A toroidal modeling tool is developed to study the runaway electron (RE) avalanche production process in tokamak plasmas, by coupling the Rosenbluth–Putvinski avalanche model (Rosenbluth and Putvinski 1997 Nucl. Fusion 37 1355) with an n = 0 magneto-hydrodynamic (MHD) solver. Initial value numerical simulations are carried out for two DIII-D discharges with different plasma shapes (one near circular, and the other with high elongation). It is found that, assuming the same level of about 1% seed current level, the Rosenbluth–Putvinski model somewhat underestimates the RE plateau current for the circular-shaped plasma, as compared with that measured in DIII-D experiments. For an elongated, higher current plasma, simulations find strong runaway current avalanche production despite the lack of measured plateau RE current in experiments. A possible reason for this discrepancy is a lack of additional RE dissipation physics in the present two-dimensional model. Systematic scans of the plasma boundary shape, at fixed pre-disruption plasma current, find that the plasma elongation helps to reduce the RE avalanche production, confirming recent results obtained with an analytic model (Flp etal 2020 J. Plasma Phys. 86 474860101). The effect of the plasma triangularity (either positive or negative), on the other hand, has a minor effect. On the physics side, the avalanche process involves two competing mechanisms associated with the electric field. On the one hand, a stronger electric field produces a higher instantaneous avalanche growth rate. On the other hand, a fast growing RE current quickly reduces the fraction of the conduction current together with the electric field, and hence a faster dissipation of the poloidal flux. As a final result of these two dynamic processes, the runaway plateau current is not always the largest with the strongest initial electric field. These results lay the foundation for future self-consistent inclusion of the MHD dynamics and the RE amplification process.

通过将 Rosenbluth-Putvinski 雪崩模型 (Rosenbluth and Putvinski 1997 Nucl. Fusion 37 1355) 与n= 0 磁流体动力学 (MHD) 求解器。对具有不同等离子体形状（一种接近圆形，另一种具有高伸长率）的两种 DIII-D 放电进行了初始值数值模拟。发现，假设大约 1% 的种子电流水平相同，与 DIII-D 实验中测量的相比，Rosenbluth-Putvinski 模型稍微低估了圆形等离子体的 RE 平台电流。对于拉长的、更高电流的等离子体，尽管在实验中缺乏测量的高原 RE 电流，但模拟发现强烈的失控电流雪崩产生。这种差异的一个可能原因是当前二维模型中缺乏额外的 RE 耗散物理。等离子体边界形状的系统扫描，在固定的预破裂等离子体电流下，等2020 J. Plasma Phys. 86474860101）。另一方面，等离子体三角形的影响（正的或负的）影响很小。在物理学方面，雪崩过程涉及与电场相关的两种相互竞争的机制。一方面，更强的电场会产生更高的瞬时雪崩增长率。另一方面，快速增长的 RE 电流会迅速降低传导电流与电场的比例，因此极向通量的耗散速度更快。作为这两个动态过程的最终结果，失控的平台电流并不总是最大的，具有最强的初始电场。这些结果为未来自洽包含 MHD 动力学和 RE 放大过程奠定了基础。