Search

Utilizing laser-wakefield accelerated (LWFA) electrons to drive aplasma-wakefield accelerator (PWFA) holds great promise for realizingcentimeter-scale electron accelerators providing ultra-high brightnessbeams. Recent experiments at HZDR could demonstrate for the first timesuch an electron acceleration in a nonlinear PWFA plasma wakefield. Fordriving this compact hybrid accelerator setup, high-charge electronbunches from LWFA self-truncated ionization injection were used.In this talk, we present recent results of the accompanying simulationcampaign performed with the 3D3V particle-in-cell code PIConGPU. Thesesimulations model the geometry, density distributions, laser modes, andgas dopings as determined in the experiments. The simulation conditionsresemble the experiment to a very high degree and thus provide goodcomparability between experiment and simulation. Additionally, thewealth of information provided by the in-situ data analysis of PIConGPU provides insight into the plasma dynamics, otherwise inaccessible inexperiments.From an algorithmic and computational perspective, we modeled the hybridaccelerator from start to end in a single simulation scenario. Wediscuss the associated challenges in maintaining numerical stability andexperimental comparability of these long-duration simulations.

State-of-the-art particle-in-cell simulations are becoming faster in terms of time to solution by utilizing modern hardware accelerators like GPUs and more accurate by improving the underlying algorithms. However, in order to model experiments, methods to include realistic laser pulses and gas distributions as well as efficient techniques to predict experimental observables, so-called synthetic diagnostics, need to be included in these simulations. In this talk, we present extensions to the particle-in-cell code PIConGPU that were essential to accurately model LWFA experiments based on self-truncated ionization injection performed at HZDR. We discuss the significant impact of the implementation of higher order laser modes on the plasma dynamics and the resulting acceleration process. Furthermore, we discuss in detail the advantage of efficient in situ data analysis on the example of studying electron phase space evolution and of predicting spectrally and directionally radiation emission by all particles. These improvements set the stage for quantitatively predicting the results of experiments in the near future.

Exploring new regimes, optimizing experimental setups, or quantifying sensitivity of final beam parameters on experimental parameters, represent current challenges for simulations of laser plasma accelerators. Time-to-solution and scalability are key parameters for codes to minimize turnaround times in order to scan e.g. tens of parameters such as the laser leading edge, resolve solid density target physics and run full-scale start-to-end simulations. PIConGPU reaches unprecedented performance by accelerating 100% of its computations on many-core architectures and leveraging next-generation scalable I/O. High-resolution, full-geometry studies on top-ten listed supercomputers decisively enhance predictive capabilities. PIConGPU's design allows for utilizing various compute architectures, including modern X86 and ARM CPUs and GPUs with a single, adaptable code base. Users can now run PIConGPU on almost any machine, either by easy recompiling or using predefined Docker images, and everybody can download, use and contribute to the code without extensive knowledge in compute architectures. We highlight latest additions to PIConGPU such as scalable file I/O via a new openPMD-API including ADIOS2 support for on the fly loosely coupled data analysis, live visualization with particle and field rendering, non-standard Gaussian laser pulses via Laguerre modes, in-situ X-ray scattering image generation, and an pythonic simulation setup interface.

Extreme field gradients intrinsic to relativistic laser plasma interactions enable compact MeV proton accelerators with unique bunch characteristics, yet complicate direct proton beam control. Only complex micro-engineering of the plasma accelerator itself and limited adoption of conventional beam optics, so far provided access to global beam parameters as direction and divergence. Here we present a novel, counter-intuitive, yet readily applicable all-optical approach to imprint detailed spatial information from the driving laser pulse to the proton bunch. In a series of experiments, the spatial profile of the energetic proton bunch was found to exhibit identical features as the fraction of the laser pulse passing around a target of limited size. The formation of quasi-static electric fields in the beam path by ionization of residual gas in the experimental chamber results in asynchronous information transfer between the laser pulse and the naturally delayed proton bunch. Essentially acting as a programmable memory, these fields provide access to a new level of proton beam manipulation.

Supplementary materials for our paper "Spectral Control via Multi-Species Effects in PW-Class Laser-Ion Acceleration". Additional high-resolution, raw HDF5 files using the openPMD standard (DOI:10.5281/zenodo.1167843) increase simulation output data to 4.7 TByte and are available from the corresponding author upon reasonable request.&nbsp

PIConGPU is an open source, multi-platform particle-in-cell code scaling to the fastest supercomputers in the TOP500 list. We present the architecture, novel developments, and workflows that enable high-precision, fast turn-around computations on Exascale-machines. Furthermore, we present our strategies to handle extreme data flows from thousands of GPUs for analysis with in situ processing and open data formats (openPMD). PIConGPU is since recently furthermore natively controlled by a Python Jupyter interface and we research just-in-time kernel generation for C++ with our Cling-CUDA extensions.

Extreme field gradients intrinsic to relativistic laser plasma interactions enable compact MeV proton accelerators with unique bunch characteristics, yet complicate direct proton beam control. Only complex micro-engineering of the plasma accelerator itself and limited adoption of conventional beam optics, so far provided access to global beam parameters as direction and divergence. Here we present a novel, counter-intuitive, yet readily applicable all-optical approach to imprint detailed spatial information from the driving laser pulse to the proton bunch. In a series of experiments, the spatial profile of the energetic proton bunch was found to exhibit identical features as the fraction of the laser pulse passing around a target of limited size. The formation of quasi-static electric fields in the beam path by ionization of residual gas in the experimental chamber results in asynchronous information transfer between the laser pulse and the naturally delayed proton bunch. Essentially acting as a programmable memory, these fields provide access to a new level of proton beam manipulation.

Laser-driven ion acceleration promises to provide a compact solution for demanding applications like particle therapy, proton radiography or inertial confinement research. Controlling the beam parameters to achieve these goals is currently pushing the frontier of laser driven particle accelerators. We present an overview of recent achievements at the high power ultra-short pulse laser source DRACO at the HZDR in Dresden (Germany). The laser system was recently upgraded by new front end components and an additional Petawatt (PW) amplifier stage, finally providing high contrast pulses of 30J within 30fs at 1 Hz pulse repetition rate. The performance of the plasma acceleration is strongly dependent on the complex pre-plasma formation process at the target front surface which is determined by the temporal intensity contrast. Plasma mirror setups have proven to be a valuable tool to significantly improve the temporal contrast by reducing pre-pulse intensity and steepening the rising edge of the main laser pulse. Re-collimating single plasma mirror devices have therefore been implemented into the Draco laser beam lines, enabling investigation of laser proton acceleration and proton energy scaling within the TNSA regime using ultra-thin foil targets. The results of the simultaneously measured proton emission energies in laser forward direction, laser backward direction and the temporal contrast, measured on a single-shot base by means of self-referenced spectral interferometry with extended time excursion (SRSI-ETE) at unprecedented dynamic and temporal range, will be presented.

Laser-driven ion acceleration promises to provide a compact solution for demanding applications like particle therapy, proton radiography or inertial confinement research. Controlling the beam parameters to achieve these goals is currently pushing the frontier of laser driven particle accelerators. We present an overview of recent achievements at the high power ultra-short pulse laser source DRACO at the HZDR in Dresden (Germany). The laser system was recently upgraded by new front end components and an additional Petawatt (PW) amplifier stage, finally providing high contrast pulses of 30J within 30fs at 1 Hz pulse repetition rate. The performance of the plasma acceleration is strongly dependent on the complex pre-plasma formation process at the target front surface which is determined by the temporal intensity contrast. Plasma mirror setups have proven to be a valuable tool to significantly improve the temporal contrast by reducing pre-pulse intensity and steepening the rising edge of the main laser pulse. Re-collimating single plasma mirror devices have therefore been implemented into the Draco laser beam lines, enabling investigation of laser proton acceleration and proton energy scaling within the TNSA regime using ultra-thin foil targets. The results of the simultaneously measured proton emission energies in laser forward direction, laser backward direction and the temporal contrast, measured on a single-shot base by means of self-referenced spectral interferometry with extended time excursion (SRSI-ETE) at unprecedented dynamic and temporal range, will be presented.

Extreme field gradients intrinsic to relativistic laser plasma interactions enable compact MeV proton accelerators with unique bunch characteristics, yet complicate direct proton beam control. Only complex micro-engineering of the plasma accelerator itself and limited adoption of conventional beam optics, so far provided access to global beam parameters as direction and divergence. Here we present a novel, counter-intuitive, yet readily applicable all-optical approach to imprint detailed spatial information from the driving laser pulse to the proton bunch. In a series of experiments, the spatial profile of the energetic proton bunch was found to exhibit identical features as the fraction of the laser pulse passing around a target of limited size. The formation of quasi-static electric fields in the beam path by ionization of residual gas in the experimental chamber results in asynchronous information transfer between the laser pulse and the naturally delayed proton bunch. Essentially acting as a programmable memory, these fields provide access to a new level of proton beam manipulation.

We present a scalable GPU-based software framework for simulating photon scattering processes of X-ray beams in matter using Monte-Carlo methods. These simulations enable us to predict SAXS signals for experiments at upcoming superlative research facilities like the European XFEL. Often the expected outcome of SAXS experiments is produced by a Fourier Transform of a static 2D electron density distribution. Our new framework provides the opportunity to simulate the probing of femtosecond timescale 3D3V electron dynamics with single and multiple scattering and is extendable by more complex physics processes like laser absorption, atomic excitation and de-excitation to further enhance its predictive capability. As a foundation we use libPMacc, a powerful particle-mesh accelerator library that is also used by PIConGPU, the reportedly fastest fully-relativistic 3D3V particle-in-cell code in the world.

We present our scientific roadmap towards in-situ modeling of non-LTE interactions of XFEL type X-rays with solid density plasmas using a symbiosis of our performance portable, open source, 3D3V particle-in-cell (PIC) code PIConGPU and its X-ray tracing prototype ParaTAXIS. Treating radiation transport via various atomic processes will enable us to synthesize detector signals and gain predictive capabilities for upcoming pump-probe experiments at the European XFEL. With the world’s fastest particle-in-cell code PIConGPU and the raw computational power of the largest high performance computers we open up the possibility for large-scale case studies of unprecedented repeatability.

In laser-generated plasmas the free electron density is a crucial parameter for plasma dynamics. Therefore, to model its spatial and temporal evolution the adequate treatment of ionization is vital. This poster presents the work in progress on numerical field ionization methods implemented in the world's fastest 3D3V electromagnetic particle-in-cell code PIConGPU. Thus, computing a value for the systematic error via repeating simulations with varying ionization schemes is in reach. With high performance computing we can give a range of validity for predictions of pump-probe experiments with high power lasers and X-ray free electron lasers.

Laser-driven solid density plasmas can be used to generate highly energetic electrons and ions. Diagnosing properties within those plasmas at nm length scales and down to fs timescales is crucial in understanding the involved processes. This has recently become feasible through the advent of X-Ray Free Electron Lasers (XFELs). For instance, XFELs allow imaging the electron density distribution within plasmas via Small Angle X-Ray Scattering (SAXS). We present a scalable GPU-based software framework for simulating photon scattering processes of X-ray beams in matter using Monte-Carlo methods. These simulations enable us to produce synthetic SAXS signals from the interaction of a modeled X-ray pulse with an arbitrarily complex, 3D electron density distribution obtained e.g. from detailed particle-in-cell simulations. Additionally, we present radiation transport methods in our 3D3V fully-relativistic PIC code PIConGPU. These methods enhance modeling of self-imaging of solid-density plasmas and lay the foundation for in-situ simulations of pump-probe experiments. Our new framework allows for single and multiple scattering and is extendable to include complex physics processes like ionization, atomic excitation and de-excitation along the photon path to further enhance its predictive capability.

Particle-in-Cell (PIC) codes are the the working horses of computational modeling in plasma physics, as they describe even the most turbulent kinetic processes from first-principles. But great applicability comes with great computational demand, requiring leadership-scale HPC systems in order to model full 3D geometries with solid-density. This poster shows the open software architecture of the world's fastest PIC code PIConGPU which is addressing these demands for the community. With a blazingly short time-to-solution, GPU-powered high-performance computing opens unique opportunities for studying transient plasma processes, e.g. by including XFEL photon distributions from simex_platform. Trading speed for enhanced predictive capabilities, we including collisional-radiative non-LTE models from SCFLY into the electro-magnetic PIC cycle.

In laser-generated plasmas the free electron density is a crucial parameter for plasma dynamics. Therefore, to model its spatial and temporal evolution the adequate treatment of ionization is vital. This poster presents the work in progress on numerical field ionization methods implemented in the world's fastest 3D3V electromagnetic particle-in-cell code PIConGPU. Thus, computing a value for the systematic error via repeating simulations with varying ionization schemes is in reach. With high performance computing we can give a range of validity for predictions of pump-probe experiments with high power lasers and X-ray free electron lasers.

Laser-driven solid density plasmas can be used to generate highly energetic electrons and ions. Diagnosing properties within those plasmas at nm length scales and down to fs timescales is crucial in understanding the involved processes. This has recently become feasible through the advent of X-Ray Free Electron Lasers (XFELs). For instance, XFELs allow imaging the electron density distribution within plasmas via Small Angle X-Ray Scattering (SAXS). We present a scalable GPU-based software framework for simulating photon scattering processes of X-ray beams in matter using Monte-Carlo methods. These simulations enable us to produce synthetic SAXS signals from the interaction of a modeled X-ray pulse with an arbitrarily complex, 3D electron density distribution obtained e.g. from detailed particle-in-cell simulations. Additionally, we present radiation transport methods in our 3D3V fully-relativistic PIC code PIConGPU. These methods enhance modeling of self-imaging of solid-density plasmas and lay the foundation for in-situ simulations of pump-probe experiments. Our new framework allows for single and multiple scattering and is extendable to include complex physics processes like ionization, atomic excitation, and de-excitation along the photon path to further enhance its predictive capability.

We present our scientific roadmap towards in-situ modeling of non-LTE interactions of XFEL type X-rays with solid density plasmas using a symbiosis of our performance portable, open source, 3D3V particle-in-cell (PIC) code PIConGPU and its X-ray tracing prototype ParaTAXIS. Treating radiation transport via various atomic processes will enable us to synthesize detector signals and gain predictive capabilities for upcoming pump-probe experiments at the European XFEL. With the world’s fastest particle-in-cell code PIConGPU and the raw computational power of the largest high performance computers we open up the possibility for large-scale case studies of unprecedented repeatability.

Particle-in-Cell (PIC) codes are the the working horses of computational modeling in plasma physics, as they describe even the most turbulent kinetic processes from first-principles. But great applicability comes with great computational demand, requiring leadership-scale HPC systems in order to model full 3D geometries with solid-density. This poster shows the open software architecture of the world's fastest PIC code PIConGPU which is addressing these demands for the community. With a blazingly short time-to-solution, GPU-powered high-performance computing opens unique opportunities for studying transient plasma processes, e.g. by including XFEL photon distributions from simex_platform. Trading speed for enhanced predictive capabilities, we including collisional-radiative non-LTE models from SCFLY into the electro-magnetic PIC cycle.

PIConGPU is reportedly the fastest electro-magnetic particle-in-cell code in the world in terms of sustained Flop/s. Its computational power does not only enable 3D3V simulations with unprecedented detail and fast time-to-solution but also allows improving predictive capabilities of simulations by estimating stochastic and systematic errors via repeated simulations. Synthetic in-situ diagnostics drive the exploration of high-detail simulations whose signatures, e.g. emitted radiation spectra, would be inaccessible in post-processing due to sheer data size. Upcoming experiments at the European XFEL require us to take PIConGPU even one step further: modeling XFEL-matter interaction on top of laser-driven particle acceleration processes requires the introduction of non-trivial X-ray photon scattering, photon generation and advanced non-LTE atomic models. Coupled with an open-science centered strategy, from open performance portable source code over open standardized data formats to documented workflows for PByte scale simulation we strive towards a new quality of predictive, reproducible simulations within our community.

We present a scalable GPU-based software framework for simulating photon scattering processes of X-ray beams in matter using Monte-Carlo methods. These simulations enable us to predict SAXS signals for experiments at upcoming superlative research facilities like the European XFEL. Often the expected outcome of SAXS experiments is produced by a Fourier Transform of a static 2D electron density distribution. Our new framework provides the opportunity to simulate the probing of femtosecond timescale 3D3V electron dynamics with single and multiple scattering and is extendable by more complex physics processes like laser absorption, atomic excitation and de-excitation to further enhance its predictive capability. As a foundation we use libPMacc, a powerful particle-mesh accelerator library that is also used by PIConGPU, the reportedly fastest fully-relativistic 3D3V particle-in-cell code in the world.

Laser-driven solid density plasmas can be used to generate highly energetic electrons and ions. Diagnosing properties within those plasmas at nm length scales and down to fs timescales is crucial in understanding the involved processes. This has recently become feasible through the advent of X-Ray Free Electron Lasers (XFELs). For instance, XFELs allow imaging the electron density distribution within plasmas via Small Angle X-Ray Scattering (SAXS). We present a scalable GPU-based software framework for simulating photon scattering processes of X-ray beams in matter using Monte-Carlo methods. These simulations enable us to produce synthetic SAXS signals from the interaction of a modeled X-ray pulse with an arbitrarily complex, 3D electron density distribution obtained e.g. from detailed particle-in-cell simulations. Additionally, we present radiation transport methods in our 3D3V fully-relativistic PIC code PIConGPU. These methods enhance modeling of self-imaging of solid-density plasmas and lay the foundation for in-situ simulations of pump-probe experiments. Our new framework allows for single and multiple scattering and is extendable to include complex physics processes like ionization, atomic excitation and de-excitation along the photon path to further enhance its predictive capability.