AM 11/03 1

VI Classifications of Weldments

Good Bad

AM 11/03 2

Outline

nn TWI Classification systemTWI Classification systemnn AISC classification systemAISC classification systemnn An alternative classification systemAn alternative classification system

AM 11/03 3

B - Longitudinal butt

AM 11/03 4

C - longit. Butt w/ Reinforcement

AM 11/03 5

C - Transverse Butt, Machined

AM 11/03 6

D - Long. Butt w/ Start-stop

AM 11/03 7

D - Butt weld w/ good toe

AM 11/03 8

E - Butt weld w/ bad toe

AM 11/03 9

F2 - Trans. butt in rolled section

AM 11/03 10

F - Trans. Butt w/ backing strip

AM 11/03 11

F - Attachments on plate face

AM 11/03 12

F - Groove welded cruciform

AM 11/03 13

G - Attachments near edge

AM 11/03 14

F2 - Load carrying fillet weld

AM 11/03 15

W - Fillet weld metal

AM 11/03 16

TWI - Classification system

AM 11/03 17

Application of TWI system

AM 11/03 18

Application of TWI system

AM 11/03 19

TWI rules

AM 11/03 20

Outline

nn TWI Classification systemTWI Classification system

nn AISC classification systemAISC classification systemnn An alternative classification systemAn alternative classification system

AM 11/03 21

10 810 710 610 510 41

10

100

Fatigue Life, N (cycles)

A

B

C

D

E

F

Weld Category

AISC classification system

AM 11/03 22

10 710 610 510 410 310 210 0

10 1

10 2

10 3

run-outsfailures

Fatigue life, N (cycles)

S = 239.55 * N^-0.15296, r = -0.50264, s = 0.13542

logN = 8.1064 - 1.6518*logS, r = -0.45032, s = 0.44503

99% survival with 50% confidence

99% survival with 95% confidence

AISC category B and C

The good welds!

AM 11/03 23

10 710 610 510 410 310 0

10 1

10 2

10 3

failuresrun outs

Fatigue life, N (cycles)

S = 159.20 * N^-0.16648, r^2 = 0.173, s = 0.1718

logN = 6.8974 - 1.03751*logS, r = -0.2387, s = 0.4293

99% survival with 50% confidence

99% survival with 95% confidence

AISC category D and E

The bad welds!

AM 11/03 24

10 710 610 510 410 310 0

10 1

10 2

10 3

A-S99%

B-S99%C-S99%D-S99%

E-S99%F-S99%

Fatigue life, N (cycles)

A

C

B

D

E

F

99% survival with 95% confidence

AISC category best fit lines to data

AM 11/03 25

6.3 A Suggestion for the Classification of Weldments

Good Bad

AM 11/03 26

Ripple

Toe

Root

Termination

Weld stress concentrations

Personal view

AM 11/03 27

4

6

8

10

30

50

70

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

Fatig

ue L

ife, N

(cyc

les)

Nominal Stress Range, ∆S (MPa)

Ripple

AM 11/03 28

Groove welded butt joints

4

6

8

10

30

50

70

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

Fatig

ue L

ife, N

(cyc

les)

Nominal Stress Range, ∆S (MPa)

AM 11/03 29

Non-load-carrying fillet welds

4

6

8

10

30

50

70

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

Fat

igue

Lif

e, N

(cyc

les)

Nominal Stress Range, ∆S (MPa)

AM 11/03 30

Transverse attachments

4

6

8

10

30

50

70

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

Fat

igue

Life

, N (c

ycle

s)

Nominal Stress Range, ∆S (MPa)

AM 11/03 31

Longitudinal attachments

4

6

8

10

30

50

70

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

Fatig

ue L

ife, N

(cyc

les)

Nominal Stress Range, ∆S (MPa)

AM 11/03 32

40

60

80

100

300

500

105 106 107

RippleGroove WeldsNon-Load-Carrying Fillet WeldsLoad-Carrying Fillet WeldsTerminations

∆S

(MPa

)

Fatigue Life, N (cycles)

Good, bad, maverick

AM 11/03 33

10

100

105

106

107

Fatigue Life, N (cycles)

Nom

inal

Str

ess

Ran

ge, S

(ksi

)

Terminations

Partial Penetration Load-Carrying FilletTermination

RippleNon-Load-Carrying Fillet

Transverse Groove Welded Butt

Good

Bad

Maverick

Nom

inal

Str

ess

Ran

ge, S

(MP

a)

500

50

100

Good welds, Bad welds, Mavericks

AM 11/03 34

C

Detail #3Kf = 1.87

²S design = 27.0 ksi.

C Full-Penetration GrooveWeld: Detail #10

Kf =2.12²S design = 23.3 ksi.

C

Non-Load Carrying FilletWeld: Detail #25

Kf =2.23²S design = 22.1 ksi.

The good welds: weld toe failures

AM 11/03 35

C,CsLoad Carrying Fillet Weld:

Detail #20Kf =3.12

²S design = 17.5 ksi.

C

Weld Termination:Detail #30Kf =3.27

²S design = 14.5 ksi.

The bad welds: terminations

AM 11/03 36

Fatigue severity of terminations

n Starts and stops introduce weld discontinuities.n Residual stresses very high.n 3-D stress concentrations effects.

C

AM 11/03 37

Examples of terminations

C

C

a

b

C

L-PL-P

LP

C

CC

AM 11/03 38

C

P• Partial penetration

weldment. Theamount of penetrationis generally unknown.

C, Cs

C, Cs

• Undercut. Failureoccurs at the undercuton wrap-around weldwhere the arris of theplate is melted. theamount of theundercut is generallyunknown.

The mavericks: something undefined

AM 11/03 39

C

C

• Multiple failure sites.The stress distributionin the joint changes asfatigue cracks initiateand grow at variouslocations.

C

L

P

L-PL-P

• Local stresses uncertainbecause of structuralredundancy

The mavericks: complex components

AM 11/03 40

Summaryn Viewed from the perspective of crack

initiation site, weldments could be grouped into four categories: ripple, toe, root, and terminations.

n This classification scheme leads to the conclusion that there are “Good”and “Bad” weldments and “Mavericks” - weldments for which a critical piece of information is unavailable and hence their behavior is undefined.