Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 1/12
Organization (S): EDF-R & D/AMA
Handbook of Utilization
U2.03 booklet: Thermomechanical
Document: U2.03.05
Note of use for calculations of welding
Summary
The objective of this note is to give information necessary so that a user can realize
multirun calculations of welding with Code_Aster. It is based on an example of welding of piping
on the 13 ways. The first 2 master keys of this example constitute a case-test Aster [V7.42.100].
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 2/12
Count
matters
1 thermal Modeling of welding ..................................................................................................... 3
1.1 General information ...................................................................................................................................... 3
1.2 Modeling of the contribution of heat ................................................................................................ 3
1.3 Case of a piping welded by pulsated process TIG (card 3488) ............................................. 4
1.3.1 Methodology: choice of an approach 2D or 3D ................................................................... 4
1.3.2 Approach thermal 3D in pointer on the plate 3D developed of the tube ................ 4
1.3.3 Approach thermal 2D on the tube ....................................................................................... 5
1.3.3.1 Imposed temperatures ............................................................................................ 5
1.3.3.2 Heat flow .......................................................................................................... 5
1.3.4 Conclusion ............................................................................................................................. 6
2 Modeling of welding in Aster .................................................................................................... 7
2.1 Grid of the weld beads ................................................................................................... 7
2.2 Thermal calculation ............................................................................................................................. 7
2.2.1 Modelings associated with the master keys .................................................................................... 7
2.2.2 Boundary conditions (convection and radiation) ............................................................. 8
2.2.3 Prolongation of the fields .................................................................................................... 8
2.3 Mechanical calculation ........................................................................................................................... 9
2.3.1 Modelings associated with the master keys .................................................................................... 9
2.3.2 Fastening of the welded zone .................................................................................................... 9
2.3.3 Mechanical calculation of a master key ........................................................................................... 10
3 ........................................................................................................................................... Conclusion 11
4 Bibliography ........................................................................................................................................ 12
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 3/12
1
Thermal modeling of welding
1.1 General
The modeling of an operation of welding requires the good knowledge of the process with
to simulate, in particular that of the parameters of welding. Moreover, phenomena to be taken in
account are many and complex: contribution of the molten metal, flow generated by the arc unit +
electrode, effect of gas protecting the bath melted, etc…
The principal difficulty of the thermal modeling of welding is the way in which one takes into account
the contribution of heat. In front of the great number of operational data accessible (energy from
welding, speed of the source, scrolling speed of the wire of contribution, output of the process…), it is necessary
in general to adopt a simplified method. These methods will be described in the § according to.
simulations carried out within the framework of a collaboratif work EDF-ECA-Framatome [bib1] showed
that other phenomena concerned (heat exchange, change of state, convection of the bath
melted…) are suitably taken into account by the computer codes by finite elements.
In the case which we present in the continuation, one has at the same time a good knowledge of
parameters of welding and elements of retiming like macrographies and the cycles
thermics. A correct thermal simulation is thus possible. The problem is then to choose
a method which will make it possible to represent accurately the contribution of heat due to the Arc unit -
Electrode - Métal of contribution.
If one does not have elements of retiming, the digital simulation of an operation of
welding can be carried out in a predictive way using calculations simplified of Rosenthal type. For
more details, one will consult [bib1]
1.2
Modeling of the contribution of heat
Two methods are possible:
·
the first consists in imposing cycles of temperature on the matter which one deposits. These
imposed temperatures can be applied either to the only cord deposited, or on
the cord unit deposited more molten zone. This method applies easily only to
two-dimensional problems. If one imposes a temperature in the filler, it
calculation proceeds in 3 phases:
1) Temperature imposed in the cord deposited until a temperature of molten bath
higher than the melting point.
2) Maintenance of the constant temperature during a time characteristic and increase of
thermal conductivity for temperatures higher than the melting point so
to find the molten zone.
3) Cooling with exchange by convection and radiation
The thermal cycle applied to the filler can result either from a calculation 3D or in
using calculations of the Rosenthal type to determine the maintenance and boarding times.
·
the second method, which is that recommended now, consists in imposing a heat flow on
weld bead modelled. It can this time apply in 2D and 3D and present
the advantage of modelling only the filler (it is not necessary to know the zone
melted). On the other hand, it is difficult to fix bus of the choices are to be carried out on the space distribution
flow (surface, voluminal) and the temporal distribution of flow.
The chock of flow can be controlled on the basis of calculations 2D of the Rosenthal type or by using them
experimental thermal cycles. In order to find the molten zone and to take account of
the homogenization of the temperature due to the movements of the bath melted, one increases beyond
melting point of a factor 100 thermal conductivity in the added metal.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 4/12
1.3
Case of a piping welded by pulsated process TIG (card 3488)
Here one knows a priori the form of the cords and the zones melted thanks to macrographies. It is
nevertheless difficult to use the method No 1 (approach in imposed temperature) insofar as
the chamfer contains 13 welding layers. To model the molten zones associated with the 13 cords
would have asked too much work in term of grid. This method is applicable only if the number
cords remains limited or if macro deposits are considered.
The other methods were tested but require all of the chocks with the results
experimental and give unequal results. Thus, approach in imposed temperature
considering the contribution of heat that in the filler over-estimates the molten zones and the cycles
thermics. This method is too calorific.
It is thus the approach in imposed heat flux which is retained. The application of the quantity of heat
Qr presents two alternatives: the application of surface or voluminal Qr. If one is considered
axisymmetric model of piping, the quantity of Qr heat is applied in all the meshs
modelled cord and one represent a flow 3D thus. In the case of a model 3D, it is difficult of
to consider voluminal Qr because one precisely does not know the form of the heat source.
It is thus surface Qr on the free edge of the cord that it is to better retain. However, Qr
surface is representative only if the cords deposited are not too thick, which is the case here.
1.3.1 Methodology: choice of an approach 2D or 3D
In any rigor, the process of welding is strictly 3D, the contribution of heat and possibly of
matter being mobile and constant speed. Numerical calculations should thus hold account of it.
However, complex and expensive calculations 3D being, they are seldom implemented and one limits oneself
with the 2D. Modeling 2D implies an important simplification: one neglects the effect speed of
welding and one suppose that the cord is deposited simultaneously over the entire length of the chamfer.
However, in order to carry out the chock of the yield coefficient of the process, calculations 3D in
locate mobile were also led for 2 master keys of the chamfer: the master key of root and one
pass current (master key 13).
The chock of master key 13 was then applied to all the other current passes.
The methodology retained for the thermal simulation of pulsated process TIG is thus the following one:
·
chock of the yield coefficient of the process for master keys 1 to 13 thanks to calculations
3D in pointer on master keys 1 and 13,
·
transposition with the model 2D by preserving the yield coefficient and by applying one
quantity of heat to the meshs of the modelled cord. The temporal distribution this flow is
fixed on the results of calculations 3D,
·
validation of this approach to the thermal simulation of master keys 2 to 13.
1.3.2 Approach thermal 3D in pointer on the plate 3D developed of the tube
The chock of the method is done on the transverse form of the zone melted by adjusting the coefficient
of output of the process. Several iterative calculations make it possible to have an acceptable molten zone.
For the master key of root, the yield coefficient is worth 0.65 then.
UI
The quantity of heat applied is worth Qr =
with:
S
·
U, the voltage welding taken equalizes with 11V,
·
I, the intensity of welding taken equalizes with 200A,
·
S, the surface of the source which is worth R2 with R=5mm.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 5/12
= 0 6
. 5
thus Q
W mm
R = 18 21
2
.
/
For master key 13, one considers a circular heat source of R=3.5mm radius, therefore lower than
that of the master key of root. Following the chock, the yield coefficient is worth 0.55 and.
Q
W mm
R = 32 87
2
.
/
1.3.3 Approach thermal 2D on the tube
The axisymmetric approach 2D is easy to implement but presents the disadvantage of not
to take into account the effect speed and to consider that each cord is deposited “of only one
blow “. Comparisons between approaches 2D and 3D [bib3] nevertheless showed the maid
representativeness of the approach 2D.
1.3.3.1 Temperatures
imposed
This method is the first method described in [§2.2]. The contribution of heat is modelled by one
cycle thermal resulting from calculation in pointer. This cycle is applied to the nodes of the added metal.
It is noted that calculations 2D over-estimate the molten zone and maximum thermal cycles.
The approach in imposed temperatures is thus too energy.
1.3.3.2 Heat flow
This method is the second method described in [§2.2] and is that recommended now.
The application of the heat flow Qr is carried out on the meshs of the cord deposited. This heat flow
is voluminal, a.c. D. per circumferential unit of length (J/mm3). Flow is given by:
UI
Qr =
with v: speed of the source.
Sv
Calculation 2D disregarding speed of the source, it is necessary to distribute this heat flow in
function of time: a descent and maintenance, boarding time.
Q * (W/MM3
R
)
0
T
time (S)
1
t2
T3
Q = Q *
1
× (T + T - T
R
R
)
2 3
2
1
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 6/12
It is necessary to determine moments T1, t2 and T3 to find the molten zones and the thermal cycles.
Several distributions corresponding to more or less long rise and fall times have
summer compared. In fact finally the following distributions were retained [bib3]:
Q * (W/MM3
R
)
pass from root
time (S)
7.
12.8
13.8
Q * (W/MM3
R
)
pass 13
time (S)
7.
10.4
11.4
1.3.4 Conclusion
Within the framework of the thermal simulation of welding by pulsated process TIG of a piping in
stainless steel 316L, the thermal history of the process can be represented by a model 2D
axisymmetric, even if retiming is better by considering an approach 3D in pointer which
account of the effect speed of the heat source takes. The methodology retained in 2D is one
approach in heat flow which requires to fix the yield coefficient of the process as well as
temporal distribution this flow. It is also necessary to know either the forms of the melted zones, or
thermal cycles of the process studied in order to allow this chock. It makes it possible not to model
melted zones, which is an important advantage in the case of a great number of master keys.
The principal characteristics of simulation are as follows:
·
heat flow applied in the metal added according to time (gone up in 7s, times of
maintenance of 5.8s for the master key of root and 3.4s for a current master key and descent in
1s),
·
the yield coefficient of the process is fixed at = 0 6
. 5 for the master key of root and with
= 05
. 5 for the current master keys,
·
the conductivity of the added metal is increased between 1500°C and 1700°C of a factor 100 so
to take into account a homogenization of the temperatures in the bath melted,
·
one takes into account the heat of fusion-solidification,
·
the thermal characteristics vary with the temperature.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 7/12
2
Modeling of welding in Aster
2.1
Grid of the weld beads
The weld beads can be with a grid in a more or less complex way.
There are 3 possible choices, while going from most complicated towards simplest:
·
one can choose to respect at the same time the volume and the form of the master key. The shape of the cords
being curved, one will have to net surfaces on curved board, therefore to use finite elements
at least of degree 2,
·
one respects only the volume of the master key, the cords being of triangular form or
quadrangular. In this case, one can use linear elements,
·
the weld beads are quadrangular, while trying to respect volume as well as possible
of each master key.
Comparisons were made in the case of the tubular model. It proves that the results of
mechanical calculations differ very little from one grid to another. Nevertheless, the comparison with
coarser grid is delicate, thermal being different to it.
One can however note that, to thermics equivalent, the curved grid does not bring anything significant
on the level of the results compared to the polygonal grid.
2.2 Calculation
thermics
2.2.1 Modelings associated with the master keys
To simulate welding multirun, one carries out a transitory nonlinear thermal calculation, passes
by master key, by adding to each master key in the thermal model corresponding the finite elements
modelling the weld bead deposited during the master key. Thus, each master key I has one
thermal model including/understanding the weld beads of numbers 1 to I. There are thus models
thermics encased with the following direction:
if mothi indicates the thermal model of master key I
and mothj indicates the thermal model of the master key J
then mothi mothj if I < J.
This poses a problem at the time of the sequence of thermal calculations, the fields of temperature of
model mothi not being defined in all the nodes of the model corresponding to the following master key i+1.
It is thus necessary to carry out a prolongation of the computed fields from one model to another. (see [§3.2.3]).
Note:
Another solution consists in considering one model containing all the passes and
“artificially to decontaminate” the cords not yet deposited in their imposing one
null thermal conductivity. This artifice can cause light numerical oscillations
temperature due to the discontinuity of conductivity to the interfaces between the cords.
It is nevertheless this method which is used in the cast-test [V7.42.100]. It allows
to save the stage of prolongation of the fields [§2.2.3].
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 8/12
2.2.2 Boundary conditions (convection and radiation)
For the modeling of the convectif and radiative exchange, one chose to make appear:
·
during the rise and the maintenance of Qr: convectif and radiative exchange on all the borders
except that of the cord of the current master key,
·
during cooling: same boundary conditions with in more the border of the cord
current master key (to take into account the weld bead have been just deposited).
The figures of boundary conditions (meshs of edge support of the boundary conditions) are with
to reactualize with each master key on the level of the chamfer, cords piling up the ones on the others. It
is necessary to have envisaged this operation upon the departure, at the moment it grid, i.e. to have
created as many figures as there are master keys.
2.2.3 Prolongation of the fields
The prolongation of the fields of temperature (and possibly of metallurgy) is necessary to
end of each master key so that those are defined on the model of the following master key. One proceeds in
4 stages:
·
one starts by creating a field of temperature to ambient (T20) on all the grid by
command CREA_CHAMP (operation “AFFE”),
·
one extends the first field of temperature calculated to master key I (sequence number 1) in
supplementing by T20 on new meshs (CREA_CHAMP operations “EXTR” then “ASSE”),
·
one stores this field in a new structure of data of the evol_ther type by
order CREA_RESU,
·
one makes a loop on the remaining sequence numbers and one repeats operations 2 and 3 for
each sequence number by enriching the structure of data created in 3 (key words reuse of
CREA_RESU).
Example:
# EVOTH1: EVOL_THER ON MODEL MOTH1 KNOWN EAST
# ONE WANTS TO CALCULATE EVOTH2 ON “A LARGER” MODEL MOTH2 THAN MOTH1.
# the 2 MODELS ARE BASED ON the SAME GRID MALL
# EXTENSION OF THE FIELDS OF TEMPERATURE BY 20 DEGREES C:
T20=CREA_CHAMP (OPERATION=' AFFE', TYPE_CHAM=' NOEU_TEMP_R',
MAILLAGE=MAIL, AFFE=_F (TOUT=' OUI', NOM_CMP = (“TEMP”,),
VALE = (20.,)) )
# EXTENSION OF FIELD TCH1 IN TCH2 ON the FIRST SEQUENCE NUMBER:
TCH1=CREA_CHAMP (OPERATION=' EXTR', TYPE_CHAM=' NOEU_TEMP_R',
RESULTAT=EVOTH1, NOM_CHAM=' TEMP', NUME_ORDRE=0,)
TCH2=CREA_CHAMP (OPERATION=' ASSE', TYPE_CHAM=' NOEU_TEMP_R',
MAILLAGE=MAIL, ASSE= (_F (ALL = “YES”, CHAM_GD = T20),
_F (ALL = “YES”, CHAM_GD = TCH1),))
EVOTH12 = CREA_RESU (TYPE_RESU=' EVOL_THER', NOM_CHAM=' TEMP',
AFFE= (_F (CHAM_GD=TCH2, LIST_INST= LPAS,
NUME_INIT=0, NUME_FIN=0,),))
TO DESTROY (CONCEPT=_F (NOM= (“TCH1”, “TCH2”),),);
# EXTENSION OF FIELDS TCH1 IN TCH2 ON the OTHER SEQUENCE NUMBERS:
for I in arranges (325):
iordr=i+1;
TCH1=CREA_CHAMP (OPERATION=' EXTR', TYPE_CHAM=' NOEU_TEMP_R',
RESULTAT=EVOTH1,
NOM_CHAM=' TEMP',
NUME_ORDRE=iordr,)
TCH2=CREA_CHAMP (OPERATION=' ASSE', TYPE_CHAM=' NOEU_TEMP_R',
MAILLAGE=MAIL, ASSE= (_F (ALL = “YES”, CHAM_GD = T20),
_F (ALL = “YES”, CHAM_GD = TCH1),))
EVOTH12=CREA_RESU (reuse=EVOTH12, TYPE_RESU=' EVOL_THER
, NOM_CHAM=' TEMP',
AFFE= (_F (CHAM_GD=TCH2, LIST_INST=LPAS, NUME_INIT=iordr,
NUME_FIN=iordr,),))
TO DESTROY (CONCEPT=_F (NOM= (“TCH1”, “TCH2”,),),);
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 9/12
For the metallurgy fields, the sequence of the commands is the same one but the types of
field are different.
# SAME PROCESSING FOR THE METALLURGY FIELDS
MOTH13 = AFFE_MODELE (GRID = MALL, AFFE= _F (GROUP_MA = (“PASSE13”,),
PHENOMENON = “THERMAL”, MODELING = “AXIS”,))
MINIT=CREA_CHAMP (OPERATION=' AFFE', TYPE_CHAM=' CART_NEUT_R',
MODELE=MOTH13, AFFE=_F (TOUT=' OUI',
NOM_CMP = (“X1”, “X2”, “X3”, “X4”, “X5”,), VALE = (1., 0., 0., 0., 10.,)) )
# EXTENSION OF FIELDS MCH1 IN MCH2 ON the OTHER SEQUENCE NUMBERS:
for iordr in arranges (325):
MCH1=CREA_CHAMP (OPERATION=' EXTR', TYPE_CHAM=' ELGA_VARI_R',
RESULTAT=EVOTH1, NOM_CHAM=' META_ELGA_TEMP', NUME_ORDRE=iordr,)
MCH2=CREA_CHAMP (OPERATION=' ASSE', TYPE_CHAM=' ELGA_VARI_R',
MODELE=MOTH13, PROL_ZERO=' OUI',
ASSE= (_F (TOUT=' OUI', NOM_CMP= (“X1”, “X2”, “X3”, “X4”, “X5”,),
NOM_CMP_RESU= (“V1”, “V2”, “V3”, “V4”, “V5”,) CHAM_GD=MINIT,),
_F (TOUT=' OUI',
CHAM_GD=MCH1,),))
EVOTH12=CREA_RESU (reuse=EVOTH12, TYPE_RESU=' EVOL_THER',
NOM_CHAM=' META_ELGA_TEMP',
AFFE= (_F (CHAM_GD=MCH2, LIST_INST= LPAS,
NUME_INIT=iordr,
NUME_FIN=iordr,),))
TO DESTROY (CONCEPT=_F (NOM= (“MCH1”, “MCH2”,),),);
2.3
Mechanical calculation
2.3.1 Modelings associated with the master keys
Contrary to thermal calculations, it is advised to use the same mechanical model for
all passes. This model will include/understand all the cords, the cords not deposited being
decontaminated artificially in their affecting a very weak Young modulus (E =
-
10 the 11th reality in
practical). The interest of such a technique is that the soft cords become deformed with the chamfer,
allowing to take again the following master key on the geometry deformed without having mending of meshes to make.
It should nevertheless be taken care that the not activated cords preserve a realistic form during
calculations. If it is not the case, they should be re-meshed.
For the cord deposited, the real mechanical characteristics are imposed to him when this one has
reached the melting point. Thus, in the phase of heating, the cord is still fictitious.
This technique is preferable with that consisting in duplicating all the nodes of the interfaces of
cords and to impose connections in increment of displacement between ddl. Indeed, the latter
technique, even if it reproduces reality rather accurately, has as a disadvantage of involving
whimsical deformations, cords not being attached to the structure at the beginning of each master key. Of
more, the setting in data of the connections is heavy and their expensive taking into account in time CPU.
2.3.2 Fastening of the welded zone
The axisymmetric modeling of welding on a tube supposes implicitly wrongly that welding has
place simultaneously on all the circumference of the tube, therefore that the temperature rises everywhere in
chamfer. In reality, the heat source progresses towards part of structure remained cold,
who attaches obligatorily the welded zone. The part, on the level of the heat source, thus cannot
to dilate freely. This effect of autobridage must grow blurred when diffuse heat and disappear with
run of the phase of cooling.
To cure this problem, one can force an axial fastening on the tube, only in the phase of
heating. One thus prevents the tube from freely dilating with the heating, on the other hand it is free
to deform with cooling.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 10/12
The implementation in Aster is done in the following way:
·
one models the supports by elements of edge (segments) which one blocks them
displacements. These elements have nodes confused geometrically with the nodes
in with respect to the border of the tube,
·
one puts in contact the supports and the elements of edge of the border (key word CONTACT of
AFFE_CHAR_MECA),
·
at the end of each master key, when the geometry is reactualized, i.e. one replaces
initial geometry by the deformed geometry, these fictitious supports also should be reactualized
by repositioning them on the deformed border, as the 3 diagrams show it
below:
initial supports
chamfer
deformed after master key 1
reactualized supports
2.3.3 Mechanical calculation of a master key
The mechanical model used is that which can take into account the effects of the transformations
metallurgical. The law of behavior used is élasto-visco-plastic. The model is isotropic
with a function threshold of the Von Mises type and a nonlinear isotropic work hardening with restoration
viscous of work hardening. One does not take account of the phenomena of plasticity of transformation
and of metallurgical restoration of work hardening (law of behavior META_VNL in Aster). The effect
kinematic work hardening was not looked at but it can be taken into account.
The increments of deformations used for the incremental relation of behavior are them
linearized deformations of the increment of displacement in the reactualized geometry (large
displacements, small deformations). It is option PETIT_REAC of STAT_NON_LINE (the large ones
deformations are possible but are generally not necessary)
The convergence of the method of Newton is difficult at the beginning of cooling and it is necessary
to use the linear algorithm of search to improve convergence. (key word RECH_LINEAIRE
STAT_NON_LINE by using the default values).
At the end of each master key, one reactualizes the grid, i.e. one replaces the initial grid by
deformed geometry. (operator MODI_MAILLAGE key word factor DEFORME) and they are reactualized
supports.
The processing of the plastic incompressibility poses problem by generating oscillations of
important constraints, in particular of the trace. The use of under-integrated elements QUAD8 does not have
not made it possible to solve the problem because the grids comprise many elements TRIA6 in
the plasticized zones for which one did not have a version under not integrated. One recommends to use them
new incompressible elements (modelings PLAN_INCO, AXIS_INCO and 3d_INCO), which have
given promising results from this point of view. These elements will be available in version 6.3
of Aster.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 11/12
3 Conclusion
The digital simulation of a test of welding on tube on the 13 ways made it possible to release one
method for calculation whose principal points can be summarized as follows:
·
grid: each cord must be with a grid by a group of meshs. It is not necessary
to have a very precise representation of the shape of the cords. On the other hand, the respect of
volume and of the position of the cord in the chamfer is important.
·
thermics: the modeling of the contribution of heat is the essential point. Method
recommended consists in imposing a heat flow on the meshs of the cords deposited. This flow
is constant spaces some and function of time. To determine the temporal dependence, it
is necessary to proceed to a thermal retiming starting from experimental data (evolutions of
temperature, zones molten) or in the absence of data on the process of welding and of
simplified calculations.
thermal retiming has an important effect on the final mechanical results, them
deformations and residual stresses being sensitive as much to the value that with
distribution of the heat source.
·
mechanics: in the case of a high number of master keys, the displacements cumulated in
chamfer are important and it is preferable to make a calculation in great displacements with
reactualization of the grid at the end of each master key. The assumption of the great deformations
is on the other hand not necessary.
in the case of a tube, the axisymmetric modeling of welding requires the catch
in account of boundary conditions particular, more precisely axial fastening
zone welded to take into account the fact that the torch progresses towards one
part of structure remained cold. This fastening is essential to obtain
correct values of the contracting of the chamfer.
·
Aster modeling: one recommends to build encased thermal models
container that cords actually deposited and to prolong the computed fields of one
model with the other. On the other hand, it is preferable to have only one mechanical model comprising
the totality of the cords upon the departure, cords being decontaminated artificially in their
affecting a Young modulus quasi-no one. In this way, one avoids having to re-mesh them
weld beads as the chamfer becomes deformed.
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Code_Aster ®
Version
6.3
Titrate:
Note of use for calculations of welding
Date:
01/10/03
Author (S):
X. DESROCHES Key
:
U2.03.05-A Page
: 12/12
4 Bibliography
[1]
F. WAECKEL: Synthesis of thermal modelings of an operation of welding
realized in the co-operative card 3449. Note EDF DER HI-74/95/028/0
[2]
F. WAECKEL, L. BIRONNEAU: Mechanical simulations of an operation of welding
multipass on autobridée plate. Note EDF DER HI-74/96/006/0
[3]
C. WOOD: Thermal simulation of welding multirun by process TIG pulsated of one
stainless steel piping. Note FRAMATOME EER cd. 1509
[4]
J. DEVAUX
: Tripartite card 3488. Digital simulation of welding. Calculations
thermo mechanics of master keys 1 to 5. Note FRAMATOME L/99.4564
[5]
X. DESROCHES: Digital simulation of a test of welding on tube on the 13 ways. Note
EDF DER HI-75/00/016/A
[6]
X. DESROCHES, A. RAZAKANAIVO, C. WOOD, pH. GILLES, J. KICHENIN, Y. LEJAIL:
Synthesis of the FC3488: “Simulation of welding multirun: experimental validation”.
Note Technique ECA DER/SERI/LCS/01/4027
Handbook of Utilization
U2.03 booklet: Thermomechanical
HT-66/03/002/A
Outline document