Beruflich Dokumente
Kultur Dokumente
CHAPTER 5
the major mode of heat transfer and heat convection is less significant as far
as the temperature field in the welded body is concerned. The partial
differential equation for transient heat conduction is
wT
Uc .N T f
wt
(5.1)
where the density ( U ), specific heat (c) and the thermal conductivity ( N ) are
dependent on temperature (T). t is the time and f represents the additional
heat-generation function in the body.
N T
G
q (5.2)
³Tref c(W ) dW
T
H (5.3)
dH
c (5.4)
dT
From Equations (5.1) and (5.4), one can write the apparent heat
capacity equation in the form
wH
U .N T f
wt
(5.5)
hc T Tref e f sb T 4 Tref4
wT
N T .n N
G
wn
qn (5.6)
where the first term is convective heat loss and hc is the heat transfer
coefficient. The second term is the heat loss due to radiation and sb is Stefan-
Bolzmann’s constant and e f is the emissivity factor. The second term is a
nonlinear boundary condition. The total heat loss in Equation (5.6) can be
written in a format more convenient for finite element implementation as
K
QS t weld QS
(5.9)
V I t weld VI
where QS is the heat generating rate and tweld is the welding time and K is the
thermal efficiency. The Gaussian-distributed heat source (see Figure 5.1) can
be used to simulate the welding heat source to give a better prediction of the
temperature field near the source center.
q x, y q0 exp k r 2 (5.10)
‘k’ determines the concentration of the heat source. It is also known as the
distribution parameter representing the width of the Gaussian distribution
curve. Higher value of k corresponds to a more concentrated heat source.
S
³ ³ q x, y dx dy q0 §¨ ³ exp( kx 2 ) dx · §¨ ³ exp( ky 2 ) dy ·
f f f f
© f ¹̧ © f ¹̧
QS q0 (5.11)
f f k
density, i.e.,
The heat input parameter (k) can be evaluated from the heat source
radius as
| 2
ln (0.05) 2.9957 3
k (5.13)
rb2 rb2 rb
Using Equations (5.11) and (5.13) in Equation (5.10), one can write
§ r2 ·
¨¨ 3 2 ¸¸
3QS
S rb2 © rb ¹
q( x, y ) exp (5.14)
The time between the onset of welding and the end of the cooling to
ambient temperature can be divided into sufficiently small intervals so that
the temperature and thermal stresses for each interval may be regarded as
50
The calculation starts with time t=0 and the thermal stress was
calculated for the initial temperature distribution of the welded components.
At the next time step, the thermal stress increment was added to the initial
stress at step t =0. The magnitude of the cumulative thermal stress was
limited to the yield strength of the material at actual temperatures. It should be
noted that at each step, the forces caused by the induced thermal stresses must
be in equilibrium. This procedure was repeated until the last step at which the
thermal stress is that at ambient temperature, i.e. the residual stress. This
numerical procedure for residual stress evaluation involves adding together
the incremental thermal stresses, previous thermal stresses and the
equilibrium stresses.
P u i , kk O P u k , ki 3O 2P DT,i 0 (5.15)
H ij , kl H kl , ij H ik , jl H jl , ik 0 (5.16)
EQ
O
(1 Q ) (1 2Q )
(5.17)
P
2 (1 Q )
E
(5.18)
H ij
1
2
ui, j u j , i , is the strain; E is the Young’s modulus; and Q is the
Poisson’s ratio. These equations, together with the defined boundary
conditions provide the residual stress field in the welded joints. The term
‘simulation’ is often used synonymously with modeling, but there are
differences in meaning (Lindgren 2001). A simulation should imitate the
internal processes and not merely the result of the thing being simulated. This
gives an association into a simulation as a model that imitates the evolution in
time of a studied process. For example, a simplified model directly giving
residual stresses due to a welding procedure will not qualify as a welding
simulation. However, the term ‘simulation’ is often used to denote the actual
computation. Simulation errors will then be those errors related to the solution
of nonlinear equations as well as the time stepping procedure.
C Si Mn P S Cr Mo Cu
0.06-0.15 0.5 0.4-0.7 0.035 0.03 2.0-2.5 0.9-1.1 0.3
Figure 5.3 shows the contour plot and Figures 5.4 and 5.5 shows
the graphical plot of temperature variation from the weld center line to the
edge of the plate along Y-direction (that is along the length of the plate). The
results indicate that the plate is undergoing significant temperature variation.
At the beginning, the temperature reduction in the area close to the weld axis
shows the quenching effect. Figure 5.6 shows the contour plot and Figure 5.7
shows a comparison of the residual stress ( V x ) distribution perpendicular to
the weld of 2.25Cr1Mo steel plate obtained from the present 2D plane stress
analysis and 3D finite element analysis (Tahami and Sorkhabi 2009). The 2D
analysis result varies from 257 MPa (tensile) to -181 MPa (compressive) and
reaches to zero is in good agreement with those obtained from 3D FEA
results. Figures 5.8 and 5.9 give the stresses and strains at mid section
perpendicular to the weld.
56
700
650 at 25 sec
at 35 sec
600
at 55 sec
Temperature (K)
550
500
450
400
350
300
250
0 25 50 75 100 125 150
Distance from weld (mm)
Figure 5.4 Variation of temperature from weld center line to the edge
of the 2.25Cr1Mo steel plate along its length direction
(at t=25, 35 & 55 secs)
57
500
at 100 sec
450 at 500 sec
Temperature (K)
400
350
300
250
0 25 50 75 100 125 150
Distance from weld (mm)
Figure 5.5 Variation of temperature from weld center line to the edge
of the 2.25Cr1Mo steel plate along its length direction
(at t=100 & 500 secs)
300
250 2D FEA
Residual stress (MPa) 200
3D FEA (Tahami and Sorkhabi 2009)
150
100
50
0
-50 0 15 30 45 60 75 90 105 120 135 150
-100
-150
-200
Distance from weld (mm)
Figure 5.7 Residual stress, ( V x )distribution from the weld centre line to
the edge of the 2.25Cr1Mo steel plate along its length
direction
(Stamenkovic and Vasovic 2009). Figures 5.15 and 5.16 give the stresses and
strains at mid section perpendicular to the weld.
2000
at 10 sec
at 15 sec
1500
Temperature (K)
at 20 sec
1000
500
0
0 20 40 60 80 100
Distance from weld (mm)
Figure 5.11 Variation of temperature from weld center line to the edge
of the ASTM 36 steel plate along its length direction
(at t = 10, 15 & 20 secs)
62
600
at 50 sec
550
at 100 sec
Temperature (K)
500
450
400
350
300
250
0 20 40 60 80 100
Distance from weld (mm)
Figure 5.12 Variation of temperature from weld center line to the edge
of the ASTM 36 steel plate along its length direction
(at t = 50 & 100 secs)
400
2D-FEA
300 3D-FEA (Stamenkovic and Vasovic 2009)
Residual stress (MPa)
100
0
0 20 40 60 80 100
-100
-200
-300
Distance from weld (mm)
the quenching effect. Figure 5.19 shows the comparison of thermal history at
the node at weld center line of mid plate for 2D and 3D analysis. Figure 5.20
shows the contour plot of residual stress component ( V x ). Figure 5.21 shows
the comparison of 2D and 3D analysis of the longitudinal residual stress ( V x )
from the weld center line to the edge of the plate along transverse direction,
which varies from 346 MPa (tensile) to -195 MPa (compressive) in 2D while
400 MPa (tensile) to -195 MPa (compressive) in 3D analysis. Figure 5.22
shows the comparison of 2D and 3D analysis of the transverse residual stress
( V y ) from the weld center line to the edge of the plate along transverse
direction, which varies from 200 MPa (tensile) to -50 MPa (compressive) in
2D while 170 MPa (tensile) to -50 MPa (compressive) in 3D analysis. Figure
5.23 shows the comparison of 2D and 3D analysis of the longitudinal residual
stress ( V x ) along the weld line along longitudinal direction, which varies from
350 MPa (tensile) to 0 MPa in 2D and 350 MPa (tensile) to 0 MPa in 3D
analysis. Figure 5.24 shows the comparison of 2D and 3D analysis of the
transverse residual stress ( V y ) along the weld line along longitudinal
direction, which varies from 225 MPa (tensile) to -300 MPa (compressive) in
2D and 160 MPa (tensile) to -370 MPa in 3D analysis. Figure 5.24 shows that
the tensile stresses were developed in the weld zone. These tensile stresses
gradually decrease in the transverse direction away from the weld center line
and become compressive towards the edge of the plate. The peak tensile
residual stress estimates from the present 2D FEA is in good agreement with
those obtained from 3D FEA results (Deshpande et al 2011).
66
1600
1400 at 31.44 sec
1200
Temp ( 0 C)
1000
800
600
400
200
0
0 10 20 30 40 50
Transverse Distance (mm)
1600
1400
3D (Deshpande et al 2011)
Temperature - 0C 1200 2D
1000
800
600
400
200
0
0 20 40 60 80 100
Time(sec)
450
350 2D
250
Stress (Vx) MPa
3D (Deshpande et al 2011)
150
50
-50 0 10 20 30 40 50
-150
-250
Transverse Distance from weld (mm)
Figure 5.21 Comparison of residual stress, V x from the weld center line
to the edge of the plate along its transverse direction
250
2D
200
3D (Deshpande et al 2011)
Stress (Vy) MPa
150
100
50
0
0 10 20 30 40 50
Transverse Distance from weld (mm)
Figure 5.22 Comparison of residual stress, V y from the weld center line
400
350
Stress (Vx) MPa 300
250
200
150 2D
100
3D (Deshpande et al 2011)
50
0
-50 0 20 40 60 80 100
Longitudinal Distance (mm)
300
200
Stress (Vy) MPa
100
0
-100 0 20 40 60 80 100
2D
-200
3D (Deshpande et al 2011)
-300
-400
Longitudinal Distance (mm)
direction
71
Figure 5.25 shows the contour plot and Figure 5.26 shows the
comparison of temperature distribution at 10 sec from the weld center line to
the edge of the plate along Y-direction (that is along the length of the plate).
The peak temperature reaches up to 1973K for the present analysis and 2112K
in 2D FEA, are in good agreement. The results indicate that the plate is
undergoing significant temperature variation. At the beginning, the
temperature reduction in the area close to the weld axis shows the quenching
effect. Figure 5.27 shows the residual stress contour plot (V x ) for 3D
analysis. Figure 5.28 shows the residual stress ( V x ) distribution along top,
middle and bottom surfaces perpendicular to the weld obtained from the
present analysis, and it is compared with experiment result (Stamenkovic and
Vasovic 2009) and 2D finite element analysis. The analysis result varies from
383 MPa (tensile) to -78 MPa (compressive) and reaches to zero at top
72
surface, while it varies from 380.8 MPa (tensile) to -77.5 MPa (compressive)
and reaches to zero at middle surface, also varies from 388.4 MPa (tensile) to
-99 MPa (compressive) and reaches to zero at bottom surface, while it varies
from 380.5 MPa (tensile) to -190 MPa (compressive) and reaches to zero in
experiment and from 389 MPa (tensile) to -132 MPa (compressive) and
reaches to zero in 2D FEA results, is in good agreement. Figure 5.29 shows
the residual stress ( V x ) distribution along top, middle and bottom surfaces
along the weld line obtained from the present analysis, and is compared with
2D finite element analysis results. The analysis result varies from 386 MPa
(tensile) to -272 MPa (compressive) at top surface, while it varies from 394
MPa (tensile) to -132 MPa (compressive) at middle surface, also varies from
385.6 MPa (tensile) to -131.4 MPa (compressive) at bottom surface while it
varies from 387 MPa (tensile) to -1 MPa (compressive) in 2D FEA results.
Figure 5.30 shows the residual stress ( V y ) distribution for top, middle and
bottom surfaces along the weld line obtained from the present analysis and is
compared with 2D finite element analysis results. The analysis result varies
from 80 MPa (tensile) to -543 MPa (compressive), while it varies from 108
MPa (tensile) to -497 MPa (compressive) at middle surface, also varies from
83.5 MPa (tensile) to -434 MPa (compressive) at bottom surface, while it
varies from 161 MPa (tensile) to -343 MPa (compressive) in 2D FEA results,
is in good agreement. Figure 5.31 shows the residual stress ( V x ) distribution
through thickness at weld center line. The result varies from 388 MPa to 384
MPa (tensile). Figure 5.32 shows the residual stress ( V y ) distribution
through thickness at weld center line. The result varies from 72.5 MPa to 50.8
MPa (tensile).
73
2250
2000 2D
1750 3D
Temperature, 0K
1500
1250
1000
750
500
250
0
0 20 40 60 80 100
Distance from weld (mm)
500
3D Top
400
3D Middle
300 3D Bottom
Stress (Vx), MPa
2D
200
Test (Stamenkovic and Vasovic 2009)
100
0
-100 0 20 40 60 80 100
-200
-300
Distance from weld (mm)
Figure 5.28 Comparison of residual stress, V x from the weld center line
to the edge of the plate along its transverse direction
75
400
300
3D Top
200 3D Middle
Stress (Vx), MPa
3D Bottom
100
2D
0
0 20 40 60 80 100
-100
-200
-300
Distance from weld (mm)
200
100
0
Stress (Vy), MPa
-100 0 20 40 60 80 100
-200
3D Top
-300
3D Middle
-400 3D Bottom
-500 2D
-600
Distance along weld (mm)
390
388
Stress (Vx), MPa
386
384
382
380
378
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Distance from top surface to bottom (mm)
80
75
70
Stress (Vy), MPa
65
60
55
50
45
40
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Distance from top surface to bottom (mm)
center line
77