Announcement

Collapse
No announcement yet.

Scraping Question - 3 Squares Method

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

  • Scraping Question - 3 Squares Method

    So I'm slowly tooling up to scrape my shaper. One of the items I will need is a large, accurate square. Enco has some large cast iron squares on sale. I'm thinking either something like this:

    http://www.use-enco.com/CGI/INSRIT?P...PMAKA=418-4145

    or

    http://www.use-enco.com/CGI/INSRIT?P...PMAKA=418-4345

    Now, I know that I can scrape them "perfectly" square if I have three of them. But I don't want to have to buy three of these beasts.

    So is it possible to get away with two 6 by 6 squares and one big 12 by 12 square? Basically I would treat the big one like it was 6 by 6 but with an added step of scraping each surface flat after referencing one of the other squares. I.e. I indicate off of one of the other squares and the surface plate. Scrape as needed. Turn over and scrape the entire side into the plane that I just established using the small square. Turn around and reference the other square, etc.

    Or should I just get one and fiddle with indicators to check for squareness as I go?

  • #2
    have you got some sort of reference square now that you can set an indicator to with confidence? (....You either need a reliable square or something reliably parallel that you can get square to set the indicator then you scrape those blocks. If you don't have a good square to start with you can either make one or tune up an existing one....but it needs to be either box or blade shape; something with two vertical and parallel surfaces.....but you probably know all that)

    If so, I don't find scraping square using an indicator to be too much of a bother....scrape one flat, start scraping the second flat while checking squareness with the indicator....it'll be good practice as there'll be lots of scraping ahead of you where you're checking multiple relationships .

    Scraping big projects is the marathon run of shop work, just finishing is an accomplishment. Those not in the run look at those in it as possibly insane and don't understand the motivation.....when scraping a big project i to look at myself as possibly insane and not fully comprehending the motivation.

    of course take lots of pics
    Last edited by Mcgyver; 12-01-2010, 06:40 PM.
    .

    Comment


    • #3
      Squaring off

      Get one of these - a "Framing Level". Accuracy is 0.02mm/m ie 1:50,000 or ~ 0.00024" per foot.

      Should be available on eBay.





      Or make some cylinder squares which I used to "square-up" my standard angle plates.

      It all turned out (sorry) very well:




      http://i200.photobucket.com/albums/a.../Squaring5.jpg

      http://i200.photobucket.com/albums/a.../Squaring4.jpg

      Comment


      • #4
        Several years ago on PM, Forrest Addy posted about checking squares with an indicator and surface gauge.
        Hopefully he will see this, and repost that advice. I've got it in favorite places on AOL, which I can't access for awhile as my computer took a big dump the other day, and I'm just currently limping along.
        Harry

        Comment


        • #5
          There is a method that uses two squares instead of three described in Franklin Jones' Machine Shop Training Course. If you add "side by side" as well as face to face to your repertoire you don't need a third piece.

          Basic idea would be to and clamp them together side by side, then lap or scrape the perpendicular sides together. If they mate both side to side as well as face to face while sitting on the plate in all orientations, then they are square. For this to work, though, the sides need to be parallel. They were also cutting all 4 sides of the "square" in addition to the two unimportant sides which where flat and parallel to start with. They had similar methods for making straight edges, 4 sided squares, cubes, and 60 degree triangles using two pieces instead of three.

          You could easily cut a pair of squares that had a vertical surface that was tilted to the side, say 10 degrees, when cut side by side. When face to face, they would mate perfectly. When you flipped the mating surface down onto the surface plate, the new vertical faces would be flat against each other but leaning to either side. If the sides are parallel and you check that they are, then this error will show up. It will also show up if you flip one of the blocks.

          Note that lapping in this context refers to proper lapping against a flat lap or a piece of sandpaper on a flat surface, not throwing some abrasive between two surfaces being generated and rubbing them together.

          If you have a surface plate and another precision flat surface which you can mount at approximately right angles to the surface plate and adjust the tilt, you can mate that with a single square then use an indicator on a squaring stand to test that both the vertical flat surfaces form the same angle with the surface plate. This doesn't take into account side to side angle, if that is important in your application.

          If you don't have Connelly's book Machine Tool Reconditioning and Applications of Hand Scraping, you are going to have a harder time completing your project.

          Comment


          • #6
            Shaping up a shaper for shaping up

            Originally posted by Fasttrack
            So I'm slowly tooling up to scrape my shaper. One of the items I will need is a large, accurate square. Enco has some large cast iron squares on sale. I'm thinking either something like this:

            http://www.use-enco.com/CGI/INSRIT?P...PMAKA=418-4145

            http://www.use-enco.com/CGI/INSRIT?P...PMAKA=418-4345

            Now, I know that I can scrape them "perfectly" square if I have three of them. But I don't want to have to buy three of these beasts.

            So is it possible to get away with two 6 by 6 squares and one big 12 by 12 square? Basically I would treat the big one like it was 6 by 6 but with an added step of scraping each surface flat after referencing one of the other squares. I.e. I indicate off of one of the other squares and the surface plate. Scrape as needed. Turn over and scrape the entire side into the plane that I just established using the small square. Turn around and reference the other square, etc.

            Or should I just get one and fiddle with indicators to check for squareness as I go?
            Some of the squares are pretty heavy:
            http://www.use-enco.com/CGI/INPDFF?P...MITEM=418-4345

            A square near the weight of your shaper vise means that you will have countered part of the table vertical deflection under load as compared to squaring just a bare/unloaded table. It will get closer to the situation that will prevail when you use the shaper to "make stuff".

            If it were me and my shaper, my requirement would be that a dial indicator fixed to the ram and in contact with the table would show minimal dial deflection as the ram swept over the table (hand-powered - of course).

            The ram and the table top both need to be square to the vertical dove-tail guide that holds the shaper traverse slide which in turn holds the table. That dove-tail needs to be very true.

            I'd be looking at the ram dove-tail first.

            Comment


            • #7
              OT - this is a 26" shaper. My vise is too heavy for me to pick up, and I can lift 150 comfortably and 200 with a little grunt. A 104 lb square shouldn't be too ridiculous.

              Thanks for all the suggestions. Beckley, I did read a description by both TGTool and Forrest Addy about measuring a square just using an indicator and surface gauge.

              I'm leaning towards getting a selection of squares and scraping them all to "master" quality (O'Connely puts that at 35-40 spots per square inch, as I recall). I'd like something that is self checking, which is what is attractive about the three squares method. It just seems like there is less room for error since I'm not making any measurements.

              But will the three squares method work with different sized squares??? I think it ought to, but I'm not entirely confident about this whole endeavor, yet.


              Also ... any thoughts about which square to get? I like the bigger ones, but the Interstate one seems to have a better design than those little wimpy ribs. What do you guys think?

              Also, I do have some ground wrist pins out of a diesel - the edges are ground as well. I'm told they make great cylindrical squares, but I've no idea how square they actually are.

              I should point out that these squares from Enco are already machined to be pretty decent. Even the cheap ones are 0.001 in 6", so I shouldn't have to do any machining. Just scraping.
              Last edited by Fasttrack; 12-01-2010, 10:49 PM.

              Comment


              • #8
                I bought a granite square from CDCO. It checked out just fine against the 3 good old American granite squares I have,when blue was sparingly applied and the squares brought together on a granite surface plate. Surprisingly accurate,and not too expensive.

                CDCO also sells that framing square. I had bought one of those also,and it checked out well against my Starrett master level,and granite squares.

                Comment


                • #9
                  Originally posted by Fasttrack
                  I'm leaning towards getting a selection of squares and scraping them all to "master" quality (O'Connely puts that at 35-40 spots per square inch, as I recall). I'd like something that is self checking, which is what is attractive about the three squares method.
                  man, I'd not want to go that route....i guess you don't know until you do it, but it seems like soooo much scraping and iterations. If you create square via the indicator method, and you're using a quality tenths indicator in good shape (not sticky) and the surface plate is this accurate, you could do square to a 10th over 12".....that's better than any square you're going to buy and approaches the max resolution you can get via scraping.

                  Unfortunately even unfinished box parallel castings are expensive in large sizes. I've got a 2" thick slice of 12*12 cast iron that one day will be a square i hope - waiting until the cnc is done so i can mill a pretty ribbed/lattice structure to lighten it. I wonder you you could scrounge a similar off cut....or even if, choke, cough outright buying it would be reasonable? Working with this shape makes the indicator method much easier imo
                  .

                  Comment


                  • #10
                    I'll look for that square checking how-to but until then consider this:

                    http://www.shars.com/product_categor...granite+square

                    They're probably cheaper than buying three 12" tool maker's knees.

                    Adding. A good hard square is a good item to have. If you can calibrate it and know the amount and direction of the error you can compensate for the error and thereby work closer that the basic tool would otherwise allow. Note: this technique does not work for everything but it works for squares and other simple tools.
                    Last edited by Forrest Addy; 12-02-2010, 05:18 AM.

                    Comment


                    • #11
                      Why not just buy a machinist square and use it?

                      I just picked up a Brown and Sharpe 20" for $50..

                      In that Mike Stet bridgeport rebuilding DVD that is all he uses to check squareness..
                      Precision takes time.

                      Comment


                      • #12
                        I'll admit there is something very appealing about making your own precision masters, gauges, etc.

                        I guess I need to do some thinking.

                        Comment


                        • #13
                          Itsokayonnaday

                          Originally posted by gwilson
                          I bought a granite square from CDCO. It checked out just fine against the 3 good old American granite squares I have,when blue was sparingly applied and the squares brought together on a granite surface plate. Surprisingly accurate,and not too expensive.

                          CDCO also sells that framing square. I had bought one of those also,and it checked out well against my Starrett master level,and granite squares.
                          Well well GW.

                          I'd have thought that would have had the "nutthinz az good az good old 'merican tools and iron = 'speshlly "Chinese" junk" etc. Brigade off and baying by now.

                          But for what its worth it accords with my experience as well as they are quite adequate for my purposes.

                          Comment


                          • #14
                            All square

                            Originally posted by Forrest Addy
                            I'll look for that square checking how-to but until then consider this:

                            http://www.shars.com/product_categor...granite+square

                            They're probably cheaper than buying three 12" tool maker's knees.

                            Adding. A good hard square is a good item to have. If you can calibrate it and know the amount and direction of the error you can compensate for the error and thereby work closer that the basic tool would otherwise allow. Note: this technique does not work for everything but it works for squares and other simple tools.
                            Thanks for that post Forrest as I will have a look into those squares and what the cost is to OZ is.

                            As the edge faces are square to each other and the main faces, using two of them together as a unit makes for a lot more uses and possibilities.

                            Comment


                            • #15
                              I'll admit there is something very appealing about making your own precision masters, gauges, etc.

                              I guess I need to do some thinking.
                              buying a big machinist square or an import granite square are good options, but they hold some difference to either generation or how i would approach it, scraping with an indicator, which are worth considering.....

                              1) you just don't know how how accurate either the machinist square or shars square is . Machinists squares go for little money vs new because you never know if they've dropped or otherwise out of tolerance....I bet they're not cheap sold calibrated and warrantied as such by a reputable seller. Buying an import square, you're faced with trusting their claim. A new Starrett or Mitutoyo with a cert is also an act of trust I suppose, but they have the credibility that most would trust their cert. The import (not even a brand) lacks credibility; you will not know for sure and some shapes can't be tested via indicator (without a master to set it to)

                              and 2) neither will be as accurate as what he could do himself. You might not need or want that accuracy but i understand why someone would. When things are done on a comparative basis, each generation can lose a little so the end result will be a reflection of how good the start was....having a very accurate square to a tenth or two over a foot would be a nice thing to have for this sort of work - basically for checking other squares.

                              There's also the intangible that it is very rewarding to make a high quality tool, something as good or better to what can be bought.
                              Last edited by Mcgyver; 12-02-2010, 11:04 AM.
                              .

                              Comment

                              Working...
                              X