PDA

View Full Version : Driving floating end (ballscrew)

blackburn mark
08-09-2010, 11:03 PM
i might have missed something obvious here...... it there any reason not to drive the floating end of a ballscrew (600mm) removing the need for the floating bearing

the only things i can think of is backlash from any torsional twist ill get over the 600mm and a slightly shorter life on my stepper nose bearing

irving2008
08-09-2010, 11:16 PM
thermal expansion... normally you drive the fixed end and the floating end can move to accomodate...

blackburn mark
08-09-2010, 11:25 PM
thermal expansion... normally you drive the fixed end and the floating end can move to accomodate...

cheers irving... i need to do some research on how much exspansion i might get... my couplers are the slitted aluminium type and will allow for some exspansion

iv designed a slight error into my build and driving the floating end might let me get away with it without to much of a re-work :redface:

irving2008
08-09-2010, 11:50 PM
coefficient of linear expansion of steel is about 13 - 17 x 10^-6 m per m per degC. i.e. a 600mm length will expand max 0.01mm per degC, so allowing for normal working environment say +20degC +/- 10degC you can expect a variation of 0.2mm.

Force = delta_L * E * A/L where E=Youngs modulus (207x 10^9 N/m^2), A = area in m (2e-4 for a 16mm screw), L = length, delta_L = change in length

=0.2 * 207e9 * 2e-4 / 600 = 13800N... not a force your stepper would be happy with!

blackburn mark
09-09-2010, 12:05 AM
jesus irving!! dose your mind ever sleep?

i was starting to sweat over the figures... id have still been at it till the sun came up

0.2mm... thats a bit depressing... i was hoping the aluminium frame would off set some of it but at almost twice the exspansion of steel (with a bit of carbon in it) im still looking at about 0.2mm i think

back to the drawing board

cheers for doing the maths irving xxxxxx

10-09-2010, 11:51 AM
coefficient of linear expansion of steel is about 13 - 17 x 10^-6 m per m per degC. i.e. a 600mm length will expand max 0.01mm per degC, so allowing for normal working environment say +20degC +/- 10degC you can expect a variation of 0.2mm.

Force = delta_L * E * A/L where E=Youngs modulus (207x 10^9 N/m^2), A = area in m (2e-4 for a 16mm screw), L = length, delta_L = change in length

=0.2 * 207e9 * 2e-4 / 600 = 13800N... not a force your stepper would be happy with!

Bloody ell irving, it makes me chuckle every time i read a post off you like this, it remind me of James May off top gear when i kicks in with his maths :lol: not thats theres anything wrong with that... i wish i could understand all that lot.:tup:

irving2008
10-09-2010, 01:05 PM
:lol: well I've been compared with worse ;) though James May's 'maths' on Top Gear often leaves a lot to be desired!

My take on this is that there is nothing wrong with getting a bit of rigour into the thinking.... and this is a good example to show how sometimes the little oversight can have big consequences...

more than happy to explain it...

Gary
10-09-2010, 04:55 PM
I have had customers wanting to drive from the floating end, and while it is not ideal, it will most likely be ok.
All stepper motors have a wavy washer or two behind the back bearing so this can take up at least 0.5mm.
The coupling will also help.
On another note the thermal expansion would not be that much unless you are really working the ballscrew and it starts to heat up, a Z axis is not that dynamic with most moves happening over a small area, and not up and down over the whole stroke.

blackburn mark
10-09-2010, 05:43 PM
cheers gary,, im going to push on with the revised plan... it solves enough of a problem to take the risk

as far as the temp goes im sure the screws will see a 20 degree shift in a shed in the uk

soooooo...... with the wavy washers plus an easy .5mm in the coupling im kind of confident:eek:
2988

irving2008
10-09-2010, 06:12 PM
Agreed the z-axis wouldnt be.. but this is a 600mm screw... I didnt get it was the Z-axis anywhere...

Though pragmatically I agree with Gary, the coupler will ease some of it... though a typical Nema23 stepper is spec'd at a maximum axial force of 15N so although the coupler will compress out to take up some motion it will still transmit a significant force to the stepper shaft. Those wavy washers might not stay wavy for long! I'd consider putting the screw under slight tension at a lower temp to give some flexibility when/if the temp rises....

blackburn mark
10-09-2010, 07:03 PM
ok iv just run some tests on the coupler... with 10N it compresses by .43mm im assuming simmilar under tension
the full travel of the coupler is just over 1mm fully compressed
im thinking unless i take a blow torch to the ballscrew (y axis) i should be ok

if not it should run long enough to machine myself out of trouble

i thought id designed this machine to perfection :) however... i snapped up some cheap 20mm profile rails for the Z axis instead of the 15mm in the design and its had more of a knock on effect than i exspected :(

iv got this funny feeling ill be seeing a few more of those before im done :)

routercnc
11-09-2010, 06:38 AM
I guess you could always loosen and re-tighten the grub screws between the coupler and the ballscrew twice a year to take out any stress. Or you could mount the stepper on a sliding arrangement so that rotations were prevented but it could slide back and forth 0.5mm or so.

Jonathan
11-09-2010, 11:39 AM
Or you could mount the stepper on a sliding arrangement so that rotations were prevented but it could slide back and forth 0.5mm or so.

Maybe put a rubber O-ring on each of the 4 bolts that mount the stepper motor. I did that on my rotary table to compensate for a little radial misalignment and it works fine.
It could also damp vibrations...

John S
11-09-2010, 11:46 AM
It'll work - just do it.

If all these MDF, draw slides, all thread screw jobbies work what's the problem.

Never let the decimal point stand in your way.

i2i
11-09-2010, 11:47 AM
you can get rubber mounts for steppers, that simply mount between the stepper and the frame, giving a slightly flexible yet resilient mount.

John S
11-09-2010, 11:48 AM
Maybe put a rubber O-ring on each of the 4 bolts that mount the stepper motor. I did that on my rotary table to compensate for a little radial misalignment and it works fine.
It could also damp vibrations...

Oldham couplings were designed to take out radial misalignment, Those lovejoy couplings are a total waste of time and space for CNC machines, they are better off driving sewerage pumps.

Jonathan
11-09-2010, 11:55 AM
Oldham couplings were designed to take out radial misalignment...

Yes I know...didn't have time to make an Oldham coupling at the time for the rotary table and haven't bothered since because it works fine. It's only a very very slight misalignment.