Sie sind auf Seite 1von 67

Recent Developments in Analysis and Testing

of Spot and Seam Welds


Dr. Shicheng Zhang

Brief review
The recent developments are characterised by the progress in
local approach which was first proposed by Pook in 1975.
The key to the local approach is the determination of local
stress parameter (LSP).
Different concepts are under development to get LSPs at spot
welds, especially for the automotive structures.
Various specimens are proposed for spot weld testing.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Q&A about local stress parameters at spot welds


What are LSPs at spot welds?
They are structural stress, notch stress, S.I.F., J-integral,
CTOD and any describer of stress concentration.
Why do we need the LSPs?
Because they condensate fatigue test data.
How to determine the LSPs?
They can be numerically calculated (stress solver),
analytically
approximated (stress formulas) and experimentally measured
(strain gauge technique).

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Structural stresses around a spot weld

Stress solver

Structural stress is
plate theory stress without singularity
linearly distributed over sheet thickness
output stress of shell elements (FEM)
Structural stress is denoted as (=r ui ).

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Notch stress around a spot weld

Stress solver

Notch stress is denoted as k .


Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Stress solver

A stress solver for spot welds


Input : ij
Stress components around a spot weld

The Stress Solver


A package of analytic solutions to spot welds

Output : r , k , K, J, ,

Structural stress, Notch stress, S.I.F., J-integral, Crack growth angle, T-stress

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

A package of solutions to spot welds

Stress solver

r max{ui ,li}

k n

4 3

{ui uo li lo [2(ui li )2 (uo lo )2

2(ui li uiuo uilo uoli lilo )]1/ 2 }


1 3
KI [ (ui uo li lo ) 5 2 (qu ql )] t
6 2
1
2
2
2
KII [ (ui li ) (qu ql )] t
Keq KI2 KII KIII
4
3 5
2
(ui li ) t
KIII
2

(12 )t 2
2
[4(ui li uili ) (uo lo )2 2(uiuo uilo uoli lilo )]
48E

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

A package of solutions to spot welds

Stress solver

arctan 3(ui li )
ui uo li lo

3uo 2ui lo
4
T T
ui li uo ui li lo
4
T 32li uo
lo lo
4
T

uo

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Finite element model

Stress solver

Diameter of the spoke pattern = Nugget diameter


Diameter of the central beam element = Nugget diameter

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Stress solver

Application of the stress solver

Table 1. Stress intensity factors KI and KII at the leading vertex and KIII at the
side vertex of the weld spot, all in N/mm , under tensile-shear force F=1 kN
with nugget diameter d=5 mm and sheet thickness t=1 mm
3/2

Authors
Pook [2], Analytic approximation
Radaj and Sonsino[23], FEM
Smith and Cooper [15], FEM
Swellam et al. [16], FEM
Yuuki and Ohira [17], FEM+BEM
Zhang [24], Analytic approximation
Zhang, FEM

KI
163.3
60.0
63.2
64.2
55.5
55.1
77.9

KII
199.8
147.8
132.4
134.6
123.2
127.3
130.0

KIII
65.4
99.0
90.3
90.0
86.9

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

10

Application of the stress solver

Stress solver

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

11

Stress solver

Interface-force-based estimations
Fx
My
1.872K 2
dt
dt

Rupp et al.:

Maddox:

2Fx 6M y
r
2
dt t w

Sheppard:

ij

Zhang:

Fij

t1

Mz
Fz FyMy
Fx Mx

6Mij

FAi

2
ti
ti2

(i, j 1,2)

4F 6M 4Fz
dt dt2 d 2

4F
3 19
(1
k
dt
8

(F

Fx2 Fy2 ; M Mx M

t 6M
) 2 (1
dt

3F 2 3M 5 2Fz
KI
2d t dt t 3d t

2
3

2F
KII
d t

2
y

5d
) 4Fz (1
d 2
3t 2
t

KIII

2F 2 2Mz
2
d t d t

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

12

Stress formulas

Tensile-shear

1.273F
r
dt

4F
r
dt

1.273F
t
(10.430 )
k
dt

0.694F
Keq
dt

Keq KI2 K

2
II

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

13

Stress formulas

Hat-profile

0.637aMt
r
dtA
0.637aMt (10.347 t )
k
dtA

0.392aMt
Keq
Ad t
Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

14

Double-cup

Stress formulas
1.273F
0.0938D
(cos
sin)
r
dt
t
1.273F
0.0938D
[cos
sin
k
dt
t
0.0458D
t
(0.430cos
sin) ]
t

0.318F
2
D2 2
Keq
(4.750cos 0.105 2 sin
t
dt
D
0.281 sin 2)1/ 2
t

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

15

Double-U

Stress formulas
1.273F
(b d)
[sin

cos

cossin]
r
4t
dt
4t

arctan
(b d) tan
1.273F
t
[(10.282
) sincos
k
dt

(b d)

4t

(10.651

) cossin]

4t(10.282 t / )
arctan
(b d)(10.651 t / ) tan

0.637F
[sin2 cos2
Keq
dt

2 (b d)2 2
cos sin2 ]1/ 2
12t2

=0 or /2
Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

16

Double-U

Stress formulas
1.273F
(b d)
[sincos
cossin]
(1)
dt
4t
t
t
1.273F
(b d)
[(1 0.282 ) sincos
(10.651 ) cossin] (2)
k
dt
4t

0.637F
2 (b d)2
2
2
(3)
cos2 sin2 ]1/ 2
[sin cos
Keq
2
12t
dt

FEM:
Spoke pattern + Stress solver

Eq. (1)

Eq. (2)

Eq. (3)

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

17

Fatigue test data condensation


across different specimens

Stress formulas

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

18

Sources of scatter of fatigue data

Local geometry:
- Sheet thickness t;
- Spot diameter d;
- Notch radius .

Material:
- Composition;
- Micro-structure;
- Dislocation density.

Stress formulas

Welding effects:
- Residual stress;
- Material inhomogeneity;
- Welding defects.

Load/boundary condition:
- Stress ratio R;
- Definition of failure;
- Test conditions.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

19

Optimization of spot weld distribution

Stress formulas

Durability reserve for every spot weld:

F
R 336
dt

N / mm3 / 2

Critical spot welds:


Spot welds with R<0 are dangerous.
Removable spot welds:
Spot welds with R 336 may be
saved.
Optimization:
Homogeneous distribution of Rvalue.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

20

Strain gauge technique

Stress distribution around


a spot or similar weld

ij (r,) 0 [ An (r) cos(n)

n1

Bn (r) sin(n)]

ij1 (r) ij3 (r)

ij (r,)

cos

2
ij1 (r) ij3 (r)
cos2

2
ij2 (r) ij4 (r)
sin
2
ij2 (r) ij4 (r)
sin2
2

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

21

Non-destructive testing
of spot welds

Strain gauge technique


Nugget

Inside stress at nugget edge is decisive.


The stress is inaccessible to measure.
Destructive measures may be taken.
Non-destructive method is desired.

Inside stress

Strain gauge

Strain gauge

Destructive:

(Japan)

Hole

Strain gauge

(BMW, Germany)

Removed (BAM, Germany)


Strain gauge

Non-destructive:

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

22

A strain gauge solution


to spot-welded lap joint

Strain gauge technique

r Eds
[(2ds d)uo 4uo ]
2(1 2 )d

Ed t
KI I s 2 {[
12(1)d

uo (uo2 uo ) / x

3(3ds 2d) 5 2t
]uo 3 3uo}
2
3

2II Eds t 2d
s d

[(

)uo uo ]
KII
2
(12 )d 8
9 5
k

1
k Eds
{(2ds d)uo 4uo
2(12 )d
4 3

[(3ds 2d)uo
6uo 3[(2ds d)
uo 4uo ]2 [(3ds 2d)uo 6uo ]2 ]}

J Etds2

2
2 {3[(2ds d)uo 4uo ]2 [(3ds 2d)uo 6uo ]2}
192(1 )d

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

23

Virtual testing

Strain gauge technique


uo (uo2 uo ) / x

Table 4. LSPs determined by the strain gauge solution in the virtual


testing (VT) compared with their finite element (FE) results for a tensileshear specimen with d=5 mm, t=1 mm and =0.2 mm under tensileshear force of F=1000 N; stresses are in N/mm , SIFs in N/mm and Jintegral in N/mm
2

Strain gauge solution

3/2

J-Integral

KI

KII

374.5

77.9

130.0

622.1

0.089

77.9

130.0

622.5

0.089

Finite element simulation 374.4

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

24

Monitoring LSPs

Strain gauge technique


r k KI KII J

t,u,F
Strain gauge solution

Strain gauge apparatus

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

25

Conclusions
1. A stress solver for spot welds is generally developed.
a package of analytic solutions to spot welds is available.
the solutions provide an analysis tool for spot welds.
2. Approximate stress formulas for spot weld specimens are given.
LSPs at spot welds are directly related to the applied load.
the formulas are well suited for correlating fatigue data of spot welds.
3. A strain gauge technique is developed for tensile-shear spot welds.
LSPs at a lap spot weld are determined by two outer strain gauges.
the technique contributes to non-destructive testing of spot welds.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

26

Limitations
The results presented are mainly valid for brittle fracture, high-cycle
fatigue and high-speed crash of spot welds.
For ultimate failure, low-cycle fatigue and low-speed crash, the results
are hardly applicable although they not necessarily always fail.
The material heterogeneity, residual stress and defects due to welding
are not considered.
The results are primarily linear solutions and large deformation, finite
strain and large plasticity are not considered.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

27

Motivation and Objective

Motivation:
Automotive structures with a large number of spot welds
need simplified finite element modeling of spot welds.
A uniform spot weld model is desired for both NVH and fatigue simulations.
Objective:
Development of a simplified spot weld model capable of
delivering reliable natural frequency and modes

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

28

Test Coupons and Compoments

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

29

Coupon Size and Distribution of Spot Welds

a) 320 x 20 x 2

b) 320 x 20 x 2

c) 320 x 80 x 2

d) 320 x 80 x 2

length x width x thickness

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

30

Simplified Spot Weld Model

Beam element

z
y
x

Rigid bar element (RBE2)

Independent nodes
Dependent nodes

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

31

Comparison Test vs. Simulation

Mode Type
1
2
3
4
5
6
7
8

sym
torsion
asym
horizontal (y)
torsion
sym
torsion
asym

Exper.
90,62
278,12
478,75
860,00

FE- FEM/Exp
Calcul.
88,24
0,974
233,23
279,40
1,005
342,39
443,53
457,93
0,957
702,44
867,77
1,009

M ode
1
2
3
4
5
6
7
8

Type
sym
torsion
asym
torsion
sym
torsion
asym
torsion

Exper. FE-Calcul.
99,37
266,67
291,25
283,25
457,09
523,70
892,50
-

FEM /Exp
0,965

506,95
860,80
874,61
1040,74

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

0,973
0,968
0,980
-

32

Comparison Test vs. Simulation

M ode Type
1 sym
2 asym
3 sym
4
horizontal (y)
5
asym

Exper.
96,87
281,87
516,20
863,70

FECalcul.
95,92
281,32
509,93
597,05
862,18

FEM /Exp

Mode Type

0,990
0,998
0,988
0,998

1 sym
2 asym
3 sym
4 asym
5 horizontal (y)

Exper.
111,25
267,50
603,70
811,20
-

FE- FEM/Exp
Calcul.
110,31 0,992
266,64 0,997
579,41
0,960
804,23
0,991
934,66
-

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

33

Comparison Test vs. Simulation

Ruber suspension

Mode
1
2
3
4
5
6
7
8
9

Type
sym
(vertical)
sym
(horizontal)
asym
(vertical)
asym
(horizontal)
Ring mode
Torsion
Torsion
Torsion
Ring mode

Resonance
Damping
frequency
in %
in Hz
Experi. Calcul. Experi. Calcul.
145,8
141,3
0,13 0,121
209,2

208,9

0,11

0,157

382,7

371,0

0,38

0,384

526,3

528,4

0,24

0,228

535,5
550,3
557,3
605,7
664,3

536,7
544,3
565,5
614,3
670,5

0,36
0,11
0,06
0,07
0,43

0,372
0,139
0,068
0,105
0,491

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

34

Mass - Spring System for Analysis

Vibration

d 2u
du
Keu FA cos t
M e 2 Re
dt
dt

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

35

Evaluation of Damping Value

Principle of full width at half maximum


for evaluating damping values

d 2u
du
K e u FA cos t
Me
2 Re
dt
dt

FA

uA
2

Re (M e

Ke ) 2

1
f2 f1
d
z2 z1
Q
f0
Damping value

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

Re
Ke M e

36

Comparison of Relative Damping Values


Ruber suspension
Coupon

Measured Sheet
Coupon Node spacing l
damping d thickness length L of basic mode
t
250x20x1
0,112%
1.0
250
138.0
320x20x2 0,043%
2.0
320
178.0
900x30x10 0,0273%
10.0
900
500.0

t
l
L
2

Case

Experiment

Analysis

d1 / d2
d2 / d10
d1 / d10

2,6

2,4

1,6

1,6

4,1

3,8

d1 l1t2
2
d2 l2 t1

l1 L1
l2 L2
2

d2
l2 t10

2
d10 2l10 t2

l2 L2
l10 L10

d1
l1t10
l1 L1

d10 2l10 t12 l10 L10

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

37

Conclusions

A simplified spot weld model is developed for NVH simulations.

The model is validated by the coupon and component tests.

The same model may also be used for fatigue predictions with additional
data recovery.

Automatic generation of the model is supported at least


by preprocessors like ANSA and MEDINA.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

38

Contents
Introduction
Joint Type in Consideration
Principle of SSZ/MSZ Method
Derivation of SSZ/MSZ Stresses
Derivation of Master S-N Curves
Comparison with FEMFAT Standard
Application Example
Weld data for Scenario Analysis
Potentials and Limitations of the Method

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

39

Joint Type in Consideration


Y/T-Joint

Overlap Joint

Butt Joint

Single Fillet

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

40

Principle of SSZ/MSZ Method


t2

s3
s1

s21

GP-force

t21
t1

t1

s22

Line force =
Procedure:
1) Identification of weld line
2) Definition of critical locations
3) Determination of line forces
4) Calculation of structural stresses
5) Estimation of notch stresses

SSZ =

Nodal force (GP-force)


Element length
Line force
Sheet thickness

MSZ = SSZNotch factor


Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

41

Determination of Line Forces Based on Grid-Point


Forces (Nodal Forces)

Averaged over
Elements
Nodes
Averaged
on Nodes
Line forces and moments
Method of Dong,
Battelle

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

42

Line Forces Averaged over Elements


Linearly distributed line force fE to be determined by gridpoint forces FN:

FN,E3

FN,E1

fE
Grid Point

FN,E2

E1
E3

E2

Weld line

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

43

Total GP-Forces Weighted by Element Length


FN,total=FN,E1+FN,E2+FN,E3

l
F
F
1
N,E1 N,total l l
1 2

l
F
F
2
N,E2 N,total l l
1 2
N

l1
E1

l2
E3

E2

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

44

Converting of Forces into Line Forces


FN2,E1

FN1,E1

fN2,E1

fN1,E1
N1

N2

l1

2
f N1,E1 2FN1, E1 FN2, E1
l1

2
f N2,E1 2FN2, E1 FN1, E1
l1

E1

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

45

Validation of Line Force Calculation


A panel of size 30mm x 50mm with different meshes
under a transverse load is analyzed by NASTRAN;
Grid-Point-Forces (Nodal Forces)
along the reference line are evaluated;
Line forces are calculated and compared.
Reference line
30
mm

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

46

Validation of Line Force Calculation


Distorted mesh
Element size of 10 mm
5 mm
2.5 mm

Result:
Line forces averaged over elements
are less sensitive to mesh quality.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

47

Derivation of Structural Stresses according to Zhang


(SSZ)
f3 m3 q

Geometry:
Sheet thickness:
t1, t2
Throat thickness:
a

Joint angle:

Weld angle:
d
Weld penetration rate: sin
Model size:
h1, h2, h3t2

t2

h3

m2
f2

3s3

s 212
s 22

q2
t1

a
d

s1
f1
q1

h2

m1

h1

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

48

Derivation of SSZ (Example: S21)


Simple beam theory:

f 6m
s m b 2
t
t
Effective height:

t2 sin
t t21 a
sin
Line force and moment:

f f3 cos() q3 sin()
m m3 c1 f3 sin() q3 cos()c2 f3 cos() q3 sin()

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

49

Derivation of SSZ (Example: S21)


f 6m
s 21 2
t
t
f cos() q3 sin()
3
t sin
a 2
sin
6m3 c1 f3 sin() q3 cos()c2 f3 cos() q3 sin()

2
t2 sin

a

sin

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

50

Derivation of SSZ (S1, S22 and S3)


Similarly:

t1
t2
a

6 m1 q1 h1
2 tan 2 sin sin
f1

s1
t1
t12
t (1 2) t1

6 m2 q2 h2 2
2 sin 2 tan
f
2
s 22
2
t1
t1

6 m3 q3 h3
f3
s3
t2

a
t2
t1

2 tan 2 sin sin()
2
t2

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

51

MSZ Notch Stresses


Analytic Approximation of Notch Effects in Seam Welds by

t j
k j s j 1 s
j

sj: SSZ stresses


kj: MSZ notch stresses (approximate)
j : Notch radius
ti : Effective thickness
s : Factor

Basis: Creager solution in Fracture Mechanics and its application to welded


joints, see:
[1] Tada, H., Paris, P. and Irwin, G. 1985. The stress analysis of cracks Handbook,
Paris Productions Incorporated and Del Research Corporation.
[2] Zhang, S. 1997. Stress intensities at spot welds. International Journal of Fracture 88, pp. 167-185.
[3] Zhang, S. 2001. Fracture mechanics solutions to spot welds. International Journal of Fracture 112, pp. 247-274.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

52


t j

1
s

k j s j
j

Determination of tj

tj

s1
s2
s21

s22
s3

Y-joint
single fillet
t1

a
t1
t2

Overlap joint

Butt joint

t1

t1
t1

Y-joint
double fillets
t1

t2 sin a t2 sin
sin
t1

a
t
cos

t2

t2

2
2 2

s4
s5
s6
s7

t2
t1
t2

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

53

Determination of s-Factor

t j
k j s j 1 s
j

Y-joint
single fillet

Overlap joint

Butt joint

Y-joint
double fillets

1.0

0.58

0.85

1.0

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

54

Concept of j =0.05t

t j
k j s j 1 s
j

j=0.05t, t takes the following values:

s1
s2
s21
s22
s3
s4
s5
s6
s7

Y-joint
single fillet
t1

Overlap joint

Butt joint

t1

t1
t1

(t1+ t2)/2
t1
t2

(t1+ t2)/2
t1
t2

t2
t2

Y-joint
double fillets
t1

t2
t1
t2

Fictitious notch radius is 5% of sheet thickness (relative value)


instead of 0.05mm (absolute value).
Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

55

Example of Fatigue Test


104

T-Joint (Alu)
Load case:
Load angle: 0

Failure criterion:
2% loss of stiffness (piston stroke)
that corresponds approximately
to crack initiation

5103

103

5102

AppliedForceAmplitudeFa[N]
Experiment

102
104

105

Cycles to Failure N

106

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

107

56

Correlation of Fatigue Test Data


1E+05

Fatigue test data

Y90F0DG25

Fatigue Test: 18 series


Material: AlMg3Mn

Y90F0DG40
Y90F0DG70

Y90F90DG25

1E+04

Y90F90DG40
Y90F90DG70
Y45F0DG90

Different joint types and loads:


Y45F90DG90

1E+03

IAXDG100
IAXDG70

in
1E+02
Appliedload
inNN

IAXVS07

Kraftamplitude

IAXVS15
IBIDG100

IBIDG70
IBIVS07
1E+01
1E+03

1E+05

1E+04

IBIVS15

1E+06

1E+07

Cycles to failure
Schwingspielzahl N

UAXDG100
UBIDG100

1E+05
Y90F0DG25
Y90F0DG40
Y90F0DG70
Y90F90DG25
Y90F90DG40
Y90F90DG70
Y45F0DG90
Y45F90DG90
IAXDG100
IAXDG70
IAXVS07
IAXVS15
IBIDG100

Psurvival =
90% 50%10%

1E+04

Different weld geometries:

1E+03

1E+02
MSZStressinMPa

Correlated by MSZ stress

MSZ-KerbspannungsamplitudeinMPa

1E+01
1E+03

1E+04

1E+05
Schwingspielzahl N

1E+06

Cycles to failure

1E+07

IBIDG70
IBIVS07
IBIVS15
UAXDG100
UBIDG100

MSZ Stress in MPa

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

57

Master S-N Curves


1E+05

Master S-N Curve


for Aluminium Alloys:

1E+04

1E+03

MSZinMPa
1E+02

MSZ-KerbspannungsamplitudeinMPa
1E+01
1E+03

1E+04

1E+06

1E+07

1E+06

1E+07

Schwingspielzahl N

1E+05

Master S-N Curve


for Steels:

1E+05

1E+04

1E+03

MSZinMPa
1E+02

MSZ-KerbspannungsamplitudeinMPa
1E+01
1E+03

1E+04

1E+05

Schwingspielzahl N

Cycles to failure

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

58

Two New Modules in Principle


SSZ-Method:
SSZ Stresses

Notch Factors
from FEMFAT Standard

Notch Factors
from MSZ-Approximation

S-N Curves
from FEMFAT Standard

Master S-N Curves


(User Defined)

MSZ-Method:
SSZ Stresses

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

59

Convention of SSZ in Seam Welds

Stress Ratio for


Determination of
Notch Stress*

Stresses for Damage


Calculations

T/Y-Joint, Single
Fillet

S1/S4; S3/S6; S22/S5

S1; S3; S21


(S6*); S22

T/Y-Joint,
Double Fillets

S1/S8; S3/S7; S5/S9

S1; S3; S5; S7

Overlap Joint
Butt Joint

S1/S4; S3/S21; S22/S5 S1; S3; S21; S22


S1/S2; S3/S4

S1; S2; S3; S4

* Only for SSZ Method


Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

60

Comparison of Methods in FEMFAT


FEMFAT 4.7

SSZ-Method

MSZ-Method

Type of Stresses

Element stresses

Structural stresses derived


from grid-point forces

Applicable Stress
Components

, ||,

Determination of Notch
Factors

Damage Calculation

Notch factors from weld database; Membrane


and bending stress will be determined using
the stress ratio of the top and bottom side of
shell elements

Analytically estimated
notch factors

Compare notch stresses with joint-typedependent S-N curves from weld database

Compare notch stresses


with a Master S-N curve

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

61

Application Example on Coupon Level


Comparison of the assessment methods
100.000

SSZ-Method
MSZ-Method

10.000

FEMFAT 4.7

1.000

0.100

Damage[-]
0.010

0.001

0.000
Joint
Joint
Joint
TY/Joint
doubleY- LapJointButtJointTY/Joint
doubleY- LapJointButtJointTY/Joint
doubleY- LapJointButtJoint
middle of weld
N5

weld start
N2

weld end
N11

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

62

Scenario Analysis (Y-Joint, Single Fillet)


Good, Normal, Bad Weld Quality
Parameter:
a-Value: a (t1 t2 ) / 2
Weld angle:
Weld penetration:
d
sin
t2

Default Value:
0,6
0,5
0,3

Normal Weld Quality

Extreme Values:
0,5 0,707
0,4 0,6
0,1 0,5
Bad Weld Quality Good Weld Quality

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

63

Scenario Analysis (Y-Joint, Double Fillets)


Good, Normal, Bad Weld Quality
Parameter:
a-Value 1: a1 1 (t1 t2 ) / 2
a-Value 2: a2 2 (t1 t2 ) / 2
Weld angle 1:
1 1
Weld angle 2: 2 2

Default Value:
1 0,6
2 0,6
1 0,5
2 0,5

Normal Weld Quality

Extreme Values:
0,5 1 0,707
0,5 2 0,707
0,4 1 0,6
0,4 2 0,6
Bad Weld Quality Good Weld Quality

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

64

Scenario Analysis (Overlap Joint)


Good, Normal, Bad Weld Quality
Parameter:
a-Value:
a t2
Weld angle:

Weld penetration:

Default Value:
0,6
45
0

Extreme Values:
0,5 0,707
3045
0,25 0,25

a t2 cos
d

t2

Normal Weld Quality

Bad Weld Quality Good Weld Quality

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

65

Scenario Analysis (Butt Joint)


Good, Normal, Bad Weld Quality
Parameter:
Extreme Values:
Default Value:
1,0
Weld width:
b t1
0,8 2,0
Weld reinforcement: e 1t1
1 0
0,2 1 0,2
Sheet offset:
v 2 (t2 t1 ) / 2 2 1,0 (1,0 2 1,0) *
Toe cut depth:
c1 k1t1
k1 0,1
0,2 k1 0
Root cut depth:
0,2 k2 0,1
c2 k2t1
k2 0,1
Toe cur depth:
c3 k3t2
k3 0,1
0,2 k3 0
Root cut depth:
0,2 k4 0,1
c4 k4t2
k4 0,1
* not a weld quality parameter
Normal Weld Quality

Bad Weld Quality Good Weld Quality

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

66

Potentials and Limitations of the Method


Potentials:
The SSZ/MSZ method allows weld geometry to be considered
explicitly in fatigue assessment.
Failure locations, notably, at weld root and toe, are clearly
indicated by the SSZ/MSZ stresses.
Weld quality can be evaluated by scenario analysis in three
classes of good, normal and bad weld quality.

Limitations:
MSZ notch stresses are only approximations.
SSZ/MSZ method is still not sufficiently validated in FEMFAT.
The relation between weld quality and weld geometry should be
further verified by experiment or experience.

Source: SAE Papers 2001-01-0432 and 2005-01-0905, Detroit, USA; FEMFAT User Meeting, May 9-11, 2007, Steyr, Austria

67

Das könnte Ihnen auch gefallen