Computational projects in Accelerator Physics

[ersa17]

Hello. I am not sure if this question is relevant to be asked here but anyway please help me if possible. I have just completed my undergraduate degree in Applied Physics and I am interested to pursue my MS in accelerator physics. Since I haven't taken any course in accelerator physics, atomic physics or nuclear physics(not offered by my university), I have been taking some online courses in Coursera on the basics of particle accelerators.

Now I am interested to do some computational projects in this subject to develop good programming skills and understand the field of accelerators. But every time I search for some computational projects, I end up with the links referring to huge projects in CERN, SLAC, etc. Can anybody refer me to some beginner computational project topics that can be done in accelerator physics? I'd appreciate it if anyone could guide me with this.

 

[ZapperZ]

Hello. I am not sure if this question is relevant to be asked here but anyway please help me if possible. I have just completed my undergraduate degree in Applied Physics and I am interested to pursue my MS in accelerator physics. Since I haven't taken any course in accelerator physics, atomic physics or nuclear physics(not offered by my university), I have been taking some online courses in Coursera on the basics of particle accelerators.

Now I am interested to do some computational projects in this subject to develop good programming skills and understand the field of accelerators. But every time I search for some computational projects, I end up with the links referring to huge projects in CERN, SLAC, etc. Can anybody refer me to some beginner computational project topics that can be done in accelerator physics? I'd appreciate it if anyone could guide me with this.

First of all, you need to be enrolled in a graduate program at a school that support a major or specialization in Accelerator Science. Then, with the blessing of your advisor, you should enroll in one of the Particle Accelerator school that is offered throughout the year (especially over the summer). Most schools do not have the faculty or course offering in Accelerator Science, which is why the particle physics community offers these for-credit courses for students majoring in Accelerator Science. Such school is actually offered in various parts of the world. In the US, it is run by the USPAS.

If you look closely at the various course offering throughout various sessions, you'll see that there is quite an emphasis on the computational aspect of Accelerator science.

Please note that you should not confuse accelerator science with being "particle physics" or "nuclear physics". They are not synonymous! 95% of particle accelerators in the world has nothing to do with those two areas of physics.

If you have not seen this thread, you should also start here:

https://www.physicsforums.com/threads/accelerator-physics-a-field-where-jobs-go-begging.410271/

Zz.

 

[ersa17]

First of all, you need to be enrolled in a graduate program at a school that support a major or specialization in Accelerator Science. Then, with the blessing of your advisor, you should enroll in one of the Particle Accelerator school that is offered throughout the year (especially over the summer). Most schools do not have the faculty or course offering in Accelerator Science, which is why the particle physics community offers these for-credit courses for students majoring in Accelerator Science. Such school is actually offered in various parts of the world. In the US, it is run by the USPAS.

If you look closely at the various course offering throughout various sessions, you'll see that there is quite an emphasis on the computational aspect of Accelerator science.

Please note that you should not confuse accelerator science with being "particle physics" or "nuclear physics". They are not synonymous! 95% of particle accelerators in the world has nothing to do with those two areas of physics.

If you have not seen this thread, you should also start here:

https://www.physicsforums.com/threads/accelerator-physics-a-field-where-jobs-go-begging.410271/

Zz.

I am aware that accelerator physics is not related to nuclear physics or particle physics. I am interested in this field because it focuses mostly on computation, engineering, and physics. I have applied to a few schools in USA and Canada offering an accelerator physics program. Unfortunately got rejected in some. I might be trying again next term but till then I want to get some basic experience in this field. This is why I was wondering if there is some small beginner coding project that one can try to develop some computation skill and at the same time get acquainted with accelerators. Anyway thank you for your reply.

 

[ZapperZ]

I am aware that accelerator physics is not related to nuclear physics or particle physics. I am interested in this field because it focuses mostly on computation, engineering, and physics. I have applied to a few schools in USA and Canada offering an accelerator physics program. Unfortunately got rejected in some. I might be trying again next term but till then I want to get some basic experience in this field. This is why I was wondering if there is some small beginner coding project that one can try to develop some computation skill and at the same time get acquainted with accelerators. Anyway thank you for your reply.

Knowing the methods of finite-element analysis can be very useful if you wish to do some computational study on your own, especially if you have access to various finite element software packages such as COMSOL.

Zz.

 

[ersa17]

Knowing the methods of finite-element analysis can be very useful if you wish to do some computational study on your own, especially if you have access to various finite element software packages such as COMSOL.

Zz.

I did want to learn COMSOL but haven't tried yet. Thanks for your suggestion. I have just done some basic codes in FDM but I should start with FEM now.

 

[JonasKK]

Hi,
First of all, I think it would be good for you to get a better feeling of the field of accelerator physics. Since you aren't (as I understand) enrolled at a school with an accelerator program, I would recommend you to start looking into the basis of accelerator physics yourself. I interpret your interest as being more into beam dynamics (the properties and control of beam of charged particles) and less about all the technological aspects of accelerators (magnets, cavities, vacuum etc.). Please correct me if I'm wrong.

Most accelerator education is done through the large collaborative schools such as USPAS, CERN Accelerator School (CAS) and Joint Universities Accelerator School as ZapperZ correctly pointed out. Here are two links with lectures for the last two editions of the Introduction to Accelerator Physics schools for CAS and JUAS:
https://cas.web.cern.ch/schools/vysoke-tatry-2019
https://indico.cern.ch/event/850755/contributions/
Since you have a undergraduate degree in applied physics, you should be able to attack these topics head on (it is relatively straight forward math). What you want to focus on in the beginning is transverse beam dynamics, then move onto longitudinal beam dynamics.

You specifically mentioned computational projects. If you are indeed interested in beam dynamics, you shouldn't bother with COMSOL. Instead, you should try and play with one of the many accelerator physics codes available [1]. The most widely used codes currently are MAD(-X) (used generally in the proton synchrotron community), Elegant, and Accelerator Toolbox (latter two mostly the synchrotron light source community). These codes are used for designing accelerator layouts ('lattices') and to predict beam behaviour. Accelerator Toolbox is made for MATLAB and (in my opinion) the most accessible of the three codes listed here. A python version is under development though.

[1] https://en.wikipedia.org/wiki/Accelerator_physics_codes

 

[ersa17]

Hi,
First of all, I think it would be good for you to get a better feeling of the field of accelerator physics. Since you aren't (as I understand) enrolled at a school with an accelerator program, I would recommend you to start looking into the basis of accelerator physics yourself. I interpret your interest as being more into beam dynamics (the properties and control of beam of charged particles) and less about all the technological aspects of accelerators (magnets, cavities, vacuum etc.). Please correct me if I'm wrong.

Most accelerator education is done through the large collaborative schools such as USPAS, CERN Accelerator School (CAS) and Joint Universities Accelerator School as ZapperZ correctly pointed out. Here are two links with lectures for the last two editions of the Introduction to Accelerator Physics schools for CAS and JUAS:
https://cas.web.cern.ch/schools/vysoke-tatry-2019
https://indico.cern.ch/event/850755/contributions/
Since you have a undergraduate degree in applied physics, you should be able to attack these topics head on (it is relatively straight forward math). What you want to focus on in the beginning is transverse beam dynamics, then move onto longitudinal beam dynamics.

You specifically mentioned computational projects. If you are indeed interested in beam dynamics, you shouldn't bother with COMSOL. Instead, you should try and play with one of the many accelerator physics codes available [1]. The most widely used codes currently are MAD(-X) (used generally in the proton synchrotron community), Elegant, and Accelerator Toolbox (latter two mostly the synchrotron light source community). These codes are used for designing accelerator layouts ('lattices') and to predict beam behaviour. Accelerator Toolbox is made for MATLAB and (in my opinion) the most accessible of the three codes listed here. A python version is under development though.

[1] https://en.wikipedia.org/wiki/Accelerator_physics_codes

I really appreciate your reply. I am not sure about my particular interest but both beam dynamics and technological aspects seem interesting to me. Haven't gone into depth of any. I have just taken a few courses available online.
About the accelerator physics code, do we need to write actual codes or is it just about playing with different parameters?

 

[JonasKK]

About the accelerator physics code, do we need to write actual codes or is it just about playing with different parameters?

I can mostly speak for Accelerator Toolbox (AT), since that is the main tool for my own work. How it normally works is that one must first define a "lattice", which is the accelerator structure so to say. In AT, there is already implemented all the standard magnetic elements, so one defines several quadrupole, which is characterized by the length and strength of the magnet. Then one defines several bending magnets, again characterized by its length and strength (bending angle). Drift-spaces ("empty spaces") are also defines.
The lattice is then defined as, e.g.: quadrupole1, drift1, quadrupole2, drift2, bending magnet, drift2, quadrupole1... This is the lattice. When one has defined the lattice, there are already functions implemented to calculate all the optical ("twiss" or "Courant-Snyder" functions), emittance etc. of the beam. Essentially, AT can calculate more or less all the properties of the beam directly from whatever lattice you define.

There is a nice little training program for AT, which might give you some inspiration on what to do:
https://github.com/atcollab/ATtraining
Anyway, before you head into playing with AT or any other code, you should definitely learn a bit about the physics and mathematics behind the code first. The CAS and JUAS lectures I posted before also contains lectures on the technological parts, not just beam dynamics. JUAS even has a full course dedicated to the technology of accelerators. Enjoy the reading!

 

[ersa17]

I can mostly speak for Accelerator Toolbox (AT), since that is the main tool for my own work. How it normally works is that one must first define a "lattice", which is the accelerator structure so to say. In AT, there is already implemented all the standard magnetic elements, so one defines several quadrupole, which is characterized by the length and strength of the magnet. Then one defines several bending magnets, again characterized by its length and strength (bending angle). Drift-spaces ("empty spaces") are also defines.
The lattice is then defined as, e.g.: quadrupole1, drift1, quadrupole2, drift2, bending magnet, drift2, quadrupole1... This is the lattice. When one has defined the lattice, there are already functions implemented to calculate all the optical ("twiss" or "Courant-Snyder" functions), emittance etc. of the beam. Essentially, AT can calculate more or less all the properties of the beam directly from whatever lattice you define.

There is a nice little training program for AT, which might give you some inspiration on what to do:
https://github.com/atcollab/ATtraining
Anyway, before you head into playing with AT or any other code, you should definitely learn a bit about the physics and mathematics behind the code first. The CAS and JUAS lectures I posted before also contains lectures on the technological parts, not just beam dynamics. JUAS even has a full course dedicated to the technology of accelerators. Enjoy the reading!

Thank you.