Optimum roller profile?

[angusinalberta]

From a link on another thread I see a set of concave rollers.
(http://www.bikecult.com/works/rollers.html)

Their apparent purpose is to keep the rider centered. Do riders find
that this self centering action occurs in practice? Or does the opposite
occur?

The reason for the question is the fact that things like bandsaws,
conveyor belts and belt sanders all use crowned rollers to keep the
belts centered.

Assuming the physics to be similar shouldn't bike rollers be crowned as
well? Just curious.

Angus

 

[Leo Lichtman]

"angusinalberta" wrote: (clip) Assuming the physics to be similar shouldn't
bike rollers be crowned as well? (clip)
^^^^^^^^^^^^^^^^^^
Sounds reasonable, but the physics is not the same. Crowned belt rollers
produce a centering action by affecting the belt tension. Belt tension does
not have a counterpart in the contact of wheel and roller.

 

[Doug B.]

On Mon, 20 Nov 2006 20:28:50 GMT, angusinalberta <[email protected]>
wrote:

>
> From a link on another thread I see a set of concave rollers.
>(http://www.bikecult.com/works/rollers.html)
>
>Their apparent purpose is to keep the rider centered. Do riders find
>that this self centering action occurs in practice? Or does the opposite
>occur?
>
>The reason for the question is the fact that things like bandsaws,
>conveyor belts and belt sanders all use crowned rollers to keep the
>belts centered.
>
>Assuming the physics to be similar shouldn't bike rollers be crowned as
>well? Just curious.
>
>Angus

Seems like it might take some of the personal responsibility out of
riding rollers. I don't think I want my wheel bouncing off the high
side when I start to get sloppy. I'd prefer to recover (or not) on my
own.
Doug B.

 

[Kinky Cowboy]

On Mon, 20 Nov 2006 20:28:50 GMT, angusinalberta <[email protected]>
wrote:

>
> From a link on another thread I see a set of concave rollers.
>(http://www.bikecult.com/works/rollers.html)
>
>Their apparent purpose is to keep the rider centered. Do riders find
>that this self centering action occurs in practice? Or does the opposite
>occur?
>
>The reason for the question is the fact that things like bandsaws,
>conveyor belts and belt sanders all use crowned rollers to keep the
>belts centered.
>
>Assuming the physics to be similar shouldn't bike rollers be crowned as
>well? Just curious.

I was thinking about this as I was on my cylindrical rollers earlier,
and thought maybe crowned rollers would induce a degree of camber
steer towards the centre; however, this would in effect be a
counter-steer input and a single track vehicle with no pilot should
then curve off towards the outside edge to compensate. For the same
reason, it's very easy to ride no-hands round the velodrome because
the banking does much of the steering for you.

Anyway, it's a solution to a problem which doesn't exist outside the
heads of the timid; staying on the rollers isn't difficult, and coming
off them isn't particularly dangerous.

Kinky Cowboy*

*Batteries not included
May contain traces of nuts
Your milage may vary

 

[Michael Warner]

On Mon, 20 Nov 2006 23:03:08 +0000, Kinky Cowboy wrote:

> Anyway, it's a solution to a problem which doesn't exist outside the
> heads of the timid; staying on the rollers isn't difficult, and coming
> off them isn't particularly dangerous.

It's a marketing problem if people are either too scared to try them,
or return them after one or two attempts, though.

--
Home page: http://members.westnet.com.au/mvw

 

[angusinalberta]

Leo Lichtman wrote:
> "angusinalberta" wrote: (clip) Assuming the physics to be similar shouldn't
> bike rollers be crowned as well? (clip)
> ^^^^^^^^^^^^^^^^^^
> Sounds reasonable, but the physics is not the same. Crowned belt rollers
> produce a centering action by affecting the belt tension. Belt tension does
> not have a counterpart in the contact of wheel and roller.

Perhaps the physics isn't identical, but there are similarities in
behavior. I offer 2 observations from my experience on the road with a
road bike using 50mm semi-slick tires.

The first example is the raised stripes that mark the shoulder of the
highway. I'm thinking of the stripes that are 2 or 3mm high and made of
something like colored concrete. I can't ride with my tires straddling
the edge of the stripe for more than a very short distance. The bike
always gets pulled up onto the stripe, often with a noticeable jerk and
causing the bike to swerve noticeably.

The second is the grooves worn in old asphalt. When I ride along in the
bottom of the groove the bike has a tendency to continue to climb up out
of the groove if I get too far from its centerline.

These examples led me to think that if a tire tends to climb upwards it
might also tend to do the same on a crowned roller. If it makes roller
riding a little easier then what's wrong with that?

On the other hand at least one manufacturer sells rollers with a concave
profile. Some members of this list must use them or at least tried them.
How do they feel? Same as standard rollers? Easier to control? Harder?
Do they keep you centered?

If someone can confirm that concave rollers have either a self-centering
action, or perhaps make no difference at all, then obviously my
conjecture is wrong.

The desirability or usefulness of self-centering as raised by some
others is a different question altogether.

Cheers,
Angus

 

[[email protected]]

Leo Lichtman writes:

>> Assuming the physics to be similar shouldn't bike rollers be
>> crowned as well?

> Sounds reasonable, but the physics is not the same. Crowned belt
> rollers produce a centering action by affecting the belt tension.
> Belt tension does not have a counterpart in the contact of wheel and
> roller.

Not only that but the belt does not steer, it moves to the side while
oriented essentially straight ahead. If you doubt it, consider that
long span belts stay centered as well as short ones even though in
long spans belts are often twisted 180° degrees to suppress whipping..

However, if crowned rollers worked as proposed, just think of how
bicycles ridden no-hands on crowned streets would move to the center
of the road... or for that matter, track riders could not ride
no-hands on banked tracks.

Get serious!

Jobst Brandt

 

[Leo Lichtman]

<[email protected]> wrote: (clip) in
long spans belts are often twisted 180° degrees to suppress whipping.(clip)
^^^^^^^^^^^^^^^^^
I thought it was to equalize wear on both sides of the belt.

 

[[email protected]]

Leo Lichtman writes:

>> long span belts are often twisted 180? degrees to suppress whipping.
> ^^^^^^^^^^^^^^^^

> I thought it was to equalize wear on both sides of the belt.

No. they run on the same side, the belt is flipped over 180 degrees as
though the pulley were detached and turned over. The other side of the
belt is usually not the wear surface and is intended to be on the
outside anyway.

Jobst Brandt

 

[G.T.]

"angusinalberta" <[email protected]> wrote in message
news:FYI8h.349833$R63.342233@pd7urf1no...
> Leo Lichtman wrote:
> > "angusinalberta" wrote: (clip) Assuming the physics to be similar
shouldn't
> > bike rollers be crowned as well? (clip)
> > ^^^^^^^^^^^^^^^^^^
> > Sounds reasonable, but the physics is not the same. Crowned belt
rollers
> > produce a centering action by affecting the belt tension. Belt tension
does
> > not have a counterpart in the contact of wheel and roller.
>
> Perhaps the physics isn't identical, but there are similarities in
> behavior. I offer 2 observations from my experience on the road with a
> road bike using 50mm semi-slick tires.
>
> The first example is the raised stripes that mark the shoulder of the
> highway. I'm thinking of the stripes that are 2 or 3mm high and made of
> something like colored concrete. I can't ride with my tires straddling
> the edge of the stripe for more than a very short distance. The bike
> always gets pulled up onto the stripe, often with a noticeable jerk and
> causing the bike to swerve noticeably.
>
> The second is the grooves worn in old asphalt. When I ride along in the
> bottom of the groove the bike has a tendency to continue to climb up out
> of the groove if I get too far from its centerline.
>

I've never experienced either of those.

Greg

 

[Paul Hobson]

angusinalberta wrote:
>
> From a link on another thread I see a set of concave rollers.
> (http://www.bikecult.com/works/rollers.html)

From the site:
"Show-Offs, Skilled Stunts, and Striptease?
Cycling Champion Lorette Burke undresses her skirt while pedaling a
chrome Paramount bicycle on wooden rollers In 1948."

hawt

\\paul

 

[Leo Lichtman]

<[email protected]> wrote: No. they run on the same side,
the belt is flipped over 180 degrees as though the pulley were detached and
turned over. The other side of the belt is usually not the wear surface and
is intended to be on the outside anyway.
^^^^^^^^^^^^^^^^
And all this time I have been thinking the belts were twisted into a Moebius
strip.

 

[[email protected]]

Leo Lichtman writes:

>> No. they run on the same side, the belt is flipped over 180 degrees
>> as though the pulley were detached and turned over. The other side
>> of the belt is usually not the wear surface and is intended to be
>> on the outside anyway.

> And all this time I have been thinking the belts were twisted into a
> Moebius strip.

That is possible with some belts that have a link pin connection but
that would not make the shape seen in the days of old in factories or
on tractor-to-thresher belts that slide past one another at mid span
in a long twist. The Möbius strip would have a twist in one run only
and serve no useful purpose.

The reason a twisted belt is more stable is that increased tension in
one run is matched by decreased tension in the other, and because they
more or less touch at midspan, they can't begin to whip or swing, the
effects being roughly equal and opposite.

Jobst Brandt