The repository contains the code and results for a simulation of a chain of beads. The beads are linked together by a spring and constrainted to a 2D plane. The first bead is pulled with a constant force. The simulations show that some rigidity in the bond angle is necessary in this simplified context to have the chain fountain effect.
I wanted to simulated the chain fountain effect after watching the videos of on the dispute between ElectroBOOM and Steve Mould. You can find the videos here:
ElectroBOOM: https://www.youtube.com/watch?v=hx2LEqTQT4E
Steve Mould: https://www.youtube.com/watch?v=qTLR7FwXUU4
The chain of beads is initialized in a line on the x-axis. The first bead of the chain is at the position (0,0). The beads are connected to the next and previous one in the chain by a spring with constant kl. Initially, all the beads are spaced by the bond lenght d0. At the start of the simulation, the first bead is pulled at a fixed acceleration g downward in the y-axis direction.
In addition, the chain rigidity can be tuned with an harmonic bond angle potential with constant kt between all triplet of connected beads. The motion can be quickstarted by setting the velocity of the first bead vy0 upward in the y-axis direction.
- No ridigidy.
- Starting from a standstill.
- Bond angle potential.
- Starting from a standstill.
- Increased bond angle constant.
- Starting from a standstill.
- No ridigidy.
- First bead has some velocity upward in the y-axis direction.
- Bond angle potential.
- First bead has some velocity upward in the y-axis direction.
- Decreased bond angle constant.
- First bead has some velocity upward in the y-axis direction.
The simulation was performed using PyTorch (https://pytorch.org/) for the automatic gradient. The plots are generated with matplotlib (https://matplotlib.org/). The gifs are generated with imageio.