Water Flow on Surface


@davidsmavrov, what was the result with the particles from FlexHopper, I wonder?

(David S. Mavrov) #22

Not much different from what I get with the other sollutions. The problem with the direction is solvable. It doesn’t really matter all that much which tool I will use. What I am stuck at is the actual amount of water I will get at a certain rain fall. Then I use a solution very similar to your first post to calculate the pipes needed.

(David S. Mavrov) #23

Actually at this stage this idea turns out to be pretty clever. No matter which solution one choses to use for the flow paths, their end points tend to collect at the same locations. All one needs to do after that is use Cull Duplicates component, to cull the duplicate points, count the number via the Valance output, then calculate what % of the total amount of points is this. And yeah… pretty neat way to estemate the amount of water you will have at a given collector. Thanks man. For now this works very well :slight_smile:


I worked on a project that dealt with this question at the landscape scale. It was called EpiFlow Link to project brief.

The primary metric of concern was peak flow calculated by the “Rational Method” to help size green infrastructure, and predict the impact of development on site hydrology. In that case we spent more time thinking about integrating runoff coefficients and storm size/duration in order to inform storm volumes, but getting the area of the watersheds is a simpler operation.

We aggregated mesh faces that had their flow path terminated in the same location. Using a method like this could tell you drainage area for each of your gutters.

(David S. Mavrov) #25

Amazing! Thank you for sharing! Can’t wait till the weekend so I can have a good look at this!


Is what you’ve done here with the RGB index possible with current GH Components?

I’m after a simple colour map showing gradients across a roof? to see where water might pool? I was thinking simply linking the normal vector of a sampled point to a colour map?

What is the simplest way you can see to do this with GH - or even Rhino?

I can see the potential for water volume being calculated - as you say, by following the flow and adding up all the water per minute / or per second multiplied by the volume overall across the entire roof. That would be a very powerful tool for roofing even in this simplistic approach - that would be adequate for validation.

(David Rutten) #28

Not really, creating bitmaps is impossible without some scripting or plugin. You can create coloured meshes in vanilla grasshopper, but that’s about it.

Still, creating bitmaps by setting pixels is not that hard in C#, it’s the geometric part of the algorithm which takes up all the code.

(Yehezkiel Wiliardy) #30

Theres a recent development by DTU about stormwater runoff simulation. https://github.com/livestock3d/livestock

(David Rutten) #31

That sim even takes time into account, nice.

And they have tests and a code of conduct, I’m very impressed.


What about using something similar to the surface curvature analysis tool to visualise gradient and topology?

(David Rutten) #33

This would be mesh vertex colours. You can do that in Grasshopper. No way to turn it into a bitmap though unless you count screenshotting the viewport.


This project is actually what I was referring to up here:

It’s pretty dope indeed…


Mesh vertex colours worked well and I was also able to calculate rainfall using simple GH components!