Difference between revisions of "DDVCS and TCS event generator"

Revision as of 02:23, 3 June 2016

Principle

This event generator was developed for JLab@12 GeV TCS and DDVCS future experiments. The main options are the generation of:

- "TCS" events with a real photon beam,

- "TCS" events with an electron beam, and a quasi-real photon emitted (default cut is Q2<0.3 GeV2) originating from interaction with the target or from bremsstrahlung in the target,

- "DDVCS" events where the final virtual photon decays into a muons pair,

- "DDVCS" events where the final virtual photon decays into an electrons pair.

These processes also interfer with Bethe-Heitler (BH). Each generated event contains weight for TCS (or DDVCS) cross section, BH cross section, TCS (or DDVCS) + BH cross section. The weights are calculated before the use of the generator and saved in data grids. To avoid strong discretisation effects, a linear interpolation is done between the adjacent bins (5 dim for TCS and 7 dim for DDVCS).

The generator also gives a "weight" for beam and target spin asymmetries (warning: in development). In the case of DDVCS, linearly polarized electron beam spin asymmetry is provided. In the case of TCS, all beam+target combinations of single and double spin asymmetries are provided. These weights are indicative and should only be use to give an estimation of the spin asymmetry for a given event kinematic.

In the case of TCS with a quasi-real photon, two weights are given: the corresponding weight for photoproduction, and the cross section accounting the photon emission.

All cross sections are calculated according to VGG model <ref> M. Guidal (private communication) </ref>. Reference articles for these processes are, for DDVCS: <ref> M. Guidal and M.Vanderhaeghen, "Double deeply virtual Compton scattering off the nucleon" Phys. Rev. Lett. 90 (2003) </ref>, and for TCS: <ref> M. Boër, M. Guidal and M. Vanderhaeghen, "Timelike Compton scattering off the proton and generalized parton distributions" Eur. Phys. J. A 51 (2015) no.8, 103 </ref>.

Kinematics

The generator is developed for some specific JLab experiment, also the kinematic range it handle was choosen according to these experiments. The default beam energy is 11 GeV for an electron beam, and in a range between 5 and 11 GeV for the case of TCS with a real or a quasi-real photon beam. Limits of the generator are given in the user's files.

Default is a proton target. The lenght of the target is an user choice. (other targets in development).

For TCS, the user can choose the beam type (electron or photon) and the range in energy of the photon. The user also defines the range in Q'2 (emitted photon virtuality), t, theta (polar decay angles for the lepton pair). In case of an electron beam, its default energy is 11 GeV. The cross section also depends on phi (azimuthal decay angles for the lepton pair), ranging by default in [0°, 360°].

For DDVCS, the beam energy is 11 GeV by default. The user can define the range in Xbj, Q2 (initial photon virtuality), Q'2 (final photon virtuality), t, theta_cm (equivalent as TCS theta angle). The cross section also depends on phi_LH (corresponding to DVCS angle), and phi_cm (corresponding to TCS angle).

Usage

To run the code: ./DEEPGen (process) process = tcs or ddvcs

input files: TCS.input and DDVCS.input. The user can change any parameter, and define the kinematics and the number of events in ouput file.

cross sections grid: Data/table_tcs.dat, Data/table_ddvcs.dat.

output file: DEEPGen.root. Contains TCS_Tree (only for tcs option), DDVCS_Tree (only for DDVCS option) and Debug_Tree (a verbose mode Tree with all the 500 first events four-vectors and parameters for each step of the generation).

Output

The root format output file contains the four-vectors of the final state particle, where the energy and momentum are provided in the "lab" frame. It also contains the weights of events and kinematic variables.

Main variables contained in the Tree.

TCS

- ALV_minus_lab, ALV_plus_lab: final lepton pair (array, 0=E, 1, 2, 3 = px, py, pz)

- ALV_gamma_in, ALV_el_in (if electron beam): initial particles, photon and/or electron from the beam

- ALV_Recoil_lab: final proton

- ALV_el_out: scattered electron in case of an electron beam

- Q2, yy, phi_beam, theta_beam: virtuality of the initial photon, beam energy loss fraction, scatteting angle of the electron in case of an electron beam and a quasi-real photon

- theta_gamma: polar angle of the quasi-real photon in case of an electron beam

- RotAxis: angle between lab x-axis and reaction plane

- tt = t, Qp2 = Q'2, Egamma: kinematics

- ttmin = t_min: maximal value of t allowed by the kinematic

- CosThetagg: cosine of the angle between initial and final photon

- Phi_CMV and Theta_CMV: angles of the electron from the decay of the virtual photons, between e- direction and the reaction plane

- Flux_qr, Flux_bmr: quasi real photon flux and bremsstrahlung flux for the given kinematic

- W_TCS, W_BH, W_tot_unpol: unpolarized cross section for TCS, BH, TCS+BH

- cross_TCS, cross_BH, cross_tot_unpol: cross section accounting the photon flux

- EventNumber: event number in the file. The last event correspond number correspond to the user request of number of events in a file.

- TrueEventNumber: number of the event in the generation process. The last event gives the total number of events that have been generated in a hypercubical space in kinematic variables to provide in the ouput file the required number of events by the user. Used for normalization.

- Flags: see below.

DDVCS

- Almost the same format than TCS, see above.

- Xbj, Q2...: kinematics

remark: asymmetries turned off for the moment.

Normalization

Use a phase space factor equal to the product of the range of each kinematic variable used to calculate the cross section. Depends on user defined phase space.

Flags

Some flags are given as information to prevent numerical divergencies.

- FlagEdge: only informative, the higher this number is, the closer it is from kinematic limits of the process or of the cross section grid.

- FlagStab: should be = 0, otherwise it indicates possible numerical divergences when calculating the cross sections.

- FlagVar: should be as low as possible. If > 1000, it indicates that the weight is not correct (numerical divergence). Otherwise it indicates that the cross section had strong variation between the given kinematic bin (of the grid) and its N-1 and N+1.

Where to find the generator

The generator is still under developpement, it will be provided once a stable version is available. For any question, please contact: marie@boer.fr.

References

Template:Reflist

• M. Guidal (private communication).
• M. Guidal and M.Vanderhaeghen, "Double deeply virtual Compton scattering off the nucleon" Phys. Rev. Lett. 90 (2003).
• M. Boër, M. Guidal and M. Vanderhaeghen, "Timelike Compton scattering off the proton and generalized parton distributions" Eur. Phys. J. A 51 (2015) no.8, 103.