radius size with small arc sections

[David Kick]

Does anyone know an easy method to get accurate measurements on radius size with small arc sections? It seems like anything less than a full circle gets varying repeatability. I am trying to measure a radius of 2mm at a 90° arc sweep.
Thanks for this Forum.
David Kick

[Wingman]

Profile.
Set Origin to Center and take Radial Point hits. Reprt the Radial Point.

[bw_bob]

What the wingman said....I do it all the time.

[Underspec]

How do you set origin to center if you don't have a location?

[bw_bob]

How do you set origin to center if you don't have a location?

You have to have the coordinates before this works of course.....

[Wingman]

How do you set origin to center if you don't have a location?

Get the location from the print.

[Underspec]

I should of been more clear. If your print has no location for this radius, then bascially you're out of luck?

[bw_bob]

Get the location from the print.

There are alot of times the radius is not controlled by a FCF for profile & therefore not dimensioned (just a radius callout). In this case I call the engineer and tell him if he wants it checked I need the coordinates

[Wingman]

There are alot of times the radius is not controlled by a FCF for profile & therefore not dimensioned (just a radius callout). In this case I call the engineer and tell him if he wants it checked I need the coordinates

Radius Gage
If no Rad Gage then you need a location for it.
Can you figure out where it is from the adjacent surfaces?

[Rob_K]

Re-measure sets origin; before that you must set circle so close that machine gets that 90°; more points will help, too

[Wingman]

Re-measure sets origin; before that you must set circle so close that machine gets that 90°; more points will help, too

Re-meas will use the feature to set origin. If feature is not correct then origin is not correct. More points can help and hurt - I have seen it go both ways. With a small arc you need to do a profile. Most of the time they are trying to control the surface for clearance. All you need to do is say where the surface IS compared to where it SHOULD be. The actual calculation of the radius is NOT a good indication of whether this is true or not when you are sampling a small section of an arc. In many cases CMM cannot tell you where the radius is or how big it is. It CAN tell you where the surface is at though.

[Rob_K]

Wingman is right with a profile, except if there MUST be a circle as output, no matter of uglynes of a surface(been there)

[Wingman]

Wingman is right with a profile, except if there MUST be a circle as output, no matter of uglynes of a surface(been there)

I don't agree. If you use Radial Points and report the Radius as MIN and MAX they will never, ever, ever know the difference.

If pushed I tell them I can give them an unreliable, misleading load of crap or I can tell them where the surface is at. Take your pick.

[David Kick]

Thanks for the responses. I will let my boss read this and see if he wants to invest the time to measure this.
David Kick

[Matthew D. Hoedeman]

Well, something to try, if all he needs is the radius value:

1) Measure 2 lines, one on either side of the radius
2) Intersec the lines
3) Make an alignment, using one of those lines as the rotate and set the origin to the intersection point.

This will let the part move all over the place (within resaon, of course) and give him a brand-spanking new origin that SHOULD be in perfect realtionship to the location of the radius. Then, simply program the section of radius so that the hits are good and go to town.

Another option:
Measure the lines as before, and intersect them for a point. Measure an AUTOFETURE circle that is a RELATIVE MEASURE feature to that intersection point. Same results without the need for the alignment.