I'm not that sure that it is that simple... Imagine that the surface has the shape of an "egg box" on a table ( slightly above the table) , and the non intersecting plane is the table top. What would be the curve ?
To define a line, you will need at least 2 pieces of data; 2 points or 1 point and a vector. So unless there are 2 "points closest to a non intersecting plane" you will need to provide some more information. Besides, you say that want a "line along a surface". Unless the surface is also a plane (or some very special situations like parallel to the axis of a cylinder or cone) any curve "along a surface" will be a Spline and NOT a Line.
Perhaps you will need to provide us a model and some more information on exactly what it is that you're attempting to create.
John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA Siemens PLM: UG/NX Museum:
To an Engineer, the glass is twice as big as it needs to be.
If I'm visualizing this correctly, I'd try using the isocline curve command. For the vector, choose something parallel to the plane (not the plane normal) and enter 0 (zero) for the angle. This will create a curve (or curves) on the surface that have the highest 'elevation'.
Attached is a , from a math point of view, very simple surface. NX7.5 format.
It is degree 2 single patch in both u and v.
But very difficult ? to create the curve.
Regards,
Tomas
I created 2 isocline curves (choosing the x axis for 1 curve, the y axis for the other), then measured the minimum distance from the datum plane to the surface (using the "create line" option). The minimum distance line hits the intersection of the 2 isocline curves. I then created an arbitrary datum plane and used it to create an intersection curve on the surface. The minimum distance line from the datum plane to the intersection curve also hits one of the isocline curves.
Thanks to our wonderful filtering software I can't post or view files. (well I could view pictures at home but that's about it)
At this point the question has been rendered academic to me since I realized the surfaces I'm dealing with are constant conic so pick two ponits draw a line and Bobs your uncle.
However, seeing as I've run into this before, I'm still intersted in a solution. (Which an issocline cuve perpendicular to the plane shuould accomplish unless I'm mistaken) (thanks cowski)
toost, It seems to me (correct me it I'm wrong) that you'd get the "high" points to the surface regardless of the vector as long as it's paralell to the initial plane. (I see what you're saying now cowski)
When creating the isocline curve(s) you will need to select a vector of interest parallel to the plane. If you select the vector normal to the plane, NX will not generate the curve of closest approach.
When you wrote "an isocline cuve perpendicular to the plane", I assumed you meant picking the vector perpendicular to the plane to create the isocline curve. I apologize if I assumed incorrectly.
Hm, i'm not sure that the method is reliable.
I edited the surface such that it is simpler, now it's degree 2x1 and single patch on both u anv.
It is in other words linear in one direction. ( The isocline should be "straight as a line")
Then i drew a sketch on the plane and in that i drew 2 lines to use as vector definitions for 2 Isocline curves. The two isoclines are totally different and none of them intersects the minimum distance point. See attached 7.5 model.
![[hourglass]](/data/assets/smilies/hourglass.gif "[hourglass] [hourglass]")