I am working on a Trimaran float hull in Grasshopper.
I have created two surfaces in grasshopper, the top one is created with surface network command and the below is lofted. I have used the same set of curves for both surfaces.
As you can see on the images the curve network suface looks better. Is any differens in quality in real life between the surfaces oris it a tolerance issue?
Best regards Stefan TörnblomSidskrov T30 grasshopper 2021-11-08.gh (10.1 KB)
The lower surface looks like it may have been lofted with straight sections between the section curves. I don’t use Grasshopper. In regular Rhino the Loft command has several different “Styles” including using straight sections.
Whether the difference matters “in real life” depends on how the hulls would be constructed.
yes I agree. I wanted to use grasshopper to make the surface parametric.
I wonder if it possible to arrange the curve list in a way that creates a continius loft over all five cross section curves.
Hi Stefan -
Most likely, yes. You’ll have to post your gh file for a real answer.
I’m moving this thread to the Grasshopper category.
As you can see in the Gh file the output surface is one surface, not four separate surfaces.
Ex grasshopper sidskrov T30 nytt.3dm (264.5 KB)
Sorry, but here it is.
Best regards Stefan
Hej Stefan -
Thanks. Note that internalizing means that you can use the Grasshopper file without needing a Rhino file.
At any rate, back to the issue -
I’m confused as to what the intended output is - should it be one continuous surface or 4 different surfaces?
It should be one contiuous and fair hull surface. The surface is created from a keel line, a shear line and five parametric cross sections. The control points for the cross section curves comes from intersecting the transverse surfaces with the two red guide lines. To alter the shape and volume of the hull I move the control poinst for the two red guide lines.
The curve network command results in a very nice surface but it is very dence has a lot of control points.
The Loft component in Grasshopper has an option for type of loft. From Loft Options - Grasshopper Surface - Component for Grasshopper | Grasshopper Docs
Type T Loft type (0=Normal, 1=Loose, 2=Tight, 3=Straight, 5=Uniform) Integer
You want a Normal, Tight or Uniform loft. The results look like you have a Straight loft.
In Rhino if CreaseSplitting is enabled a surface created using Loft with the Straight option will be split into a polysurface with individual surfaces between each adjacent section curves. If CreaseSplitting is disabled the result of Loft with the Straight option should be a single surface with degree 1 in the direction from one section curve to the next.
Here you can see the result.
Bottom surface Rhino tight loft
middle surface Grasshopper loft
Top surface Grasshopper network surface.
I get a bit confused, is the lofted surface god enough for mould making.
Hi Stefan -
If it’s the smoothness of the zebra stripes that you are wondering about, you can change the mesh settings in the Zebra dialog to get those to be smoother.
The result of Loft with Normal, Tight and Uniform options should be a surface with continuous curvature between sections. There should not be any kinks in the Zebra analysis between sections.
You may need to refine the mesh settings for Zebra. Too coarse a mesh can cause kinks.
Ok But i use the Zebra to check the quality of the surface, how do I know when the Zebra mech is of, or if the surface is bad.
Refine the mesh. If it looks better then the cause was the mesh. It is does not look better with a refined mesh then the cause is the surface.
This is what I get when I open the file you posted above:
Then run a Zebra analysis and run the ***What__ command on the surfaces :
You can see that the surface at the top (produced from Network Surface) produces a much denser analysis mesh and looks smoother in the Zebra analysis.
If you adjust the settings for the analysis mesh until the two meshes have a similar polygon count, the quality of the Zebra analysis looks about the same for both surfaces:
Fairly dense meshes - I just tried a few settings until I found that these produced meshes with the same polygon count for both surfaces.
These are the analysis mesh settings I used to get the above (adjust the Simple Settings to max then click the Detailed button and set Maximum Distance: Edge To Surface to 0.01):
Ok I will try that.
Thank you for all the information. I think you solved my concerns.
I adjusted the Zebra mech according to your suggestions. It worked, the Zebra stripes looked perfect even on the lofted surface.
Then I projected a line on the two surfaces, the network surface and onto the lofted surface.
As you can see ,the curvature graph is smother on the network surface than on the lofted surface, as the Zebra told me in the first images I enclosed in my first post.
Best regards Stefan
Remember to hook the Orca3D component to your Grasshopper definition. This will allow you to see real- time the hydrostatic parameters of your hull model.