Having a bit of fun optimising a wind propeller car with Monolith
Richard Ahlfeld, Ph.D.
Founder & CEO of Monolith | Engineering and Intractable Physics solved with Machine Learning
Hi all, I was looking for new ways to pass the time indoors during the lockdown and thought to myself: I should use Monolith to build something cool. To save budget, I decided to start small: by building the fastest wind propeller car. Below is a preview of what I managed to build in about an hour :)
Step 1: buy some super cheap components (electric motor, some wheels, batteries, etc for example from a science kit for kids )
Step 2: assemble a basic car and attach the motor to it. Then test the simple fan design that was included! Result: not driving. The fan does not create enough thrust to get the heavy batteries moving. Disappointing, but they already mentioned that on the amazon reviews, so no surprise there.
Step 3: Use the Monolith propeller design app and ask the optimiser to find an engine that maximes the thrust! Below a picture of the autoencoder we already have in the platform to show how Monolith can change the design ( we already have this app which is why I selected this use case).
Step 4: Download the STL file of the optimal file from Monolith and have a local on-demand 3D printing service manufacture it. I used https://makeitquick.co.uk/
Step 5: Fit the car with the 3D printed blade and try the driving test again:
Voilà, it moves! :)
This was all still super simple and basic, I am now looking for something more ambitious. Let me know if you have any ideas :)
Mechanical Simulation Engineer, materials modelling expert
4 年That's a great way to demonstrate how the software works! well done.
Unit Director Health, PIABO Communications
4 年super cool. Step 6: Feed that design back to the manufacturer to make all the other kids out there happy as well :)
The power of Monolith: from fighter jets to wind driven vehicles
Professor in Multidisciplinary Design Optimization, Group leader of Sustainable Aerostructures & Interactions
4 年Excellent ! next is lego assembling optimization ;)