Simulating Measuring a Chamfer with a Ball Bearing

[ifoundgoldbug]

Greetings all,

I have a part that was hand drafted in the 1960's and in order to measure a chamfer they drop the part into a fixture with clearance below the part to accommodate a ball bearing placed into the chamfer finally they measure from the top of the ball bearing to the bottom of the part. This is the only dimension i have controlling this chamfer other than the degrees. For clarity the bottom of the ball bearing drops below the bottom surface of the part. If this were solidworks I could just mate a sphere to the inside surface of the chamfer and measure from there but I am uncertain how to replicate that functionality in PC-DMIS or if such behavior is even possible.

I also attempted to solve with linear algebra but there are at least 3 different variables that I would have to consider (part thickness, chamfer angle, & chamfer diameter) and that was more measurement error than I am comfortable with. Also given the age of this part I will not get a drawing update from my customer.

Any help or advice would be greatly appreciated Thank you very much for your time.

[Douglas]

I had to do this on the end cap for a ball valve in the past, that top internal chamfer was a seal surface and had a .002" tolerance on diameter, for me it was the top of the part I used not the bottom is the only difference.

[gibsonridge]

have you tried measuring a cone and constructing a diameter at the cone per the depth required on your vintage print ?

[Douglas]

can you probe a line on the chamfer for the angle? can you probe the chamker at all? you are making it sound like it might be a chamfer where 2 hole sizes meet... a step drill, that would be hard to probe

[ifoundgoldbug]

Douglas While I can probe a line and get the angle. This is not a instance where 2 holes meet it is something like | \O/ | where the O is the ball bearing and the \ / is the chamfer depth i am trying to measure. The drawing only gives me the dimension from the top of tht ball O to the bottom of the part.

[Douglas]

verify each individually... probe the line and you got the angle. Code is not the best practice in this case is my opinion, when I faced the same thing on a drawing they called it "QC CHECK" on the drawing so that is how I checked it. I assume you are probing the bottom of the part elsewhere in the program? I would put a ball on the part just like the drawing shows and probe it like a sphere to get its center, then do a distance with the radius compensated....done

[Douglas]

Are you able to probe the chamfer directly? or are you able to only probe the ball bearing?

It sounds like on this drawing only the chamfer angle, and ball loaction from bottom of part are dimensioned, why would you care about cone diameter at all?
Drop the same ball in the part and measure it the way the drawing asks.... that gives you location and you only have chamfer angle to measure

[RandomJerk]

Meas cone, I'm calling it CONE1 in this aircode.

Assign/CONE_ANG==CONE1.A/2
ASSIGN/GBALL==(Your gage ball radius)
ASSIGN/3ANG==90-CONE_ANG
ASSIGN/BRAD=COS(DEG2RAD(CONE_ANG))*GBALL
ASSIGN/3LEG=SQRT(GBALL^2-BRAD^2)

CIR1 =FEAT/CIRCLE,CARTESIAN,OUT,NO
THEO/<-0.1386,0.08,-0.1411>,<0,0,1>,0.7547
ACTL/<-0.1386,0.08,-0.1411>,<0,0,1>,0.7547
CONSTR/CIRCLE,CONE,CONE1,DIAMETER,[SIZE=16px][B]BRAD*2[/B][/SIZE]

A1 =ALIGNMENT/START,RECALL:STARTUP,LIST=YES
ALIGNMENT/TRANS,XAXIS,CIR1
ALIGNMENT/TRANS,YAXIS,CIR1
ALIGNMENT/TRANS,ZAXIS,CIR1
ALIGNMENT/END

ASSIGN/ZOFF=3LEG-GBALL

BALL_BOTTOM=GENERIC/POINT,DEPENDENT,CARTESIAN,$
NOM/XYZ,<0,0,0>,$
MEAS/XYZ,<0,0,ZOFF>,$
NOM/IJK,<0,0,1>,$
MEAS/IJK,<0,0,1>

[JEFMAN]

Some threads about it here :
https://www.pcdmisforum.com/forum/pc...-inside-a-cone
https://www.pcdmisforum.com/forum/pc...les-from-cones


Or searche cone + sphere...

[Douglas]

out of curiosity what is the bevel angle? Smaller cone angles will introduce more error through constructions... for example if it is a 10° taper you'd never construct a diameter accurately as you might with a 45° chamfer... if it is less angle like that the ball check is holding a fairly tight diameter tolerance


Edit: also what is the length of the chamfer? and the tolerance on ball height? with a 45° and enough length you could probably probe the cone directly and code it... with a smaller cone angle and shorter bevel good luck with that, use the ball

[AndersI]

PC-DMIS can construct a circle with a given diameter, tangent to two lines - this is the 2D equivalent of dropping a sphere into a cone. See for example https://www.pcdmisforum.com/forum/pc...520#post435520

[gibsonridge]

this is where i was trying to go , the sphere diameter and angle should give most of your info

[Douglas]

I've done that exactly in place of pins being used to check splines and it worked well... but in this case for small cone angles and short bevel length, very slight diameter errors would move that ball quite a bit along the cone axis. You might be able to program it and probe it but it may be functionally more accurate to probe the ball sitting on the chamfer, and measure it the way the drawing shows it.This would be my primary consideration, if it should be programmed rather than how to program it