Shifting grid

Replies to This Discussion

Permalink Reply by Canhui Chen on April 21, 2015 at 10:07pm

Hi Alan,

What are your trying to achieve? What's your design intent? With different different design goal you should write your script in different ways which the outcomes could look similar at some stage but would make huge difference when you try to further develop them. 

Did you just want a brick pattern? with mortar gaps ? 

I uploaded a shorter definition for you to look at, but like I said how you do it really depends on what you want to achieve in your design later.

Attachments:

• reeftroubleshoot Response.gh, 15 KB

Permalink Reply by ALan Chan on April 21, 2015 at 10:34pm

I want to achieve the pattern of the Iwamoto Scott - MoMA/Ps1 Reef if you look from the top.The grid shifts like half a cell every row and the entire grid follows a curved/wavy path, not just a rigid grid that runs straight. 

Anyways, thank you very much for your response Chen!!! Much appreciated!

Permalink Reply by Canhui Chen on April 21, 2015 at 10:53pm

I had a look. here are a few suggestions:

Their grid is fluid and dynamic. You need to accommodate this nature from the beginning of the script. If you don't have this flexibility, you will find it extremely painful to do the variations later on for the matrix and you won't be able to get any interesting outcomes.

instead of using those straight lines, try to start your script by making a set of dynamic grids.

you can do 'tween curves' and make two grids curves and then use 'ccx' curve curve intersections to make that point grid. 

Or you can have a normal grid and then deform it by using filed evaluations (take the vector at those points and use them to move the points).

after you have the grid of points just use cull pattern and data management to get the staggering pattern.  after you have the control of the point list you can either shatter the curves of just redraw the lines from them. 

have a go and feel free to post your progress and ask for help again when you are getting stuck again.

 

Permalink Reply by Canhui Chen on April 21, 2015 at 11:06pm

This is what I meant

Permalink Reply by ALan Chan on April 22, 2015 at 6:15pm

Attachments:

• reef troubleshoot 2015 04 27.gh, 38 KB
• reef trouble shoot.3dm, 29 KB

Permalink Reply by ALan Chan on April 22, 2015 at 6:16pm

I have done the it! Getting the pattern on every second row is the hardest because I had to somehow remove the branches so that the loft can start at the second row instead of the first. Am I doing it correct?

Permalink Reply by ALan Chan on April 23, 2015 at 1:06am

Hey Chen, why is the lofted surfaces twist like this? Sorry for the gasshopper file is a bit messy. The definition that made this is grouped in purple.

Attachments:

• B4.gh, 62 KB

Permalink Reply by Canhui Chen on April 27, 2015 at 8:29am

Hi Alan,

Where is the 3DM file? Also I didn't know you replied because I didn't get any email (only got it for your first question)

did you post this on air group discussion? because if you are doing it that way I will get emails whenever you reply.

You problem to me looks like the starting point of the upper curves are different from the lower curves. 

it depends on what you want, but in this case' square to square' you can just find the center point of the bottom curve, move it up along the surface normal at those points (or the average plane normal of your bottom curves, do discontinuity then fit plane through points) and then extrude the squares to the moved points and you will have a pyramid shape. then you can move the plane you made earlier up and trim the pyramid to make the openings.