Metafor

ULiege - Aerospace & Mechanical Engineering

User Tools

Site Tools

You are here: metafor.ltas.ulg.ac.be » doc » Documentation » integration » general » Managing N.-R. Iterations
Trace:
doc:user:integration:general:mim_tim

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
doc:user:integration:general:mim_tim [2013/07/12 15:26] – jorisdoc:user:integration:general:mim_tim [2024/03/11 19:04] (current) – papeleux
Line 1: Line 1:
 +====== Managing N.-R. Iterations ======
  
 +Mechanical iterations parameters are imposed by functions associated to ''metafor.getMechanicalIterationManager()'', when thermal iterations parameters are imposed by functions associated to  ''metafor.getThermalIterationManager()''.
 +
 +    mim = metafor.getMechanicalIterationManager() 
 +    tim = metafor.getThermalIterationManager()
 +    
 +    mim/tim.setResidualComputationMethod(meth)    
 +    mim/tim.setVerbose()     
 +    mim/tim.setResidualTolerance(prec)
 +    mim/tim.setMaxNbOfIterations(itma)
 +    mim/tim.setNbOfIterationsForStiffUpdate(irea)
 +    mim/tim.setMaxNbOfLineSearchIterations(lsma)
 +    mim/tim.setLineSearchTolerance1(lsp1)
 +    mim/tim.setLineSearchTolerance2(lsp2)
 +    mim.setPredictorComputationMethod(predMeth)
 +    mim.setConstrainedDofsVAComputationMethod(VACMeth)
 +    mim/tim.setPreviousResidualsDecreaseCriterion()
 +    mim/tim.setCpuDependency(cpuDep)
 +    mim/tim.setForceOneIte(bool)  
 +where ''meth'' defines the method used to compute the $Residual$.
 +
 +=== Method1ResidualComputation(limitNormFactor) ===
 +<note important>important : since revision 2946 Method1ResidualComputation is **not anymore** the default value</note>
 +
 +The residual is adimensionalized by the norm of external forces and internal reactions .
 +First :
 +
 +$$Rmin = \frac{||FreeExternalForces|| + ||CstrInternalForces||+||CstrInertialForces||}{nreac}  $$
 +
 +If 
 +$$ Rmin < limitNormFactor $$
 +
 +$Rmin$ is set to
 +$$ Rmin = limitNormFactor $$
 +
 +and the residual is adimensionalized :
 +
 +$$Residual = \frac{||FreeResidual||}{ndofs*Rmin}$$
 +
 +where
 +
 +$ndofs$ : total number of degrees of freedom
 +
 +$nreac$ : number of fixed degrees of freedom (having an associated reaction)
 +
 +=== Method2ResidualComputation(limitNormFactor) ===
 +
 +The residual is adimensionalized by the square of the norm of external forces and internal reactions. First:
 +
 +$$Rmin = \frac{\sqrt{||FreeExternalForces||^2 + ||CstrInternalForces||^2+||CstrInertialForces||^2}}{nreac}$$
 +
 +If
 +$$ Rmin < limitNormFactor $$
 +then 
 +$$ Rmin = limitNormFactor $$
 +
 +and the residual is adimensionalized :
 +
 +$$Residual = \frac{||FreeResidual||}{ndofs*Rmin}$$
 +
 +where
 +
 +$ndofs$ : total number of degrees of freedom
 +
 +$nreac$ : number of fixed degrees of freedom (having an associated reaction)
 +
 +=== Method3ResidualComputation(limitNormFactor) ===
 +
 +The residual is adimensionalized by the norm of external and inertial forces, and internal reactions (**value to use preferentially for thermal simulations. Careful, the user must define it**). 
 +
 +If 
 +$$ ||FreeExternalForces|| < limitNormFactor $$
 +then 
 +$$ Residual = max(|FreeResidual|) $$
 +Else,
 +$$Rmin = ||FreeExternalForces|| + ||CstrInternalForces||+||CstrInertialForces|| $$
 +and
 +$$Residual = \frac{||FreeResidual||}{ndofs*Rmin}$$
 +
 +=== Method4ResidualComputation(limitNormFactor) (**default value**) ===
 +<note important>important : since revision 2946 Method4ResidualComputation is the new **default value**</note>
 +This method is the same as ''Method1ResidualComputation()'' but without dividing by the number of dofs nor nreac, so
 +
 +$$Rmin = ||FreeExternalForces|| + ||CstrInternalForces||+||CstrInertialForces|| $$
 +
 +If
 +$$ Rmin < limitNormFactor$$
 +then 
 +$$ Rmin = limitNormFactor $$
 +
 +and the residual is adimensionalized :
 +
 +$$Residual = \frac{||FreeResidual||}{Rmin}$$
 +
 +=== Method5ResidualComputation(adimFactor) ===
 +
 +The residual is adimensionalized by the factor ''adimFactor'', given by the user, without dividing by the number of dofs.
 +
 +$$Residual = \frac{||FreeResidual||}{adimFactor }$$
 +
 +=== Method6ResidualComputation(adimFactor) ===
 +
 +The residual is adimensionalized by the factor ''adimFactor'', given by the user, and by the number of dofs.
 +$$Residual = \frac{||FreeResidual||}{ndofs * adimFactor }$$
 +
 +where
 +
 +$ndofs$ : total number of dofs
 +
 +=== Parameters ===
 +^ Parameter ^ Default value ^  Description  ^ 
 +| ''limitNormFactor'' |  1.0  | Value from which the residual is no longer divided by $$Rmin$$ (to adjust as a function of the loading using verbose output)|
 +| ''adimFactor'' |  -  | Fixed adimensionalization factor for methods #5 and 6 (no default value, the factor must be defined for these methods)|
 +| ''prec'' |  1.0E-4  | accuracy that the residual must reach |
 +| ''itma'' |  7  | maximal number of Newton-Raphson iterations (this number if corrected internally if the automatic criterion is used to update the matrix). Iterations are stopped if the residual does not decrease over 6 iterations|
 +| ''irea'' |  1  |= 1 : stiffness matrix updated for each iteration|
 +|     |     | > 1 : stiffness matrix is updated automatically if the iteration number is lower than ''irea''. The criterion is based on a convergence speed and a CPU time ratio (requires ''MDE_ICPU=1'')|
 +| ''lsma'' |  0  | maximal line-search iteration number|
 +| ''lsp1'' |  1.0  | accuracy that the line-search residual must reach  |
 +| ''lsp2'' |  1.E-8  | stop criterion when the line search parameter is not updated enough  |
 +| ''predMeth'' |  ''EXTRAPOLATION_DEFAULT''  | ''EXTRAPOLATION_DEFAULT'' = Extrapolation of the scheme by default (for example method to compute the displacement predictor by Newmark formula in the case of  [[doc:user:integration:scheme:dynimpl#la_famille_alpha-generalisee|alpha-generalized scheme]])  |
 +|              |                        | ''EXTRAPOLATION_MRUA'' = method to compute the displacement predictor by MRUA formula (for now only possible for [[doc:user:integration:scheme:dynimpl#la_famille_alpha-generalisee|alpha-generalized algorithms]]) |
 +|              |                        | ''EXTRAPOLATION_ZERO'' = New positions are equal to the old ones, speed (and acceleration for dynamic schemes) equal to zero (for now only possible for  [[doc:user:integration:scheme:quasistatique|quasi-static algorithms]] and [[doc:user:integration:scheme:dynimpl#la_famille_alpha-generalisee|alpha-generalized algorithms]]) . |
 +| ''VACMeth'' | ''VAC_SCHEMECONSISTANT'' | ''VAC_SCHEMECONSISTANT'' : Calculation of speed and acceleration with Newmark formula | 
 +|             | | ''VAC_BACKWARDEULER'' : Calculation of speed and acceleration using Backward Euler (to stabilize Newmark with dofs with a non zero fixed displacement. See JPP's thesis pg VIII.28) |
 +| ''cpuDep'' | ''False'' | Specify if the criterion for updating the stiffness matrix (irea > 1) depends on the User CPU time. (Metafor Version > 2422). | 
 +# ''forceOneIte'' | ''False'' | If True, Force the resolution of at least one iteration (mechanical or thermal according to the iteration manager even if the tolerance is allready met at iteration 0. |
 + 
doc/user/integration/general/mim_tim.txt · Last modified: 2024/03/11 19:04 by papeleux
Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki