RE: re: Re: rE: Bearing mounts/machine design (stackup3.gif)
Posted by
ballendo@y...
on 2000-10-25 14:17:41 UTC
Wally K wrote:
I'm not Jan, but YES, more distance between the bearings. As I
mentioned before the drawings are "squished" A LOT from side to side.
Grab the left and right and "stretch" and you'll be looking closer to
reality. And YES, still more distance between bearings (than
presently shown) will need to be added, even after this "mental
stretch". Something else the drawings leave out is the projection of
the lead screw beyond the housing so that it may be driven.(pulley,
gear, etc.) Tension / compression below.
List,
This is getting a whole lot more complicated than it is!
We have a shaft which rotates. We want to support it so it can rotate
as fast as WE NEED it to. We want to be sure the shaft does not move
endwise more than WE WANT it to. We want the shaft to be able to
support the load WE HAVE without whipping and/or bending(column
loading).
We know that for a given distance between supports, shaft diameter,
rotation speed, and support(mount) type(s) we will get more or less
of the things WE NEED above.
Some standard mount definitions:
RIGID- Two bearings spaced some distance apart and contained in the
same housing which (generally)do not allow ANY end play, and due to
the distance between the bearings, impart a "stiffening" support to
the shaft.
FIXED- Two bearings placed back-to-back in the same housing which
(generally) do not allow end play, and due to the "2 point" contact
on the shaft, impart a SMALLER "stiffening" support (compared to the
rigid type).
SIMPLE- One bearing placed in a housing which (generally)allows end
play(within the bearings' mfg. clearance), and due to the single
point of contact with the shaft, imparts almost NO stiffening support
to the shaft.
COMBINATION- Three or more bearings used in combinations of the above
types, within ONE housing, to achieve the desired results.(this will
apply mostly to spindle applications, and is not usually used for
leadscrews, as the other 3 types are sufficient).
NOTE: the TYPE of bearings used (plane, plain, tapered roller,
angular contact, dual-row, etc.)does not DETERMINE the mount "type".
In an earlier post, I described how to use cheap plain ball bearings
with shims to create a FIXED style mount. Even though commercial
FIXED mounts use face ground, angular contact, matched pair
bearings.
Now that we have the problem expressed, and the typical solutions
defined, we can begin to choose a solution to meet OUR NEEDS. We can
also look at others' solutions to the same problem for guidance.
For a given distance/dia./speed, the progression from least support
(to the shaft) to most support is:
00. SIMPLE-none
0. FIXED-none
1. SIMPLE-SIMPLE
2. FIXED-SIMPLE
3. RIGID-SIMPLE
4. FIXED-FIXED
5. RIGID-FIXED
6. RIGID-RIGID
4-7. COMBINATION-xxxxx (This will be for specialised applications,
usually machine spindles)
By far the most common in motion applications are #2 and #3.
Numbers 5 and 6 are used where a combination of either HEAVY loads,
LONG distance between support, or HIGH rotational speeds are
expected. Number 4 is used sometimes, where #5 is overkill(or won't
fit); or where it is easier to purchase or install than #3.
Re: tension or compression of the shaft:
It depends on what you're doing!
As things rotate, friction creates heat, which causes expansion of
the parts in the system. This must be taken into account by the
designer, and will depend on machine structure, materials, and the
speeds/usage expected.
For example, a low end 4 x 8 sign router of aluminum or welded
angle/channel will probably have enough "give" in the structure to
account for any expansion in the shaft and mounts. NOT TRUE for the
high speed/high load/ constant running of a production machining
center made of cast iron or heavy steel weldments.
So again, the designer has to take the expected usage (the WE NEEDS/
WANTS above) into account and make an appropriate selection.
In the real world of machines, this problem has been solved in many
ways that do not fit the commercial standards described above!
For instance, you can take a 1/2 inch leadscrew, turn one inch of one
end to the ID of a plain ball bearing, say .375(which will allow you
to put a pulley on to hold the bearing in place and drive the
shaft).Turn the other end similarly, and use a shaft collar to hold
the bearing. Now, if each bearing is placed in a housing, and the
housings(which are SIMPLE mounts)are bolted to a flat surface of
steel or aluminum,they can be "adjusted" (in relationship to each
other) until all end-play is gone. Very cheap solution, and as long
as the load/speed/distance is not too great, WILL work just fine.
Another common one on mill/drill or sherline retrofits is to use two
thrust bearings on either side of the casting and let the shaft
be "centered" by the plane(bronze or cast iron) bearing/hole already
there. The other end(being short) is not supported at all! And it
works. For some needs. Will it last as long(as a more fully supported
shaft)? Maybe not. But it can make accurate parts, and many do just
that, daily.
Ron, who started this thread, wants to know the "standard way"
of "doing" end mounts for leadscrews. We have covered that and the
drawing is at least "technically" correct(for a RIGID mount).The
proportions are still way off.
But there's something else I have been trying to "inject" into this
thread, as a result of watching beginners (and being one) in many
areas of my life:
1. I get interested.
2. I get excited.
3. I get started.
4. I get scared.
5. Out of my fear, I over-compensate my needs and abilities.
Jan (Troll) correctly points out that for a 12' 1.500 screw, You had
better do it right! (in a faraway place, to boot: this I gather from
his other posts).
But for a lot of what we do, it is not necessary to use the "top-of-
the-line" solution because we are afraid if we don't...
Hope this helps.
Ballendo
>Jan,I would like to know more about your ballscrew mounting method.Wally,
>Instead of getting very wordy, is it like method A, B, or, C. In
>the drawings stackup2, and 3 gif, but with more distance between the
>bearings. Also i am getting confused if a ballscrew is supposed to
>be held in compression or in tension.
>Wally K.
I'm not Jan, but YES, more distance between the bearings. As I
mentioned before the drawings are "squished" A LOT from side to side.
Grab the left and right and "stretch" and you'll be looking closer to
reality. And YES, still more distance between bearings (than
presently shown) will need to be added, even after this "mental
stretch". Something else the drawings leave out is the projection of
the lead screw beyond the housing so that it may be driven.(pulley,
gear, etc.) Tension / compression below.
List,
This is getting a whole lot more complicated than it is!
We have a shaft which rotates. We want to support it so it can rotate
as fast as WE NEED it to. We want to be sure the shaft does not move
endwise more than WE WANT it to. We want the shaft to be able to
support the load WE HAVE without whipping and/or bending(column
loading).
We know that for a given distance between supports, shaft diameter,
rotation speed, and support(mount) type(s) we will get more or less
of the things WE NEED above.
Some standard mount definitions:
RIGID- Two bearings spaced some distance apart and contained in the
same housing which (generally)do not allow ANY end play, and due to
the distance between the bearings, impart a "stiffening" support to
the shaft.
FIXED- Two bearings placed back-to-back in the same housing which
(generally) do not allow end play, and due to the "2 point" contact
on the shaft, impart a SMALLER "stiffening" support (compared to the
rigid type).
SIMPLE- One bearing placed in a housing which (generally)allows end
play(within the bearings' mfg. clearance), and due to the single
point of contact with the shaft, imparts almost NO stiffening support
to the shaft.
COMBINATION- Three or more bearings used in combinations of the above
types, within ONE housing, to achieve the desired results.(this will
apply mostly to spindle applications, and is not usually used for
leadscrews, as the other 3 types are sufficient).
NOTE: the TYPE of bearings used (plane, plain, tapered roller,
angular contact, dual-row, etc.)does not DETERMINE the mount "type".
In an earlier post, I described how to use cheap plain ball bearings
with shims to create a FIXED style mount. Even though commercial
FIXED mounts use face ground, angular contact, matched pair
bearings.
Now that we have the problem expressed, and the typical solutions
defined, we can begin to choose a solution to meet OUR NEEDS. We can
also look at others' solutions to the same problem for guidance.
For a given distance/dia./speed, the progression from least support
(to the shaft) to most support is:
00. SIMPLE-none
0. FIXED-none
1. SIMPLE-SIMPLE
2. FIXED-SIMPLE
3. RIGID-SIMPLE
4. FIXED-FIXED
5. RIGID-FIXED
6. RIGID-RIGID
4-7. COMBINATION-xxxxx (This will be for specialised applications,
usually machine spindles)
By far the most common in motion applications are #2 and #3.
Numbers 5 and 6 are used where a combination of either HEAVY loads,
LONG distance between support, or HIGH rotational speeds are
expected. Number 4 is used sometimes, where #5 is overkill(or won't
fit); or where it is easier to purchase or install than #3.
Re: tension or compression of the shaft:
It depends on what you're doing!
As things rotate, friction creates heat, which causes expansion of
the parts in the system. This must be taken into account by the
designer, and will depend on machine structure, materials, and the
speeds/usage expected.
For example, a low end 4 x 8 sign router of aluminum or welded
angle/channel will probably have enough "give" in the structure to
account for any expansion in the shaft and mounts. NOT TRUE for the
high speed/high load/ constant running of a production machining
center made of cast iron or heavy steel weldments.
So again, the designer has to take the expected usage (the WE NEEDS/
WANTS above) into account and make an appropriate selection.
In the real world of machines, this problem has been solved in many
ways that do not fit the commercial standards described above!
For instance, you can take a 1/2 inch leadscrew, turn one inch of one
end to the ID of a plain ball bearing, say .375(which will allow you
to put a pulley on to hold the bearing in place and drive the
shaft).Turn the other end similarly, and use a shaft collar to hold
the bearing. Now, if each bearing is placed in a housing, and the
housings(which are SIMPLE mounts)are bolted to a flat surface of
steel or aluminum,they can be "adjusted" (in relationship to each
other) until all end-play is gone. Very cheap solution, and as long
as the load/speed/distance is not too great, WILL work just fine.
Another common one on mill/drill or sherline retrofits is to use two
thrust bearings on either side of the casting and let the shaft
be "centered" by the plane(bronze or cast iron) bearing/hole already
there. The other end(being short) is not supported at all! And it
works. For some needs. Will it last as long(as a more fully supported
shaft)? Maybe not. But it can make accurate parts, and many do just
that, daily.
Ron, who started this thread, wants to know the "standard way"
of "doing" end mounts for leadscrews. We have covered that and the
drawing is at least "technically" correct(for a RIGID mount).The
proportions are still way off.
But there's something else I have been trying to "inject" into this
thread, as a result of watching beginners (and being one) in many
areas of my life:
1. I get interested.
2. I get excited.
3. I get started.
4. I get scared.
5. Out of my fear, I over-compensate my needs and abilities.
Jan (Troll) correctly points out that for a 12' 1.500 screw, You had
better do it right! (in a faraway place, to boot: this I gather from
his other posts).
But for a lot of what we do, it is not necessary to use the "top-of-
the-line" solution because we are afraid if we don't...
Hope this helps.
Ballendo