Spline Measurement

[[To...]]

I've got a part with a female spline. The drawing calls out the minor diameter as datum A, and it is used as the datum for checking runout of some other features-a couple of surfaces and diameters. (Yes, I know that the minor diameter isn't really the functional part of a spline, but that's my drawing).

Not sure what would be the best way to measure and evaluate this. I don't have any repeatability on subsequent runout measurements. The spline minor is about Ø34.5mm, and about 32mm long.

I measure a cylinder with 5 circle paths (5mm apart, 25mm total length), using an 8mm probe, speed 1.0, and evaluate with max inscribed. But I keep toying with constraints, LSQ, and outlier mode, and mulitple iterations, but I don't have a good grasp. Basically, throw out the outer points, and keep the inner points. I feel like I'm just pushing buttons hoping for a random happy outcome.

Any thoughts?

[[Cl...]]

Seems like a large diameter stylus to probe that size (34.5mm) diameter.
Have you tried constructing a 3d line from the 5 circle paths instead of a cylinder?

[[De...]]

Use a self centering point inside one tooth space (evaluation set to midpoint) to find a gap - for simplicity this can be set to run in manual mode in the features setting editor / characteristics settings editor.

Create an alignment using the center point of the spline, and this point for rotation with a special value of 1/2 angular pitch calculated with the formula
(360 / Number of teeth) / 2

The special value combined with using the self centering point to create rotation will mean that the position of the line will be aligned with whichever axis is used in your planar rotation in your alignment.

Create a 2D line using this alignment.

Because your 2D Line uses this alignment, it means your 2D line location will be at 0 in one axis and at 1/2 the minor diameter in the other axis. then you just have to set your third axis to the locations of the length of the spline. and make sure your normal angle points to the center of the spine for an internal or away from center for an external to control the probing direction of the line.

Create a pattern centered on the center of the spline, using rotational pattern for scallop of the 2D line with the angular pitch of one tooth and repeating the number of teeth.

Recall all points from pattern into cylinder for Minor Diameter.

[[To...]]

Thanks Derek-that gives me something to work with.

I used to use something sort of like that in umess-find the spline "clocking" and set an alignment. Then a circle, with 26 equally spaced points (say for a 26 tooth spline), then another circle a little lower, and so on for the length of the spline-but it took quite a while per part. This way seems like it would be quicker-26 scanned lines vs 130 single probings. With round features, I never seem to think of lines.

Clarke-the 3d line from the circles gave a very similar result, a few microns difference. My thinking with the big probe was to skip over scanning the flanks of the spline I wasn't interested in, and try to make a smoother scan-less bouncy.

I guess I figured there might be an easier way of just scanning the circles sections and filtering.

[[Cl...]]

🙂