how to straighten an acme leadscrew ?

[belgrado]

Dear cnczone community ,

I am building a machine with 16mm diam /4mm pitch metric acme rolled leadscrews on it .(nice quality )
The x axis is 1400mm long and the Y 600m long .
They don't seem to be really straight in some area's.probably because they are rolled .
If anybody out there could give me some tips on how to get them as good as possible .
I read something about hanging them vertical and heating them with a torch would help .
Thanks in advance for replying

Paul

[Mcgyver]

how much are they out? if its extreme return them or contact the manufacturer regarding what the tolerances should be, but its hard to know whether your have a problem without knowing the magnitude of the problem

heating it seems like a bad idea, 1) you'll anneal them, 2) whatever shape it is in now is the equilibrium of the forces in the steel. when you heat them it will move, you may end up with a pretzel. even you had it commercially normalized and didn't care that it was annealed, it would not come out straight. maybe the idea some localized heating, the way sheet metal bashers uses heat to change the shape of metal, not sure I'd want go there though

I've never done it but shafts are generally straighted using a arbor or hydraulic press. maybe someone here has experience and can give better instruction, but I'd do it careful with indicators and in a few steps, press so much, measure, press again etc.

[belgrado]

Thank you Mcgyver,

So it's better to put my torch back in its place .
I took the leadscrew off my machine ,put it on a flat piece of marble and the deflection is about 0.5 mm on the total lenght of 1400 mm .
I put it on two blocks and tried to bend it straight by hand and it seems to go better .It is not a bend in one place but more like a general curvature .
When I use it on the machine with the nut it gives me a bit more friction + wobble when it's positioned in the middle .(I only turned it manually )
Now is there any procedure to follow to get it as straight as possible ?

Greetings ,

Paul

[fkaCarel]

Manufacturers tolerance (Maedler) = 0.8mm per 1000mm. So maximum allowed runout for 1400mm = 1.12mm. You're lucky! It probably will sag 0.5mm under it's own weight. Better check that the nut/bearing alignment is good at both ends and live, or better appreciate it

[belgrado]

fkaCarel,

Thanks (bedankt )for the info . I'll try to align the bearing blocks as perfect as possible .
Maybe when running the machine for a while the leadscrew wil get better .I was also thinking of adding a second nut 240 mm distance from the other nut .(the width of the gantry base )
This would support the leadscrew a bit more but it will give me more friction .


Paul

[mxtras]

The second nut could be used to remove backlash, too.

If the .5mm is over the length as mentioned, it seems that you could put it between a set of V-blocks (one at each end) and just using hand pressure, straighten the length using a dial indicator in the mid-point to find the highest point. This is how I would do such a task with such a diameter.

Scott

[NEATman]

Paul-
It is possible to straighten the screw stock with 2 "V-Blocks", a granite surface plate and an arbor press, but it takes a while to get the proper feel for it. You need to roll the screw on the granite, and look for the "light" between the screw and the granite, or use a feeler gage. You can then bend it back to the proper shape when you find the high spots. A dial indicator on the arbor press will help with the deflection required to straighten a certain bend.

Another idea is that the second nut needen't be a nut at all, it could be a bushing that rides on the OD of the thread. One additional trickier idea is to have moving bushings that suppirt the screw. It would consist of a bushing over the screw, and a support that is mounted to the linear guides of your machine. A magnet would hold the block in contact with your moving carriage, and as it got to a certain distance from the end (say 1/3 travel) it would hit a stop, and get left behind as the carriage continued down the travel. This way the unsupported length of screw is 1/3 of your travel.

Neatman

Neatman

[One of Many]

Paul,

Rolled acme stock or any rolled rod can end up with a helical wow along its centerline.

Since the diameter is relatively pliable. I might be inclined to grip this leadscrew in some true bored aluminum chuck jaws with maybe a couple inches sticking out. Check and correct the runout(revolving by hand) in 50-100mm stages or some such interval. This process may need to be done from each end, going past the mid point to equal out opposing directions. Then worry about a bend(if any) at midpoint once both ends are reasonably straight.

One concern is that a severe localized bend may distort the thread pitch at that position.

DC

[belgrado]

Scott , Neatman & DC.

Thank you for all the ideas and info .
I already started to bend it by hand pressure on some v block substitutes ( 3cut-off s of structural aluminum )
It seems to work but it's very tricky .I looked at some thicker diameter metric acme but they seem to have a deflection too .
I will try "One of many's" approach and start at both ends and make minor corrections every 100 mm.
The idea of the extra moving support sounds good but I'd have to modify too much on my actual structure.
I will try the second nut or a bushing of the outside diameter .
I will be using bronze nuts and I was thinking of slicing them in half hand make some backlash adjustement . I will have to measure backlash once in a while and adjust it since it doesn't adjusts itself .The only thing that worries me is that if the pitch is not even all over the rod it might get stuck at some points .
Maybe I should try to compensate backlash by sofware as close as possible .
This machine will mainly be used for wood .
I'll put some pics of my "to be machine " on cnczone soon .

Thanks ,
Paul