Which filtering and outlier settings for measuring Knurling?

[[Jo...]]

I am trying to measure knurling on a bolt, so outer diameter. I cannot find anything in the cookbook or help menus to direct me on how to set the outliers and filters. Does anyone have any suggestions?

[[Jo...]]

Hello Jordan. I hesitated to answer this because I'm still relatively new to Calypso, therefore not super confident about my 'expertise'. But I'll offer this since you haven't received another reply. I don't think you need to use any special filter/outlier settings. The standard recommended settings for your stylus and feature type/size should work. What you may want to do is change your evaluation settings depending on the characteristic type. For example, if you're looking for the functional size of an OD, set your evaluation settings to Minimum Circumscribed. That should tell you what is the smallest hole that OD will fit into. I know that one of the programs I run, we use linear scans (vertical) across the internal threads. When you look at the scan, the peaks and valleys of the threads are clearly visible in the graphical representation. But when I check those diameters with the Max Inscribed setting, I get the same result that I do when checking the same holes with gage pins. If you need to check location of the bolt, you probably want to use LSQ and constrain the Normal Vector. Hopefully you have already resolved this but if not, try this. 

[[Cl...]]

 

If your intent on scanning it, I'd set the evaluation to outer tangential.

Edited Tuesday at 03:39 PM

[[Ch...]]

I've measured a lot of knurling in the past and it depends on what they need to know.  Are you trying to measure size and depth of knurling, for example, width of each knurl and depth/height of each knurl?   Or do you only care about the total outside diameter of the feature.  

Size and depth of knurling is more difficult on the cmm.  That's because your probe size will probably be larger than the knurling.  A good tip is to use a profile measuring machine.  Something like a Mitutoyo FormTracer or even the SJ410 will work.  

If you only care about the OD of the knurled surface, then you can use a circle set to Outer Tangential. I would use a high point density and aggressive filter because you will probably get lots of noise as the probe bounces off of each peak and valley.  

[[Ke...]]

Most settings will depend on the size of your bolt/knurling; and I don't usually need to inspect bolt knurling BUT I do similar measurements on splines... some of my go-to practices are:

Using a "larger" than usual probe diameter (for a smoother scan)

Scan slowly & Collect a lot of data

Set Outlier Elimination to "1" for Inside Workpiece, and "3" for Outside Part (idk why they use different terms lol)

You may need to include some adjacent points.

Also: Open your feature - right click -  and select "Show actual points" and "Show macked actual points"...after your scan, look to make sure that you are collecting and removing an appropriate number of points.

But... as mentioned, Outer Tangential/Minimum Circumscribed may be suffucient for your needs

[[Ga...]]

To all who have mentioned MCC/Outer Tangential evaluations, is there a difference in what these settings do in this scenario?

I would default to MCC given that it's an outer diameter, but I honestly don't know much about outer tangential other than I use it to simulate the way an imperfect bottom plane of a part would sit on the glass of a Keyence when verifying my values with CT. I have never considered using it on a diameter. (I genuinely don't know much about filter/outlier/evaluation/constraint settings despite doing this for the last 5 years and taking basic and advanced calypso classes)

 

[[Ke...]]

Please sign in to view this quote.

 

Calypso Algorithms - Evaluation Methods.pdf

[[Cl...]]

I don't know enough to comment on this exact question. Bear one thing in mind regarding filters.

Circles, slots (Closed features) typically use Gauss and the UPR setting is based on size. The cookbook goes into detail.