hexagon logo

Error when creating a constructed width

Hello, I am fairly new to the inspection field and am having some trouble with setting up my datums for a part program.

There are several slots cut out on a rectangular plate, and the distance between the edges of the slots are the datums. So, in an attempt to create a constructed width to use as my datums, I created lines by taking hits along the edge and selected the option for them to projected onto a feature (a plane along the top face of the plate). I have a picture from the program shown below:



As you can see, the first constructed width was created without a hitch, but the for the second one (from Lin_A1_2 to Lin_A1_4), PC-DMIS gives the error "Surface vectors are not opposed" and refuses to make the feature. I'm not really sure what this means and am hoping someone can help me through this.

If it helps, here are the directional vectors for each of the four lines:
Lin_A1_1:
I: 0.002825
J: -0.9999959
K: -0.000436
Lin_A1_2:
I: -0.9999944
J: -0.0031568
K: -0.0011256
Lin_A1_3:
I: 0.0014157
J: 0.9999989
K: 0.0004408
Lin_A1_4:
I: 0.9999808
J: -0.0060896
K: 0.0011215

Only the holes in the picture were machined, the datum features were torch cut and the top of the plate was not faced off (i.e. it was rusted over). I know that's not ideal for making datums, but it was how the print said to do it, and the holes were cut after being lined up along those edges. I'm hoping that the error has nothing to do with that fact.
Parents Reply Children
  • I used a laser tracker for those measurements, not an ARM, so I'm not sure if/how the hits would have directional vectors... Would the feature set of the two lines behave the same as a constructed width for GD&T purposes? Thank you for your help.

    Edit: I checked the hit target values for the lines, and you were right about the vectors being an issue. I'll try to explore this more on my own.

    Edit #2: I was able to fix the issue for the lines in the photo, but fixing the problem for the whole part will be quite tedious. Is there any way to control or predict what the hit vectors will be during initial measurements?