1. What’s new between 9.1 and 9.2#

1.1. New orders#

1.1.1. CALCUL#

The object of this new command is to calculate the elementary tangent matrices, the elementary internal forces and to integrate the law of behavior (cf. [9.1.7]).

Its principle is, knowing U-, SIGMA -, -, VARI - and dU, we calculate:

  • SIGMA +, VARI +, elementary tangent matrices, elementary residue vectors if OPTION =( “COMPORTEMENT”, “FORC_INT_ELEM”, “MATR_TANG_ELEM”)

  • SIGMA +, VARI +, the elementary residue vectors if OPTION =( “COMPORTEMENT”, “FORC_INT_ELEM”)

1.1.2. FACTORISER and RESOUDRE replace FACT_GRAD/FACT_LDLT and RESO_GRAD/RESO_LDLT#

Commands FACT_GRAD and FACT_LDLT have merged into one: FACTORISER.

The same goes for RESO_GRAD and RESO_LDLT which merged in RESOUDRE (cf. [9.1.14]).

The MATR_FACT keyword in RESO_LDLT becomes MATR.

The MATR_ASSE keyword in RESO_GRAD becomes MATR, and MATR_FACT becomes MATR_PREC.

1.2. General changes#

1.2.1. Control variables#

The use of temperature as a control variable in mechanics must now be done using AFFE_MATERIAU/AFFE_VARC/NOM_VARC =” TEMP “instead of AFFE_CHAR_MECA/TEMP_CALCULEE.

The old TEMP_CALCULE syntax will be gone in the next release.

See the reference temperature note § AFFE_MATERIAU/TEMP_REF.

1.2.2. Behavior ELAS_VMIS_PUIS#

This equivalent behavior of VMIS_ISOT_PUIS in linear elasticity has been available since 9.1.15.

The ELAS_VMIS_PUIS keyword (an analogue of ELAS_VMIS_TRAC) appears in commands STAT_NON_LINE, DYNA_NON_LINE, and CALC_G.

1.2.3. Parallel distributed MUMPS solver#

This evolution consists in calling the solver MUMPS in a distributed parallel version as opposed to the centralized parallel version. In both cases, the resolution of the linear system is done in parallel on several nodes and processors of the machine. In the distributed version, the calculation of the elementary matrices is distributed to the various processors, then the assembly and finally the resolution of the system. The distribution method is chosen using the PARALLELISME keyword.

The operators concerned are DYNA_LINE_TRAN, DYNA_NON_LINE, MECA_STATIQUE,, THER_LINEAIRE,, THER_NON_LINE, STAT_NON_LINE.

Friction is excluded from the scope of use. The development is still subject to validation

The solver is only available on the centralized server for now (see [9.1.16]).

1.3. Resorptions#

1.3.1. FACT_GRAD/FACT_LDLT and RESO_GRAD/RESO_LDLT#

Replaced by FACTORISER and RESOUDRE (cf. [9.1.14]).

1.4. Changed orders#

1.4.1. AFFE_CHAR_CINE#

EVOL_IMPO new:

  • Allows you to define the result of a previous calculation as a kinematic load. The targeted application is structural zoom where the movements obtained during the first calculation are imposed on the zoom border (cf. [9.1.14]).

1.4.2. AFFE_CHAR_MECA/AFFE_CHAR_MECA_F#

PRES_REP/FISSUREnouveau:

  • Allows you to apply pressure to a non-meshed crack, defined by DEFI_FISS_XFEM (see [9.1.8]).

CONTACT/GROUP_MA_FONDnouveau:

  • Together with the use of METHODE =” VERIF “, this keyword makes it possible to define the crack background on which it is useless to check the interpenetration (cf. [9.1.11]).

LIAISON_SOLIDE/TRAN, ANGL_NAUT, CENTREnouveaux:

  • These keywords make it possible to impose the displacement of an undeformable part of the model (cf. [9.1.13]).

APPL_FORC_xxxxsupprimés:

  • These keywords were used for loading GRAPPE_FLUIDE (see [9.1.22]).

1.4.3. AFFE_MATERIAU#

AFFE_VARC/FONC_INST new:

  • This keyword was added in order to apply a loading, thermal for example, whose evolution in time is different from that of mechanical calculation. For example, mechanical calculation can make cycles on a thermal evolution. The function provided then makes it possible to pass from loading time to mechanical calculation time (cf. [9.1.5]).

TEMP_REFplus by default value:

In order to rule out the risk of false results with poor consideration of the reference temperature, a calculation is considered to be thermo-mechanical if there is a temperature field on the element in question. In this case it is mandatory to enter both the reference temperature AND the thermal expansion coefficient. If one of these two values is absent, the code will stop in a fatal error with a warning message for the user (see [9.1.15]).

1.4.4. AFFE_MODELE#

MODELISATION: GRILLE_EXCENTRE replaces GRILLE:

  • Consideration of eccentricity (cf. [9.1.4]).

MODELISATION: xx_ XFEM_CONT new:

  • Finite elements xx_ XFEM no longer have a degree of freedom for contact to deal effectively with non-contact problems. The finite elements xx_ XFEM_CONT are added for this (cf. [9.1.11]).

1.4.5. CALC_FONCTION#

FRACTILEnouveau:

  • Calculation of the fractile of sheets or functions (cf. [9.1.8]).

PUISSANCE/EXPOSANTréel:

  • It is now possible to take the real power of a function (cf. [9.1.9]).

1.4.6. CALC_MATR_ELEM#

MECA_GYRO new:

  • Calculation of the gyroscopic damping matrix for the calculation of tree lines (see [9.1.16]).

1.4.7. CALC_TABLE#

VALE_xxxmodifiés:

  • In IMPR_TABLE and CALC_TABLE, it is possible to filter the rows in a table. By repeating the keyword factor FILTRE, an AND is applied between the filters. We can now apply an OR by giving several values to the VALE_xxx keywords (cf. [9.1.3]).

1.4.8. COMB_SISM_MODAL#

MULTI_APPUI new:

  • Allows the taking into account of correlated excitations in multi-support. The user now has the choice between MONO_APPUI =” OUI “, MULTI_APPUI =””, =” CORRELE “(new option) or” DECORRELE “, which corresponds to the option available previously (see [9.1.18]).

1.4.9. CREA_MAILLAGE#

HEXA20_27nouveau:

  • Allows you to transform the HEXA20 elements of a mesh into HEXA27 elements (see [9.1.6]).

1.4.10. CREA_RESU#

OPERATION =” ASSE “new:

  • Allows you to assemble several thermal calculation results, possibly by translating them. This operation can also be completed with AFFE_MATERIAU/AFFE_VARC, FONC_INST to repeat a thermal loading (cf. [9.1.5]).

1.4.11. DEBUT and POURSUITE#

RESERVE_CPUnouvelle value by default:

  • This keyword allows you to reserve a fraction of the total time of a study (requested in astk) at the end of the calculation for system tasks (compression, file copying…). The value is set by default to 10% of the requested time within a limit of 900 seconds (see [9.1.20).

1.4.12. DEFI_FONCTION#

ABSCISSE/ORDONNEEnouveaux:

  • Equivalent to the keywords LIST_PARA and LIST_RESU except that we pass lists of values instead of list_r8 concepts (see [9.1.9]).

1.4.13. DEFI_LIST_ENTI#

OPERATION = “NUME_ORDRE “new:

  • This makes it possible to extract the order numbers of a result concept within a given interval. You can then use this list in all commands that accept the LIST_ORDRE keyword (see [9.1.16]).

1.4.14. DEFI_MATERIAU#

RELATION_KIT/THM_DIFFUnouveau:

  • For unsaturated models and coupling laws LIQU_xxx_GAZ, the user can define his law by choosing either HYDR_UTIL as before, or HYDR_VGM: Mualem Van-Genuchten’s law. In the latter case, he must define 5 parameters (cf. [9.1.4]).

LETKnouveau:

  • This is a viscoplastic law by Lagle and Klein to model the behavior of argillite, the host rock of a storage site (cf. [9.1.17]).

HUJEUXnouveaux parameters:

  • This is the cyclical extension of Hujeux’s law (cf. [9.1.18]).

1.4.15. DYNA_TRAN_MODAL#

NB_MODE_DIAG removed:

  • This keyword was not used (see [9.1.11]).

1.4.16. FORMULE#

VALE_Cnouveau:

  • Offers the possibility of defining a formula with a complex value (cf. [9.1.1]).

1.4.17. IMPR_RESU#

PARTIEétendu:

  • Allows you to select the real or imaginary part that will be printed (see [9.1.2]).

GMSH/TYPE_CHAMP new:

  • Allows you to choose the type of field and the components to be printed in format GMSH: scalar (the default), vector (VECT_2D, VECT_3D) or tensor (TENS_2D, TENS_3D) (cf. [9.1.10]).

1.4.18. IMPR_TABLE#

Cf. CALC_TABLE.

1.4.19. LIRE_RESU#

TYPE_RESU =” EVOL_VARC “new:

  • Allows you to read fields in MED format that will then be used as control variables: temperature, irradiation… (cf. [9.1.17]).

COMP_INCR new:

  • When reading a result, the COMP_INCR keyword allows you to recreate behavior information that cannot be stored in MED format (see [9.1.18]).

1.4.20. MACRO_MODE_MECA and NORM_MODE#

MASS_INER removed:

  • To facilitate the use of the order and avoid the risk of errors, the mass is calculated by NORM_MODE (see [9.1.18]).

1.4.21. MACR_CARA_POUTRE#

GROUP_NOnouveau:

  • Same function as the existing keyword NOEUD (see [9.1.9]).

1.4.22. POST_CHAM_XFEM#

NOM_CHAMétendu:

  • Possibility to post-process internal variable fields (cf. [9.1.4]).

1.4.23. POST_ELEM#

MINMAXnouveau:

  • Allows you to extract the extremes of a component of a field in a table possibly over time. The table contains the location (mesh or node) of the extrema (cf. [9.1.18]).

1.4.24. POST_MAIL_XFEM#

PREF_NŒUD_X /M/P, PREF_MAILLE_Xnouveaux:

  • Allows you to impose the prefix of the node names created by the command (see [9.1.5]).

1.4.25. POST_RCCM#

NUME_PASSAGE removed:

The use of NUME_PASSAGE is no longer possible for type TUYAUTERIE. In*Code_Aster, we only treat simple paths of passage for the moment (cf. [9.1.22]).

1.4.26. TEST_FICHIER (developer command)#

NB_CHIFFRE, EPSILONsupprimés, **** TYPE_TEST, NB_VALE, VALEnouveaux: **

  • See order documentation (see [9.1.21]).

1.4.27. THER_LINEAIRE, THER_NON_LINE, THER_NON_LINE_MO#

Syntax approximation with STAT_NON_LINE (cf. [9.1.19]):

  • TEMP_INIT/NUME_INIT becomes ETAT_INIT/NUME_ORDRE.

  • Under INCREMENT, NUME_INIT and NUME_FIN become NUME_INST_INIT and NUME_INST_FIN.

The redistribution of the time step is available and works under the same conditions as STAT_NON_LINE (cf. SUBD_METHODE).

For THER_NON_LINEuniquement, OBSERVATION is available with the same syntax and operation as in STAT_NON_LINE.

In THER_NON_LINE_MO, the only change is for ETAT_INIT/NUME_ORDRE (because there is no time step list).