Issue of a profile line A|B|C

[[Xa...]]

Is there some way to fix this issue?

As you can see in the picture the software in some zones (at the extreme of the section and in the 90-degree intersections) seems as computes the length of the vector incorrectly and then the deviation. you can see in the below picture:

 

 

The same point in ABC reports a dN of 0.263mm and in the profile line computes 0.472mm.. Do you know how to fix it? to avoid these directions incorrectly calculated in these zones.

 

Thank you

 

Edited June 6

[[Ch...]]

I would assume that your nominal section is smaller than your actual one (or alternativly lies away from the actual one in the alignment defined from your datum system). Due to the definition of the ISO 1101 standard that "the extracted profile line shall be contained between two equidistant lines enveloping circles, the centres of which are situated on a line having the theoretically exact geometrical form.", that results in a big deviation values if some of your actual points can only project to one of the end point of your nominal section. Yes that is not the values you are interesting but unfortunetely this is the way how the line profile is defined in the ISO standards.

If you think this is not the case please be free to contact your local partner so that we can have a deeper look on your project.

 

Best regards

Christoph Schult

[[Xa...]]

As you mentioned, the nominal section is smaller than the actual one, and the worst deviation calculated by the software is a diagonal vector with respect to the normal of the nominal surface.

In fact, if we enlarge the nominal section, the software calculates it correctly as you can see in the image... From my point of view the software should really be improved to handle this type of situation, and somehow ensure the accurate calculation of deviations at the ends of the sections, because unfortunately, some people only look at the final value of a profile without considering how it is being calculated and whether it is being calculated correctly. For example, in this case, the worst deviation value is practically halved... from 0.472 to 0.264mm just adding manually more nominal section to compare correctly.

 

[[Mi...]]

Did you create a nominal section first and applied a measuring principle? Which measuring principle did you choose? The choice of measuring principle affects your actual section and whether there are excessive points selected and the end points.

[[Xa...]]

Hello Michael,

First, we create the nominal section and then apply the measurement principle. This is a referenced construction because we are interested in creating the section at a specific height related to plane A. And in the actual part, we follow the same criteria, with this measurement principle.

We have detected this problem since we started using GOM many years ago, from version 7.5. For this reason, to avoid those miscalculated deviations whether because the nominal section is too short to match the real one or vice versa, and the software calculates a strange vector we always prefer to use points instead of a continuous profile.

[[Mi...]]

Hello Xavi, maybe you can try to switch your measuring principle to "actual section" since "referenced construction" can result in a shift between the actual and  nominal section if the alignment between the mesh and CAD is not optimal in the region of the section. Another workaround could be to use a local alignment to make sure that the mesh and CAD are closer to each other around your section/ line profile.