Dylan P. Brennan, Principal Scientist
Providing Scientific Solutions and Data Analytics to the Fusion Energy Sciences Community With Advanced Theoretical Modeling, Computational Analyses, and Machine Learning
Most Recent Publications
M.D. Pandya, B.E. Chapman, K.J. McCollam, R.A. Myers, J.S. Sarff, B.S. Victor, D.P. Brennan, D.L. Brower, J. Chen, W.X. Ding, C.T. Holcomb, N.C. Logan and E.J. Strait, “Early internal detection of magnetic tearing and implications for tokamak magnetohydrodynamic stability,” Phys. Plasmas 31, 070706 (2024).
Cihan Akcay, John Finn, Dylan Brennan, K.E.J. Olofsson and A.J. Cole, “Probabilistic locked mode predictor in the presence of a resistive wall and finite island saturation in tokamaks,” Phys. Plasmas 31, 032301 (2024).
An isosurface of the perturbed n = 1 pressure in orange, surrounded by an isosurface of the perturbed n = 2 toroidal current density in green later in a simulation. This indicates that the n = 1 perturbed pressure is dominated by the 1/1 mode while the n = 2 perturbed toroidal current density is dominated by the 3/2 mode. [Brennan et al, Nucl. Fusion 45, 1178 (2005)]
The vortices formed late in the ballooning dominant ELM evolution (a) and (b), and the linear n=20 mode with the largest energy (c). Filaments form and separate from the main plasma. [Brennan et al, J. Phys. : Conf. Ser. 46, 63 (2006)]
(a) Electron flux in momentum space at t = 4.0s including the whistler wave diffusion. (b) RE density growth (decay) rate as a function of E/ECH. The Red dots are from the simulation results with wave diffusion, and the red line is a linear regression. The green dots are from the simulation results without wave diffusion, and the green dashed line is the growth rate calculated from Rosenbluth and Putvinski (1997), with critical electric field Ec = 1.82 ECH from C. Liu et al, PPCF 59, 024003 (2017). [C.Liu et al, Phys. Rev. Lett. 120, 265001 (2018)].
Stability results from an RDCON analysis of n= 1 modes as the plasma is compressed. Ideal instability is shown in red, resistive instability in green/yellow, and stability in dark blue. Dotted white contours indicate plasma core temperature in keV. Safety factor values qmin and q95 are indicated by black and white dashed contours. Magenta contours show shaft current in MA. Two compression trajectories from nonlinear simulations with the VAC code are shown in light blue. [Brennan et al, Nucl. Fusion 61, 046047 (2021)]
CORSICA equilibria showing compression of plasma by a converging liquid metal shell. Contours indicate plasma current intensity, J·B/B^2. The solid metal is dark gray and liquid metal is light gray. [Brennan et al, Nucl. Fusion 61, 046047 (2021)]
Sparse data overlaid on the locking probability from a transfer-learning trained NN. The x’s are the sampled locked solutions and the o’s are the sampled unlocked solutions. The boundary of the hysteretic region (black solid) and the 50% probability contour (dashed) are also shown. [Akcay et al, Phys. Plasmas 31, 032301 (2024)]