Swing arm bearing

Green Gear - Bike Friday has lots of experience

"JT" <[email hidden]> wrote in message
news:[email hidden]...
| I want to put a swing arm onto a bike I'm building, but what bushing,
what's
| the best aproach?
|
| Any know-how out there would be much appreciated.
|
| Regards JT
|
|

Depending on the weight of the rider(s) and how hard you intend to use it,
my friends and I use 1/2"(or 12mm) or 3/8" precision bearings and a grade 8
pivot bolt. Make/get a piece of steel tube and press the bearings into the
end with a compression tube in the middle. The 1/2" & 12mm would be for the
heavier/harder rider while the 3/8" would be for lighter and more moderate
riders (say 210lbs and under ?).

"JT" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

I want to put a swing arm onto a bike I'm building, but what bushing,

what's

Quoted message said:

the best aproach?

Any know-how out there would be much appreciated.

Regards JT

"Sticker Jim" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Depending on the weight of the rider(s) and how hard you intend to use it,
my friends and I use 1/2"(or 12mm) or 3/8" precision bearings and a grade

8

Quoted message said:

pivot bolt.

Here's one vote that says precision bearings and Grade 8 are vast
overkill/expense.

First, figure that the suspension spring/damper element itself has
stiction/viscous friction to begin with (some commercial suspension elements
have a LOT of striction), so trying to minimize it at the swing arm pivot
joint is pointless.

Second, figure the loads. Rider and vehicle weight are largely supported by
the spring/damper, and the direct radial loads on the bearings derive
primarily from traction and braking, which is to say they are not that
great. The significant loads on a swing arm derive from the side loads
from wind and cornering. Though the external forces may be modest, they are
applied over lever arms the size of ,say, the wheel radius or the swing arm
length, and the resulting moments must be taken out across the side-to-side
span of the pivot joint. Notice that these loads in a trike swing axle
pivot might be 5 or 10 times those generated in a two-wheeler. These
applied moments are related to the other critical design issue, wheel
alignment, which I'll discuss below.

Third, consider the application. Ball bearings are designed and built to
spin, and they're not particularly happy making small reciprocating motions
as in a suspension joint. Now, in this case the loads are low, and I'm not
claiming that they'll fail, but they're overkill. Look around at production
automotive applications - a world of 100,000 mile warrantys and micro-penny
counting. I can't think of a place where ball or roller bearings are used
like that. Where low friction reciprocating joints are required, you
sometimes see needle bearings, where the displacement allows for a turn or
two of the rolling element.

Fourth, alignment. Fore and aft alignment is not important, side-to-side
medium important. What is critical are the wheel alignments: 1.) toe
(angular displacement of the rear wheel about a vertical axis), and 2.)
camber (angular displacement of the rear wheel about the longitudinal axis
of the vehicle.), both relative to the position of the front wheel. These
alignments are established by the relative positions of the two swing arm
pivot bearings. Initial misalignment, structural flexibility, excessive
clearances, or wear here will kill the design.

So, considering all of this, what does the bearing look like? Considering
the small motions, it'll have a bushing (well-greased bronze or Delrin if
the loads are right) running in a thin-wall housing on a thin wall largish
diameter shaft. The downside here is the weld or braze distortion you get
when you join the housing to the frame. If you have a shop, it's pretty
straightforward to bore the housing after welding, hard without the shop.

But the really nifty solution would be to rip the design off race cars. Use
a pair of spherical joints threaded into the ends of the swing arms.
Compute the load ranges and pick a ball joint technology (bronze, Delrin,
etc.) that fits. Make the span of the joint wide enough to get the loads
and the alignment sensitivity down.

If you're building with angle-iron-hack-saw-and-stick-welder, you can use
four or five spherical joints and forget about alignment altogether until
you get to the mechanical assembly.

Use bolts soft enough to drill out the centers. When you size the ball
joint for load, you won't need anything like the full bolt area for shear.
Before hitech materials arrived, it was common for race car designers to
drill out bolt centers to save weight. It's a little mysterious why this
technique hasn't percolated down into recumbents.

Too, it baffles me why people insist that Grade 8 bolts are the "best" for
anything. For moving vehicles "best" is usually AN.

Just free advice,
Fred Klingener

"Fred Klingener" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

"Sticker Jim" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Depending on the weight of the rider(s) and how hard you intend to use

it,

Quoted message said:

Quoted message said:

my friends and I use 1/2"(or 12mm) or 3/8" precision bearings and a

grade

Quoted message said:

8

Quoted message said:

pivot bolt.

Here's one vote that says precision bearings and Grade 8 are vast
overkill/expense.

First, figure that the suspension spring/damper element itself has
stiction/viscous friction to begin with (some commercial suspension

elements

Quoted message said:

have a LOT of striction), so trying to minimize it at the swing arm pivot
joint is pointless.

Second, figure the loads. Rider and vehicle weight are largely supported

by

Quoted message said:

the spring/damper, and the direct radial loads on the bearings derive
primarily from traction and braking, which is to say they are not that
great. The significant loads on a swing arm derive from the side loads
from wind and cornering. Though the external forces may be modest, they

are

Quoted message said:

applied over lever arms the size of ,say, the wheel radius or the swing

arm

Quoted message said:

length, and the resulting moments must be taken out across the

side-to-side

Quoted message said:

span of the pivot joint. Notice that these loads in a trike swing axle
pivot might be 5 or 10 times those generated in a two-wheeler. These
applied moments are related to the other critical design issue, wheel
alignment, which I'll discuss below.

Third, consider the application. Ball bearings are designed and built to
spin, and they're not particularly happy making small reciprocating

motions

Quoted message said:

as in a suspension joint. Now, in this case the loads are low, and I'm

not

Quoted message said:

claiming that they'll fail, but they're overkill. Look around at

production

Quoted message said:

automotive applications - a world of 100,000 mile warrantys and

micro-penny

Quoted message said:

counting. I can't think of a place where ball or roller bearings are used
like that. Where low friction reciprocating joints are required, you
sometimes see needle bearings, where the displacement allows for a turn or
two of the rolling element.

Fourth, alignment. Fore and aft alignment is not important, side-to-side
medium important. What is critical are the wheel alignments: 1.) toe
(angular displacement of the rear wheel about a vertical axis), and 2.)
camber (angular displacement of the rear wheel about the longitudinal axis
of the vehicle.), both relative to the position of the front wheel. These
alignments are established by the relative positions of the two swing arm
pivot bearings. Initial misalignment, structural flexibility, excessive
clearances, or wear here will kill the design.

So, considering all of this, what does the bearing look like? Considering
the small motions, it'll have a bushing (well-greased bronze or Delrin if
the loads are right) running in a thin-wall housing on a thin wall largish
diameter shaft. The downside here is the weld or braze distortion you get
when you join the housing to the frame. If you have a shop, it's pretty
straightforward to bore the housing after welding, hard without the shop.

But the really nifty solution would be to rip the design off race cars.

Use

Quoted message said:

a pair of spherical joints threaded into the ends of the swing arms.
Compute the load ranges and pick a ball joint technology (bronze, Delrin,
etc.) that fits. Make the span of the joint wide enough to get the loads
and the alignment sensitivity down.

If you're building with angle-iron-hack-saw-and-stick-welder, you can use
four or five spherical joints and forget about alignment altogether until
you get to the mechanical assembly.

Use bolts soft enough to drill out the centers. When you size the ball
joint for load, you won't need anything like the full bolt area for shear.
Before hitech materials arrived, it was common for race car designers to
drill out bolt centers to save weight. It's a little mysterious why this
technique hasn't percolated down into recumbents.

Too, it baffles me why people insist that Grade 8 bolts are the "best" for
anything. For moving vehicles "best" is usually AN.

Just free advice,
Fred Klingener

Good points, but . . .

First, I'm not an engineer. I try to take into consideration the direction
of the loads but I have no idea what the loads are, nor do I know what the
bearings, bolts, steel, etc loads are. What I do know about is rider abuse.
I usually try to idiot-proof anything I build but I am sometimes surprised.

I understand what role the bearings play and what they are designed for but
for lack of anything better/more applicable at the same price, I use the
bearings. They have precision dimensions, fit easily in standard tubing (no
machining), they're hardened and easy to come by. Sintered bushings are
easy to get and cheap but also much softer. I've seen guys that used
bushings for similar pivot points and they get elongated in the small
section where they are constantly under load and that doesn't include the
wrenching/yawing motion that the lateral forces from a trike would add.

I get precision bearings at a pretty good price (I think) of about $2.50 Cdn
ea - that's $5 for two 12mm or 1/2" bearings - 3/8" are less. A grade 8
bolt is less than $2 and a Gr 5 is cheaper still. They have more than
enough strength BUT I'm interested in hardness, not shear or tensile str.
I've seen and had stainless and Gr 2 bolts bend and shear (even Gr 5 bolts),
especially on chain pulleys when someone gets on and tries to power away in
high gear. So, for ~$12 I have a pretty sweet pivot point not counting
maybe another $1 in washers and a nyloc nut. No slop, and no play in this
assembly either.

Now, let's switch to your idea of the spherical type rod end. The cheapest
ones I can get are $13 Cdn and I don't really like them because there is
already play in them. That's for 5/16" too. If I get 3/8" and up it's over
$15 each and that is for what I perceive to be low end rod ends. They
aren't stainless and don't have any kind of protective boot. I have no idea
how strong they are either (even the parts store doesn't know/can't be
bothered to find out). So, I'm at $30 already and haven't bought any nuts,
bolts or washers AND I don't trust the joint already. I have no idea what
good joints would cost and the time it would take to track them down would
be more time than it would take to build a pivot the way I do now. Also,
all the current parts I use were found at the same place I buy the rest of
my hardware so there was no searching required for them.

In the end, I'll buy a Gr 8 bolt and spend an extra .50 cents to a dollar on
it because it's peace of mind. The precision bearings are the same, AND
cheap and easy to replace (but I haven't heard of any of my friends needing
to replace theirs yet and that's with almost 2 years running them at ~200 to
400 km a week. Good enough for me.

Reliable, inexpensive, TIGHT pivots are what is stopping me from building a
full suspension trike. I've seen some commercial offerings of full
suspension and even brand new, there's play in them, too much for me to be
happy with.

Lastly, while I am not much of a sheep, I am a little pragmatic when it
comes to pivot point design. Of all the bikes and trikes I've seen with
rear suspension, none of them use spherical ball. I'm sure there's a reason
why, but I'm not wasting my time finding out because I'll be busy building
pivots the way I do. As I become 100% confident in the strength,
reliability and durability of the pivots, I'll start downsizing components
(which the 3/8" bolt is an example) and again, I'll watch it 'til I'm happy
with it. It would be pretty easy to tear apart a commercial offering of a
pivot point and measure everything, but I don't have one and don't know of
anyone that would let me temporarily disassemble theirs. I'm not too
worried though as from the looks of them, I decided to start with a bolt and
2 bearings, and go from there. In the end, the experimentation and
overbuilding only amounts to a few dollars per trike. I'm fine with that.
I'd rather be overbuilt/overkilled than have it fall apart.

YMMV