Nominal in dimensions changes iself

[derrickcoyne]

I have a cylinder that we're trying to dial in, and our machining tech showed me a report yesterday where the nominal was different than what was expected. after doing a bit of digging, it appears that something causes this nominal to change in the reporting part of the program. this program was run several times back to back and some of the time the nominal is the correct number, others it changes. I set the program to read-only for the off shifts last night, but it happened anyway. I inserted a new dimension that is set up exactly the same, and it also changes. The nominal of the actual cylinder does not change, only the dimension. Any thoughts on what could cause this and how to stop it? my SPC data is completely hosed by this. Thanks in advance.

[RIDER]

Depending on which end of the cylinder is the "worst" the nominal is going to change. The end point of the cylinder is not in the same y location as the start point.

[Rich P]

What rider stated

Check your drawing, the auto cylinder is at a 55 deg (35 deg) angle? typically when I have had features like this there is an intersection point defined on drawing that has Basic dimension from and will report worse condition along the axis and as stated that location changes along the basic for an angled feature.
Construct an intersection that matches the Basic(s) and define the True Position from that, you may also need to figure out some form control on a feature like this depending what is going on and process to manufacture feature.

[NinjaBadger]

It's not always possible (if you're not working to a print, or it's a rubbish drawing) but typically I'll rotate my co-ord system so I'm looking square on the the hole, that way you can easily tell the machinist which way it needs moving.

Of course the inclined cylinder might be the issue in this case (and should be the first thing you look at), but there are other factors which can cause nominals to change (incomplete alignments and incorrect feature theoretical elsewhere in the program) but it's more likely to be what RIDER and RichP have said.

[vpt.se]

I believe that what you are seeing is the coordinates to your cylinders END point (calculated by PC-DMIS using the vector and length of the cylinder). The START point is defined in the autofeature dialog.

If you need these values to be consistent, create circles (intersection plane/cylinder) at either the start or end of the cylinder (or both) and do a TP for the circles instead of the cylinder.

[louisd]

Is your current alignment matching your xactmeasure ABC datum structure, when that feature is measured?
If not, your theo nominal on the feature can -and commonly will be- different from the nominals as prescribed by your datum structure in xact.

If you want to lock down a value, you can add a +0 to the theo values of the feature (this is a bandaid). I don't think you can use this method within xactmeasure's nominal values however.

[KChristian]

Did you turn on the void detection toggle?
Sometimes this turns on by its self and really messes up the feature measurement. Try toggling it off and see what happens.