Simulation for IAXO- and CAST-like GridPix X-ray detectors

This Garfield++ and Degrad based tool can simulate events for X-ray detectors that are based on GridPix. It simulates the absorption of the X-rays in the gas volume, the track of the photoelectron including its scattering, the drift of the secondary electron towards the grid if the GridPix and the gas amplification. The events are stored in the same format as TOS stores frame data, so that the same analysis tools (like TimepixAnalysis) can be used. Additionally, the photon conversion points and the starting points of the secondary electrons are stored.

Requirements

To use this tool Garfield++ and Degrad are required:

Garfield++

Garfield++ can be installed by following the installation guide. If you already have an installation of Garfield++ make sure that you use commit 57ffa3f9 or newer, as there was a bug in the photon transport before.

If you are using CVMFS you can find a compatible Garfield version in the LCG 103 Release.

Degrad

For Degrad you need to put an executable of it into the folder where you run the simulation. You can download the source code here and you should use version 3.15 or newer. Degrad can be compiled with (for version 3.15)

gfortran -o degrad.run degrad-3.15.f

Depending on the version of the compiler and degrad it can run into an error like this:

degrad-3.15.f:2683:55:

 2683 |      /QION,PEQION,EION,NION,QATT,NATT,QNULL,NNULL,SCLN,NC0,EC0,WK,EFL,
      |                                                       1
Error: Rank mismatch in argument 'nco' at (1) (scalar and rank-1)
degrad-3.15.f:2683:59:

 2683 |      /QION,PEQION,EION,NION,QATT,NATT,QNULL,NNULL,SCLN,NC0,EC0,WK,EFL,
      |                                                           1
Error: Rank mismatch in argument 'eco' at (1) (scalar and rank-1)

As a workaround add -fallow-argument-mismatch for the compiler.

Warning: By default degrad emits the photoelectron in a random direction on the perpendicular plane to the photon direction based on a uniform distribution. For the simulation it is needed that photoelectrons are always emitted in the same direction on this plane. The simulation will rotate the event afterwards based on the polarization settings. Therefore there needs to be a small change in degrad (here for version 3.15): In lines 24780 and 34303 delete *R3.

HDF5

For the Timepix3 interface of the simulation HDF5 is needed. Installing the packages libhdf5-dev and hdf5-tools should be sufficient.

Usage

To use the simulation tool it must be compiled in a first step. Therefore create a new folder "build" and change into it:

mkdir build
cd build

Then cmake is used to generate the makefile (make sure that the compiled Degrad executable (degrad.run) is in this folder):

cmake <path to simulation folder>

Then the tool need to be compiled:

make

Now the tool is ready to be used. It has several arguments that define the detector parameters. The tool can be stated with:

./simulation <path> <job> <absorption> <approach> <length> <energy> <gas1> <gas2> <percentage1> <percentage2> <temperature> <pressure> <field> <polarization> <angle_offset> <amp_scaling> <amp_gain> <amp_width> <events> <degrad_output> <tar> <asic>

These are the parameters:

• path defines the path to a gas file which is used for the drift of the secondary electrons. If no path is provided automatically a new gasfile is generated at the start. This can then be used in the future for simulations with the same gas mixture, the same temperature and pressure and the same drift field.
• job defines a number for this simulation. It is used as run number in the output files.
• absorption defines which simulation approach is used for the simulation of the photon absorption. With 0 for each event it is simulated with Garfield. With 1 it it simulated at the start of the program for 100000 photons with Garfield, fitted with an exponential function and then per event the absorption point is drawn from this fit.
• approach defines which simulation approach is used for the simulation of drift and diffusion. With 0 Garfields AvalancheMC is used. With 1 Garfields AvalancheMicroscopic is used and with 2 a monte carlo simulation based on the diffusion parameter is used.
• length defines the length of the drift cylinder in cm
• energy defines the energy of the incoming photons in eV.
• gas1 defines the name of the first gas component. For possible gases see below.
• gas2 defines the name of the second gas component. For possible gases see below.
• percentage1 defines the percentage of gas1 in the gas mixture.
• percentage2 defines the percentage of gas2 in the gas mixture.
• temperature defines the temperature of the gas in Celsius.
• pressure defines the pressure of the gas in Torr.
• field defines the driftfield in V/cm.
• polarization defines the degree of polarization between 0.0 and 1.0 (linear polarization in x-direction)
• angle_offset defines the direction of the polarisation plane in radian
• amp_scaling, amp_gain and amp_width are the fit parameters of the polya distribution for the simulation of the gas gain. If all are set to 0 then just the number of primary electrons per pixel without amplification is counted.
• events defines the number of events that should be simulated
• degrad_output defines if degrad in and out files are stored (1) or not (0)
• tar defines if the events are packed into a tar.gz file (1) or not(0)
• asic defines if the data output is as for a Timepix (1) or a Timepix3 (3). With 0 both outputs are used.

Gases

The following gases are supported:

Name as argument	Secondary name
CF4
Ar	Argon
He	Helium
He3	Helium-3
Ne	Neon
Kr	Krypton
Xe	Xenon
CH4	Methane
Ethane
Propane
Isobutane
CO2
Neo-Pentane
H2O	Water
O	Oxygen
N	Nitrogen
Nitric Oxide
Nitrous Oxide
Ethene
Acetylene
H2	Hydrogen
D2	Deuterium
CO	Carbon Monoxide
Methylal
DME
Reid Step Model
Maxwell Model
Reid Ramp Model
C2F6
SF6
NH3	Ammonia
C3H6	Propene
Cylopropane
CH3OH	Methanol
C2H5OH	Ethanol
C3H7OH	Iso-Propanol
Cs	Cesium
F	Flourine
CS2
COS
CD4
BF3	Boron
C2HF5	C2H2F4
TMA
C3H7OH	N-Propanol
CHF3
CF3BR
C3F8
O3	Ozone
Hg	Mercury
H2S
N-Butane
N-Pentane
CCL4

Get parameters from the gasfile

When compiling the simulation a second script is compiled that can be used to read the drift velocity and the diffusion coefficients from it. It can be used by:

./gasparameters <path>

Here path should be a gasfile.

Planned for the future

• Consideration of z-boost of photoelectrons
• Optional gas gain simulation based on tracking
• Monte carlo of the Grid collection efficiency
• Consideration of absorptions in detector windows
• Support for field maps