Announcement

Collapse
No announcement yet.

Scraping a straight edge longer than your surface plate?

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

  • Scraping a straight edge longer than your surface plate?

    Hi folks,

    I guess this is a question for Forrest or Richard but I'd appreciate any advice in general. My scraping has been coming along nicely for some months now, not least because of some good advice received here.

    Having repaired numerous small parts - lathe tailstock, compound slide, angle plates, vices, etc - I'm beginning to eye up my RF45 clone mill for a complete, bottom-to-top rebuild and scrape. If I don't loose my mind I'll follow by doing my 1970's surface grinder and 10x24 lathe too.

    In order to do so I'm going to buy a 24" surface plate so that I can accommodate some of the larger elements - as well as a good grade 12" square so I can get the column and X/Y perpendicularity fixed.

    The one thing I'm struggling with are the table ways. Being 38" long they exceed the diagonal on the 24" surface plate.

    I think I have two possible solutions:-

    1) Buy a 40" straight edge casting and attempt to scrape it in using the (smaller!) surface plate, though it makes me a bit nervous not being able to spot or check for stability in one pass. If the table is unstable it might be hard to understand and I don't know if this is a recipe for disaster?

    2) Try to buy a granite or CI straight edge which is already sufficient and doesn't need scraping in. Since it seems the cost of a casting is near enough the cost of a 1000mm granite straight edge, the only problem is most of them don't have a knife edge to reach in to the dovetails.

    Would appreciate any advice particularly on whether (1) above is possible or not.

    Cheers,
    Rich

  • #2
    Buy three castings long enough and scrape them flat yourself?

    Edit: Same with the square. Or make an adjustable square from two scraped straight edges.
    Amount of experience is in direct proportion to the value of broken equipment.

    Comment


    • #3
      First you will need to know how good is your surface plate. I would suggest a Grade A or AA. I usually order a Laboratory Grade AAA. It should be sitting on 3-points and level. I have tested plates in the field by laying a straight edge like a yard-stick or a sheared piece of sheet metal about 36" long by 2" wide.

      You will need to have a precision level with an accuracy of .0005" / 12" or better. Put the straightedge diagonally on your plate and place the level in the middle, it's side against the S-E and zero it. Then move level 1/2 it's length outward and note the reading. Then go back to the middle and repeat going in the opposite direction. Move the straight-edge (S-E) on the other corner to corner and repeat. Find the straightest plane of the plate. You can test the entire plate in this manner. I call it "Graphing the Plate.

      Then place your longer S-E on the plate with it's center in the center of the plate. Both ends will hang over the edge. Cut 6 feeler gauges of the same thickness place them under the straight-edge and see if they are tight. Then move the S-E so the ends no longer hang over the edges. This will tell you what part of you S-E is lowest. Blue up your plate and scrape the lowest area first and the move the S-E and scrape the area that is now high. Scrape the high area only (never ouch where you have finished scraped) and when the low finished area begins to touch stop scraping the high end (s) I like to scrape a S-E .00005" / 12" with 40 PPI. You will need to get it at least .0002" / 12.

      Then repeat on the other high end. If you have a precision level you can also flip the S-E over and set it in a vise or 2 vises and level it. Then move the level along the S-E to see if it is flat. Once you have the S-E finished you will lap scrape the grinder in the same manner. Find the lowest area and scrape there first and bring the rest down to it's level. I would do the flat on the table first and then lap scrape the V's and use a gage to get the V parallel to the flat. You will need to also set the table on 3 points to be sure it's not twisted when you scrape it. It is difficult to explain sometimes by writing it. If you can't figure it out, please call long distance....a very long distance. 001 651 338 8141. But email me when your going to call.
      Good Luck. Rich

      Comment


      • #4
        Why not buy a granite 40" surface plate and a 40" granite straight edge - both new and Grade AA or better.

        Michael Ward (VP magazine author) - aka McGyver- here covered the scraping you intend - dove-tails included - in a 20-part (or there-abouts) series of scraping - which were excellent.

        If you don't have the relevent copies then PM/fax George Bullis - the Editor of VP magazines and the moderator of/for this forum for back copies of the magazines.

        I'd suggest that it would be best if you read (and re-read) those articles and thoroughly understood the principles and processes before laying a scraper on the mill components.

        Comment


        • #5
          I was answering his question "Scraping a straight edge longer than your surface plate?"

          Loply and others may as well know how in case they have a 100" way and a 48" straightedge. I always teach "you use a straightedge longer then the way your scraping", but sometimes it isn't possible. One thing I forgot to mention never scrape an area more then a few inches longer then your straight-edge, once you finish the low area move out 1/2 the distance of the S-E and continue out as I stated before and using a level to double check the flatness. Never slide the S-E as longer then a few inches because lets say the way is convex, as you move along the S-E will rock like a rocking chair and give you a false reading. Lap scraping is not easy to do. I just had to scrape the table of a Drake Thread Grinder. It had a 84" long table and I was using a 72" Camel-Back. It was tedious but it worked.

          Comment


          • #6
            Originally posted by oldtiffie View Post
            Why not buy a granite 40" surface plate and a 40" granite straight edge - both new and Grade AA or better.
            Because they'd cost a years wages! 24" surface plate is expensive enough, I dread to think what a 40" one would cost.

            Maybe I can do a bit of compromise here - 24" plate diagonally is 30", so I can probably make a 33" to 34" straight edge on it.
            Then if I can scrape the 38" ways using that...? I know it's pushing the limits... But is probably possible?

            I do have a precision level, I forget what accuracy it is but it's a £300 thing, British made and good quality.

            Cheers,
            Rich

            Comment


            • #7
              I would suggest starting with Connelly's book "Machine Tool Reconditioning" first as it's far cheaper than all of the above. There are methods (water levels etc.) for working large machines with reasonable sized tools. Think of the giant lathes, slotters, VBM's etc used at the turn of the century for marine steam engines, etc...

              Edit to add link: http://www.machinetoolpublications.com/about.cfm
              Last edited by WCPenney; 11-23-2012, 07:20 AM.

              Comment


              • #8
                Originally Posted by oldtiffie

                Why not buy a granite 40" surface plate and a 40" granite straight edge - both new and Grade AA or better.
                Originally posted by loply View Post
                Because they'd cost a years wages! 24" surface plate is expensive enough, I dread to think what a 40" one would cost.

                Maybe I can do a bit of compromise here - 24" plate diagonally is 30", so I can probably make a 33" to 34" straight edge on it.
                Then if I can scrape the 38" ways using that...? I know it's pushing the limits... But is probably possible?

                I do have a precision level, I forget what accuracy it is but it's a £300 thing, British made and good quality.

                Cheers,
                Rich
                Maybe so - maybe no.

                http://www.shars.com/products/view/2..._Surface_Plate

                http://www.shars.com/product_categor...lates_Grade_AA

                Granite straight edges:

                http://www.shars.com/products/view/2..._Straight_Edge

                http://www.shars.com/product_categor...+straight+edge

                May seem expensive on dollar alone consideration but given the accuracy (new) you are ready to hit the ground running and be on your way as soon as they land - providing you've read Michael Ward's (aka McGyver here) recent "Scraping" articles in VP publication "Home Shop Machinist".

                If I needed them I'd buy them.

                But I don't need 'em and won't be buying them as I don't have a need for scraping.

                I have a copy of Connelly's book "Machine Tool Reconditioning" and its one of the worst written books I've read as while the content is OK the style of writing is terrible.

                I'd much rather read Michael Ward's articles as in my opininion they are precisely what it required and his approach, explanations, photoes and text are very well done - and they are a very good read as well.

                Comment


                • #9
                  Originally posted by Richard King View Post
                  First you will need to know how good is your surface plate. I would suggest a Grade A or AA. I usually order a Laboratory Grade AAA. It should be sitting on 3-points and level. I have tested plates in the field by laying a straight edge like a yard-stick or a sheared piece of sheet metal about 36" long by 2" wide.

                  You will need to have a precision level with an accuracy of .0005" / 12" or better. Put the straightedge diagonally on your plate and place the level in the middle, it's side against the S-E and zero it. Then move level 1/2 it's length outward and note the reading. Then go back to the middle and repeat going in the opposite direction. Move the straight-edge (S-E) on the other corner to corner and repeat. Find the straightest plane of the plate. You can test the entire plate in this manner. I call it "Graphing the Plate.

                  Then place your longer S-E on the plate with it's center in the center of the plate. Both ends will hang over the edge. Cut 6 feeler gauges of the same thickness place them under the straight-edge and see if they are tight. Then move the S-E so the ends no longer hang over the edges. This will tell you what part of you S-E is lowest. Blue up your plate and scrape the lowest area first and the move the S-E and scrape the area that is now high. Scrape the high area only (never ouch where you have finished scraped) and when the low finished area begins to touch stop scraping the high end (s) I like to scrape a S-E .00005" / 12" with 40 PPI. You will need to get it at least .0002" / 12.

                  Then repeat on the other high end. If you have a precision level you can also flip the S-E over and set it in a vise or 2 vises and level it. Then move the level along the S-E to see if it is flat. Once you have the S-E finished you will lap scrape the grinder in the same manner. Find the lowest area and scrape there first and bring the rest down to it's level. I would do the flat on the table first and then lap scrape the V's and use a gage to get the V parallel to the flat. You will need to also set the table on 3 points to be sure it's not twisted when you scrape it. It is difficult to explain sometimes by writing it. If you can't figure it out, please call long distance....a very long distance. 001 651 338 8141. But email me when your going to call.
                  Good Luck. Rich
                  Thanks, Rich, for the "blow by blow" description of how to deal with a longer SE. When I bought my plate, I thought I had a 60 inch camel back SE, so I ordered a 3 x 4 plate, knowing the diagonals were 5 ft. - it wasn't until I got the plate and put the SE on it that I realized I had a 72 in SE..... It was kind of obvious:



                  I have not needed it yet (I also have a 24 and a 48 in camelback and I took the lathe beds to be ground), so it is still unscraped. There is one rather deep (maybe .010) scratch from mishandling before I took possession that will probably remain, as, even with the Biax, I cannot contemplate scraping off .010 over the full 72 inches and Commerce Grinding did not want to attempt grinding it.... I think it would take less time to machine off the offending material on the mill in two or three setups, and then scrape it flat from there.

                  Would the taunt wire method yield better results than my 5 ft diagonal?
                  Last edited by TexasTurnado; 11-23-2012, 07:04 PM. Reason: typo
                  TexasTurnado

                  Comment


                  • #10
                    Why not have the plate and straight edge (re) calibrated.

                    http://www.tru-stone.com/pages/smp.asp#prod

                    At least you will know the accuracy (or lack of it) of your references.

                    Comment


                    • #11
                      To: Texas Turnado...
                      I would leave the camel back alone as it is only cosmetic. But you're retired and you have tons of time on your hand, right? Ha,ha.

                      For all that don' know Texas T, he is an accomplished rebuilder and I believe he could also teach scraping.

                      Comment


                      • #12
                        Wanting to scrape something (anything?) infers a need for known accuracy which in turn infers and requires a known reference which in turn has been calibrated and graded to specific standards.

                        Larger plates and staright edges are permitted to be less accurate than smaller ones.

                        The job cannot be more accurate than the reference and may even be less so.

                        It is quite possible that say a straight-edge scraped to a plate of known accuarcy (ie calibrated/graded) will no more and possibly less accurate than the plate is was scraped to.

                        The best/only way to determine the second-order "flat" definitive accuracy is to have it graded/calibrated.

                        If a known non-calibrated/graded straight edge is used to scrape a say mill table ("third order") that the table will be no more accurate than the second order straight edge.

                        Ambient conditions in the shop may also affect to result if the reference and the job being scraper to it vary with chasnges in the shop ambient conditions (ie temperature aboveor below 20 deg C / 60 deg F and humitity). As an extreme, what is OK at say 5.00pm one day may not be as correct at 9.00am the day after.

                        Here is a link to Mitutoyo USA granite surface plates - grades and accuracy.

                        Accuracy is given as +/- millionths of an inch (ie 1000 miliontht = 0.0001" ie "1 tenth".

                        http://www.mitutoyo.com/TerminalMerc...spx?group=1156

                        For comparison, here is the link to Shars black granite plates:

                        http://www.shars.com/product_categor...lates_Grade_AA
                        Last edited by oldtiffie; 11-24-2012, 05:53 AM.

                        Comment


                        • #13
                          Is it possible to pay someone with the proper equipment to scrape the few parts that exceed your equipment capability? May be a comprimise. Scrape what you can and farm out what you can't.

                          Sorry I didn't answer your question...because I don't know the answer. I would expect you could get enough accuracy on a 38" part with a 34" diagonal...but thats a pure guess.
                          If the women don't find ya handsome, they should at least find ya handy

                          https://www.facebook.com/WDHSTechClub

                          Comment


                          • #14
                            A former student bought new Shar's granite and blade square off E-Bay. Took them to the community college where he was taking some classes and the instructor measured both of them on the CMM and both were out of the certificate of accuracy sent with them. He had issues returning them.

                            On the note about temperature; The ideal temp is 68F, but constant temp is what I look for. You should not use a S-E or heck a machine after you turn off the heat or AC at night and flip it on in the morning as the iron will change. One other thing to watch out for is a window and the sun shines in at hits one side and heats it up.

                            I've seen where a heater blows hot air on one side of a machine and heats it up and the specs change. When I take a straight edge to a job-site, I let it normalize overnight so the iron expands the same all over. I have had issues with humidity too when I worked in tropical areas. I have scraped straight edges and Jig Bores for years and held .00005" never taking my straight-edges to a lab to have them calibrated. It's like the folks who now sell lasers and claim their is no better way to do it.

                            I always say to them, it's amazing then how we ever did anything before they were invented. I taught a scraping class at a GM plant once where they had a Hamar Laser. We had a contest as to who could level / align the same machine faster. They adjusted it out of alignment both times as near to the same each time.

                            I used a King-Way which uses level vials and I won. I use lasers now too on longer machines but use the old methods on small plates, S-E's and machines. Only after ISO BS was pushed onto us that calibration for a rebuilder was needed to trace the measuring tools. Having to have dial indicator or level tested when we use them to compare most of the time and not measure a specific size, etc.

                            I am not a curmudgeon who says there is no place for calibration, but in a Machine tool rebuilder's shop, smll shop owner or home machinists shop that knows how to use a precision level, knows how to check the "Rotation of Points" (hinge) knows about the Kenematic mount principal (3 points) and respects and has knowledge of temperature control all the modern gizmo's are not needed in non ISO shops.
                            Last edited by Richard King; 11-24-2012, 08:58 AM.

                            Comment


                            • #15
                              Originally posted by oldtiffie View Post
                              Accuracy is given as +/- millionths of an inch (ie 1000 miliontht = 0.0001" ie "1 tenth".
                              1000 Millionths = 1 thousandth (a thou.)
                              100 Millionths = 1 ten thousandth (a tenth)

                              Larry S
                              Larry Swearingen
                              Fort Wayne, IN
                              New Hoosier

                              Comment

                              Working...
                              X