manelippert/bachelorthesis-shearingrate-convergence

Bachelorthesis about the Topic of Size Convergence of the ExB Staircase Pattern in Flux Tube Simulations of Ion Temperature Gradient Driven Turbulence

bachelorthesis-shearingrate-convergence's Issues

Brief Communications in Physics of Plasma

-> : corrections
-->: corrections reviewer

Preperation

• Look up style of brief communications
• Make a mindmap how to elustrate the results
• Read some brief communications

Content

Titlepage

• Title, Author
  -> new title for stand alone paper
• Add github repo in notes
• Title: "Convergence of shearing rate ExB wavelength with boxsize in gradient driven simulation"

Abstract

• Write short form of the paper
• --> Remove Digits from Mesoscale

Introduction

• Work so far in paper [1, Abstract, Section I, Section IV, Take thoughts from end of Section VI]
  -> the paper should be an standalone paper so motivated more why this is intresting as subject
  -> shearing rate linked to propagation of lawine
  -> GISELA work (?)
  -> length scale important for physics
• Recap work from [1]

Theory

• shearing rate theory (only Formula)
• fourier trafo of data
• note that all modes of wexbmax has to be almost zero for stabil result [additional graphic (?)]
  -> One mode has to be maximal e.g the dominate wavelength zonal mode
  -> More important is that a structure is developed even within the turbulence

Methods

• Methodes and Used Resolution for adiabatic electrons [1, Section II]
• GKW
• S6 Standard Resolution in 6th Order (Nvpar = 48)
• Nummerical disspation -> [1, linked in sidenote (?)]
• geometry and gradient length
• explain variables (normalized time units etc.) [1]
• --> Include collision into account with paper from Arne Weikl 2017

Discussion of Results

• wavelength of shearingrate with boxsize [wait for 4x1 (?)]
• explain raidial movement for better visiability in plot
• long runs can still result in stable modes
• wavelength of boxsize 3x3
  -> do not argument with the wavelength it more than enough with same scale
• further binormal increased box sizes
• show with plot of wexbmax that staicase structure has to be fully developed [call back 1 to solidify thesis of waltz role]
• note the fast stabilzation for binormal increased boxsizes
• --> include 1.5x1.5, 2.5x2.5 into isotropic box scan
• --> statistical description of 3x3 for different initial conditions (default cosine2, additional noise, cosine5)
• --> diagnostic of density and energy for temperature gradient

Conclusion

• wavelength converges with boxsize
• fully developed staircase structure shows when simulation is stable
• boxsize well chosen
  -> scale length converges so not the box size is well chosen more the scale length
• mention more the structure
• --> Remove Digits from Mesoscale

Additions

• References

Writing

• Write function to count words including graphics in latex (max_value = 3500) and pages
• Exclude comparison in gradient length when not needed
• Link to paper [1] for everything more specific
• -> longest wavelength zonal mode not used correctly -> dominate wavelength zonal mode in Fourier spectrum
• -> remove ref index and just say that ones in the paper -> which reference?
• -> change $j$ to $n_\mathrm{ZF}$ and $k_j$ to $k_{n_\mathrm{ZF}}$

Plots

• Color wavelength right....
• -> Change $k_i$ to $k_j$
• Change Plot heigth
• Make sure graphics can be readable and printed in black and white
• Share axis in subplots
• -> Plot 3x1 (---) in radial turbulence plot
• -> Remove selection plot from paper
• -> Shearing rate plot of binormal increased boxsize
• -> Shearing rate plot of binormal scan
• -> Put shearing rate plots in one subplot graphic with legend on the left of the 3 times boxsizes
• -> Change colors in shearing rate plot
• -> Change $t_{4\times 1}^R$ into $4\times 1$
• -> Rename figures for paper with (1, 2, 3, ....) like they appear in text

Review

• Write answer for the referees as LaTeX

Start x2go Session

Use x2go for remote use of university computers.

Make rlt = 6.2 with 2x2

Rlt = 6.2 is nearby the 'finite heat flux threshold' and we want to show if zonal flows will get generated

New File Appendix

• Convert all existing files from boxsize1 to boxsize1x1
• Remove grid from foldername
• Redo pictures with new file appendix
• Dynamic filename for pictures

Write Shearing rate, zonal flow potential into data.h5

Bachelor Thesis

ToDo

Plots

• Redo all pictures in new style and axis and increase font size -> include label function in plot function?
• Add linear growth rate into wexb plots

Writing

Declaration

• Declaration regarding brief communication

Dedication

• Create list of people
• make picture of Blacky (my cat)

Abstract

• Add work from thesis
• German Version

Introduction

• Rename into Motivation
• Reread introduction for minor corrections

Theory

• Write about ITG Turbulence and Zonal Flow generation

• Geometry (Hamada)

• Shearing rate -> Brief Communication

• Heatflux threshold and Heat Conduction Coefficent

Methods

• Include standard box size parameter in setup chapter as well statement for theory implemented in GKW
• More about GKW and Resolution used (change Nvpar = 64)
• The cluster
• Guide for slurm_monitor.py

Results

• Resoltion part with plots
• Increased boxsize scan -> brief communication
• Heat flux threshold -> brief communication

Closure

• Add rewind of work on python script and resolution

Additional

• Include Brief Communication in Bachelor Thesis and in Appendix (source file)
• Literature
• Reference file needs rework

Additional Diagnostics

Activate for the cases in #42 the diagnostics:

• xy_kyzero_dens --> s[0] is relevant
• xy_kyzero_ene_perp
• xy_kyzero_ene_par

h5py split data

write pyhton script wich splits the data.

Run gkw code

Learn to run gkw code and understand its parameter used in input.dat

Look up fourier spectrum of s6 data

Statistical validation of Box Scan

Use 3x3 box size and variate the initial condition:

• finit = cosine2(default), noise, cosine5
• amp_init = 1000

As inspiration use paper from Florian Rath 2022

Increase boxsize in y domain

y domain

• 2x2 -> Convergence
• 3x3 -> Convergence

VS Code Remote Server

Establish Remote Server Connection with VS Code for

ssh [email protected]
ssh [email protected]

Mirror repository to Bitbucket

Comparison Plots

Near finite heat threshold

• Rlt = 6.0 and Rlt = 6.3 (Ns = 16, Nvpar = 64, Nmu = 9)
• Rlt = 6.0 and Rlt = 6.2 for boxsize2x2

Resolution

• Ns = 12/16 (Nvpar = 48/64, Nmu = 9)
• Nvpar = 16/32/48/64 (Ns = 16, Nmu = 9)
• Nmu = 6/9 (Ns = 16, Nvpar = 48/64)
• dtim = 0.020 and dtim = 0.025

Boxsize

• boxsize2x1 and boxsize2x2
• boxsize3x1 and boxsize3x3
• boxsize 1x1, boxsize2x2, boxsize3x3
• boxsize1x1, boxsize2x1, boxsize3x1 and boxsize4x1 as an Subplot with each boxsize

Shearing Rate

• Plot with different boxsizes that expands
• Plot with turbulent modes
• Plot vor boxsize3x3 with 4 times wavelength

Plots

• Create new Subplot look -> Done but individual plots needed for publication
• Increase fontsize
• Discuss Layout

Redo all pictures with daynamic file name

Upload data back to server and cut it down

Data will be from now on on the server and will be cutted only to the pieces that are needed.

Increase Gradient Length for binormal boxsize

simulations for boxsize 3x3 with

• rlt = 6.2 -> Convergence $k_3$
• rlt = 6.4 -> Turbulent -> Consistent with results of [1]

The goal is to determine the if the step down in the heat flux does increase for bigger gradient length.

Redo plot function mean_shearing_rate

Make a subplot of the spectrum with a certain stepsize and remove mplitude plot (rendundant)

Do this dynamically!

Latex-Files Bachelor Thesis

Learn vim

Learn vim shortcuts and navigation

Seperate Notebook into data folders

• for each dataset a notbook for itself
• Link into README.md with pictures

Literature

Host GitHub Repository

Compile gkw code

Try to run gkw code on local machine with testfiles

Write Shell/python script to restart gkw

To Fix

• Clean Script
• Generalize function
• Increase documentation
• WATING -> WAITING
• Change Backup Routine and output-type
• No ERROR-Mail
• Add Progess-Bar Output
• Change Outline (?)
• Add parse into file
• Move nohup to screen -> Both possible
• Add runtime progressbar (?)
• Add job name in start message
• define jobscript right
• write status.txt in file
• Add print() statement
• Add jobinformation in progressoutput
• Fix space of message in table
• Fix output of compiler in verbose mode
• Fix append mode in status.txt
• Fix FDS.dat not found but simulation is running
• Fix multiple CONTROL outputs

Additional

• Make local backup of folder with option 'local'
• Scan for Restart files DM1, DM2 -> Function Florian
• Scan for Zeros in slurm.out at the end of the file
• Scan output.dat for Run successfully for the next step to validate that a run is successful
• Increase default refresh rate to 300s
• Fix spacing for nodes in table
• Investiagate write process of multiple files -> Increased refresh rate fixed problem
• kill option for monitor process
• Continue writing status.txt file
• Rewrite progressbar -> Show basic infos instead of nothing
• Write Error Message with slurm-XXXXXX.out
• Check for DM2/DM1 for Timesteps -> Install Dependencies automatically for this feature (?)
• Create Classes to declutter code (?) -> classes do not get applied in this case of script
• Check close option for gkwdata.h5 before restart
• Check time stamp of FDS and gkwdata.h5
• Check Dumpfiles for Timestep

Create Standard Resolution 6th order (S6)

Recreate results of [1] for (S6) with rtl = 6.0 and rtl = 6.3

Write data auf NAS - TPV Peeters

• Data to NAS
• data.h5 for certain
• note to delete gkwdata.h5 if necessary, but with clear instruction what data will get lost
• include hdf5_extract in note message
• clone repository
• copy Ns12 and Ns16/Nvpar16 from local disk to NAS
• write shearing rate, zonal flow, max amp into data.h5
• Sync FDS restart file
• Upload created data as .csv
• Git Pull on machine in university
• Clean server directory
• Write new data to NAS
• Data 3x1.5 is broken (local 23000 time step on server 20000) which version? The unbroken file

Find optimal resolution

Start with Full-Time Bachelor Work

Notify Arthur Peeters and Florian Rath when I'm ready to start full-time with the work on the bachelor thesis and with the weekly meetings

```python``` script for evalution of h5-data

Generate Automatic Changlog

Create Plot for second long mode

get fourier plot if fft with index 2 for that rewrite function def get_max_shearingrate(wexb, time) in zonalflow.py

Increase boxsize

radial domain

• 2x1 -> Convergence $k_2$
• 3x1 -> Convergence $k_3$
• 4x1 -> Convergence $k_4$

isotropic domain

• 1.5x1.5 -> Convergence $k_3$
• 2x2 -> Convergence $k_2$
• 2.5x2.5 -> Convergence $k_3$
• 3x3 -> Convergence $k_4$

binormal domain

• 3x1.5 -> Convergence $k_3$, $k_4$ -> Leads to the question if box size should be smaller
• 3x2.5 -> Convergence $k_3$, $k_4$
• 3x5 -> Convergence $k_4$

Read Lectures of Wolfgang Suttrop

• Einführung in Plasmaphysiks
• Einführung in Fusionsforschung

Login Linux Account and Command line

Automatic Login with ssh

Look up staircase structure for increased boxsizes

Publish Repository

• Fix Format and Linking
• Add correct links to papers (Not planned)
• Remove Suttrop lecture
• Add back to top link (not planned)
• Add literature folder to .gitignore
• Fix grafic

Comparison between dtim 0.025 and dtim 0.020

Write Journal into Markdown

• Write Days into journal
• Make a calender in markdown and put under a drop down menu in README.md

Clone gwk

Clone gwk from Bitbucket and link it in the repository