mp-advec-diff: Parameter description

Contents

• base

• time stepping

• adaptive meshing

• advection diffusion

• output

• profiling

---

🔷 base

Parameter	Type	Default	Description
case name	string	not_initialized	Sets the base name for the application that will be fed to the problem type.
dimension	integer	2	Defines the dimension of the problem
number	string	double	Floating point number format. Currently, only ‘double’ is explicitely instantiated. Allowed values: - double
global refinements	integer	1	Defines the number of initial global refinements
do print parameters	boolean	True	Set this parameter to true to list parameters in output
verbosity level	integer	1	Sets the verbosity level of the console output: 0: silent: for non-robust tests and benchmark runs; 1: minimal: for robust tests; 2: detailed; 3: full
fe	object		See table

base: fe

Parameter	Type	Default	Description
type	string	FE_Q	Finite Element.FE_Q: hexahedral continuous finite element with polynomial degree p; FE_SimplexP: tetrahedral continuous finite element with polynomial degree p; FE_Q_iso_Q1: hexahedral continuous finite element with p subdivisions containing linear elements; FE_DGQ: hexahedral discontinuous finite element with polynomial degree p Allowed values: - not_initialized - FE_Q - FE_SimplexP - FE_Q_iso_Q1 - FE_DGQ
degree	integer	1	Defines the degree p of the finite element type. If “type” is “FE_Q_iso_Q1” this parameter defines the number of subdivisions.

---

🔷 time stepping

Parameter	Type	Default	Description
start time	number	0.0	Defines the start time for the solution of the levelset problem
end time	number	1.0	Sets the end time for the solution of the levelset problem
time step size	number	0.01	Sets the step size for time stepping. For non-uniform time stepping, this parameter determines the size of the first time step.
max n steps	integer	10000000	Sets the maximum number of melt_pool steps
time step size function	string	0.0*t	Set an analytical function to determine the time step size. For the prediction of the new time increment, the old time is used.

---

🔷 adaptive meshing

Parameter	Type	Default	Description
do amr	boolean	False	Set this parameter to true to activate adaptive meshing
do not modify boundary cells	boolean	False	Set this parameter to true to not refine/coarsen along boundaries.
upper perc to refine	number	0.0	Defines the (upper) percentage of elements that should be refined
lower perc to coarsen	number	0.0	Defines the (lower) percentage of elements that should be coarsened
max grid refinement level	integer	12	Defines the number of maximum refinement steps one grid cell will be undergone.
min grid refinement level	integer	-1	Defines the number of minimum refinement steps one grid cell will be undergone.
n initial refinement cycles	integer	0	Defines the number of initial refinements.
every n step	integer	1	Defines at every nth step the amr should be performed.
min cells marked to refine	integer	1	Minimum number of cells that must be marked for refinement/coarsening before the mesh is updated.
min indicator threshold to refine cell	number	0.0	Minimum indicator value required for a cell to be considered for refinement.
solution transfer average values	boolean	False	Set this parameter to true to average the contribututions to the same DoF coming from different cells during solution transfer.

---

🔷 advection diffusion

Parameter	Type	Default	Description
fe	object		See table
convection stabilization	object		See table
diffusivity	number	0.0	Defines the diffusivity for the advection diffusion equation
implementation	string	meltpooldg	Choose the corresponding implementation of the advection diffusion operation. Allowed values: - meltpooldg - adaflo
enable time dependent bc	boolean	False	Set this parameter to true to enable time-dependent bc.
predictor	object		See table
linear solver	object		See table
time integration	object		See table

advection diffusion: fe

Parameter	Type	Default	Description
type	string	not_initialized	Finite Element.FE_Q: hexahedral continuous finite element with polynomial degree p; FE_SimplexP: tetrahedral continuous finite element with polynomial degree p; FE_Q_iso_Q1: hexahedral continuous finite element with p subdivisions containing linear elements; FE_DGQ: hexahedral discontinuous finite element with polynomial degree p Allowed values: - not_initialized - FE_Q - FE_SimplexP - FE_Q_iso_Q1 - FE_DGQ
degree	integer	-1	Defines the degree p of the finite element type. If “type” is “FE_Q_iso_Q1” this parameter defines the number of subdivisions.

advection diffusion: convection stabilization

Parameter	Type	Default	Description
type	string	none	Defines the type for convection stabilization. Allowed values: - none - SUPG
coefficient	number	-1.0	Defines the stabilization coefficient for convection. (default velocity-dependent).

advection diffusion: predictor

Parameter	Type	Default	Description
type	string	none	Choose a predictor type: none: use old value as initial guess; zero: se zeros as initial guess; linear_extrapolation: calculate the predictor by a linear combination from the two old solution vectors; least_squares_projection: least squares projection (WIP) Allowed values: - none - zero - linear_extrapolation - least_squares_projection
n old solutions	integer	2	Choose the number of old solution vectors considered.This parameter is only relevant for least squares projection.For all other predictors, this parameter will be set appropriately.

advection diffusion: linear solver

Parameter	Type	Default	Description
solver type	string	GMRES	Set this parameter for choosing an iterative linear solver type. Allowed values: - CG - GMRES
preconditioner type	string	Diagonal	Set this parameter for choosing a preconditioner type. Allowed values: - Identity - AMG - ILU - Diagonal
max iterations	integer	10000	Set the maximum number of iterations for solving the linear system of equations.
rel tolerance	number	1e-12	Set the relative tolerance for a successful solution of the linear system of equations.
abs tolerance	number	1e-20	Set the absolute tolerance for a successful solution of the linear system of equations.
do matrix free	boolean	True	Set this parameter if a matrix free solution procedure should be performed.
monitor type	string	none	Set the monitor type of the linear solver. Allowed values: - none - reduced - all

advection diffusion: time integration

Parameter	Type	Default	Description
type	string	crank_nicolson	Name of the time integration scheme. Allowed values: - not_initialized - LSRK_stage_1_order_1 - LSRK_stage_3_order_3 - LSRK_stage_5_order_4 - LSRK_stage_7_order_4 - LSRK_stage_9_order_5 - RKC_n_stages - implicit_euler - explicit_euler - crank_nicolson - bdf_1 - bdf_2 - bdf_3 - bdf_4 - bdf_5 - bdf_6 - imex
preconditioner update frequency	integer	100	Frequency at which the preconditioner gets updated.
RKC n stages	integer	5	Number of stages for the RKC scheme.
epsilon	number	0.153846	Damping parameter for the RKC scheme.
nlsolve	object		See table
linear solver	object		See table

advection diffusion: time integration: nlsolve

Parameter	Type	Default	Description
max nonlinear iterations	integer	10	Set the number of maximum nonlinear iterations with standard tolerances.
field correction tolerance	number	1e-10	Set the tolerance for the maximum allowed correction of the unknown field.
residual tolerance	number	1e-09	Set the tolerance for the maximum allowed residual of the nonlinear system.
max nonlinear iterations alt	integer	0	Set the number of maximum nonlinear iterations with alternative tolerances.
field correction tolerance alt	number	1e-09	Set the alternative tolerance for the maximum allowed correction of the unknown field.
residual tolerance alt	number	1e-08	Set the alternative tolerance for the maximum allowed residual of the nonlinear system.
use relative tolerances	boolean	False	Set to true to interpret the residual tolerances relative to the norm of the initial residual and the field correction tolerances relative to the norm of the current solution. If the respective reference norm is zero, the tolerance is applied as an absolute value.
verbosity level	integer	-1	Set to one for detailed solver output.

advection diffusion: time integration: linear solver

Parameter	Type	Default	Description
solver type	string	GMRES	Set this parameter for choosing an iterative linear solver type. Allowed values: - CG - GMRES
preconditioner type	string	Identity	Set this parameter for choosing a preconditioner type. Allowed values: - Identity - AMG - ILU - Diagonal
max iterations	integer	10000	Set the maximum number of iterations for solving the linear system of equations.
rel tolerance	number	1e-12	Set the relative tolerance for a successful solution of the linear system of equations.
abs tolerance	number	1e-20	Set the absolute tolerance for a successful solution of the linear system of equations.
do matrix free	boolean	True	Set this parameter if a matrix free solution procedure should be performed.
monitor type	string	none	Set the monitor type of the linear solver. Allowed values: - none - reduced - all

---

🔷 output

Parameter	Type	Default	Description
directory	string	./	Sets the base directory for all output.
write frequency	integer	1	Every n timestep that should be written
write time step size	number	1.79769e+308	Write output output every given time step. If this parameter is set, the output write frequency is deactivated.
output variables	string	all	Specify variables that you request to output.
do user defined postprocessing	boolean	False	Set this parameter to true to enable user defined postprocessing.
paraview	object		See table
particles	object		See table

output: paraview

Parameter	Type	Default	Description
enable	boolean	False	Set this parameter to true to activate paraview output.
filename	string	solution	Sets the base name for paraview output files.
n digits timestep	integer	4	Number of digits for the frame number of the vtu-file.
print boundary id	boolean	False	Set this parameter to true to output a vtu-file with the boundary id.
output subdomains	boolean	False	Set this parameter to true to output the subdomain ranks.
output material id	boolean	False	Set to true to output the material id.
write higher order cells	boolean	True	Set this parameter to false to write bi- or trilinear data only. Set this parameter to true to write higher order cell data. Note: higher order cell data can only be written for hexahedron meshes and 2 or 3 dimensions.
n groups	integer	1	Number of parallel written vtu-files.
n patches	integer	0	Control number of patches to enable high-order.

output: particles

Parameter	Type	Default	Description
enable	boolean	False	Set this parameter to true to activate particle paraview output.
filename	string	particle	Sets the base name for particle output files.

---

🔷 profiling

Parameter	Type	Default	Description
verbosity	string	none	Sets the verbosity of the profiling information to be collected. If set to ‘none’, no profiling information is collected. If set to ‘basic’, only data which is cheap to collect is recorded. If set to ‘detailed’, also data which is expensive to collect is recorded. It is recommended to enable detailed profiling only when necessary for in-depth analysis or debugging purposes. Allowed values: - none - basic - detailed
write time step size	number	10.0	Write profiling output every given time step size. If this parameter is set, the specified parameter for write frequency is overwritten.
time type	string	real	Choose the type of time measure to write profiling information. Allowed values: - real - simulation

---