Measuring less that 20 degrees of a circle

  • Filter
  • Time
  • Show
Clear All
new posts

  • Measuring less that 20 degrees of a circle

    Is there an amount of error that comes with measuring smaller circumference area on a circle? say like only measuring 15 degrees compared to 360 deg. of the circumference.

  • #2
    Yes, there are many factors involved with how much error and why, but the simplest solution I have found is get a ring gage or pin that is as close to the diameter you are trying to check as possible and then measure the same arc on the artifact. That will show you how accurate or inaccurate your measurement is likely to be.

    For example if I need to check 10 degrees of a 6" diameter, I might grab a calibrated 5.995" ring and measure a 10 degree segment of the ring. How close my results are to 5.995 tells me how accurately I can measure the part, (ball park speaking). HTH (hope this helps)
    sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery


    • #3
      PC-DMIS FOR CMM 4.1


      • #4
        Will You Get A More Accurate Reading The More Points You Take On A Radii? Please Explain
        PC-DMIS FOR CMM 4.1


        • #5
          Yes you will gain accuracy with more points and also if you can measure a cylinder instead of a circle, sometimes that helps. You might try the different algorythems as well. If you open autocircle and press F1 the help menu will open. In the advanced area you have a choice of how the diameter is calculated, least squares, max inscribed, etc. The help menus do a good job of explaining these formulas, how they work and why they work best with more hits.
          sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery


          • #6
            Measuring an arc

            First of all, you really can't measure an arc accurately of 20 deg....

            You can measure points on the arc accurately, individual points.

            You can use cad to pick each point, or use Polar coordinates, which require you to make an origin at the center pt of where the arc is supposed to be.

            See attached spreadsheet for inaccuracy calcs.
            Attached Files
            Jim Jewell


            • #7
              The Engineering Firms I have been associated with have a general rule of thumb that says if there is less that 270° of a circle then they will use profile. While many people may consider that a bit conservative there is a point that measuring a Radius to obtain a circle becomes a bad idea. It is not just the machine / accuracy. The form of the curve also influences the reported diameter and center point. It is not unreasonable to find the center point off 10% of the radius on a very small portion. Grabbing 300 points on 10° of radius is just an exercise in futility as it will certainly bite you at some point.
              I have personally found PC DMIS to be one of the best at figuring out where a circle is and what size it is when only a portion of the circle is available. Still, just as a personal choice and arbitrary assignment, I use Radial Points (emulate Profile) if I have less than 180° of a circle. With machined parts you can probably cut this down to a lesser degree. With formed parts you will want to increase the degree. I check injection molded plastic right now and 180° works for me.
              Bill Jarrells
              A lie can travel half way around the world while the truth is putting on its shoes. - Mark Twain


              • #8
                Wingman, I wish more prints were like that. Have you seen this?


                • #9
                  Nice. Does a real good job of explaining it. I just saved that site. Nice. Thanks
                  Bill Jarrells
                  A lie can travel half way around the world while the truth is putting on its shoes. - Mark Twain


                  • #10
                    That is a good explanation, but 270 degrees? How was that determined? I've typically worked with 90 to 120 degrees, depending on the radius size. I'm more curious as to why than anything. More knowledge for me.

                    Edit- I just looked at the tec-ease site. It gives a good explanation for needing 180 degrees of arc.
                    Last edited by John Kingston; 06-07-2007, 10:37 AM.
                    When in doubt, post code. A second set of eyes might see something you missed.


                    • #11
                      If you only take 4 hits, you are only checking 270º...
                      Links to my utilities for PCDMIS


                      • #12
                        270° provides two diameters at 90° to each other (a criss cross).
                        Bill Jarrells
                        A lie can travel half way around the world while the truth is putting on its shoes. - Mark Twain


                        • #13
                          I was told by tech support that PC-DMIS needs at least 120Deg of an arc to get an accurate center. We use a minimum of 90 here.
                          sigpic GDTPS - 0584


                          • #14

                            Earlier in the week I got some good information from these guys while trying to check a 360" (Yep 30 ft) radius with only 6° of arc available. I set an alignment to the CL of the radius then reported the radial points. I used the average of the radial points in my report. This was an appliance part with ±0.020" tolerance so averaging the points will be sufficient in this instance.

                            LINK TO THREAD

                            In the Basic clas I took the question of checking radii came up. The instructor advised taking many points then, depending on what you need which algorithym to use.

                            If you are looking for location use FIXED_RAD
                            If the radius value is critical use MIN_SEP

                            I have found using an odd number of hits, 5, 7, 9 etc. returns the best results on my machine. I'm guessing the controller read position better when both axes are moving. Just a hunch though, no real proof to offer.


                            Edit: If you happen to be checking a 30 ft radius as well Craiger actually has a 30' radius gage.
                            B&S Mistral
                            3.207 Beta on XP

                            Older'n dirt


                            • #15
                              I replied to that post. We get those all the time and I always move to Center and take radial points and report the MIN and MAX as a Radius. I leave the Center as SET. However, the different calculations thing is nice. Have never considered those for a Radius. JKust the points. Hmm.

                              The number of degrees is dependant on the type of material and process you are using. Even 120 won't work for all material and all occasions and the Tech was making some assumptions unless he knew the material, processes, size, point density, tolerance etc. Just a seat of the pants belief. The shape / form of the Radius can and will throw off your measurement. With machined parts this is not as big an issue as it is with formed parts or some plastic parts. I would no way emphatically trust a 90° or even a 120° radius to give me an accurate center and size with some of the Radius we get on some of the parts we do here. Would not work. Other Radius measurements are OK with 90° on our parts although I try to use a RAD Gage when that is all I have. Depends on the circumstance.
                              We make fan shrouds for automotive and some of the truck shrouds are huge and flimsy. Radius in many instances are mis-shapen and would certainly report incorrectly even when every point on their surface falls within the required zone.
                              If 90 works for you GREAT, you are lucky
                              Like I said, We arbitrarily set it at 180 here. Less than 180 and we use Radial Points MIN MAX (emulate Profile). I add in the location tolerance to the radial tolerance as well to correctly define where the surface is actually allowed to be. After reading the article in posted earlier I might internally set it to 270.
                              Last edited by Wingman; 06-07-2007, 02:53 PM. Reason: Left out
                              Bill Jarrells
                              A lie can travel half way around the world while the truth is putting on its shoes. - Mark Twain


                              Related Topics


                              • jjager
                                Measuring ring gauges
                                by jjager
                                Hello all,
                                Couple days ago my boss asked if we could measure ring gauges on our CMM.. After the last couple days playing around with it I cant...
                                01-09-2014, 12:33 PM
                              • ADAM WALLACE
                                Minuimum circumscribed circle
                                by ADAM WALLACE
                                I'm trying to measure a outside dia. circle using min_circ. I only have 170 degree of arc I can measure but at the end of measurement I get a best fit...
                                11-16-2017, 12:07 PM
                              • infinisys
                                31.00 In Ring
                                by infinisys
                                I have to measure this ring with 90 deg. CMM can go only 28.00in i am @ A90
                                for all my hits so i can touch the ring with probe what is easy way...
                                09-19-2012, 10:56 PM
                              • PCDEMON
                                diameter errors
                                by PCDEMON
                                I am new to PCDMIS and have a measurement error I can not explain or correct. Im a beginner at this so any help would be greatly appreciated.
                                12-18-2008, 03:23 PM
                              • ClayOgre
                                set ring?
                                by ClayOgre
                                I seem to remember someone on here mentioning using a set ring gage for probe qualifications. If this is the case...why would one do this and how is...
                                02-26-2008, 09:21 AM