DNA2MHD/simulationlog.txt at main · akashukla/DNA2MHD · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Simulation 458/459:
DNA2MHD cyclotron wave simulations with RK4 vs DORPI
Energy conserved only slightly better with RK4 (~1 difference in error)

02/06/24

Allowed mode decomposiition for all initial condition simulations

Simulation 460/461:
Helical Navier Stokes simulation with DORPI vs RK4
Energy conserved only slightly better with RK4 (~1 difference in error)

02/07/24

Wrote linear harmonic oscillator simulation

Simulation 462/463:
Helical 3ND harmonic oscillator simulation with DORPI vs RK4, 1200 its
Energy conserved about 20-30x better with DORPI

Simulation 464/465:
DNA2MHD cyclotron linear HMHD wave simulations with DORPI vs RK4, 12000 its
Energy conserved about 20-30x better with DORPI

02/12/24

Wrote in condition to set - modes equal to conjugates for kx = 0 (ensure reality)
Thinking about adding in ics to set v weakly equal to zero or weakly normal to the boundaries, but not sure if that would hold for later times

Simulation 467:
DORPI nonlinear DNA2MHD simulation at dt = 0.001 to compare energy conservation against 458
Full cyclotron initial condition
Error ratios dropped by a factor of 10^7-8 from what I recall
Does the smaller dt affect energy conservation though - if it doesn't, there could be a gauge or boundary issue

Simulation 470:
DORPI nonlinear DNA2MHD simulation to test energy conservation with parallel to edge boundary condition imposed
Parallel boundary condition didn't work
Energy not conserved

Simulations 471+472:
With 128 modes in each direction instead of 64
Larger overall mode in 471

Simulations 473+474:
256 modes in perpendicular directions, 32 in z direction
Energy conservation got worse, about the same for RK4 and DP
Was this just due to decreasing max kz?

Simulations 485+486:
64^3 modes with 2x padding in all directions and a centered convolution
testing if "centered convolution" had lower energy growth and fixed the issue
or if the preliminary testing (20 its) was right in that the energy growth was more
485 DORPI + 486 RK4 - test if converge right with this and if growth worse
Gain about 0.01 energy for both cases - not converging right

Simulations 487+488:
64^3 modes with 2x padding in all directions, uncentered convolution
testing if only 3/2 padding wasn't enough