clylinder probe accuracy

[[Ro...]]

We have a part that needs to check the Inside diameter using a cylinder probe. ( we have a cylinder 2.0mm ruby) when I check the ID in different "Z" heights I get different numbers ( up to .002 inch difference) I calibrate it on the calibration sphere and then use a ring gage to check, have to adjust radius of cylinder sometimes to get a good number with the ring gage. Is there a another probe I can use, or I'am I doing something wrong. The ID is shaped with a radius and angle then they intersect this is where I check for size.
Thanks in advance

[[To...]]

The problem with cylinder probes is that their axis might not be perfectly parallel to the axis of motion which is why you're having an issue with your results on the ring gauge. You're probably hitting at the top of the cylinder on one side of the ring and at the bottom of the cylinder on the other side of the ring.

Instead of using a ring gauge, try using your reference sphere. You already precisely know the radius and it will only make contact at the equator.(or very close to it)

[[An...]]

Tom's right,
Your measuring ring will not be 100% horizontal or right-angled to the cylinder axis!
With this you provoke an edge measurement in places.

[[Da...]]

Can you show a sketch of the part you're trying to check ? There might be a better way.

[[Ro...]]

Dave, here is a pic on what I have to check. Thanks Anybody have another idea without using cylinder probe need a diameter (arc length about 180 degrees)

PICDoc1.pdf

[[Da...]]

OOOH that's a tough one ! The cylinder probe won't make contact at the "INTERSECTION " of the ID cone and the radius. That's actually a few degrees below the thru bore size. If you have a NICE engineering department you could ask for a diameter on the angle at a Gage line , and then report the radius centerline height. The actual intersection of a radius and the angle is probably not actually important. It is also VERY hard to nail down because , depending on manufacturing process , it is probably not actually a tangent point, or an actual intersection...

[[Ri...]]

Not sure this will work but if you measure the radius'd portion as a torus then measure the cone, you will have the Cone Calculation construction available. Put a check mark in for Position Definition and in the Diameter field use this formula

-- getActual("Torus1").diameter- getActual("Torus1").diameter2

You will get the result of where the cone reaches the most inside diameter of the torus. Take the calculated result to the CAD guys and confirm the result

[[Ri...]]

Another possibility would be to measure the radius'd portion as a torus using the small circle scan path strategy.

Then create a circle with one of the circle paths.

Then create a 3D Line feature with the created circle X, Y, and Z coordinates and the Half Cone Angle for the A1 or A2 , whichever is appropriate.

Then intersect the 3D Line with the cone shell.

Take that number to the CAD guys and see if they like that.

[[To...]]

Just thinking out loud here. How about creating several sections on your cad model. Then, generating lines and circles from the sections. Then, creating intersections of each line/circle combo.

The length of your lines will be impacted by the diameter of your probe. Depending on how you set your section plane, you might be able create good usable lines without having to do too much tweaking of the line segment.

This method may not be too stable as you'd be creating your diameter from minimal points on a short arc segment.

[[Da...]]

Chances are that there are either 2 intersections , OR, there are NO intersections.
Circles and lines are always tangent on the model, but never in real life.

That being said finding the actual LOCATION of the cone IS something we can do , very accurately. Finding the location of the torus small radius is not that accurate.

The 3d line creation, That Richard mentioned, using the actual 1/2 angle of the cone and originating at the circles center point will point to where the intersection SHOULD be. ( the line would look like a perpendicular from the cones surface and ending at the torus small diameter center ). You'll have to do this several times around the 180 degrees of arc.

[[To...]]

Please sign in to view this quote.

True Dat

[[Ri...]]

What is the tolerance for the intersection we are trying to get? Also, how dependable is the manufacturing process for this collection of features?

[[Ri...]]

After taking a closer look at the image provided, it's possible Tangents might work. It require the radii at the top and bottom of the cone be measured. Also, Tangent provides the point of Tangent of two circles and shows the line between the two points but the values displayed are only for one point. It is very easy to get the other point by swapping the two circles. I've included screen captures.

[[Ri...]]

Take note of the Z nominal value for both Tangent constructions. Also, next time I will size the image before uploading.