Collaboration diagram for Time step options and variables:

Variables
integer(c_int), pointer, save	ntpabs
	Absolute time step number for previous calculation. More...

integer(c_int), pointer, save	ntcabs
	Current absolute time step number. In case of restart, this is equal to ntpabs + number of new iterations. More...

integer(c_int), pointer, save	ntmabs
	Maximum absolute time step number. More...

integer(c_int), pointer, save	ntinit
	Number of time steps for initalization (for all steps between 0 and ntinit, pressure is re-set to 0 before prediction correction). More...

real(c_double), pointer, save	ttpabs
	Absolute time value for previous calculation. More...

real(c_double), pointer, save	ttcabs
	Current absolute time. More...

real(c_double), pointer, save	ttmabs
	Maximum absolute time. More...

integer(c_int), pointer, save	iptlro
	Clip the time step with respect to the buoyant effects. More...

integer(c_int), pointer, save	idtvar
	option for a variable time step More...

real(c_double), pointer, save	dtref
	Reference time step. More...

real(c_double), pointer, save	coumax
	maximum Courant number (when idtvar is different from 0) More...

real(c_double), pointer, save	cflmmx
	maximum Courant number for the continuity equation in compressible model More...

real(c_double), pointer, save	foumax
	maximum Fourier number (when idtvar is different from 0) More...

real(c_double), pointer, save	varrdt
	maximum allowed relative increase in the calculated time step value between two successive time steps (to ensure stability, any decrease in the time step is immediate and without limit). Useful when idtvar is different from 0. More...

real(c_double), pointer, save	dtmin
	lower limit for the calculated time step when idtvar is different from 0. Take dtmin = min (ld/ud, sqrt(lt/(gdelta rho/rho)), ...) More...

real(c_double), pointer, save	dtmax
	upper limit for the calculated time step when idtvar is different from 0. Take dtmax = max (ld/ud, sqrt(lt/(gdelta rho/rho)), ...) More...

double precision, dimension(nvarmx), save	cdtvar
	multiplicator coefficient for the time step of each variable More...

real(c_double), pointer, save	relxst
	relaxation coefficient for the steady algorithm relxst = 1 : no relaxation. More...

Detailed Description

Variable Documentation

◆ cdtvar

double precision, dimension(nvarmx), save cdtvar

multiplicator coefficient for the time step of each variable

useless for u,v,w,p
for k,e the same value is taken (value of k)
for Rij, e the same value is taken (value of r11)
Hence, the time step used when solving the evolution equation for the variable is the time step used for the dynamic equations (velocity/pressure) multiplied by cdtvar. The size of the array cdtvar is nvar. For instance, the multiplicative coefficient applied to the scalar 2 is cdtvar(isca(2))). Yet, the value of cdtvar for the velocity components and the pressure is not used. Also, although it is possible to change the value of cdtvar for the turbulent variables, it is highly not recommended.

◆ cflmmx

real(c_double), pointer, save cflmmx

maximum Courant number for the continuity equation in compressible model

◆ coumax

real(c_double), pointer, save coumax

maximum Courant number (when idtvar is different from 0)

◆ dtmax

real(c_double), pointer, save dtmax

upper limit for the calculated time step when idtvar is different from 0.
Take dtmax = max (ld/ud, sqrt(lt/(gdelta rho/rho)), ...)

◆ dtmin

real(c_double), pointer, save dtmin

lower limit for the calculated time step when idtvar is different from 0.
Take dtmin = min (ld/ud, sqrt(lt/(gdelta rho/rho)), ...)

◆ dtref

real(c_double), pointer, save dtref

Reference time step.

This is the time step value used in the case of a calculation run with a uniform and constant time step, i.e. idtvar =0 (restart calculation or not). It is the value used to initialize the time step in the case of an initial calculation run with a non-constant time step(idtvar=1 or 2). It is also the value used to initialise the time step in the case of a restart calculation in which the type of time step has been changed (for instance, idtvar=1 in the new calculation and idtvar = 0 or 2 in the previous calculation).
See Time step modification for examples.

◆ foumax

real(c_double), pointer, save foumax

maximum Fourier number (when idtvar is different from 0)

◆ idtvar

integer(c_int), pointer, save idtvar

option for a variable time step

-1: steady algorithm
0: constant time step
1: time step constant in space but variable in time
2: variable time step in space and in time If the numerical scheme is a second-order in time, only the option 0 is allowed.

◆ iptlro

integer(c_int), pointer, save iptlro

Clip the time step with respect to the buoyant effects.

When density gradients and gravity are present, a local thermal time step can be calculated, based on the Brunt-Vaisala frequency. In numerical simulations, it is usually wise for the time step to be lower than this limit, otherwise numerical instabilities may appear.
iptlro indicates whether the time step should be limited to the local thermal time step (=1) or not (=0).
When iptlro=1, the log shows the number of cells where the time step has been clipped due to the thermal criterion, as well as the maximum ratio between the time step and the maximum thermal time step. If idtvar=0, since the time step is fixed and cannot be clipped, this ratio can be greater than 1. When idtvar > 0, this ratio will be less than 1, except if the constraint dtmin has prevented the code from reaching a sufficiently low value for dt. Useful when density gradients and gravity are present.

◆ ntcabs

integer(c_int), pointer, save ntcabs

Current absolute time step number. In case of restart, this is equal to ntpabs + number of new iterations.

◆ ntinit

integer(c_int), pointer, save ntinit

Number of time steps for initalization (for all steps between 0 and ntinit, pressure is re-set to 0 before prediction correction).

◆ ntmabs

integer(c_int), pointer, save ntmabs

Maximum absolute time step number.

For the restart calculations, ntmabs takes into account the number of time steps of the previous calculations. For instance, after a first calculation of 3 time steps, a restart file of 2 time steps is realised by setting ntmabs = 3+2 = 5

◆ ntpabs

integer(c_int), pointer, save ntpabs

Absolute time step number for previous calculation.

In the case of a restart calculation, ntpabs is read from the restart file. Otherwise, it is initialised to 0 ntpabs is initialised automatically by the code, its value is not to be modified by the user.

◆ relxst

real(c_double), pointer, save relxst

relaxation coefficient for the steady algorithm relxst = 1 : no relaxation.

◆ ttcabs

real(c_double), pointer, save ttcabs

Current absolute time.

For the restart calculations, ttcabs takes into account the physical time of the previous calculations.
If the time step is uniform (idtvar = 0 or 1), ttcabs increases of dt (value of the time step) at each iteration. If the time step is non-uniform (idtvar=2), ttcabs increases of dtref at each time step.
ttcabs} is initialised and updated automatically by the code, its value is not to be modified by the user.

◆ ttmabs

real(c_double), pointer, save ttmabs

Maximum absolute time.

◆ ttpabs

real(c_double), pointer, save ttpabs

Absolute time value for previous calculation.

In the case of a restart calculation, ttpabs is read from the restart file. Otherwise it is initialised to 0.
ttpabs is initialised automatically by the code, its value is not to be modified by the user.

◆ varrdt

real(c_double), pointer, save varrdt

maximum allowed relative increase in the calculated time step value between two successive time steps (to ensure stability, any decrease in the time step is immediate and without limit).
Useful when idtvar is different from 0.

Variables