Hmmm… While I am operating in total ignorance of the specifics, it seems reasonable to me that you could define initial measurements as two lines defined by three points on a plane set at a particular position and rotation on the model. (Circumferences would just be four lines on the same plane.)
To ‘drape’ the measurement over the model,
- The endpoints are be moved on a line parallel to the surface normal of the model until they are each resting on the surface.
- The centre point of each line is made an endpoint, creating two lines from each one.
- Repeat steps 1 & 2 until there are no more points that can be moved more a set value. (e.g. 2mm)
- The measurement will be the sum of the lengths of the line segments.
Does that sound reasonable?