Operands ========= Operands CHAM_MATER --------------------- .. code-block:: text ♦ CHAM_MATER = mater_field The name of the material field assigned on the mo model. .. warning:: **Note:** Make sure that the chmat was created using the same mesh as that of the ``model`` model. Otherwise, a fatal error will occur. Operand METHODE_STAB = 'SRM' ----------------------------- .. code-block:: text ♦ METHODE_STAB = 'SRM' By choosing method SRM, the slope stability analysis is based on the non-linear solver STAT_NON_LINE in order to search iteratively for the safety factor until the calculation diverges. Consequently, the entries of this last one command are also required by the macro command. Operand MODELE ~~~~~~~~~~~~~~~~ .. code-block:: text ♦ MODELE = model Name of the model whose elements are the subject of mechanical calculation. With model THM (see [:ref:`R7.01.11 `]), the macro command takes into account the effect hydraulics during the stability analysis, which especially favors the evaluation the stability of dams and dikes. .. warning:: On the other hand, models requiring the definition of elementary characteristics via The AFFE_CARA_ELEM command is not authorized by the CALC_STAB_PENTE macro command. Operand GROUP_MA/TOUT ~~~~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ♦/GROUP_MA = Grma /TOUT = 'OUI' List of mesh groups defining the zone where algorithm SRM applies. If we analyze the stability of the entire model, we configure the operand TOUT = 'OUI'. In case the mesh groups differ from those used in Assigning the material to the mesh, the macro command will identify the materials affecting zone SRM based on the input material field ``chmat``, and calculate the groups of the intersecting cells between zone SRM and the zone of allocation of the material in question in order to affect the degraded material in zone SRM by the principle of overload. Keyword EXCIT ~~~~~~~~~~~~~ .. code-block:: text ♦ EXCIT = _F () Keyword factor used to describe a load at each occurrence (stress and boundary condition), and possibly a multiplying coefficient and/or a type of load. The configuration of this factor keyword is identical to the STAT_NON_LINE command (see [:ref:`u4.51.03 `]), **except that load types are limited between 'FIXE_CSTE', 'SUIV', and 'DIDI '**. .. important:: The load types related to feature PILOTAGE have been eliminated in the macro command CALC_STAB_PENTE since they are in contradiction with the SRM algorithm. Keywords INCREMENT, CONVERGENCE, and COMPORTEMENT ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ♦ INCREMENT = _F () Keyword: factor defining the time intervals taken in the method incremental (see STAT_NON_LINE [:ref:`u4.51.03 `]) .. code-block:: text ♦ CONVERGENCE = _F () Keyword: factor for controlling convergence criteria during the non-linear calculation performed by the command STAT_NON_LINE ([:ref:`u4.51.03 `]). .. code-block:: text ♦ COMPORTEMENT = _F () Keyword factor defining behavioral relationships. It is a keyword that is common to several orders. Its syntaxes are described in document [:ref:`U4.51.11 `]. **Note:** 1. The use of the option of automatic division of the defined time steps by the command DEFI_LIST_INST with the operand ECHEC/ACTION = 'DECOUPE' (in the key word factor INCREMENT) is highly recommended. .. to be reworded below in order to better capture the moment of collapse of the slopes, especially since the law of behavior DRUCKER_PRAGER avoid the possible singularity of the tangent matrix caused by the load surface From the law of behavior MOHR_COULOMB automatic division of the increment at several levels excludes The numerical increment factor that probably leads to the discrepancy of the nonlinear calculation and therefore reassures the relevance of the stability analysis result. Likewise, it is also favorable to increase the ITER_GLOB_MAXI operand in the CONVERGENCE keyword to ensure that the discrepancy of calculus is caused mainly by failure in the continuous medium. 2. In the keyword factor COMPORTEMENT it is possible to define relationships different behaviors on different parts of the model. However, similar to the treatment of the material field in zone SRM, macro command CALC_STAB_PENTE looks at behavioral relationships in zone SRM so that there is one and only one behavior among MOHR_COULOMB and DRUCKER_PRAGER in zone SRM. If not, a fatal error will occur. In addition, the definition of behavioral relationships on meshes is not allowed by the macro command CALC_STAB_PENTE. FS keyword ~~~~~~~~~~ .. code-block:: text ◊ FS= _F () Keyword factor defining the parameters of algorithm SRM, including the value initial of the safety factor, the initial increment, and the refinement method. Operand FS_INIT ^^^^^^^^^^^^^^^^ .. code-block:: text ◊ FS_INIT = f0 This operand makes it possible to fill in the initial value of the safety factor. **Note:** 1. This is the coefficient of reduction in shear strength properties for the first iteration SRM. So the closer this initial value is to the factor of real security, the less the calculation will take. This value is often obtained by empirical method. or by prior testing, but can never exceed the real value, otherwise the calculation will diverge at the first iteration SRM and a fatal error will be produced. ``f0`` must be strictly positive and the value by default is 1.0. 2. In case of the fatal error informing that ``f0`` is too small, you should not intuitively step Increase it and restart calculation SRM. It is favorable to examine at first the configuration of the non-linear calculation, since it is probable that the discrepancy of the calculation to the first iteration SRM is due to a misconfiguration of the solver STAT_NON_LINE. Operand INCR_INIT ^^^^^^^^^^^^^^^^^^^ .. code-block:: text ◊ INCR_INIT = p0 This operand allows you to define the initial increment of the safety factor. Operand RESI_MAXI ^^^^^^^^^^^^^^^^^^^ .. code-block:: text ◊ RESI_MAXI = fine This operand makes it possible to define the maximum residue of the safety factor. The safety factor finally obtained is ``pfin`` close to its real value. This operand therefore defines the precision of the stability calculation. Its value by default is fixed at 0.01 according to the habit of the hydraulic profession. The significant difference between p0 and ``pfin`` will lead to a long calculation time if we do not impose a step the number of refinements using the ITER_RAFF_LINE keyword. Operand ITER_MAXI ^^^^^^^^^^^^^^^^^^^ .. code-block:: text ◊ ITER_MAXI = iter_max This operand defines the maximum number of iterations SRM. It is used to avoid an inefficient calculation due to the inappropriate configuration of the FS keyword. The calculation stops with the fatal error if iter_max is reached without the algorithm having converged. Operand METHODE ^^^^^^^^^^^^^^^^ .. code-block:: text ◊ METHODE = /' EXPONENTIELLE ' /' LINEAIRE ' This operand defines the law of variation of the increment of the safety factor. Two options are available: 1. **' EXPONENTIELLE '**: Exponential change The safety factor increment approaches the maximum residue exponentially. At the nth refinement, the increment is calculated by: :math:`{p}_{n}={p}_{0}{2}^{-n}` until :math:`{p}_{N}<{p}_{\mathit{fin}}`. 2. **' LINEAIRE '**: Linear variation The safety factor increment approaches the maximum residue linearly. In this case an additional ITER_RAFF_LINE operand is required. At the nth refinement, the increment is calculated by: until :math:`{p}_{N}={p}_{\mathit{fin}}`. :math:`{p}_{n}={p}_{0}-n\dfrac{{p}_{0}-{p}_{\mathit{fin}}}{N}` The advantage of the exponential method is that the calculation converges more quickly, while the linear method better controls the accuracy of the result. The exponential variation of the increment is especially adapted to the calculation with big INCR_INIT. The following method would simplify the choice between the two laws of variation of the increment. Let :math:`N` be the desired number of refinements, we choose the linear law if: :math:`\dfrac{{p}_{0}-{p}_{\mathit{fin}}}{{\mathrm{log}}_{2}({p}_{0})-{\mathrm{log}}_{2}({p}_{\mathit{fin}})}`]. The user should ensure that the slope profile provided includes the bands to search for entry and exit points, and that it meets the needs mentioned above. It is advisable to choose a slope profile that contains the entire slope as well. only a portion of the flattened surfaces upstream and downstream. An example of the slope profile is shown in :numref:`fig1-panorama-lem`. .. figure:: images/10000000000007EA0000034BD7652DBA91FA4C04.jpg :name: fig1-panorama-lem :width: 80% Geometric parameters illustration LEM .. _RefImage_10000000000007EA0000034BD7652DBA91FA4C04.jpg: Figure 1:. Operand NB_TRANCHE ~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ◊ NB _ TRANCHE =/10 /Net This operand defines the number of slices dividing the slope. Note that calculation LEM will be all the slower the more you increase the number of slices. However, the number of slices must be increased. if we analyze the stability of stratified structures including the fracture surface is probably non-circular. Operands X1_ MINI, X1_ MAXI, X2_, X2_ MINI and X2_ MAXI ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ♦ X1 _ MIN I = x1_min ♦ X1 _ MAX I = x1_max ♦ X2 _ MIN I = x2_min ♦ X2 _ MAX I = x2_max These operands define the search bands for the x-axis of the points. entry and exit from the potential rupture surface. By convention, ``x1_min`` and ``x2_min`` should be less than ``x1_max`` and ``x2_max`` Respectively. An example of the definition of the X1 and X2 is shown in :numref:`fig1-panorama-lem`. Keyword RAFF_MAIL ~~~~~~~~~~~~~~~~~~ .. code-block:: text RAFF_MAIL = _F ( ◊ NB_RAFF_MAXI =/4 /N raff ◊ RAFF_CRIT_STAB =/1e-3 /raff_ crit_stab ) Keyword factor defining the parameters related to the mesh refinement algorithm on the edge of the slices. Refinement via macro command MACR_ADAP_MAIL is controlled by the proximity field measuring the distance between the nodes and the outline of the slices. The complete procedure is shown in [:ref:`R7.05.02 `]. Operand NB_RAFF_MAXI defines the maximum refinement number. This operand makes it possible to limit the number of refinements in order to be able to consult the FS result more quickly in exchange for the precision of the calculation if the FS convergence criterion is not met. The RAFF_CRIT_STAB operand defines the criterion for stabilizing the safety factor during the refinement of the mesh. The algorithm stops if the absolute value of FS variation is less than ``raff_crit_stab``. .. _RefNumPara__16641_1595115202: Operand METHODE_LEM = 'BISHOP' or 'FELLENIUS' ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ◊ METHODE_LEM = /' BISHOP ' /' FELLENIUS ' The METHODE_LEM operand allows you to switch between the different procedures stability calculation using the LEM method. This operand determines at the same time the objective function of the optimization algorithm. When METHODE_LEM = 'BISHOP' or 'FELLENIUS', we do the assumption that the fracture surface is circular and the factor is calculated of security assuming that the interaction forces between the slices are horizontal (BISHOP) or none (FELLENIUS). Since the surface is circular, we look for the critical surface by scanning all combinations of entry and exit points. For each combination, we automatically calculate the minimizing radius FS. Two approaches are available to determine radius test bands: * **Option 1**: We scan all the rays that form a breaking surface kinematically admissible. * **Option 2**: We define the test band manually by defining the horizontal lines tangential to the fracture surfaces. This search option speeds up the calculation if the experienced user provides a reduced area where the fracture surface is likely to be located. We also take advantage of this option to examine the surfaces around that given by the result. The [:ref:`R7.05.02 `] document explains the optimization algorithm in more detail for **Bishop simplified** and **Fellenius** procedures. Operands NB_POINT_1 and NB_POINT_2 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ .. code-block:: text ◊ NB_POINT_1 =/5 /nb_point_1 ◊ NB_POINT_2 =/5 /nb_point_2 The NB_POINT_1 and NB_POINT_2 operands define the numbers of the intermediate points between the entry and exit search bands respectively. For example, we test all the points x1: :math:`{x}_{1}^{i}={x}_{1\mathit{min}}+({x}_{1\mathit{max}}-{x}_{\mathit{min}})/({N}_{1}-1)\ast i\text{}\forall i=\mathrm{0,1},\mathrm{...},{N}_{1}-1` Y_ MINI and Y_ MAXI operands ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ .. code-block:: text ◊ Y_ MIN I = y_min ◊ Y_ MAXI = y_max The operands Y_ MINI and Y_ MAXI define the minimum and maximum ordinates of horizontal lines tangential to the fracture surfaces tested. These operands will be needed if we control manually search for the radius (option 2 mentioned in the section :ref:`4.3.6 `). An example of configuring Y_ MINI and Y_ MAXI is shown in `Figure 1`_. **Note**: Make sure y_min is less than :math:`\mathit{min}\{x{1}_{\mathit{max}},x{2}_{\mathit{min}}\}`. In fact, if the entry and exit points are located on both sides of the tangential line, this combination will be ignored since it will be impossible to create a tangential fracture surface that crosses its tangent line. Operand METHODE_LEM = 'SPENCER' or 'MORGENSTERN - PRICE' ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .. code-block:: text ◊ METHODE_LEM = /' SPENCER ' /' MOGENSTERN - PRICE ' When METHODE_LEM = 'SPENCER' or 'MORGENSTERN - PRICE', we make the assumption that the fracture surface is multi-linear and calculates the safety factor by solving both the balance of strengths and full moments of all the slices. The search for the critical surface is based on the reinforced fireworks algorithm (EFWA) which is part of intelligent swarm algorithms, and therefore requires additional parameters to set up algorithm EFWA. The [:ref:`R7.05.02 `] document explains the EFWA algorithm in detail. .. note:: **Note:** When evaluating the FS of potential fracture surfaces generated by EFWA, it is likely that the surface is physically illegal and that the condition of convergence of fixed point iterations is not satisfied, which does not prevent the location of the global critical surface. So in this case we attribute to this Illegal surface a very large FS (with the emission of an information message) and continues to test other fracture surfaces. In order to prevent EFWA from generating this type of illegal surface, it is better to increase the number of slices so that the surfaces are smooth. However, it must be checked at the same time that the mesh is quite fine, which reassures the existence stitches in which all the knots are included in a reduced size range. Keyword ALGO_EFWA ^^^^^^^^^^^^^^^^^^ .. code-block:: text ALGO_EFWA = _F ( ◊ NB_STAB_MAXI =/10 /stab_maxi ◊ CRIT_STAB =/1st- 3 /efwa_resi ◊ MARGE_PENTE =/0.1 /margin ) Keyword factor defining the parameters related to the optimization algorithm EFWA. .. code-block:: text ◊ ETAT_INIT = tabfs This operand takes as input the output table of the macrocommand CALC_STAB_PENTE, which allows you to initialize algorithm EFWA by adding fireworks to the first generation the position of the critical surface resulting from the previous calculation. This operand is useful in the following two circumstances: 1. Since EFWA is a probabilistic algorithm, reaching the global optimum is not mathematically guaranteed after a limited number of iterations. So CALC_STAB_PENTE offers the possibility of recovering the result before reaching the true one. optimum and to repeat the calculation if necessary. 2. In order to reduce the number of iterations, we can take the result of the procedures designed for circular surfaces (Bishop or Fellenius) as the initial state. This operand makes it possible to introduce a circular surface in the form of a table containing the x-axes of the entry and exit points and the ordinates of the intermediate points. You will therefore have to convert the Bishop or Fellenius output in accordance with the description format. of the non-circular surface before moving on to EFWA. .. code-block:: text ◊ ITER_MAXI =/1st 2 /iter_max The ITER_MAXI operand defines the maximum number of iterations for EFWA. After ``iter_max`` iterations, the algorithm stops even if the minimum FS value has not yet been obtained. .. code-block:: text ♦ A = a ◊ N = /5 /n ◊ M = /40 /m ◊ MG = /5 /mg ◊ SA =/0.04 /Her ◊ SB =/0.8 /sb The operands A, N, M, MG, SA, SB define the parameters related to the generation of ordinary and Gaussian sparks in EFWA. The meaning of these operands is as follows: * Operand A: the total amplitude of the fireworks explosion. It depends on the characteristic dimension of the slope. For example, let H be the height of the slope, it is reasonable to estimate approximately the value of A as: :math:`A=\dfrac{m}{\mathit{nb_tranches}}H` * The N operand: number of fireworks in each generation. * The operand M: total number of ordinary sparks generated when all fires explode fireworks. * The MG operand: number of Gaussian sparks generated when fireworks explode. The challenge of generating Gaussian sparks is to improve the diversity of sparks distributed in the state variable space, which is the main advantage of EFWA compared to other swarm algorithms. * The SA operand: multiplier coefficient determining the lower limit of the number of sparks generated during the explosion of a fireworks display. * The SB operand: multiplication coefficient determining the upper limit of the number of sparks generated during the explosion of a fireworks display. Operand CHAM_DEFO ------------------ .. code-block:: text ◊ CHAM_DEFO =CO (chamdef) Name of the concept ``evol_noli`` containing the field visualizing the fracture surface. 1. **In case METHODE_STAB = 'SRM':** The result ``chamdef`` contains the cumulative plastic deformation field (component V1 of field VARI_NOEU) to be produced by the macro-command. It is possible to print this field using the IMPR_RESU command at the end of the execution of the macro-command to visualize the sliding surface, which corresponds to the zone where the cumulative plastic deformation is significant. 2. **In case METHODE_STAB = 'LEM':** The result ``chamdef`` contains the indicative displacement field (NOEU_DEPL) visualizing the slippery part of the slope. The value 1 is assigned to the DX component if the node is part in the sliding part, and 0 otherwise.