Runout on spline

[[To...]]

I have a part with an OD spline. They call out the OD and Runout to the ID. Runout uses Minimum Feature evaluation so I get all of the undulations of the probe scanning over the teeth. I tried using OTE but it got worse and seems to include the valleys in the evaluation.

Next, I went into the feature and changed to OTE, then created a new circle and use Recall One Feature, which should create a circle with the location and diameter but with no form error. The results got better but no plot is generated.

I'm sure they're only interested of the runout on the OD. Do you have any ideas?

UPDATE: I just noticed the results were the same on both evaluations....

Min Feature Runout.JPGOTE Runout.JPGRecall One Feature Runout.JPG

[[Ri...]]

Crank the Point Density high - like 5:1 normal.

Outlier Elimination - 1:3
Adjacent Points - 5
Iterations: 3

GL

[[Ph...]]

Give this a try...

You'll need to create a coordinate system that is rotationally aligned (self centered point) to the teeth.

• Scan OD as previous attempts

• Recall feature points from OD into a new circle.

[list=]

Turn on “manual point masking” within newly created circle and adjust angle range to capture only the points on the OD
of one tooth.

[list=]

Add rotational pattern to circle based on number of teeth.

[list=]

Recall points from each of the patterned circles into a new feature.

[list=]

Check runout

I'd run this a few times to verify repeatability.

[[To...]]

Please sign in to view this quote.

Thanks Phillip, I love your creativeness. Unfortunately, the self-centering point may take too long. I have a target cycle time of 15-20 seconds with an XXT sensor to measure the ID with 3 circles and the OD with 1 circle. If the customer is ready to buy 2 or 3 machines to meet the rate, I may give your idea a shot. 😃

[[To...]]

Please sign in to view this quote.

Outlier Elimination - 1:3 Inside:Outside?

Standard Pre-filters???

UPDATE..... I did 3 Outside and 1 Inside. Haven't looked at runout because I was so excited over first iteration.,..


UPDATE 2 added runout plot

Thank You Richard....

Shoe.JPGShoe 2.JPG

[[Ri...]]

[[Ri...]]

Please sign in to view this quote.

Yeah!!! This is my favorite way to measure splines, and threads (if you want to do linear measurements).

I always do a countercheck to ensure that I kicked the correct points out by looking at the OTE vs the LSQ. The diameters should be close (within the form measurement).

Edit: Don't forget to crank up the number of points you are scanning because as you will see you don't have enough points to filter. You're results should get better. 🙂

[[To...]]

Please sign in to view this quote.

The problem is when I speed it up to try an achieve the required cycle time, Calypso cuts down the number of points.