Recalling separate curves into one curve

[[Jo...]]

I have a program that has a profile callout that goes from point A to Point B.  Originally I scan this as sever curves because that is what engineering wanted... i e one curve from point A to Point B across several features (radius, cone, radius, diameter..etc).  They then came back and asked to separate the curves into individual sections... I did this and for YEARS this was fine.   Now they want it back to original 1 scan.  Program has evolved to where I do not want to use the original probe I used to get it in one scan.  I tried to recall all the individual scan sections into 1 curve, but I am getting random "noise" on the curves.   I am using the same filters on the recall... if I increase the filter, i filter out the noise, but then the complete scans do not match the individual scans.  Several part runs do, some multiple complete scans have this noise but the rest of the scans match... its all random... 

NOISE

NO NOISE

both are from the same scan.. bottom one is the actual scan, the top with the noise is when i recall it. 

[[Ma...]]

Are you recalling calculated points or all?

Also do you use masking points beggining/end of a scan? You can show your masked ( red ) points on actual scan and i think it will corelate.

When i do scans, i always mask curve points at beggining and end of a scan manually - keeping those points for me to review

[[Jo...]]

I do not mask other than the global "point masking by time interval, but not in the settings of the feature.  the bottom scan is the one segment (curve) that is originally evaluated as a profile of a line.. the top scan is the exact same segment, but recalled with the other curves.  I have tried both selecting and deselecting the "filter option" that is now available in 2025 when recalling points.. both give me the same error.   I have tried increasing the filters in the eval..which does get rid of the noise, but does not match the original scan.  Engineering wants to look at both.  this particular segment, the nominal points set to 425 with point scan set at 1000

[[Iv...]]

Engineers 🙄 My buddy said, engineers should go back to operating trains... He said it to me 🥲

Please sign in to view this username.

 

• Are you using the classic or new profile?
• How are you creating the nominals for the recalled curve?
  • My preferred method is using the connect nominal curves
  • If using the "re-calculate nominal geometry" option from the recall feature points pop-up, make sure there is no actual data in your curve
• I would recommend using the curve results points option when recalling the feature points. This would pull the actual curve points rather than the raw scanned points (generally 3x your actual points)

 

Edited yesterday at 04:31 PM

[[Jo...]]

I have done this several ways and i am still getting random noise... 

I have created noms 3 different ways

1) from recall points , then recalculate.... in order to do this no data can be in the features yet

2) I created a text file with all the nominal points and either "read into" within the curve or use the points set

3) used the cad model to generate the nominals

 

once that was done, i recalled the points using both non-filtered data and checked the filtered data. 

 

Every one of those things creates the same random "noise" .... sometimes it is at the end of one curve some times its well into the curve....

[[Iv...]]

Are you using the new GD&T Line Profile or the Classic Line Profile?

[[Jo...]]

classic for both... cant really use the new one for this type of eval.

[[Jo...]]

but i am also looking at the Raw Data.. I see the noise it that even without the GDT evaluation

 

[[Jo...]]

I am trying to recall 3 curves..... with all the errors I was getting I did the following...    

1) i copied the recalled curve (no actual data anywhere when i did this but the same nominal curve), 5 times making 5 new features

2) I then de-selected curves just to select 1 single curve for three of the new features

3) I then de-selected curves in two of the new features to just two... one feature has the curve at the beginning and middle, the other has the curve at middle and end

All of those came out with no errors... its only when I have selected all three, that the random error shows up...not just at the ends of one curve... some are in the middle of the the curve itself.....