VII Improving the fatigue life of weldments

AM 11/03

Outline
n
n n n n

Obvious things to do
Problems the weld toe Fatigue life Improvement Strategies Light and heavy industry weldments Improving the bad weldments

AM 11/03

Crude! (bad)

AM 11/03

Better

AM 11/03

Bad - planar weld discontinuities

AM 11/03

Outline
n

Obvious things to do

n
n n n

Problems the weld toe

Fatigue life Improvement Strategies Light and heavy industry weldments Improving the bad weldments

AM 11/03

Weld toe is a stress concentration

AM 11/03

Slag entrapments at toe?

Virtually eliminates fatigue crack initiation life NI

AM 11/03

Cold-lap defects at weld toe

Nominal

Perfect

Welding procedure A

Welding procedure B

Cold laps virtually eliminate the fatigue crack initiation life (NI)
AM 11/03

Such weldments may have an appreciable fatigue crack initiation life (NI)
9

Cold lap at a weld toe

r

Weld Metal

Weld Toe Location Without Cold-Lap Defect

Heat Affected Zone

Curved Path

Base Metal
Vertical Path

Loading Direction

AM 11/03

10

Effect of cold lap depth

10
K Weld geometry correction factor, MK

2-D FEM

D = 0.0 mm, MK CG, curved path from weld toe location D = 1.0 mm, MK CG, curved path from cold-lap defect D = 1.0 mm, MK CG, vertical path from cold-lap defect

0.001

0.01 0.1 Crack length / main plate thickness, (a/T)

11

AM 11/03

Effect of flank angle

10 2-D FEM
K

Weld geometry correction factor, MK

= 30o = 45o = 60o

D = 1.0 mm

0.001

0.01 0.1 Crack length / main plate thickness, (a/T)

12

AM 11/03

Effect of cold root radius

10

Weld geometry correction factor, MK

D = 0.00 mm, weld toe location D = 2.00 mm, r cl = 0.146 mm D = 2.00 mm, r cl = 0.025 mm

0.001

0.01

0.1

Crack length / main plate thickness, (a/T)

AM 11/03 13

Recent study on rail welds

Toe Radius, R Flank Angle,

Weld Reinforcement Height, H

Base Metal Thickness, Tb Weld Collar Width, W

AM 11/03

14

Geometric Parameters

Fin Length, Lf Flank Angle,

Fin Thickness, Tf

Cold Lap Length, Lcl

AM 11/03

15

Weld with a Fin and a Cold Lap

AM 11/03

16

Deformed Shape

AM 11/03

17

Nominal Weld Geometry

4.5 4

R=.5mm

3.5

R=1mm R=3mm R=6mm

2.5

1.5

1 0 10 20 30 40 50 60 70 80 90

Flank Angle, (Deg.)

AM 11/03

18

Fin length
7

Elastic Stress Concentration Factor, Kt (-)

3 Fin Thickness = .5mm Fin Thickness=1mm 2 Fin Thickness=2mm

1 0 2 4 6 8 10 12 14

Fin Length, Lf (mm)

AM 11/03

19

Fins and Cold Laps

7

Cold Lap; NO Fin

Cold Lap with Fin 0.5 mm Thick Cold Lap with Fin 1 mm Thick Cold Lap with Fin 2 mm Thick

For All Tests R=3mm and =30

Defect Free -4 -2 0 2 4 6 8

Cold Lap Length, L (mm) cl

AM 11/03

20

Predicted effect of SuBM

60 Fatigue Strength at 1E+07 cycles, Sa (ksi) Hot Rolled Steel 50 40 30 Over Stressed Q & T Steel

Trends in Ideal 1.0-in plate thickness, non-load carrying cruciform weldments fatigue strength. R=0 Welding residual stresses = 50% of SYBM Sfab SYBM

Stress Relieved As Welded, S = 0 fab

20 10

As Welded, S = SyBM fab 0 50 100 150 200

Base Metal Ultimate Strength, Su (ksi) BM

AM 11/03

21

Outline
n n

Obvious things to do Problems the weld toe

Fatigue life Improvement Strategies

Light and heavy industry weldments Improving the bad weldments

n n

AM 11/03

22

Good - grind off reinforcement

AM 11/03

23

Good - burr grind weld toe

AM 11/03

24

Very good - full face grinding

AM 11/03

25

Shot peened weld toe

AM 11/03

26

Remelted weld toe (laser)

AM 11/03

27

Improvement Strategies

TWI suggestions as to weld improvement procedures

AM 11/03

28

Improvement strategies
100 Fatigue Strength (ksi) at 106 cycles Shot Peen Residual stress
15 30 45 15 30 45

Over Stressed plate Plain Stress Relieved

10 Geometry

15 30 45

As Welded

Base metal strength

1
AM 11/03

100 Base metal UTS (ksi)

1000
29

Example

IP model predictions and experimental data.

AM 11/03

30

ASTM A 36 butt weldment

AM 11/03

31

ASTM A 514 butt weldment

AM 11/03

32

Outline
n n n

Obvious things to do Problems the weld toe Fatigue life Improvement Strategies

Light and heavy industry weldments

Improving the bad weldments

AM 11/03

33

Light, heavy industry weldments

Light industry weldments are presumed to be fabricated from 1/2 or smaller plate and not to have large fabrication stresses. Heavy industry weldments are presumed to be fabricated from larger than 1 thick plates and to possess large fabrication stresses.

AM 11/03

34

Light industry weldments

100

Remote Stress Range, S (ksi)

Ideal

t = 0.5-in. (12mm); R = 0 Without Fabrication Stresses

Nominal

10

Toe Ground (radius = 0.1 in.) Over Stressed Stress Relieved Weld Profile (flank angle 20 o) As Welded
1

10

10

10

10
T

10

Fatigue Life, N

(cycles)

AM 11/03

35

Heavy industry weldments

100

Remote Stress Range, S (ksi)

t = 2.0-in. (50 mm); R = 0 Ideal Nominal

10

With Fabrication Stresses

Toe Ground (radius = 0.1 in.) Over Stressed Stress Relieved Weld Profile (flank angle 20 As Welded
1 10
4

10

10

10
T

10

Fatigue Life, N
AM 11/03

(cycles)
36

Outline
n n n n

Obvious things to do Problems the weld toe Fatigue life Improvement Strategies Light and heavy industry weldments

Improving the bad weldments

AM 11/03

37

Good weldment
Transverse attachment

Weldment with a transverse attachment

AM 11/03

38

Bad weldment
Longitudinal attachment

Weldments with longitudinal attachments have a low fatigue resistance because of the presence of weld terminations. Starts and stops introduce weld discontinuities. Residual stresses very high. 3-D stress concentrations effects
AM 11/03 39

Examples of terminations

C
P

L-P L-P
L

b a

AM 11/03

40

Placement of stress diffuser

Longitudinal attachment

Stress diffuser

Longitudinal attachment with stress diffusers

AM 11/03

41

3-D FEM modeling

AM 11/03

42

Effectiveness of a stress diffuser

Longitudinal attachment
AM 11/03

Longitudinal attachment with stress diffuser

43

Effect on MK and NP
Fatigue Crack Propagation Life, N P (cycles) 107 L.A. L.A. with stress diffusers Transverse attachment

Weld geometry correction factor, MK

4 3

L.A. L.A. with stress diffusers Transverse attachment

106

1 0.001

105

0.01 0.1 Crack length / main plate thickness, (a/T)

0.002 0.02 0.2 Initial crack length / main plate thickness, (a o /T)

AM 11/03

44

Fatigue test results

500 L.A., Series 1 L.A., Series 2 L.A., Series 3 L.A. with stress diffusers, Series 1 L.A. with stress diffusers, Series 3

Remote Stress Range, S (MPa)

300

100 80 60 40

105

106 Fatigue Life, N (cycles)

10 7

AM 11/03

45

Fatigue test results

500

300

Transverse attachments, database Longitudinal attachments, database LA specimens, Procedures 1 and 2 LA specimens, Procedure 3 SD specimens, Procedure 1 SD specimens, Procedure 3

100 80 60 40

105

106

107

Fatigue Life, N (cycles)

AM 11/03

46

Summary
The fatigue strength of Ideal weldments can be much improved; whereas, the fatigue strength of Nominal weldments cannot. Weld toe grinding or weld profile control works best for Ideal weldments at short lives. Beware of corrosion pitting. Smaller Ideal weldments are more susceptible to improvement than larger weldments. Fabrication stresses are critically important.

AM 11/03

47

Summary
The behaviors of light and heavy industry weldments are dissimilar. Stress relief annealing and over-stressing works best for Ideal weldment at long lives. Beware of compressive overloads. Fatigue behavior of weldments and effective life improvement methods depends upon weldment size and weld quality Stress-diffuser can substantially improve the fatigue life of terminations without post-weld processing.

AM 11/03

48