From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: shaft drive
Date: 5 Oct 1998 20:57:17 GMT

Bob Michael writes:

> A friend and I were pondering about shaft drives the other day and were
> wondering if anyone had ever tried one on a bike. With the new composites
> and alloys for the shaft and gear housings, new internal hubs and the
> like, and the motorcycle-like rear suspensions, isn't it inevitable?

It's all been done and a long time ago.  I think the killer is the
thought of a flat tire but that is only one of the problems.  Torque
on a bicycle is higher than most automobile engines while the
rotational speed is about 100 times lower.  This becomes apparent when
you stand up on such a drive shaft that turns into a full suspension
pedal with plenty of spring.  The bevel gears that change the axis of
rotation at the BB and rear axle ought to be larger than those of a
car, that weigh several pounds.

With a hub gear, which is something I haven't seen on the shaft drive
bicycles, the bevel gear at the BB should be about as big as a
moderate chainwheel to get the shaft spinning and out of the high
torque range.  Of course this would be way to heavy so it isn't done.
The gear drive requires precision for the gears to properly mesh and
not eat themselves up and this requires precision bearings...
completely sealed from the weather.  After you solve these problems,
the whole idea gradually loses its appeal.  It was tried often enough.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: drela@mit.edu (Mark Drela)
Newsgroups: rec.bicycles.tech
Subject: Re: shaft drive
Date: 6 Oct 1998 00:44:05 GMT

In article <6vbbrd$mfm$3@hplms2.hpl.hp.com>, jbrandt@hpl.hp.com (Jobst
Brandt) writes:

> Bob Michael writes:
>
> > A friend and I were pondering about shaft drives the other day and were
> > wondering if anyone had ever tried one on a bike. With the new composites
> > and alloys for the shaft and gear housings, new internal hubs and the
> > like, and the motorcycle-like rear suspensions, isn't it inevitable?
>
> It's all been done and a long time ago.  I think the killer is the
> thought of a flat tire but that is only one of the problems.  Torque
> on a bicycle is higher than most automobile engines while the
> rotational speed is about 100 times lower.  This becomes apparent when
> you stand up on such a drive shaft that turns into a full suspension
> pedal with plenty of spring.  The bevel gears that change the axis of
> rotation at the BB and rear axle ought to be larger than those of a
> car, that weigh several pounds.


Good guess.

Here's a data point on this:

The Deadalus human-powered aircraft used two 90-deg gearboxes.
Their total weight was several pounds.  They were finicky at
first, but quite reliable once the bugs were worked out.
They were far less robust than a 1/2" chain system, but they
didn't have to be, since standing on the pedals was impossible
in the recumbent position.  A system comparable in strength
to a chain would have weighed at least 6 lb, I'd guess.

The gearbox at the pedals had a 2:3 speedup, and the other
one was 1:1.  The biggest factor which drives up weight is
the gear ratio deviating from 1:1, since one of the gears
must then be larger, which gives a longer shaft, larger housing,
etc. So the lightest solution splits the overall ratio more or
less evenly between the two gearboxes.  The weight of the
shaft is secondary.  The Daedalus had a 4-foot long, 1.25" diameter
carbon fiber shaft between the gearboxes (CF is very easy for
this application).  It took 2/3 of pedal torque and weighed
5.5 ounces.

Here are the specs in case anyone is curious:

2:3 box:  10-pitch spiral-bevel gears, 2.4" dia main, 1.6" pinion.
1:1 box:  12-pitch spiral-bevel gears, 1.6" dia both



                   Mark Drela
             _______________________________
    o/LO  .'
     O  .'  Gravity-Powered Technologies Lab
      .'  MIT Aero-Astro Department  37-475
     '

From: drela@mit.edu (Mark Drela)
Newsgroups: rec.bicycles.tech
Subject: Re: shaft drive
Date: 7 Oct 1998 22:52:17 GMT

In article <F0FCBM.Cu4@world.std.com>, dp@world.std.com (Jeff DelPapa) writes:
> In article <6vbp4l$is@senator-bedfellow.MIT.EDU>,
> Mark Drela <drela@mit.edu> wrote:
> >
> >The Deadalus human-powered aircraft used two 90-deg gearboxes.
> >Their total weight was several pounds.  They were finicky at
> >first, but quite reliable once the bugs were worked out.
>
> Actually, I think what saved them is that the propeller kept the peak
> loads down.  You couldn't get the high forces that a toggled knee,
> that had a seat back to press against can develop when the driven
> object has significant (frictional) coupling to the outside world.

Having a prop does not significantly reduce the worst-case
pedal loads you can generate.  At low speed and high prop pitch
you couldn't push it much above 40 rpm, which is practically static.


> I know some of the HP aircraft (but I don't know which ones) used the
> berg 2 cables with polycarbonate bars chain substitute, and had a lot
> of success with it.  Its a lot lighter than chain, but can't be
> derailed.  It is possible to form a loop out of straight section,
> unlike belts.  It worked in the propellor drive application.  (lasted
> at least as long as the airfame was designed to)

The Berg chain works OK, but is prone to derailments
when the loop is twisted through the necessary 90 degrees
between pedals and prop shaft.  Derailment is a disaster
in an HPA over water.  This is the main reason why we went
to gearboxes on Daedalus --- for the reliability, not weight
or efficiency.

The Berg chain was used on all other MIT HPAs.  We threw
chains all the time.  It's a mere nuisance over land.



> The fragments I saw at Garber (the smisonian restoration facility)
> showed some sort of chain like drive used from the cranks.  The
> gearboxes must have been further down the line.  The backup plane is
> nearby, I will have to go look again.

You must have been looking at some other plane.
Daedalus never had anything resembling a chain.



> Its also interesting that Decavitator (an airscrew driven hydrofoil)
> used chain drive, and the designers were involved with Deadalus...
> (and they used comparable weight concious construction, as the
> dockside main spar collapse showed.)
> What were the gears made from?  I assume the case was carbon fiber.

The Decavitator used 1/4" pitch chain, on CNC-milled Al sprockets.


The 1/4" pitch chain is actually quite robust.
It's very light, and might help a smidgeon on
a hill-climb record assault.


                   Mark Drela
             _______________________________
    o/LO  .'
     O  .'  Gravity-Powered Technologies Lab
      .'  MIT Aero-Astro Department  37-475
     '