help surface modeling

[LuPo_81]

I need to model an object using single surfaces that I will later have to extend, trim and solidify.

as seen from the attached image I must generate surfaces from the blue profile (which comes from a toroidal projection) such that in each point I give a angle of 15° regarding the normal surface.

I had thought of using sweep-> surfaces using the blue profile as a trajectory and as a section a straight at 15°, so that the section, turning along the trajectory, maintains its angle of 15° compared to the surface.

Now: apart from the fact that not all segments of the blue profile are selectable as trajectory (why?), but then selected a piece of trajectory that suits me it is "impossible to set the sketch plan".

Yet someone else did it, as visible in the second image!

Any ideas?

ps: use pro/e wildfire 4.0

 

[maxopus]

If I understand your need, you have to use the offset function.
realize the sketch at a certain distance from the surface (no matter how much, the important thing is that you do not intersect it).
then apply the disform offset function and decide the depth and angle of sform to be applied to the lateral surfaces.
it does not let you select all segments because obviously not being tangent among them would create overlaps in the trend of the sweep.
If you want you can post the zipped file so I'll take a look at it.

 

[LuPo_81]

I managed to do it anyway. in practice I projected the curves generated by the toroidal projection on the toroidal surface itself, although already there are, along the normal to the surface itself. I had to do it for every single section of the surface because it does not allow me to project curves on chains of surfaces, but only curves on a single surface, perhaps because it cannot find the normal at the same time on multiple surfaces: mad:.
I then generated the surfaces using the curves projected as trajectories, obviously not using all the curves together, but using the same chains generated during projection, although the individual segments are not tangent.
But so it is very long, if I have more complex objects becomes a workmanship!

I don't understand how you say, I'm sending you the zipped file, if you have a little time, would you show me what you mean? Maybe it's faster than you say.

 

[maxopus]

I managed to do it anyway. in practice I projected the curves generated by the toroidal projection on the toroidal surface itself, although already there are, along the normal to the surface itself. I had to do it for every single section of the surface because it does not allow me to project curves on chains of surfaces, but only curves on a single surface, perhaps because it cannot find the normal at the same time on multiple surfaces: mad:.
I then generated the surfaces using the curves projected as trajectories, obviously not using all the curves together, but using the same chains generated during projection, although the individual segments are not tangent.
But so it is very long, if I have more complex objects becomes a workmanship!

I don't understand how you say, I'm sending you the zipped file, if you have a little time, would you show me what you mean? Maybe it's faster than you say.

See if that's okay with you.
Look how short the feature tree is, it took more or less 40 seconds.

 

[LuPo_81]

the idea is great, but it is of fundamental importance that the starting curves are generated by the toroidal projection and not by a sketch plan ... mmmm ... unless I project the same toroidal projection on a plane and then make the offset. I have to study this.
Thank you.