Announcement

Collapse
No announcement yet.

What's wrong with my threading technique?

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

  • Originally posted by tomato coupe View Post
    Sorry, but I have to ask: Why do you argue about something you have no experience with and have limited knowledge of? What's the point? Do you just like to argue?
    because it was posted as reasoning here, supposedly relevant to the thread, and it did not appear to make sense, some of it was provably wrong. I wanted to find out what the deal was and whether it was actually relevant.

    I have found out what it's about (thank you), and that it is "marginally relevant" in that at least it talks about helix angle, which is probably part of the OP's problem, but he is not using that type, so that there is no other direct relevance.

    If I had used something similar to the OP's tooling, that would not make me any more expert about it, and the erroneous information in the link would still have been just as erroneous and confusing.

    Unless we have the maker and part numbers of the OP's tooling, we cannot know for sure whether his insert tooling is specific to one helix direction. It probably is, but we do not KNOW that, so knowing a lot about those OTHER inserts THAT HE IS APPARENTLY NOT USING is of pretty limited helpfulness.

    Maybe txfireguy can tell us what he IS using, so we can see for sure if it is possibly causing a problem.
    Last edited by J Tiers; 03-30-2018, 09:32 PM.
    CNC machines only go through the motions.

    Ideas expressed may be mine, or from anyone else in the universe.
    Not responsible for clerical errors. Or those made by lay people either.
    Number formats and units may be chosen at random depending on what day it is.
    I reserve the right to use a number system with any integer base without prior notice.
    Generalizations are understood to be "often" true, but not true in every case.

    Comment


    • Originally posted by txfireguy2003 View Post
      Okay, so I went back to basics and cut a thread the normal way. My starting OD was a bit undersized at 0.310, but it worked anyway. I fed the compound (set to 30 degrees) in 10 thou on the radius and made a cut, then another 10 and made a cut. Threads starting to look pretty decent, so I tried the knob on them. To my surprise, it went on a short distance, maybe two or three threads before jamming. I continued to cut, taking 1-2 thou each pass, and a spring cut after each pass (sometimes I'd get a few more chips, others nothing) until I got to a total of 25 thou on the compound slide. I test fit the knob after each consecutive cut and with each pass, it went on further and further before jamming. Finally, at 25 thou on the cross slide, I was able to thread all the way to the shoulder. Still though, it was somewhat sloppy at first, then got tighter toward the end, but I suppose that's fairly normal, plus the slop could easily be in the female part.....
      So your compound set at 30 deg, and you fed in by a total of 25 thou? And at the end of that it fit?

      What tooling did you use for this... pics maybe?

      You said 25 thou "on the radius". Does that mean the actual feed on the compound was 25 thou, or that the actual feed on the compound was about 30 thou? The 30 thou at 30 deg would be 25 thou actual radial distance.

      No, sloppy at first, and tight at the end is not so normal. Unless you are just saying that you could wiggle it more at first than you could later when it was threaded in farther, and NOT that it was actually "tight" as in hard to turn, at the end. (I am assuming you did not just run into the tapered thread that the end of the tap left)

      if everything is the right size, then it should just thread in nicely. The actual pitch diameter determines how close a fit it is.

      Can you measure pitch diameter? And if you can, what do you measure on the part? How about the PD of the tap you used (just as a "tie-breaker)?

      Did you de-burr before trying the part in the nut?

      If the tap has the truncated correct UN form, and you did not deburr, and did not knock down the sharp V of the thread you cut, then there might be interference between the thread crests of the part you cut, and the presumably correct flatted threadform that the tap cut.

      If you have that happen, you may end up cutting the thread too deep to make it fit, and end up with a sloppy fit. The best way is to measure the pitch diameter, and cut just until that is within spec. Then get rid of burrs and make sure the major diameter is also within spec.
      Last edited by J Tiers; 03-30-2018, 09:31 PM.
      CNC machines only go through the motions.

      Ideas expressed may be mine, or from anyone else in the universe.
      Not responsible for clerical errors. Or those made by lay people either.
      Number formats and units may be chosen at random depending on what day it is.
      I reserve the right to use a number system with any integer base without prior notice.
      Generalizations are understood to be "often" true, but not true in every case.

      Comment


      • Ta da! Yes, there will be burrs.

        It seems to me that almost all types of error have been eliminated by the OP. The pitch is correct. He checked that. The threads in the knob are apparently deep enough. My suggestion that he could be running into partial threads is probably not the case. He is seeing more resistance when he screws it in deeper. So, I doubt that the part is springing due to the cutting forces.

        What does that leave? As the famous detective was prone to say, when you have eliminated everything else, whatever remains, no matter how unlikely, must be the problem. So, I ask, is there anything that has not been eliminated besides burrs? I can only think of one more thing. The internal threads may have some crud in them. The only thing I can think of besides burrs is that the internal threads need to be cleaned. Run the tap back in there two or three times or pump some solvent into them from a parts washer.



        Originally posted by danlb View Post
        TL;DR There were burrs but it fit anyway.

        Full story.

        I just single pointed a 1 inch long 5/16-24 (5/16 UNF) thread with unknown aluminum.

        I turned the shaft down to .3105 to start. All measurements were with my caliper since extreme accuracy was not called for. I verified that there was no taper. I turned a run out groove .027 deep at the chuck side.

        I used a TCMT21.50 coated carbide insert. Compound set perpendicular to the lathe axis. Infeed via cross-slide. About 50 rpm.

        Multiple passes were starting with .010 and ending with a spring pass. . Nothing came off the spring pass.

        I measured the diameter again. It was 0.3150. Wow. That's 0045 bigger than when I started. Would it interfere? Good question.

        I rummaged in my screw and nut bins for a suitable screw and found I had none. I eventually found one in the 12-24 bin. It looked right but I verified it was really 5/16-24 by threading it onto a tap. It went on without cutting into the nut. The tap measured .3155 across the cutting edges.

        It fit well. No wobble. No binding.

        Paul is correct. There will be burrs, even if they are too small to see. I did not expect that. They should be cleaned up.
        Paul A.
        SE Texas

        And if you look REAL close at an analog signal,
        You will find that it has discrete steps.

        Comment


        • There is probably no need to try to diagnose his success.

          I am also curious about the insert that he was using on the back side.

          Dan
          At the end of the project, there is a profound difference between spare parts and left over parts.

          Location: SF East Bay.

          Comment


          • Originally posted by danlb View Post
            There is probably no need to try to diagnose his success.

            ...

            Dan
            He would probably like to know what went wrong with the original plan and what went right with the conventional approach. And maybe how to make threads that are less sloppy, and do not jam even when they are a snugger fit.

            He probably needs to be able to measure the pitch diameter. It;s pretty hard to diagnose ANYTHING with threading if you have absolutely zero idea what the pitch diameter is.... you have no clue if you made a good thread of the desired size, or not.

            If you just want nut A to pretty nearly fit thread B, you don't have to care.... But you don't always get to make both parts, and may not ever get "nut A" to check with, just a thread description. So then you have to do it by the book, to the PD and the major diameter..
            CNC machines only go through the motions.

            Ideas expressed may be mine, or from anyone else in the universe.
            Not responsible for clerical errors. Or those made by lay people either.
            Number formats and units may be chosen at random depending on what day it is.
            I reserve the right to use a number system with any integer base without prior notice.
            Generalizations are understood to be "often" true, but not true in every case.

            Comment


            • Originally posted by Paul Alciatore View Post
              Ta da! Yes, there will be burrs.

              It seems to me that almost all types of error have been eliminated by the OP. The pitch is correct. He checked that. The threads in the knob are apparently deep enough. My suggestion that he could be running into partial threads is probably not the case. He is seeing more resistance when he screws it in deeper. So, I doubt that the part is springing due to the cutting forces.

              What does that leave? As the famous detective was prone to say, when you have eliminated everything else, whatever remains, no matter how unlikely, must be the problem. So, I ask, is there anything that has not been eliminated besides burrs? I can only think of one more thing. The internal threads may have some crud in them. The only thing I can think of besides burrs is that the internal threads need to be cleaned. Run the tap back in there two or three times or pump some solvent into them from a parts washer.
              I'm not feeling any burrs when I run my fingers over it, but I suppose they could still be present. I caught grief for touching the tops with a file, so..... any other suggestions?

              I'm sure there was some junk in the female tread, as I have another copy of the sane knob that wouldn't thread all the way on until I ran the tap through it again.

              I agree, I need a method to measure pitch diameter, and I'll get one soon.

              Sent from my SAMSUNG-SM-G891A using Tapatalk

              Comment


              • TXfireguy; A couple of problems with pitch diameter... First is that it's slow and awkward. But that's OK, a lot of what we do is slow and awkward. Many people will say that checking the PD is a matter of QC to make sure that you did is within standards. They use either proper inserts and turning to the proper depths or they sneak up on it and use go/nogo gauges to tell them when to stop cutting the thread.

                Second is that there are not easy ways to check the PD of female threads. Thread wires won't work. Internal thread mikes are expensive. That leaves cutting to depth and validating with go/nogo gauges.

                I'd not worry too much about that handle that still wiggled when only one or two threads were engaged IF that's the only problem. Most threaded items have a tapered lead in. That includes your screw. When you have a taper there is nothing to support the far end of the screw for the first two threads. Get it screwed in one diameter and the slop should pretty much level out for the balance of the engagement. Then the slop will be the clearance between the threads.

                As for the use of a file to remove burrs... You can use one, but you have to then check the MD as you do it. As soon as the diameter is back to the starting point the burr is gone and if you file any further you are reducing the thread engagement.



                Dan
                At the end of the project, there is a profound difference between spare parts and left over parts.

                Location: SF East Bay.

                Comment


                • Originally posted by txfireguy2003 View Post
                  I'm not feeling any burrs when I run my fingers over it, but I suppose they could still be present. I caught grief for touching the tops with a file, so..... any other suggestions?
                  Use the file and ignore the purists.

                  As for checking PD, you clearly do not need to do that for every thread.

                  But you do need to be ABLE to do it, for cases where you need a measurement, such a your earlier problem with the reverse threading. A PD measurement would have quickly told you what you had. And, in that case, since you used a tap, the female threads should be according to standard. Problems measuring them are not an issue, and you already checked with other taps, as confirmation that they were at least reasonably close to spec.

                  a set of taps is actually a pretty good set of "go-gauges" for female threads, if you have no other method. They do not check for oversize, as a "not-go" gauge would, but they do check for being no smaller than the right size.

                  If you can find them, and decide you want the capability, thread micrometers are available. Each one covers a range of thread pitches, so one might cover all your work. They zero similarly to regular mics, but can also be checked aganst a known good thread or a gauge. And, especially for gun work, they can be set from one thread, and then used to size another thread to the same exact size. considerably faster than using the wires, but can be less accurate.

                  Burrs, or bad thread form, can throw a thread mic off, because unlike wires, it uses a sort of tiny v-block anvil and a cone-ended spindle that are intended to fit the thread tightly. They fit against a thread of good form perfectly, but a bad threadform they will give a bad number for. And they might be held away from the thread if there are excessive burrs. Most have some sort of recess in the V-anvil so a bit of a burr is not a problem.
                  Last edited by J Tiers; 03-31-2018, 12:57 AM.
                  CNC machines only go through the motions.

                  Ideas expressed may be mine, or from anyone else in the universe.
                  Not responsible for clerical errors. Or those made by lay people either.
                  Number formats and units may be chosen at random depending on what day it is.
                  I reserve the right to use a number system with any integer base without prior notice.
                  Generalizations are understood to be "often" true, but not true in every case.

                  Comment


                  • It might be worthwhile to make a casting of the female threads in the knob, using wax or some other material that can be removed without binding or sticking. Then you can use external thread measurement tools to determine the PD and other features of the hole.

                    I used a wire brush to clean up the threads I cut on the lathe. I also may have used a file or some abrasive cloth or Scotch-Bright. I think I used a standard nut to check the threads, but I didn't remove burrs, so I cut a little deeper. That resulted in threads that were too loose once I deburred the piece.
                    http://pauleschoen.com/pix/PM08_P76_P54.png
                    Paul , P S Technology, Inc. and MrTibbs
                    USA Maryland 21030

                    Comment


                    • Originally posted by txfireguy2003 View Post
                      I'm not feeling any burrs when I run my fingers over it, but I suppose they could still be present. I caught grief for touching the tops with a file, so..... any other suggestions?

                      I'm sure there was some junk in the female tread, as I have another copy of the sane knob that wouldn't thread all the way on until I ran the tap through it again.

                      I agree, I need a method to measure pitch diameter, and I'll get one soon.

                      Sent from my SAMSUNG-SM-G891A using Tapatalk
                      When you de-burr the thread crests use a small three sided file held so that the length of the file is in line with the helix of the thread, and tipped so it does not contact the thread flanks. I usually make two passes, tipping the file so it favors one side of the thread then the other, with the cutting face about 15 degrees off the thread axis. Also, I prefer to hold the file "backwards" and de-burr threads with the lathe running in reverse, letting the file ride from the chuck out to the end of the part. This is an operation best performed while wearing short sleeves, as the file is held with the right hand with the left hand guiding the file handle on the back side of the spindle. Be extra careful when reaching over a running spindle. Spindle speed is slow, about 100rpm. Don't pull the file towards you, just let it ride on the thread. I do this with the lathe in reverse so if the file does catch on the part and is pulled from my hand the file will get thrown away from, not towards me. Even if you take a bit off the major diameter while de-burring its only going to partially reduce the % thread contact, just as you would do with an oversize drill when tapping difficult material.

                      One of the most critical dimensions on a thread is the pitch diameter, so you must be able to measure it within the accuracy requirements of the job at hand. While they can be a bit awkward to use at first, a set of thread wires is a good start to producing threads where some sort of interchangeable manufacture is needed. Wire sets, complete with sizing calculations and look up charts are readily available from an internet search. If you have repeat jobs of the same thread pitch, then a thread micrometer is a bit faster and a little easier to use. They're also a bit more money.

                      Comment


                      • Somebody's going to call me a purist, but I'll say it anyway. Using a triangular file on the groove of the threads negates the value of using wires (or thread mic) to check the pitch diameter.

                        Here's why.

                        The PD is measured at the point in the thread where the metal is the same thickness as the groove. ( http://www.katofastening.com/article/science.html ) In a properly cut UN or ISO thread that point is 1/2 of the distance between the top and bottom of that theoretical sharp V thread that the UN thread is based on. In a male thread, it's 3/8th of the thread height less than the major diameter. On the female thread it's 1/4 height more than the minor diameter.

                        When you cut a thread properly, the groove is very consistent and measuring at any point will give the same reading. Once you run a triangle file over it, any change in pressure on the file will change the depth of the groove (and thus the width of the groove) at that point. The end result is that you will have (to one extent or another) a drunken thread. The reading at any point is only good for that point.

                        It's further complicated by the mechanism by which the "wire over threads" works. You start with the pitch and the "nominal size". Then you choose from a chart a wire that is close to the right size to touch the sides of the groove at the right spot. Then you put the three wires in the threads (2 on one side and 1 on the other) and measure the result, comparing it to a table. The problem comes in how close the wires come to the actual pitch diameter.

                        Note that in the example below there are 6 wire sizes spanning .014 inches from optimal. The max size for this example is .0375 and the smallest is .0233. Keep in mind that a wire that is .0375 is toward the top of a groove when the thread height is only 0.0361 inch. But it's still a valid measurement.

                        From http://theoreticalmachinist.com/Thre...dImperial.aspx for a 5/16-24 UNF thread.
                        MAX MIN
                        Major diameter: 0.3114 / 0.3042
                        Pitch diameter: 0.2843 / 0.2807
                        Minor diameter: 0.2618 / 0.2501
                        Over wires: 0.3204 / 0.3168 ( using a .0241 wire)

                        It is worth noting that the major diameter has a valid range of .0072 inches. The PD has a valid range of .0036, exactly 1/2 of the MD range. This is not a surprise since there is a direct geometric relationship between the PD, MD and thread form. That is, after all, why every major document says to start by turning the diameter to a specified size, then cut the thread to a given depth with a well described cutting tool. Then check with a gauge or thread mic.

                        Dan
                        At the end of the project, there is a profound difference between spare parts and left over parts.

                        Location: SF East Bay.

                        Comment


                        • Using a triangular file is FAR different from hitting the OD with a flat file. If the FILING is with a triangular file, then I join the purists.

                          I remove burrs with a flat file, and I see no issue with doing that. Never occurred to me to carefully run a thread and then file the thread flanks. That does negate (potentially) the accuracy of threading. At least if you bear down, and are not just knocking off burrs with a light touch.

                          I'll argue that the PD tolerance is independent of the MD tolerance, though. they do not need to go in lockstep.
                          Last edited by J Tiers; 03-31-2018, 08:51 AM.
                          CNC machines only go through the motions.

                          Ideas expressed may be mine, or from anyone else in the universe.
                          Not responsible for clerical errors. Or those made by lay people either.
                          Number formats and units may be chosen at random depending on what day it is.
                          I reserve the right to use a number system with any integer base without prior notice.
                          Generalizations are understood to be "often" true, but not true in every case.

                          Comment


                          • I do admit to using a small triangular file to take the burrs off, but only to take the burrs off, not to file the flanks. Don't forget that over here I'm often doing whitworth profile, which has a rounded top to the crest, so gentle application of the triangular file produces this as well, good enough for what I'm doing, which isn't commercial work.
                            'It may not always be the best policy to do what is best technically, but those responsible for policy can never form a right judgement without knowledge of what is right technically' - 'Dutch' Kindelberger

                            Comment


                            • I think I know where most of the ...don't use the file part came from..
                              It was the pic of the obviously flat crests that was posted. ThAT made some surmise he was majorly filing off the OD.

                              I too use the tipped file and a test nut, for most threading. I used to have to file the top of crests, still run it over in case there is a raised burr. I had to train myself to reduce OD a few thou , from nominal... that avoids anything piling up on the OD ..to the point of where it could drag in the nut.
                              Once you got your first thread figured out m you can go by the number and wire wheel the rest of the threads. I used laydown run straight in , for most threading, with excellent results.

                              I will suggest though, that if threading to a shoulder is a chore on the example provided, that a lot of gunsmithing work will be a real chore.

                              Comment


                              • Originally posted by J Tiers View Post

                                I'll argue that the PD tolerance is independent of the MD tolerance, though. they do not need to go in lockstep.
                                They sort of do. Consider an extreme example. If you require the MD to be exactly 0.3114 and the minor to be exactly 0.2618 and the cutter is a perfect 60 degree edge, then what is the tolerance on the PD? How can it be anything other than 0.2843?

                                Now take that same example and move the MD to 0.3104. That means that you will cut the thread .001 deeper. Now measure the PD. Remember, the PD is that point where the thread and groove are of equal width. That's the definition. Is your PD still 0.2843? Nope. It's moved down by .0005

                                What does this tell us? If you allow a range of MD to be acceptable, you have to allow a corresponding range of PD too.

                                Dan
                                At the end of the project, there is a profound difference between spare parts and left over parts.

                                Location: SF East Bay.

                                Comment

                                Working...
                                X