Parameter file#

Tandem simulation parameters are written in the toml script. Here are the key parameters and their descriptions:

  • final_time: Final simulation time in seconds.

  • mesh_file: Mesh file.

  • lib: Lua file containing material & frictional paramters.

  • scenario: Name of the specific scenario defined in the Lua library.

  • Lua material functions: Note that the mu, lam, and rho Lua functions receive the volume tag as the last function argument. If region IDs are not set in the mesh, a value of -1 is assumed.

  • type: Type of problem. Available options: [poisson | elastic/elasticity].

  • mode: Mode of SEAS simulation. Available options: [QuasiDynamic/QD | QuasiDynamicDiscreteGreen/QDGreen | FullyDynamic/FD].

  • ref_normal: Define reference normal vector.

  • up: Define up direction vector.

  • boundary_linear: Assert that boundary is a linear function of time (i.e. boundary(x, t) = f(x) t). Default = false.

  • matrix_free: Use matrix-free operators. Default = false.

  • cfl: CFL tuning parameter (typically <= 1.0). Default = 1.0.

  • mg_coarse_level: Polynomial degree of coarsest MG level. Default = 1.

  • mg_strategy: MG level selection strategy. Available options: [TwoLevel | Logarithmic | Full]. Default = TwoLevel.

Green’s function operator checkpointing#

When using mode=QDGreen, Green’s function checkpointing is parameterized by the [gf_checkpoint] section. Here are the key parameters and their descriptions:

  • prefix: Path where Green’s function operator and RHS will be checkpointed.

  • freq_cputime: CPU time (minutes) frequency between Green’s function operator checkpoints.

Time stepping checkpointing#

Time stepping checkpointing is parameterized by the [ts_checkpoint] section. Here are the key parameters and their descriptions:

  • load_directory: directory from which a checkpoint is loaded. Give path to last_checkpoint.txt to let tandem retrieve the name of the last checkpoint file.

  • save_directory: Directory where checkpoints are saved. If not specified, it defaults to checkpoint.

  • freq_step: Time step frequency between checkpoints. This determines how often checkpoints are saved based on the number of time steps. The default value is 1000 steps.

  • freq_cputime: CPU time frequency between checkpoints, specified in minutes. Checkpoints are saved based on the elapsed CPU time. The default value is 30 minutes.

  • freq_physical_time: Physical time frequency between checkpoints. This determines how often checkpoints are saved based on the physical time of the simulation. Desactivated by default.

  • storage_type: Type of storage for checkpoints. It can be one of the following:

    • none: Completely deactivate checkpointing.

    • limited: Store a finite number of unique checkpoints on disk.

    • unlimited: Store all checkpoints without any limit.

  • storage_limited_size: Number of unique checkpoints stored on disk when storage_type is set to limited. The default value is 2.

Output configurations#

Five types of outputs are available: fault output, fault probe output, fault scalar output, domain output, and domain probe output.

Common options:

  • prefix: Output file name prefix.

  • atol: Absolute tolerance for VMax. Default = 1e-50.

  • rtol: Relative tolerance for VMax. Default = 0.01.

  • t_min: Minimum time difference between samples. Default = 0.0.

  • t_max: Maximum time difference between samples. Default = 365 * 24 * 3600.

For domain output:

  • jacobian: Output Jacobian. Default = false.

For probe outputs:

  • type: Define the probe output format. Available options: [Tecplot | CSV]. Default = CSV.

  • probes: Define the names and locations of the probes.

    • name: Station name added to the prefix to give the complete output file name.

    • x: Location of the probe. Units are consistent with your mesh [km].

Commented parameter file:#

final_time = 47304000000      # Final simulation time in seconds.
mesh_file = "tutorial.msh"    # Mesh file.
lib = "tutorial.lua"          # Lua file containing material&frictional paramters.
scenario = "normal"           # Name of the specific scenario defined in the lua library.
type = "elasticity"           # Type of problem: [poisson | elastic/elasticity].
ref_normal = [1, 0]           # Reference normal vector.
boundary_linear = true        # Assert that boundary is a linear function of time (i.e. boundary(x, t) = f(x) t). Default = False.

# On-fault probe outputs
[fault_probe_output]
prefix = "fltst_"
probes = [
    { name = "dp000", x = [0.0, -0.0] },
    { name = "dp025", x = [1.2500000000000002, -2.1650635094610964] },
    { name = "dp050", x = [2.5000000000000004, -4.330127018922193] },
    { name = "dp075", x = [3.750000000000001, -6.495190528383289] },
    { name = "dp100", x = [5.000000000000001, -8.660254037844386] },
    { name = "dp125", x = [6.250000000000002, -10.825317547305483] },
    { name = "dp150", x = [7.500000000000002, -12.990381056766578] },
    { name = "dp175", x = [8.750000000000002, -15.155444566227676] },
    { name = "dp200", x = [10.000000000000002, -17.32050807568877] },
    { name = "dp250", x = [12.500000000000004, -21.650635094610966] },
    { name = "dp300", x = [15.000000000000004, -25.980762113533157] },
    { name = "dp350", x = [17.500000000000004, -30.31088913245535] }
]

# Fault outputs
[fault_output]
prefix = "output/fault"
rtol = 0.1

# Domain outputs
[domain_output]
prefix = "output/domain"
rtol = 0.1

Optionally, you can also specify the format of the output probes (domain_probe_output and fault_probe_output) in order to be able to use the HDF5 format for instance. To do so, you need to set the type parameter in the probe output section, e.g. type = "HDF5". Note that the HDF5 output format is only available for probe outputs and not for field outputs. For example,

# On-fault probe outputs
[fault_probe_output]
prefix = "fltst_"
type = "HDF5"
probes = [
    { name = "dp000", x = [0.0, -0.0] },
    { name = "dp025", x = [1.2500000000000002, -2.1650635094610964] },
    { name = "dp050", x = [2.5000000000000004, -4.330127018922193] },
    { name = "dp075", x = [3.750000000000001, -6.495190528383289] },
    { name = "dp100", x = [5.000000000000001, -8.660254037844386] },
    { name = "dp125", x = [6.250000000000002, -10.825317547305483] },
    { name = "dp150", x = [7.500000000000002, -12.990381056766578] },
    { name = "dp175", x = [8.750000000000002, -15.155444566227676] },
    { name = "dp200", x = [10.000000000000002, -17.32050807568877] },
    { name = "dp250", x = [12.500000000000004, -21.650635094610966] },
    { name = "dp300", x = [15.000000000000004, -25.980762113533157] },
    { name = "dp350", x = [17.500000000000004, -30.31088913245535] }
]

Finally, the moment rate writer is also available as an output separately but it is only available in HDF5 format. To use it, you need to use the [moment_rate_output] section. You do not need to sepcify a type currently because it is only available in HDF5 format at the moment. But you can specify the prefix for the output files and the frequency of the output using the same parameters as for the other outputs, e.g. prefix = "output/moment_rate" and t_max = 9460800 to output the moment rate every 60 minutes of CPU time. For example,

[moment_rate_output]
prefix = "output/moment_rate"
t_max = 9460800