Clustering | Fabrication File Preparation

In this post, clustering and file preparation for laser cutter will be discussed. There will be another clustering section for support structure as well but the base is similar.

Clustering

Clustering is the process of categorising data into different lists. There are different methods of clustering such as: Machine learning (K-Mean and Gaussian mixture) and Row-Column. below is a chosen generation which is clustered. Each row is categorised in one different list and each list has a different colour. It is due to the order of each pedal and row for assembling.

Additionally, Each row has a number starting from 0 at the bottom to 12 (in this example) at the top. In other words, they are numbered based on their Z location and they need to be installed in this order since assembler can’t access inside the model if it is in reverse direction. Moreover, each pedal in a row has a specific number. So, the first pedal in first row is numbered ‘0;0’ and the first pedal in row number two is numbered ‘1;0’. By the time of assembling, each pedal will be plugged based on their numbers.

File preparation for fabrication

Since Pedals are in 3D mode, they needed to be planar so laser cutter can cut the pattern. Component ‘orient’ was used for mapping pedals from their 3D location onto XY plane. First image below shows pedals which are oriented and they all kept their cluster details. Second image shows a closer view and how hexagons kept their characteristics.

Since each hexagon is different in its form, there was a need for a code to produce pattern for each pedal. In this code, curves are produced based on pedals numbers which will be engraved so in the time of assembly, each pedal will be in its right location. Additionally, Curves are divided into two different groups (colours) for preventing material from melting in the time of laser cutting.

Now that pedals are ready, next step is nesting and cost estimation.