20th Advanced Accelerator Concepts Workshop

6 Nov 2022, 15:00 → 11 Nov 2022, 17:10 America/New_York

Hyatt Regency Long Island

Mark Palmer (Brookhaven National Laboratory) , Navid Vafaei-Najafabadi (Stony Brook University)

The AAC’22 workshop is the 20th in a series of by-invitation biennial fora for intensive discussions on long-term research in advanced accelerator physics and technology. This research supports the development of capabilities for the basic sciences, from photon science to high energy physics, as well as the development of compact accelerators for industrial, medical and security applications.

AAC'22 will be organized into eight working groups covering the following topical areas:

1. Laser-Plasma Wakefield Acceleration 
2. Computation for Accelerator Physics
3. Laser and High-Gradient Structure-Based Acceleration
4. Beam-Driven Acceleration
5. Beam Sources, Monitoring, and Control
6. Laser-Plasma Acceleration of Ions
7. Radiation Generation and Advanced Concepts
8. Advanced Laser and Beam Technology and Facilities

Conference Home:  https://www.aac2022.org/

 

Sunday, 6 November

18:00 → 19:30 Welcome Reception

Monday, 7 November

10:20 → 10:40 Coffee Break

12:10 → 13:30 Lunch

13:30 → 15:00 WG2: Computation for Accelerator Physics: Session 1

Conveners: Alexey Arefiev (UC San Diego) , David Bruhwiler (RadiaSoft LLC)

13:30 10-year timetable for AAC computing

13:50 Traveling-wave electron accelerators – towards scalable laser-plasma accelerators beyond 10GeV

Traveling-wave electron acceleration (TWEAC) is an advanced laser-plasma accelerator scheme, which is neither limited by dephasing, nor by pump depletion or diffraction. Such accelerators are scalable to energies beyond 10 GeV without the need for staging and are candidates for future compact electron-positron colliders based on existing CPA lasers. TWEAC utilizes two pulse-front tilted laser pulses whose propagation directions enclose a configurable angle. The accelerating cavity is created along their overlap region in the plasma and can move at the vacuum speed of light. The oblique laser geometry enables to constantly cycle different laser beam sections through the interaction region, hence providing quasi-stationary conditions of the wakefield driver.

The TWEAC geometry enables to access to a wide range of regimes, which are customizable in cavity geometry, laser-to-electron energy efficiency and the required laser properties at different plasma densities, making the scheme suitable for high-rep rate lasers at low energies per pulse to multi-PW laser facilities. Exploring these regimes in high-fidelity simulations is computationally highly demanding, as these need to include large plasma volumes in 3D at high-resolution over an extended acceleration distance. Since even "small" test simulations need hundreds of GPUs, TWEAC simulations require exascale compute resources.

We present recent progress in TWEAC simulations and various technical advances in the 3D3V particle-in-cell code PIConGPU that enable running on the upcoming Frontier cluster, most notably support of the HIP computational backend allowing to run on AMD GPUs, as well as openPMD, PICMI and algorithmic developments. These advances are mainly driven by our participation in OLCF’s Frontier Center for Accelerated Application Readiness providing access to the hardware platform of the Frontier exascale supercomputer. We show performance data and recent applications of PIConGPU profiting from these developments.

Speaker: Dr Alexander Debus (Helmholtz-Zentrum Dresden-Rossendorf)

Debus_PIConGPU+X_AAC_2022_web.pdf

Debus_TWEAC_AAC_2022_web.pdf

PIConGPU+X_AAC_2022.pptx

TWEAC_AAC_2022_without_backup_slides.pptx

14:20 Now that we are ready for Exascale, what can we do with it? PIConGPU for next generation accelerator research.

PIConGPU, like many other codes, is ready for the next Exascale supercomputers. Heterogeneous programming as the main ingredient enables effective use of these machines. Important challenges still ahead are timely analysis of large scale simulation data and complex workflows for multi-physics simulations and machine learning.

As experimental capabilities progress and high-repetition rate sources have become widely available, the role of simulations is shifting. Predictive capabilities are put to the test more often than ever. While simulations for high intensities still push the boundaries of known physics, the task to predict experimental outcomes for sources existing now has become more and more pressing.

The question arises which steps will be necessary to achieve good comparison with experiments. We will present our approach to work closely with experiment, lessons learned from this approach, what can be done better and how Exascale computing might help.

Speaker: Michael Bussmann (Helmholtz-Zentrum Dresden - Rossendorf)

221107_aac_exascale.pptx

14:50 Discussion

15:00 → 15:30 Coffee Break

Tuesday, 8 November

10:00 → 10:30 Coffee Break

12:00 → 13:20 Lunch

15:00 → 15:30 Coffee Break/Exhibits

Wednesday, 9 November

10:00 → 10:30 Coffee Break/Exhibits

12:00 → 13:20 Lunch

Thursday, 10 November

10:00 → 10:30 Coffee Break/Exhibits

12:00 → 13:20 Lunch

13:30 → 15:00 WG2: Computation for Accelerator Physics: Session 8

Conveners: Alexey Arefiev (UC San Diego) , David Bruhwiler (RadiaSoft LLC)

13:30 Eulerian Finite-Difference Vlasov Solver with a Non-Uniform Momentum Grid

An Eulerian finite-difference method solving the Vlasov equation is developed with a static, non-uniform momentum grid. The computational cost of this transformation differs negligibly from the uniform case with the same number of grid points. A general grid parametrization is tested against classic instabilities and driven cases and is found to provide significant efficiencies over the uniform grid case. This technique allows for the distribution of computational resources based on the relativce importance of kinetic activity in phase-space while preserving variationally conserved quantities from the formal bracket. This method can be readily extended to multiple dimensions and is compatible with dynamically adapting the momentum grid.

Speaker: Prof. B. A. Shadwick (Univ of Nebraska - Lincoln)

13:45 An Alternative Approach to Incorporating Laser Pulses in Particle-in-Cell Simulations

Numerical modeling of electromagnetic waves is a critical component of particle-in-cell simulation of laser–plasma interactions. Traditionally, laser pulses have been either launched from simulated antennas or initialized in their entirety in the computational domain. Relying on the electromagnetic field update to advance the laser pulse, however, imposes needless computational expense and complexity for a number of emerging applications. As an alternative, we demonstrate that laser pulses can be incorporated using analytic expressions provided that numerical dispersion is matched. The otherwise self-consistent treatment of the plasma reproduces 3-D–like focusing in lower-dimensional simulation, enables direct examination of approximate solutions to Maxwell’s equations including Laguerre–Gaussian beams, and facilitates the modeling of novel laser pulses including the spatiotemporally shaped flying focus.

Speaker: Kathleen Weichman (University of Rochester Laboratory for Laser Energetics)

14:00 Opportunities and Issues with the Unitary Particle Pusher

By using the spinor representation of four-vectors, it is possible to write a simple expression for the momentum change of a charged particle in an arbitrary crossed field. It can be evaluated exactly if transcendental function evaluations are tolerable, or in an invariant-preserving expansion otherwise. We discuss progress in incorporating this approach into a particle-in-cell framework.

Speaker: Dr Daniel Gordon (U.S. Naval Research Laboratory)

gordon-AAC-2022.pdf

gordon-AAC-2022.pptx

14:15 Efficient algorithms for multi-level ionization of high-atomic-number gases and applications

An efficient numerical algorithm for multi-level ionization of high-atomic-number gases has been developed. It is based on analytical solutions to the system of differential equations describing evolution of ionization states. The algorithm fully resolves multiple time scales associated with ionization processes coupled to electromagnetic processes of laser-plasma interaction. The effects of the orbital quantum numbers and their projections are also examined. The algorithm efficiency is improved by using a locally reduced system of differential equations. The multi-level ionization algorithm has been implemented in SPACE, a parallel, fully relativistic, three-dimensional particle-in-cell code. In addition to Vlasov-Maxwell equation solvers, SPACE implements a novel, highly adaptive particle method for Vlasov-Poisson equations called Adaptive Particle-in-Cloud (AP-Cloud) that replaces the traditional PIC mesh with octree data structures. Verification and validation problems for the multi-level ionization algorithm in SPACE will be presented. The code has been applied to the study of ionization injection of electrons into laser-driven plasma wakefields. Comparison of simulations with BNL-ATF experiment on ionization-injection will also be discussed.

Speaker: Aiqi Cheng (Stony Brook University)

AAC2022_WG2_highZionization.pdf

AAC_Abstract_MultiIonization.docx

14:30 EZ: An Efficient, Charge Conserving Current Deposition Algorithm for Electromagnetic Particle-In-Cell Simulations

We present EZ, a novel Current Deposition algorithm for particle-in-cell simulations, which calculates the current density on the grid due to macro-particle motion within a time step by solving the electrodynamic continuity equation. Being a charge conserving hybridization of Esirkepov’s method and ZigZag, we refer to it as “EZ” as shorthand for “Esirkepov meets ZigZag”.
The talk will detail the new method and show that EZ achieves the same level of charge conservation as the commonly used method by Esirkepov, yet reaches higher performance for macro-particle assignment-functions up to third-order. These results are obtained from Simulations of a warm, relativistic plasma with PIConGPU.
In addition, guide lines for its implementation aiming at highest performance on GPUs are provided.

Speaker: Dr Klaus Steiniger (Helmholtz-Zentrum Dresden-Rossendorf)

Steiniger_EZ_web.pdf

14:45 Discussion

15:00 → 18:00 Afternoon at Leisure

Friday, 11 November

10:00 → 10:30 Coffee Break

12:00 → 13:00 Lunch

14:40 → 15:00 Coffee Break