Physics

pde Partial differential equation.
Name of the PDE class and optional parameters separated by spaces.
initCond Initial conditions.
Name of the initial condition class and optional parameters separated by spaces. Basic FuncInitCond implementations:
  • Zero — Sets zero value everywhere for each field component.
  • UserDef — User-defined initial condition. Its parameters are the specifications of the field components using mathematical formulas, in the following form: f=formula1 g=formula2 ....
  • ODEShootInit — Initial condition determined by an ordinary differential equation (see the ode parameter) which is solved using shooting method.
  • ODERelaxInit — Initial condition determined by an ordinary differential equation (see the ode parameter) which is solved using relaxation method.
  • BData — Uses a data file from a previous run.
Common parameters of FuncInitCond implementations:
  • interpolatedLevel=l — Use this parameter to interpolate the initial condition on refined meshes. An integer value l≥1 specifies the minimum refinement level to initialize using interpolation. If set to zero (interpolatedLevel=0), then no interpolation is used; each level is initialized as precisely as possible.
Examples:
initCond=UserDef f=(0.1*r1+0.5*r2+r3)/(r1+r2+r3)
initCond=BData kgm.bdata
initCond=gr.fixmp.kerrhiggs.Hunch interpolatedLevel=1
ode Initial state specified by an ordinary differential equation.
Use it with initCond=ODEShootInit or ODERelaxInit.
Example:
initCond=ODEShootInit
ode=gr.dynss.ekg.EKGODE a=7.162 b=3 c=0.08 d=100 psi=(rho>a-b)*(rho<a+b)?c*exp(d+d*b^2/((rho-a)^2-b^2)):0 psi_rho=(rho>a-b)*(rho<a+b)?-2*c*d*b^2*(rho-a)*exp(d+d*b^2/((rho-a)^2-b^2))/((rho-a)^2-b^2)^2:0
parameters.* Parameters of the PDE and the initial condition.

Basic simulation parameters

resolution Grid resolution.
Use resolution=N to specify the resolution for a unigrid, use resolution=Nx2^L to specify both the base grid resolution (N) and the maximum number of refinement levels (L). Examples:
resolution=1024
resolution=256x2^4
dtdx The Δt/Δx Courant factor.
It can be constant or a function of time. Examples:
dtdx=1
dtdx=t<16.4? 1 : t<17.58? 0.1 : 0.02
gridInt Integration method.
RK2 — Runge-Kutta 2nd order
RK4 — Runge-Kutta 4th order
ICN — Iterated Crank-Nicholson
LW2 — Lax-Wendroff
sigma The σ factor of the numerical dissipation term.
It can be coordinate dependent. Examples:
sigma=0.01
sigma=LStepSigma sigma0=0.02 sigma1=0.01 i=8
shoot.integrator Integration method for ODEShootInit.
RK2 — Runge-Kutta 2nd order
RK4 — Runge-Kutta 4th order

Special mesh refinement related parameters

amError Error function for the mesh refinement condition.
Examples:
amError=ComponentError 0
errorTolerance Error tolerance for the mesh refinement condition.
Example:
errorTolerance=1e-12
errorCheckFreq Frequency of error checking.
Example:
errorCheckFreq=8
regridFreq Maximum number of time steps without regridding.
Example:
regridFreq=16
bufferZoneSize Buffer zone size.
Example:
bufferZoneSize=2

Output file generation

dtwrite Time difference for datafile writing.
It can be a constant or a function of time. Examples:
dtwrite=0.1
dtwrite=t<1.6? 0.1 : t<1.68? 0.01 : t<1.688? 0.001 : 0.0001
tmax Output should end at this time.
It can be a constant or a function of the initial time parameter, t0. Example:
tmax=t0+16

Interactive monitoring — xgridripper

displayedComponents Comma separated list of components and grid functions to display.
Example:
displayedComponents=g00,m,r,e,y,t_proper
memoryMax Maximum number of integration steps to memorize.
The simulation can be backtracked by the specified number of steps.
memorySteps A value n means that each n-th integration step is memorized and backtracking restores the time t-nΔt.
delay Delay between time steps in milliseconds.
Use it to slow down the simulation.
breakpoints Comma separated list of times where the simulation should pause automatically.


08 May 2009, P. Csizmadia