Skip to main content

Topology optimisation

· 2 min read
Marco Pellegrino
Isak Björhag


It’s time for some topology optimization! More specifically multivariable optimization of a parametric timber truss. 🤓

In this example I have created a timber truss with 4 variables that describe the shape of it. This is a very practical way of going about a topology optimization since you can limit the possible shapes due to production limitations to achieve an optimized structure that doesn’t look like a warped alien spaceship 🚀 😊

This parametrization gave approximately 100 000 different shapes which is too many to try them all (brute force). Instead, I used Grasshopper and Galapagos to utilize some machine learning to reduce the time to find an optimal solution.

Watch the video

I also utilized the design groups in the FEM-Design API so that all truss elements had the same section as well as the above and below beam. I also instructed FEM-Design to carry out an auto design of each shape to obtain viable solutions with below 100% utilization of the cross sections. I opted to optimize with regard to timber weight to minimize material, but you can create another fitness function to optimize for price, deflection, CO2 emissions or why not a combination of them all.

🔥 If you are interested in this type of workflow don’t miss our upcoming free webinar on the topic, register here:

🌟 Free trial of FEM-Design and the API:

Watch the video 👇

Watch the video

📝 Download FEM-Design API
📝 Download StruSoft Trial Version
📰 Newsletter
🌟 Community
🙋 Help