phys389-2020-project-CRD1998 created by GitHub Classroom
The simulation attached models a proton bunch in a synchrocyclotron. There are several files required to run this simulation, which have all been detailed below.
- Particle.py - Conatins a class also called Particle, generates an object to represent a subatomic paritcle, such as a neutron. This class also contains methods returning the Lorentz factor, momentum, kinetic energy and methods to update an instances position using one of the following methods: Euler, Euler-Cromer, velocity Verlet, fourth order Runge-Kutta.
- ChargedParticle.py - Inherits from the class Particle, generates an object to represent a charged subatomic particle, such as a proton.
- EMField.py - Contains a class called EMField, this generates an object to represent to an electromagnetic field in a (synchro)cyclotron. It consists of a constant magnetic field and a time-varying electric field in the form Acos(wt+phi) bounded by two x ordinates.
- Bunch.py - An abstract base class that provides the blueprint for an object representing a bunch of ChargedParticle objects, for example a bunch of protons. It samples initial positions and velocities for these particles from a Gaussian and can return the Lorentz factor of the bunch, average position and average velocity. It also contains a method to reduce the time step used in the update methods if any of the particles in the bunch are near or in the electric field, to ensure the electric field is not stepped over.
- ProtonBunch.py - This files contains a class called ProtonBunch, which inherits from the Bunch ABC, it represents a bunch of protons.
- RecordCylotron/Synchrocyclotron.py - This is the main simulation file and should be the only one a user interacts with. This is where the particle bunches and EM fields are instantitated, as well as where the time step, duration of the simulation and theupdate to be used are declared. Once the simulation has finished, a list of the time values is written to an npz file alongside a list containing a copy of the ProtonBunch object at every time value.
To run this simulation, open the RecordCyclotron.py or RecordSynchrocyclotron.py and change the initial average kinetic energy and bunch size to your desired parameters. Changing the initial kinetic energy will likely mean that the electric magnetic field will also need changing. After these objects have adjusted, you can change the time step and length of the simulation.
A flowchart is shown at the bottom of this file to visualise the structure of the simulation.
- analyse_methods.py - First, this file will look for a npz file called "methods_data.npz" which contains data from simulating a proton in a constant magnetic field for approximately 100 revolutions. It will then plot the fractional kinetic energy and momentum of this proton against time. If the "methods_data.npz" is not found, the data file from RecordCyclotron.py will searched for, if this is also not found it will be generated and the same plots will be produced. In the case that the data file from RecordCyclotron.py needs to be generated, ensure you remove the electric field entirely by setting its field strength and width to zero so only a constant magnetic field is present. By only have a magnetic field present, both energy and momentum should be constant throughout the simulation.
- analyse_period.py - This script will first look for a pickled csv named "period_data.csv" which contains a pandas dataframe detailing how many revolutions were measured when simulating a proton in a constant magnetic field. It then averages the measured time period and take ones standard deviation as its error, which it compares to the theoretical time period. If you do not have this csv, it will be generated for you.
- analyse_phase.py - This script simulates a bunch of protons in a cyclotron for approximately 10 revolutions nine times, each iteration has a different phase shift applied to the electric field. It then plots the spread in the y-position and the spread in kineitc energy against time for each phase shift. It first looks for a file containing this data called "phase_data.npz" if you do not have this file it will be generated for you.
- analyse_synchro_phase.py - This script simulates a bunch of protons in a synchrocyclotron for approximately 10 revolutions nine times, each iteration has a different phase shift applied to the electric field. It then plots the spread in the y-position and the spread in kineitc energy against time for each phase shift, it also fits a linear curve to kinetic energy spread data, so the effects of the phase shift are more obvious. It first looks for a file containing this data called "phase_synchro_data.npz" if you do not have this file it will be generated for you.
- analyse_timestep.py - This script simulates a bunch of protons in a cyclotron for approximately 100 revolutions three times, each iteration has a different time step being used in the integrator. It then plots the fractional kinetic energy and momentum against time. It first looks for a file containing this data called "timestep_data.npz"if you do not have this file it will be generated for you.
- plot_position.py - This file looks for both the "cyclotron_data.npz" "synchrocyclotron_data.npz", any files you don't have will be generated for you. It then plots the average position of the proton bunch in the accelerator throughout the simulation. The width of the electric field as well as the direction of the magnetic field is shown on the plots produced. In total it will produce five plots; four cyclotron plots (each plot corresponds to using a different integrator) and a single plot showing the bunch's path at relativistic speeds. Note, that the mesh grid used to show the magnetic field and the shaded region showing the electric field will have to be manually overwritten in plotting code. If you change the width of the electric field for example, the shaded region will not automatically update.
- test_ChargedParticles.py - This contains several unit tests. Ensuring the methods in this class, including the ones it inherits are working as expected.
- test_EMField.py - This contains several unit tests. Ensuring the methods in this class are working as expected.
- test_ProtonBunch.py - This contains several unit tests. Ensuring the methods in this class, including the ones it inherits are working as expected.
A file named "log.py" is also included in this project, this writes to a file named "PHYS389.log" and can be used to check on the progress of longer simulations as well as follow the execution order of processes in the simulation. The data files required to run all of the analysis files have been included. If you wish to generate your own, I recommend reducing the simulation's duration and reducing the number of particles in the bunch to prevent long run times.
There is also a folder called ".vscode" this contains the configuration for discovering unit tests written in the pytest framework and the project configuration files in the ".ropeproject" folder. For this reason I recommend using Visual Studio Code, so the tests can be ran easily.
