【作者】 杨长鸿；

【导师】 蒙林；

【作者基本信息】 电子科技大学 ， 等离子体物理， 2013， 博士

【摘要】 从上世纪60年代开始，直线感应加速器得到了快速发展，数值计算方法广泛应用到直线感应加速器的研究中。由于国内缺乏有效的数值模拟研究工具，采用数值模拟方法对直线感应加速器进行研究尚属空白。为了填补这一研究空白，以及为国内直线感应加速器的研究提供新的方法，本论文将对直线感应加速器关键物理问题及数值模拟算法进行研究，主要内容包括以下几个方面。（1）研究了注入器的物理过程，同时研究出了多功能发射模型来代替注入器发射电子束。在研究注入器物理过程时，研究了光发射的物理机制，建立起光发射数值模型。并结合注入器二极管结构，实现了光阴极注入器模拟研究，多项模拟结果与实验测试结果基本一致。同时该模拟研究还获得了低发射度电子束，为未来光阴极在直线感应加速器上的应用提供支持。在研究注入器时还研究并开发了多功能发射模型。该模型能够设置多种电子束参数，这些参数涵盖了研究注入器电子束时关心的参数种类，例如：电子束能量、电子束发射度、能散以及电子束发射方向，等。多功能发射模型为研究其它物理过程提供了简单、方便、准确的宏粒子束；（2）研究了螺线管线圈与校正线圈的匹配，为实验调节提供指导。为了研究螺线管线圈磁场与校正线圈磁场的匹配，研究并开发了直线感应加速器专用螺线管线圈磁场模型和校正线圈模型。在研究磁场匹配时，通过设置螺线管线圈模型参数来模拟螺线管线圈磁轴倾斜情况，再使用校正线圈对磁轴倾斜场进行校正。经过多次数值调节校正线圈输入电流，最终校正了螺线管线圈磁场。由此从数值研究中实现了螺线管线圈磁轴倾斜的校正。此外还研究了校正线圈的另一个功能：校正电子束质心。该研究是通过建立模型，利用多功能发射模型来产生电子束偏离轴心情况，再使用校正磁场校正质心。以上研究为实验调试电子束心位置及校正螺线管线圈磁场提供指导；（3）研究了电子束在加速间隙处激励起的尾场。为了研究电子束尾场及实现电子束输运过程的研究，建立了两种加速腔模型。两种加速腔模型均能在加速间隙处建立起加速电场。通过对两种间隙模型在开放边界或金属边界情况下激励起的尾场分析，研究发现当前加速场馈入方式下的加速腔模型不能很好模拟电子束尾场。但如果改变加速场的馈入方式和改变当前加速间隙处的边界条件，加速腔就能很好的用于尾场研究。同时该研究证明了弯曲式间隙较直缝式间隙尾场阻抗小；（4）研究了电子束在直线感应加速器主加速器的输运过程。为了实现电子束输运过程研究，分析模拟所需数值模型，并设计出了计算流程图。同时，为解决计算大尺度问题，引入了“分区建模”并行算法。研究电子束在主加速器的输运过程中采用多功能发射模型代替了注入器发射，模拟了电子束在18个加速组元中加速场与磁场匹配状况下的输运过程。模拟耗费了6台电脑，历时3天的运算。该模拟结果得到了两个重要的结论：加速后电子束能量为20.2MeV，以及束包络不随时间变化。这两个现象与实验观测结果一致，同时该研究获得了一组特定情况下的匹配磁场；（5）研究了电子束在聚焦段的输运及聚焦。为了实现聚焦段的研究，首先研究并开发了磁透镜数值模型。在研究磁透镜时分别采用磁标势和磁矢势两种方法来研究了磁透镜中场的分布情况，并对两种算法的特点进行了分析。然后通过结构参数和磁场参数建立起相应的聚焦段模型，通过对磁场的调节，实现了电子束聚焦过程的模拟研究，得到了3.9mm的焦斑。此外还对四极透镜模型及其聚焦效果进行了探索性研究，为四极透镜在直线感应加速器上的应用做准备。随后对本研究在实验调试及研究中能够解决的问题做总结。并调研了下一步的研究方向。更多还原

【Abstract】 Linear induction accelerators (LIA) are constructed to produce relativistic intenseelectron beams with high energy and high qualities. Researching on linear inductionaccelerator by numerical method has been widely used abroad. However, best PICsoftware is embargo for us. University of Electronic Science and Technology of Chinacooperated with Institute of Fluid Physics, China Academy of Engineering Physics toresearch on linear induction accelerator about key physical issue and numericalsimulation algorithms.1. To research photoemission injector, photoemission module is researched anddeveloped firstly. In the photoemission module, photocathode material and laser sourceare taken into account. A relationship between photocathode and macro particle initialcondition is established. Then a photoemission injector is modeled to simulatephotoemission process. Some of the simulation result is coincident with experiment.The deviation of current density between experiment and simulation is less than7%.And the emittance of photoemission injector beams is smaller than cold emission;Multi-function emission module is researched and developed, which could set beamparameter not only energy, charge, but also emittance, energy spread and beam off-axis;2. Steering magnetic field matching with solenoid magnetic field is researched. Toresearch this physical issue, solenoid magnetic field and steering magnetic field moduleare developed firstly. When research solenoid magnetic field, solenoid axis tilt anduniform ring are taken into account. Then researching shows how to adjust and matchthem together. This research could guide researchers how to adjust steering coil currentin experiments. Steering magnetic field has another function, which could adjust beamcentroid. This physical process is researched also. In the research, multi-functionemission module is used for emit tilt beam, and try to adjust steering coil current toadjust beam centroid. The beam centroid adjusted at last;3. To research wakefield and beam transmission, two kinds accelerator cell moduleare developed. The difference between two kinds accelerator cell module is the shape ofthe accelerator gap. The accelerator gap is the most important place on researchwakefield, which could excite certain frequency wakefield. The wakefield aroundaccelerator gap are integrated and analyzed. The simulation result shows that the current gap boundary condition isn’t suitable for research wakefield. If the gap boundarycondition set as electric boundary, that would be perfect for research wakefield. In thenext study, the method of input accelerator field must choose other option. Under thissituation, electric boundary condition could be set at accelerator gap;4. Beam transmission process in the main accelerator system is researched. Andthis is the very first time researched main accelerator by numerical method in domestic.The main accelerator system is about40meters long. Based on the platform of theMPICH2message passing system, a proper method for parallel computing of moreaccelerate sections is provided and solved LIA’s great scale problems. It cost6personalcomputers,3days to simulate beam transmission in main accelerator. The simulationgot two important result are coincident with experiment:1. The waveform of beam atcertain position is fixed, which wouldn’t change vary time;2. After transmission mainaccelerator, beam energy reached20.2MeV. By the way, the simulation got a group ofmatch fields under certain condition;5. Beam transmission in downstream is researched also. Magnetic lens areresearched by magnetic scalar and vector method firstly. Each method has its advantageand disadvantage. The algorithm of magnetic scalar method is simpler than vectormethod. That when set parameters, vector method is much more complicate. But ifpermeability of magnetic lens is smaller than4000, vector method is more proper forsimulation magnetic field. When researching beam transmission in downstream,magnetic vector method is applied. Then build downstream module and set magneticfield to simulate and research beam transmission. The beam focused to a3.9mm spot inthe simulation. For future research, quadrupole lens module are researched anddeveloped also.更多还原