Northern Illinois University, DeKalb, Illinois, USA
The high-intensity, high-brightness and precision frontiers for charged particle beams are an increasingly important focus for study. Ultimately for electron beam applications, including FELs and microscopy, the quality of the source is the limiting factor in the final quality of the beam. It is imperative to understand and develop a new generation of sub-Kelvin electron sources, and the current state of PIC codes are not precise enough to adequately treat this ultracold regime. Our novel computational framework is capable of modelling electron field emission from nanoscale structures on a substrate, with the precision to handle the ultracold regime. This is accomplished by integrating a newly developed Poisson integral solver capable of treating highly curved surfaces and an innovative collisional N-body integrator to propagate the emitted electron with prescribed accuracy. The electrons are generated from a distribution that accounts for quantum confinement and material properties and propagated to the cathode surface. We will discuss the novel techniques that we have developed and implemented, and show emission characteristics for several cathode designs.
JLab, Newport News, Virginia, USA
Truncated power series algebra (TPSA) or differential algebra (DA) is often used by accelerator physicists to generate a transfer map of a dynamic system. The map then can be used in dynamic analysis of the system or in particle tracking study. TPSA/DA can also be used in some fast algorithms, eg. the fast multipole method, for collective effect simulation. This paper reports a new TPSA/DA library written in C++. This library is developed based on Dr. Lingyun Yang’s TPSA code, which has been used in MAD-X and PTC. Compared with the original code, the updated version has the following changes: (1) The memory management has been revised to improve the efficiency; (2) A new data type of DA vector is defined and supported by most frequently used operators; (3) Support of inverse trigonometric functions and hyperbolic functions for the DA vector has been added; (4) function composition is revised for better efficiency; (5) a python wrapper is provided. The code is hosted at github and available to the public.
ANL, Lemont, Illinois, USA
In particle accelerators, the tracking simulation is usually performed with symplectic integration, or linear/nonlinear transfer maps. In this paper, it is shown that the linear/nonlinear transfer maps may be represented by an artificial neural network. To solve this multivariate regression problem, both random datasets and structured datasets are explored to train the neural networks. The achieved accuracy will be discussed.
ANL, Lemont, Illinois, USA
Standard neural network algorithms are developed for classification and regression applications. In this paper, the details of the neural network algorithms are presented, together with several applications. Artificial neural network is trained to classify multi-class transfer matrix of different types of particle accelerator components. It is shown that with a fully-connected feedforward neural network, it is possible to get high accuracy of 99% on training data, validation data and test data.