Re: New page "Gears, pulleys and bearings"

[Miles]

""Prototype gears will initially be aluminium for the ring and sun
gears, and a self lubricating polymer will be used in the four planet
gears to reduce noise since the other low-noise alternative, helical
spur gears, are quite difficult to make. If it works well and will
handle the load, the weight will also be reduced compared with more
standard steel cogs. Alternatively, if the load is to excessive, it
might be possible to add a fifth planet wheel to reduce the contact
forces by 20%""

Per,

I like the idea of composite planet gears. Perhaps, for the
production version, the sun and ring gears could be hard-anodised and
sealed with PTFE or Keronite surface treatment, if the cost isn't
prohibitive?

Miles

[Per Hassel Sørensen]

It is a good idea to try to use coating and composite but would'n it
be more costly than steel in volumes?
Cheers,
Per

2008/10/15, Miles <mi...@hellonline.eclipse.co.uk>:

[Miles]

Not sure how the costings would compare. I'll do some research.

Would you have to upgrade the bearings for the axial forces on helical
gears?

Miles

[Per Hassel Sørensen]

Yes, due to bidirectional torque (motor and regen) some upgrading of the bearings or additional axial bearings would be required. The current design uses springs for stabilizing the rather low axial forces with the horizontal motors. With helical these weak springs are not sufficient. 

 

If the sound level and wear resistance is acceptable with treated aluminium and fibre reinforced polymer, straight spurs it is (like an old race car with a Moss gear box but not sounding like one! :-))

[PerBear]

They probably would need to be reinforced due to bidirectional torque
(running & braking).
Let's see if aluminium and polymer works fine!

Per

[Miles]

"" Syncro pulley sizing for the belt drive is dependent on wheel size.
The calculations done so far has used 26" wheel as reference and a
1:13 final ratio. If final ratio is kept within 1:9 a single stage
belt drive could work. but it would be benficial to have two belts
with an intermediate axle, this has the following advantages:

1. Easy to adjust belt tension by adjusting position of
intermediate axle.
2. Adjusting final ratio becomes possible by replacing pulleys on
intermediate axle only.
3. Same motor pulley and rear wheel pulley for different bike
sizes.
4. Easy to service rear wheel with quick release on intermediate
axle.
5. No need to losen motor fixture including wiring etc when
adjusting belt or wheel. ""

Per,

Do you envisage the intermediate axle being structurally attached to
the motor unit? I'm working on a design like this for my current
project. To minimise weight, it will be for specific pulley sizes -
the bearings will be eccentrically mounted, to give 2mm of adjustment
for setting belt tension.

Miles

[Per Hassel Sørensen]

Miles,

 

To make the Nohassel easy to adapt to many different cycle frames, it should be possible to disconnect motor and intermediate axle if necessary. That said, I assume that the best design would be to have the motor and axle well connected, with a good mechanism for adjustment of tension on both belts, maybe adapting your design. Do you have a drawing or picture to share?

Per Hassel

[Miles]

Per,

I'm working on a variant which will allow a greater range of
adjustment. I'll send you the model to look at, when I've roughed it
out.

On my Moulton ebike, I used the limit of the rear dropout slot to
reset the belt tension.

How about an integral adjustment for the tension on the 1st reduction
stage and adjustment in the mounting of the motor for the tension on
the 2nd stage (between inter-axle and wheel)?

Would one need to change the sizes of both pulleys on the inter-
axle......?

Would the small pulley on the inter-axle need to be within the width
of the motor casing?

Gates and Conti seem to make the best synchro belts. Gates Powergrip
GT3 5M for both stages (or 8M, for a narrow belt on the second stage)?
Or?

Miles

[Per Hassel Sørensen]

Miles,

 

I would like to see your solution before designing the pulley system for Nohassel. I also need to finish some more cycle parts (still miss some parts and a representative frame).

 

GT3 seems to have a price premium, but we should be able to get some budget pricing on the different alternatives to verify this. I think we should let the cost of belt and pulleys decide which belt to choose, as long as we can make it work within our constrains.

 

The group has received many new members, maybe someone with good knowledge on belt and pulleys could shed some light here?

 

Per Hassel

[Miles]

Per,

About 12 euros for the small belt in GT3? Better to have the higher
capacity belt and keep the tension down : )

For 8M pitch, the GT3 pulley profile is the same as HTD, for 5M it's
pretty close...

[PerBear]

Too much creates excessive wear, but too little tension can, in
addition to cog slip, cause belt vibration at high speed. The current
design has a pulley speed of 8000 RPM max (at 81 km/h!). Minimum
tension should be fairly easy to find using design manuals etc. from
the belt manufacturer.

Per

[Miles]

Some references for gear materials:

http://www.dsm.com/en_US/html/dep/products.htm
http://www.dsm.com/en_US/html/dep/ind_apps_gears.htm
http://www.dsm.com/en_US/downloads/dep/ ... s_2005.pdf
http://www.omnexus.com/tc/pa4-6/index.aspx?id=grade

Miles

[PerBear]

I have now redesigned the epicyclic gearing to use polymer for planet
gears and aluminium for ring gear. I also modified to planet carrier
to have five planets instead of four to reduce stress on polymer
cogs.

This modification reduced the weight from 2.2kg to less than 2kg, a
nice weight saving and obviously noise reduction as well. I guess the
pricing will be better also, at least for the prototype where material
costs are less important than manufacturing costs.

Hopefully the change to polymer planet gears will not bring the
lifetime down to less than 3000 hours.

I am working on 2D drawings for the prototype and will post them when
they become ready, together with 3D models of the parts.

Per

[PerBear]

I have now finished modelling a nice rear hub in 3D and put it into
the files folder. This fully detailed 3D model of the hub has been
made by "reverse engineering" two used hubs with calipher and various
gauges. Please be aware that 3D model files are in Solid Edge format,
at least for now.

Since the cycle need rim brakes on the rear wheel it is important to
centralize the rim with respect to the frame. Since the 3D hub model
is assymetric by design, giving more space for the sprockets, the
spoke flanges are shifted 15mm to the left, looking strange when
located in the frame. This is actually normal and is taken care of by
shifting the rim to the right with respect to the hub flanges when
building the wheel. This shift require higher spoke tension on the
right (sprocket) side. This is the reason many wheels have stronger
spokes on the right hand side.

Perbear