Flatness on glass windows

  • Filter
  • Time
  • Show
Clear All
new posts

  • Flatness on glass windows

    I have to check the flatness on a glass window. It is about 30 in x 19in square. Here is the spec for it.

    Flat safety glass of the previous types may have 0.8 mm (0.03 in) maximum total bow per each lineal 305mm (12 in) and each part may have a maximum overall bow in millimeters of 0.03 times the length of the part in millimeters ( 0.03 times the length of the part in feet).

    This is how I was going to check it. Take 4 points as a plane 2 on one edge for a line and then 2 points on the other edge for another line. Make my alignment then take a lot of points on the glass and create a plane and give it a flatness of .8. Has anybody had to do flatness on glass?

    I am running 3.5mr2 on a 7-10-7 with a TP 2. I do not have a scanning head for this.

    Thanks for the help

  • #2
    can the bow be both Above and Below the plane or is bow only in one direction? cause flatness wont give you bow.
    Links to my utilities for PCDMIS


    • #3
      I am not sure on the direction of the bow. I was also thinking of taking some points along the edge and giving it a tolerence of +/-.4. To me .8 is alot of tolerence and I would think you would be able to see that if you put it on a surface plate.


      • #4
        is that .8 inches or mm. It seems to me that .8inches is WAAAAYY too much, but you converted everything into inches so I wasn't sure. If it were me I would almost make a grid and check the part at .5in intervals with lines 4 inches apart and evaluate the bow in your length axis and your width axis. From what you said your tolerance is length dependant, thats why I would evaluate the bow in each axis and not overall flatness.
        Saving the world, one bad part at a time.


        • #5
          cmmmike....i used work for a window company & had to do this all the time...i used to have the glass fixtured up on its side to prevent sagging. Then i would do a patch scan(dont need scanning head) to cover the entire surface>>construct your plane>apply flatness dimension>graphically display your points...this way you can see where the glass is warped..Luckily you are dealing with small pieces of glass! Our lineals were 8ft long 4ft wide & 5mm thick... pain in the butt they were...out on the floor we would actualy stretch fishing line from one end to the other & measure the largest gap with a steel rule....it was pretty high tech stuff!!

          Oh yeah.... would report flatness in two workplanes
          Last edited by spazus_maximus; 09-25-2007, 11:21 AM.


          • #6
            Sorry about that. It is .8mm I will give the grid idea a try.


            • #7
              Spazus_maximus this is what I have done so far and you can tell me if I am correct.

              I put a piece of cardboard on top of the CMM table.
              Layed the glass on the CMM.
              Did a basic alignment of the glass. (Plane, Line, Point).
              Set up a scan with a grid of 1 in by 1 in.
              Created a plane from the scan.
              Put in a flatness dim with 0.03 for the spec. (I am doing this in inches)
              The print out a display of the scan which shows my points that are in and out.
              Does this sound about right?

              Thanks for the help.


              • #8
                Sounds good....but i would definately try to measure the glass perpendicular to the table with a A90B90.... that is if it is to be measured in its free state. By laying the glass flat on the table you may be taking the bow out & would actually be checking surface flatness.


                Related Topics