Iterative Alignment Method to Measure Bore from Lands Only on Rifled Barrels

[[Br...]]

I’m looking for input from anyone who has implemented an iterative bore-measurement approach in Calypso 2022 that isolates the lands only when establishing a datum from a rifled barrel.

On our previous Hexagon CMMs (PC-DMIS, non-scanning), we established the primary datum feature from the lands by measuring a 4-point circle. After the initial measurement, we used an IF/THEN routine:

• If the measured diameter fell outside a defined range (indicating contact on grooves), the alignment would rotate by a set increment.

• The circle would then be remeasured.

• This loop continued until all four contact points landed only on the lands.

• That resulting circle defined the bore diameter.
  We repeated the process at the opposite end of the barrel and constructed the datum axis from those two circles.

With our Zeiss machines, we’re currently scanning and using a maximum inscribed circle to define the bore. While this works, I’m interested in recreating something closer to the iterative “land-only” approach described above—ideally using Calypso logic, PCM, or another built-in method.

Has anyone done something similar in Calypso 2022? Specifically:

• Rotating an alignment iteratively based on measured diameter feedback

• Forcing point selection to land features only

• Or scripting a looped remeasurement routine to converge on a land-derived circle

Any guidance, examples, or best-practice suggestions would be greatly appreciated.

[[Ma...]]

Most of it only with PCM to have a comfort.

For others you have to do much of a work by copying features and so on.

I would recommend using base alignment to have iterative result and better / safer definitions of nominal paths / points to scan.

[[Ze...]]

How about using maximum/minimum coordinate on your scanned circle to find the high/low spots. Then use that info to refine subsequent alignments. 

[[Cl...]]

What about using fixturing to physically clock the geometry? This would allow for an alignment

on one of the lands. I would think this would allow you to scan only the surfaces you needed.

We do something very similar here. Physically clocking some very complex geometry so we can

avoid certain surfaces to measure Datum circles.

Edited Tuesday at 08:45 PM

[[Pa...]]

If your geometry allows, you could use self-center points like in a spline to find the start of a land.  You could then trace the correct nominal helix with line scans.  This would find the start of the land in just a few seconds.

[[Ke...]]

What sensor are you using for this? The methodology may be similar, but an active system like the Vast Gold or Vast XT can use a feature called flyscan which can be quite useful for things like rifling.

As a general method, create a circle with a short partial arc strategy (enough to hit 1 land whether it starts in a groove or not, but not much more than that) near one end of the barrel. Set that strategy to use single points (you may be able to just scan based on the depth of your grooves and size of probe), and set your point density/spacing enough that you’ll definitely get a good point on a land, but not so many it takes forever to run. From this feature extract a point on the land (most likely using a max point of that circle feature). Use that point as the clocking feature in an alignment, so that one of the axes points directly through the land. Use this alignment as the alignment for your datum measuring feature, so that way the start of the strategy will always adjust to be on the land that it found previously. From here you have options for how to actually check the datum feature. You could flyscan a circle,do a 360 circle set to single points with a number of points equal to your land count so each point will hit a land(then rotate by the Twist RateXDistance to do another measurement further along the barrel), or potentially use a cylinder feature with several helixes matched to the twist to traverse the entire barrel scanning the land depending on what you need.