25mm lead THK ground leadscrew - ditch or keep?

[tonyel]

Hi everyone

I am just considering designing and building a CNC router, partly due to having found this site, and partly because I found a load of THK linear bearings and two leadscrews in a scrapyard!

I have a question though: the leadscrews appear to be THK large lead double ball ground ballscrews, part number BLK2525-2.8, with a lead of 25mm, and are about 1300mm long. I am concerned that the lead is too high, and will be inaccurate and prone to being back driven. I am guessing that to some extent I can alleviate this by changing the current 18tooth pulleys such that I gear it down (about 2:1 looks feasible), and using a big stepper motor.

Does anyone have any opinion on this? Would I be better trying to sell/swap them for a lower lead type?

Any opinions/answers greatly appreciated!

Regards,

Tony

[irving2008]

hi Tony... you lucky b*****, which scrapyard was that?

25mm lead may give some issues with backdriving... you'll need to work out the reflected load/inertia back into the motor before you can make that assessment. With a microstepping drive you get the resolution even when geared down. It may be ok for X and Y, no good for Z tho.

regards,
Irving...

[tonyel]

Irving - thanks very much for taking the trouble to reply.

I couldn't believe my own luck actually - it's a scrapyard near Oxford, and I bought all the THK stuff they had. The funny thing was that, very few of my friends and family seemed to appreciate why I was so excited about the bits of heavy shiny metal I had brought home!

Do you have any idea where I could get new pulleys to replace my 18 tooth ones?

I'm designing it on Solidworks at the moment, so when I have some pics I'll post them.

The other thing is, I've got these two leadscrews, which I'm intending to use to drive together as the Y axis (is that right - the axis along which the gantry moves?). That leaves me still needing to get a couple of leadscrews for the transverse (X?) and Z axes. Do you recommend any?

Cheers,

Tony

[irving2008]

Tony,

Generally the long axis of the machine is the X and thats generally the gantry movement (or table movement in a fixed gantry). Y axis is transverse, across the gantry.

There was a thread on mycncuk recently on where to get pulleys in the UK

[ger21]

Id use em. Either as is or geared down 2:1. If you're worried about back driving, then you should be using bigger motors.

[tonyel]

What sort of forces are generated by routing aluminium? I'm not really worried about getting to high speeds, as this is only a hobby type machine. I'm thinking of doing a little FE analysis to see what sort of deflections I'll get in the structure, but I don't really know what kind of forces are involved. My guess is that it wouldn't be more than my weight (about 70kg, since you ask...).

[Al_The_Man]

Originally Posted by tonyel I couldn't believe my own luck actually - it's a scrapyard near Oxford, and I bought all the THK stuff they had. Tony

Boy, the scrap yards in Oxford must have changed since I lived there!!
But I must say, I did often get some beautifully electronic scrap from the Atom Energy Commission dumpster, which was a bonus for a struggling student.
Al.

[tonyel]

This place is magic. They've got the contract for places like Harwell and Culham labs, and a load of RAF stations. There's a Wessex helicopter (UK version of a Sikorsky S58) fuselage there at the moment, but I can't get it in the back of my car.

[irving2008]

Originally Posted by tonyel What sort of forces are generated by routing aluminium? I'm not really worried about getting to high speeds, as this is only a hobby type machine. I'm thinking of doing a little FE analysis to see what sort of deflections I'll get in the structure, but I don't really know what kind of forces are involved. My guess is that it wouldn't be more than my weight (about 70kg, since you ask...).

I did some research into this but there is little hard information. One approach is to look at the chip size and work that back to the cutter size, rotational and linear speeds needed and then equate that to the required energy which then translates to torque which based on cutter size translates to the cutting force. Its but a crude measure but allow a frig factor of 2 you can get some estimate.

So lets say a 10mm, 2 flute cutter milling aluminium taking a .5mm cut. The spindle speed required is 1000 * cutting speed)/(pi * cutter dia) = 320 * 100/10 = 3200rpm

Lets say you gear the screws down 2:1 and spin them at 2rps = 120rpm = 3000mm/m (motor = 4rps = 800 full steps/sec, full step = 25/400 = 0.0625mm, so microstep at 1/8 give better than 0.01mm resolution)

At 3000mm/min feed on a 2flute cutter the feed per tooth is 0.5mm/min and the removal rate Q = 30cc/min

Power = Q * k where k=12 for aluminium = 360W (k rises to over 17 for feeds below .2mm/min which is why you want high feed rates to reduce power input - non-intuitive or what!)

Torque = Power * 60/ (2 * pi *revs * eff) where eff is cutter efficiency of typically 80%
= 360 * 60 / (2 *3.14 * 3200 *.8) = 2.25Nm

which acting at 5mm radius = 2.25/.005 = 450N so you need to be allowing for up to 900N force on the workpiece from the cutter... so 90Kg, you weren't far off...

The electrical input will need to be Power/(cutter eff * motor eff) = 360/(.8 *.6) = 720W (so forget a Dremel!)


Don't take this verbatim, I may have made a boo boo... here is some useful guidance to milling aluminum

[tonyel]

This is incredible. You just figured all this out for me for the sheer hell of it? What can I say? Have a beer on me!

I like the sound of this 'ere 'microstepping'. Does this magically allow you to achieve less than the standard 360/200 deg/step movement on a stepper motor? (is it always 200 steps per circle in a stepper motor?)

I'm going to digest this, figure out my gantry design, and do a little analysis on the design.

I'm starting to settle on a table base made from a rectangle of 100x50x5mm mild steel box, welded together, and then I'm hoping to get a local machine shop to skim the top (depending on how accurately it ends up plane after welding) to get a true(ish) plane surface to bolt my THK X axis rails to.

I think I may have found a source of a lead screw that may do the Y axis; we have a scrap breast imaging machine at work, with a couple of lead screws in it for lifting the breasts up and down. Or is it the imaging head? Whatever. Anyhow, it looks like a good size, decent quality long screw (fnar fnar), but I suspect it wouldn't be a high accuracy class. Anyway, I'll check it out and see.

Thanks again for your research!

Cheers,

Tony

[irving2008]

tony,

You're welcome.. actually I'd figured out the math some time ago for my own... just plugged your numbers in...

For more info on stepper motor techniques there's a great tutorial at http://www.cs.uiowa.edu/~jones/step/index.html but yes, the idea of micorstepping is to interpolate between the natural steps of the motor, which is commonly 200 (1.8degrees) although other types exist.

If you want to ensure the rails are mounted on something level, consider self-levelling epoxy - there are several threads on the Zone about it.

"scrap breast imaging... couple of lead screws in it for lifting the breasts up and down" the mind boggles