Rotating Ballnut - design ideas

[teamloks]

hello im thinking to make a new cnc router for size 5x10 so my question is im going to use rotary ball nut,do i need 25mm or 32mm ballscrew for 10 feet axis and do i need high pitch screw such as 25 or 32 since im thinking rotary ball nut doesnt have critical speed maybe i can use 10mm pitch and just spin it 3000-5000 rpm with 1:1 ratio belt,
and im going to pair it with 750-1kw 3000-5000 rpm servo for it i already.

i already got quote for my supplier

rotary ballscrew 3232 10 feet=350 usd
rotary ballscrew 2525 10 feet=300 usd

my target rapid is around 1500-2000ipm

thanks

[JAZZCNC]

hello im thinking to make a new cnc router for size 5x10 so my question is im going to use rotary ball nut,do i need 25mm or 32mm ballscrew for 10 feet axis and do i need high pitch screw such as 25 or 32 since im thinking rotary ball nut doesnt have critical speed maybe i can use 10mm pitch and just spin it 3000-5000 rpm with 1:1 ratio belt,
and im going to pair it with 750-1kw 3000-5000 rpm servo for it i already.

i already got quote for my supplier

rotary ballscrew 3232 10 feet=350 usd
rotary ballscrew 2525 10 feet=300 usd

my target rapid is around 1500-2000ipm

thanks

Hi,

I'm just finishing a 10x5 machine with rotating ball nut so it's good timing to ask this question.

The first thing to mention is how much the screws sag under there own weight at 10ft long. With NO tension on the screws, it's massive but even with tension on the screws which you must have, there is still more Sag than you would probably expect.
So when the machine is at Rapid feeds and when axes are near the ends it leaves a long length of screw unsupported, so the screw can still vibrate.

On the machine I've built it uses a 25mm diameter with a 25mm pitch with a 1.5:1 ratio. I'm using 220V closed loop steppers spinning which spin between 1500-2000rpm. The machine is tuned at 30mtr/min(1200in/min) and will reach 40mtr/min((1575in/min) however I wouldn't suggest that speed and here's why.

The ball nut was never designed to spin at those rpm's so while you havent got the critical whip factor to deal with you still do have to consider the nut speed and what it was designed to spin at. Most ball nuts expect to be fixed and are designed for low spinning screws. So with Servo's spinning at 3000rpm or higher then you will have trouble with premature wear. Also, the alignment will be much more critical with higher RPM's.

My advice is to use a higher pitch then apply a ratio to lower the rotation speed of the ball nut while still getting the rapid feeds you would like. You will also get the advantage of increasing torque so you can lower the size of servos.

Now, where it gets tricky is with the Ball screw size. The larger Diameter screw will probably Sag more under its own weight so more tension will be needed. So you will need to design strong brackets that allow for the screw to be adjusted for alignment.

My suggestion is to go with a 32mm pitch and ratio of 2:1 that will give the equivalent of 16mm pitch and 1500rpm at the nut with 3000rpm motor. Giving just short 1000in/min which could be pushed up 1575in/min with 5000rpm giving 2500rpm at the nut which I would suggest is pushing the limit.

Hope this helps. Oh and this machine is vertical but that doesn't make any difference.

[Kitwn]

My advice is to use a higher pitch then apply a ratio to lower the rotation speed of the ball nut while still getting the rapid feeds you would like. You will also get the advantage of increasing torque so you can lower the size of servos.

I've read this recommendation several times but doesn't the increase in pitch cancel out the increase in torque so that the maximum linear force that can be applied to the gantry without stalling the motor remains the same? Or are you more concerned about the moment of inertia of the ballscrew itself?

Kit

[JAZZCNC]

I've read this recommendation several times but doesn't the increase in pitch cancel out the increase in torque so that the maximum linear force that can be applied to the gantry without stalling the motor remains the same? Or are you more concerned about the moment of inertia of the ballscrew itself?

Kit

Kind of yes but no, or should say maybe your misunderstanding what's being said or maybe I'm not putting it over clear enough..

Yes, 32mm pitch with 2:1 ratio will give the same linear torque as 16mm pitch 1:1 so torque is canceled by pitch difference like you say.
However, when I say Torque is doubled so the smaller motor can be used, what I'm referring to is that if 1:1 the 32mm pitch would require a larger motor to give the same linear force as 16mm pitch. So the ratio doubles the torque and allows the smaller motor to have the same linear force as 16mm pitch.

The trades off's and gains are that 32mm linear movement halves but so does rotation speed of the nut. So the pay off is that you get the same feed and torque as 16mm at 1;1 but with half the rotational speed of the nut or screw. This is the main goal to reduce whip and stress on the ballnut.

Hope that makes more sense.!

[teamloks]

HI JAZZ thanks for your insight regarding my question just talk with my ball screw supplier he suggest me to use 4040 for 3.2meter axis and 3232 for 2 meter axis..and i might follow their suggestion regarding that if budget is not constraint..anyway maybe i opt for 3232 all axis 4040 is really expensive

[Kitwn]

Yes, 32mm pitch with 2:1 ratio will give the same linear torque as 16mm pitch 1:1 so torque is canceled by pitch difference like you say.
However, when I say Torque is doubled so the smaller motor can be used, what I'm referring to is that if 1:1 the 32mm pitch would require a larger motor to give the same linear force as 16mm pitch. So the ratio doubles the torque and allows the smaller motor to have the same linear force as 16mm pitch.

Aha! The point I had missed is that your 'smaller motor' required to drive a 32mm pitch screw with 2:1 pulleys is NOT smaller than the motor required to directly drive the 16mm pitch screw. In fact it may well be the same motor, which is at the heart of what I was trying to say. Thanks for clearing that up.

Is there a case for using the direct driven screw with a bigger motor? You are saving the cost and complexity of the 2:1 drive and halving the speed required from the motor, so running it in it's higher torque region anyway?

Kit

[Voicecoil]

Reading this thread with interest (as I may need to go this route on my "big" machine soon), I noticed the concerns over rotation speed of ballnuts. This made me recall some stuff I read in the TBI catalogue some months back about different types of ballnut (differing mostly in the configuration of the return paths) and their relative speed capabilities, worth a read and maybe doing the sums they outline there.

[JAZZCNC]

Reading this thread with interest (as I may need to go this route on my "big" machine soon), I noticed the concerns over rotation speed of ballnuts. This made me recall some stuff I read in the TBI catalogue some months back about different types of ballnut (differing mostly in the configuration of the return paths) and their relative speed capabilities, worth a read and maybe doing the sums they outline there.

Certainly worth a read but what helps with rotating nut is to get shortest length Nut possible.
It's also critical that the Shaft as a Flange which is machined perfectly concentric to the Nut otherwise concentricity to the screw is misaligned so causes big vibrations.

[Voicecoil]

Certainly worth a read but what helps with rotating nut is to get shortest length Nut possible.
It's also critical that the Shaft as a Flange which is machined perfectly concentric to the Nut otherwise concentricity to the screw is misaligned so causes big vibrations.

What's the thinking behind the shortest possible nut please?

The concentricity is an obvious thing and might rule out some of the cheaper offerings if my experience on the machine I've just completed is anything to go by. Having set everything up and got it working smoothly I thought I'd spin one ballnut through 180deg to get better access to the lube nipple. Major fail, that side of the axis got really stiff, it turned out that the casing/flange of the ballnut was something like 1 or 2 degrees out of alignment with the axis of the screw.

[JAZZCNC]

What's the thinking behind the shortest possible nut please?

access/interference and balance/inertia mostly, it's no big deal and depends on the design to some degree but shorter is better in my experience.

The concentricity is an obvious thing and might rule out some of the cheaper offerings if my experience on the machine I've just completed is anything to go by. Having set everything up and got it working smoothly I thought I'd spin one ball-nut through 180deg to get better access to the lube nipple. Major fail, that side of the axis got really stiff, it turned out that the casing/flange of the ball-nut was something like 1 or 2 degrees out of alignment with the axis of the screw.

It's not always the case obvious to everyone that's why I mentioned it. However most ball-nuts, even the cheap ones, have a machined reference surface on the outer body which you can chuck upon and skim flange if needed. Thou I've not had any issues if honest.