Measuring the point at which a bore meets a countersink.

[ChairsWithWheels]

So right now I have a part that is a cylinder with about 5 levels of bores leading into each other with a countersink.
I'm looking to find the best way to measure the bottom of these bores before they lead into the countersink.

All I can think is constructing a line and then a point where the edge of the bore and line intersect, then reporting a distance from there. But was wondering if there might be a better way.

[JEFMAN]

Could that help ?
https://www.pcdmisforum.com/forum/pc...inks-and-cones
https://www.pcdmisforum.com/forum/pc...-with-2-points

[InspectorJester]

The countersink I would measure as a cone.
I would measure the bore a cylinder or multiple circles.
I would then create an intersect circle, which should be at the "top" of the countersink (if I'm interpreting that correctly).

[Jim Poehler]

I think you mean that the intersecting circle would be at the "bottom" of the countersink. The countersink (cone) and surface plane would be the "top".

[InspectorJester]

It depends on the orientation of the bore I think, but yes you're right

[RandomJerk]

Constructed circle from cone.

MEASURED_BORE=FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,OUT,LEAST_SQR
            THEO/<0,0,-0.015>,<0,0,1>,0.575,0
            ACTL/<0,0,-0.015>,<0,0,1>,0.575,0
            TARG/<0,0,-0.015>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=YES
              NUMHITS=5,DEPTH=0,PITCH=0
              SAMPLE METHOD=SAMPLE_HITS
              SAMPLE HITS=0,SPACER=0
              AVOIDANCE MOVE=NO,DISTANCE=0.3937
              FIND HOLE=DISABLED,ONERROR=NO,READ POS=NO
            SHOW HITS=NO
THIS_IS_CSINK=FEAT/CONTACT/CONE/DEFAULT,CARTESIAN,OUT
            THEO/<0,0,-0.015>,<0,0,1>,9.2543,15,0.575
            ACTL/<0,0,-0.015>,<0,0,1>,9.2543,15,0.575
            TARG/<0,0,-0.015>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=YES
              NUMHITS=5,NUMLEVELS=3,DEPTH=0.0787,END OFFSET=0.0787
              SAMPLE METHOD=SAMPLE_HITS
              SAMPLE HITS=1,SPACER=0
              AVOIDANCE MOVE=NO,DISTANCE=0.3937
              ONERROR=NO,READ POS=NO
            SHOW HITS=NO
THIS_IS_THE_DEPTH=FEAT/CIRCLE,CARTESIAN,IN,YES
            THEO/<0,0,-0.015+0>,<0,0,1>,.575+0
            ACTL/<0,0,-0.015>,<0,0,1>,0.575
            CONSTR/CIRCLE,CONE,THIS_IS_CSINK,DIAMETER,MEASURED_BORE.D

[SingularitY]

Is this method preferable over constructing a circle from the intersection on a cone and cylinder?

[michaelwildschutz]

no not really

[N8VES]

You can't just intersect a cone with a cylinder (or plane) and expect to get a circle. The ONLY time you can do that is offline in CAD. PCDMIS knows that such an intersection is actually an ellipsoid and will generate an error online. What you can do is CONSTRUCT a circle from the cone at the measured diameter of the cylinder. There are parameters within the construct circle window that lets you set them. After you create it, you can edit that size by making it extract the size directly from the measured cylinder.

[Jim Poehler]

"You can't just intersect a cone with a cylinder (or plane) and expect to get a circle. The ONLY time you can do that is offline in CAD"

This statement is absolutely false. It is one of the more common constructions I need to do to meet print specs. It works both off line and on line;.

[Mike Ruff]

You can't just intersect a cone with a cylinder (or plane) and expect to get a circle. The ONLY time you can do that is offline in CAD. PCDMIS knows that such an intersection is actually an ellipsoid and will generate an error online. What you can do is CONSTRUCT a circle from the cone at the measured diameter of the cylinder. There are parameters within the construct circle window that lets you set them. After you create it, you can edit that size by making it extract the size directly from the measured cylinder.

I completely agree with Jim Poehler . As long as the THEOs of the cone and cylinder are concentric, PC-DMIS will have absolutely no problem with the basic construction of a circle from a cone and a cylinder. I could trig it out on paper, the math is not complicated. It is true that technically if the cylinder and cone are not exactly concentric and don't have perfect form, it will not be a true circle, but this doesn't affect anything in PC-DMIS. That's like saying taking 4 points to make a circle won't work because now it isn't round. This is absolutely not the case.

[JEFMAN]

Agree with Jim Poehler and Mike Ruff .
The most problem with intersection of perfect features is about the form of the intersection, which is always perfect, even if there's an angle between axes.
In this case, constructing lines from hits (on the cone, for example), and pierce the cylinder can give an idea of the form.
I made it once for the interction of a cone and a sphere, where the cone axis had an offset from the sphere center...

[AndersI]

Agree with Jim Poehler and Mike Ruff
In this case, constructing lines from hits (on the cone, for example), and pierce the cylinder can give an idea of the form.

Good method - if you can 'guarantee' that your lines are true generatricies of the cone (or the cylinder, if you do it the other way round) . Unfortunately, it isn't obvious from the result of the measurement if the deviation comes from the cone, the cylinder, or both.