Step control parameters

@REUSE_NUMBER {nreuse}


  1. The parameters described in this section must be defined for each time step of the analysis. All parameters are optional and are set to a default values at the beginning of the simulation. A new definition of a parameter overrides any previous definition. The overall structure of the time stepping procedure is as follows
  2. Begin loop over time steps
    Read new step control parameters as controlled by nreuse
    Begin loop over rejections
    Begin loop over iterations
    Factorize stiffness matrix as controlled by nfact
    if solution is converged: end loop over iterations;
    else: reject time step, select new time step size, start next iteration.
    End loop over iterations
    if solution is acceptable: end loop over rejections;
    else: adjust time step size or contact conditions, start next rejection.
    End loop over rejections
    End loop over time steps
  3. The time step size Δt is the time step increment used in the time stepping procedure (loop over the time steps). (Default value: Δt = 1 ms)
  4. The step control parameters for a time step are often the same for a large number of consecutive time step. To avoid the repetition of identical data for many time steps, the step control parameters defined in at a specific time step will be reused nreuse times. For instance, if nreuse = 100, the present step control parameters will be used for 100 time steps. A new @STEP_CONTROL_PARAMETER_NAME section should then be appear to define the step control parameters for time step 101. (Default value: nreuse = 0).
  5. At each time step, the solution is deemed to be acceptable or not. If not acceptable, the solution is rejected. The solution can be rejected a maximum of maxrej times. (Default value: maxrej = 5.) A solution is not acceptable when
    • The solution fails to converge within the allowable number of iterations. In this case, the solution is rejected and a new iteration loop is started with a new time step size. If time adaptivity is used, the new time step size is selected by the time adaptivity procedure. Otherwise, the time step is halved. In all cases, the new time step size must satisfy the bounds defined in the finite element control parameters.
    • Specific conditions are met: typically, a contact condition becomes active. In this case, the solution is rejected and a new iteration loop is started with a contact condition.
  6. The factorization strategy for the solution of the nonlinear equations of motion is controlled by the parameter nfact. (Default value: nfact = 5.)
    • If nfact = 1 a full Newton iteration scheme is used, i.e., the global stiffness matrix is computed and factorized at each iteration;
    • If nfact > 1 a modified Newton iteration scheme is used, i.e., the global stiffness matrix is computed and factorized every nfact iterations.
  7. The loop over iteration will perform a maximum of itmax iterations. If convergence is not reached within this number of iterations, the time step is rejected. (Default value: itmax = 12).
  8. Exit of the loop over iterations is controlled by a convergence criterion that is defined by four parameters ConvNormType, εconv, Lref, Fref, and Eref.
  9. The conditioning of the governing equations of the problem is improved by scaling the constraint equations of the problem. Parameters kave and mave are used to normalize the constraint equations of the problem. (Default value: kave = 1.0 10+06, mave = 0.0)
  10. It is often the case that the stiffness matrix describing the static behavior of multibody systems is singular. Indeed, the stiffness matrix associated with a system presenting rigid body modes is singular. It is possible to remove these singularities by adding fictitious spring connections to the ground during the solution process. (Default value: pf = 0.0, i.e., singularities will not be removed during the solution process)
  11. When a time step is successfully completed, the equilibrium configuration of the system is archived in a file. This file can become very large if the results are archived at each time step. The archival frequency parameter, narv, controls this archival rate. (Default value: narv = 1).
  12. The last four parameters, neig, eigpflag, @GYROSCOPIC_TERMS, and ρ control the eigen analysis procedure.