I am trying to create a floor pattern. I have this line diagram laid out of how i would like small mosaic chips to follow the flow. I need a set to follow the lines exactly and i need another set to fill in the gaps between the lines in a similar direct as the surrounding. My inspiratio to create the pattern is attached below.
this post might help, was a solution posted just now on another topic, its basically the same
thing, very similar, can take this approach and modify it to work for you
1 Like
just a question for better understanding… your sketch looks fundamentally different from the image
in the image tiles have always fixed width, suggesting what could be an offset-based solution
while in your sketch it looks like they have variable width: the very same curved flow is much thicker at one end and becomes progressively smaller on the opposite end, suggesting what could be a tween/isocurve-based solution
which of the two? 
4 Likes
running bond.gh (32.7 KB)
4 Likes
The Pattern i am trying to get is the one with a thicker end which progressively gets smaller on the opposite end.
If this is going to be built, your stonemason is going to hate you! Every tile will have a unique width, and will vary a lot.
When I’ve worked with professional mosaic-makers in the past, they have a set of standard stone sizes that are either square or come in sticks which they cut to length as they lay a line. Constantly varying widths like this would involve a lot of wastage.
I have to ask though: why draw every tile? On a site this big, it will be millions and millions of lines. Rhino will be brought to its knees…
3 Likes
Its true that they would have a specific size to use maybe a 1”x2” chip that they lay but because of hoe intricate the pattern is we want to make sure we get our material combinations and flow of the curves corresponding to the material correct, The way i imagine it is as the width between two lines reduces the number of chips between each will reduce as well instead of having 4 rows the narrow end can just have 1 chip making it more like a fan. Once we have the design finalized we can have each fan produced as a sheet and then laid on site.
Ok, so more like this? (I’ve done squares, but they could be longer, reducing the number of tiles in each row)
In which case, we only really need to generate the “orbital” coursing lines and leave the “radial” joints up to the mason, who can cut the tiles to suit.
So, final question before I ask to see your source file, do the coursing lines bend (red), or are they straight (blue)?
Straight will be easy to lay, curved will look nice but will mean more cut tiles.
(I have to go now, so won’t be able to look at this thread for a while)
The coursing line follow the red pattern
Ok, so now you have to share your source file, or at least part of it…
Flooring Trial.7z (13.0 MB)
Here’s the rhino file with the line diagram in it.
if you don’t want to sweat too much and are open to using plugins, a very fast method could be to use Pufferfish and have all the areas to be composed by exactly 4 edges, like this:
Mosaic_Flow_Pufferfish.gh (16.7 KB)
there is for sure a similar way to do without using Pufferfish (likely using something like sweep2 or network surface + isocurve on normalized surfaces)
1 Like
The output I initially attached in the thread was made using isocurves and surfaces, but I see that using pufferfish would definitely have been a better way to get the same result. Now the next task where I am stuck is is to further divide the individual fans into mosaics following the same curve direction.
with some “poetic license” here and there to correct tiny pieces (which I’m not entirely sure why they do exist in first place… 
) here is a dirty sketch:
Mosaic_Flow_Pufferfish_2.gh (21.0 KB)
1 Like
This is great, I’ll try this method and let you know if it works! I think it should get me very close to my desired result. Thankyou so much!!
Although The final result for fabrication would need the divisions reducing in number as we move from the wider gap to the narrower gap.
I’m just going to come out and say it: this is very stupid! Offset the boundaries as the image exemplifies and @Tom_Newsom alludes to.
In the image and what I am trying are different, to create something like the image I’d offset and then divide it in equal segments which is basically a take on “concentric circles” but with organic shapes. What I want to do is essentially create a fan/shell pattern and fill it in with mosaics all going in one direction. Each fin has 4 sides as @inno realized in the pufferfish method what I am stuck at is reducing the number of mosaics as we move towards the tapered edge, @Tom_Newsom captured that perfectly in the sketch he posted.
Okay, yeah. That does make more sense. However inno’s Pufferfish solution does not lay out the tiles linearly in one direction. The key here would be to find the line of best fit to define the paving direction, so that you get the blue lines from this sketch:
Attempting to follow the red lines is silly.
If they’re concentric and not too tightly curved it’ll be ok. The reference images have gentle curves to them and come out just fine in mosaic. Although it’ll all be othogonal to the main curve. I’m going to mock this up but I get the feeling it won’t be the look @Rohan4 is going for…