Graphics anomaly with Form plot?

[[Ri...]]

I am using form to plot and print in PiWeb for circle that has two broken scans.

When I add this to the form characteristic, the plot looks weird, like its twisted at both ends.

If I plot the same radius, but only use a single arc scan, it plots fine.

THe reported form for dual arc scans and single arc scans are .0004".

Attached is the form plot as shown on the screen (and report), the second image is the actual scanned path.

Calypso v2017

 

Any help is appreciated.

As an FYI, these plots and PiWeb are doing strange things. The plots setup to view and print change, all by themselves.

Edited August 14

[[Ma...]]

Is there a checkbox for compensating segments? On curves it does.

[[Ri...]]

Unfortunately, these CMM's do not have Curve and I do not see an option for Joining mutli circle segments as one.

 

[[Ma...]]

What filter do you have there or evaluation. For me it looks like bad gauss :-)

Can you post like it looks like with only one path points?

Form characteristic will use actual origin of feature as nominal, so you would need to compensate for that while showing only one path.

Once again - what form deviation is reporting feature and what numbers are shown on CAD view? isn't that just a big scale of deviations?

Edited August 14

[[Da...]]

What does the part look like and how is it machined? Looks like two radii with center offset.

[[Ri...]]

Please sign in to view this quote.

It is gauss, I played with other filters, but the result was the same.

With one pass, it displays normally.

Please sign in to view this quote.

The part has a 16ra finish and using a trace machine, it doesn't display any abnormalities and records a .000421" form which correlates with the form reported by Calypso.

Attached is with a single scan.

[[Ma...]]

Is it scanned with same probe? If not, then i would suggest calibration.

How much actuals of origin differs between: only left scan / only right scan / both scans?

[[Ri...]]

One probe that scans all of the sphere for reporting.

No articulations, Z Down.

The band shown in the prior image is .0005" max. The minor wobbles are what creates the .00004" form.

And I just noticed I typoed the original value for reported form, it was .0000421"

Edited August 15

[[An...]]

Does grouping the circle paths as one make any difference in the calculations?

I initially was thinking it was jsut the probe entering and leaving the feature but that doesn't make sense for the piece to sink back in like you shown. I assume this is a typical metal part?

[[Ma...]]

You might want to turn off the Point Masking Time which is set to 2.5 for these scans. It might smooth the entrance and exit of the scans.

[[Ri...]]

Mark, I will look into that.