Hi - hope this is the right forum.
Advice needed on max length before sag is a concern with 16mm diam rolled ballscrews (isel). I'm not worried so much about whip as I'll only be driving them Max 5 revs/sec.
Thanks
Ian
Printable View
Hi - hope this is the right forum.
Advice needed on max length before sag is a concern with 16mm diam rolled ballscrews (isel). I'm not worried so much about whip as I'll only be driving them Max 5 revs/sec.
Thanks
Ian
I looked for a ball screw calculator/formula, but I wasn't able to find one... so we can do a little logical assumption to help you come up with an answer. The stanards book I have says that the root diamater of a 16mm ball screw is about 13.4mm. So, if you do a calculation for the deflection of a 16mm steel shaft and for a 13.4mm steel shaft, the deflection of your screw will fall somewhere inbetween these two numbers.
To play on the safe side, you could just skip doing the 16mm shaft and only do the calculations for the 13.4mm shaft.
Sag = L^4 / S
L = length
S = Sag Coefficient
S = E x I x 384 / (5 x sw)
E = Modulus of Elasticity
sw = PI x diam^2 / (4 x density)
If you want to save yourself some math... pacific bearing's catalog has a lot of the math done for you and it's in chart format. So...
Sag = L^4 / S (getting S from a table)
So.. here are the entries for a 14mm shaft:
S = 2.45 x 10^12 for hardened steel
S = 2.31 x 10^12 for stianless
you pick your material
Don't forget to use millimeters for length.
Use the critical speed calculator here.Quote:
Originally Posted by stirling
http://www.nookindustries.com/ball/BallCalculators.cfm
Put 13 in for the root diameter, and 300rpm, and it'll give you a max length. It varies depending on the mounting.
Thanks to both of you. I calcualted from schbrownie's figures that with a 1m screw I'd have 0.4mm of sag, which is intuitively reasonable I feel. Wheras with a 2m screw (the length I actually need) it shot up to 6.5mm which I think is probably too much - though I don't really know to be honest.
(thought: if I never reach the critical speed - does 6.5mm of sag over 2m actually matter? - maybe that's where compressive force comes in but see fixed-fixed below - wow this gets complicated)
From gerry's calculator (I assume the resonant speed is where whip occurs or perhaps maximises) I'm within bounds. Of course the fact that the sag is probably unacceptable negates this anyway.
Whilst researching this I came accross the four accepted methods of ballscrew support again i.e. the so called fixed-free, simple-simple, fixed-simple and fixed-fixed. I currently use the fixed-simple scheme in the form of a double angular contact bearing for the fixed (driven) end and a fully floating bearing at the other end. I use this because I was told that if you use fixed-fixed then this leaves no mechanism for screw expansion/contraction with temperature change. Any thoughts on this please? I ask because the fixed-fixed system (e.g a double angular contact at both ends) appears from the data to give higher performance for a given screw.
FWIW I looked at techno isel's technical papers but I'm sure there's a problem with their critical speed/screw length/diameter chart. It doesn't appear to make any sense to me at all. For those interested it's at http://www.techno-isel.com/LMC/H860/PDF/H860P019.pdf page 22.
Thanks
Ian
You can use the calculator on this site for the exact length with the machined ends, you can calculate travel length and total length : https://www.vallder.com/en/c/ball-sc...om-size-detail
ShopSabre uses 12+ foot ballscrews on their IS 612 if I'm not mistaken. They rotate the ballnut rather than the screw to reduce some of the problems, but is this still a significant compromise? What kinds of problems can one expect to see on such a machine?
There is nothing wrong with this at all. The advantage is that there is no issue with ballscrew whip. Its not very hard to do, even use a couple of wheel bearingsQuote:
Originally Posted by GJeff
Do you know of any other CNC makers of large machines (4x8+ travel) that use two giant ballscrews on the longest axis? It does seem like a pretty good option to have them fixed and the nut rotating. I guess the real downside is just how much very long ballscrews cost.
Sorry I don't . I built a plasma cutter and am very happy with the accuracy of the rack and pinion drives I use. You have not said how big you arr going and what sort of machine. Mine is 4' x 4' and I cut some M5 clearance holes the other day at 2200 mm/min which I thought were amazingly accurate for a home built machine.Quote:
Originally Posted by GJeff
GJeff, We have a few Thermwood large format routers at work. 5' x 10' and they have dual ballscrew on longest axis. I havent measured but the screws seem like 2510 styles and are belt driven screws. The ballscrews does have 2 'whip preventers' on each that prevents the screws from whipping. We cut at 300IPM with faster rapids (unsure on rapids)
Hi Ginger, what does a whip preventer look like? Peter
Not sure on the real name but it is close fitting and is probably to limit the amount of sagging that can happen under the weight of the screw and by limiting the sagging you can limit the whip effect in theory (at least according to me)
These are double ended machined plastic supports, 1 end in contact with the screw and as the gantry moves across it the ballnut rotates the block so the other end is in contact once the gantry has cleared the preventer. Hopefully the pictures explains it.
The ballscrews on closer inspection seems to be 2525
That is the coolest idea i have seen recently !
:cheers: