Postprocessing According To The RCC-M

Version
Code_Aster default
Titre : Post-traitement selon le RCC-M Date : 10/06/2011 Page : 1/37
Responsable : The-Hiep CHAU Clé : R7.04.03 Révision : 6503
Postprocessing according to the RCC-M
Résumé:
Operator POST_RCCM [U4.67.04] allows to check the criteria of level 0 and certain criteria of level A of the
B3200 chapter of the RCC - M, for modelizations of continuums 2D or 3D. He also allows the computation of
the criteria of fatigue of the B3600 chapter in postprocessing of computations of pipework.
The criteria of level 0 aim at securing the material against the damage of excessive strain, plastic instability
and elastic and elastoplastic instability. These criteria require the computation of the equivalent stresses of
membrane Pm , local membrane Pl and membrane plus bending PmPb . These parameters are
computed by command POST_RCCM, and are compared with limiting values.
The criteria of level A aim as for them at securing the material against the damage of progressive strain and
fatigue. Except fatigue, they require the computation of the amplitude of variation of stress linearized, noted
Sn , and possibly of the quantity Sn* and the thermal ratchet. For fatigue, they require in more computation
of the amplitude of variation of stress in a point, noted Sp . A fatigue model (standard Wohler) then makes it
possible to deduct the factor of it of use by taking of to account the various groups of situations defined by the
user.
In addition, operator POST_RCCM allows of compute the factor of priming the level of a singular area, within the
meaning of appendix ZD of the RCC-M. This computation is available for modelizations of continuums 2D or
3D.
Warning : The translation process used on this website is a "Machine Translation". It may be imprecise and inaccurate in whole or in part and is
provided as a convenience.
Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

Version
Code_Aster default
Contents
1 Introduction4..........................................................................................................................................
2 Mise in data in Code_Aster5.................................................................................................................
2.1 Données géométriques5.................................................................................................................
2.2 Données of loading - Type EVOLUTION5......................................................................................
2.3 Données of loading - Type UNITAIRE6..........................................................................................
2.4 Données matKe mixteériau6..........................................................................................................
2.5 Hypothèses simplificatrices6..........................................................................................................
2.6 Computations carried out by POST_RCCM7.................................................................................
3 Criteria of level 0 (key word PM_PB)................................................................................................. 8
3.1 Criteria of level 0 specified by................................................................................... the RCC-M8
3.1.1 general primary Equivalent stress of membrane primary......................................................
Pm8 3.1.2 Equivalent stress of local membrane Pl8......................................................................
3.1.3 primary Equivalent stress of membrane+flexion Pmb (or Plb)........................................... 8
3.2 Computations carried out in the case ` EVOLUTION'...................................................................9
3.3 Computations carried out in the case ` UNITAIRE'.......................................................................9
4 Criteria of level A for the results of the type ` UNITAIRE'.................................................................10
4.1 Definition of the criteria in........................................................................................ the RCC-M10
4.1.1 Computation of Sn10.............................................................................................................
4.1.2 Computation of Sn*11...........................................................................................................
4.1.3 Computation of the ratchet thermique11...............................................................................
4.1.4 Computation of the factor of usage11...................................................................................
4.1.4.1 Principle général11....................................................................................................
4.1.4.2 Computation of the factors of use élémentaires12....................................................
4.2 Hypothèse simplifying in the case ` UNITAIRE'..........................................................................13
4.3 Computations carried out with the option ` SN'...........................................................................14
4.4 Computations carried out with the option ` FATIGUE'................................................................15
4.4.1 Combination of the situations inside each group of situations15...........................................
4.4.1.1 Computation of Sp (p, Q)..................................................................................... 15
4.4.1.2 Computation of Sn (p, Q)..................................................................................... 16
4.4.1.3 Computation of Spméca (p, Q) and Spther (p, Q)................................................ 16
4.4.1.4 Computation of alt (p, Q)...................................................................................... 16
4.4.2 Prise in account of the séisme17...........................................................................................
4.4.3 Situations of transition between groups of situations17.........................................................
4.4.4 Stockage18............................................................................................................................
4.4.5 Computation of the factor of total18 use...............................................................................

Version
Code_Aster default
5 Criteria of level A for the results of the type ` EVOLUTION'.............................................................20

5.1 Definition of the critères20..............................................................................................................
5.1.1 Computation of Sn20............................................................................................................
5.1.4 Computation of the factor of use according to appendix ZH21020.......................................
5.1.5 Computation of the factor of priming ..........................................................................20.5.2
Computations carried out with the option ` SN'.................................................................................21
5.3 Computations with fatigue with the option ` FATIGUE_ZH210' .....................................................
22.5.3.1 Computation of the factors of use élémentaires23...........................................................
5.3.2 Computation of the factor of use total23...............................................................................
5.3.3 Exemple24............................................................................................................................
5.4 Computations of the factor of priming with the option ` AMORCAGE'.......................................25
6 Criteria of fatigue for the analysis simplified of the pipework according to the RCC-M preliminary......
B360026 6.1 Computation of all the states of chargement26...............................................................
6.1.1 Computations of the states of loading statiques26................................................................
6.1.2 Computation of the loadings sismiques26.............................................................................
6.1.3 Computation of the transients thermiques27.........................................................................
6.2 Computations of the amplitudes of variation of stresses ...............................................................
28.6.2.1 Principle of the méthode28...............................................................................................
6.2.2 Computation of the combinations of loading (I, J) inside each group of situations29...........
6.2.2.1 Notations and définitions29.......................................................................................
6.2.2.2 Computation of .....................................................................................................29
6.2.2.3 Computation of .....................................................................................................30
6.2.2.4 Computation of .....................................................................................................30
6.2.2.5 Cas of............................................................................................... under-cycles31
6.2.3 Computation of the combinations of loading (I, J) for the situations of transition between
group of situations31.............................................................................................................
6.3 Computation of the factor of use ...............................................................................................32
7 Bibliographie34......................................................................................................................................
8 Description versions of the document34...............................................................................................

Version
Code_Aster default
1 Introduction
the RCC-M [bib1] described the general rules of analysis of the behavior of the materials of level 1 of
the Nuclear power plants. These rules aim at ensuring the components of power plants REP of the
sufficient safety margins with respect to the various types of damage to which they could be exposed
because of the loadings which are applied to them: excessive strain and plastic instability, elastic or
elastoplastic instability, progressive strain under the effect of repeated stresses, fatigue (progressive
cracking), brutal fracture…
In Code_Aster, it is possible to carry out two types of computation:
• computation of the criteria of level 0 and A of the B3200 paragraph in postprocessing of
computations on structures 2D or 3D;
•computation of the criteria of fatigue of the B3600 paragraph in postprocessing of computations
of pipework.
The criteria of B3200 correspond in operator POST_RCCM to results (TYPE_RESU_MECA) of types `

EVOLUTION' and “UNITAIRE”. Their computation is detailed in chapters 3 4 and 5.
The criterion of fatigue of B3600 corresponds in operator POST_RCCM to the results of the type `
PIPEWORK'. Their computation is detailed in chapter 6.

Version
Code_Aster default
2 In data in Code_Aster This

chapter bets aims pointing out some basic definitions associated with the B3200 chapter with the
RCC-M and at describing the adaptations necessary to Code_Aster, which are justified in [bib2]. The
criteria of the §B3200 correspond in operator POST_RCCM to results (TYPE_RESU_MECA) of types `
EVOLUTION' and ` UNITAIRE'.
2.1 Geometrical data

the user of the RCC-M must distinguish in his structure the areas of major discontinuity, the areas of
minor discontinuity and the areas comprising of the geometrical singularities. These last require a
specific processing (described with the §5.45.4).
Current area (except geometrical singularity):

The RCC-M defines “segments of bearing” which are used to linearize the stresses. These segments
are, out of areas of discontinuity, of the generally normal segments on the median surface of the wall,
and in the areas of discontinuity, the shortest segments making it possible to join the 2 sides of the
wall.
The user of Code_Aster must thus define all the sections of structure where computations of
postprocessing will be made (it is him which knows if these sections pass by current areas, or areas of
geometrical discontinuity). In practice, one works:
• maybe on an existing segment in the mesh;
• maybe on a segment provided by INTE_MAIL_2D or INTE_MAIL_3D ;
• maybe on a segment defined in MACR_LIGN_COUPE.
One systematically computes all the criteria at the two ends of the segment.
Geometrical singularity:
Areas of local discontinuities whose geometrical contour present of the abrupt variations are the seat
of acute stress concentrations. In this case, the common methods associated with the current areas
are unsuited and one introduces the concept of factor of priming. This parameter must be computed
on a circle (of imposed radius, depend on the material) around the singularity.
The user must thus define this cut line. In practice, one works:
• maybe on a circular cut line existing in the mesh;
• maybe on a circular cut line defined in MACR_LIGN_COUPE.
It is reminded the meeting that this option is available only for the loadings of the type EVOLUTION.
2.2 Data of loading - Type EVOLUTION

the option ` EVOLUTION' is well adapted to computations on a component subjected to few
situations of loading and not of earthquake. The user of the RCC-M must give the number of
occurrences of each situation of operation (for example: heating of the boiler, hot stop, etc.). A
situation of operation can be broken up into transients, i.e. evolutions of the total parameters of
operation (pressure, temperature) according to time.
In Code_Aster, one treats mechanical results (produced by MECA_STATIQUE or STAT_NON_LINE),
therefore transients. For each transient, the stress fields are with being provided on the segment of
analysis to times of discretization of computation via arrays created by call to POST_RELEVE_T or
MACR_LIGN_COUPE.
Several types of results can be necessary for each transient: stresses for the loadings
thermomechanical (TABL_RESU_MECA), forced for the thermal loading only (TABL_SIGM_THER),
forced for the direct loading of compression (TABL_RESU_PRES) and forced for the singular areas
(TABL_SIGM_THETA).
This option is that which leads to the most precise results. It indeed requires to introduce any
simplifying assumption neither on the definition of the loadings, nor on the computation of the various
criteria of level 0 or of level A.

Version
Code_Aster default
Par ailleurs, this option allows of compute the factor of priming the level of a singular area, within the
meaning of appendix ZD of the RCC-M.
2.3 Data of loading - Type UNITAIRE

the option ` UNITAIRE' is well adapted to computations on a component subjected to many
situations, with possibly of the earthquake. Several groups of situations can also be defined, with
situations of transition between these groups.
Each situation is described by two stabilized mechanical states (A and B, possibly related to seismic
loadings) and a thermal transient.
Each stabilized mechanical state is described starting from a pressure P and of a load vector force
{ }
F x  i  , F y  i  , F z  i  , M x  i  , M y  i  , M z  i  . The tensors of the stresses are reconstituted by linear
combination starting from the tensors of the stresses associated with each unit loading. One notes

 avec ∈{ XX,YY,ZZ,XY,XZ,YZ } the components of the tensor of the stresses associated with
U
the unit loading U . The computation of the tensor of the stresses corresponding to a mechanical
loading pertaining in a stabilized state is then obtained in the following way:
 i= F x i. Fx_ U  F y i . Fy

_U
F z i.  Fz_ U 
M x i . Mx My Mz P
 _ U M y i.   _ U M z i.   _ U  P i.  _ U
The use of this option requires the preliminary computation of the stress fields for the 7 elementary
loadings and of the stress fields for each thermal transient; these fields are with being provided on the
segment of analysis to times of discretization of computation via arrays.
Note:
1) For the bypasses, it is also possible to use method UNITAIRE , by defining two
tensors of forces respectively associated with the body and the pipe.
2) Contrary to type EVOLUTION , it is necessary here to introduce certain
assumptions for the fatigue analysis, cf § 4.2 .
3) It is advisable to warn the user that the choice to separate the mechanical parts
and thermals leads to a desired conservatism which can prove to be
excessive.
2.4 Material characteristics

Les material characteristics necessary to computation are the following ones:
• Sm : acceptable value (tabulée in the RCC-M Annexe Z1).
• Sy : conventional limit of elasticity (tabulée in the RCC-M Annexe Z1 2.1).
• m n : constant material for the computation of Que (defined in the RCC-M B3234.6)
• Ec E : elasticity moduli (for the correction of the curve of fatigue, additional Z1).
• Curves of fatigue of the material: according to the additional RCC-M Z1.
•Distance d to the geometrical singularity and model of priming of the material (as defined and
tabulées in appendix ZD2200 of the RCC-M) for the computation of the factor of priming.
2.5 Simplifying assumptions

Dans the RCC-M, the user must be able to say, after analysis of the results of computation, if the
principal directions in a given point are fixed or if they turn in the course of time.
On the other hand, in command POST_RCCM, one can not make an assumption. One will consider only
the case where the principal directions are unspecified.

Version
Code_Aster default
Moreover, the user must be able to classify the stresses in the following categories:
•General primary education of membrane: Pm
•Primary education of local membrane: Pl
•Primary education of bending: Pb
•Thermal expansion: Pe
•Secondary: Q
•Of point: F
This choice cannot be made by POST_RCCM. Only the user can qualify a stress field (“Primaire”,
“secondary”, or add it of both). The criteria which are to be checked are computed starting from stress
fields (constant or function of time) provided by the user. It is him which ensures coherence between
the computation of these fields and the criteria applied.
However, to fix the ideas, classification is simpler in the following cases:
• a constant or variable loading with force or imposed pressure is primary, except for certain very
particular structures,
• a loading constant or variable with imposed displacement is in theory, secondary (except in the
case of “the effect of spring”),
•a thermal loading permanent or transitory is in theory secondary.
On the other hand, the combination of these types of loadings leads to a result which cannot be
qualified any more of primary education or secondary. According to the criteria, the user will be able to
thus have to break up his loadings.
2.6 Computations carried out by POST_RCCM

One describes here the operation of command POST_RCCM allowing to carry out the computation of
certain criteria RCC-M B3200 of levels 0 and A. the realization described here does not take into
account all the criteria of B3200 and could be supplemented (for example for other levels of criteria).
The main data is the segment (of bearing) where computations will be carried out. It is the user who
chooses the segment and which has the responsibility to find that for which the quantities intervening
in the criteria are maximum. The automatic search for this segment carrying out the maximum is a
difficult problem, and is not programmed.
After having computed one or more results by MECA_STATIQUE or STAT_NON_LINE, the user must
extract the stresses on the segment from analysis by POST_RELEVE_T or MACR_LIGN_COUPE. To
finish, the user asks for the computation of the criteria by operator POST_RCCM.
Three types of criteria are accessible each one by a key word factor:
• criteria of level 0 by key word PM_PB,

• of the criteria of level A (except fatigue) by key word SN,
•of the criteria of fatigue (also of level A) by key words FATIGUE (for ` UNITAIRE') or
FATIGUE_ZH210 (for ` EVOLUTION'),
In addition, for the results of the type ` EVOLUTION' only, it is possible to check the criterion of
priming (criterion of level A) in a singular area (key word AMORCAGE).
3 Criteria of level 0 (key word PM_PB)

3.1 Criteria of level 0 specified by the RCC-M
Les criteria of level 0 aim at securing the material against the damage of excessive strain, plastic
instability and elastic and elastoplastic instability. They must be checked by the situation of reference

Version
Code_Aster default
(see B3121 and B3151). These criteria require the computation of the equivalent stresses
P m , P l , P b which are below defined.
3.1.1 General primary equivalent stress of membrane Pm

Etant given the primary stress of the situation of reference (1st category) and a segment located out of
an area of major discontinuity. In each point end of this segment length l , one computes:
l
1
P m =max∥σ moy
ij ∥ σ moy
∫ σ ds ,
ij =
t l 0 ij
où ∥.∥ désigne la norme de Tresca :∥τ ij∥=max ∣τ I −τ J ∣
I ,J
τI I =1,3 sont les contraintes principales 
The criterion is written (B3233.1):
P m ≤S m
S m is the acceptable equivalent stress, tabulée in the Z1 appendix of the RCC-M. S m is defined in
computation by operand SM of the key word factor RCCM (or RCCM_FO) of DEFI_MATERIAU. It can be
a function of the temperature.
3.1.2 Primary equivalent stress of local membrane Pl

Etant given the primary stress of the situation of reference (1e category) and a segment located in an
area of major discontinuity, the definition of P l is identical to that of P m on this segment.
The criterion is written (B3233.2):

P l ≤1 .5 S m
3.1.3 Primary equivalent stress of membrane+flexion Pmb (or Plb)

Étant given the primary stress of the situation of reference (1st category) and a segment (directed). In
each point end of this segment length l , (ends corresponding to the skins offsite and internal), one
computes:
P m =max ∥σ ijmoy∥ P b =max∥σ ijfle∥ P mb =max∥σ lin

ij ∥
t t t
l l
1
σ ijmoy = ∫ σ ij ds
l 0
σ ijfle =
6
2∫
l 0
l
 
s− σ ij ds
2
σ lin moy fle
ij = σ ij ± σ ij
σ lin moy fle

ij  s=0 = σ ij − σ ij
σ lin moy fle
ij  s=l  = σ ij  σ ij
The criteria are written (B3233.3):

P mb ≤1. 5 S m
P Ib≤1. 5 S m
3.2 Computations carried out in the case ` EVOLUTION'

The array of the stresses comprises either only one time step, or a complete transient (nb_inst time
step). In this last case, one will seek the maximum, compared to the list of the sequence numbers,
different the terms intervening in the criteria.

Version
Code_Aster default
It is to the user to know if one computes Pm (forced general of membrane: out of areas of
geometrical singularity) or Pl (forced local of membrane: in the singularities). From the recordings of
stresses provided, one thus computes a membrane stress.
The algorithm is the following:

• On all the sequence numbers n=1, nb_inst
• extraction of the stresses at time t
• on each end of the segment
•computation of P m  t  P b t  , P mb  t , s=0  and P mb  t , s=l  by integration on the
segment
l
1
σ moy
ij  t = ∫ σ ij  t  ds , P m t =∥σ moy
ij  t ∥
,
l 0
l
6
σ ijfle  t =2∫
l 0
l
 
s− σ ij  t  ds , P b  t =∥σ ijfle  t ∥
2
P mb  t , s=0 =max∥σ moy fle
ij  t −σ ij t ∥ P mb  t , s=l =max∥σ moy fle
ij t σ ij  t ∥
t t
• Recherche of the maximum of P m  t  , P mb  t , s=0  and P mb  t , s=l 
• Output and storage in the array of the result.
The limiting values are respectively S m and 1,5 S m , S m being working stress function of the
material and the temperature, given by key word SM_KE_RCCM of behavior FATIGUE in
DEFI_MATERIAU.
Note:
The thermal stresses are of secondary type and do not have to thus be taken into account in
the computation of the criteria of level 0. In POST_RCCM , if TABL_RESU_MECA and
TABL_SIGM_THER are present simultaneously, it is supposed that result TABL_RESU_MECA
corresponds to the thermomechanical complete loading, and one thus cuts off the stresses to
him of thermal origin.
3.3 Computations carried out in the case ` UNITAIRE'

While noting A and B the two stabilized mechanical states, the computation of P m is done in the
following way:
P m =max {∥σ moy
ij ∥ ,∥σ ij ∥ }
A moy B
and the same for P b and P mb .

Version
Code_Aster default
4 Criteria of level A for the results of the type ` UNITAIRE'
4.1 Definition of the criteria in the RCC-M

Les criteria of level A aim at securing the material against the damage of progressive strain and
progressive cracking. For the results of the type ` UNITAIRE', four type of criteria can be checked:
1. Amplitude of variation of the linearized stresses S n (option ` SN');

2. Amplitude of variation of S *n (option ` SN');
3. Computation of the thermal ratchet (option ` SN');

4. Computation of the factor of use in fatigue (option ` FATIGUE').
These various parameters and the associated criteria are described below such as defined in the
RCC-M. In the next paragraph, one introduces a simplifying assumption before detailing their
computation in Code_Aster.
4.1.1 Sn computation
One takes into account the more secondary primary stresses and the stresses resulting from opposed
thermal dilations: P l P b P E Q who thus represents the linearized stresses associated with all
the loading (mechanical and thermal).
The points of computation are the two ends of the segment. In each point end of this segment length
l , one computes S n according to the B3232.6 paragraph:
 max ∥ σ lin

lin
S n= max ij  t 1  − σ ij  t 2  ∥
t1 t2
l l
σ ijmoy =
1
∫ σ ds
l 0 ij
σ ijfle =
6
∫
l2 0  s−
l
2  σ ij ds σ lin moy fle
ij = σ ij ± σ ij
σ ijlin  s= 0  = σ moy fle

ij − σ ij
σ lin moy
ij  s= l  = σ ij  σ ij
fle
The criterion of total adaptation is written (B3234.2):

S n ≤3 S m
S m being the working stress function of the material and temperature, given by operand SM of the
key word factor RCCM (or RCCM_FO) of DEFI_MATERIAU.
If this criterion is not checked, one can practise the simplified elastoplastic analysis of B3234.3. The
three following operations should be carried out:
• check the criterion:
S *n ≤3 S m
• make an elastoplastic correction ( Ke1 ) in the analysis with fatigue,
• check the criterion of the thermal ratchet (B3234.8) in the current parts of the cylindrical shells
(and pipes) subjected to a pressure and a gradient of cyclic temperature.

Version
Code_Aster default
4.1.2 Computation of Sn*

*
One notes S n the amplitude S n computed without taking into account stresses bending of origin
thermal. One translates this definition by:
- At the origin of the segment:
 max ∥  σ lin  t 1   −  σ linij  t 2   σ ijfleth  t 2  ∥

fleth
S *n= max ij  t 1   σ ij
t1 t2
- At the end of the segment:
 max ∥  σ lin  t 1   −  σ linij  t 2  − σ ijfleth  t 2  ∥

fleth
S *n= max ij  t 1  − σ ij
t1 t2
l
with
l
6
σ ijfleth= 2 ∫
0
 s−
l
2  σ th
ij ds
.
σ th
ij only comes from a computation carried out with the thermal loading (i.e. that one removes from
complete computation, having led to the value of S n , all loadings other than the thermal loading).
4.1.3 Computation of the thermal ratchet

the wall of an aircraft subjected simultaneously to a constant pressure and cyclic variations of
temperature can undergo large deformation under thermal ratchet. It is about a particular mechanism
of progressive strain in which the strain roughly increases same quantity with each cycle.
The condition to respect is written below and relates to the acceptable maximum value of the
amplitude of variation of the stress of thermal origin, cf B3234.8. It refers to the case of a shell with
symmetry of revolution loaded by a constant internal pressure. One notes:
σ θ , acceptable maximum value of the amplitude of variation of the stress of thermal origin, in
the same way calculated that Sp
σ m , maximum of the average (or) membrane stress general due to the pressure
S y , conventional limit of elasticity read on tables Z I 2.1, for the maximum temperature reached
during the cycle.
σθ σm
The criterion is form: σ θ= f  σ m , S y  . While posing y '= and x= , one has
Sy Sy
- if the temperature variation is linear through the wall:
y ’ =1/ x for 0x≤0.5
y ’ =41− x 0.5x1
- if the temperature variation is parabolic in the wall:
y ’ =5,21−x  for 0.615≤x1
and for x0,615 : linear interpolation enters the points: x=0,3 ; 0,4 ;0,5 and
y ’ =4,65 ; 3,55 ; 2,7 .
4.1.4 Computation of the factor of general

4.1.4.1 Principe use
the general principle of computation to fatigue consists in combining each situation 2 to 2 and making
sure that the factor of total use thus defined is lower than 1.

Version
Code_Aster default
In a schematic way, the calculation algorithm defined in the B3200 paragraph of the RCC-M is the
following:
1) Computation of the elementary factor of use of each combination of situation:
a) the combination between two situations p and Q rests on the definition of two
fictitious transients:
Transient 1 between the maximum (over time t ) of the maximum of
 σ p  t  , σ q  t   and the minimum of the minimum of
 σ p t  , σ q  t  ;
Transient 2 between the minimum of the maximum and the maximum of
minimum.b)
Computation of the elementary factor of use associated with each of the two
fictitious transients using the fatigue model (cf paragraph below):
u  p , q =u Transitoire 1u Transitoire 2
c) À la fin de this stage, one thus has a matrix n×n (where n is the number of
situations) factors of Identification
2) U tot =0
3) use of the combination k , l  more penalizing (maximum factor of use):
tot tot
U =U min n k , n l .u  k , l  , where n k and n l are the numbers
of occurrences of the situations k and l
4) Réactualisation amongst occurrences of the situations k and l :
n k =n k −min  n k , nl  and n l =n l −min  n k , n l 
5) Retour at stage 3 until exhaustion of all the occurrences.
The definition of the two fictitious transients constitutes a delicate stage of this algorithm. The rule is
different according to whether the principal directions are fixed or variable.
- Case of the fixed principal stresses: according to the B3234.5 paragraph and [6], the fictitious
transients for the combination of the situations p and q are below defined, by noting σ i 1≤i≤3 the
principal stresses and S ij=σ i −σ j their difference.
Fictitious transient 1:
[ S ijmax  1 ] pq= Max  S ijmax  p  ; S ijmax  q 
[ S ijmin  1 ] pq =Min  S ijmin  p  ; S min
ij q  
Fictitious transient 2:
[ S ijmax  2  ]pq =Max  Min  S max
ij  p  ; S ij  q   ,Max  S ij  p  ; S ij  q   
max min min
[ S ijmin  2 ]pq =Min  Min  S max

ij  p  ; S ij q   ,Max  S ij  p  ; S ij  q   
max min min
- Case of the variable principal stresses: the B3234.5 paragraph indicates that one “overrides S ij
by S ’ ij , by taking care to preserve at each principal direction which turns its identity”. The
computation of S ’ ij is detailed in § B 3232.6:
 S 'ij t , k = σ 'i t , k − σ 'j t , k with  σ '  t,k =σ  t  −σ  t k 
time t k corresponds to a extremum of the situation considered. It is necessary to maximize the
quantity  S ij t , k on all the extrema of the situation.
'
4.1.4.2 Computation of the elementary factors of use

Version
Code_Aster default
the factor of use elementary for a combination of situations  p , q is computed by introducing into
the curve of fatigue of the material (curved of Wöhler) the amplitude of variation of the alternate
stresses S ' alt  p , q  .
This amplitude is defined starting from the amplitude of variation of the linearized and Sn p ,q
total stresses S p  p ,q  (even definition that S n  p , q  , but with the stresses in the beginning or the
extremity).
Two formulas are proposed (cf §B3234.6):
• KE_MECA :
1 E
S ' alt  p , q = . c . K e  S n  p , q   . S p  p ,q 
2 E
with:
- E c : Young modulus of reference for the construction of the curve of Wöhler;
- K e the elastoplastic concentration factor:
∣
1 si S n  p , q  ≤3 . S m
1−n
K e  S n  p , q  = 1 n.  m−1  . 3. S
S n  p , q
m
−1
 si 3 . S m S n  p , q 3. m . S m
1
si S n  p , q  ≥3 . m . S m
n
where m and n are coefficients dependent on the material.
• KE_MIXTE : since the modifying 1997 of the RCC-M [1], one can choose another formula,
based on a decomposition of S ' alt :
1 E
2 E e 
S ' alt  p , q = . c . K meca  S n  p , q   . S meca
p  p , q  K ther
e
 S n  p , q   . S ther
p  p ,q 
with:
- K meca  S n  p , q   is equal to K e definite above,

e
∣  
1
1,86 1−
1,66S n
- K ther
e  S n  p , q   =max ,
Sm
1
meca
- S p  p , q  represents the quantity S p , amplitude of variation on mechanical behalf
of the stresses computed on the basis of stress of mechanical origin: pressure, inertia loading,
earthquake (inertial and displacements of anchorage), thermal expansion.
ther
- S p  p , q  the quantity computed S p starting from the mechanical stresses only
generated by the thermal transients represents.

Version
Code_Aster default
4.2 Simplifying assumption in the case ` UNITAIRE'

In the method ` UNITAIRE' developed in Code_Aster, the situations are defined in a way simplified
by two stabilized states and a thermal transient. It is then not possible to work directly on each time of
the situations and of the assumptions must be introduced, in particular to identify the two fictitious
transients which couple two situations.
One thus will suppose that times which correspond to the extrema of the amplitude of variation
of the stresses between two situations are also times corresponding to the extrema of the
thermal stresses of each situation.
In other words:
t1 , t 2

max∥σ ther , p  t 1  −σ ther , q  t 2 ∥≈max ∥σ ther , p  t max min min max

p −σ ther ,q  t q ∥ ,∥σ ther , p  t p  −σ ther ,q  t q ∥
max min
where time t i (resp. t i ) corresponds to time when the thermal transient of situation I
reached its maximum (resp. minimum).
This assumption is checked when the principal directions are fixed. In the industrial cases which
concern us, the principal directions of the thermal loading vary a priori little in the course of time. This
assumption thus seems licit.
Times t max
p
min
and t p are thus identified beforehand for all the situations p : they correspond to the
extrema of the thermal stress associated with p , within the meaning of an equivalent stress of
Tresca signed by the trace of the stresses:
t min
p = Arg min ∥σ ther , p  t ∥. sgn Tr  σ ther , p t    
t max
p = Arg max ∥σ ther , p  t ∥.sgn  Tr  σ ther , p  t    
Note:
max min
The identification of times t p and t p is made independently for the two ends of the
segment of analysis. One does not differentiate here the two couples from extreme times not to
weigh down the presentation of the algorithms.
4.3 Computations carried out with the option ` SN'

One notes  1 p and  2 p the mechanical stresses associated with the two states stabilized with the
situation p ;  ther , p t  the thermal transient associated with this situation; and t max
p t min
p
extreme times of this thermal transient such as definite above.
The parameter S n for the situation p is then computed as follows:

S n =max ∥ lin lin lin max lin min lin lin lin max lin min
1 p − 2 p  ther , p  t p − ther , p  t p ∥,∥ 2 p − 1 p  ther , p  t p − ther , p  t p ∥ 

Version
Code_Aster default
*
If a thermal transient is defined (i.e key word NUME_RESU_THER is indicated), the computation of S n
for the situation p is done in the same way:

S *n =max ∥σ 1lin −σ lin
p
moy max moy min lin lin moy max moy min
2 p σ ther , p t p −σ ther , p  t p ∥,∥σ 2 p −σ 1 p σ ther , p t p −σ ther , p  t p ∥ 
computation is carried out if operand NUME_RESU_THER is present in the case ` UNITAIRE'. It is also
necessary beforehand to have defined the conventional limit of elasticity for the maximum
temperature reached during the cycle is by operand SY_MAX of POST_RCCM ; maybe by operand SY
02 of key word RCCM in DEFI_MATERIAU [U4.43.01]. If no elastic limit is defined, the computation of
the thermal ratchet is impossible. The stresses due to the loading of pressure are extracted from the
array provided under TABL_PRES.
In table result appear, for each end of each segment of analysis, elastic limit SY, the amplitude of
variation of the stress of origin thermal SP_THER, the maximum of general membrane stress due to
pressure SIGM_M_PRES and two acceptable values maximum of the amplitude of variation of the
thermal stress calculated either by supposing a linear variation in temperature in wall
(VALE_MAXI_LINE), or by supposing a parabolic variation in temperature in wall
(VALE_MAXI_PARAB).
4.4 Computations carried out with the option ` FATIGUE'

The computation of the factor of use requires as a preliminary the computation of the amplitude of
variation of the stresses for each combination of situations. This computation is carried out
successively for the situations inside each group with or without earthquake, then for the combinations
of situations of transition between groups of situations.
The amplitudes of variation of the stresses make it possible to identify the elementary factors of use
associated with all the combinations. One uses then a method of office plurality of the elementary
factors of use, based on the assumption of the linear office plurality of the damage, to obtain the factor
of total use.
4.4.1 Combination of the situations inside each group of Calcul

4.4.1.1 situations of Sp (p, Q)
One initially considers the combination between the states i p and j q . These two states can be combined in
two manners, characterized by S pi p , j q ,t max min
p ,tq 
max min
and S p  i p , j q ,t q , t p  :
{∥σ i σ ther , p  t max min

p −σ j −σ ther ,q  t q ∥
S pi p , j q ,t max min
p , t q =max
p
max
q
∥σ j σ ther , p  t p −σ i −σ ther ,q  t min

q ∥
q p
{∥σ i σ ther , p  t max min

q −σ j −σ ther ,q  t p ∥
S p  i p , j q ,t qmax , t min
p =max
p q
∥σ j q σ ther , p  t max min

q −σ i p −σ ther ,q  t p ∥
One can then define the two fictitious transients which combine the states i p and j q :
S 1p i p , j q =max  S p i p , j q , t max
q p  , S p  i p , jq , t p , t q 
, t min max min
S 2p i p , j q =min  S p i p , j q , t max
q p  , S p  i p , j q , t p , t q 
, t min max min
S 1p And it S 2p of the combination of situations  p , q are then defined by maximization on the four possible
combinations of stabilized states:
S 1p  p , q =max S 1p i p , j q  and S 2p  p , q =max S 2p i p , j q 
i p , jq i p , jq

Version
Code_Aster default
the case however also should be considered where the combination more penalizing thermal stresses
corresponds to both extrema of the same situation (in other words the maximum of the maximum and the
1 2
minimum of the minimum are in the same situation). One thus modifies the definition of S p and S p as
follows:
1 1 2
If S p  p , q S p  p , p  , then S p  p , q =S p  p , p  and S p  p , q =S p  q , q  ;
If S 1p  p , q S p q , q  , then S 1p  p , q =S p  q , q  and S 2p  p , q =S p  p , p  .
Note:
1) In the typical case where the two situations are in pure mechanics, the
1 2
preceding definition would result in taking S p and S p equal, which would be
2
unnecessarily conservative. In this case, one definite thus S p as agent with
the second combination of stabilized states more penalizing. In the same way,
the formulation is adapted if one of the situations does not have a thermal
transient.
2) There would be other manners of defining the two fictitious transients, but the
method presented here seems most general and remains always conservative.
Thus, if the combination more penalizing thermal stresses corresponds to both
extrema of the same situation, then computed S p it corresponds directly to the
definition of the RCC-M. In the same way if the mechanical loading is negligible.
The level of conservatism of this formulation increases with the weight of the
mechanical loading compared to the thermal loading.
4.4.1.2 Sn computation (p, Q)

the amplitude S n associated with the combination with situations  p , q is computed like the
parameter S 1p above, that is to say:

S n  p , q =max S n  p , p  , S n  q ,q  , max S n  i p , j q 
i p , jq 
4.4.1.3 Computation of Spméca (p, Q) and Spther (p, Q)
Dans le cas où one uses method KE_MIXTE, it is necessary to break up the amplitude of variation of
the stresses into a mechanical part and a thermal part. For the definition of S méca
p
ther
(resp. of S p ),
the RCC-M (§B3234.6) leaves freedom between (cf §4.1.4.2):
- take the mechanical share (resp. thermal) of the amplitude of the maximum stresses between the
two transients;
- take the maximum value of the amplitude of the mechanical stresses (resp. thermals) during these
transients.
It is this last method, more conservative but simpler to implement, which was retained. For the
mechanical share, there are six combinations of possible stabilized states:
∥σ i −σ j q∥ for i p , j q ∈{1,2} ; ∥σ 2 −σ 1 p∥ ; ∥σ 2 −σ 1q∥ .
p p q
1 méca
The fictitious mechanical transient 1, characterized by S p , corresponds to the combination more
penalizing. If this combination utilizes two stabilized states of the same situation, then the fictitious
mechanical transient 2 corresponds to the combination of the two stabilized states of the other
2 méca
situation; if not corresponds S p it to the second combination more penalizing.
1 ther 1 méca
The thermal fictitious transients S p and S p are defined same manner, by combining times
corresponding to the extrema of the thermal transients.
4.4.1.4 Computation of alt (p, Q)

Version
Code_Aster default
the computation of S ' alt  p , q  was carried out, for each couple of situations starting from the
1 2
amplitude of variation of the stresses of the two fictitious transients S p  p ,q  and S p  p ,q  , and
the amplitude of variation of the linearized stresses S n  p , q  .
Case KE_MECA :
1 E
S ' 1alt  p , q = . c . K e  S n  p , q   . S 1p  p ,q 
2 E
1 E
S ' 2alt  p , q = . c . K e  S n  p , q   . S 2p  p ,q 
2 E
The Young modulus of reference ( E c ) is provided by the user in DEFI_MATERIAU, under key word
E_REFE, of the key word factor FATIGUE. The parameters m and n are provided in
DEFI_MATERIAU, under key words M_KE and N_KE, of the key word factor RCCM. If key words
TEMP_REF_A and TEMP_REF_B are present, S n is interpolated for this temperature (which must
correspond to the average temperature of the transient). If not, S n is taken with room temperature.
Case KE_MIXTE :
1 E
2 E e 
S ' 1alt  p , q = . c . K meca  S n  p , q   . S 1pmeca  p , q  K ther
e
 S n  p , q   . S 1pther  p , q  
1 E
2 E e 
S ' 2alt  p , q = . c . K meca  S n  p , q   . S 2pmeca  p , q  K ther
e
 S n  p ,q   . S 2pther  p , q  
One computes finally FU  p ,q  the factor of use elementary associated with the combination of the
situations p and Q, defined starting from the curve of fatigue of the material N adm = f  S alt  :
FU  p , q = f  S ' 1alt  p , q   f  S ' 2alt  p , q  
4.4.2 Taken into account of the earthquake

the second phase consists with compute the amplitudes of stresses which correspond to the
combinations of all the stabilized states pertaining to the situations of a given group, of taking of
account the seismic loadings.
S S
The seismic loadings are not signed. Each component of the tensor of the stresses ( M X M Y ,…)
can thus take two values (positive and negative). At the time of the superposition of a loading not
signed with a signed loading, the RCC-M forces to retain on each component a sign such as the
computed stress (in fact S ' alt ) is raised. The tensor of the stresses cash six components, there thus
exist 64 combinations of signs to be examined.
1 2
The parameters S p  p ,q , S  S p  p ,q , S  , S n  p , q , S  and S ' alt  p , q , S  with earthquake are
computed same manner as without earthquake, but by maximizing the amplitude of the stresses
compared to the sign of each component of the tensor of the stresses. For example:
 
S p  p , p , S =max ∥± σ 1 −σ 2 σ ther , p  t max
p p
p  −σ ther , p  t p  2 ∑ ±M X σ M ±M Y σ M . ..∥

min S S
x Y 
Maximization relating here to the combination of the signs of the components of the tensor of the
stresses of the situation of earthquake. σ M X indicate the tensor of the stresses associated with a unit
loading according to M X . One notes FU  p ,q , S  the factor of use elementary associated with
the combination of the situations p and q . One also computes the factor of use associated with the
under-cycles of earthquake FU  S  , characterized by S p  S  :

Version
Code_Aster default

S p  S =max ∥2 ∑ ±M SX σ M ± M YS σ M .. .∥
X Y 
Note:
Factor 2 on the seismic component of the preceding statement appears, because the quantity
computed here is an amplitude .
4.4.3 Situations of transition between groups of Pour

situations each situation of transition of a group with another, one considers all the combinations
 p , q with p pertaining to the first group (of dimension N ) and q pertaining to the second group
(of dimension M ). For each combination, FU  p , q  is obtained in the same way that previously.
One still builds a matrix (rectangular) containing all them FU  p , q  , with which one associates the
number of occurrences of the situation of transition.
Indeed, two situations are combinable only if they belong to the same group or if there exists a
situation of transition between the groups to which they belong. In this last case, one will associate
with the combination  p , q the number of occurrences of the situation of transition. If the situation
of transition belongs to the one of the two groups (what is not excluded a priori), it is naturally
combined with the other situations of this group, then is used for the combination of the situations of
its group with the situations of the group in relation.
4.4.4 Pour
storage to carry out the computation of the factor of total use, the elementary factors of use previously
computed and the associated numbers of occurrences are stored in a square matrix containing all the
elementary factors of use FU except earthquake, for all the possible combinations of situations
(inside each group of situations, and between two groups if there exists a situation of transition). The
matrix has as a dimension the sum amongst situations of all the groups:
FU Groupe 1 Group 2 Groupe 3
Situ … Situ J … … … Situ L … … . Situ N Situ
1 N
Groups 1 Situ 1 FU … … … 0 0 0 0 … … … …
Situ I FU (I, … 0 0 0 0 … … FU (I, …
J) N)
… … 0 0 0 0 … … … …
… … 0 0 0 0 … … … …
Groupe 2 …. … … … … 0 0 0 0
Situ K … FU (K, … 0 0 0 0
L)
… … … 0 0 0 0
…. SYM … 0 0 0 0
Group 3 …. … … … …
Situ m … FU (m, …
N)
… … …
Situ N …
Dans the table above, one associates value 0 with the combinations of states between groups 1 and 2
and groups 1 and 3, because there does not exist situation of transition between these groups. On the
other hand there are some between groups 1 and 3, one thus associates with each combination of
states of groups 1 and 3 the value of FU .
The number of the states of loading will have to be built starting from the number of situation (single)
and the relative number (1 or 2) of the stabilized state of the situation, to obtain a univocal dialup of
the states of loading.
In the case of the under-cycles, only the diagonal term is filled.
This matrix is also to build for the values of FU  i , j , S  which take into account of earthquake.
There are thus two matrixes giving all the possible values of FU .

Version
Code_Aster default
In correspondence of this matrix, one can associate a table giving the number of occurrences, the
number of situation and the number of group of each stabilized state, and a table giving the number of
occurrences of the situations of transition:
Situati groups Under- nocc group 1 groups Npass

on cycle 2
1 1 0 N1 1 1 0
2 1 0 Nor 1 2 0
3 2 0 Nk 1 3 npass
… … … … 2 1 0
… 3 nsous 0 2 2 0
… … … … 2 3 0
N 3 0 N … … …
4.4.5 Calcul of the factor of total use

À l'issue of the preceding stages, one thus has:
• the matrix [ N , N ] of the factors of fascinating FU  p ,q , S  use of account the seismic
loading;
• the matrix [ N , N ] of the factors of use FU  p ,q  without earthquake.
One notes:
n p the number of cycles associated with the situation p
n q the number of cycles associated with the situation q
N s the number of occurrences of the earthquake (in general only SNA is considered in second
category)
n s many under-cycles associated with each earthquake.
n pass many cycles associated with a possible situation with transition between p and Q if these
situations do not belong to the same group, but if there exists a situation of transition
between the two.
If N S 0 , one selects N S /2 the combinations more penalizing, i.e. N S /2 the combinations

k , l  leading to the greatest values of FU k ,l  , without taking into account the
earthquake.
For each one of these N S /2 combinations:
• One computes the factor of use u 1 k , l associated with S ' alt  k , l ,T , S  :

u 1  k , l  =FU  k , l , S  .
• One takes into account n s the under-cycles of the earthquake by:
u 2  k , l  =  2 . n s −1 ∗FU  S  .
• One obtains then: u k , l =u1 k , lu2 k , l
• One starts again this computation until exhaustion of N s /2 the combinations more penalizing.
The computation of the factor of use is then continued without taking into account the earthquake:
If N s=0 , or after having taken into account the earthquake for N S /2 the most unfavourable
combinations:
• One selects the combination  p , q leading to the maximum value of FU  p , q  , on all the
combinations such as the number of occurrences n 0 is non-zero, by noting
n 0 =min {n p , n q , npass } if n pass is non-zero, or n 0 =min {n p , n q } if not.
• One increments the factor of total use: U = U + n 0∗FU  p , q 

Version
Code_Aster default
• One replaces:
n p by n p−n0 
nq n q−n0 
if it is about a situation of transition, n pass by n pass−n0 
if n k =0 , the column and the line k of the matrix of S ' alt i , j are settings with 0.
if n l=0 , the column and the line L of the matrix of S ' alt i , j are settings with 0.
•The loop is repeated until complete exhaustion amongst cycles.
Note:
Appendix ZI of code RCC-M defines the curve of Wöhler until an amplitude of stress
minimum corresponding to one life duration of 106 cycles. If the computed value S ' alt
for a combination i , j of stabilized state is lower than this amplitude minimum, the
factor of use is equal to 0 for the combination i , j considered.
Cases of the under-cycles

Les under-cycles correspond either to the taking into account of the under-cycles related to the
earthquake, or with situations for which key word COMBINABLE='NON' was well informed. In both
case, one computes the amplitude of stresses while utilizing only the stresses related to these under-
cycles (not of combination of states of loading apart from this situation). For the computation of S ' alt ,
it is necessary to use the factor K e which corresponds to the main situation from which the under-
cycle is resulting.

Version
Code_Aster default
5 Criteria of level A for the results of the type ` EVOLUTION'
5.1 Definition of the criteria

Les criteria of level A aim at securing the material against the damage of progressive strain and
progressive cracking. For the results of the type ` EVOLUTION', four types of criteria can be
checked:
1. Amplitude of variation of the linearized stresses S n (option ` SN');
2. Amplitude of variation of S *n (option ` SN');
3. Computation of the thermal ratchet (option ` SN');

4. Computation of the factor of use in fatigue (option ` FATIGUE_ZH210'); (without
earthquake)
5. Computation of the factor of priming for the singular areas (option 'AMORCAGE').
These various parameters and the associated criteria are below defined.
the amplitude of variation of the linearized stresses S n and the associated criterion are defined in the
§4.1.14.1.1.

the amplitude of variation of the stresses linearized without taking into account bending stresses and
the associated criterion are defined in the §4.1.24.1.2.

Les requirements relating to the thermal ratchet are defined in the §4.1.34.1.3.
5.1.4 Computation of the factor of use according to appendix ZH210

Les requirements relating to computation of the factor of use are defined in the §4.1.44.1.4.
Selected method for the results of the type ` ` EVOLUTION'' corresponds to appendix ZH210 of the
RCC-M. It consists in “forgetting” the concept of situation and to combine states of loadings directly,
who are significant times of all the transients where the stresses pass by a local extremum.
The main advantage of this method is to consider all the possible under-cycles automatically: it is not
necessary to identify the fictitious transients combining the situations between them. Its disadvantage
is the number of computations to be carried out if one does not restrict all the times used in
computation.
5.1.5 Computation of the factor of priming

Les areas of local discontinuities whose contour present of the abrupt variations are the seat of acute
stress concentrations. In this case, the concept of factor of use defined in the preceding chapter is not
adapted any more and it should be overridden by the concept of factor of priming (B3234.7).
The factor of priming is computed starting from the amplitude of variation of the stress in structure at a
distance d of the singularity, and a model of priming. The procedure of analysis is defined in
appendix ZD2200. The distance d and the models from priming are characteristics material and are
tabulées in table ZD2300.

Version
Code_Aster default
Appear 5.1.5-a : definition of the local coordinate system
the model of priming defined in the RCC-M is form:

     d =A_AMORC . N a B_AMORC
with N a the number of acceptable cycles and     the amplitude of variation of the shear
stresses, in the local coordinate system, with the distance d from the singularity.
The model of priming developed in operator POST_RCCM takes into account the report of load R of
the loading, as recommended in RSE-M (ref. 7):
B_AMORC
  eff  d =A_AMORC . N a 
with the following relation between amplitude of variation of the stresses real     and effective
  eff :
  
  eff = .
R
1−
R_AMORC
The parameters of the model of priming (A_AMORC, B_AMORC, R_AMORC) and distance it to singularity
D_AMORC are to be defined under the key word factor RCCM of DEFI_MATERIAU.
Note:
To use a model of priming such as defined in the RCC-M, that is to say without taking into account
of the report of load, it is enough to define a R_AMORC large (1000 for example).
5.2 Computations carried out with the option ` SN'

One notes nb_inst the number of times selected in the transient considered.
The calculation algorithm of S n is the following:
• on the group of the sequence numbers, n 1=1, nb_inst

• Extraction of time t 1

Version
Code_Aster default
lin
• Computation of σ lin
ij t 1 , s=0  and  ij  t 1 , s=l 
• Pour n 2 varying from n 11 with nb_inst
• formula of time t2
lin
• computation of σ ij t 2 , s=0  and σ lin
ij t 2 , s=l  of
σ ij t 2 , s=0 −σ ij t 1 , s=0 and σ lin
lin lin lin
ij t 2 , s=l −σ ij t 1 , s=l 
• computation of the principal directions and the criterion of Tresca:
 σ linij  t 2 , s=0  −σ linij  t 1 , s=0  Eq . Tresca and  σ linij  t 2 , s=l −σ linij  t 1 , s=l   Eq . Tresca
• thus seeks maximum of S n with each Sortie
• end and storage in the array the result.
Note:
The quantity S n computed here corresponds to an amplitude. It is thus essential that all the
system statuses are considered, including the states with null stress (for example cold stop: null
pressure and applied moments and room temperature).

This computation is carried out if operand TABL_SIGM_THER is present. Only the user ensures the
coherence of the data, i.e. this result must be produced by a thermomechanical computation under
thermal loading only, knowing that the result given by TABL_RESU_MECA can be due to a
combination of this thermal loading with other loadings. It is necessary thus in particular that times of
arrays TABL_RESU_MECA and TABL_SIGM_THER correspond.
The algorithm is identical to the precedent but relates to two stress fields.
computation is carried out if operands TABL_SIGM_THER and TABL_RESU_PRES are present. It is

also necessary beforehand to have defined the conventional limit of elasticity for the maximum
temperature reached during the cycle is by operand SY_MAX of POST_RCCM ; maybe by operand SY
02 of key word RCCM in DEFI_MATERIAU [U4.43.01]. If no elastic limit is defined, the computation of
the thermal ratchet is impossible.
In table result appear, for each end of each segment of analysis, elastic limit SY, the amplitude of
variation of the stress of origin thermal SP_THER, the maximum of general membrane stress due to
pressure SIGM_M_PRES and two acceptable values maximum of the amplitude of variation of the
thermal stress calculated either by supposing a linear variation in temperature in wall
(VALE_MAXI_LINE), or by supposing a parabolic variation in temperature in wall
(VALE_MAXI_PARAB).
5.3 Computations with fatigue with the option ` FATIGUE_ZH210'

The principle of method ZH210 is to define, for each transient, a set “of states of loading”, which are
significant times where the stresses pass by a local extremum. By defaults, in Code_Aster, all times of
computation are used. One associates with each one of them the number of occurrences Nocc of
the transient. The definition is thus:
State of loading = {urgent, stress tensor, many occurrences}.
Then, one builds the set of all the states of loading by sweeping all the transients. At the end of the
day, the concept of transient is forgotten: one does not work any more but on a set of states of

Version
Code_Aster default
loading. One computes then the elementary factors of use associated with all the combinations taken
two to two. One uses then a method of office plurality of the elementary factors of use, based on the
assumption of the linear office plurality of the damage, to obtain the factor of total use.
Note:
The algorithm describes here is similar to that of POST_FATIGUE . More precisely, the algorithm
used in POST_FATIGUE is a restriction on the uniaxial case of method ZH210. Indeed, the data
of command POST-FATIGUE is a scalar function of time, whereas POST_RCCM treats stress
tensors functions of time.
5.3.1 Computation of the elementary factors of use

A each end of the segment, for any couple of states of loading k and l , one calculates the quantities
S p  k ,l  and S n  k , l  definite by:

S p  k ,l  =  σ ij  k  −σ ij  l   Eq .Tresca  
S n  k ,l =  σ lin
ij  k  −σ ij  l   Eq . Tresca
lin

For the computation of S alt  k , l  , two formulas are proposed (cf §B3234.6):
•the original method (KE_MECA ) which does not make distinction between the mechanical share and the
thermal share :
1
S alt  k , l  = E  S n  k , l   S p  k ,l 
2
where:
∣
1 si S n ≤ 3. S m
1− n
K e  S n = 1 n .  m− 1  .  Sn
3.Sm
−1
 si 3 . S m S n 3 . m . S m
1
si S n ≥ 3. m . S m
n
•method KE_MIXTE introduced into the modifying 1997 of the RCC-M [1] which is based on a
decomposition of S alt between the mechanical share and the thermal share :
1 E
S alt  k , l  = . c . K meca  S n  k ,l   . S meca
 p  k , l  K ther  S n  k , l   . S ther
p k ,l 
2 E e e
with:
- K meca  S n  k , l   equal to K e is defined above,

e
∣  
1
1,86 1−
1,66S n
- K ther
e  S n  k , l   =max ,
Sm
1
meca
- S p  k , l  represents the quantity S p , amplitude of variation on mechanical behalf
of the stresses computed on the basis of stress of mechanical origin: pressure, inertia loading,
earthquake (inertial and displacements of anchorage), thermal expansion.
-
S ther
p  k ,l 
represents the quantity S p computed starting from the mechanical
stresses only generated by the thermal transients.

Version
Code_Aster default
The Young modulus of reference of the curve of fatigue E c corresponds to operand E_REFE of the
key word factor FATIGUE in DEFI_MATERIAU. S m , m and n parameters material which can
depend on the temperature, are provided by the user in DEFI_MATERIAU, under key words SM, M_KE
and N_KE, of the key word factor RCCM (or RCCM_FO).
 
The curve of fatigue N adm = f S alt is a function defined by DEFI_FONCTION, and introduced into
DEFI_MATERIAU by key word WOHLER of the key word factor FATIGUE. It allows of compute the
acceptable number of cycles N adm  k , l  associated with S alt  k , l  , then the factor of elementary
use:
1
u  k , l = .
N adm  k , l 
This computation is carried out for each combination of two states of loading. One thus obtains
(always for each end of the segment) a symmetric matrix u  k , l  , of order the number of states of
loading N tot .
5.3.2 Computation of the factor of total use

the calculation algorithm of the factor of total use, for each end of the cut line, is the following:
1) u tot =0
2) u max =0
3) Loops i=1… N tot (search of the maximum in the table)
If Nocc i 0 :
Loops j=i1 … N tot
If Nocc  j 0 and u  i , j u max :
u max = u i , j  , m=i , n= j
4) Nocc  m , n =min  Nocc  m  , Nocc  n  
5) u tot =u tot  Nocc  m , n ∗u  m ,n 
6) Reactualization amongst occurrences:
Nocc  m =Nocc  m −Nocc  m ,n 
Nocc n = Nocc n −Nocc  m ,n 
7) Return to the beginning of the procedure until elimination of all the Remarques
occurrences:
• If the number of times defined for each transient is large, the CPU time can be prohibitory. It
is thus necessary to be able to restrict it. It is what is made in POST_FATIGUE , by a tri
preliminary of times. One eliminates times such as the scalar function is linear to keep only the
ends of the line segments. One eliminates also the very small variations. Here, in multiaxial
situation, the sorting is more delicate. The concept of stresses proportional could be used, but
in practice the user can define itself the list of times (key word NUME_ORDRE ).
• By this method, one is sure not to forget no under-cycle. On the other hand, it is desirable to
eliminate times which do not correspond to local extrema, because they could generate
factitious under-cycles, increasing the factor of use (these times are only used for the
numerical discretization of the mechanical or thermal problem).
•With the option ` FATIGUE_ZH210' , the combinations of transients are taken into account
*
in the computation of S n and of S n .

Version
Code_Aster default
5.3.3 Example
This paragraph aims at illustrating the calculation algorithm of the factor of use on a simple example,
drawn from the case test of elementary validation rccm01a [V1.01.107]. It is supposed that there are
three situations of two time steps each one, the number of occurrences being respectively of 1.5 and
10.
The matrix of the factors of use as computed by the first part of the algorithm is given below. To
reduce the presentation, except the upper part of the symmetric matrix is written.
J 1 2 3 4 5 6
I Nocc 1 1 5 5 10 10
1 1 0 1.10-4 0 3.10-4 2.10-4 1.10-4
2 1 0 1,104 2.10-4 1,104 0
3 5 0 3.10-4 2.10-4 1.10-4
4 5 0 1.10-4 2.10-4
5 10 0 1.10-4
6 10 0
Tableau 5.3.3-1 : Initial matrix of the factors of use
the combination more penalizing is {i=1, j=4 } , of which the number of occurrences is 1:
tot −4
FA =01∗3 .10
The numbers of occurrences are upgraded: Nocc  1 =0 Nocc  4 =4 . The matrix of the factors
of use is upgraded; if the line i or the column j has a number of occurrences no one, it is resetting.
J 1 2 3 4 5 6
I Nocc 0 1 5 4 10 10
1 0
2 1 0 1.10-4 2.10-4 1.10-4 0
3 5 0 3.10-4 2.10-4 1.10-4
4 4 0 1.10-4 2.10-4
5 10 0 1.10-4
6 10 0
Tableau 5.3.3-2 : Stamp factors of use - iteration 1 of computation
computation continues same manner: the combination more penalizing is now {i=3, j=4 } , of which
the number of occurrences is 4:
FAtot =1∗3 .10−4 4∗3 .10−4
The penalizing combinations are then successively {i=3, j=5 } number of occurrences 1;
{i=5, j=6 } of many occurrences 9.
The factor of total use is then:
FAtot =1∗3 .10−4 4∗3 .10−4 1∗2 .10−4 9∗1. 10−4 = 2,6 . 10−3
5.4 Computations of the factor of priming with the option ` AMORCAGE'

The computation of the factor of priming is carried out with option AMORCAGE, for results of the type
'EVOLUTION' only. The array expected in input, under key word TABL_SIGM_THETA, corresponds to
the recording of the stresses on a circular cut line (of radius D_AMORC) around the singularity. The
stresses must be expressed in the local coordinate system.
Such an array can be created using command MACR_LIGN_COUPE. In the fatigue analysis (cf
§5.35.3), one considers that all provided times correspond to extrema of the transient. In addition, the
concept of transient is forgotten and one does not work any more but on a set of states of loading.

Version
Code_Aster default
One notes N occ i the number of occurrences associated with the state with loading i and N tot the
full number of states of loading.
The calculation algorithm is then the following:
Loops on the points of the Vérification
•cut line which the point is at the distance D of the singularity
•Boucle i=1… N tot
    i
Extraction de
Boucle j=i1 … N tot
Extraction de    j 
Calcul of      i , j=∣    j −    i ∣
min     i  ,     j  
Computation of R i , j =
max      i ,   j  
    i , j 
Computation of
  eff i , j =
R i , j 
1−
R_AMORC
Computation amongst acceptable cycles and of the elementary factor of
priming
1/B_AMORC
N a  i , j =    eff i , j 
A_AMORC 
1
FA i , j=
N ai , j 
À la fin de this first part, one thus has a matrix of the factors of priming of all the combinations of
states of loading. The size of the matrix is N tot ×N tot but only the part above the diagonal is
indicated.
The calculation algorithm of the factor of total priming, for a point given on the cut line, is then the
following:
1. FAmax =0
2. FAtot =0
3. Loops i=1… N tot (search of the maximum in the table)
If Nocc i 0 :
Loops j=i1 … N tot
If Nocc  j 0 and FA i , j FA max :
FAmax = FA  i , j  , m=i , n= j
4. Nocc  m , n =min  Nocc m  , Nocc  n  
5. FAtot =FA tot Nocc  m , n ∗FA  m , n 
6.Reactualization amongst occurrences:
Nocc  m =Nocc  m −Nocc  m ,n 
Nocc n = Nocc  n −Nocc m ,n 
7.Return to the beginning of the procedure until elimination of all the occurrences
À la fin de this algorithm, one thus has the factor of priming for each point (i.e for each angle) of the
cut line.

Version
Code_Aster default
6 Criteria of fatigue for the analysis simplified of the

pipework according to the RCC-M B3600
De même as for the case ` UNITAIRE' described previously, it is of use in B3600 to define each
situation as the transition of a state stabilized A (agent with an internal pressure given in the line of
pipework, a given uniform temperature, and fixed mechanical loadings) in a state stabilized B (with
constant loadings different from the precedents). One associates with this situation a thermal
transient. Stresses of thermal origin can also be taken into account in computation.
The processing which is described here is carried out for each node of each mesh of the line of
pipework considered. The result obtained will be thus a factor of use (total or partial) for each node of
each mesh required by the user.
6.1 Preliminary computation of all the states of loading

Pour each node of each mesh, the present stage consists with compute, for all the situations, the
moments relative in each stabilized state (by cumulating the various loadings which intervene).
6.1.1 Computations of the static states of loading

One treats the results of static computations (field EFGE_ELNO or SIEF_ELNO) for the stabilized states
of the list of the situations undergone by the line.
A stabilized state can be defined by a list of loading case, each load being signed. In this case, the
torsors of the stabilized state are obtained by algebraic summation of the torsors of each loading case:
M i= M i CHAR 1 M i CHAR 2 . .. i∈ { x; y; z }
The loadings are for example opposed thermal dilation, the displacement of anchorage.
6.1.2 Computation of the seismic loadings

the seismic loading breaks up into 2 parts:
•An inertial part
Elle is computed by imposing on all of the anchorages the same motion characterized by the
spectrum envelope of the various floor spectrums, in the horizontal directions X and Y on the
one hand, and vertical Z on the other hand (in the total reference). With this intention, the
command COMB_SISM_MODAL is used, which produces generalized forces which correspond
to each direction of earthquake as well as the quadratic office plurality of these forces.
The inertial contribution of the earthquake to the component i of the moment is written:
M i_S_DYN = ∑  Mj i_S_DYN  spectre j  
2
 i,j  ∈ { { x ; y ; z } ; { X ;Y ; Z } }
with Mi_S_DYN (spectrej) moment in direction I resulting from the dynamic loading in the direction J.
This office plurality is made directly by COMB_SISM_MODAL.
•A quasi-static part
Elle is estimated by imposing static differential displacements corresponding to maximum of the differences of
seismic motions of the points of anchorage in the course of time. Computations are thus carried out for each
unit loading (a computation by displacement in a direction given for an end of the line).
One notes N ANC the number of points of anchorage of structure. The quasi-static contribution of differential
displacements of anchorage to component I of the moment is written:

Version
Code_Aster default
∑
N_ANC
M i_S_ANC =  M i_S_ANC k 2
k =1
with M i_S_ANC k i the component
ème of the moment corresponding to k the ème
displacement of
anchorage.
Combination of the inertial components and differentials due to the earthquake:
i The ème component resultant is obtained by quadratic average of i the inertial and differential component
ème:
 2
M i_S =  M i_S_ANC    M i_S_DYN 
2
i ∈{ x ; y ; z }
what returns in fact to carry out it average quadratic of every inertial and differential moment,

2
M i_S = ∑  M i_S_ANC
k
  M i_S_DYN 2 i ∈{ x ; y ; z }
k=1,N_ANC
Pour the user, the situation of earthquake is defined by the results list corresponding to the inertial
response and the responses with the displacement of N ANC the successive points of anchorage. The
recombination by quadratic average is made directly by operator POST_RCCM.
6.1.3 Computation of the thermal transients

Les loadings of type “heat gradient in the thickness” are broken up into three parts, cf Appear 6.1.3-a :
- a constant value which is the mean of the temperature:
1 t /2
T moy = ∫ T  y . dy , where t corresponds to the nominal thickness of the wall.
t −t /2
- a linear distribution of null average (moment of order 1):
12 t /2
V= ∫ y . T  y . dy
t 2 −t /2
- a nonlinear distribution of null average and null moment compared to average fiber.
For each one of the transients and each section of pipework of the line (and each junction), one thus
realizes as a preliminary, according to the geometrical complexity of the problem studied a thermal
computation 2D or 3D.
Each computation is then stripped in order to extract, for each time of the transient, the temperature on
the selected section and the means (moments of order 0 and 1). This operation can be made for
example using two calls to POST_RELEVE_T (OPERATION = ` EXTRACTION' and OPERATION = `
MOYENNE').
In the case of a discontinuity of material or a junction, one computes the average temperature (noted
T a and T b ) on the two sides of the junction. In practice, the areas a and b will correspond to
segments chosen by the user in POST_RELEVE_T, and the produced arrays will be associated with the
two adjacent meshes having jointly the node which corresponds to the junction.

Version
Code_Aster default
Appear 6.1.3-a : Decomposition of the distribution of temperature in the thickness of the wall (figure extracted
the RCC-M, §B3653.4)
6.2 Computations of the amplitudes of variation of the Principe

6.2.1 stresses of the method
Les amplitudes of variation of the stresses are defined in the B3653 paragraph of the RCC-M for
combinations between two times or two states of loadings. While noting t i and t j these two times,
one has in a schematic way for the amplitude of variation of a quantity S :
mécanique mécanique thermique thermique
S  t i , t j =S t i −S  t j S  t i −S t j 
In the method as developed in Code_Aster, the situations are defined in a way simplified by two
stabilized states and a thermal transient: it is then not possible to work directly on each time of the
situations and of the assumptions must be introduced.
One thus considers all the combinations i , j with i , j∈1, 2,.. , N , 1, 2, .. , N  , N being the
number of states stabilized except earthquake (i.e. 2 times the number of situations of the group). Are
two stabilized states, i and j , belonging respectively to the situations p and q . The amplitude of
variation S will then be computed in the following way:
S  i , j =S mécanique  i−S mécanique  j max   S thermique  p  ,  S thermique  q  
by noting  S thermique  p  the amplitude of variation of the thermal stress of the transient p .
Note:
[1] It is important to note that the amplitude of variation of the stresses is
done by maximizing the amplitude of the thermal stresses for each
thermal transient independently one of the other. Method of calculation
for the case ` PIPEWORK' is thus different from that adopted for the
case ` UNITAIRE'.
[2] As indicated in the B3653.2 paragraph, all the system statuses must be
considered, including the states with null stress (for example cold stop:
null pressure and applied moments and ambient temperature) .

Version
Code_Aster default
6.2.2 Computation of the combinations of loading (I, J) inside each group of

situations
the objective is to build, for each group of situation, a symmetric square matrix containing all the
amplitudes of variation of the alternating load S ' alt  i , j  , with i and j two stabilized states
respectively associated with the situations p and q . This computation requires the preliminary
computation of the quantities S p (amplitude of the total stress) and S n (amplitude of the linearized
stress).
6.2.2.1 Notations and definitions

One notes:
C1 C2 C3 = Stress indexes provided to the §B3680 of RCC-M
K1 K2 ,
K3
E = Elasticity modulus of the pipework to room temperature
 = Poisson's ratio
 = Coefficient of thermal expansion of the pipework to room temperature
E ab = average Elasticity modulus enters the two areas separated by a discontinuity to
the room temperature
D0 = Diamètre external of the pipework
t = nominal Thickness from the wall
π
I = Moment of inertia of the pipework: I=
64

D 20−  D20 −2t 
2

M i i , j  = Variation of moment resulting from the various loadings of the situations to which
belong the stabilized states i and j :
M i  i , j =  M X
2 2
 i − M X  i     M Y  i −M Y  i    M Z i −M Z i 
2
P0  i , j  = Difference in pressure between the states i and j

T a  t k , tl  , = Amplitude of variation of the average temperatures in the areas a and b
T b tk , tl between times t k and t l
T o  t k ,t l  , = Amplitude of variation of the temperatures on the level of the offsite wall / intern
enters times t k and t l
T i  t k ,t l 
ΔT 1  t k , t l  = Amplitude of the variation between two times of the difference in temperature
between the walls internal and offsite, for an equivalent linear distribution of the
temperature:
12 t /2
ΔT 1  t k , t l = ∫ y .T  t k ,t l   y . dy = V  t k −V  t l 
t 2 −t /2
ΔT 2  t k ,t l  = nonlinear Partie of the distribution in the thickness of wall of the amplitude of
variation in the temperature enters times t k and t l :
∣
∣T o  t k ,t l −T moy  t k ,t l ∣−∣1/2 ΔT 1  t k , t l ∣
ΔT 2  t k ,t l  =max ∣T  t , t −T  t ,t ∣−∣1/2 ΔT  t ,t ∣
i k l moy k l 1 k l
0
6.2.2.2 Computation of S p i , j 

Version
Code_Aster default
the amplitude of variation of the total stresses S p for the pipework is defined in equation 11 of the
§B3653 of the RCC-M. One computes:
∣P 0  i , j ∣. D 0 D 1
S p  i , j ,t p =K 1 .C 1 .
2. t
K 2 .C 2 . 0 . M i  i , j  
2. I 2.  1−ν 
. K 3 . E . α .∣ΔT t kp , t lp ∣
1  
1
k  
 K 3 .C 3 . E ab .∣α a . T a t p , t lp −α b . T b t kp ,t lp ∣
1−ν 
. E . α .∣ΔT 2 t kp , t lp ∣  
t p
k
p
,tl  two unspecified times of the transient associated with the situation indicate p . In the event
of discontinuity of matter or a junction, the terms ΔT 1 and ΔT 2 to retain are those associated with
the section more penalizing.
One computes in the same way S p  i , j ,t q  with the thermal transient associated with the situation
q . The amplitude S p for the combination i , j is then:
{
S p  i , j =max max  S p  i , j ,t p   , max  S p  i , j , t q  
t qk , t ql   t ,t 
q
k
q
l
}
6.2.2.3 Computation of S n p , q 
the amplitude of variation of the stresses linearized S n for the pipework is defined in equation 10 of
the §B3653 of the RCC-M. One computes:
∣P 0  i , j ∣. D0 D0 1
S n  i , j ,t p =C 1 .
2.t 2. I
C 2 . . M i  i , j 
2 .  1−ν  
. E . α .∣ΔT 1 t p , t p ∣
k l 
 
C 3 . E ab .∣α a . T a t , t l −α b . T b t kp , t lp ∣
p p
k  
∣P 0  i , j ∣. D 0 D 1
S n  i , j ,t q =C 1 . C 2 . 0 . M i  i , j   . E . α .∣ΔT 1 t q , t ql ∣  
2 .t 2. I 2 .  1−ν  k
 
C 3 . E ab .∣α a . T a t q , t ql −α b .T b t qk ,t ql ∣
k  
{
S n  i , j =max max  S n  i , j , t p   , max  S n  i , j , t q  
t qk ,t ql  t ,t 
q
k
q
l
}
One computes then S n  p , q  =max S n  i , j  , for i and j sweeping all the stabilized states of the
i,j
two situations p and q (4 possible combinations).
6.2.2.4 Computation of S ' alt  i , j 

Deux formulas are proposed to define the amplitude of variation S ' alt  i , j  between the states i and
j :
• KE_MECA : it is the original method, only available in the previous versions to version 7.2:
1 E
S ' alt  i , j  = . c . K e  S n  p , q   . S p  i , j 
2 E
with:
- E c : Young modulus of reference for the construction of the curve of Wöhler, provided
by the user in DEFI_MATERIAU, under key word E_REFE, of the key word factor FATIGUE.

Version
Code_Aster default
- E : Smaller of the Young moduli used for the computation of the states i and j , i.e.
evaluated with the temperatures of these stabilized states.
∣
1 si S n  p , q  ≤3. S m
1−n

Sn  p , q 
K e  S n  p , q  = 1 n .  m−1  . 3 . S
m
−1
 si 3 . S m S n  p ,q  3 .m . S m
1
si S n  p , q  ≥3. m . S m
n
with m and n depend on the material, and provided by the user in DEFI_MATERIAU, under
key words M_KE and N_KE, of the key word factor RCCM. If key words TEMP_REF_A and
TEMP_REF_B are present, S n is interpolated for this temperature (which must correspond to
the average temperature of the transient). If not, S n is taken with room temperature.
• KE_MIXTE : since the modifying 1997 of the RCC-M, one can choose another formula, based
on a decomposition of S alt :
1 E
2 E e
S ' alt  i , j  = . c . K meca  S n  p , q   . S meca
p  i , j  K ther
e
 S n  p , q   . S ther
p i , j  
with:
- K meca  S n  p , q   is equal to K e definite above
e
 
1
1, 86 1−
Sn
- K ther
e  S n  p , q  =max∣ 1, 66
Sm
1
- S meca
p i , j  the amplitude of variation on mechanical behalf of the quantity represents Sp
, between the states i and j . It is computed on the basis of stress of mechanical origin:
pressure, inertia loading, earthquake (inertial and displacements of anchorage), thermal
expansion.
S ther
- p  i , j  the amplitude of variation on thermal behalf of the quantity
S p , between the
states i and j (terms dependent on T a T b , ΔT 1 and ΔT 2 in the definition of the
§6.2.2.2).
6.2.2.5 Cases of the under-cycles
Les under-cycles correspond either to the taking into account of the under-cycles related to the
earthquake, or with situations for which key word COMBINABLE='NON' was well informed. In both
case, one computes the amplitude of stresses while utilizing only the stresses related to these under-
cycles (not of combination of states of loading apart from this situation). For the computation of S ’ alt ,
it is necessary to use the factor K e which corresponds to the main situation from which the under-
cycle is resulting.
6.2.3 Computation of the combinations of loading (I, J) for the situations of transition
between group of situations
Deux states of loading are combinable only if they belong to the same situation or if there exists a
situation of transition between the groups to which they belong. In this last case, one will associate

Version
Code_Aster default
with the combination i , j the number of occurrences of the situation of transition. If the situation of
transition belongs to the one of the two groups (what is not excluded a priori), it is naturally combined
with the other situations of this group, then is used for the combination of the situations of its group
with the situations of the group in relation.
For each situation of transition of a group with another, one thus considers all the combinations i , j
with i pertaining to the first group (of dimension N ) and j pertaining to the second group (of
dimension M ). For each combination, S ' alt i , j is obtained in the same way that previously and
one associates to him the number of occurrences of the situation of transition. One builds a matrix
(rectangular) containing all them S ' alt i , j .
6.3 Computation of the factor of use

One notes:
n k : number of cycles associated with the situation p to which belongs the stabilized state k
;
nl : number of cycles associated with the situation q to which belongs the stabilized state l ;
Ns : many occurrences of the earthquake;
ns : many under-cycles associated with each occurrence with the earthquake;
n pass : many cycles associated with a possible situation with transition enter p and q if these
situations do not belong to the same group, but if there exists a situation of transition
between the two.
For all the combinations of states of loading (inside a group of situations or associated with a
situation of transition):
If N s0 , one selects N s /2 the combinations of stabilized states k and l more penalizing,
i.e. N s /2 the combinations k , l  leading to the greatest values of S ' alt  k , l .
For each one of these N s /2 combinations:
•A) Superposition of the moments of seismic origin and the combination k , l  :

• One superimposes the loadings of earthquake to the variation of moment resulting from the
various loadings of the stabilized states k and l :
 2 2
M i = ∣M 1  k − M 1  l ∣ΔM S1   ∣M 2  k  −M 2  l ∣ ΔM S2  ∣M 3  k  − M 3  l ∣ ΔM S3 
2
with:
M x  k  and M x  l  : components in the direction x ( x ∈ { 1 ; 2 ; 3 } ) of the moments
associated with the states k and l ;
ΔM Sx : total amplitude of variation in the direction x of the moments due to the earthquake
( ΔM Sx =2M x where M x is the total resulting moment (inertia and displacements of
S S
anchorage) such as defined in the §6.1.26.1.2).
• One computes then S p and S n such as previously definite with the new value of M i (noted
respectively Sp S
and S n ) and one computes:
S
1 E
S ' alt  k , l  = . c . K e  S n  m , n   . S p  k ,l 
S
2 E S S

Version
Code_Aster default
• One computes the number of acceptable cycles N  k ,l  for the amplitude of stress
S ' alt  k , l  using the curve of Wöhler associated with the material.
S
1
• One finally computes u 1  k , l  =
N k ,l 
•B) Prise in account of 2n s −1 seismic cycles considered as under-cycles:

• Amplitude of variation of the seismic stress only:
Ec D
S ' alt  k ,l  = K e  Sn k , l . K 2. C2 0  ΔM S12 ΔM S22 ΔM S32
SC
E S
4. I
• One computes the number of acceptable cycles N SC  k ,l  for the amplitude of stress
S ' alt SC
 k ,l  . It should be noted that one uses the value K e  S n  k ,l   previously computed
S
for the main cycle.

 2n s−1 
• One finally computes u2  k , l  =
N SC  k , l 
•C) Cumul
• u  k , l =u 1  k , l  u 2  k , l 
One starts again this computation until exhaustion of N s /2 the combinations more penalizing.
The computation of the factor of use is then continued without taking into account the earthquake.
If N s=0 , or after having taken into account the earthquake for N s /2 the most unfavourable
combinations:
•One selects the combination k , l  leading to the maximum value of S ' alt  k , l  , on all the
combinations, such as the number of occurrences n 0 is non-zero, with:
n 0 =min { n k , n l , n pass } if n pass is non-zero

n 0 =min { n k , n l } if n pass One is
•null computes the number of acceptable cycles N  k ,l  for the amplitude of stress S ' alt  k , l 
, using the curve of Wöhler associated with the material.
n0
•One computes then the factor of elementary use: u  k , l = .
N k ,l 
•One overrides finally:
n k by n k −n0 
n l nl −n0 
if it is about a situation of transition, n pass by n pass−n0 
then:
if n k =0 , the column and the line corresponding at the stabilized state k of the matrix
S ' alt  i , j  are settings with 0.

Version
Code_Aster default
if n l=0 , the column and the line corresponding at the stabilized state l of the matrix
S ' alt  i , j  are settings with 0.
The loop is repeated until exhaustion amongst cycles.
Note:
Appendix ZI of code RCC-M defines the curves of Wöhler until an amplitude of stress
minimum corresponding to one life duration of 106 cycles. If the computed value S ' alt for a
combination i , j of stabilized state is lower than this amplitude minimum, the factor of
use is equal to 0 for the combination (I, J) considered. This implicitly amounts considering
the existence of a limit of endurance to 10 6 cycles.

Version
Code_Aster default
7 Bibliography
1) “RCC-M: Rules of Conception and Construction of mechanical equipment of nuclear
islands PWR. Edition, modifying” Edited June 2007 June 2000 by the
AFCEN: French association for the rules of design and construction of the
materials of the nuclear boilers.
2) Y. WADIER, J.M. PROIX, “Spécifications for an ordering of Aster allowing of the

analyses according to the rules of the RCC-M B3200”. Note
EDF/DER/HI-70/95/022/0
3) I. BAKER, K. AABADI, A.M. DONORE: “Project OAR: Description of the ` file OAR',
filesystem of feeding of data base” Note EDF / R & D / HI-75/01/008/C
4) F. CURTIT “Réalisation of a software tool for analysis to fatigue for a line of

pipework - schedule of conditions” Note EDF / R & D / HT 2/26/010/A
5) F. CURTIT “Analyse with the fatigue of an interior line VVP BR with under-thickness”
Note EDF / R & D / HT-26/00/057/A
6) “Request for interpretation IC 73 (in response with request D4507-SIS-POT n

°07/0870 of J. Pot)”, AFCEN, 2007
7) “RSE-M: Rules of Surveillance in Exploitation of mechanical equipment of nuclear

islands REP. Edition 1997, modifying 2005” Edited by the AFCEN: French
association for the rules of design and construction of the materials of the
nuclear boilers.
8 Description of the versions of the document
Version Auteur (S) Description of the amendments

Aster Organisme (S)
5 J.M. Proix, EDF-R&D/AMA initial Texte
7.4 J.M. Proix, EDF-R&D/AMA Ajout of the unit method
8.4 E. Galenne, EDF-R&D/AMA Ajout of the results of the type UNITAIRE
9.4 E. Galenne, EDF-R&D/AMA - Ajout of the computation of the factor of
priming (file 10429)
- Modification of the formulation in fatigue for
the results de1type UNITAIRE (file 12297)
10.4 E. Galenne, Chau H.T., - Ajout of method KE_MIXTE for EVOLUTION
EDF-R&D/AMA (file 12818)

Postprocessing According To The RCC-M

Uploaded by

Copyright:

Available Formats

You might also like

Postprocessing According To The RCC-M

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Postprocessing According To The RCC-M

Uploaded by

Copyright:

Available Formats

Version

Postprocessing according to the RCC-M

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

5 Criteria of level A for the results of the type ` EVOLUTION'.............................................................20

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

The criteria of B3200 correspond in operator POST_RCCM to results (TYPE_RESU_MECA) of types `

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

2 In data in Code_Aster This

2.1 Geometrical data

Current area (except geometrical singularity):

2.2 Data of loading - Type EVOLUTION

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

2.3 Data of loading - Type UNITAIRE

 i= F x i. Fx_ U  F y i . Fy

2.4 Material characteristics

2.5 Simplifying assumptions

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

However, to fix the ideas, classification is simpler in the following cases:

2.6 Computations carried out by POST_RCCM

• criteria of level 0 by key word PM_PB,

3 Criteria of level 0 (key word PM_PB)

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

3.1.1 General primary equivalent stress of membrane Pm

3.1.2 Primary equivalent stress of local membrane Pl

The criterion is written (B3233.2):

3.1.3 Primary equivalent stress of membrane+flexion Pmb (or Plb)

P m =max ∥σ ijmoy∥ P b =max∥σ ijfle∥ P mb =max∥σ lin

σ lin moy fle

The criteria are written (B3233.3):

3.2 Computations carried out in the case ` EVOLUTION'

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

The algorithm is the following:

3.3 Computations carried out in the case ` UNITAIRE'

and the same for P b and P mb .

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

4 Criteria of level A for the results of the type ` UNITAIRE'

4.1 Definition of the criteria in the RCC-M

1. Amplitude of variation of the linearized stresses S n (option ` SN');

3. Computation of the thermal ratchet (option ` SN');

σ ijlin  s= 0  = σ moy fle

The criterion of total adaptation is written (B3234.2):

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

4.1.2 Computation of Sn*

- At the origin of the segment:

 max ∥  σ lin  t 1   −  σ linij  t 2   σ ijfleth  t 2  ∥

- At the end of the segment:

 max ∥  σ lin  t 1   −  σ linij  t 2  − σ ijfleth  t 2  ∥

4.1.3 Computation of the thermal ratchet

4.1.4 Computation of the factor of general

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

[ S ijmin  2 ]pq =Min  Min  S max

4.1.4.2 Computation of the elementary factors of use

Licensed under the terms of the GNU FDL (http://www.gnu.org/copyleft/fdl.html)

Two formulas are proposed (cf §B3234.6):

- E c : Young modulus of reference for the construction of the curve of Wöhler;

- K e the elastoplastic concentration factor: