Thread Locking With Steel

Discussion in 'Design and Modeling' started by Anubis22, May 3, 2017.

  1. Anubis22
    Anubis22 Member
    I am currently working on a design that needs a thread to connect two parts. Imagine the connection of upper and lower parts in a pen and you get the exact idea. The object will be printed in shapeways stainless steel. Due to the well known accuracy limits of 3D printing, the thread is designed to be very coarse.

    The actual question is, if it is enough to have the thread designed as seen in the pictures to get a good locking once screwed in.
    Phrased differently: I don't want the parts to be too loose; once screwed in it should feel reasonably secured, but still be easy to unscrew again if needed. With a plastic test model this worked fairly well, simply by applying a bit more force after it is loosely screwed in. But I am worried that this may not work well with the harder metal material in conjunction with the coarse thread.

    What do you think? Do I worry over nothing, just screw in with a little bit of force and it's gonna be fastened well enough (as with the plastic test or usual screws, which are much finer teethed)?

    Thread_1.png Thread_2.png

    Two possible solutions I thought of:
    • Using an o-ring / sealing ring which is getting squashed slightly and acts as a tightener <- seems like a safe rescue plan
    • Giving the bottom contact area a few small protrusions (tiny teeth) which can bite into the other contact area (would that work?)
  2. Your absolute best bet is to print the parts without the threads and machine them in afterwards. Don't be misled by the term "machined" as this can likely be done with hand tools - taps and dies.

    If printing is a must, your design may be aided by a few simple features. Based on your pen body analogy, I will refer to the female as a cap and the male as the body.
    • Truncate the crests of the threads a little bit. Keep enough thread on each part to maintain good thread face-to-face contact, but thin tips on printed parts are unlikely to be reproduced in fine enough precision to be useful. Excessive pruning is unwarranted, as you must maintain sufficient contact area on the working faces of the threads.
    • Leave sufficient axial gap between the threads to assure easy assembly. That is to say, thread thickness at the pitch diameter should be slightly less than half of the thread pitch.
    • Coarser threads are better, though not to the point of becoming a spiral nail. Six threads per inch is about as fine as I would dare but your mileage may vary, depending on material.
    • A shoulder on the body at the base of the thread with a corresponding bore in the cap will aid alignment at final closure. Anything at or above the major thread diameter will do. A similar approach at the tip of the body and root of the cap is also a good idea. This maked the threads no longer responsible for alignment, only axial force.
    • Acme standard and square thread profiles may offer additional strength and improve printability. Don't forget to stick to the rules about minimum wall thickness and you should be ok.
    Finally, don't ever trust 3D printed parts for strength. Shapeways offers any printed item strictly as non-structural or decorative in nature. It will (likely) hold itself up, but do not trust it as you would an ordinary mechanical part. Each printed part is unique, no matter how many times the same part has been printed. Testing of any one part does not validate the next part's strength. These are not engineering quality parts and should be used with that caveat strictly and strongly in mind.

    Good luck, happy hunting, and be sure to let us know how it turned out.
  3. Anubis22
    Anubis22 Member
    Thanks alot for your answer!

    Cutting the threads afterwards would surely be the best course of action, but unfortunately I lack all of the needed equipment. I do not even have a vise or something comparable to hold the piece securely while working on it. Besides I have read that the shapeways steel can be fairly hard to work with (though opinions seem to differ on the subject).

    Don't worry about strength. The parts in question do not play any kind of safety critical role. In case of failure the worst that could possibly happen is that I'd be sad.

    The current design amounts to 10 threads per inch. When you imagine the cross-section of the thread as a pyramid it is 1mm wide at the base and tapers to 0.6mm at the crest, with a height of 0.75mm. The diameter of the whole thread is ~10mm.
    The thread does not have to be compatible with any standard, only with it's printed counterpart. The version I outlined above is currently sitting on my desk, printed in Alumide and works fairly well. It does screw in and out effortlessly right out of the box. With the margins I used it is even a tad too loose for my taste. Only gripe is that it does not lock in a satisfactory manner without the need to apply a bit of extra force in the end (effectively compressing the material ever so slightly). This works well enough in Alumide, but I imagine it becoming worse when printed in steel, hence the question.

    I am not sure I understood your 4th point the way you meant it. So the shoulder would just be a protrusion of some form on the base of the male thread part (not on the thread itself) with a form-fitting hole in the female counterpart, so it slides in tightly? Where exactly would you place them and should they have a specific form? Maybe you could elaborate on that and correct if I understood wrong?

    Do you think the idea of using a sealing ring on the base of the male part (which would have an outer diameter slightly bigger than the female part's inner wall) to get a tight fit in the end is serviceable? If so how many un-screwing and re-screwings do you think the ring would survive?
  4. Sounds like you've got something vaguely resembling an Acme thread form. That should be plenty robust for a long service life.

    The "4th point"...
    Using your graphic above as a reference, left hand side of the image (the male portion):

    The graphic shows the thread "running out" towards the base, reverting to its minor diameter. Instead, have a short shoulder at the base at least as large as the thread's major diameter. Similarly, have a short stub "pin" at most as large as the thread's minor diameter.

    For the female, have a corresponding counterbore (open end) and pilot bore (interior end) to suit the shoulder and pin of the male, respectively.

    These two straight cylinders act as the centralizing feature so that the thread can be a slightly looser fit - preferable, given the relatively poor tolerance quality for printed parts. (Yes, they're fairly accurate, but not as compared to a machined part. I think you get my drift.)

    If you're still having trouble visualizing it, let me know & I'll whip up a picture for you.​

    An o-ring fitted into a groove at the base of the male feature (that is to say, not the tip) should do well for gripping the female bore. Lifespan will be heavily dependent upon the roughness of the mating bore. The o-ring can also be fitted to the female instead, depending on your need. You will easily find o-ring groove specifications online. Go with the specs for a static application, and perhaps make the groove slightly - just slightly - deeper than called for. Take a look at the actual cross sectional dimension of the ring, the groove depth, and the corresponding bore or shaft diameter, then take into consideration the printing variation. Industrial applications are designed around retaining significant pressure. Your application probably doesn't need such a tight fit. (Keep in mind we're talking in the range of thousandths, maybe 10 to 15 of them, of an inch difference. The smaller the cross section, the finnikier the fit. Also, give a small lead-in chamfer or radius to the part not holding the o-ring so that it doesn't scar the o-ring on assembly. Don't feel too put out if the o-ring doesn't last a long time/many cycles. 10mm o-rings are plenty cheap.

    I recommend you have a look around your area for a small machine shop. Nothing fancy, just some little Mom & Pop operation will do. Little jobs involving minor machining, such as tapping your threads, should cost nearly nothing, as that sort of thing is a 10 minute concern for them. Light jobs, like machining the mating bores & shoulders of this example, are a fairly trivial matter. Even a reasonably equipped automotive mechanic's shop might have all you need. Get to know them. Bring them donuts. Or beer & wings. Whatever the local favorite happens to be. Befriend them. You never know how they might help you if you do.
  5. Anubis22
    Anubis22 Member
    I don't know the standard's specifications, but going by a quick google search it seems to follow the same philosophy indeed.

    That would be great!

    Good to hear another opinion on the matter. No the fit wouldn't need to be watertight or otherwise perfect by any means. As long as it just increases the friction enough to give a confident feel that the thread will not loosen randomly I would be happy. I figured the surface would be important to the rings life, small damage can escalate quickly. I could smooth it out manually for better results and as you said replacements are easy to come by.

    Thanks for the tip. I should take a look around and see if I can find a suitable place to handle such things.

    I find your answers to be well thought out and substantiated; do you work with mechanics like these on a regular basis by any chance?
  6. I'm not in the office today, but I'll crank out an image after the holiday.

    There's much that can be done with simple tools to get the results you need. You don't need a mill or a lathe to shape materials. A dowel and a piece of sandpaper are plenty for smoothing out the bore, for example.

    Thanks for the compliment! I work as a machine designer in a fabrication and machine shop - CAD draftsman, engineering technologist, and chief electron wrangler.
  7. Anubis22
    Anubis22 Member
    Cool, I am sure it will help me understand it better.

    Oh, when I said to find a place to handle these things I actually meant to refer to cutting the threads and other tasks that involve more specialized equipment or require some dedicated workspace, not the sentence I wrote before. In hindsight that was formulated in a misleading way. I would certainly smooth the surface out myself. I actually even got some quality precision files for work like that. I am just missing an actual workspace and the will to invest in specialized tools for the time being.

    Ah I expected something along those lines. Sounds like a very interesting job description to me. Would you mind to go a bit more in depth about that, maybe per pm? I would like to hear more about your responsibilities, daily tasks and what you think of the job personnally.