"form" of geometry best fit

[[Mi...]]

Is there an easy way to calculate how good a geometry best fit is? Like, report the max deviation from the resulting coordinate system of all the features the system is being fit to?

The way I do it now is to add a position characteristic to each of the features being used, and report the max of those, but that's a little tedious.

[[Jo...]]

Isn't this really what profile does ? size, form, position ?

[[Ri...]]

The Default Printout has a lot of useful information.

[[Mi...]]

Please sign in to view this quote.

Perhaps, but I'm not sure I can do it with the licenses we have. I have 8 separate plane features, and I'm trying to get a handle on how good their positions wrt each other are, if that makes sense. I think if I had an FFS license I could combine them into a single surface and evaluate form, but I do not. Geometry best-fit is the alternative I came up with, but maybe there's another way.

Please sign in to view this quote.

Thanks, I'll check it out.

[[Er...]]

What exactly are you asking for? The shift/rotation values? Because the fit is as good as it can be, (minimizing the sum of squared error) with a gaussian fit.

Sounds like you can do a TP of all the planes to your bf geometry coordinate system. (you might need to use the "special" function to rotate and/or offset for every plane if its "3-dimensional".) That will give you the positional deviation of the planes with the "error" distributed "evenly" over all planes.

The form error of the plane doesnt change, it is what ever it is. Unless its a profile tolerance callout to some drf. But then its not only form, Its also location and/or orientation involved. And even then, the actual form error is still the same.

[[Mi...]]

Please sign in to view this quote.

Basically, I need the form if it were evaluated as one continuous surface. Like I said, I think free-form surfaces were meant to accomplish this, but I can't use that.

Please sign in to view this quote.

Sounds like what I'm doing now, which is reassuring. Thanks.

[[Mi...]]

Alright, so I did both the above suggestions: check the compact printout, and create position characteristics for each of the features and check the worst one. The problem is, they disagree.

According to the compact printout, the geometry best fit alignment has form of 0.0015". Of the 9 position callouts I created (8 planes and one diameter, all the features defining the alignment), the worst is .0006".

Any ideas why?

[[Jo...]]

Please sign in to view this quote.

check evaluation settings/filtering

[[Mi...]]

Please sign in to view this quote.

I have everything filtered at the feature level, so I wouldn't think that would lead to the difference. Could it be that the alignment is "reaching back" to the raw data by default? I don't really know how to check, sorry.

EDIT: Just checked and it looks like they're all evaluated the same way. I'll probably just go with the Compact Printout output. It seems more likely to be correct.

[[Er...]]

The form in the compact printout is the form error from a perfect geometrical shape. As in distance between highest and lowest point in the plane. Its not comparable to either TP or profile of a line/plane. Since that also inculdes location and/or orientation to the other planes in your case. Its also the biggest deviation (max or min) times two.

So dont mix those up.