Announcement

Collapse
No announcement yet.

Beware parallelograms (or un-square parallels)

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

  • Beware parallelograms (or un-square parallels)

    I was squaring some stock up on the mill, but it just wasn't working - sides were not square, opposing faces not parallel - what was going on? I'd never produced stock this bad.

    Much head scratching and checking later, one thing I'd taken as a given (how naive of me) was that the top/bottom of the parallels would be square to the sides. Below is the parallel sat vertically on the surface plate with a square brought up to it:

    Click image for larger version

Name:	20200510_unsquare-parallel-4598.jpg
Views:	223
Size:	201.0 KB
ID:	1875103

    Oh yes that explains a lot, I can see two ways how that was messing me up:

    Click image for larger version

Name:	vice-workpiece-parallels.jpg
Views:	204
Size:	13.7 KB
ID:	1875104
    The left hand image, the workpiece is sitting on the not very well defined deburred edge, on the right image the parallel is sitting in the vices grinding releif between its fixed jaw and its bed.

    Of course Mr Murphy was in the shop during this, this pair of parallels was the worst in the set. Checking the rest they ranged form being visually spot on to these.

    I swapped to a thicker pair from a different set, which were visibly square and suddenly I could now produce square/parallel stock again - if somewhat undersize now!

    So I'm now thinking how common is it for parallel sets to be all over the place like this, despite the impressive parallelism figures quoted for the sets only Suburban specify squareness - and I don't have enough body parts to sell to be buying those any time soon.

    Naively, as I've never owned or run a surface grinder, I would have thought if you are going to trouble of achieving parallelism, squareness to a reasonably high degree (on thinner parallels) was almost guaranteed.

    Thinking further - how to quantify/measure squareness of thin (1/8") parallels like these? For thicker parallels I can see clamping to a known square object (123 block) on the surface plate and running an indicator across the top, but across 1/8" isn't much of a sweep.

    Andy

  • #2
    That seems really bad, and you've done a great presentation of the pitfalls. Who made/sold them? Best imo to name names, hold em accountable etc. For sure you should not have to regrind/rework tools you buy.

    As for quantifying it, usually squareness is stated as out so much over a such and such a distance. You've done that and while not to a tenth of course, when its this bad the rule has quantified enough to know they should be returned. You could clamp to a square/angle plate known to be good and indicate across the top for a more precise quantification

    Like probably all of us, i use the 1/8 parallels, but I also have a lot of the tradition wide pairs that eliminate this problem...a problem that until this post it didn't imagine existed. I mean you';d have to really try get them that far off lol
    Last edited by Mcgyver; 05-15-2020, 06:32 AM.
    in Toronto Ontario - where are you?

    Comment


    • #3
      No question if new they would be going back - but I've had them several years, used (abused) them on drill press before I got the mill.

      If I was 100% sure who they came from, yup no problem naming, but as time has passed I can't be 100% sure, would have been one of the UK importers of no-name product from the east.

      Comment


      • #4
        Originally posted by ATW View Post
        I was squaring some stock up on the mill, but it just wasn't working - sides were not square, opposing faces not parallel - what was going on? I'd never produced stock this bad.

        Much head scratching and checking later, one thing I'd taken as a given (how naive of me) was that the top/bottom of the parallels would be square to the sides. Below is the parallel sat vertically on the surface plate with a square brought up to it:

        (...snip...)

        Andy
        Interesting. As a matter of curiosity, how much variance were you getting in your work piece(s)?

        I'd probably move quickly from head scratching to chalking it up to bad technique.



        Lynn (Huntsville, AL)

        Comment


        • #5
          Opposing faces of the block were something like 0.1mm (4th) out of parallel over 100mm (4"). Typically I'd expect (and be happy with) 0.01~0.02mm variation.

          Comment


          • #6
            Hmmm, Makes one think. I have 3 set of the import 1/8" thick parrallelles at work, and while I've mic'd them all and they're all withing a "tenth" or two from nominal size, I've never check them for square like that. Never experience any out of square blocks that weren't user error, but now you have me curious, for curiosity sake.

            I DID run into this problem with some O1 flat ground a few times in the last few years. Usually it comes nice and square, but the last one I remember was 0.75x0.25" and was out 0.007" across the short side. Usually when I mill pockets for slides I'll make them 0.01" under so we can tickle the flatground to a nice sliding fit, but this whole batch (3 or 4 36" sticks) were all useless. I had to mill some from larger pieces. Flat ground doesn't mean square ground....

            Comment


            • #7
              very good post. usually parallels have a larger deviation (even in absolute terms) in width/height than in length. lets say a 8x150 mm parallel might have 3µ in 150 mm and 5µ in 8 mm. it might be different if you get a set for $500. its an interesting question why that is. and, btw, parallels have a higher precision on the flat side than on the narrow side, at least when made to the IT standard, simply because its shorter.

              however, when used in a milling vice it doesnt matter much. the movement of the part when clamping is much larger. its advisable to know your vice. often it acts in a different way than you might think. e.g. a part will sit flat in my hydraulic vice, if i tighten just beyond the "hydraulic point", tap it down with medium force and proceed to tighten to the third mark (3 tons?). anything else and its not flat.

              Comment


              • #8
                It's something we normally take for granted, like the 2-4-6 blocks that had me going around in circles when testing the mill for squareness. They turned out to have at least 0.020" slant in the 4" side.

                Comment


                • #9
                  OK, I wrote this first paragraph before I noticed that you do not remember the source. I still want to say it, so here it is anyway.

                  The one in your photo does not show any signs of wear or abuse. I suspect yours are just as good as mine which are new enough to have only been used a few times. SO, do not feel guilty about naming names. IMHO, the standard, 30 day warranty is a crock. If I make and sell something, I am ALWAYS willing to take it back if there is a defect. If they don't care and hide behind a time limited warranty, then they deserve to be named. They are perfectly free to make it good by taking your defective ones back and either replacing them with a set that is checked to be correct or to refund your money. If they are not willing to do that for a product that is so obviously defective (manufacturing defect), even after several years, then they deserve the bad publicity.

                  I think I am going to go check my set of parallels.

                  As for how to check 1/16" thick parallels, I think your own photo answers that. Just let them sit on the edge, on a surface plate and sneak up on them with a square as you have done. It does seem to work. Some sort of careful clamping after it reaches it's own position could be used if you need to make actual measurements. Then you could carefully use feeler gauges or an protractor head. Or a low power microscope with a calibrated reticle could be used to get some numbers.



                  Originally posted by ATW View Post
                  No question if new they would be going back - but I've had them several years, used (abused) them on drill press before I got the mill.

                  If I was 100% sure who they came from, yup no problem naming, but as time has passed I can't be 100% sure, would have been one of the UK importers of no-name product from the east.
                  Paul A.
                  SE Texas

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

                  Comment


                  • #10
                    I have a set of cheap 1mm thick parallels, I'm not going to worry about their edge squareness., but the main set which go up to 6mm thick, are worth a check.

                    Comment


                    • #11
                      despite the impressive parallelism figures quoted for the sets only Suburban specify squareness
                      I looked at Suburban's specs. If I am reading it correctly, the only spec. that I see is for parallelism over the 6" length. I do not see any specs. on the squareness. I think Suburban products are very high precision and am not questioning their quality, just saying that I do not see that spec. I suspect that any really impressive squareness spec. would be hard to verify and would add a lot to their already high prices.

                      And I would think that the way to make a pair of parallels would be to clamp them against a right angle fixture on the surface grinder table. So how does that go wrong? One would think that even a nominally good, right angle fixture, would be a lot better than the parallel in the photo. Failure to dress the wheel often perhaps?

                      All of this highlights the reason why high precision tools and parts can be quite expensive. A machinist may do all the right things, but if the tools and measuring equipment are not also verified in all respects, then the result can be out of tolerance. And we scratch out heads as to what went wrong.
                      Paul A.
                      SE Texas

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

                      Comment


                      • #12
                        Sorry, should have qualified the Suburban tool reference further, they do "2-way" and "4-way" parallels, yup the "2-way" its just parallelism, the "4-way" they add squareness (and a fair few dollars to the price):

                        https://www.subtool.com/st/p_suburba...els.html#pairs

                        Comment


                        • #13
                          You didn't mention who they were made by.

                          Sounds like cheap mass ground import crap quality.

                          I've bought ground tool steel and found that the edges are not square to the sides. Happens all the time, but that's tool steel not being sold as parallels.

                          JL.................

                          Comment

                          Working...
                          X