# Design Honeycomb structure

Honeycomb structure with a thickness of 0.5, the hollow hexagon is materialized(This means that the hollow structure becomes a solid, yellow marker. ) by the trajectory（short-line segment representation）.

I’ve implemented a curve that abstracts the hollow hexagons through the area, but when I use Solve intersection events for two curves, I can only achieve one intersection at a time. Now that the design idea has been modified, it is necessary to consider the materialization of the area through which the short line passes. If the short line segment intersects the hexagonal edge, it is cumbersome. So please ask your friends for help.

Dear forum friends, this is a modified model and method obtained by combining two friends for @HS_Kim and @diff-arch. In my model, the solidified hollow structure is controlled by a short line. But I use Solve intersection events for two curves, I can only achieve one intersection at a time. But there are a lot of lines in my picture that intersect the hexagonal sides. So I need to modify it in my Grasshopper program.
Thanks again for all the people.

I have known these friends who are very nice, but I have a little difficult in how to build the accurate model.
So I come here for posting again.

Please do so in the topics you have already started. Friends will appreciate this.

This is my last two cents…

Honeycomb Structure_reV2.gh (11.7 KB)

Dear Kim
I need to modify my Honeycomb structure, please check the figure.
You will find that the hollow hexagon in the Honeycomb structure is materialized (This means that the hollow structure becomes a solid, yellow marker. ) by the trajectory（short-line segment representation）.
If the trajectory is a line, you’ve told me how to do it before.
Now that the design idea has been modified, it is necessary to consider the materialization of the area through which the short line passes. If the short line segment intersects the hexagonal edge, it is cumbersome. So please ask your friends for help.

Honeycomb Structure_V12.gh (17.6 KB)

Hi @shuxin,

I hope this is finally is what you are looking for. It is exactly what you portrayed in your sketch.

Have a nice weekend.

honeycomb_structure_v13.gh (41.0 KB)

Your creation process is very detailed, thank you for your help. It is for my structural design, compared to @HS_Kim , a lot of places do not understand. So is the design process simplified?

This is the final modeling file I got based on your method, but there seems to be a warning. Seeing that you want to build the hexagonal model I need, you also use the Cluster tool specifically. Seeing your help, it is worth learning.

Thanks again.

Honeycomb Structure_V16.gh (16.7 KB)

Hi @shuxin,

You don’t need the custom hexgrid cluster, but keep in mind that your sketch shows another hexgrid than what the vanilla `HexGrid` component produces. In your diagram, the hexagon cells point upwards, whereas `HexGrid` produces cells with a lateral cell ‘direction’. It is really up to you to stay faithful to your original idea or not. Both should work with the rest of the script though.
My idea was to show you how to construct a hexagonal grid starting from a single cell.

The warning stems from you, trying to perform a solid union on solids that aren’t even intersecting, which is impossible!

Furthermore, generally speaking, for a boolean union to work properly, a good practice would be to sort your data into tree branches of intersecting geometries, before inputting it into `SUnion`. This way, the component knows what to try to union and the probability of errors, being raised, is decreased.

Thanks for your advance. When I try to sort my data into tree branches of intersecting geometries before inputting it into `SUnion`, there is no warning in my program now.
In the custom hexgrid cluster, if I want to turn a single cell into a triangle or a quad, I find that the polygon structure produced is messy.
Honeycomb Structure_V12.gh (17.6 KB)

Regarding my hollow hexagonal materialization: the red box as shown below, since it is a random parameter control generation, this will happen if two consecutive ones are made, so it is foreseeable to do 4, and 5 will have more of this.

Because the number of hexagonal matrices is limited, the more physicalized sampling, the easier it is to cause interlacing.
My thoughts first through a series of random points, and then find the closest point to the random point, sorting and filtering