Irregular feature to report

[Troy Tatters]

Attached screen capture shows a sheet metal feature that I need to report the feature center location on. The biggest issue is that the feature is shorter than the radius values on each end.

Any ideas how to measure it so I can report it?

[Bill **********]

Quick and dirty. Circle in both end features and a mid point between them. You will probably need to find the circles in polar mode to have any kind of accuracy.

Bill

[Troy Tatters]

That's what I normally do, but in this case, the radius is longer than the feature, so I can't execute 2 circles without interference. (The arc center is past the end of the opening)

[MrComment]

If your using Auto Circles to find the 2 radii make sure that you have find hole and Move options turned off on both features, and the initial and perm hits are going to have to be set to 0. This means you should level to this feature by measuring the workplane of this feature in a seperate step if you havent already. This should avoid any unnecessary moves to the center of each radius.

If all else fails another approach would be to take Auto vector points inside of each radii (at least 3 each). follow up by constructing each radius off of those features.

Once you know the center point of the resulting circles you can always find the midpoint between them, its also good practice to verify that your in the correct workplane too.

Good luck

[cmmguy]

Attached screen capture shows a sheet metal feature that I need to report the feature center location on. The biggest issue is that the feature is shorter than the radius values on each end.

Any ideas how to measure it so I can report it?

What about the two bent down sides?? They look like an important part of the feature. Shouldnt they be included in measurement also? What are their functions?

[Troy Tatters]

I don't think this is what it was intended for, but it works. I used the auto square slot to the outside trim (tangent to tabs) then manually offset the edges to the nominal values for the formed tabs and reduced the radius size and increased the depth (I'm using a shank) so the tab surfaces are hit true.

I'm still open for a better way to do this. I need X,Y and Z as this detail is part of an assembly and out in space.

[Troy Tatters]

The detail is just 1 small part of a prototype instrument panel assy. The detail itself has been inspected on a fixture. I need to report only the position of the feature as part of the assy. No fixture is available at this time

[cmmguy]

In that case then the last method you described would work, or

Measure the sides as lines, find the midline and then take a single point on each end that is coincident with that mid line. Keep it simple. Use a little extra prehit for the end points.

[Matthew D. Hoedeman]

In that case then the last method you described would work, or

Measure the sides as lines, find the midline and then take a single point on each end that is coincident with that mid line. Keep it simple. Use a little extra prehit for the end points.

IF you knew the nominals, you could type in the info for a square slot and let Pcdmis do it for you, use the WIDTH measure, then it will do the lines on each side and the 'mid-points' on the ends.

[JamesMannes]

If there are any further questions on this, could we see a snapshot of the print? That would help a lot.