Curvature graph wishes

I’d love to see a little more functionality for the Curvature Graph.

My wishes are-

  1. Separate display scales for the U and V directions of a surface (see image1 for reason why - V is visible U is not).
  2. Ability to increase the number of graphs without having to increase the iso-curve display.
  3. Ability to choose between iso-curve related graphs and ones which cut through the surface perpendicular to X,Y and Z planes at a given cut distance (see image2).
  4. Not a wish, but, I seem to be getting some dodgy display issues in the object properties panel while trying to mess with the iso-curve settings while using the curvature graph (see image3).









While I’m at it, let’s have another wish…

  1. Could the Curvature display be limited to one viewport? Ideally the perspective one. This leaves the others available for easy control point selection and visibility of changes to the Iso-curves etc.

Pic below shows the coloured curvature display in all four viewports -


Hi Rob - thanks, I’ll get this on the pile.


1 Like

Hi Pascal,

It is so sad that this hasn’t been implemented yet. This is brought up over and over again.
A client of mine made this as a plugin for themselves which they don’t want to share or sell, which I completely understand, but what makes surface fairing so much easier.
The current curvature graph for surfaces is not useful at all.

Hi Gerard - does your client’s plug-in retain the UV orientation of the graphs or does it use some other (XYZ, or User defined, say) orientation?


Afaik they also have XYZ graphs as this is what we need in the marine industry. They have it rollin for a couple of years now so I cannot remember well (@menno).

Call me greedy but I like the sound of all three of those…(all with independent scales, density and colours).

1 Like

We have created a dedicated plug-in for ship hull fairing where the curvature can be shown in the following ways

  • Along a line of control points
  • Along intersections in the X-, Y- and Z-direction (in ship parlance: sections, buttocks and waterlines)
  • Gaussian curvature on the hull with iso-curvature lines

These curvature views help the user in creating a smooth/fair hull surface. Below is a typical view with dedicated viewports for the sections/buttocks/waterlines, including curvature on the active control point. The perspective viewport shows the gaussian curvature and the Fairing viewport shows the curvature along the active line of control points.


Well Pascal,

The Marin plugin of Menno is what I meant.
Thank you Menno for informing us with the nice images.
This is what we need in the marine industry when it comes to fairing a ship hull.

May I also add that it’s not just hulls that need fairing. Curvature analysis is also important for superstructure elements too (only saying this because I’d hate to end up with something that was purely oriented towards hulls, so flexibility to use this over all manner of surfaces at different angles etc would be great).

You are right about other applications of this functionality. Also other industries might need this as it is essential analysis functionality for freeform surfacing.

1 Like

Hi Rob - here’s a start

Is the display glitch the blank density setting?



1 Like

Brilliant, thank you Pascal!

Yes, when the surface was selected a whole load of the properties went blank and were difficult to get back. It happened a bit randomly though.

I’m also having some general display issues with the viewports not updating during interactive commands (like extend surface and analyse direction) and also during some general viewport view rotations. It seems to ‘stick’ with whatever is on the screen at the time of the first click. I have to force it to regen by clicking on the viewport name and then it updates. I’ve updated display drivers but it’s still the same. However, I am running a very old graphics card with a minuscule amount of vram, perhaps that’s the issue??

Thank you Pascal,

What I miss in the myjetbrains post is the option to have only one single curvature graph on one isocurve, surface edge or xy plane for example.
Otherwise we still see graphs like down below which are useless. In the marine industry, we judge the quality of surfaces often in one plan view.

Hi Gerard - thanks - as a workaround in V6, and maybe a better solution since the location is completely up to the user, you can extract an isocurve with History on, so you can display the graph for just that curve.


Thanks Pascal, I didn’t knew that _ExtractIsocurve is able to have record history.

That brings me to the amazing fact that still contour and section are not having ‘Record History’ possibilities. Can you please explain why it is not developed yet?

Hi Gerard,
Good points.
That said, it is very easy to set up a contour history, just make a line, duplicate with history on and project with history on, then put original curves on a hidden layer and the result from project on a locked layer and you are good to go.

Maybe the reason history isn’t implemented for contour is that it produces multiple results that are not linked to each other, and what should Rhino do if a user tries to delete one of the result curves? should they be considered a group of objects, or multiple single curves?

Shouldn’t matter. Currently if you use Copy with History on, and then you delete one of the copies, the rest still update when the original is modified…


Thanks Jorgen,

Your trick is something we do here too with _Intersect of a corrugated U-shaped surface object.

I can imagine of several issues with the contour command. However, Orca3D has good functionality with their dynamic sections.

I’ve got a feeling the ‘wish’ pile has got a little low, so here are a couple more to top it up…

  1. Could there be a transparency slider for the coloured curvature analysis? This could allow us to easily ‘see through’ the coloured surface and select control points.

  2. Could the specular highlights be removed from the surface? They tend to get in the way of viewing the colours and can make some areas look as if they are distorted slightly when they’re not. Of course this is solved by rotating the view so the highlights move to a different place, but I just feel they’re not needed in this situation and are not adding anything to the analysis.

See images below - first image shows what could be interpreted as possible surface distortions, second image shows model rotated and looks fine.

Many thanks!