Modeling of a leaf structure using Dynamo and TSpline

A model is created taking inspiration from the aesthetics of the structure of a leaf. The general modeling process and critical challenges I came across during coding are also outlined in this article.

(Above pictures are cited from https://bybio.wordpress.com/tag/water-lily-reproduction, https://commons.wikimedia.org/wiki/File:Green_leaf_leaves.jpg)

1.Initial Personal Sketch / Idea

2.Modeling process

First, outline curves are determined as below;

Then, by using these outline curves a reference surface is created;

The leaf is separated into two sides which will later have different patterns. The division is made by locating the midrib and vein curves on the surface;

After all the basic geometry above is ready, three scripts with different methodologies are produced to model;

·        side 1 - the veins,

·        side 2 – the quadrilateral part of the blade,

·        side 2 – the triangular part of the blade

The locations and results of all parts are as below;

In the final script, all the above parts are presented together with the midrib and the two parts of the blade.

3.Challenging issues

A characteristic of an organic form is smooth edges and curvy surfaces. These can be achieved using an option in the TSpline function node if a proper meshing has been done in advance. However, the smoothing process can take a long time and requires lots of computer memory. For this reason, I split out each of the blade patterns to run in different scripts and output into three different .tsm files, recombining in the final script.

In the final script of this model, there are five pieces of TSpline surfaces produced separately in the same view. Unfortunately, this is not the final result though as this is not a single one-piece watertight TSpline surface. To combine all pieces, some refining border edges and welding process is still required, which I will address in the future.

An interesting challenge in this model has been creating the uniform and randomised patterns on the same surface. To do this required extensive coding to achieve.

Another challenge was synchronising the panel distribution to give a consistent shape on the two layers of the pattern, and improve the aesthetic characteristics with the holes aligned.

This article hopefully demonstrates that creating a model of a structure with biomimicry is feasible. Further to this it has been prepared with no packages required for the coding.

This article and the ideas contained within are the work of the author only and are not related to any group or consultant. The ideas expressed are illustrative and for information only.