MOLLER GEANT4 Simulations

I am going to suggest chapters and possibly sections (an outline) for the MOLLER GEANT4 documentation. People will need to provide the text for the different sections. People should also feel free to suggest a different outline or set-up.

Overview

MOLLER

DOxygen Documentation

Getting Started

Quick start (see important details below):

In a terminal window (on ifarml6, for example) type:

mkdir ~/scratch

mkdir ~/scratch/ROOTfiles (you may actually want to make this a link to somewhere with more space)

setenv MOLLERGEANTDIR <directory where you would like to have the simulation/moller12gev>

(maybe in /w/halla/parity/disk2/)

cd <directory where you would like to have the simulation>

svn checkout https://jlabsvn.jlab.org/svnroot/moller12gev/mollersim/trunk moller12gev

cd moller12gev

source env_jlabcue

gmake clean

gmake

cd analysis

gmake clean

gmake

cd ../

change number of events in run_ee.mac to <10k and type bin/Linux-g++/moller_sim prerun.mac run_ee.mac

change analysis/inifiles/sample.ini to point to rootfile that is eventually created in ~/scratch/ROOTfiles

(may need to change number of events to match run_ee.mac)

analysis/mollerClass analysis/inifiles/sample.ini

eog plots/moller_myr.gif (for example)

Getting the simulation

In a terminal window, type:

svn checkout https://jlabsvn.jlab.org/svnroot/moller12gev/mollersim/trunk moller12gev

Get the field maps:

Proposal: upstream, hybrid

Default: upstream, hybrid

-OR- link to some central location, on the farm machines:

The defaults are currently (both in cylindrical coordinates):

ln -s /w/halla/parity/disk2/jmammei/tosca_stuff/MAPFILES/blockyHybrid_3.0.txt fieldmap_back.txt

ln -s /w/halla/parity/disk2/jmammei/tosca_stuff/MAPFILES/blockyUpstream_1.1.txt fieldmap_front.txt

Setting the environment variables

You'll have to make sure your environment variables are set correctly. If you are running the simulation on your
own machine, for now you'll have to set the environment variables yourself, but be sure to also change the version
of GEANT in the local_moller_settings.dat file.

On ifarml1 or ifarml6, type:

source env_jlabcue

Batch Mode

To run in batch mode:

Create a scratch directory in your home directory with a link to a ROOTfiles directory

Create your own .mac file or modify (or modify/use run_ee.mac, run_ep.mac or run_ine.mac)

Type, for example: bin/Linux-g++/moller_sim prerun.mac run_ee.mac

Note: The default .mac files are intended to run on the farm using jsub, but to test you can change the number of
events to 10k instead of 100k, for example.

Interactive Mode

Right now the visualization works pretty well. It uses OpenQt (thanks Wouter!).
I haven't played with it myself much, but you can use the mouse to change views and
you can enter commands in the gui, or use the help to click the commands.

To use the visualization on the farm machines, in the simulation directory:

Type:

bin/Linux-g++/moller_sim prerun.mac

To exit the interactive mode, type:

exit

Note: To use the visualization, at least from a Windows machine, I had to make sure my X-server was compiled

with the OpenGL libraries. I used Xming-mesa, and it works well.

Running on farm

If you want to run multiple simulations at the same time, or even one simulation with more than about 10k events,
you'll want to use the JLab Auger. There is a batch submission script in the batch directory that will write the
necessary files and submit the jobs. The script performs all the steps you could do by hand (see below).

If you haven't already, create a scratch directory in your home directory with a link to a ROOTfiles directory

To submit a job to the farm "by hand":

Create your own .mac file (or modify/use run_ee.mac, run_ep.mac or run_ine.mac)

Create your own .csh file (or modify/use runsim_ee.csh, runsim_ep.csh or runsim_ine.csh)

Create your own .jsub file (or modify/use eesim.jsub, epsim.jsub, inesim.jsub)

Then type, for example:

jsub eesim.jsub

Note: If you want to receive email notification of the completion of your job(s) you have to change the line

in the jsub file:

MAIL: user@jlab.org -> your username at jlab or another email address

Look at output

From moller12gev directory:

If you haven't compiled the analysis code yet:

cd analysis

make

cd ../

Then:

Modify sample.ini file in the analysis/initfiles directory to have the name(s) of the rootfile(s)

From terminal: ./analysis/mollerClass analysis/inifiles/sample.ini

The output and plot files should be created in your output/ and plots/ folders

--> see more details in the README file

Geometry

Currently the geometry is not very sophisticated. It consists of the hydrogen of the target, with no aluminum cell.

Target

Maybe Anna could work on improving the target geometry?

• Add target windows

Beamline

Right now there is no beamline in the simulation; the plan is to transport the scattered particles in vacuum; the
details will come as we decide whether to place the coils in the beamline or in a petal beamline.

Collimators

The collimators can be defined in a collimator geometry file which can be specified in the prerun.mac file. The default
collimators are defined in svn as the default_coll.dat in the geometry directory.

Spectrometer

Currently there are no coils in the simulations, but the fields are present. The field maps can be specified in the
prerun.mac file.

Detectors

Right now the detectors can be defined in the collimator geometry file which can be specified in the prerun.mac file.

Benchmarks of the Simulation

Return to MOLLER at 11 GeV E09-005