[python] Re: Pedal Induced Steering

On other lists/pages/forums that refer to recumbents, it's also called PIS
or pedal induced steering.  This may be all over the map a bit, but
hopefully I get the information across that I want to, wihtou being too
epic.

I think there's a lot more going on than just frame geometry of a bent with
regards to the phenomena.  In my experience from building 17 or 18 bents of
different designs and modifying 5 or 6 other bents, the majority of the
issue is the rider and how smooth they ride and how much effort they want to
improve their riding skill/style.  Even on normal uprights I have seen
people go down the street as if they were a swimming fish, and again, this
has nothing to do with the bike and everything to do with the rider.  Even
with the same rider on the same bike, if they ride with no hands, the PIS
goes away.

Now, that's not to say that some designs are not easier to ride smoothly
than others, but I have yet to ride a recumbent (I've ridden close to 30
different styles and shapes of recumbents in the last 3 years) that COULDN'T
be ridden smoothly.  On conventionally steered bents like SWBs, or on SWB
tadpole trikes, if they have close to 50% of the weight on or near the front
wheel(s), they have a tendency to produce pedal steer fairly easy because of
the weight over the front wheels (and the front wheel(s) contact patch ends
up being more like the pivot for a teeter-totter) and also because the
weight is constantly shifting forward and back, which adds to the "pulsing".
Unless the bent is SO front heavy that it wants to tip forward even while
sitting still on  flat ground, even those can be ridden WITHOUT pedal steer
IF the rider expends the effort to concentrate on spinning smoothly and
evenly.  I've ridden SWB tadpole trikes that were so front heavy that on
even a moderate down-hill with very light braking, they would start to lift
the rear wheel off the ground, but again, with effort, PIS could be
eliminated.

My brother is a grinder and uses a TON of body language when riding.  DF,
high racer, trike - doesn't matter.  Riding behind or beside him, you can
see his arms moving around, his shoulders moving up and down and his hips
twisting left to right.  On an upright, this "dirty' pedalling doesn't seem
to produce as pronounced of an effect because subtle weight shifts along the
entire height of the body can take it out of the equation.  On a recumbent,
especially ones with laid back seats, all of this extra body movement turns
into steering input because you can't use body language to compensate.  When
my brother rides my one trike, he gets a ton of PIS.  He tells me it's
"twitchy".  Even though he constantly squirms all over the place, when I
tell him the 'twichiness" is because he's not sitting still, he adamantly
claims he's as steady as a rock.  While pedalling, if I tell him to let go
of the handle bars of the trike, it immediately stops swerving and tracks as
straight as a train on rails.  This is because by letting go of the
handlebars, he has lost the leverage points that he uses to push and pull
himself around in the seat while pedalling.  When I ride this trike, it
tracks perfectly and in fact, when I raced it on the velodrome last time, I
let go of the handlebars around the banked corners and let the trike pick
it's own line based on the speed I was doing.  It worked fine.  Also, on
most of the trikes I build, by riding with no hands, you can steer the trike
by leaning left or right.  This wasn't intentional, it just seems to work,
and again, there is no pedal steer while doing it.  Even on bents that I can
ride smoothly with no pedal steer, if I exceed a certain pedal RPM, my
pedalling technique starts to get sloppy and pedal steer (not to mention a
"pulsing" sound coming from the tires) is generally the result.

The nice thing about non-center steered bents is that the "power system" and
the "steering system" are separate and they can remain that way unless
sloppy riding habits combine them.  On the python, it's obviously a little
more complicated because those two systems overlap.  You can also steer with
your hips, your shoulders and even slight down-pressure on the "handlebars",
but it seems, at least to Marcel and I, that most of our steering is done
with our legs.  While accelerating hard, we still tend to steer with our
legs/feet (even though we don't want to) but we can counter that steering
input by using our hips and shoulders and try to keep the python going
straighter and take the power lost from going left and right, and try to
redirect it back forward.  Marcel spent a lot of time on racing road bikes
with clipless pedals and learned how to pedal very smoothly a long time ago.
as a result, he is extremely smooth on the python, which enabled him to not
only learn to ride it in a matter of minutes, but also enabled him to race
it on ice as well as race it quickly SLALOMING on ice :)  When our legs get
tired though, we both get pedal steer.  When my legs start getting tired I
actually INDUCE pedal steer.  Why?  Well, when I'm tired and trying to ride
perfectly straight, it takes a fair bit of concentration to keep it going
perfectly straight and react to small steering corrections.  If I use pedal
steer on purpose (and I'm talking just a small amount to make a small
swerving motion) I'm always either in a slight left turn or slight right
turn.  The act of turning helps stabilize the python for me and it's easier
to ride because I don't need to try to stay in a "neutral" upright position.
So with always being in a slight, gentle turn, it makes it much more
relaxing to ride.  Obviously, I don't do this when I'm in tight traffic or
when I see cars coming up behind me :)

At any rate, in the end, pedal steer, I believe, is a fact of life on a
recumbent because of their design.  Regardless of the design of the bent, I
have yet to ride one that can't be ridden smoothly if you really want to.
Conversely, I have yet to see a bent get ridden by a new rider that DOESN'T
have pedal steer, regardless of how smooth or easy it is to ride.  :)  It's
a natural biomechanical function for our entire bodies to move in rhythm
with the motion of our legs.  Only through conscious effort and practice can
we reduce and eliminate it.  While some bents might be "easier" (relative
term . . ) to ride than others, to get ride of pedal steer, you just need to
find the rhythm of the bent and adapt yourself to it, and the pedal steer
will subside.  People who ride on rollers a lot are extremely smooth riders
because in most cases the rollers are only 12 to 15" wide.  Sloppy riders
who swerve a lot either learn to control extraneous body movement or they go
off the rollers a lot.  On the high racer I just finished, I used a recycled
road bike frame (http://www.fleettrikes.com/high%20racer%20stage%206.jpg).
If I don't pedal smooth and allow my shoulders to roll and arms to move in
sync with my legs, I can get heavy pedal steer.  If I pedal circles and
relax my arms, it tracks straight as an arrow.

Try this experiment.  Find a chair that spins fairly easily.  Lean back in
it and pick your feet up free of the chair base and the ground.  Hold your
arms out at shoulder height and slowly alternate arms "punching" the air in
front of you.  Now, in the same position, but this time hold your arms down
at the sides of the chair, and once again alternate arms punching downwards.
In both instances the chair will oscillate slightly left and right.  Even
with simple motions like that, the body wants to "do it's thing" and shift
to accommodate the arm action, not to mention the various laws of physics
applying themselves.  The difference in those two motions is that the
vertical punching is akin to riding a DF where the body, weight shift and
subtle handle bar movement can compensate for that yawing action, but on a
recumbent your body doesn't have the same range of mobility to compensate
for it.  Un;ess there are serious alignment problems with the frame or it is
actually cracked/damaged somehow, I think just about any pedal steer can be
mediated by the riders' technique.

> Hello all,
>
> I would like to forward a mail from Doug Burton,
> concerning TT and Cruzbike frame geometry and
> the recent "Pedalling Steering Interferences" discussion:
>
> ======================================
>
> Sent: Saturday, April 22, 2006 5:01 PM
>
> Dear Jurgen,
>
> I have been very interested in your development work,
> and have followed your research for the Airbike and
> Python. Thank you for being generous with what you've
> learned and for stimulating the discussion of
> front-wheel-drive bicycle dynamics.
>
> I am particularly interested in the PSI discussion on
> the "Modifications" page of your Python website.
>
> I have been working for about a year on variations on
> the Cruzbike kit, which converts a standard Y-frame
> mountain bike into a FWD recumbent. You may know of
> it. http://www.Cruzbike.com.
>
> I have built two versions of the Cruzbike which have
> given me some experiences around the PSI discussion. I
> have included photos of them annotated along the lines
> of the other photos on your Modifications page. If
> they can further stimulate discussion of the PSI
> topic, feel free to use them.
>
> Cruzbike 1 had a large distance between the hip line
> and the steering axis. It had noteable PSI. I later
> modified the seat mount to allow the seat to slide
> forward on the top tube. The closer the hip line moved
> to the steering axis, the less-noticeable the PSI
> became. Note also, persuant to the statement about the
> "A-B" distance in the "TT Improved Jurgen Maier"
> photo, that Cruzbike 1 has a shorter
> crank-to-contact-patch dimension when the hip line was
> far from the steering axis. My initial impression is
> the hip line effect is stronger than the "A-B"
> dimension effect.
>
> Cruzbike 2 was built to use the learnings from
> Cruzbike 1, add larger wheels, and reduce weight. The
> closer hip line, in concert with very wide handlebars
> (673mm wide) has reduced the PSI, but it can still be
> felt very slightly.
>
> I am very very interested in what was learned with the
> TT Improved Jurgen Maier bike, as my next Cruzbike
> experiment is aimed at evaluating the "A-B" dimension
> effect on the perception of PSI.
>
> I would welcome your opinions and a discussion of the
> dynamics of these interesting bicycles.
>
> Be well,
>
> Doug Burton
>
> Richmond, Virginia, USA

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

This is the Python Mailinglist

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

Listmaster: Jürgen Mages jmages@xxxxxx

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

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

Other related posts: