Hey John,

This is advice from strictly empirical knowledge.

Looks OK, probably either way, and there should be no noticeable
deflection of the "dowel" assuming the "support" is truly
"fixed".....but......

If the "load element" will be of an oscillating nature, rather than
rotational beyond 45 degrees, then use of Torrington style needle
bearings will EVENTUALLY lead to "tracking" on the shaft, flattening
of the needles and then seizing on the shaft, as the needles will
always be tracking and loading only in a repeating pattern. Even worse
if the force is not constant (eg..it "bangs")

I would consider using a plain or oil-lite bushing if that is the
case. At least when they wear, they don't "jam", and with the sizes
and weight shown will easily do the job.

In either instance, what is the purpose for "two" bearings. Even with
the needle type, there is no sense reducing the available contact area
is there? Why not a bearing or bushing that is of maximum allowable
size, projecting form the bearing housing at least a few thous towards
the fixed support, to provide a "rub point". And you also have not
indicated at all what will keep any of this from "walking off" the
shaft/dowel.

Take care.

Brian Lawson,
Bothwell, Ontario.



On 8 Nov 2005 16:52:19 -0800, "John2005"
wrote:

Hello everyone,

I would like to ask for some advice regarding a cantilever mounted
bearing housing.

Due to various design constraints, I have two choices as to how I
implement a cantilevered mounted bearing housing. I have uploaded two
simple dimensioned jpeg images for reference, at the following site.
You can Save or print the images as needed.

http://www.ice9.zoomshare.com

The housing has two drawn-cup needle roller bearings pressed into it,
and the housing oscillates rotationally on a stationary shaft (i.e., a
hardened
steel dowel pin). As shown in the drawings, I can use a 7/16" OD dowel
with the bearings spaced closer together, or I can use a 3/8" OD dowel
with the bearings spaced further apart.


1. I mainly need to know the correct way to model how the bearing
forces act on the
shaft, due to the housing load, so I can determine how far the end of
the shaft & housing will deflect, and whether the dowel can withstand
the stresses without taking a permanent set.

The maximum load on the housing will probably be about 130 pounds, but
I would like for the dowel to be able to withstand a 200 pound housing
load if possible, for a safety factor.

I can calculate the force at each bearing, but I am not sure exactly
how the force actually acts on the dowel. It seems to me that the load
on the bearings will almost be a torque
moment, where the the shaft is being bent between the contact points of

the two bearings, with bearing #2 pushing down, and bearing #1 pushing
almost upwardly.

Here is the formula I used to calculate the bearing loads, the letters
are shown with the corresponding dimensions, on the two drawings.

Where (LA) = the housing load

Load on bearing #1 = (LA) * B / A

Load on bearing #2 = (LA) * C / A

2. The bearing shaft is a hardened steel pull dowel pin, made from
C1541, 4037, or 4140 steel (thats all the info I can get from Mcmaster
Carr). The single shear strength for the pin is 130,000 PSI. The pins
have a core hardness of Rockwell C47-58, and a surface hardness of C60
(they meet ASME B18.8.2 standards).

Since the dowel is hardened I am not sure what the maximum yield
strength is, I know tensile goes up with hardening but I don't have any

information on the yield strength of the hardened dowel pin. I am
hoping someone can shed some light on this issue.

I have a beam design program I can use to help determine stress and
deflection of the dowel, but I am not sure If I should model a torque
moment with the rotation axis between the two bearings, or perhaps a
combination of a torque moment and vertical forces.

Using a housing load of 130 pounds, the formula given above, and a
downward force on both bearing #1 & #2, it seems the 3/8" OD dowel has
slightly less stress than the 7/16" dowel, but it deflects about .001"
further.

The end of the dowel is 1.26" from the cantilever support. The loaded
end of the housing is 1.48" from the cantilever support, and there is a

.031" space between the housing and the cantilever support.

I have a 3/8" OD, lever "connection socket" that screws into the
bearing housing, perpendicular to the housing, right next to the
cantilever steel support. The advantage of housing #1 is that I can
"step" the bearing bore in the housing so that I have a little more
housing material to thread the connection socket into, since bearing #
1 is moved out from under the connection socket, and closer to the load

end of the housing.

The advantage of housing #2 is that the bearings are spaced further
apart, and this helps reduce housing deflection due to bearing
misalignment. The trade off is that I have a little less housing
material to thread the connection socket into, since the socket threads

in, right on top of bearing #1.

The housing is oscillated by hand, and never gets hot. The radial
clearance between the bearing ID and the shaft OD will be .0002" Min.
to .002" maximum.

If there is no way to model this to get a close estimate, I would
appreciate your gut feelings as to which method is best, as far keeping

the deflections at the load end of the housing to a minimum, and
avoiding overloading the dowel so that it does not incur a permanent
set due to the housing load. Obviously, if the dowels had the same OD
in each case, then housing #2 would be best, since the bearings are
spaced further apart.

I think that either arrangement could take 130 pounds, but I am not
sure about 200 pounds. I would like to be able to estimate the maximum
housing load the dowel pin could withstand without taking a permanent
set, and make a close estimate on the deflection of the dowel and
housing.

I want to keep the deviations of the housing end from it's longitudinal
axis to a minimum, whether the deviation comes from shaft deflection or
bearing misalignment. However, I need to be sure that the dowel is not
going to be stressed past it's yield point, so that it springs back to
it's original position when the housing load is removed, and does not
take a permanent set.

Thanks for your help.
John