Capable True Postion

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

  • Capable True Postion

    *disclaimer* have read the previous huge discussion on this subject that basically says cPK on TruPos is ok for dimensional reference but stats are no good for process improvement. i think thats what i got out of reading the discussion.

    Customer sends us this report and says we need to get the X deviations closer to zero to meet "capability"?? we have moved the punch and button in the die twice already to try to get the stupid hole dead nuts zero for them the True Position has never been out of spec.

    the hole is a clearance hole for a welded nut. according to their project manager they are using the hole to locate the nut on thier weld fixture, wich IMHO is wrong to stack-up the tolerance like that, they should use a pin that is postioned to zero to loacate the nut.

    8PC SAMPLE:
    DEVIATIONS in X
    CIR1-BX CMMTRUE POSITION OF CIRCLE CIR1 UNITS=MM
    -0.257 -0.248 -0.269 -0.253 -0.258 -0.256 -0.251 -0.271

    DEVIATIONS in Y
    CIR1-BY CMMTRUE POSITION OF CIRCLE CIR1 UNITS=MM
    -0.051 -0.040 -0.046 -0.061 -0.017 -0.036 -0.008 -0.058

    TRUE POSITION TOLERANCE = 0.5mm (they are also giving 0.433mm of bonus from hole size plus dat-b- dat-c- max mat'l cond.)

    what i am asking for is a way to explain to the customer (in laymans terms) why they shouldn't use capability on TruPos. TIA

  • #2
    True Position does NOT take into account the DIRECTION of deviation. For any hole, there are at minimum 4 directions of possible deviation (lets say, X-, X+, Y- and Y+).
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      Originally posted by Matthew D. Hoedeman View Post
      True Position does NOT take into account the DIRECTION of deviation. For any hole, there are at minimum 4 directions of possible deviation (lets say, X-, X+, Y- and Y+).
      not picking up what you're putt'in down??? can u explain why, in simpler terms.

      how do i tell the customer that thier cape study on the True Postion of this hole is BS. we are about to move this punch and button for the 3rd friking time. thnx again.

      Comment


      • #4
        Originally posted by rupture View Post
        not picking up what you're putt'in down??? can u explain why, in simpler terms.

        how do i tell the customer that thier cape study on the True Postion of this hole is BS. we are about to move this punch and button for the 3rd friking time. thnx again.
        Direction of deviation, pretty simple:

        Check #1: Xdev -1.0 Ydev 0.0 TP = 2.0
        Check #2: Xdev +1.0 Ydev 0.0 TP = 2.0
        Check #3: Xdev 0.0 Ydev -1.0 TP = 2.0
        Check #4: Xdev 0.0 Ydev +1.0 TP = 2.0

        Each of these will give a TP of exactly 2.0 and will give a range of ZERO, TP will show the hole as NOT HAVING MOVED at all when in reality, it moved 2mm in X and 2mm in Y. The first hole was off to the "LEFT", the second hole was off to the "RIGHT" the third hole was off to the "FRONT" and the forth hole was off to the "REAR" yet the TP will show that the hole HAS NOT MOVED and that all 4 checked exactly the same. The DIRECTION of the deviation does NOT get included in the TP deviation so it is not included in the SPC.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #5
          We use the components for true position capability. Just reverse the math to obtain your tolerances

          Comment


          • #6
            The real issue is that the customer is using a clearance hole to locate the weldnut. They are relying on you to locate the clearance hole to the weld nut tolerance instead of the weld fixture. Tisk, Tisk, Tisk.

            You are correct in saying that the nut should be held in the weld fixture.

            Good explanation Matt.

            If the deviation is verified to be all in the same direction for every part then Cpk can be calculated correctly.

            What Cpk value are you obtaining?
            sigpic GDTPS - 0584

            Comment


            • #7
              they are telling us to get the hole closer to zero because it is not meeting Cpk, all of the deviations are in the same direction and are consistent, so i guess what i am asking for is an explanation of how they are using miss-using Capabilty in this case. correct me if im wrong but i think their cape study is [email protected] here is what they sent us.

              also should the bonus tolerance be added to the Cape calculation??
              Attached Files
              Last edited by rupture; 07-09-2007, 12:53 PM.

              Comment


              • #8
                Originally posted by rupture View Post
                they are telling us to get the hole closer to zero because it is not meeting Cpk, all of the deviations are in the same direction and are consistent, so i guess what i am asking for is an explanation of how they are using miss-using Capabilty in this case. correct me if im wrong but i think their cape study is [email protected] here is what they sent us.

                also should the bonus tolerance be added to the Cape calculation??

                If it gets bonus then it should be added...it's part of the tolerance
                sigpic.....Its called golf because all the other 4 letter words were taken

                Comment


                • #9
                  You CAN do a cap study of the "X" deviations OR the "Y" deviations.
                  These must be calculated separately.
                  Lately, it occurs to me
                  What a long, strange trip it's been.

                  2017 R1 (Offline programming)

                  Comment


                  • #10
                    Originally posted by John Riggins View Post
                    You CAN do a cap study of the "X" deviations OR the "Y" deviations.
                    These must be calculated separately.
                    how do you split up the TP tolerance for X and Y?

                    Comment


                    • #11
                      What version are you running? Can you please fill out your User CP profile. That would help us better try to answer your questions.

                      Let me get Hoedeman going again here. Up to version V4.0, PC-DMIS can NOT take MMC on the datums into account. That means they are always put at RFS. In V4.1 it worked, mostly. In V4.2 they seem to have it correct. So PC-DMIS will "wiggle" the datum structure to minimize TP error based on the VC that exists on your datums (be sure to read the ASME Y14.5 section on simultaneous requirements though).

                      So what does this mean? If you switch to 4.2 and use the new style GD&T (XactMeasure as they call it), you can use the virtual conditions on your datums and you will instantaneously see your TP go to (close to) zero. And it is perfectly legit too! Your customer is wrong in trying to have you evaluate the FCF to RFS. You are allowed to use the VC on the datums so use it to your advantage.


                      Jan.
                      ***************************
                      PC-DMIS/NC 2010MR3; 15 December 2010; running on 18 machine tools.
                      Romer Infinite; PC-DMIS 2010 MR3; 15 December 2010.

                      Comment


                      • #12
                        Originally posted by Jan d. View Post
                        What version are you running? Can you please fill out your User CP profile. That would help us better try to answer your questions.

                        Let me get Hoedeman going again here. Up to version V4.0, PC-DMIS can NOT take MMC on the datums into account. That means they are always put at RFS. In V4.1 it worked, mostly. In V4.2 they seem to have it correct. So PC-DMIS will "wiggle" the datum structure to minimize TP error based on the VC that exists on your datums (be sure to read the ASME Y14.5 section on simultaneous requirements though).

                        So what does this mean? If you switch to 4.2 and use the new style GD&T (XactMeasure as they call it), you can use the virtual conditions on your datums and you will instantaneously see your TP go to (close to) zero. And it is perfectly legit too! Your customer is wrong in trying to have you evaluate the FCF to RFS. You are allowed to use the VC on the datums so use it to your advantage.


                        Jan.
                        OK, the 4-head wants to get me going. Here's an interesting bit of info for you:

                        BEFORE they change the diameter axis in Pcdmis from D to DF, Datapage WOULD included the MMC for the feature. When they changed Pcdmis to DF, they didn't change Datapage so Datapage no longer included the diameter of the feature and no longer added the bonus to the tolerance. Sweet! Sort of like, "Gee, I am a genitic engineer, I'm going to add a thumb to the right hand of right-handed people and a thumb to the left hand of left-handed people, but we will continue to make gloves with 4 fingers and a thumb."
                        sigpic
                        Originally posted by AndersI
                        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                        Comment


                        • #13
                          Here we go again with this business about not being able to use CAP on TP... It is legit.
                          You can use the components, if you want, but why limit yourself when TP gives you more manufacturing room??
                          Isnt that the whole purpose of TP anyway?
                          Links to my utilities for PCDMIS

                          Comment


                          • #14
                            Originally posted by Jan d. View Post
                            What version are you running? Can you please fill out your User CP profile. That would help us better try to answer your questions.

                            Let me get Hoedeman going again here. Up to version V4.0, PC-DMIS can NOT take MMC on the datums into account. That means they are always put at RFS. In V4.1 it worked, mostly. In V4.2 they seem to have it correct. So PC-DMIS will "wiggle" the datum structure to minimize TP error based on the VC that exists on your datums (be sure to read the ASME Y14.5 section on simultaneous requirements though).

                            So what does this mean? If you switch to 4.2 and use the new style GD&T (XactMeasure as they call it), you can use the virtual conditions on your datums and you will instantaneously see your TP go to (close to) zero. And it is perfectly legit too! Your customer is wrong in trying to have you evaluate the FCF to RFS. You are allowed to use the VC on the datums so use it to your advantage.



                            Jan.
                            i would rather use an outside program than pc-dmis for my TP calculations, that way i can use the holes sizes i get from pin gage measurements for bonus rather than trusting the CMM to hit the high points of the shear.

                            Comment


                            • #15
                              Not to throw gas on the fire...but have you looked at the 'raw' stock you are shipping...my experiance has shown that most customers don't really know how to measure things in general...

                              Having said all that, I agree with the consensus here - TP cap studies are a waste b/c the direction is not included...
                              "Perfection is not attainable, but if we chase perfection, we can catch excellence" - Vince Lombardi

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X