Radius Not Enough % of a Circle

FYI, this is a 50° radius. What do you find is the minimum degrees for accurate/repeatable results? I'm finding about 60° or more is OK.

Do you happen to have Radius gages in your facility? You could quickly verify with them that it is a radius of .4. I know there is also the options to measure a Fixed Radius Circle in PC-DMIS, but I haven't messed with it. 

https://docs.hexagonmi.com/pcdmis/2023.2/en/helpcenter/mergedProjects/core/16_const_topics/Best_Fit_Type_for_Circle.htm

You could also try getting the arc / chord length. I've seen people post about arc length calculations for circles less than 180°. Then calculate it into a Radius.

https://www.omnicalculator.com/math/arc-length

I attached a pdf someone once shared about a small arc.

PDF

Have you explored any of this stuff? 
https://nexus.hexagon.com/community/public/pc-dmis/f/off-topic-shop-talk/129440/small-arc-segment-deviation-graphic/736723

Also what tolerance are you trying to hold? Do you have other tools that can measure it, maybe key in the dimension? 

And of course our links we share come full circle(unlike this guy's radius)

Haha! 
I miss having access to a good vision system for all those "too small radius" parts!
Hot wax sticks and lots of practice always won in the end and we had good parts. 

Haha, same as the last thread. True, maybe in this example I just use a radius gage They are trying to hold ±.005 tolerance on a 50° radius though.

I'll definitely have to look into this "fixed radius" thing! Sometimes they put locational dimensions on these 50° radii. It sounds like the "fixed radius" might help for that?

One time there was a ±.001 tolerance on both the size and location...

We actually have a Tesa Visio 300 DCC that we hardly ever turn on anymore. Since the Vision systems I believe are 32bit, and we couldn't go beyond Windows 7 at that time before I worked here, so now it's an off-network PC with a "camera" that uses a HASP. Poor Visio.

Yes I think Fixed Radius would help with the locational dimension. Since in the help file it states: "Only the circle's position is allowed to vary."

I noticed that too! That PDF is lengthy though. I wonder what happens if I simply change my autocircle to fixed radius.

I'll test it out on my program that has radius sizes and locations. I guess it can't hurt...the dimensions are already "inaccurate" according to my report comment. I'll just change it to a fixed radius and do a few test runs and maybe measure the part using a comparator and a radius overlay chart.