ML_bne
Using machine learning methods to predict Brazil Nut Effect output.
Simulation
The data input of the model are obtained from simulation of Brazil Nut Effect simulation based on molecular dynamic method.
The demo can be accessed here: https://raw.githack.com/miqbalrp/ML_bne/main/simulation/src/bne_single.html
Previous Studies
Some of related research:
Kesuma et al 2016: "As the system vibrated at Γ = 5, the intruder rise time observed as a function of contactopy from the initial configurations. It shows that the rise time in average tends to decrease as the number of contactopy increases." https://iopscience.iop.org/article/10.1088/1755-1315/31/1/012001
Ain et al 2016: If the top of the bed is already HCP, BNE can be blocked (the intruder can no longer rise). https://iopscience.io.org/article/10.1088/1742-6596/739/1/012135Putra et al 2019: The network of the bed changes as the intruder rises. https://iopscience.iop.org/article/10.1088/1757-899X/546/5/052057
Clamarra et al 2006: Relation between density, diameter, and Γ to BNE/RBNE. https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.96.058001
Work Report
2021/05/30
- Simulation modification to perform determined (non-random) variations in vibration frequency, bed density and intruder density with fix initial position of system.
- Simulation: https://raw.githack.com/miqbalrp/ML_bne/main/simulation/src/bne_single_rho.html
- Initial position:
- Densities are varied from 700, 720, 740, ... 880
- Frequencies are varied from 5, 10, 15, ...
- Relation of intruder and bed density with normalized acceleration can be shown below (for f=5,10,15):
