Inciter: Example setup for marching a transonic problem to steady-state
This example runs the ALECG hydrodynamics algorithm marching the solution to steady state using a local time-stepping procedure. The problem is initialized with a free-stream Mach number of M=0.84
and an angle of attack of 3.06 degrees.
Surface mesh for computing the flow around the F16, nelem=6M, npoin=1M, see below.
Control file: f16.lua
inciter = { title = "F16", ttyi = 10, -- TTY output interval cfl = 0.5, steady_state = true, residual = 1.0e-5, rescomp = 1, scheme = "alecg", partitioning = "mj", compflow = { physics = "euler" }, ic = { density = 1.225, -- density of air at STP, kg/m3 pressure = 1.0e+5, -- N/m^2 -- sound speed: sqrt(1.4*1.0e+5/1.225) = 338.06 m/s -- free stream Mach number: M = 0.84 -- angle of attack: 3.06 degrees -- u = M * a * cos(3.06*pi/180) = 283.57 -- v = M * a * sin(3.06*pi/180) = 15.159 velocity = { 0.0, 15.159, -283.57 } }, material = { { gamma = { 1.4 } } }, bc = { { symmetry = { 1, 2, 7 }, farfield = { 12 }, pressure = 1.0e+5, density = 1.225, velocity = {0.0, 15.159, -283.57 } } }, field_output = { interval = 1000, nodevar = { "density", "pressure", "x-velocity", "y-velocity", "z-velocity", "specific_total_energy" }, sideset = { 1, 2, 7 } }, diagnostics = { interval = 1, format = "scientific", error = "l2", precision = 12 } }
Run on 34 CPUs
./charmrun +p34 Main/inciter -v -i ../../tmp/f16b_tet.exo -c ../../tmp/f16.lua
The simulation took about three and a half minutes using 34 CPUs.
Solution
Pressure contours.