Announcement

Collapse
No announcement yet.

Leadscrew nuts

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Evan
    replied
    So all in all, you probably have between 7 tenths and a thou of backlash, which is great, but it's not zero
    That's where the compliance of plastic enters the picture. It only has to comply by no more than .0001" per pitch revolution to accomodate that error according to your calculation and still give zero backlash. In practice rolled thread leadscrews are usually quite a lot better than the maximum allowed tolerance.

    commercial acetal nuts come with sophisticated anti-backlash mechanisms:
    Including mine. The nut, not yet split,can be compressed by the mechanism on the housing. There are two per leadscrew and the distance between them is adjustable as well.



    As I said, even though it has zero clearance as made, I don't expect it to stay that way without adjustment.

    Leave a comment:


  • lazlo
    replied
    That spec is the maximum error per foot, not the error over the length of the nut.
    OK, let's calculate the backlash for your acetal nut:

    The actual 3C acme spec is +/- 0.002" per foot. That works out to 0.002" *2 (if the tap and nut are minimum/maximum) / 12" = 0.000333" (3.3 tenths per inch).

    Assuming your nut is about an inch long, your maximum theoretical backlash would work out to 3.3 tenths (which is actually pretty darn good).

    However, the manufacturer of your screw has to add that much clearance tolerance to the flanks of the screw threads to make sure it will clear the worst-case nut. This will add another 3.3 tenths to the combined nut/screw combination.

    So now we're up to 6.6 tenths...

    But this calculation assumes that the included thread angle for both the nut and the screw is exactly 29 degrees (the Acme thread standard). There's going to be a manufacturing tolerance to the included angle depending on where you got the screws, and this adds to the pitch error clearance above.

    Then you need to add the pitch error introduced when you tapped the Acetal. As I'm sure you experienced, Acetal tends to drag/pull on the tap, stretching the threads as you tap.

    So all in all, you probably have between 7 tenths and a thou of backlash, which is great, but it's not zero

    All commercial acetal nuts come with sophisticated anti-backlash mechanisms:

    http://www.kerkmotion.com/learn/kerk...h-overview.asp

    If it was as simple as using a 3C (or 4c, or 5C) tap on a 3C screw, do you think the manufacturers would bother with the anti-backlash mechanism?
    Last edited by lazlo; 07-01-2006, 07:55 PM.

    Leave a comment:


  • Evan
    replied
    It's a nice dream, but don't think that will ever work with any load applied.
    I'm not counting on it. It just happens to end up that way when made the way I did. As I also said before, wear will produce clearance and I am providing for that. Not only is each nut adjustable for compression against the screw but I am using a pair of nuts on each screw that can be adjusted for backlash.

    Leave a comment:


  • Peter N
    replied
    Actually I don't think Evan has too much too worry about with movement on the acetal nuts.
    Although bronze has a modulus about 40 times greater than acetal the yield strength is only about 10 times greater. Acetal is a very stiff material and has a yield strength of around 65-70 Megapascals, so particularly with this bit of geometry it is not going to deflect very easily at all.

    Plastics are actually very good in compression and shouldn't be dismissed so lightly.

    Peter

    Leave a comment:


  • Mcgyver
    replied
    thanks lazlo, and BadDog thats what I was getting at - the backlash might not be in the clearance with a plastic nut, but in the distance it takes for the compression to reach equilibrium on the side of the thread when direction is reversed. i know ballscrews have back lash and the various techniques to remove most of it, either for ballscrews or acme.

    I'm making a quick and dirty prelude to a real cnc and have used two plastic nuts such that i can adjust the tension between, but it is for very light use - ie milling circuit boards. that same compresability of the plastic nut that allows for it to be in contact with both sides of the thread may in fact allow for enough movement that it's problematic - don't really know, just speculating.

    Leave a comment:


  • Guest's Avatar
    Guest replied
    {Plastics are compliant and will return to shape when pushed, metal deforms and stays that way.} Ya got that right, up to a point of permanent deformation of the plastics; but you sure about zero-zero backlash? It's a nice dream, but don't think that will ever work with any load applied.

    Leave a comment:


  • Evan
    replied
    Nope. The compliance that allows for zero backlash isn't over the entire length of the nut. It's from thread to thread and within each individual portion of the thread crest. Trying to move the entire nut in relation to the screw would require displacing a total of 1/2" of material per inch in shear. That isn't going to happen under any normal load but the compliance of the material allows it to fill the thread of the leadscrew so there is no gap.

    Once again, acetal nuts aren't some sort of easy way out, they are widely used in industry now. Before when I said the impact resistance was very high I didn't mean relative to other plastics. The impact resistance of acetal and nylatron is higher than metals commonly used in the same applications. Plastics are compliant and will return to shape when pushed, metal deforms and stays that way.

    Leave a comment:


  • Guest's Avatar
    Guest replied
    Well Bad Dog hit that nail on the head.

    Leave a comment:


  • BadDog
    replied
    Seems to me that the same flex allowing for “zero clearance”, and consequently “zero back lash”, also allows for more problematic lack of consistency. While a metal nut does not compress very much under load, plastics typically will start moving with relatively light loads (depending on durometer, volume, and all that). With metal, you can have 5, 10, or 20+ thou of back lash, but you can compensate (at least for manuals) since it is consistent. Once you take it up, you are ready to roll. But with plastic, even if it has no measurable back lash with no load, the minute you start pushing with up milling or holding back down milling, or whatever, you are going to get varying degrees of shift/compress depending on the size of the bite being taken, cutter sharpness, and the resistance of the material being machined.

    Leave a comment:


  • Evan
    replied
    Evan's 3C screw stock, for example has a typical pitch error of 0.002" per foot. So once you've traversed any distance down the screw, the crests of the threads can mismatch the valleys of the nut by 2 thousandths of an inch.
    That spec is the maximum error per foot, not the error over the length of the nut. The leadscrew nut may be out of position by up to .002 after a foot but the pitch at that point won't be .002" out over the length of the nut. That error spec translates to a maximum error of .00017" per inch of nut length.

    In the case of the acetal the compliance of the material allows for zero backlash. In the case of the bronze the oil film serves the same purpose as it basically has nowhere to go except down a long very tight spiral labyrinth in order to escape under load. It's similar to the hydrodynamic oil film that supports the spindle of my SB9 with plain bearings.

    Leave a comment:


  • lazlo
    replied
    Jon,

    Your gear hobbing setup is really neat -- I'm guessing that's home brewed?

    Robert

    Leave a comment:


  • lazlo
    replied
    if it has zero backlash, there must be at least some force acting on both sides of thread, correct?
    McGyver, your intuition is correct -- it's physically impossible to make a zero backlash nut. You need at least some clearance on the threads, or the nut won't spin. Even if you miraculously managed to get the thread forms on the screw and nut to be identical, there is still significant pitch error on the threads, which means the center-to-center alignment varies, and the nut would still bind without clearance.

    Evan's 3C screw stock, for example has a typical pitch error of 0.002" per foot. So once you've traversed any distance down the screw, the crests of the threads can mismatch the valleys of the nut by 2 thousandths of an inch.

    To put this in perspective, toolroom lathes have leadscrews with a Schlesinger limit of 0.0016" per foot, which is nearly the precision of 4C ground acme stock (very expensive!).

    ACME screw stock goes up to at least 5C precision -- Kerk stocks ground 4C and 5C screws, for example, and the thread to nut tolerances get smaller, but it's never zero clearance (which would be necessary for zero backlash).

    Even ballscrews have backlash (although they use balls running in a curved track instead of sliding surfaces), which is why high-end CNC setups use preloaded duplex ballscrew nuts.

    In any event, none of this matters for building a hobby CNC machine -- a lot of folks on CNCZone have made leadscrews with low-backlash nuts threaded in Acetal with a tap made from the same leadscrew stock, and it works fine.

    Leave a comment:


  • Mcgyver
    replied
    Evan, take this is a question not a statement - trying to understand plastics and zero backlash.

    if it has zero backlash, there must be at least some force acting on both sides of thread, correct? When I've tapped plastic before, the tapped hole ends up feeling slightly tight on the threaded rod or bolt. I've assumed that this is because the tap while its cutting compresses or displaces the plastic slightly as opposed to cutting a perfect profile, hence when the tap is removed is springs back and is tight fitting thread. I can see how this would create the force on both sides of the thread, it ends up being a very light interference fit, but not enough that it won't turn.

    So if the above is correct, this is why its able to turn with both sides of the thread in contact whereas with metals you wouldn't have this compressibility factor and it if was in contact with both sides of thread at once, it would seize - the compressible nature of the plastic takes the place of clearance??

    If that's the case, and the plastic is compressible enough that it allows it to turn with zero clearance, wouldn't you then have a problem similar to that of backlash because of this compressibility? ie there would still be an element of movement between the nut and screw when shifted from one direction to the other. In a metal set up its becuase of clearance, with plastic because its not as rigid a material and the transfer of force from one direction to the other changes which side of the the thread is comrpressed?

    Sorry if this makes no sense, it did when i typed it

    Leave a comment:


  • Evan
    replied
    ou mean low-backlash, right? If the nuts were zero backlash, they would have no thread clearance, and they wouldn't spin
    Actually, I mean no backlash. The nuts are a zero-zero fit. They spin just fine with only a small amount of friction. As soon as a film of oil is introduced it takes up any slack that may exist and acts as a squeeze film damper. This will also account for future wear, at least for a while. I do have measures incorporated to account for wear. If you look at the acetal nut housing you will see that it is split on one side. This allows for the nut to be split on one side axially and the housing to be compressed by a simple mechanism using a screw.

    As for the Nylatron I also have a nut made from that that fits the same housing. I plan on experimenting with various alternatives. The nuts are easy to change.

    Leave a comment:


  • lazlo
    replied
    Looks like Evan's been hanging out at CNCZone

    One of the materials I am experimenting with for nuts is acetal copolymer. This is similar to Delrin which is acetal homopolymer.
    Delrin AF is the thermoplastic often used in industrial acme leadscrew applications. It's Delrin embedded with 13% PTFE Fiber. Nylatron GSM (Nylon embedded with molybdenum disulphide) is another common thermoplastic used in low-backlash nuts. I have several sheets and rounds of both that I use for bushings, nuts and low-stress gears.

    both the acetal and bronze nuts I made have zero backlash as made.
    You mean low-backlash, right? If the nuts were zero backlash, they would have no thread clearance, and they wouldn't spin

    That's why your 3C screw stock has a tolerance of +0.0/ -0.0011, and why your tap made from this stock works. If both pieces (screw and nut) had any positive tolerance, the screw and nut combination would bind.

    If you want even lower backlash, you can find a bunch of posts on CNCZone where they split half of the acetal nut radially in three or four very thin slices, and then wrap the split sections with O-rings. This gives you a poor-man's anti-backlash nut.

    The industrial anti-backlash nuts typically have two paired nuts held on a captive secondary screw with a very coarse pitch, and an radially wound spring between them, that forces the nuts to unwind apart and take up any backlash. This works better than a spring in compression between the two nuts (which is a lot easier to build), since if you work the table in the direction of the compressed spring (as opposed to the wound spring), you lose some of the backlash reduction in that direction.
    Last edited by lazlo; 07-01-2006, 12:56 PM.

    Leave a comment:

Working...
X