build and the huge increase in cost for
stronger steel does not justify (in my mind) saving 3 or 4 pounds in frame
weight that is going to be absorbed by things like fairings, racks, fenders,
bags, lights, etc etc etc.
Yes, the Chinese steel is rubbish, but the Aussie stuff is top notch :-)
Is the tie rod coming off the end of your rocker arm a little further to the
outside than the mounting point on item 35?  I'm thinking I'll find out what
the total deflection of the rocker and offset it to the outside of the
swingarm mount about 30% just so there are no extreme angles that the heim
joint can't handle.
IT probably should be a little to the outside, given that the rocker arm arc is always pulling inwards when it is at its most forward and rear position.
Very cool, I like the suspended and hinged rocker arm. I have 12" of #60
durometer coming, but I have a shock that I want to use instead.  I think
it'll look cool to have a shock absorber mounted sideways and acting as
suspension.  I guess I'd have to make sure I tune the suspension then to
hold the proper angles with rider weight on it.  As for the bearings and
housings, are you sure you really need them?  I tinkered with small engine
repair and mini-bikes and motorcycles and some of the old ones had nothing
more than a bolt through a 1/8" wall steel tube for the swingarm suspension
and flat washers on the ends for thrust bushings.  I have also ridden a
number of homebuild suspension trikes with the same setup and there was no
play at all.  Even after time, if play does develope, seeing as how the
rocker arm pivots are always under load, they should take up the play too.
It's just a prototype, so I think I'll just use a thick walled tube with a
bolt and some grease for now.  Actually, two 3/8" rod ends should work for
pivots too mounted to 1/4" steel.  Actually that is likely the way I'll do
it - much simpler and it'll be doing what the rod ends are pretty much
designed for.  Hopefully the tilting tadpole will be ready for powder by the
end of the weekend and then I can sit down and see what I can come up with
for a 3D design so I can get a working set of dimensions for my existing
Yes it probably is over kill, but I got the bearing housings back last week, and boy they look cool -- photo's attached.


JPEG image

JPEG image

Other related posts: