Circle Paths vs Helix Path

[[To...]]

I was checking position of one cylinder to another "in-line" cylinder on a demo part and used a 4 helix path on both to show off another measurement technique. However, when I did a test run 10 times, the repeatability was around.0025", which was terrible. I changed the strategies to 4 circle paths, using the same speed and step widths and the repeatability range was .00016", which I thought was pretty good for a star probe on a DuraMax.

Anyone else have an opinion on helix paths on cylinders?

[[Jo...]]

I have an opinion, based on observation. The helix seems to have issues providing a stable vector. Compare the vectors between those.

[[Je...]]

.

I have noticed the same issues on helix patterns for RDS / XXT setups, especially if there is any significant weight to the probe tip. I don't notice the same error on VAST sensors. They seem to handle helix patterns just fine.

.

[[Da...]]

I always use circle paths.

[[Ow...]]

Yeah, as known, circle paths have the ability to overlap 360° and remove unwanted probe deflection at the start of the scan.
When looking for wear or distortion in a cylinder bore, I'll use several circle paths and 4 to 6 linear paths down the length of the cylinder as the vectors are better controlled with the linear paths and you can set the evaluation range of the lines to remove start error by choosing special settings in the strategy. Makes a cool looking plot too. 😉

[[Ün...]]

In a previous training, the technician stated that measuring helically according to the size of the cylinder gives a more reliable result. He also suggested that the Spiral rises in the +Z direction, starting from -Z. (because it's RDS)

I mostly choose depending on the size of the cylinder and the probe size I use.

If I measure diameter with 4 scans, I measure with 6-7 helical scans.

[[Ro...]]

Helix is never the correct answer, for anything.

[[Er...]]

My experience measuring cylinders with XXT...

Helix is only for threads.
Circles are best for diameter.
Lines are best for position.

Math wise it makes sense to me after seeing the data. Especially using CAD evaluation. If I measure three circles on a cylinder, that is only three points along the axis and prone to variation.

If I need better repeatability I will add lines to the strategy.

[[Da...]]

Always seemed like a SALES tool... Haha. But you can run a worm major ! Looks cool ! 🤣

[[Je...]]

.

We had a high-profile military customer who swore by helix paths. They never were able to argue why they were so effective, though. I mostly swear at them rather than by them. In theory they sound sophisticated, but they don't reach their potential due to the mechanics of a CMM.

.

[[Jo...]]

Please sign in to view this quote.

I use them in two situations:
scanning completely thru a part to get "pin" diameter, where there could be burr or flash on edge(constrain vector)
synced with thread crest to get runout value(slowly, and ONLY at full thread height)