JP Geant2004

67 %
33 %
Information about JP Geant2004
Education

Published on January 4, 2008

Author: Samantha

Source: authorstream.com

Simulation of Particle Cascades for the Acoustic Detection of Ultra High Energy Neutrinos:  Simulation of Particle Cascades for the Acoustic Detection of Ultra High Energy Neutrinos Jonathan Perkin University of Sheffield Acoustic Cosmic Ray Neutrino Experiment Overview:  Overview ACoRNE What we require from G4 and how we use it Up to what energy can we trust G4? Computation time and I/O Summary And finally… simulation visualisation Acoustic Cosmic Ray Neutrino Experiment Acoustic Cosmic Ray Neutrino Experiment:  Acoustic detection of ultra high energy (UHE) neutrinos Acoustic Cosmic Ray Neutrino Experiment Acoustic Cosmic Ray Neutrino Experiment It is possible to detect the interaction of a UHE (>1021eV) neutrino via the particle cascade induced by a deep inelastic scatter in seawater. Experimentally we hope to push the energy threshold down to ~1019eV A quasi-instantaneous deposition of thermal energy, with a gaussian radial distribution will produce a distinctive bi-polar pulse. Acoustic Cosmic Ray Neutrino Experiment:  Acoustic detection of ultra high energy (UHE) neutrinos Acoustic Cosmic Ray Neutrino Experiment Acoustic Cosmic Ray Neutrino Experiment For E = 1020eV, 95% of the cascade energy is contained within a cylinder of length 20m and radius 20cm. The energy deposition can be considered as a continuous distribution of individual heating centres Radiation is emitted coherently along the cascade axis – leading to a confinement of the signal to a narrow pancake due to a superposition of wavelets. This is analogous to the diffraction of light through a narrow slit Why do we need GEANT4?:  → To simulate the particle cascade Acoustic Cosmic Ray Neutrino Experiment Why do we need GEANT4? A complete understanding of the geometry of the particle showers is required if we are to successfully simulate UHE neutrino interactions. The pressure wave t location r as a function of time t is given by: Where E(r’) is the thermal energy. de/dz + de/dr = acoustic pulse How do we do it?:  High Energy Calorimetry Acoustic Cosmic Ray Neutrino Experiment How do we do it? Assume, that following a UHE neutrino interaction, we have one energetic lepton or hadron that is the precursor to the particle cascade. This will carry roughly ¼ of the original neutrino energy. (cascade induced by a 1GeV pi+ in pure water, volume = 0.06m × 0.06m × 10m. High energy calorimetry physics lists used.) How do we do it?:  High Energy Calorimetry Acoustic Cosmic Ray Neutrino Experiment How do we do it? Although a neutrino interaction such as the ones we are modelling takes place in seawater we are restricted to simulations that run in pure water containing salt atoms. There is currently no facility in G4 that allows for a detector composed of a liquid with free ions in solution. Development of a 1TeV e- shower with (a) and without (b) salt in the water a b How do we do it?:  High Energy Calorimetry (contd…) Acoustic Cosmic Ray Neutrino Experiment How do we do it? Using G4VSteppingAction we bin the total energy deposited at the halfway point of the step into a 3-dimensional array. x-y energy plot (summed over z bins) x-z energy plot (summed over y bins) float energyDeposit[nxBins][nyBins][nzBins] High energy calorimetry Physics List:  What is the highest energy we can trust G4? Acoustic Cosmic Ray Neutrino Experiment High energy calorimetry Physics List The QGSP Physics list uses: High energy calorimetry Physics List:  What is the highest energy we can trust G4? Acoustic Cosmic Ray Neutrino Experiment High energy calorimetry Physics List The QGSP Physics list is valid only up to 100TeV (105GeV). In order to produce simulations beyond this energy one has to modify the physics lists: G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSCPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100TeV G4QGSPNeutronBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPPiKBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV G4QGSPProtonBuilder.cc: theModel->SetMaxEnergy(10*PeV); // 100 TeV Increase maximum energy to 10PeV… High energy calorimetry Physics List:  What is the highest energy we can trust G4? Acoustic Cosmic Ray Neutrino Experiment High energy calorimetry Physics List …code runs but becomes unstable for hadronic interactions: GetHadronicProcess: no model found for this energy range At PeV energies the cross sections for pair production and bremsstrahlung are suppressed because the interaction length becomes larger than the atomic spacing. This causes the electron to undergo multiple scatters and elongates the shower → Landau-Pomeranchuk-Migdal (LPM) effect. - this is not yet fully implemented in G4 (the cross sections are I included for electron bremsstrahlung but not gamma conversion) Modification of the shower shape has a direct influence on the shape of the accompanying acoustic signal Re: UHE Hadronic Showers (Jonathan Perkin) Keywords: high-energy Date: Thu, 12 Feb 2004 14:31:17 GMT From: michel maire <michel.maire@lapp.in2p3.fr> User Jonathan Perkin wrote: >> Hi, >> >> I am interested in simulating EM and Hadronic showers upto energies of >> 10^17 - 10^21 eV for the incoming particle (e- or pi) in water. For now >> the EM showers seem to work ok with the high energy calorimetery physics >> lists (but the computation time is very long - 200 hrs for 10^17 eV!). I >> am not sure about LPM effect though? >> LPM is included in electron Brems, not yet in GammaConversion. See Physics Reference Manual. Notice that in V6.0 all physics tables have maximum bin at 100 TeV by default. For your application, you must add in PhysicsList : G4LossTableManager::Instance()->SetMaxEnergyForMuons(1000*PeV) Re: UHE Hadronic Showers (Jonathan Perkin) Keywords: high-energy Date: Mon, 16 Feb 2004 14:33:25 GMT From: Wellisch, J.P. <hpw@geant4.com> Hi, to get to the energies you would like to have is non-trivial. We will need to include semi-hard scattering, mini-jets, and hard matrix elements, etc.. We do not plan in the short range to address all of it. Many greetings, Hans-Peter. Physics Lists:  What would we like from G4? Acoustic Cosmic Ray Neutrino Experiment Physics Lists As mentioned, we are forced to consider the development of the particle shower from one energetic primary particle with ~¼ E. We need set of physics lists defined well beyond 100TeV that with clear statements about their range of validity. We would like to be able to model the neutrino interaction and thus generate a distribution of excited leptonic and hadronic states as progenitors of the particle cascade. Development of high energy neutrino interactions is currently limited to atmospheric interactions for use in extensive air shower (EAS) simulations. e.g. CORSIKA. In the mean time we use an existing toolkit to simulate the neutrino interaction and feed the primaries into G4, modelling the progression of separate sub showers? Computation time and I/O:  Energy of primary particle versus computation time Acoustic Cosmic Ray Neutrino Experiment Computation time and I/O Behaviour of simulation time with respect to the energy of the ‘gun’ particle is linear → increase particle energy by factor of 10, simulation time increases by factor of 10. ~2 weeks for a 108GeV electron shower. Computation time and I/O:  How to store the data?.... Acoustic Cosmic Ray Neutrino Experiment Computation time and I/O We want to be able to store the contents of each bin in our 3D array. We have a few options for the format of the data and how much we store: ASCII (plaintext) This allows for the energy content of the whole array to be saved. The file is written in full at the end of the simulation run. There is one entry per bin corresponding to the energy of that bin. Computation time and I/O:  How to store the data?.... Acoustic Cosmic Ray Neutrino Experiment Computation time and I/O ROOT ntuples: The ROOT ntuple has the following structure: ntuple = new TNtuple("ntuple","energydensity","x:y:z:energy"); The ntuple is filled at every instance of the G4VUserStepping function with the x,y,z coordinates of the step centre and the total energy deposited. Computation time and I/O:  How to store the data?.... Acoustic Cosmic Ray Neutrino Experiment Computation time and I/O ROOT histograms Geant4 has can link directly to the ROOT analysis toolkit. Hence ROOT objects can be created within a simulation. One can significantly reduce the I/O load on the network and improve computation time by generating ROOT histograms `on the fly’, retaining only the desired distributions, e.g. de/dr and de/dz. However the raw shower data is not retained. If one wishes to plot a new distribution, it must be reconstructed from the existing plots (not always possible) or the simulation requires recompiling and re-running. Computation time and I/O:  How to store the data?.... Acoustic Cosmic Ray Neutrino Experiment Computation time and I/O Performance stats for a 100 TeV electron shower… wanted to include the I/O load on the network here but it is broken!  Summary:  We are interested in the thermal energy density occurring from a UHE (>1021eV) neutrino interaction in seawater We are pushing Geant4 to the limits in terms of energy We use the QGSP high energy calorimetry physics lists, there is no plan to develop these beyond 100TeV at present. Some imporant process are not included → LPM effect We cannot model the neutrino interaction and so are limited to the assumption that one energetic particle represents the exited hadronic or leptonic progenitor of the shower The simulations are very CPU/time intensive ~2weeks at 108GeV discounting delays from network I/O The ROOT analysis toolkit has been integrated with the cascade simulation. Acoustic Cosmic Ray Neutrino Experiment Summary And finally…:  Simulation visualisation Acoustic Cosmic Ray Neutrino Experiment And finally… Visualisation of the response of a hypothetical array of 1260 hydrophones.

Add a comment

Related presentations