[python] Re: Python trike plans

  • From: George Durbridge <gdurbrid@xxxxxxxxxxxxxx>
  • To: python@xxxxxxxxxxxxx
  • Date: Thu, 24 Dec 2009 10:35:23 +1100

Stefan,

Markus is right about the construction of your design C.  If you erect a
A-shaped frame on a rear axle, supporting a pivot behind your head, with
another pivot under the seat, you have great freedom in locating the
pivot axis: it can go through your body, and be angled in ways which are
impossible with with the usual Python under-seat pivot.  You could even
arrange positive trail, as your diagram shows.  In addition, the
construction is simpler, the loads in the bearings will be much lower,
and ground clearance can be much greater for a given seat height.  A few
similar designs have been made, such as the STrike and some beach
trikes.  The same idea was applied to a bike in Wim van Wijnen's Ecocar
2000.

I suggest that you not try it, however, as the usual STRike layout has
poor dynamics.  For a trike to be stable under cornering forces, the
centre of gravity (CoG) should be as low as possible.  Modern bike tyres
will generate about 1g force, but trikes will typically tip over at much
lower cornering forces.  (Except that it would also affect braking
power, some trikes could be made safer by fitting them with worse tyres,
because then they would slide at the limit, instead of tipping.)  My own
trike, which is a touring tadpole (2 front wheels), brakes at nearly 1g,
but tips up sideways at about 0.65g.  This problem is worst with delta
trikes (2 rear wheels) under brakes, when the limit on cornering force
can be less than 0.5g.

If it moves while cornering, the CoG of a trike should move to the
inside of the corner, and/or downwards.  The usual STRike layout has the
opposite characteristic.  That layout is like a hammock: the pivot axis
passes above the rider's centre of gravity, from his head to the front
of the seat.  That makes the trike self-righting: gravity returns the
seat to the centre and the steering to the straight ahead, as soon as
the rider stops forcing the trike to turn.  Although the rider appears
to lean into a corner, in the STrike layout his weight moves to the
outside of the corner and upwards, which drastically reduces the
cornering force the trike will generate before it rolls over.

You could produce the opposite effect by putting the pivot axis below
the rider's CoG, though the trike would be unstable and I am not aware
of an example.  Imagine that instead of having four legs, your chair was
clamped to a slanting rail which pivoted near your feet and behind your
back.  To steer, the rider would lean into the turn, shifting his CoG
down and laterally towards the apex of the turn.  This action would
increase cornering power, perhaps greatly, but the rider would need to
use his own strength to straighten up (and even to steer straight),
perhaps assisted by springs.

As between A and B, both Jurgen and Esko Meriluoto (Hipparion) seem to
be persuaded that it is best to keep the steering pivot axis as close to
your own hip joints as you can.

Regards, and merry Christmas and a happy new year to all listies,

George

On Wed, 2009-12-23 at 20:50 +0100, Stefan Bartels wrote:
> Dear folks, after a pause caused by illness and hard work I now have
> all the wheels and materials together to begin building my trike.
> I can do three variants which I made sketches of. In the third version
> the seat swings on the beam, the others are a bit more conventional.
> Where should I put the hinge? Which design will be most stable and
> have an acceptable turning radius?
> All three wheels are to be 20".
> Your ideas and advice would be most welcome!
> And have a happy Christmas!
> Stefan
> 
> address is http://www.free-archers.de/files/galleries/246/recumbent%
> 20trike%20variants.jpg
> trike


============================================================

This is the Python Mailinglist

http://www.freelists.org/list/python

Listmaster: J�rgen Mages jmages@xxxxxx

To unsubscribe send an empty mail to 
python-request@xxxxxxxxxxxxx
with 'unsubscribe' in the subject field.

============================================================

Other related posts: