feature set question

Would need to see the routine.  What do you mean by "bad results" - is the measured profile value different for each case?  They should be identical if you're using the same hit data.  If it's just that the graphics look messed up then that's understandable because there is no intelligent point ordering - it just goes dot-to-dot between the hits and features in whatever order they were selected. 

the values are showing profiles OOT out to 9 inch's. the part is only 4 inches. I can change the order I put into the feature set and it will change the results. I am not changing the hit points. Just the order i create the feature set with. 

I do not normally program this way. I would ether do each feature as its only profile or take points on all the features and create the feature scan that way.  the customer is requiring us to do it this way. 

I'd need the program to investigate, are you able to send it to me - the .prg and the .cad?

also this is 2020 R1. 

let me see what Ican do. Unfortunately, this part is for the US DOD and they tend to be a little strict. 

Getting the program and CAD model is probably a non-starter then because I'm based in the UK.

If the problem only occurs in 2020 R1 (I'm assuming you're using XactMeasure dimensioning not legacy) then I'm not too concerned.  As you may be aware, we completely replaced XactMeasure with the Geometric Tolerance command in 2020 R2.  Geometric tolerance commands take the raw hit data directly from the features and re-calculate everything so I wouldn't expect to see this problem in newer versions of PC-DMIS.

correct. ITAR and things. I am trying to see if I can at least send you a screen shot of the feature set.

I have only seen it in 2020 r1 with Xactmeasure. In 2023 R2 I do not have this issue. This customer requires us to use 2020 R1. 

Is there a way in 202 to get results we can trust? 

A screenshot won't be much use, I wanted to check to see whether it was using the hits or the feature centroids.  Make sure the feature references in the SET command include the hit syntax...

SCN1 =FEAT/SET,CARTESIAN
THEO/<41.378,30.09,0>,<0,0,1>
ACTL/<44.984,29.359,-0.001>,<0,0,1>
CONSTR/SET,BASIC,PLN1.HIT[1..PLN1.NUMHITS],CYL1.HIT[1..CYL1.NUMHITS],,

Rather than just...

SCN1 =FEAT/SET,CARTESIAN
THEO/<41.378,30.09,0>,<0,0,1>
ACTL/<44.984,29.359,-0.001>,<0,0,1>
CONSTR/SET,BASIC,PLN1,CYL1,,

A screenshot won't be much use, I wanted to check to see whether it was using the hits or the feature centroids.  Make sure the feature references in the SET command include the hit syntax...

SCN1 =FEAT/SET,CARTESIAN
THEO/<41.378,30.09,0>,<0,0,1>
ACTL/<44.984,29.359,-0.001>,<0,0,1>
CONSTR/SET,BASIC,PLN1.HIT[1..PLN1.NUMHITS],CYL1.HIT[1..CYL1.NUMHITS],,

Rather than just...

SCN1 =FEAT/SET,CARTESIAN
THEO/<41.378,30.09,0>,<0,0,1>
ACTL/<44.984,29.359,-0.001>,<0,0,1>
CONSTR/SET,BASIC,PLN1,CYL1,,

the other option might be to convert those planes & cylinders into points instead of planes and cylinders.  then, if needed, construct planes and cylinders from the points to report the planes and cylinders as planes and cylinders and make a feature set of the individual points for the profile.

Matthew that is how I normally do it using points. However this customer wants it done this way.