SIGNIFICANCE

AND
CONTROL OF LAMELLAR
TEARING OF STEEL PLATE
IN THE SHIPBUILDING
INDUSTRY

r.

.+

This
document
hasbeanapproved
releasa
andsala;
its
fOrpublic
distribution
isunlimitad.

SHIP STRUCTURE
1979

COMMITTEE

SSC-290
FINAL REPORT
on
Project SR-1Z50
Lamel Tar Tearing

SIGNIFICANCE

AND CONTROL OF LAMELLAR

TEARING OF STEEL PLATE IN THE

SHIPBUILDING

INDUSTRY

by

J. Sommella
Gibbs & Cox, Inc.
under
Department of Transportation
U. S. Coast Guard
Contract No. DOT-CG-74355-A

TWS

docwwnt
sale;

has hen approved for pub lie w lease and

its distribution is unlimited.

U. S. Coast Guard Headquarters

!Jashington, D.C..
1979

Ad&es
Cormpndence
to:

M&r Agenci~:
Guwd
Umtish& Ccmt

S92retoiy,
Ship
structure
Colmnittes
U).Coxt
Guad
Hmdqumtew(G-A4W

Naval
% S@ems
Command
Militury
S9ditl
Conmland
Afmhne
A~ffm
united
sfote3
G13&imi
Swvwy
Ammwn
Burwu
of
tip~g

W&@on,D.C.
20599

AnInteragency
Advisory
Commitk
Dedicatal
toImproving
theSbucture
ofShips
SR-1250
JULY
1979
TheShip
Structure
Corrnnittee
recognized
theneed
of
evaluating
available
information
onpreventing
larellar
tearing
inm3rine
stmctures.
While
theincidenee
oflamellar
tearing
inship
structures
hasbeen
lW,itismore
common
inmobile
andfixed
platforms
ofthetype
used
intheoffshore
mineral
Itsoccumen=
results
ekplorat
ionandproduction
industry.
in03s
tlyrepairs
,and,
insome
instances,
fabrication
delays.
A project
wasundertaken
to develop
a guide
describing
thefactors
which
contribute
toandinfluence
lamellar
tearing,
a review
oftheprocedures
used
todetermine
susceptibility
to
lamellar
tearing,
andmthods
forpost-welding
detection
and
repair
oflamellar
tears.Theresults
ofthis
effo~t
are
contained
inthis
report.
Anycomments
OFrequests
foradditional
copies
arewelmne.
&&Z@~He
Rear
Admiral.
U.S.Coast
Guard
Chairman,
Ship
?Lructure
(hnnittee

.._- --

.-..

. ....

Technical
Report
Documentation
Poge
1.ReportNo.

3. Rectp!ents
CatalogNo.

SSC-290
4. Title apdSubt,
tle
+--

TVACCCST
~
-

I 5, RtportDo~o

SIGNIFICANCE
ANDCONTROL
OFLAMELLAR
TEARING
OFSTEELPLATEINTHESHIPBUILDING
INDUSTRY

k-II

8. Performing
O,gonization
ReportNo.
I

J. SOMMELLA
.

18521(1-146)

9. Perlorm
IrBg
Organ
izottonNomeandAddress

10. werkun,t N6, (T RAIS)

GIBBS& COX,INC.
40Rector
Street
NewYork,N.Y.10006

11. Controct
orGrantNo.

SR-1250

13.
Type01 Report
andPeriodCovered

Agent Name andAddress

12. Sponsoring

ShipS{ructure
Committee
Office
ofMerchant
Marine
Safety
----U.S.CoastGuardHeadqua~ters
Washington,
D.C.
20590

15, Supplementary
Not,,

16.Ah.?rac?

Lamellar
tearing
isa separation
intheparent
orbasemetal
caused
by
through-thickness
strains.
These
strains
areusually
induced
byweld
metal
shrinkage
under
conditions
ofhighrestraint.
Thismanual
provides
specific
recommendations
forcontrolling
lamellar
tearing
inthetypes
ofsteels
used
intheconstruction
ofshiFs
andoffshore
platforms.
A brief
description
ofthecharacteristics
andmechanism
ofIamellar
tearing
shows
thatforIamellar
tearing
tooccurthere
mustbea critical
combination
ofmaterial
susceptibility,
andwelding
procedures
andjointdesign
which
permit
thedevelopment
ofhighthrough-thickness
strains.
Tee(T)and
corner
joints,
used
extensively
inshipsandoffshore
structures,
arethe
twobasic
joint
configurations
mostsusceptible
tolamellar
tearing.
However,
theincidence
oflamellar
tearing
hasbeenextremely
rareinshipbuilding.
Theproblem
oflamellar
tearing
isconsiderably
moresignificant
innmbile
andfixed
offshore
drilling
platforms
whichusethick
plates
in
highly
restrained
T-and
cruciform
joints.
Thefactors
whichcontribute
toandinfluence
lamellar
tearing
aregrouped
intothree
.categories:
joint
design
, material
selection
andfabrication
procedures.
Foreachparameter
recommendations
arepresented
forreducing
theriskoflamellar
tearing.
Inquiries
madetothemajor
shipclassificationsocieties
indicate
thatthemostsuccessful
andcost-effective
method
ofpreventing
lamellar
tearing
istheuseofsteels
withimproved
throughthickness
(Z-direction)
properties
atsusceptible
connections.
Nethods
fortheposr-welding
detection
andrepair
oflamellar
tears
are
reviewed
asarethetestprocedures
developed
todatefordetermining
the
susceptibility
ofsteel
plates
tola!wllar
tearing.

17, KeyWords

18. Dlstribu!ion
Statement

Document
isavailable
tothepublic
LAMELLAR
TEARING SHIPBUILDING
liJELDING
theNational
Technical
ULTRASONIC
TESTINGthrough
Information
Service,
Springfield,
VA 22161
(UT)
STEELS
.1
.
19. Security
CIUSSII.
(o(thisrepe,!)

UNCLASSIFIED

FormDOT F 1700.7
[8-72)

20. Security
ClasSIl,(ofthispogc)

UNCLASSIFIED
Reproduction
ofcomplctcd
pogeouthorized

-ii-

21.No.of~ages

70

22. Price

SHIPSTRUCTURE
COMMITTEE
STRUCTURE
COMMITTEE
isconstituted
toprosecute
a research
TheSHIP
program
toimprove
thehullstructures
ofships
andother
marine
structures
byanextension
ofknowledge
pertaining
todesign,
materials
andmethods
of
construction.
RADMH.H.Bell (Chairman)
Mr.M.Pitkin
Chief,
Office
ofMerchant
Marine Assistant
Administrator
for
Safety
Commercial
Development
U.S.Coast
Guard
Headquarters Maritime
Administration

14r.
P.M.Palermo
Assistant
forStructures
Naval
Ship
Engineering
Center
Naval
SeaSystems
Command

Mr.R.B.Kzahl
Chief,
Branch
ofMarine
Oiland
GasOperations
U.S.Geological
Survey

Mr.W.N.Hannan
VicePresident
American
Bureau
ofShipping

Mr.C.J.whitestone
Chief
Engineer
Military
Sealift
Command

LCDR
T.H.Robinson,
U.S.Coast
Guard
(Secretary)
SHIP
STRUCTURE
SUBCOMMITTEE
TheSHIP
STRUCTURE
SUBCOMMITTEE
actsfortheShipStructure
Committee
ontechnical
matters
byproviding
technical
coordination
forthe
determination
ofgoals
andobjectives
oftheprogram,
andbyevaluating
and
interpreting
theresults
interms
ofstructural
design,
construction
and
operation.
U.S.COAST
GUARD
MILITARY
SEALIFT
COh!MAND
Cdr.J.
C.Card
Mr.T.W.Chapman
LcdrS.H.Davis
Mr.A.B.Stavovy
CaptC.B.Glass
Mr.D.Stein
Dr.W.C.Dietz
Mr.J.Torresen
NAVAL
SEASYSTEMS
COMMAND
AMERICAN
BUREAU
OFSHIPPING
Mr.R.Chiu
Dr.H.Y,Jan
Mr.D.Liu
Mr.R.Johnson
Mr.I.L.Stern
Mr.G.Sorkin
(Chairman)
Mr.J:B.OBrien
(Contracts
Admin.)Mr.S.G.Stiansen
U.S.GEOLOGICAL
SURVEY
Mr.R.Giangerelli
Mr.J.Gregory
INTERNATIONAL
SHIPSTRUCTURES
CONGRESS
Prof.
J.H.Evans
- Liaison
AMliRICAN
IRON
& STEEL
INSTITUTE
Mr.R.H.Sterne
- Liaison
STATE
UNIV.
OFNEWYORKMARTTTME
COLLEGE
~r.
W.R.Porter
Liaison
SOCIETY
0)?
NAVAL
ARCHITECTS
&
MARIKE
ENGINEERS
U.S.COAST
GUARD
ACADEMY
Mr.A.B.StavovyLiaison
Capt
W.C.Nolan
- Liaison
WELDING
RESEARCH
COUNCIL
U.S.NAVAL
ACADEMY
Dr.R.Ba.ttacharyya
- Liaison
Mr.K.H.Koopman
- Liaison
U.S.NERCHANT
MARINE
ACADEMY
Dr.Chin-Be.
Kim- Liaison

MARITIME
ADMINISTRATION
Mr.F.J..
Dashnaw
Mr.N.O.Hammer
Mr.F.Seibold
Mr.M.Touma
NATIONAL
ACADEMY
OFSCIENCES
SHIP
RESL4RCH
COMMITTEE
Mr.O.H.Oakley
- Liaison
Mr.R.W.Rumke
- Liaison

-iii-

METR!CCONVERSIONFACTORS
.. . -

llp~raximate
Conuersians
to MetricMeasures
symbol

When
youknow

Mul!ipl~
by

ToFind

h
yd
!,

2.5

mches
Ieet

30
0.9
1.6

..
=-

yards
m,ltis

centtnml.rs
centrmw?rs
rim
1.,,
k,hm,w,rs

m,
cm

,,,
h

AFIEA

m
(12

II,,
J

.,,,,., ,,,.1,..s
.<,,,.,,.fklul

6.5
0.09

ln~
d

Approximate
Conversions
fromMel!icMeasures
Swlrhal When
YouKnow

symttol

LE~GTH
I,,

.
.
.
.
.

rnillimemrs
ce.timctecs
maers
mm,,
kilmlulc,
s

nw
ml
m
m
km

.,IU.,, L. ,11,h,,.,

.,,,,.,,1. .,,,1.
.Ip,l,
r 1,1!
1,...
.,,,!..,

Ihm-
b,,

311,.1
1,,,15
i20m3
110

gaI
f(j
vd3

W,S,,0LW5
1.,1
!14!%,.1s
Ill,,d,4,VXS
U,I,S
P,llts
quarts
oallons
cubic{eel
cubicyards

0.95
3,e

0.76

cubicnml=s

0,47

0.03

cub,c
mews

,,,!
mi
,(d
1
t
I
I
n?
~3

TEMPERATURE
[CXSCI)
Fahrmha
it
lempsraom

5/9 (aitel
subtract
ig
321

,Wwd.
.I!<I,L
1,,.

.--. s
=~

.
Mull
,t
,1,!,
s
m,ll,
tlh.,s
,,,,11,1,0,(
s
I,wrs
IImm
k,WS
Iumrs

5
1s
30
0.24

d~
m,J

VOLUME
Im,ll,
liw,s
1,1,,,
s
1,1,,,
s
I,mr,
C,dmc
.*1.,5
Cuticmeters

ITd
I
I
I
m3
~3

U,03
7.1
I.(W
0.26
35
1.3

ll,lld*,11314
U,.
,I,,lu.,

41,,,

w,,f~
,.

,,1
{,.,I
I1~
W13

!,,,IIW,
.,,1,,,.i..,,,
cub,,y,,tds

TEMPERATUff
E (exact]
Celsius
temperat.
re

c
Celsms
temp-sralum

.,
Ii

,,,,! 4..

!1
h$l
1

0.45
0.9

VOLUME

I
4

In

MASS [wcigfrl}

2s

+M,,,d
s

3.3
1.1
0.6

AREA

MASS
{weight)
,,,,...s

0.04
0.4

0.8
w,,.,,.
r,,d%
+.,,,,,,,,,
lm,
b!s
7.6
,,.,,,
s
o.
a

symbo

roFind

LENGTH

s!

. - 5

.
2

Multioly
by

9/6 (Ihe.
add32)

.1 ,. z 2.54$etilcllv>.
Fzxother
enact
Cm*vMsr.,ls
andmredn
la,ledmbl
es,seeNOShi,
s.. h,bl.236,
urmt,
ofWe,ah!s
tmdLkes=
es. PIIce$2.25,
SDCalalwNCJ.
C13.1O:2E5.

F
-40
t;
$:

0
*
1
-20

98.6
32
160
L20
80
40
,
1 1 1 $ 1 1 1 1
1 I
00
60
40
20
D
37

F
212
2W
4
100
Oc

,,1

TABLEOFCONTENTS
PAGENO.

SECTION

vii- ix

GLOSSARY
LISTOFABBREVIATIONS

ix

THEWELDING
INSTITUTE
OFGREATBRITAIN
LETTER
OFRELEASE
1

AUSTRALIAN
WELDING
RESEARCH
ASSOCIATION
INTRODUCTION

DeSCriptiOn
OF LAMELLAR
TEARING
(LT)
2.1
2.2
2.3
2.4
2.5

WhatIsLamellar
Tearing?
WhereDoes
Lamellar
Tearing
Occur?
WhatIsTheExtent
ofTearing?
WhatDoesA Lamellar
TearLookLike?
HowIsLamellar
Tearing
Normally
Observed
Or Detected?

FACTORS
CONTRIBUTING
TOLAMELLAR
TEARING
3.1 UnderWhatConditions
DoesLamellar
Tearing
Occur?
3.2 ByWhatMeansDoesLamellar
Tearing
Occur?
3.3 Design
Factors
3.4 Material
Factors
3.5 Fabrication
Factors

OCCURRENCE
OF LAMELLAR
TEARING
4.1 WhatTypesofStructures
AreSusceptible
To Lamellar
Tearing?
4.2 HowOftenDoesLamellar
Tearing
Occur?
4.3 WhenDoesLamellar
Tearing
Occur?
4.4 WhatTypesOf Steel
AreSusceptible
to
Lamellar
Tearing?
4.5 WhatSteelThicknesses
AreSusceptible
to
Lamellar
Tearing?

SIGNIFICANCE
OFLAMELLAR
TEARING

xi
I-l&1-2
2-1- 2-3
2-1
2-1
2-2
2-2
2-3
3-1- 3-7
3-1
3-1
3-2
34
3-6
4-1- 4-lo
4-1
4-8
4-8
4-9
4-10
5-1& 5-2

5.1 Static
LoadCondition
5.2 Dynamic
LoadCondition
5.3 Fatigue

5-1
5-1
5-2

CONTROL
OF LAMELLAR
TEARING

6-1- 6-16
6-1

6.1 JointDesign

6.1.1Avoidance
ofExcessive
Through-Thickness
6-1
Strains
6.1.2Reduction
ofJointRestraint
6-3
6-5
6.1.3Reduction
ofComponent
Restraint
6-6
6.1.4Selection
ofWeldMaterial
-v-

(Centd)
TABLEOFCONTENTS
TITLE

SECTION

6.2 Material
Selection
6.2.1GradeofSteel
6.2.2Product
Type
ofSusceptible
6.3.1 LayoutandForming

Components
6.3.2Welding
Process
Fit-up
andJigging
6.3.3JointPreparation,
6.3.4Welding
Conditions
6.3.4.1Preheat
6.3.4.2Deposition
Rate
6.3.4.3Interpass
Temperature
6.3.4.4WeldSizeandShape
6.3.5
Welding
Techniques
6.3.5.1RunSequence
6.3.5.2Buttering
6.3.5.3__
InSitu
-Buttering
6.3.5.4Peenlng
6.3.5.5Welding
andFabrication
Sequence
6.3.5.6Intermediate
Stress
Relief

6-7
6-7
6-10
6-10

6.3 Fabrication

PAGENO.

DETECTION
ANDREPAIR
OF LAMELLAR
TEARING
AFTER
WELDING
7.1 When1s
Non-Destructive
Testing
ForLamellar
Tearing
Required?
7.2 WhichNon-Destructive
Testing
Methods
Are
Applicable?
7.3 Ultrasonic
Testing
(UT)OfWelded
Joints
7.4 Repair
ofLamellar
Tears

6-10
6-10
6-11
6-11
6-11
6-12
6-12
6-12
6-13
6-13
6-13
6-15
6-15
6-15
6-16
7-1- 7-5
7-1
7-1
7-1
7-3

TESTSFORDETERMINING
THESUSCEPTIBILITY
OF
STEELPLATES
TO LAMELLAR
TEARING

8-1- 8-3

REFERENCES

9-1& 9-2

10

1o-1- 10-3

BIBLIOGRAPHY

APPENDICES
A

MECHANISM
OF LAMELLAR
TEARING

A-1- A77

ORIGINS
OF INCLUSIONS

B-1- B-10

-vi-

GLOSSARY
having
different
mechanical
properties
in
ANISOTROPIC
- notisotropic,
i.e.,
different
directions.
BASEMETAL- thebasic
mill-rolled
material
tobewelded.
existing
duetorigidity
ofthevarious
eleCOMPONENT
RESTRAINT
- restraint
mentsof a jointorconnection.
CONNECTION
- complete
assembly
consisting
of thevarious
jointsmakingup
thetotalunit.
CONSUMABLES
- thefiller
metaladdedinmak
ng a welded
joint- usuallyin
theformofelectrodes
orwelding
rods.
between
thematerial
matrixand
DECOHESION
- separation
along
theinterface
an inclusion.
DISCONTINUITIES
- lackofhomogeneous
characteristics
caused
bynonmetallic
inclusions,
cracks,
tears,
e~c.
DUCTILITY
- ability
of a material
subjected
tostress
to undergo
permanent
deformation
intheplastic
rangepriortorupture.
ELECTRODE
STRENGTH
- usually
theminimum
tensile
strength
ofdeposited
weld
material.
ELECTRODE
MATCHING
- thepractice
ofproviding
electrode
strength
equalto
thebasemetaltensile
strength.
- percentage
elongation
measured
ina standard
tensiontestand
ELONGATION
usedasa measure
ofductility.
HEAT-AFFECTED
ZONE(HAZ)
- portion
ofthebasemetaladjacent
to thefusion
lineoftheweld,whichisnotmelted
butis heatedduringweldingto a
temperature
highenoughto modifythemechanical
properties
of microstructure.
INTERPASS
TEMPERATURE
- in multiple-pass
welds,thetemperature
of the
deposited
weldmetalbefore
thenextpassisstarted.
ISOTROPIC
- having
thesamemechanical
properties
indifferent
directions.

JOINT
- junction
oftwoormorestructural
members
whichareto bejoined;
a
single
element
of a connection.

inthebasematerial
caused
by induced
strains
LAMELLAR
TEARING
- separation
inthethrough-thickness
direction
duetoweldshrinkage.
- largediscontinuity
in rolledsteelproducts
resulting
from
LAMINATION
flattening
andelongating
of inclusions
or voidsduringtherolling
process
- usually
a layer
ofnonmetallic
inclusions.
-vii--

MATRIX
MATERIAL
- themajorcontinuous
substance
of a metalas opposedto
inclusions
orparticles
ofmaterials
having
dissimilar
characteris:lcs.
strength,
yieldstress,
percentage
elongaMECHANICAL
PROPERTIES
- tensile
tion,reduction
of area,
etc.
MEMBER
RESTRAINT
- restraint
in closure
memberwhereinherent
rigidity
requires
weldshrinkage
tobeabsorbed
bythebasemetal.
MULTI-PASS
WELDS- weldsrequiring
morethanonepasstocomplete
deposition
ofrequired
weldmaterial.
particles
ofcompounds
insteel
matrix;
NONMETALLIC
INCLUSIONS
- microscopic
silicates
andaluminum
oxides.
principally
sulfides,
PARENT
METAL- thebasic
mill-rolled
material
tobewelded.
PEENING
- themechanical
working
oftheweldbeadsbymeansof lightimpact
blowstotheweldsurface
toreduce
residual
stress.
PLANAR
DISCONTINUITIES
- discontinuities
having
majordimensions
of length
andbreadth
ina plane,
i.e.,likea flatplate.
PREHEATING
- theapplication
ofheattothebasemetalimmediately
before
welding.
PREPARATION
- geometry
of a jointdetailincluding
theedgebeveliroot
opening,
andbackup.
percentage
reduction
incross-sectional
REDUCTION
OFAREA(RA)- themaximum
areameasured
ina standard
tension
testatthepointofrupture
andused
as a measure
ofductility.
RESTRAINT
- resistance
ofthejointorconnection
tomovement
of anykind.
ROLLING
(orX)DIRECTION
- direction
thathotrolledstructural
material
rolling
direction
for
travels
through
theforming
rolls- ortheprincipal
crossrolled
material.

Transverse
(Y)Direction
/

Rolling
(X)Direction Through,1
Thictiess
(~)
Direction
!

-viii.-

STRAIN
- deformation
perunitoforiginal
lengthcausedby changesin appliedforces.
STRESS
- forceperunitofcross-sectional
area.
THROUGH-THICKNESS
(orZ) DIRECTION
- perpendicular
to theplaneof the
rolled
surface.
TRANSVERSE
(orY)DIRECTION
- perpendicular
totherolling
direction
in the
planeofthematerial.
elements
ofwelding
(u,iually
a written
proWELDING
PROCEDURE
- thedetailed
cedure)
whichdefine
theprocess,
voltage,
current,
speed,
electrode
type
preheat,
sequence
andanyother
andsize,position,
edgepreparation,
related
factors
required
foranacceptable
weld.
WELDING
SEQUENCE
- theorderinwhichweldsaremadein a particular
weldmenttominimize
distortion,
to compensate
forshrinkage
andto reduce
internal
stresses.
ULTIMATE
STRESS
- maximum
stress
attained
before
rupture
ofthematerial.
UTMATERIAL
- material
ultrasonically
inspected
in itsentirety
priorto
fabrication.
YIELDPOINT- thepointon a stress-strain
curvewhereelongation
occurs
withverylittle
increase
instress.

LISTOFABBREVIATIONS
HAZ -

heat-affected
zone

LT

lamellar
tearing

RAz -

percentage
reduction
of areaintheZ direction

UT

ultrasonic
testing

-ix-

THE WELDING INSTITUTE

PLEASEREPLYTO RESEARCHLABORATORY AEINGTON HALL ABINGTON CAMBRIDGE CB1 6AL
Telephone
CAMBRIDGE0223 891162 Telegrams
WELDASERCH
CAMBRIDGETelex81183
LONDONOFFICE 54 PRINCESGATE EXHIBITION ROAD LONDONSW72PG
WELDINSTLONDONS.W.7
Telephone
01-5848556 Telegrams

Ourref: TGD/wH/50

9thJanuary
,1.979.

Mr.J.Sommella,
GibbsandCoxInc.
,
40Rector
Street,
NewYork10006,
USA.

DearMr.SOmJElla,
Dr.Dolbyhasaskedme toreplyto yourletter
of
concerning
thereport
Significance
andControl
Lamellar
Tearing
ofSteelPlatein theShipbuilding
We feelyoushould
be congratulated
on
Industry.
having
drawnthepublished
material
together
verywell
andproduced
a readable
document
whichclears
up a
number
of common
misunderstandings
(e.g.thedistinction
between
a lamination
anda lamell.ar
tear),
andgives
soundadvice
on avoidance
andrepairs
of lamellar
toyouruseof
tearing.We thushaveno objection
someofourmaterial
inyourmanual
initspresent
form.
Yourssincerely,

T.G.DAVEY
Materials
Department.

-x-

+.$

TXALIM WELIMG HESEARCHANOCIATION

lffll I!US
118

+,
.

-,,.. .
,?
~,,... . .,,
. ,, .,,-

AL F~EO ST REET, VI L50NSP0 INT, r

~,W.206i

TELEPHONE: 922-3711

AV:pd397/62

8thDecember,
1978
Mr J. Sonunella
GibbsG CoxInc.
40Rector
Street
NEWYORK N.Y. 10006
DearMrSommella

18521- Significance
andControl
ofLamellar
Tearing
ofSteelPlates
intheShipbuilding
Industry
SSCProject
No.SR-250
Thankyouforthecopyof theabovedocument
whichI perused
with

interest.
I am happyto giveyouformalauthorisztion
to include
thematerial
reproduced
fromAWRATechnical
Note610nbe~alf?
of theAssociation.

Yourssincerely,
:- .

L x%+
/

DrA.Vetters
DIRECTOR

-xi
-

z
o

4
m

i WY

r-

1.

INTRODUCTION

Lamellar
tearing
isa separation
intheparent
orbasemetalcausedby
through-thickness
strains.
Suchstrains
areinduced
primarily
byweldmetal
shrinkage
underconditions
ofhighrestraint.
Whilethe lamellar
tearing
phenomenon
hasbeenrecognized
bywelding
experts
foroverthirty
years,
the
incidence
of lamellar
tearing
inshipbuilding
hasbeenextremely
rarefor
shipsunderconstruction
or inservice.
Thelowrateof occurrence
should
notbeliethefactthatlamellar
tearing
canbe a potentially
significant
problem
whenitoccurs
incritical
connections.
Theincidence
of lamellar
tearing
isconsiderably
moresignificant
inmobile
andfixedoffshore
drillingplatforms.
Thesearecomplexstructures
whichusethickplatesin
highly
restrained
T andcruciform
joints.
Wheredetected,
lamellar
tearingcanresultin oftendifficult
and
costly
repairs
andsubsequent
construction
delays.Withtheproperselectionofjointdesigns,
materials,
andwelding
procedures,
theoccurrence
of
lamellar
tearing
canbeminimized
andcontrolled.
Theintent
ofthismanual
istoprovide
theengineer
ordesigner
withspecific
recommendations
for
controlling
lamellar
tearing
inthetypesofsteels
usedintheconstruction
of shipsandoffshore
platforms.
A briefdescription
ofthecharacteristics
andmechanism
of lamellar
tearing
isprovided
togivea basicunderstanding
of thecomplexities
of theproblemandtherationale
forthe subsequent
recommendations
foritsprevention.
Methodsof detecting
andrepairing
lamellar
tearsafter
welding
arealsopresented.
Thefollowing
organizations
havegenerously
provided
dataon their
experience
withlamellar
tearing
andmethods
foritscontrol:
* American
Bureau
ofShipping
o Bureau
Veritas

DetNorske
Veritas

@ Germanisher
Lloyd
s Lloyds
Register
ofShipping
@ Nippon
KajiKyokai

Avondale
Shipyards,
Inc.

Continental
OilCompany

Lukens
SteelCompany

1-1

Mostof thepublished
literature
on lamellar
tearingaddresses
in
detail
thefactors
andmechanisms
whichcontribute
to lamellar
tearing
while
reviewing
control
procedures,
particularly
weldingtechniques,
in a more
cursorymanner.
Theprimary
exceptions
tothisgeneralization
areTechnical Note6 - Control
ofLamellar
Tearing
published
bytheAustralian
WeldingResearch
Association
andLamellar
Tearing
inWelded
SteelFabrication
published
byTheWelding
Institute.
Withtheirpermission
portions
oftheir
previously
published
material
hasbeenincorporated
inthismanualandthe
excellence
oftheirwork
andtheirgenerosity
inpermitting
itsuse is acknowledged.
Special
thanksis alsoduetheAustralian
Welding
Research
Association
forpermitting
thereproduction
inthismanualof Appendices
A
andB oftheir
Technical
Note6.

1-2

2. DESCRIPTION
OFLAMELLAR
TEARING
(LT)
2.1 WhatIsLamellar
Tearing?
Lamellar
tearingin steelfabrications
is theseparation
of
parent
orbasemetal,
primarily
inplanes
parallel
to therollingplaneof theplate,dueto
highthrough-thickness
strains.
Thehighstrains
inthethroughthickness
direction
areusually
induced
by localized
weldmetal
shrinkage
at hicihly
.
. restrained

NOTE

Lamellar
tearing
should
notbeconfused
with
laminations
whch arediscontinuities
inroled
steelproducts
resulting
fromflattening
and
elongating
of inclusions
orvoidsduring
the
rolling
process.
2.2 WhereDoesLamellar
Tearing
Occur?

Remote
Underbe
LTNear

Thetearing
alwayslieswithin
the basemetal,usuallyjust
outside the visible heataffectedzone (HAZ),and is
ionLine generally
parallel
to theweld
fusion
boundary.The location
mayvaryfromwithinthe lower
HAZtowellintothebasemetal
thickness.
Thetearingmaybe
subsurface
anddif.HAZcompletely
ficultto detector readily
RootLT
visible
on exposedplateedges
or at thetoe androotof the
weld.

IIIIIIH-H--

-w-

lNumbers
inbrackets
designate
References
inSection
9.
2-1

2.3 WhatIsTheExtent
ofTearing?
toseveral
Lamellar
tearsmayvaryinlength
froma fewmillimeters
equalto thesizeof the
meters
andhavea widthapproximately
crack
ofthefracture
mayvaryfroma hairline
weld.Thethickness
to approximately
1 mm.
2.4 WhatDoesA Lamellar
TearLookLike?
lartearsexhibitunique
Lamel
appearance
characteristics
which
enable
themtobedistinguished
fromotherformsof cracking,
includingcracksin the HAZ
Whena tear
caused
byhydrogen.
reaches a surface or is
sectioned,
itgenerally
appears
as a straight
linein thebase
metalparallel
tothedirection
ofrolling
oftheplate.
Z Direction
1

//

1/

Thecross-section
is step-like
withlongitudinal
terraces
that
aresubstantially
longerthan
thetransverse
depth.

Thefracture
surface
is fibrous
or woody in appearancewith
little
orno discoloration
unlesstheteariscorroded
orhas
beensubject
to hightemperatures. Theflatfibrousterraceslieparallel
to theplate
surface,
withstepsor shear
wallsbetween
terraces
approximatelynormalto the plate
surface.

NOTE
Thecharacteristic
fibrous
orwoodyappearance
ofthe
fracture
surface
together
withtheterraced
profile
and
location
within
thebasematerial
distinguishes
a lamellar
tearfromotherformsofcracking.
2-2

2.5 HowIsLamellar
Tearing
Normally
Observed
Or Detected?
Lamellar
tearswhichpropagate
to thesurfacecanbe detected
by
visual,
dyepenetration
andmagnetic
particle
inspection
techniques.However,
sincemostlamellar
tearsarecompletely
subsurface,
thesedetection
methods
areof limited
usefulness.
Ultrasonictesting
hasbeenfoundto be themosteffective
methodof
detecting
sub-surface
tears. A moredetailed
discussion
of the
detection
andrepair
of lamellar
tearing
ispresented
inSection
7.

2-3

CONTRIBUTING
TOLAMELLAR
TEARING
3. FACTORS

3.1 UnderWhatConditions
DoesLamellar
Tearing
Occur?
For.lamellar
tearing
tooccurthefollowing
threeessential
conditionsmustbesatisifed:
o Thematerial
mustbe susceptible
to tearing.Thatis,the
basematerial
Intheregion
ofthejointmusthavepoorductility
intheZ-(through-thickness)
direction.

Thewelding
procedures
mustproduce
strains
whichactthrough
thejointacross
theplate
thickness,
thatis,through-thicknessstrains.
Sucha condition
existswhentheweldfusion
boundary
isroughly
parallel
tothesurface
ofthebaseplate.

@ Thejointdesign
mustpermit
thedevelopment
ofhighthroughthickness
strains.Thesestrainsusually
resultfromweld
metalshrinkage
inthejointbutcanbe increased
by strains
developed
fromreaction
withotherjoints
inrestrained
structures.

NOTE
Forlamellar
tearing
tooccurtheremustbea critical
combination
ofmaterial
susceptibility,
andwelding
procedures
andjointdesign
whichpermit
thedevelopment
of highthrough-thickness
strains.

3.2 ByWhatMeansDoesLamellar
Tearing
Occur?
Lamellar
tearing
isgenerally
believed
tooccurin threedistinct
phases.During
thefirstphasevoidsareformedusuallyby decohesion
orfracture
ofsingle
elongated
nonmetallic
inclusions
or
groups
of inclusions
lyingparallel
to therollingplaneof the
additional
voidinitiation
mechanisms
havebeen
plate.Although
reported,
thedecohesion
ofmicroscopic
inclusions
is considered
theprimary
initiation
mechanism.
Thefirstphaseprobably
takes
placein theelasticrangewherethestressrequiredfor the
initial
decohesion
willbedependent
onthetype,shapeanddistribution
of inclusions
andtheproperties
ofthematerial
matrix.

3-1

Terrace

Shear
Wall

-s
Decohesion
at Inclusions

Inthesecond
phasetheinitiated
voidsortearsonthesameplane
extend
andjoinbymeansof neckingor microvoicl
coalescence
to
formterraces.
Theelongation
andlinkupof adjacent
inclusions
iscaused
by increased
strains
duetocooling
ofpreviously
depositedweldrunsand/or
thedepositing
of additional
weldmetal, As
thestrains
increase
theligaments
ofmatrix
material
betweenthe
inclusions
become
fullyplastic
andthevoidsincrease
in sizeby
ductile
tearing.
Further
straining
inthethirdandfinalphaseconnects
theterracesondifferent
levels
byductile
shearing
ofthevertical
walls
between
theterraces.
Theformation
oftheshear
wallscreates
the
characteristic
step-like
appearance
ofthecompleted
lamellar
tear.
Additional
information
onthemechanism
of lamellar
tearing
ispresented
inAppendix
A. Thefactors
whichcontribute
to andinfluencelamellar
tearingmay be groupedintothreecategories:
design,
material
andfabrication.
3.3 Design
Factors
Thesusceptibility
ofa structural
component
orjointto lamellar
tearing
isaffected
bythosedesign
factors
whichdetermine
theinternal
resistance
ofthejointandtheresulting
accumulation
of
weldmetalshrinkage
strain
inthethrough-thickness
direction.
The
principal
design
factors
whichinfluence
theriskof lamellar
tearingare:
3-2

,..
.
.-;
.
,..,
A
.

~-.
. .,;,
~:.,.,
.,:,,
,:.
:

* WeldOrientation.
Jointconfigurations
whichorientthe
weldfusion
boundary
parallel
to thedirection
of rolling
of thebasemetalpromote
the
development
ofthrough-thicknessstrains.Tee (T) and
cornerjoints,theprimary
examples
ofsuchjoints,are
usedextensively
inshipsand
offshore
structures,

o JointRestraint.
Thelevel
ofjointrestraint
is an important
factor
indetermining
theamount
andconcentration
of strainat
the connection
andis influenced
by thesize,balance,and
distribution
of theweld. Weldswhicharelargerthanthose
required
to accommodate
thedesign
loads
unnecessarily
increase
theweldshrinkage
strainsas do theunwarranted
useof wide
groove
angles
andfullpenetration
weldsin placeof properly
sizedfillet
welds.Inmultipass
welds,
thesizeoftheweldbead
determines
thenumber
ofpasses
required
tofillthejoint. The
smaller
thebeadsize,thegreater
thenumber
ofrequired
passes
andthehigher
theweldshrinkage
strains.

>

Jointdesigns
withlargesinglesidedweldscauseunsymmetric
strainsto concentrate
on the
sideoftheweld. Double-sided
weldsreduceand balancethe
shrinkage
strains
witha resultant decreasein the risk of
lamellar
tearing.

Flanged
Gusset
Plate

@ Component
Restraint.Structuralcomponents
fabricated
of
thickand/orcurvedplates,
and stiffenedwith heavy
brackets
or gussetshaveinherently
morerestraint
inthe
through-thickness
direction
thancomponents
fabricated
of
unstiffened,thin, flat
plates.

3-3
-

F1
Examples
of.highcomponent
restraint
usually
canbe foundat
themulti-column
connections
or
nodejoints
ofmobileandfixed
offshore
structures.

o WeldMetalStrength.
Whentheyieldpointoftheweldmetalis
significantly
higher
thanthatof thebasemetal,
allof the
weldshrinkage
strains
mustbe accommodated
by thebasemetal
matrix.Theconcentration
ofthestrain
inthebasemetalincreases
theriskof lamellar
tearing.Weldmetalis usually
matched
tothebasemetalonthebasisofequivalent
tensile
strengths.However,
weldmetalswhichmatchthe tensile
strength
of thebasematerial
generally
havesignificantly
higher
yieldpoints
thanthebasematerial.
3.4 Material
Factors
A detailed
discussion
ofthemetallurgical
factors
whichinfluence
thesusceptibility
of rolledsteelplatesto lamellar
tearing
wouldbetoovoluminous
to include
ina practical
guidance
manual
fordesigners
andengineers.
However,
an understanding
of the
fundamental
metallurgical
considerations
isnecessary
toobtain
an
appreciation
ofthecomplexity
oftheproblem
andtheunderlying
rationale
forthecontrol
methods
presented
in Section
6. Additional
information
onthematerial
factorsinfluencing
lamellar
tearing
maybefoundinAppendix
B andtheselected
workslisted
inthebibliography.
Lamellar
tearing
isdirectionally
sensitive
andatleast
partially
dependent
on thethrough-thickness
propertiesof the base
material.
Theanisotropy
ofhot-rolled
steelplates
usually
producesthegreatest
strength
andductility
inthelongitudinal
and
transverse
directions
withsignificantly
lessductility
in the
through-thickness
direction.
The susceptibility
of carbonand
low-alloy
steels
to lamellar
tearingis primarily
dependent
on
theselowthrough-thickness
(Z-direction)
ductilities.
Thetype,
number,
shapeanddistribution
ofthenonmetallic
inclusions,
as
wellasthematrix
properties
of theparticular
gradeof steel,
aregenerally
considered
responsible
forthereduction
inductility intheZ-direction.

3-4

,
z.

Allnormal
quality
structural
steels
forhullandmarineapplicationscontain
quantities.of
exogenous
andindigenous
inclusions.
Exogenous
inclusions
usually
consist
of ladlerefractory,
ingot
scum,orslagthatisoccasionally
trappedin the ingotduring
solidification.
Theyareusually
largeincomparison
to indigenousinclusions
andwhenlocated
closetothesurface
of a rolled
platesignificantly
increase
thesusceptibility
of theplateto
lamellar
tearing.
Indigenous
inclusions
areformed
asa result
ofthechemical
reactionof elements
in thesteelor elements
addedto thesteel
usually
during
deoxidation.
Thenumber
anddistribution
of indigenous inclusions
dependsonthe steelgradeanditschemical
composition,
thedeoxiciation
procedure,
themeltingtechnique,
Whenthe
position
intheingot,
andthehotworking
temperature.
ingotisrolled
toforma plateorsection
theinclusions
areprogressively
elongated
andflattened
to varying
degreesto form
plates
orstringers
parallel
totheplatesurface.
Material
which
hashighconcentrations
ofelongated
orflattened
inclusions
will
havelower
through-thickness
ductility
anda greater
susceptibility to lamellar
tearing.
Thedominant
inclusions
aresulfides
andoxides
withthedeoxidationpractice
determining
thetypeofeach,
inclusion
present.
For
comparison
purposes
deoxidation
practices
areusuallyclassified
intwocategories:
non-aluminum
treated
andaluminum
treated.In
semi-orfully-killed
non-aluminum
treatedsteelssilicates
and
TypeI manganese
sulfides
aretheprimarytypesof inclusions.
TypeIImanganese
sulfides
andalumina
aretheprincipal
inclusionsinfully-killed
aluminum
treatedsteelswhileTypeIII
manganese
sulfides
andalumina
inclusions
predominate
in fullykilledwithexcessaluminum
materials.In the non-aluminum
treated
steels
thesilicates
become
moreelongated
thanthesulfidesduring
hotrolling
andareprimarily
responsible
forthe
reduction
inZ-direction
(ST)ductility.However,
in aluminum
deoxidized
steeltherodshapedmanganese
sulfideinclusions
become
highly
elongated
during
rolling
andaretheprimarycause
ofthelowZ-direction
ductility.
Highconcentrations
orelongated clusters
ofmanganese
sulfides
andalumina
canalsoproduce
locally
poorZ-direction
ductility
in non-aluminum
andaluminum
treated
steels,
respectively.
Manyoftheearlier
workson metallurgical
aspects
of lamellar
tearing
emphasized
theimportance
ofsulfur
content
andinclusion
shapecontrol
to improve
through-thickness
ductility
as measured
by thepercentage
reduction
of areaintheshort-transverse
direction.Fora reduction
of areahigher
than25 percent(a levelat
whichtheriskof lamellar
tearing
issignificantly
reduced),the
sulfur
content
mustbe lowerthan0.010percent.Addition
of
.
rare-earth
(RE)metalreducestheresidual
sulfurlevelswhile
alsopreventing
theformation
ofmanganese
sulfides
andsilicates,
forming
instead
onlysmallglobular
shapedRE-containing
inclufornon-alunlinum
treated
steels,
wheresilicates
sions.However,
areprimarily
responsible
forreducing
theZ-direction
ductility,
3-5

thesulfur
content
alone
cangiveaninadequate
indication
of the
susceptibility
of thematerial
to lamellar
tearing.Improved
manufacturing
processes
suchaselectroslag
remelting
andcalciumargon-blowing
mayalsobeusedtoreduce
themaximum
sulfur
levels
and/or
remove
mostofthenonmetallic
inclusions.
The properties
of thesteelmatrixarealsoimportant
in all
phases
oftearing.
Forsteels
witha low-strength,
highly
ductile
matrix~the
material
attheedgesof inclusions
candeformplastically
without
propagating
thefractures
or voidsformedby the
decohesion
ofthenonmetallic
inclusions
andthematrix
materials.
Inhigher
strength
steels,
thethrough-thickness
ductility
decreases
whilethehigher
yieldstrength
of thematrixmaterial
permits
thedevelopment
ofhigh-strain
levels
across
an inclusion
before
thematrixyields.Thesehigherstrainlevelsin turn
facilitate
theextension
andjoiningof adjacent
voidsin the
second
phaseoftearing.
Ferrite-pearlite
banding
in thesteelmatrixhasalsobeenreported
tocausebothinitiation
andpropagation
of lamellar
tears,
partially
because
theferrite
hasa lower
cleavage
fracture
stress
thanthepearlite. Strainaging,hydrogen
embrittlement
and
differences
inthethermal
expansion
between
the inclusions
and
thesteel
matrix
allcontribute
in somedegreeto thesusceptisteels
bility
ofsteelplates
to lamellar
tearing.Susceptible
withhighbrittle
fracture
transition
temperatures
showimproved
resistance
totearing
whenpreheated
abovethebrittlefracture
transition
temperature
before
welding
[2].
3.5 Fabrication
Factors
Fabrication
practices,
particularly
welding
variables,
helptodetermine
thelevel
ofjointrestraint
andtheresulting
riskof
Factors
whichaffect
lamellar
tearing
susceptilamellar
tearing.
bility
include
preheat
temperature,
heatinputlevel,
beador run
sequence,
andfabrication
sequence.
Increasing
preheatandheat
inputlevels
arereported
to increase
thepostweld
ductility
of
themetalwitha correspondi~y
improvement
intearing
resistance.
Explanations
fortheapparently
lower
riskoftearing
withhigher
preheat
andheatinput
welding
processes
arevariedandinclude
increased
weldpenetration
andweldmetaldeposition
rate,reduced
rateofpostweldcocling
andproduction
of a wider,softerand
tougher
HAZ. Increased
penetration
canintercept
andbluntexistinglaminations
whilehigherdeposition
ratesdecrease
therequired
number
ofweldrunsandthesubsequent
numberof strain
incooling
ratespermits
stress
relaxation
cycles.Thereduction
Theuseofhighandthedevelopment
ofsmaller
strain
gradients.
er heatinput
processes
willalsoproducelowerstrength
welds
whichwillaccommodate
moreoftheshrinkage
strain.In addition
to improving
thepostweld
ductility
of thematerial,
preheating
mayretard
thepropagation
of lamellar
tearingby raisingthe
temperature
ofthesusceptible
material
aboveitsbrittle
fracture
transition
temperature.
3-6

Basedontheseresearch
results
the influence
of higherpreheat
andheatinputweldingprocesses
on the incidence
of lamellar
tearing
appears
substantial.
However,
reports
of fabricator
experience
indicate
little
ornodiscernible
success
withincreasing
preheat
orheatinput
within
a givenwelding
process.
Onthecontrary,higherpreheatandheatinputlevelsmay increase
the
amount
ofsubcritical
tearing
andcontraction
strains.
Thesequence
ofdepositing
theweldbeadsorrunscansignificantly affect
thelevelandconcentration
of shrinkage
strainsnear
the HA2 andparallel
to thedirection
of rollingof thebase
plate.Whenfabricating
double-sided
T jointsunsymmetrical
depositing
oftheweldmetalcancausestrains
to concentrate
on
thesideoftheweld.Symmetrical
depositi[jn
of therunswill
reduce
andsomewhat
balance
theweldshrinkage
strains.

In multi-joint
components
thefabrication
orwelding
sequence
can
affect
therestraint
level
ofeachjointatthetimeof welding.
Theriskof lamellar
tearing
increases
whenthemoresusceptible
joints
aremadetowards
theendofthefabrication
sequence
when
themaximum
restraint
ofthestructure
isbeingapproached.

3-7

4. OCCURRENCE
OF LAMELLAR
TEARING
4.1 WhatTypesofStructures
AreSusceptible
toLamellar
Tearing?
Lamellar
tearing
usually
occursat highlyrestrained
jointsin
largewelded
structures,
Therestraint
maybe imposed
by a masTee
sivecomponent
orby a smaller
onewhichhasbeenstiffened.
(T)andcorner
joints
arethetwobasicjointconfigurations
most
susceptible
to lamellar
tearing. Thecruciform
jointis considered
a moresevere
formoftheT jointsincetherestraint
of
thebaseplateinwayoftheweldishigher.Thesusceptibility
of thesejoints
reflects
thefactthattheinternal
restraint
of
thejointin thethrouqh-thickness
direction
is sufficient
to
causetheweldshrinkag;
strains
toexceed
theductility
limits
of
thebasemetal.

Withtheexception
of cruciform
joints,
T joints
withsingleor
double-sided
full-penetration
weldshavethegreatest
incidenceoftearing.
T joints
with
simple
filletor partialpenetration
ratherthanfull-penetration
weldsappear
to present
lessriskasdobalanced
doublesidedweldscompared
to large
single-sided
welds.

In cornerjoints,
tearingcan
occurin one or more planes
throughthebaseplatethickness.Thetearsoftenextend
to
the exposedplateedge where
they are eithervisible or
readilydetected
by standard
non-destructive
testing
methods
suchasdyepenetration
or magneticparticle
inspection.

4-1

...

:FV=lE!!:$3~
t<lgmm

boundary
isata largeangle
tothe~latesurface.However,tearinghasbeenreported
inbuttweldsofthick
plates
(ta 19mm)
withan xgroove.

t
t

t~19mm

In the heavyfabrication
andconstruction
industriesj
lamellar
tearing
iscommonly
reported
tooccurin thefollowing
typesof
structures:

o Nozzle
or insert
setthrough
a rlald~late.Tearingcan
occur in a rolled p-late
nozzleor penetratorset
through
a vessel
shellplate
orendwall,or in a fabricatedinsert
inthewebof a
largegirder.Forexample,
a
Vierendeel
girder
fabricated
of heavyplatesections
with
a ringstiffener
setintothe
webopening
issusceptible
to
lamellar
tearing
in thering
Inallcases,
any
stiffener.
tearingwilloccuronly in
thenozzle
or insert
plate.

%,e, :
P1ate

4-2

s
Shellof Cylin&ical
Vessel
/
RigidEndPlate

Stiffeners
or end closure
strucPIatesIncylindrical
tures.Shellplatesof cylindrical
structures
which
areinwayoftheendclosure
platesor heavyinternal
stiffeners
aresusceptible
to
lamellar
tearing.In structuresofthistypethetears
canbecompletely
subsurface
anddifficult
todetect.

Boxstructures
andstiffened
jointssuchas beam-to-column.
Structures
inthiscategory
rangefromsimple
boxcolumns
to large
structural
configurations
withcomplex
multi-member
connections.

/------[

Heavy
Lifting
Lug

lc)

4-3

.....

..

For
highly
susceptible
materia.,
11.
tearing
hasbeenreported
in
apparentlylow restraint
situations
suchaspullout
of
lifting
lugsandin flangeto-web
connections
in fabricatedI-beams.Theriskof
lamellar
tearingin apparentlysimple,unrestrained
jointsmakesit essential
thatforcritical
components,
suchas liftinglugs,post
weldinspection
fortearing
be performed
andoftenaccompanied
by a reduction
in the
through-thickness
service
loads;

Documented
casesof lamellar
tearing
areextremely
rareforships
However,
isolated
instances
of
underconstruction
or inservice.
lamellar
tearinghavebeenreported
in thefollowing
typesof
CVK/innerbottom,
CVKriderplate/transstructural
connections:
versebulkheadg
deckstringer
plate/side
shellsheerstrake,
container
buttress
supports
andthick-walled
boxgirdersof large
container
ships.

nger

Innerbottom
-

Il!!!r

Note Unusual
Configuration

SheerStrake

DeckStringer/Sheer
Strake

(
\

CVK/Innerbottom

w
Transverse
Bulkhead
RiderPlate

.CVK

CVKRiderPlate/Transverse
Bulkhead
heavystern
frameweldBulkhead
or innerbottom
heelconnections,
mentsandthickwebframeflange
to longitudinal
bulkhead
connectionsinlarge
tankers
arealsoconsidered
susceptible
to Iamellar
tearing,
although
noactual
failures
hav@beenrePorted

4-4

Tank
Top
Bulkhead
or
Cofferdam

Bulkhead
or
Cofferdam
a

Shel
1 Plating
\
Plating
$

Itisimportant
tonotethatallofthesesusceptible
shipdetails
areessentially
variations
ofthebasiccorner,
T andcruciform
joints
described
inthebeginning
of thissectionas beingthe
mostsusceptible
to lamellar
tearing.
Theweldsmaybe eitherof
thedouble
continuous
filletor bevelgroovetypewithfullor
partial
penetration.
To date,lamellar
tearinghasnotbeenresponsible
foreither
Whileitcannot
be
numerous
orcritical
failures
inshipbuilding.
considered
a serious
problem
basedontherateofoccurrence,
the
designer
orengineer
mustbeaware
thatlamellar
tearing
canbe a
potentially
significant
problem
whenitoccurs
incritical
connections,
suchasbulkhead
orcofferdam
heels.
Intheseareas)proceduresforthecontrol,
detection
and,Ifnecessary,
therepairof
lamellar
tearsshould
be implemented.
Wherethe increase
in the
sizeofships
results
in structural
assemblies
fabricated
from
thicker
plates,
theriskof lamellar
tearing
in jointswhichare
4-5

..

acceptable
whenfabricated
of thinner
material
shouldbe reevaluated.
Whilethesignificance
of lamellar
tearing
shouldnot
be underestimated,
theextent
oftheproblem
shouldnotbe exaggerated
tothepointthatexpensive
materials,
andfabrication
and
inspection
procedures
areunnecessarily
specified.
Theproblems
of Iamellar
tearing
inmarine
structures
areconsiderably
moresignificant
intheconstruction
of mobileandfixed
offshore
drilling
platforms.
Theconfiguration
of thesestructuresis verycomplex
withtheuseof thickplatesin highly
stressed
welded
T andcruciform
joic:s.Thesejoints
usually
take
theformofmulti-column
connections
ornodejoints
atwhichtubes
of large
diameter
andthicksection
passthroughor aresurface
welded
to another
tubewithfull-penetration
fillet
welds.

L
L

4-6

HeavyWallPipe
At Intersection

..
k

Sample
structural
connections
ofcolumn
stabilized
and
self-elevatingmobile
offshore
unitsandfixedjacket
typeplatforms
which
aresusceptible
to lamellar
tearing
include:
Column
Stabilized
Units
1.

Intersection
ofvertical
columns
andupperandlower
hulls.

2. Majorintersections
ofhorizontal
andvertical
braceswith
themselves
andwiththevertical
column.

3. Portions
ofdeckplating,
heavyflanges,
andbulkheads
withintheupperhullorplatform
whichformi~ox
orI typesupporting
structure.
Self-Elevating
Units
1.

Jackhousesupporting
structure
andbottomfootingstructure.

2. Vertical
columns
in wayof the intersection
withthemat
structure.
3. Combinations
of deck,side,bottomandbulkhead
plating
within
theupperhullwhichformboxor I typesupporting
structure.
4-7

JacketTypeFixedPlatforms
1.

Deckto legcanintersections,

2. Majorintersections
ofhorizontal
andvertical
braceswith
themselves
andwiththevertical
cancolumns.
4.2 HowOftenDoesLamellar
Tearing
Occur?
Larnellar
tearing
hasbeenestimated
tooccur
insignificantly-less
thanonepercent
of allweldments.
Thefrequency
of occurrence
increases
slightly
forlargeweldedstructures
fabricated
of
plates
orsections
over25to 30mm inthickness
underconditions
of highrestraint
inthethrough-thickness
direction.
For applications
whichdonotsatisfy
theessential
conditions
of material
susceptibility
andthrough-thickness
strains
dueto weldingprocedures
andjointconfiguration,the
riskof lamellar
tearingis
negligible.
Thefrequency
of lamellar
tearing
in theconstruction
of ships,
andmobile
andfixedoffshore
structures
isdifficult
toestimate.
Replies
toquestionnaires
senttotheworlds
majorclassification
societies
indicate
thattheincidence
of lamellar
tearing
inshipbuilding
issmall.Isolated
casesof lamellar
tearingin such
connections
asthedeckstringer
plate/side
shellsheerstrake
havebeenvirtually
eliminated
by theuseof improved
weldand
jointdetails.
Theproblem
of lamellar
tearing
is considerably
moreseriousin
theconstruction
ofmobile
andfixedoffshore
drilling
platforms.
Thegreater
susceptibility
totearing
ofthelarge
number
ofhighlyrestrained
T andcruciform
joints
inthesestructures
increases
thefrequency
withwhichlamellar
tearingoccurswhennormal
structural
quality
steel(sulfur
content
> 0.020%by weight)is
used.Thefrequency
of tearingis reducedsignificantly
when
steels
withimproved
through-thickness
properties
areusedinconjunction
withrevised
welding
procedures
andjointdesigns.One
oilcompany
whichfabricates
15to20fixedoffshore
structures
a
yearestimates
theirfrequency
of lamellar
tearingat lessthan
oneperyear.
4.3 WhenDoesLamellar
Tearing
Occur?
Lamellar
tearing
usually
occurs
during
fabricationjoften
atanadvanced
stagewherpthemaximum
level
ofrestraint
is approached.
Thereisconsiderable
disagreement
in the literature
concerning
thetimeandtemperature
attheonsetof tearing.Somereports
indicate
thatlamellar
tearing
is initiated
shortlyafteradditional
weldmetalis deposited
overprevious
beadswhichhave
cooled
tothepointofdeveloping
weldshrinkage
strainssufficienttocausedecohesion
at the interface
between
microscopic
nonmetallic
inclusions
andthesurrounding
matrix.Otherreports
conclude,
however,
thattearing
isanambient
temperature,
delayed
cold-cracking
phenomenon.
4-8

4.4 WhatTypesofSteel
AreSusceptible
ToLamellar
Tearing?
Lamellar
tearing
hasbeenencountered
primarily
innormalquality
structural
steelplates
ofthecarbon,
carbon-manganese
andlowalloytypes.Thesteel
maybe inthenormalized,
as-rolled,
controlled-rolled
orquenched
andtempered
condition,
or be fineor
coarse
grain.Examples
oftypical
.American
SocietyforTesting
andMaterials
(ASTM)
andAmerican
Bureau
ofShipping
steelspecifications
withreported
histories
of lamellar
tearing
include
[3]:
SpecificationTypeof Product
ABSAH36
ASTMA36
ASTMA283
ASTMA285
ASTMA515
ASTMA516

Higher
Strength
HullStructural
Steel
Structural
Carbon
Steel
Structural
Carbon
Steel- LowandIntermediate
Tensile
Strength
Pressue
VesselCarbonSteel- LowandIntermediate
Tensile
Strength
Pressure
Vessel
Carbon
Steel- ForIntermediate
andHigher
Temperatures
Pressure
Vessel
Carbon
Steel- ForModerate
and
LowerTemperatures

Nonmetallic
inclusions
havebeenshown
tobeprimarily
responsible
forlowthrough-thickness
ductility
andthevoidinitiation
phase
andthedeoxidation
method
usedinthesteelof lamellar
tearing;
making
process
determines
thetypesof inclusions
presentin the
steel.Theearlier
literature
indicated
thataluminum
treatedsemikilled
steels
could
beexpected
tohavebetterresistance
to
lamellar
tearing
thansilicon-treated
semiskilled
steels.However,
recent
studies
report
thatthelamellar
tearing
mechanism
is too
complex
tosimply
relate
susceptibility
tosteelgradeor inclutreated
orsemiskilled
steels
cannotbe
siontype.Hence,aluminum
considered
moreor lesssusceptible
totearing
thannon-aluminum
treated
orfullykilled
steels.
Intheory,
thereisanincreased
riskof lamellar
tearing
withincreasing
strength
levels.Forhigher
strength
steel,
thethroughthickness
ductility
decreases
whilethegreater
strength
propertiesofthesteels
matrix
material
willpermit
thedevelopment
of
higher
elastic
strain
across
aninclusion
before
thematrix
itself
yields.Theincreased
susceptibility
ofhigher
strength
steels
is
offset
insomecasesbytheincreased
flexure
duringweldingdue
to theuseof thesmaller
thicknesses
permitted
by thehigher
Somehigher
quality,
high-strength
alloysteels,such
strength.
asHY-80,
HY-1OO,
HY-130
andHY-180,
haveshown
minimum
susceptibility
todecohesion
cracking.
However,
thisresult
isattributed
to theincreased
cleanliness
(reduced
inclusion
content)
of these
special-purpose
steels
produced
by electric
furnacesteelmaking,
coupled
withvacuum
degassing.

4-9

Lamellar
tearing
occurs
primarily
inrolled
structural
plates,
to
a lesser
degree
inrolled
sections
andrarely
inforgings.
Steel
castings
arenotsusceptible
totearing.
4.5 WhatSteel
Thicknesses
AreSusceptible
To Lamellar
Tearing?
Lamellar
tearing
hasoccurred
inplates
ranging
inthickness
from
10to200mm,withthemostcommon
incidence
beinginplates
25to
60mm thick.Thinplatesusuallyhavelowerductility
inthe
through-thicknesss
direction
thanthickerplatesdue to the
greater
deformation
of inclusions
inthinplatesduringrolling.
However,
theydonotnecessarily
exhibita greaterincidence
of
tearing,
since
flexure
ofthethinner
platestendsto limitthe
strains
inthethrough-thickness
direction.Exceptions
to this
generalization
arerolledplatenozzles,
cruciform
jointsand
highly
stiffened
structural
configurations
whichlimit
theflexure
rfthethinner
plates.

4-10

,.,
.:

OFLAMELLAR
TEARING
5. SIGNIFICANCE
Wheredetected,
lamellar
tearing
canresult
incostlyrepairsandfabrication
delays.Thesignificance
ofundetected
orunrepaired
tearson the
service
performance
ofthestructure
varies
withthetypeof loading.The
following
sections
evaluate
theeffects
of lamellar
tearing
on thestatic,
dynamic
andfatigue
modesof loading.

5.1 Static LoadCondition

Theextremely
fewreported
incidence
of lamellar
tearing
failures
inservice
indicates
thatthestrains
developed
during
welding
are
morelikely
tocausetearingthanthestaticdesignor service
strains
ashighas2%havebeenreported
during
loads.Localized
welding
andtheimmediate
postwelding
cool-down
period.By comparison,the
offset
strain
level,
corresponding
to theyieldpoint
stresses
are
ofmoststructural
steels,
isonly0.2%.Sincedesign
always
significantly
lower
thantheyieldstress
ofthematerial,
thestrains
encountered
in serviceareat mostonly10%of the
strains
developed
during
welding.
Preliminary
results
ofresearch
doneintheUnited
Kingdom
atTheWelding
Institute
indicate
that
evenin caseswhereextensive
tearingis initiallypresent
through-thickness
staticstresslevelsgreater
thantheyield
strength
ofthebasemetalarerequired
to extendthetearsto
complete
failure.
Hgwever,
astheextent
of the initial
tearing
increases,
thestress
levels
necessary
topromote
failure
decrease
[3].
Lamellar
tearing
isreported
tohavenoeffect
ontheservice
performance
of jointsstressed
primarily
in compression
in the
through-thickness
direction.
Injoints
subject
to shear,
theserviceperformance
willnotbediminished
provided
thereis sufficientareaintheremaining
ligaments
between
thetears.In areas
of extensive
tearing,the
maximum
shear-load
capacity
ofthejoints
maybereduced.

5.2 Dynamic
LoadCondition
Verylittle
information
isfoundintheliterature
concerning
the
effect
of lamellar
tearing
ontheability
of a structure
to withstanddynamic
loads.A fewstudies
report
reduced
CharpyV-notch
impact
energies
anddynamic
tearingproperties
in thethroughthickness
direction.
Shocktestsperformed
bytheBritish
Navyon
full-penetration
weldedT-joints,
fabricated
of HY 80 andtwo
grades
ofC-Mnandlowalloysteels~showed
thatlamellar
tearing
couldbe initiated
bydynamic
loads.Ofthethreesteelstested,
onlytheHY80,withitsgreater
Z-direction
ductility,
failedto
develop
lamellar
tears.Although
notconclusive,
thesereports
wouldseemto indicate
thatmaterials
withlowerZ-direction
properties
aremoresusceptible
to lamellar
tearingwhenexposed
to dynamic
loads.Conversely,
thepresence
of undetected
tears
canonlyincrease
theriskoffailure
during
dynamic
loading.
5-1

5.3 Fatigue
Forlowcyclefatigue,
existing
lamellar
tearswillgradually
extendandmayultimately
result
incomplete
failureas thenumber
of cycles
approaches
thedesign
limit
[1]. However,
in practice,
catastrophic
failure
may be avoided
by thetransfer
of loadto
Stress
concentrations
attheroot
othermembers
ofthestructure.
or toeoftheweldmaybemoredetrimental
thanexisting
tearsor
is exposedto
poorZ-direction
ductility
when he structure
high-cycle
fatigue
(greater
than10t cycles)
[3].

5-2

OFLAMELLAR
TEARING
6. CONTROL
Sincedesign,
material
andfabrication
factors
contribute
to lamellartearing,
control
oftearing
mustaddress
thesesameparameters.Itisevident
fromSections
2 and3 thatthecausesof
lamellar
tearingareapplicable
to generic
typesof weldments
whichareindependent
of thespecific
endproduct.It matters
little
whether
thesusceptible
weldments
arein a skyscraper,
or largeoffshore
structure.
nuclear
powerplant,
super
tanker,
Accordingly,
mostofthefollowing
recommendations
forthecontrol
of lamellar
tearing
inthemarine
industry
arepresented
in their
mostfundamental
form.Itisimperative
thatthenavalarchitect
or designer
usejudgement
to arriveat theoptimum
balanceof
material
selection
andcosteffective
fabrication
jointdesign,
procedures
suitable
fortheapplication.

6..1 Joint Design

Theavoidance
andcontrol
of lamellar
tearing
mustbeginat the
design
stage.Thedesignof susceptible
jointssuchas those
showninSection
4.should
beoptimized
wherepracticable
to:

Avoidexcessive
through-thickness
strains

Reduce
jointrestraint

o Reduce
component
restraint

Allowfortheuseof low-strength
weldmetals

6.1,1Avoidance
ofExcessive
Through-Thickness
Strains
Methods
foravoiding
thecreation
ofweldshrinkage
strains
inthe
through-thickness
direction
include:
.,*:
.
.
.
,...+.
..,
.4
.
,.,,.,
..::
..-....
,:#,
.-.:..-:
.
.

iiii?

Susceptible

Improved

6-1

Welding
between
theends
of platesratherthanon
thesurface
ofthesusceptiblematerial. This
welding
technique
directs
theshrinkage
strainsin
the X or Y directions
rather
thaninthecritical Z direction
and may
require
the useof electroslag
welding.

Electroslag
Weld
:

.7

A*
.
Susceptible

Improved

Electroslag
Weld

.,.-.
.

Susceptible

~f.
,*F?.
:

Improved

LeastImproved
6-2

Orienting
theweldfusion
boundary
at an angleto
thesurface
ofthesusceptibleplate.Largebevel
angles
offerlessriskof
tearing,
buttheedgepreparation
costandthevolumeofweldmetalrequired
is alsohigherthan for
smaller edge angles.
Selection
of a cost-effectiveanglemustconsider
thesusceptibility
of the
plate,the importance
of
the connectionand the
relative
costof fabrication.

Susceptible

. Replacing
ofdouble-sided,
full-penetration
welds
withsymmetrical
fillet
or
partial-penetration
welds
tominimize
thevolumeof
weldmaterial
andreduce
thestrain
intheZ-direction.Thetotalshrinkage
ofthefillet
weldsoccurs
at anoblique
anqle
tothe
platesurface
th~rebv
furtherreducing
the~train
component
intheZ-direction.

Improved

. Usingcastings
orforgings
in some criticalT and
cruciform
joints
toeliminatethecriticalwelds
andanyriskof lamellar
tearing.Thismethodis
expensive,
involves
considerably
morewelding,
andis generally
usedin
highly
critical
situations
inpressure
vessels.

Forging
or
Casting

6.1.2Reduction
ofJointRestraint
Methods
ofreducing
jointrestraint
include:

ii

Susceptible
3

Improved

13t

_l-

Improved
6-3

Reducing
thesizeof the
weldby notusingwelds
larger
thannecessary
to
transfer
thecalculated
designloads. Forexample, full-penetration
weldsatthedeckstringer
plate/sheer
strake
connectioncanoftenbere~laced
by smaller
partial
~eri~tration
orfillet
welds.

0 Joining
plates
of different thicknesses
so that
theweldsizemay be reducedbyplacing
itinthe
thinner
plate.

Susceptible

Susceptible

Susceptible

Improved

!==3

o Replacing
largesinglesidedweldswithbalanced
double-sidedwelds in
ordertoeliminate
theunsymmetric
concentration
of
strain.

Improved

Improved

6-4

1
Susceptible

Improved

Selecting
weldconfigurationswhichdistribute
the
weldmetalovermoreof the
surface
of thesusceptible
plate.The useof smaller
weldsizesof longerlength
or doublefilletsin place
of fullpenetration
welds
reduces
thevolumeof weld
metal and diffuses the
shrinkage
strainsover a
larger
areaofthesusceptibleplate.

o Othermethods
include
specifyinglow yield strength
weldconsumables
andtheuse
of buttering.
Thesemethods
are discussed in other
sections.

6.1.3Reduction
ofComponent
Restraint
Component
restraint
cansometimes
be reducedby modifying
the
structural
configuration
orscantlings.
Methods
ofdecreasing
the
levelofrestraint
include:

Avoidcomplex,
multi-member
connections.
Thisprohibition
is
notalways
practical
instructures
suchas fixedandmobile
offshore
drilling
units.

o Minimize
memberstiffness
by usingscantlings
of minimum
thickness.
o Useflatplates
instead
ofcurved
members
wherever
possible.
o Donotusestiffeners,
brackets
orgussets
notspecifically
required
bythedesign
calculations.
Scantlings
andwelding
of allauxiliary
stiffening
should
bethesmallest
required
to suitthedesign
loads.
6-5

-.

o In cruciform
jointsstag-

gerthememberson oppositesidesofthesusceptibleplate.Thismethod
isnotalways
desirable
in
highly
loaded
joints.

s If possible,use lower
strength
material
forthe
member
causing
thestrain
inthethrough-ti.ickness
direction.

6.1.4Selection
ofWeldMaterial

al
To accommodate
moreoftheweldshrinkage
strain
intheweldmetld
welding
consumables
whichmatchtheylese
select,
wherepossible,
strength
rather
thanthetensile
strength
ofthesusc@p\i~le
b?11
calculations
ofthestresses
across
theJointW1le
plate.Detail
usually
havetobeprepared
tojustify
theuseof lowertensi
in
Low-hydrogen
consumables
arerecommended
strength
consumables.
orderto avoidembrittlement
oftheheat-affected
zone.
6-6

6.2 Material
Selection
6.2.1GradeofSteel
Inquiries
madetothemajorshipclassification
societies
indicate
thatthemostsuccessful
andcost-effective
methodof preventing
lamellar
tearing
istheuseofsteels
withimproved
through-thickness(Z-direction)
properties
atsusceptible
connections.
Improved
Z-grade
steels
havebeenusedprimarily
in theconstruction
of
fixedandmobile
offshore
structures.
Thelimited
useto dateof
theZ-grade
materials
inshipbuilding
reflects
thelimjted
occurrenceof lamellar
tearingin theconstruction
of conventional
shipsandthefactthatmanyofthesusceptible
connections
(such
asthegunwale)
havebeeneasily
corrected
bymodifying
thejoint
However,
Bureau
Veritas
has
configuration
andwelding
procedures.
reported
thattwoshipyards
haveputstrakesof special
Z-grade
plates
inthetanktopofLNGships[4]. Theseplates
areusedat
thecritical
intersections
oftheheelsofcofferdams
whichform
thesecondary
containment
boundary
fortheliquefied
gas.
Specifications
forsteelsto be usedin critical
components
of
offshore
structures
have,in
thepast,specified
maximum
sulfur
content,minimum
Z-direction
tensile
strength,
minimumZ-direction
percentage
elongation
andreduction
in area(RAz),andmaximum
allowable
inclusion
content.However,
highyieldandultimate
tensile-strength
valuesin theZ-direction
do notnecessarily
reduce
theriskof lamellar
tearing.The percentage
elongation
measured
byconventional
tensile-test
procedures
alsodoesnot
provide
a reliable
measure
oftearing
susceptibility,
sinceit may
includedeviations
causedby theformation
of smallfissures
Furthermore,
thesmallgauge
adjacent
tononmetallic
inclusions.
length
of samples
takenfromthinplates
makesit verydifficult
to measureelongation
in theZ-direction
withanyacceptable
degree
of accuracy.
Whilesulfur
content
cangivean indication
of thesusceptibility
of aluminum
deoxidized
steels,
itisnotapplicable
tonon-aluminum
treated
steels
wheresilicates
areprimarily
responsible
forreducing
theZ-direction
ductility.The
measure
of inclusion
contentby theprefabrication
ultrasonic
inspection
ofthesteelplates
hasby itself
beeninadequate
for
assessing
theriskof lamellar
tearing.
At present$the
percentage
reduction
in areain theZ-direction
(RAz)isthemostpractical
andaccurate
measure
ofmaterial
susceptibility.
Reports
published
bytheWelding
Institute
showgood
correlation
between
measured
RAzandobserved
incidence
of lamellartearing
[5].RAzisbeingincreasingly
usedbythemajorship
classification
societies
todefine
andapprove
Z-grade
steelsfor
useinshipsandoffshore
structures.
Theserequirements
define

6-7

uptothreeZ gradeplatecategories
whichvaryaccording
totheir
minimum
guaranteed
meanvalueandtheminimum
individual
valueof
example
fromBureauVeritas
rulesforoffRAz. Thefollowing
shoreplatforms
istypical
[6]:
GradeCategory

Individual
Minimum
Guaranteed Minimum
RAzValue
RAz,MeanValue
*

Z15

15%

10%

Z25

25%

15%

Z35

35%

25%

* Meanobtained
fromthreetests.
Duetothecomplex
interrelationships
between
thefactorswhich
cancauseandcontrol
lamellar
tearing,
theselection
ofcandidate
sitesandZ gradecategory
isusually
leftto thediscretion
of
thedesigner
subject
to theclassification
societys
approval
during
design
review.
Whenselecting
materials
forsusceptible
components
in offshore
structures
thefollowing
requirements
should
beobserved:
o Atjoints
connecting
structural
elements
whichareessential
to theintegrity
ofthestructure,
andwhicharesubject
to
highstresses
inthethrough-thickness
direction,
specify
.
steels
witha minimum
guaranteed
meanvalueRAzof 25%anda
minimum
individual
RAzvalueof15%. Examples
ofsusceptible
connections
inmobileandfixedoffshore
structures
were
notedinSection
4.1. It isnotedthatsomemarineclassification
societies
andmajoroilcompanies
require
minimum
mean
RAzvalues
of30%to 35%forcritical
applications
suchas
thenodeplates
inoffshore
drilling
rigs.
s Foraluminum
treated
steels}
thesulfurcontentshouldnot
exceed
0.01%by weight.

Prefabrication
ultrasonic
inspection
of thesteelplatesto
be usedinsusceptible
connections
willnotgivean adequate
indication
ofthematerials
resistance
or susceptibility
to
lamellar
tearing.
Whereultrasonic
inspection
isto be used
to indicate
thenumber
andsizeof laminations
or inclusions,
theplateshould
becontinuously
tested
along
thelinesof a
meshgrid100mm square.Alledgesshould
be inspected
fora
widthequalto1-1/2timestheplatethickness
or 100mm,
whichever
isgreater.

6-8

o Thefollowing
figure,
published
by theAustralian
Welding
Research
Association
[1],maybeusedas a generalguidein
selecting
minimumRAz valuesformaterial
to be used in
structural
elements
susceptible
to lamellar
tearing.The
designer
mayvarytherequired
minimum
meanRAzvaluewith
theimportance
ofthecomponent
totheoverallintegrity
of
thestructure
andthelevel
ofrestraint
of theconnection.
Forextiple,
material
usedinthehighlyrestrained
nodes
connecting
critical
membersof an offshore
drillingrig
shouldhavea minimummeanRAzvalueof 25%. The risk
categories
shownon thefigurearebasedon a qualitative
evaluation
of therecommendations
of bothindustrial
and
marine
references
(seeAppendix
B)rather
thana statistical
forecast
oftheprobability
of a lamellar
tearoccurring.
Similarly,
thelevels
ofjointrestraint
cannotbe equated
withan accepted
quantitative
measureof restraint.The
figure
isa plicable
tosteels
witha minimum
yieldstress
of
f (58,000
40.8kg/mm
PSI)or less. The mean RAz values
correspond
to thoseobtained
using6.4mm diametertest
specimens.
AsnotedinAppendix
Bjthediameter
of thetest
specimen
issignificant
whenquoting
RAzvalues.

30
25
Negligible
Risk
20

Mean
RAz
(%)

15

10
5

Low

Medium
Joint
Restraint

High

Withtheuseofelectroslag
remelting
or calcium-argon-blowing
essentially
allgradesof hull
steelmanufacturing
processes,
steelusedinshipsandoffshore
structures
canbepurchased
with
Typical
lytelectroslag
remelting
improved
Z-direction
properties.

6-9

---

willdouble
thepriceperpoundof anABSAH32gradesteel. The
costofmanufacturing
thesamegradebythecalcium-argon-blowing
process
isconsiderably
lessexpensive,
adding
onlyapproximately
3 centstotheper.
poundcostofthebasicqrade.However+
calciumprocessed
steels
canonlyobtain
a min~mum
RAzof25 percent,
whilesteels
produced
byESRcanobtain
minimum
RAz valuesof 30
Z-grade
steels
areonlyused
percent
ormore.Sincetheimproved
thetotalextracostperstructure
locally
atsusceptible
joints,
fortheimproved
materials
isoftenlessthanthecostof a single
repair
andtheassociated
construction
delays.
6.2.2Product
Type
replacing
rolledsteelplateswithotherless
Wherepractical,
susceptible
typesofsteelproducts,
suchas castings
andfor9ings,willdecrease
theriskof lamellar
tearing.
6.3 Fabrication
6.3.1Layout
andForming
ofSusceptible
Components
Whenfabricating
components
outofplates
susceptible
to lamellar
tearing
thefollowing
practices
canbeusedtoreduce
theriskof
tearing:
.. PlatePosition
- avoid
making
heavyattachment
weldsat the
center
oftheplatewidth,
theextremeedgesof platewith
as-rolled
oruninspected
flamecutedges,andareasof the
platewhereultrasonic
inspection
indicates
heavyconcentraCaution
attheedgesofplatesis wartionsof inclusions.
ranted
bythefactthatthematerial
at an as-rolled
edge
usually
haslessthrough-thickness
ductility
thantherestof
theplate.Theheateffects
of a cutting
torchcanresult
in
thedecohesion
of inclusions
inthesteel
matrix.

Direction
ofRolling
- components
withhighriskof lamellar
tearing,
suchasheavylifting
eyesattached
tothickplates,
should
beoriented
withtheweldaxisatrightanglesto the
primary
rolling
direction
ofthesusceptible
plate.

PlateForming
- thickcold-formed
plates
aremoresusceptible
to lamellar
tearing
and,wherepractical,
should
notbe used
incomponents
requiring
large
welds.

6.3.2Welding
Process
Withtheexception
ofelectroslag
welding
mostconventional
weldingprocesses
aresusceptible
to lamellar
tearing.
Thefrequency
of occurrence
inprocesses
whichutilize
higher
heatinputisless
thanthosewhchhavea relatively
low-heat
input.Thisismost
likely
duetothedeeper
penetration,
reduced
hardness
intheHAZ,
andthesmaller
strain
gradients
encountered
withhigher
heatinput. Recommended
welding
processes
inorderofdecreasing
preference(orincreasing
susceptibility
totearing)
are:
6-10

o Electroslag

Submerged-arc

o Gasshielded
metal-arc
(MIGorC02)andflux-cored
arc
welding
s Manual
metal-arc
- lowhydrogen
electrodes
o Manual
metal-arc
- non-low
hydrogen
electrodes
6,3.3JointPreparation,
Fit-up
andJigging
Otherthantherequirements
conforming
to normalgoodpractice,
jointpreparations
tocontrol
lamellar
tearing
should
reflectthe
improved
jointdesigns
discussed
inSection
6.1.Mainly
thejoint
preparation
shouldprovide
for a balanced
weldwitha fusion
boundary
whichisnotparallel
tothesur~dce
of thesusceptible
plate.Widegroove
angles
whichincrease
distortion
andstrain
should
be avoided
andthedepthoftheweldshouldbe limited
to
thatnecessary
fortherequired
weldthroatthickness.Fillet
weldsorpartial
penetration
weldsshould
begivenpreference
over
fullpenetration
welds.

J!lk

Tightfit-upandheavyjigs
whichinhibitlateralweld
shrinkage
should
be avoided.
flamesoft Theuseof anundressed
Steel cut surfaceor soft-steel
willpermit
conWire wirespacers
traction
of theweldmetal
without
producing
highconcentrations
ofstrain.
Copper
wire should not be used
becauseitmaycontaminate
theweldmetal. Larqeqaps
whichincrease
thevolum~of
weldmetalshouldalsobe
avoided.

0.8nun

to1.6InrJ

6.3.4Welding
Conditions
6.3.4.1Preheat
As notedinSection
3.5,the
useofpreheating
tocontrol
lamellar
tearing
canbebothbeneficial
andharmful.
However,
wheresusceptible
joints
arepreheated
either
tocontrol
lamellar
tearing
or tosatisfy
otherwelding
requirements,
suchas theprevention
of hydrogen
cracking,
thefollowing
considerations
shouldbe observed:

6-11

, Avoidthecreation
of additional
orconcentrated
contraction
strains
byheating
allcomponents
aroundthejointforan
equaldistance
andto approximately
thesametemperature.

A preheat
temperature
of approximately
100Cor greateris
considered
themosteffective.

6.3.4.2Deposition
Rate
Welding
processes
withhighweld-metal
deposition
ratesarepredeposition
ratesdecrease
thenumberof weld
ferred.Thehigher
runsnecessary
tocomplete
theweldwitha corresponding
decrease
inthenumber
of strain
cycles.Sincedeposition
ratesareprimarily
a function
oftheheatinputof theweldingprocess,
the
listandranking
ofpreferred
welding
processes
arethesameas
present~d
inSection
6.3.2.
6.3.4.3Interpass
Temperature
Maintaining
proper
interpass
temperature
isnecessary
to prevent
excessive
cooling
ofpreviously
deposited
weldmetalbetween
runs.
Repeated
heating
andcooling
cycles
mayunnecessarily
increase
the
totalshrinkage
strains.
Recommended
practices
include:
# Donotpermit
theinterpass
temperature
togobelowthepreheattemperature
untilallwelding
onthejointiscompleted.
o AS innormal
welding
procedures,
avoidveryhighinterpass
temperatures
whichmayunfavorably
altertheproperties
of
thesteel.

Allowcompleted
joints
tocoolslowly
andevenly
in orderto
prevent
excessive
thermal
strains.

6.3.4.4WeldSizeandShape
To decrease
theriskof lamellar
tearing
insusceptible
joints
the
following
considerations
ofweldsizeandshapeshouldhe implemented:
. Usetheminimum
weldsizecompatible
withthedesignloads
andstress
distributions
across
thejoint.Oftenexcessive
weldsizesarechosen
arbitrarily
whenthestrength
requirementsacross
thejointareunknown.

Weldswithdeeppenetration
anduneven
shapepermit
thediffusion
ofthecontraction
strains
intomoreofthesusceptiblematerial
andavoid
concentrations.

o Increasing
thelength
ofthelegonthebaseplatewillalso
distribute
thestrains
overmoreofthebasemetal.

6-12

6.3.5Welding
Techniques
6.3.5.1RunSequence
Proper
sequencing
ofweldrunswillhelpto reducethe leveland
concentration
oftheweldshrinkage
strains.
Applicable
methods
include:
o Minimize
thenumberof weldrunsin orderto reducethe
number
ofheatcycles.

Deposit
a layer
ofweldmetalonthesurface
ofthesusceptibleplateprior
tomaking
connecting
runsbetween
thecomponents.Theseinitial
runsshould
bedonein accordance
with
therecommended
procedures
forbuttering
andinsitubutter.inginSections
6.3.5.2
and6.3.5.3,
respectively.

0 Strainconcentrations
in
symmetrically
configured
T
jointscanbe reducedby
depositing
weldrunsinan
alternating,balanced
sequence.

.-

6.3.5.2Buttering
Buttering
consists
ofdepositing
oneormorelayersof lowyield
strength
weldmetaldirectly
on thesurfaceor in a gouged-out
areaof thesusceptible
plate.The purpose
of the buttered
layer(s)
isto accommodate
theweldshrinkage
strains
byspreading
themmoreuniformly
throughthe lowerstrengthweld metal.
Buttering
alsodisplaces
theheat-affected
zoneawayfrom the
susceptible
parent
metal. In ge.neral~
buttering
hasbeenvery
successful
inpreventing
lamellar
tearing
innewweldments
andin
therepair
ofexisting
tears.Points
toconsider
whenusingthe
buttering
technique
include:
6-13

0 The butteredlayer(s)
should
be5 to10mm thick
and extend15 to 25 mm
beyond
eachweldtoe.

...

layeris
9 Whenthebuttered
to beapplied
in a groove,
thegrooveshouldbe approximately
5 mm deepand
extendunder the full
widthofthebuttering.
0

Wherebuttering
is to be
used in placeof steel
with improvedthroughthickness
properties,
the
relative
costsshouldbe
thoroughly
evaluated
prior
to fabrication.
Theyieldstrength
of the
weldmetalshould
be less
thanthestrength
of the
baseplate.
Submerged
arcwelding
with
low-hydrogen
consumables
shouldbe usedwhenever
possible
to obtaingood
penetration
ofthebutteringlayer(s)
andto avoid
thebuildup
ofhydrogen
in
theweld. The levelof
heatinput
shouldalsobe
carefully
regulated.

6-14

6.3.5.3InSituButtering
Thistechnique
isa modification to the run sequence
ratherthan an additional
preweldpreparationas is
conventional
buttering.In
insitu
thefirst
. .. buttering
weldrunsaredeposited
on
thebaseplatepriorto completing
th,?
connecting
runs.
Thismethcl.i
hasprovensuccessful
indiffusing
theweld
shrinkag~
strains
ata negligibleincrease
infabrication
cost
.
6.3.5.4Peening
Peening,
thecontrolled
workingof theweldbeadsby meansof
lightimpact
blowstotheweldsurface
toreduce
residual
tensile
stress,
hasnotproven
successful
incontrolling
lamellar
tearing.
Excessive
peening
cancauselossoftoughness
andcracking
in the
weldmetal.Although
nota viable
method
ofreducing
theriskof
lamell,
tearing,
peening
ingeneral
doesnotincrease
theriskof
tearingin susceptible
material.If employed,
thefirstand
closing
runsshouldnotbe peened.Onereportindicates
that
~~~ning
ofthelastweldrunmaycontribute
to Iamellar
tearing
LIJ.

6.3.5.5Welding
andFabrication
Sequence
Therestraint
level
of a welded
jointis greatly
influenced
bythe
sequence
inwhichtheweldsinthejointaremadeandby thefabrication
seauence
of ad.iacent
com~onents.
The Ollow
ng factors
should
beconsidered
whknpreparing
fabrication
dingschedandwe
ules:

In multi-joint
components,
themoresuscept
blejo ntsshould
bemadefirst.

@ Completel!l
weldsubassemblies
~riorto finalassemblv
. to
limitthe-number
ofcritical
joints.

Minimize
strain
accumulation
bywelding
fromareaof maximum
restraint
tofreeedgesorotherareas
ofminimum
restraint.

Forindividual
joints,
sequence
thewelding
sothatthe level
of restraint
willbeminimized
forthelargest
welds.

o Minimize
thesizeandnumberof tackweldsusedto hold
components
together
during
welding.
6-15

To reducerestraintat
critical
areas,such as
corners,
leavea portion
of theconnection
between
lessrestrained
components
unwelded
untilthecriticalweldsarecompleted,
as indicated
by D.

D approx.
150mm
6.3.5.6Intermediate
Stress
Relief
Theuseof intermediate
heattreatment
toreduce
residual
stresses
hasnotbeenparticularly
successful
incontrolling
lamellar
tearing. Largemembers
whichcannot
be placedin a furnacerequire
localized
heating
whichcan increase
thecontraction
strains
during
cooling.
Heattreatment
maycauseadditional
decohesion
of
thereby
increasing
theindicainclusions
insusceptible
material,
tionsof lamellar
tearsduring
subsequent
ultrasonic
inspection.

6-16

7. DETECTION
ANDREPAIR
OF LAMELLAR
TEARING
AFTERWELDING
7.1 WhenIsNon-Destructive
Testing
forLamellar
Tearing
Recommended?
Normal
weldinspection
requirements
andprocedures
areadequate
wheretheriskof lamel.lar..
tearing
issmall.However,
wherethe
riskof tearingis significant
because
of thecombination
of
welding
procedures
andjointconfiguration,
material
properties,
additional
testmethods
forthedetection
of lamellar
tearsshould
Formembers
andjointswhicharecritical
to the
be employed.
suchas thenodejointsand
overall
integrity
ofthestructure,
deckto legconnections
ofoffshore
jacket
typestructures,
supplemental
testing
isrecommended.
7.2 WhichNon-Destructive
Testing
Methods
AreApplicable?
Standard
non-destructive
testing
methodssuchas visualinspection,dyepenetration
andmagnetic
particle
inspection
aresatisfactory
forsurface
cracking
butnotforsub-surface
tears.Radiography
isgenerally
notpractical
forthedetection
of sub-surfacetearssincetheinclusions
intheplatecanmaskdefectsand
it isdifficult,
ifnotimpossible,
todirect
radiation
alongthe
tearaxis.Of alltheconventional
non-destructive
testing
methods,
ultrasonic
testing
isthemostpractical
andwidely
usedtechniquefordetecting
lamellar
tears.
7.3 Ultrasonic
Testing
(UT)OfWelded
Joints
Thepulse-echo
ultrasonic
testing
technique
is basedon the interpretation
ofreflected
ultrasonic
wavesfromthefracture
surfaceto detectlamellar
tears.The instrument
probeis both
transmitter
andreceiver.
Theultrasonic
beamisreflected
either
bythefaceplateopposite
theoneonwhichtheprobeis applied
(bottom
echo),
orpartly
atleast
by anareaof lamel
lartearing
or anyotherdefect
ofthemetal(flawecho).WhileUT methods
normally
locate
plateandwelddefects
whentheplatesurfaces
are
flatandreasonably
freefromloose
material,
it is oftendifficulttodistinguish
truelamellar
tearsfrominclusion
bandsand
otherformsofcracking.
Thematerials
thataremostsusceptible
to lamellar
tearing,
suchasthickplates
withhighconcentrations
of nonmetallic
inclusions,
contain
thetypeof defects
whchcan
attenuate
thesignals
andmakeinterpretation
difficult.Misinterpreted
ultrasonic
indicators
canoftenleadtounnecessary
and
costly
repairs.

i-

Compreslion
Wave
7-1

Bothcompression
wave and
shear(orangle)probeUT
methods
arecapable
of accuratelylocatinglamellar
tearing.However,
theusefulness
of compression
wave
techniques
islimited
toT or
corner
joints.

ShearorAngleProbe

Probefrequencies
of 2 MHz
areconsidered
suitable
for
the rapidlocation
of true
resolution
tears.Sufficient
to obtain
goodidentification
of thecharacteristic
stepped
surfaceof thetearcan be
obtained
bytheuseof4 to5
MHzprobes.Theuseofhigher probe
frequencies
together
withhighequipment
gainsettingsarereportedto increasethe likelihoodof
erroneous
indications
of lamellartears.

Details
oftechnique
andequipment
fortheultrasonic
detection
of
lamellar
tearing
areessentially
thesameas forthenominal
UT
inspection
ofwelds.Specifics
may be obtained
in theAmerican
Bureau
ofShipping
RulesforNon-destructive
Inspection
of Hull
WeldsorASTME164- Standard
forUltrasonic
Contact
Inspection
of
Weldments.
POINTS
TO NOTE

Ultrasonic
testing
should
bespecified
forhighly
restrained
welded
connections,
critical
totheintegrity
of thestructure,wheretheriskof lamellar
tearing
issignificant.

s Thereliability
oftheultrasonic
testing
method
depends
toa
greatextent
ontheability
andexperience
of theoperator.
Personnel
responsible
forconducting
ultrasonic
testsshould
befamiliar
withtheequipment
beingusedandbe properly
qualified
bytraining
andexperience
toperform
thenecessary
andto interpret
andevaluate
indications
in
calibrations,
In addition
accordance
withthetermsofthespecification.
to beingqualified
inaccordance
withtherequirements
ofthe
American
Society
ofNon-destructive
Testing
Publication
TClA- Supplement
C,Ultrasonic
Testing
Methods
orotherrecogthepersonnel
should
preferably
haveexperinizedagencies,
encein,orbe abletodemonstrate
ability
to identify,
lamellartears.

Ultrasonic
testing
should
be performed
whenallweldingon
thejointis completed
andmaximumrestraint
is reached.
Sincelamellar
tearing
hasbeenreported
to occurup to 36
hoursafter
thecompletion
of allwelding
andthecoolingof
thecomponent,
finalultrasonic
inspection
ofcritical
joints
should
beperformed
nosooner
than36hoursafterweldingon
thejointiscompleted.

7-z

Allcritical
welded
joints,
suchasthenodejointsanddeck
to legconnections
ofoffshore
structures,
shouldbe ultrasonically
inspected.

Acceptance
standards
forlamellar
tearingshouldbe established
sothatclusters
of inclusions
anddensemicrostructuralbandswhichappear
asdefect
indications
donotconstituterejectable
defects.Sincetheremovalandrepairof
minornon-critical
tearsmay do moreharmthangood,the
acceptance
criteria
should
consider
thefunctional
requirementsof thecomponent
or jointas wellas thepractical
levelofworkmanship
whichexperience
indicates
canbe obtained
inweldments
of a giventype.At present,the
welding
acceptance
criteria
ofmarineclassification
societies
and
national
codesandspecifications
do notcontainspecific
acceptance
standards
forlamellar
tearing.

@ Inorderto distinguish
lamellar
tearsfrompre-existing
defects(largeinclusions,
laminations,
etc.),the base
material
intheareaof theweldshouldbe ultrasonically
inspected
priortofabrication.
Theseinspections
shouldbe
methodically
performed
andrecorded
usinga gridsystemfor
locating
checkpoints.
s Ultrasonic
indications
of lamellar
tearing
exhibit
a characteristic
multiple
peaksignal
anda rapidchange
in depthas
theprobeismoved.Thesecharacteristics
helpto differentiatelamellar
tearsfromothercracks,
laminations
or backwallreflection.
7.4 Repair
ofLamellar
Tears
Therepair
of lamellar
tearscanbedifficult,
timeconsuming
and
costly;and,inthecaseof highlyrestrained
connections
the
repair,
canbemoredetrimental
thantheoriginal
weld. The increased
riskof lamellar
tearing
during
repair
ispartially
dueto
thegreater
overall
restraint
of thecompleted
structure.The
mechanical
andthermal
strains
induced
bytherepairweldingcan
causetearing
tooccurata greater
depthbelowtheoriginal
weld
fusion
line.Thebasic
methods
ofrepair
are:

gouging
outofthetearandreplacement
withweldmetal

cuttingoutthedefective
materialand replacingwith
material
withimproved
through-thickness
properties.
This
procedure
isoftenaccompanied
bymodifications
to theweldingprocedures
orjointdetails
inorderto avoidtheconditionswhichprecipitated
theoriginal
tear.

o providing
additional
structural
members
tocarrytheanticipatedservice
loadsacross
thejoint.

7-3

Priorto selectingrepair
procedure,
itshouldbe verified
that
thediscontinuity
isa lamellar
tear. Thenumber,location
and
extent
ofothertears,
ifany,should
alsobedetermined.
If the
crackisinfacta lamellar
tearandexceeds
thespecified
acceptancestandards,
allthefactors
whichcancontribute
tothedevelopment
of lamellar
tears,
namely
material
susceptibility,
joint
configuration
andwelding
procedures,
should
bereviewed.
Merely
gouging
outthetearandrewelding
using
theoriginal
jointgeometryandwelding
procedures
underconditions
ofpossibly
evenhigherrestraint,
willprobably
resultin new lamellar
tears. For
thisreason,
itis.often
moreeconomical
toreplace
a component
or
jointrather
thantorepair
it. Alternately,
additional
support
members
canbeprovided
to reducethe serviceloadsacross
the
joint,
thusreducing
thecriticality
of thejoint.
Wheregouging
andrewelding
is judgedacceptable)
thefollowing
procedures
should
beused:

Remove
damag~d
material
using
flamegouging,
arc-air
gouging
or grinding.
TheAustralian
Welding
Research
Association
reports
thatflamegouging
isslightly
preferred
to arc-air
gouging
duetothelessintense
thermal
gradients
andlower
thermal
induced
stresses
offlamegouging
[1].On theother
hand,a fabricator
withLTrepair
experience
reports
thatgas
gougingdevelops
toomuchheatin the weld causingthe
lamellar
tearto propagate
[7]. Herecommends
using
onlyarcairgouging
withnopreheat.
Grinding
isslowandimpractical.

0 Flame
Gouging
- the first
passofthegouging
torch
should
cutacross
theend
of theplaneoftearing
at
such an attitudeas to
minimize
thetendency
of
the tear to open up as
well as toreleaseany
tensile
stress
acrossthe
tear. The base of the
gouge should be well
rounded.
Subsequent
passesshouldremove2 to 3
mm of thematerial
below
the original
tearand3 to 4 mm beyondtheendsof thetear.
Deeper
andwidergouging
willincrease
the amountof rewelding
witha corresponding
increase
intheriskofnewtearing.
In some
cases,complete
release
ordisassembly
oftheweldedjointmaybe
required.

7-4

o Air-Arc
Gouging
- thefirst
arc-airgougeshouldbe
madeadjacent
to thetear
andextend
3 to4 mm beyond
theendofthetear.Additionalpassesshouldbe
madeparallel
to the initialgougeuntil
material
2
to3mm belowtheoriginal
tearhasbeenremoved.

Regardless
ofthegouging
method,
allfinalsurfaces
should
be
lightly
ground
andinspected
using
magnetic
particle
or dye
penetration
testing
methods
priortowelding.

Buttering
layersof lowyieldstrengthweld metalare
generally
considered
essential
priorto makingconnecting
welds.Thewelding
procedures
giveninSection
6.3should
be
employed
inordertoreduce
strains
inthethrough-thickness
direction.It is recommended
thatthe minimumpreheat
temperature
bemaintained
for8 hoursafterallweldingon
thejointiscomplete.

Therepaired
jointshould
be thoroughly
re-examined
in accordance
withtherequirements
forultrasonic
testinggiven
inSection
7.3.

-NOTE

During
therepair
process,the
actual
location
ofthelamellar
tearshould
benotedandcompared
withthatindicated
by the
ultrasonic
testinordertovalidate
thetesting
procedures.

7-5

8. TESTSFORDETERMINING
THESUSCEPTIBILITY
OF
STEELPLATES
TOLAMELLAR
TEARING
During
thelastfifteen
years,
manytestshavebeendeveloped
fordetermining
thesusceptibility
ofdifferent
gradesof steelplateto lamellar
tearing.
Withvarying
success,
thesetestsattempt
to provide
reproducible
results
whichcorrelate
wellwithknowncasesof lamellar
tearing.Each
testincorporates
different
combinations
oftheactual
jointdesign.,
componThe
entandjointrestraint,
welding
procedures,
andmaterial
properties.
difficulty
in devising
a single,universally
acceptable
testprocedure
reflects
thefactthat,
whileitispractical
torepresent
tosomedegree
the
actual
geometry
andfabrication
procedures,
thetestmaterial
may notbe
representative
ofthematerial
usedintheproduction
joint.Thethroughthickness
properties
of individual
platesarevariable
because
of the
irregular
distribution
ofnonmetallic
inclusions
in the plate.Themore
important
testsarelisted
belowbytype:

Nondestructive
Tests(without
welding)
(i) Ultrasonic
inspection
ofsusceptible
steelplates
priorto
fabrication
[8] - of limited
usefulness
indetermining
the
susceptibility
ofsteelplates
to lamellar
tearing;
useful
fordistinguishing
lamellar
tearsdetected
duringpostwelding
UT inspection
frompre-existing
defects.

s Destructive
TestsWithout
Welding
~
i-

(i) Through-thickness
tensile
test- practical
test;RAzgives
goodcorrelation
withknownincidence
oftearing.
(ii) Slice-bend
test[9]- groundsurfaceis examined
forthe
sizeandlocation
ofcracks
after
thespecimen
issubjected
to increasing
surface
strains;
practical
testwhichproducesrealistic
tears.
(iii)Special
notchedtensile
test[10]- compares
theshear
fracture
loads
onandacross
thelaminar
plane;
forpreliminaryscreening
ofsusceptible
material.
(iv) Charpy
V-notch
impact
test[11]- limited
usefulness;
can
be usedonlyforplates
55mm thickorgreater.
(v) Microscopic
countofnonmetallic
inclusions
- usesstandard
metallurgical
examination
procedures
to determine
number,
size,shapeanddistribution
ofnonmetallic
inclusions;
impractical
forproduction
test;moresuitable
forresearch.

8-1

Destructive
TestsWithWelding
(i) Welding
Institute
windowtest[12,13]- specimen
of the
susceptible
basematerial
isinserted
through
a rectangular
holeorwindow
ina restraining
plateandwelded
to forma
cruciform
joint;
observed
tearing
canoftenbe induced
by
rootcracking
notlamellar
tearing;
restraint
levelvaries
withjigandtestplatethickness;
notpractical
forproduction
use.
to 45 or 60
(ii) Cranfield
test[14]- a stemplateisbeveled
degreesandmultirun
weldedto thesteelsampleto be
tested;
severe
testconditions
areonlycapable
of identifyingmaterials
witha highresistance
to tearing;
mixed
correlation
withcasehistories.
(iii)Short-transverse
notched
bendtest[15]- complexandexpensive
testmoresuitable
toresearch.
(iv) Lehigh
restraint
test[2]- suitable
forquantitatively
measuring
lamellar
tearing
susceptibility
underlaboratory
conditions.
(v) H-type
restraint
test[16]- research
oriented
test.
(vi) Testsofprototype
welded
joint- difficult
to duplicate
theexactrestraint
levels
andweldingcondition
present
doesnottakeintoaccount
the
during
actual
production;
variability
ofproperties
inthethrough-thickness
direction;notpractical
forproduction
testing.

With
theincreasing
useofRAzbythemajorshipclassification
societiestodefine
andapprove
Z-grade
steels
forthe.use
inshipsandoffshore
structures)RAz,
asmeasured
bythrough-thickness
tensile
test,is quickly
becoming
thestandard
testfordetermining
thesusceptibility
of steel
plates
usedinthemarine
industry.
Todate,a
standard
RAztestprocedure
hasnotbeenadopted
bythedifferent
classification
societies.
Whilesimiallofthecurrently
published
RAztestprocedures
differ
larintheory,
slightly
intherequired
dimensions
andconfiguration
oftensile-test
speciandsteelmaker
should
follow
thespecific
requirements
mens. Fabricators
ofthecognizant
classification
society
ruleswhilekeeping
inmindthefoladditional
considerations:
lowing
@ TheRAzacceptance
limits
showninSection
6.2areappropriate
for
materials
withyieldstrengths
lessthan40.8kg/mm2
(58,000
psi).

Becauseof thesmallcross-sectional
areaof materialbeing
tested,
theuseofsixspecimens
willgivea statistically
better
sampling
thanthethreesamples
required
bysomerules.

o Theresults
canbegreatly
influenced
bysinglelargeinclusions
or clusters.
8-2

o RAzvalues
arestrongly
influenced
by thediam.qter
of thetest ,
specimen.
Smaller
diameter
specimens
areinfluenced
by thesize
andposition
of largeinclusions
whichdecrease
minimumvalues
whileincreasing
maximum
values.Larger
diameter
specimens
have
morelateral
restraint
whichlowers
themeanvaluesandreduces
scatter.
Accordingly,
caremustbe exercised
in comparing
RAz
values
obtained
fromdifferent
sources
ortestprocedures.

8-3

9. REFERENCES
1. Australian
Welding
Research
Association,
Control
of Lamellar
Tearing,
AWRATechnlcal
Note6,April1976.
2. Oates,
R.P.,andStout,
R.D.,AQuantitative
Weldability
Testfor
Susceptibility
toLamellar
Tearing,
Supplement
toWelding
Journal,
November
1973.
3. McEnerney,
J.W., Assessment
of Lamellar
Tearing,
Oak Ridge
National
Laboratory
Report
ORNL/NUREG/TM-171,
OakRidge,
Tennesse@,
March1978.
4. Bureau
Veritas,
Lamellar
Tearing
(Non-confidential
Technical
Note
ofGeneral
Information),
August
1974.
5. Farrar,
J.C.M.,TheProblem
ofLamellar
Tearing.
TheRelationship
Between
ShortTransverse
% RA Values
andtheIncidence
of Lamellar
Tearing,
inPhaseIVof Investigations
intoLamellar
Tearing- A
Compendium
ofReports
froma Sponsored
Research
Programmed,
The
Welding
Institute,
Cambridge,
England,
March1975.
6. Bureau
Veritas,
Rules
andRegulations
fortheConstruction
and
Classification
ofOffshore
Platforms,
1975.
7. Smith,
R.K.,Fabrication
Techniques
toControl
Lamellar
Tearing,
Australian
Welding
Research
Association,
Symposium
on Control
of
Lamellar
Tearing,
April8,1976.
8. American
Society
forTesting
andMaterials,
Standard
Specification
forStraight-Beam
Ultrasonic
Examination
of PlainandCladSteel
Plates
forSpecial
Applications,
ASTMA578.
9. Drury,
M.L.,andJubb,
J.E.M.,
Lamellar
Tearing
andtheSliceBend
Test,
Welding
Journal,
Welding
Research
Supplement,
February
1973.

10. Kanazawa,
S.,Yamato,
K.,Takeshi,
Y.,et al,OntheAssessment
of

Lamellar
Tearing
Susceptibility
ofSteel
Plate,
IIWDocIX-840-73.

11.

American
Society
forTesting
andMaterials,
Standard
Methodsfor
Notched
BarImpact
Testing
ofMetallic
Materials,
ASTME23.

12. Nippon
SteelCorporation,
Engineering
DataBrochure
No.QT21248.9,
NSCs
SteelPlates
forMarineStructures,
Section
5, Lamellar
TearResistant
Steels,
1974.

13. Farrar,
J.C.M.,andDolby,
R.E.,AnInvestigation
intoLamellar
Tearing,
MetalConstruction
andBritish
Welding
Journal,
February
1969.

9-1

D.N.,Lamellar
14. Elliott,
Tearingin Multi-Pass
FilletJoints,
Welding
Journal,
September
1969.

J.C.M.,Inclusions
15. Farrar,
andSusceptibility
toLamellar
Tearing

Supplement
toWelding
Journal,
August
of Welded
Structural
Steels,
1974.

16. Nishio,Y.,Yamamota,
Y.,Kajimota,
K., andHirozane,
T.,On

Lamellar
Tearing
inMultirun
Fillet
Welds,
Mitsubishi
HeavyIndustries,
Ltd.Technical
Review,
Vol.9, No.3,October
1972.

9-2

10. BIBLIOGRAPHY
American
Institute
ofSteelConstruction,
Commentary
on HighlyRestrained
Welded
Connections,
AISCEngineering
Journal,
ThirdQuater,
1973.
Araki,
M.,Nagao,
T.,andNarasaua,
H.,Cracking
inMultirun
Fillet
Welds,
NipponKokanTechnical
Report
- Overseas,
June1974.
Arita,
Yukio,
Kajimoto,
Katsuya
(Mitsubishi
HeavyIndustries,
Ltd.),and
Ohba,Kenui(Nippon
SteelCo.,Ltd.),
TheStudyof Lamellar
Tearingin
Offshore
Structures,
Offshore
Technology
Conference,
Houston,
Texas,May
2,1972.
Baker,
T.J., Gove,K.B.,andCharles,J.A., Inclusion
Deformation
and
Preprints
of a MetalsSociety
Toughness
Anisotropy
inHotRolled
Steels,
Meetinq:
Directionality
ofPro~erties
inWrouaht
Products.
27-28thNov.

1974,vol. 1.

Banks,
E.E.,Assessment
ofSteel
Susceptibility
toLamellar
Tearing,
Ius1ar
tralian
Welding
Research
Association,
Symposium
on Controlof Lame
Tearing,
April8,1976.
Bernard,
G.,Grumbach,
M.,andMoliexe,
F.,Inclusions
inSteelPlatesand
Mechanical
Anisotropy,
Preprints
of a Metals
Society
Meeting:
Direct
onalityofProperties
InWrought
Products,
27-28th
Nov. 1974,
vol. 2.
Chapman,
J.A., Clark,
A.,andKirkwood,
P.R.,Steels
forNorthSeaStrucpresented
atWelding
inOffshore
Structures,
Newcastle,
UK,1974.
tures,
Charleux,
J.,Metallurgical,
Welding
andInspection
Problems
Raisedby the
Construction
ofSteelOff-Shore
Platforms,
Technical
Bulletin
of the
Bureau
Veritas,
February,
1976.
Croll,
J.E., Material
Factors
Affecting
Lamellar
Tearing,
Australian
WeldingResearch
Association,
Symposium
on Controlof Lamellar
Tearing,
April8,1976.
Dolby,
R.E.,TheWeldability
of LowCarbonStructural
Steels,
Welding
Institute
Resource
Bulletin,
18(8):
pp.209-16,
August,
1977.
Farrar,
J.C.M.,andDolby,R.E.,Lamellar
Tearing
inWelded
SteelFabrication,
TheWelding
Institute,
Publication
No.SBN85300068,
1972.
Farrar,
J.C.M.,Dolby,
R.E.,andBaker,
R.G.,Lamellar
Tearingin Welded
Structural
Steels,
Welding
Journal,
July1969,pp274s-282s.
Frost-Drury,
M.L.,Quality
Assurance
andSpecifications
forLamellar
Tearing,Australian
Welding
Research
Association,
Symposium
on Controlof
Lamellar
Tearing,
April8,1976.

1o-1

Ganesh,
S.,andStout,
R.D.,Effect
ofWelding
Variables
onLamellar
TearingSusceptibility
inLehigh
Test,
Welding
Journal,
March1977,pp.39:
78s-87s.
Hancock,
J.,Cowling,
H.C.,TheEffects
ofDirectionalM.J.andMackenzie,
ityontheMechanisms
ofHoleGrowthandDuctileFailureinThreeHigh
Preprints
of a Metals
Society
Meeting:
Directionality
of
Strength
Steels,
Properties
inWrought
Products,
27-28th
Nov.1974,Vol.1.
Welding
Research
Council,
Bulletin
No.
Lamellar
Tearing,
Jubb,J.E.M.,
168,December
1971.
Lay,M.G.,TheStructural
Significance
of Lamellar
Tearing,
Australian
Welding
Research
Association,
Symposium
onControl
ofLamellar
Tearing,
April8,1976.
McCombe,
A.A.F.
andRatnapuli,
R.C.,TheEffectof Slabto
Little,
H.H.,
PlateDeformation
RatioonProperty
Directionality
of E.S.R.andContinPreprints
of a Metals
Society
Meeting:
Directionality
uously
CastSteels,
ofProperties
inWrought
Products,
27-28th
Nov.1974,Vol.1.
ofTesting
Direction
ontheMechanical
PropMorrison,
W.B.,TheInfluence
erties
ofWrought
Steel,
Preprints
of a MetalsSociety
Meeting:
Directionality
ofProperties
inWrought
Products,
27-28th
Nov.1974,Vol.2.
ofSteelPlatesinThickness
DirecH.,Properties
Nakao,
N.,andOhsuzu,
tion,
Nippon
KokanTechnical
Report-Overseas,
Decembei
1972.
Steel
Selection
forOffshore
Peterson,
Marvin
L. (Continental
OilCompany),
Journal
ofPetroleum
Technology,
March1975,pp.274-282.
Structures,
of Lamellar
Tearingby Ultrasonic
Pope,C.W.,andCard,K.J., Detection
Testing,
Australian
Welding
Research
Association,
Symposium
on Control
ofLamellar
Tearing,
April8,1976.
Tearingin PlateSteels(A Literature
Survey),
Porter,
L.F.,Lamellar
Technical
Report
1O-F-OO2(O18-5),
Research
Laboratory,
United
States
Steel
August
1975.
Rodgerson,
P.,Rowntree,
G.,andSchofield,
R.,Aspects
pf Di~@ctionali.ty
Preprints
of a Metals
Society
Meeting:
Dlrectlonallty
inPipeline
Steels,
ofProperties
inWrought
Products,
27-28th
NOV.1974,VO1.2.
andLamellar
TearSantini,
W.,Repetto,
G.,andPosini,
G., ColdCracking
oftheProbinginWelded
Carbon-Manganese
Structural
Steels.Discussion
lemsandPossible
WaysforPrevention
withParticular
Regardto Ship
Transactions
JapanWelding
Society,
Vol.2, No.2,
Building
Applications,
April1951(Welding
Research
Abroad,
October
1972).

.,

Scott,
R.,Directionality
of Mechanical
Properties
in HotRolledSteel
Plates,
Preprints
of a Metals
Society
Meeting:
Directionality
of Proper27-28th
Nov. 1974,
Vol.2.
tiesinWrought
Products,

10-2

Lamellar
Tearing
andMarine
StrucTakeshi,
Y. (Nippon
SteelCorporation),
tures,
Welding
andMetalFabrication,
December
1975,pp740-746.
Thomson,
A.D.E.,
Christopher,
P.R.,andBird,
J.,Short
Transverse
PropertiesofCertain
HighStrength
Steels,
J. Eng.
Ind.,
Transactions
ASME,90.
627-735
(Nov.
1968).
Wilson,
A.D.,TheEffect
ofAdvanced
Steelmaking
Techniques
on the InclusionsandMechanical
Properties
ofPlateSteel,
presented
atSymposium
on
Toughness
Characterization
andSpecifications
forHSLAandStructural
Steels,
AIMEAnnual
Meeting,
March6-10,1977.
Wilson,
A.D.,TheInfluence
ofThickness
andRolling
RatioontheInclusion
Behavior
inPlateSteels,
presented
attheASM1977rflaterials
Conference,
Chicago,
Illinois,
October
25-27,
1977.
Wilson,
W.G.,Minimizing
Lamellar
Tearing
by Improving
Z-Direction
Duci~ity,Welding
Journal,
Vol.12,No.1, 1975,pp.274-282.
Weld,-Gunnar
andKristoffersen,
Thorbjorn
(DetNorske
Veritas),
Development
ofMethod
forMeasuring
Susceptibility
of SteelPlateto Lamellar
Tearing,Offshore
Technology
Conference,
Houston,
Texas,
April29-May2,
1973

Research
Council
Bulletin
No.232Through
Thickness
ProperNote:Welding
tiesandLamellar
Tearing
(ABibliography),
by D.H.SkinnerandM.
Toyama
lists
over400papersrelating
to through-thickness
or Zdirection
properties
andlamellar
tearingpublished
priorto April
1977.

-.

10-3

APPENDIX
A

MECHANISM
OF LAMELLAR
TEARING

Reprinted
withthepermission
ofthe
Australian
Welding
Research
Association
fromAppendix
A ofAWRATechnical
Note6,

References
forAppendix
A arelisted
atthe
endoftheAppendix.

A-1 BASIS
Lamellar
tearing
occurs
whentensile
stresses
andstrainsapplied
to
thematerial
inthethrouqh-thickness
(Z)direction
ofthesteelexceedthe
ability
ofthesteel
towithstand
thes~stresses
andstrains,
andthereare
sufficient
inclusions
present
toresult
inthetypical
LTfracture.
Alternatively
stated,
lamellar
tearing
occurs
whenthesteelproperties
areinadequate
towithstand
thestresses
andstrains
intheZ-direction.
A-2 STEELPROPERTIES
.1 Anisotropy.
An
inherent
feature
ofwrought
steelproducts
is thedifferences
whichcanoccurinproperties
indifferent
directions,
i.e.the
steel
isanisotropic;
although
formostapplications
it canbe treatedas isoisbrought
about
bytherolling
or forgingoperatropic.Thisanisotropy
tionswhichmainlyflatteninclusions
andmaymodifythemetallurgical
structure
ofthesteel.
Mechanical
properties
in theZ-direction
.2 Mechanical
Properties.
to a varying
degree.Thetensile
usually
arereduced
strength
andductility
(asmeasured
byreduction
of areaor% elongation)
aremostaffected.In
exceptional
casestensile
strength
maybereducedby 30%andductility
to
ishardlyaffected
exceptwheninvirtually
zero.Yieldorproofstress
clusions
aregross.
.3 Cause.TheZ-direction
properties
arereduced
asa result
of:
(a) increase
inmaximum
andaverage
length
andwidthof inclusions.
(b) increased
number
of inclusions
andcloser
spacing.
(c) increased
alignment
of inclusions
onplanes.
Thiseffecthasbeenestablished
by theoretical
work usingfracture
mechanics
andproven
bytesting.
Theseproperties
mayalsobereduced
by:
(d) hydrogen
fromwelding
operations
orparent
metal;
andasmorerecently
reported
(RefAl)by:
(e) thetemperature
andstresscyclesleading
to strainageingor
similar
effects;
and
(f) priorcoldworkleading
toexhaustion
of ductility
andpossible
strain
ageing
effects
inthematrix
between
inclusions.

A-2

A-3 THROUGH-THICKNESS
TENSILE
STRESSES
ANDSTRAINS
andstrains
result
from:
.1 Cause.Thesestresses
ng,
(a) weldandparent
metalcontraction
oncooling
after
weld
(b) restraint
of thejointcomponents,
ie external
res stanceto
contraction,
(c) otherthermal
or loadinfluences,
eg.uneven
preheat;
andfrom
(d) thePoisson
effect
dueto highlongitudinal
stresses
alongthe
weldresulting
inhightriaxial
stressing.
ofMechanical
Strains.Tensilestresses
.2 Location
andstrainsvary
across
andalong
thewelded
Joint,
andmaximum
values
maybe located:
(a) atextreme
weldrunssuchasundertherootor toe (lastrun)as
inFig.A-la.Loaddiffusion
andstrain
across
theplatewillbe
concentrated
intheseareasbecause
of thenotchor stressconcentration,
particularly
if anybending
momentis involved,
as
withsingle
sidedorunbalanced
welding.
(b) inthecentral
plateareaasinFigA-lb,oftenwhenthisismore
susceptible
thannearthesurface.
(c) atweldormaterial
defects
or poorweldor penetration
shape.

Thesecanactassevere
stress
andstrain
concentrators
- seeFigs
A-lc-e.

IB
1~~~
.,

..---:

Z-strain
isgreater
at:
A whencomponent
(2) isfreeto rotate
B whencomponent
(2) isprevemed
fromrotating
Inpractice
mostLToccurs
nearfinalweldrun(Ref.K3)

I .=(3
~11
,1-===,
-- 11
,,,:!,l
iA
1
~1
~
a)

~~

e HAZS.
highhardness,
n

angle
ntrates
strain

.
Ieterootpenetration
slopeofHAZ
ratesstrain

Smallunderbead
cold
inhigh
tration

d)
ofmaximum
strain
inZ-direction.
Fi9 A-l Location
A-3

(d) justoutside
thevisible
HAZ,particularly
wheretheHAZboundary
isparallel
totheplatesurface.
Thispartof theparentmetal
almost
always
haslower
yieldstress
thantheHAZorweldmetalat
thetemperatures
whereLToccurs.
Weldandparent
metalcontraction
increases
astheweldwidth,volume
andnumber
of runsincrease.
A-4 FRACTURE

It is believed
thatfracture
or tearingoccursin thestagesshown
diagrammatically
inFig10 (from
AWRATechnical
Note6 datedApril1976),
ie:,

Principal
tensilestrength
(inZ direction)

Elastic
stress
concentration
at tips
,.

J-W:,*;.
f

Cohesion
between
steel
andinclusion

Decohesion
at inclusion
,&+v+&E7///~

/
tfeavvt31astic
strairiirig
of
ligament

plastic
zoneat
inclusion
rips

iecomplete
fracture
throuah
inclusions
andli~aments
~
,,
~&, ...
~
Heavyplastic
shearing
between
terraces

ShearWall
(located
atterraces
atdifferentlevels)

d
/
Ductile
fractureof
ligament
between
inclusions

Fig10 Mechanism
ofLT.

certainly
withinthe
.1
Onfirstencountering
significant
stress
- almost
decohesion
occurs
at the inclusion/matrix
interface.
The
elastic
range,
stress
required
willbedependent
onthetypeandshapeof inclusion
andthe
microstress
system
developed.
.2 Atthesametimeatthetipsorendsof inclusions
andotheradjacent
defects,
plastic
deformation
occurs
firstatthelarger
inclusions
or those
so closely
spaced
thatthereisinteraction
between
thetwo,i.e.
wherethe
spacing
islessthanabout
thesizeofthelarger
inclusion.
.3 Onfurther
straining
dueto furthercoolingor mostlikelydueto
between
the inclusions
becomesfully
further
welding
runs,the ligament
plastic
andthevoidsatinclusions
increase
in sizegenerally
by ductile
tearing.
.4 Withadditional
straining,
theextended
voidslinkup in planesof
iewheretheinclusions
arealigned;
andtheterrace
is
general
weakness,
oftheterrace
surface.
formed
- see-Fig
A-2. FigA-3showsthenature

A-4

FigA-2 Linking
upofvoidstoformterraces
x50.

FigA-3 Fracture
surface
ofterrace
showing
duct~le
fracture
andinclusions
XIOO.

A-5

.5 Slightly
further
straining
connects
theterraces
ondifferent
levels
by
virtually
togivecomplete
separation.
ductile
shearing
ofthewalls,
See
tearingaresmaller
~han
in
FigsA-4andA-5. Shearwallsin lamellar
mechanical
tests.
.6 Ifthematerial
forming
theligament
between
inclusions
hasexception~e.low
plane-strain
fracture
toughness
(Klc)or low
allypoorproperties,
critical
crackopening
displacement
(&c ),theseareasmay showareasof
brittle
fracture.
accentuates
local
.7 Theroleof hydrogenis notclear,butprobably
to whichhydrogen
preferentially
difstress
atvoidtips,iea position
fuses.Hydrogen
hasbeenshownexperimentally
tohavea greater
effecton
Thisismostlikely
dueto the
LT insteels
withhigher
carbon
equivalent.
increased
riskofunderbead
(orcold)cracking,
whichevenif on a microscalemaytrigger
offLT.
theconnection
system
musthaveconsiderable
strain
.8 ForLTto occur,
energy
andbe abletotransfer
thistotheareaoftearing.

closely
FigA-d Smallsheaxwalllinking
X150.
al,igr[ed
terraces

A-6

FigA-5 Electron
scanning
micrograph
giving
3--dimensional
viewofshearwallx200.

LISTOFREFERENCES
FORAPPENDIX
A
Al Takeski,
Y.,Lamellar
Tearing
andMarine
Structures,
Welding
and
MetalFabrication,
December,
1975.

A-T

APPENDIX
B

ORIGINS
OFINCLUSIONS

Reprinted
withthepermission
ofthe
Australian
Welding
Research
Association
fromAppendix
B ofAWRATechnical
Note6,

References
forAppendix
B arelisted
atthe
endoftheAppendix,

B-1 GENERAL
LToccurs
inparent
material
andhenceis greatlyinfluenced
by the
LTisalsohighly
directional
sensitive
anddependent
material
properties.
onthrough-thickness
direction
properties.
Theseproperties
andtheriskofLTdepend:
(a) primarily
on inclusions
insteel;
butalsoon
(b) thesteel
matrix
itself.
Theeffect
of inclusions
onproperties
hasbeennotedin2.3of Appentheiroriginin steelmaking
and
dixA. Thevarious
typesof inclusion,
modification
in rollingarediscussed
below,together
withtheuseof
various
material
properties
inassessing
thesusceptibility
ofsteel
toLT.
B-2 INCLUSIONS
Inconstructional
steelplates
andsections,
typicalnonmetallic
incisions
are:
.1

Sulphides,
mainly
manganese
sulphide
(MnS)
whichmaybe:
(a) TypeI, i.e.
ellipsoidal
orspherical
inshape.Theseareusually
smallinsizeandareseldom
responsible
forLT.
(b) TypeII,i.e.
flattened
lamellar
shapeas in FigB-1a. Theseare
oftena majorfactorin LT, particularly
when elongatedin
stringers.Theyarepredominant
in aluminum
treatedsilicon
killed
steels.

Othersulphides
include
therare-earth
metal(REM)
sulphides
andoxysulphides
whicharefoundin steelsspecially
treated
to givehighRAz.
Thesesulphides
arepredominantly
spherical
inshape- seeFigB-lb.
.2 Silicates,
metal-silicon-oxygen
compounds.
Inthestringer
form,asin
FigB-lc,thesehavea dominant
effect,
particularly
whenRAz is lessthan
15%(RefBl).
.3 Aluminates,
A1203orcomplex
aluminates.
.4 Mixedtypes,
generally
combinations
ofsul.phides
andsilicates
(seeFig
B-id)whichoftenarerelatively
short.
Whilesilicate
stringers
andTypeII sulphides
havegreatest
effecton
RAz,all-inclusions
maybe involved
inLT.

B-2

.
r

a)TypeIIsulphide
stringer
(inC-steel,
Si-kil
led,
Al-treated,
with3%RA-~).

..

b)REMsulphide
andoxysulphide
(inC-Mn-Nb
steel,
Si-killed
Aland
REMtreated,
with65%
RA,z)
.

w
.,..

c)Silicate
stringer
(inCMn steel,
semi-killed,
with8%RAz).

,-

d)Duplex
stringers,
Mn-S
surrounded
withsilicate
tails(inC+lnsteel,
semi-killed
with50%RA~).

X400.
inclusions
FigB-1 flicrogr~phs
Of typical
6.3

B-3 ORIGIN
OF INCLUSIONS
.1 Steelmaking
Controls.
Inclusions
existin allsteelsto someextent
andorlglnate
Insteelmaking
fromimpurities
inrawmaterials
andfromgas
reactions.
either
inthebasicoxygen
furnace
Nearthefinalstageof steelmaking,
or afterpouring
intoladles
fromopenhearth
orelectric
steelmaking
furnaces,
thegeneral
composition
ofthesteelis largely
fixed. Controlof
steelmaking
operations
tothisstageenables,
for instance,
sulphur
- the
causeofMnSinclusions
- to be limited
frequently
to lessthan0.02%in
control
sometimes
enables
0.01%to be
normal
constructional
steels.Special
achieved
- a level
wheretheriskofLT isgreatly
reduced.
Further
steelmaking
operations
whichhavean important
effecton the
type,distribution
andgeometry
of inclusions
mayinclude:
(a) de-oxidisation,
and
(b) sulphur
control.
Different
processes
appear
to haveno important
.2 Steelmaking
Process.
effect
on Inclusions
whichpromote
LT.
Thisoperation
improves
properties
byreducing
oxygen
.3 De-oxidisation.
content,
andiscarried
outby:
or
(a) addition
of de-oxidants
(ie killing) to give semi-killed
forconfully-killed
steels.Thisistheusualmethodadopted
aretheprincipal
detrolling
oxygen.Siliconandaluminum
oxidants
used.
Infully-killed
steel,alloxygenin the ladlereactswith
steels,
addedsilicon
or aluminum
toformoxides.In semi-killed
excess
oxygen
reacts
withcarbon
toformcarbon
monoxide
whichis
evolved
during
solidification
oftheingot.
(b) vacuum
degassing
usingspecialhighcapitalcostequipment
to
remove
gasesincluding
oxygen
without
theaddition
ofdeoxidants.
Thismethod
reduces
oxide-type
inclusions
andisusedforspecial
steels
only.
.4 FinalSulphur
Control.
At thefinalstageofsteelmaking
sulphurcan
befurther
controlled
bytheaddition
ofrare-earth
metals
(REM),
e.g.
mischmetal,
whichcontains
cerium
(Cc),
orbycalcium
compounds
suchashypercal.
Thisaddition,
which
mayalsobemadetotheingot
(a) reduces
sulphur
level
byremoving
sulphur
through
theslag,and
(b)
tiesUD sul~hurin REMoxv-sulDhides
whichhavehiahmeltinc!
.,
duringrollingandthus-retain
~
points:resistdeformatio~
globular,
lessharmful
shape.
B-4

,-.

Thismethod
isnowusedforspecial
steelswherehighRAzandresistancetoLT isrequired.
Further
methods
adopted
bythesteelmaker
.5 Ingot
Pouring
andTreatment.
toreduceinclusions include:
(a) adjustment
of pouring
technique,
and
(b) hottopping
of ingots.
B-4 INFLUENCE
OFROLLING
ANDHEATTREATMENT
ON INCLUSIONS
BasicEffect.Aftersolidification,
the ingotis reheated
to a high
.1
hottoslabsandthentoplateor rolledsections.
temperature
andreduced
Thisdeforms
theoriginally
globular
inclusions
to a flatandsometimes
elongated
shape- thusinfluencing
RAzvalues
andsusceptibility
toLT. The
grains
ofthematrix
arealsoelongated
inthedirection
ofrolling.
Fromtheslabstagethechange
inshapeof inclusions
depends
on:
(a) thedegree
ofrolling
orreduction
inthickness,
(b) thedirection
ofrolling,
and
(c) thetemperature
ofrolling.
J

Cropping
attheslabstageremoves
thepartof theslabcontaining
piping,
grossinclusions,
etc.
ofThickness
Reduction.
.2 Effect
Witha greater
degreeof rollingand
reduction
ofthickness,
inclusions
become
flatandhavemoreinfluence
on
RAzvalues.
.3 Effect
ofRolling
Direction.
Rolling
predominantly
inthedirection
of
straight
rolling)
elongates
theinclusions
into
theoriginal
ingotaxis(l.e,
stringers.
iewherethereisrolling
bothtransverse
andalongthe
Crossrolling,
ingotaxis,lessens
elongation
andgivesa roundershapein plan. This
leadstoreduced
RAz(Ref132)
butnotnecessarily
increased
susceptibility
toLT.

:+

ofRolling
Temperature.
Thistemperature
is important
as the
.4 Effect
plasticity
ofmanganese
sulphlde
inclusions
relative
to thesteelmatrix
whilethatof silicate
inclusions
increases
withdecreasing
temperature,
decreases.
Thus,thelower
thetemperature
rangeof working,
thegreater
theflattening
ofMnSinclusions.
Forfurther
details
seeRefB2.
Modification
to shapeof inclusions
is
.5 Influence
ofHeatTreatment.
Diffusion
annealing
hashelped,
but
notposslbl
e bynormal
heattreatment.
thereisnocleardifisnotpracticable
forconstructional
steels.Hence,
ference
between
as-rolled
ornormalised
condition.
B-5

B-5 MATRIX
FACTORS
Theeffect
ofmatrix
properties
onLT isprobably
smallwith
.1 General.
usualconstructional
steels
wheninclusions
arenumerous
andlarge.Research
andexperience
suggest
thematrixhasmoreeffectwithfew inclusions,
butthentheproblem
ofLT is less.
showsthatbothcoarse
andfinegrained
steels
.2 GrainSize.Experience
canbesusceptible
toLT.
Banding
ofrolled
plateandsections
Is reasonably
common,
.3
buttR
ISan othersegregation
of alloying
elements
do not appearto substantially
reduce
Z-direction
properties
or increase
LT.
B-6 INFLUENCEOF
INCLUSIONS
ONPROPERTIES
ANDLAMELLAR
TEARING
Theeffect
of inclusion
shape,
distribution
andsizeis discussed
in
2.3ofAppendix
A. Usingthisandtheaboveinformation
onvarious
typesof
inclusions
itispossible
to givetheveryapproximate
relationships
in
TableB-6.
TABLEB-6 INFLUENCE
OF INCLUSIONS
SteelTypes
Usually
Involved

Inclusion
Type

FK-Si+Al
Silicate
stringersDecreasingIncreasingSK,FK-Si,
riskof
RAz
2. TypeII manganese Lamellar
sulphide
stringers Tearing
FK-Si,
FK-S+A1
1.

sulphides
3. Duplex
withsilicate
tails

SK

4. Ellipsoidal
sulphides

FK-Si+Al
(thick
plate)

5. Spherical
sulphides

SK= Semi-killed
FK= Fully-killed

REMtreated
steels
Si= Silicon
killed
Al= Aluminum
treated

B-7 REDUCTION
OFAREAASAN INDICATOR
OFSUSCEPTIBILITY
TOLT
.1 Selection
on material
ofReduction
ofArea.LT is clearlydependent
properties,,
particularly
Inthel-direction.
Itisnatural
thatconventional,well-established
mechanical
testshavebeenusedto checkthematerial
testsofthematerial
havebeen
susceptibility.
Through-thickness
tensile
andtodaytheyareusedforspecification
purposes
for
usedextensively,
special
steels.
B-6

Thesetensile
testsgiveyieldstress,
tensile
strength,
% elongation
(onvarious
gaugelengths)
and% reduction
of area(RAz).Variation
of
theseproperties
isindicated
in2.3ofAppendix
A andvarious
researchers
(RefsB3andB4)haveshown
RAztobethemostdiscriminating
andaccurate
It also
material
measure
atthelowductilities
whereLT is encountered.
hasbeen-found
tocorrelate
wellwithknown
LT incidence
(RefB5). Thisis
notunreasonable
asRAzisa measure
ofthelocal
strain
occurring
at fractureandmorenearly
represents
thebehaviou~LT.
To ensure
thatthefullthickness
of material
is assessed,
standard
practice
todayistoweldextension
stubstotheplatebyfriction,
studor
manual
arcwelding
toprovide
gripsforthistensile
testing.
.2 Influence
ofTestSpecimen
onRAz. It is important
to notethatthe
testspecimen
diameter
issignificant
inquoting
RAzvalues
as:
arelikely
tobemoreinfluenced
by
(a) thesmaller
diameter
specimens
thesizeandposition
of largeinclusions,
ieminimumvaluesare
reduced
andmaximum
values
increased.

(b) thelarger
specimens,
upto approximately
20mmdiameter,
havemore
lateral
restraint
andthusgivelowermeanvaluesandreduced
Thus,6.4mmdiameter
specimens
give,inabsolute
terms,
scatter.
approximately
10%higher
R
Azmeans
than
18mmspecimens.
SeeFig
.-

50
40
RAz
o
Io

30
20
10
0

FigB-2

15
5 6.4
10
Specimen
Diameter
mm

Typical
effect
ofspecimen
diameter
onRAz
[forsteelwithapprox.
25%Wlz on 6.4~ 4).
RefsB5andB1O.
B-?

.3 Variation
ofRAzinThickness
Position.Testinghasshownthatthe
maximum
incidence
of inclusions
andfracture
location
mayvaryfromjust
belowthesurface
tothecentre
oftheplateorsection,
butusually
ismore
alsoindicates
thatplates
may be more
prevalent
inthecentre.Experience
susceptible
toLTatdifferent
thicknesses.
To provide
forthis,RAztestscheckthefullthickness.
.4 Variation
ofRAzinPlate.Investigations
(Refs
Bl,B6andB7)showno
apparently-consistent
variation
ofRAzovervarious
positions
intheplate.
The onlygeneralindication
is thatfor.theplatestestedt
here was a
slightly
greater
probability
of lowerRAz in thetopcentral
30%of the
plate.
.5 RAzValues
toReduce
Riskof LT. Farrar
(RefB5)givescorrelation
of
RAzwith-known
LT,andIndicates:
(a) inallsteels
whichencountered
LT,themeanRAzvalues(6.4mm
. diameter)
wereequaltoorbelow15%andtheminimum
values
of all
werebelow
12%.
(b) inallsuccessfully
fabricated
steels,themeanvaluesallexceeded
13%andtheminimum
values
of allexceeded
5%.
At thisstageitisnotpossible
to state
thatsteels
withhighvalues
ofRAzwillnotencounter
LT,asmuchdepends
on restraint
factors.As a
guide,
thevalues
giveninFigB-3arerecommended.
Thisdatatakesinto
account
recommendations
from:
o
o
o
o

References
B1andB8,
IIW,ie15%
to 20%minimum
mean,
Reference
B9,30%fornodeplates
inoffshore
drilling
rigs,
Current
practice,
25%minimum
mean(4or6 specimens
takenat the
1/4platewidthposition)
for.
mostapplications,
(35%by one
majoroilcompany)
and10or 15%individual
minimum.

As thescatter
canbehighwithsmalldiameter
specimens
thereis a
riskthatanoccasional
specimen
willfailbelowthespecified
minimum.
This
willhavelittle
effect
onLT,butshould
be allowed
forbytheusual
retest
clause.
B-8 EFFECT
OFOTHERMATERIAL
PROPERTIES
ONDESIGN
Thepossible
reduction
oftensile
strength
andyieldstressin theZdirection
mayinfluence
thedesign
ofthejointinaddition
to allowance
for
reduced
RAz. Thesereductions
arenormally
notgreat(see2.3of Appendix
A) andcanalmost
always
be ignored
indesign.
Extensive
service
experience
has indicated
thatconventional
design
whichisbasedonX-direction
properties
isadequate.
Thisis largely
due
to thefollowing:
B-8

(a) steels
withverylowtensile oryieldstrength
oftenwillbe detected
byLTduring
fabrication,
whereconditions
aremoresevere
thanalmost
allservice
conditions.
(b) tensile
strength
ofmostconstructional
steels
intheworstconditionintheZ-direction
isgreater
thantheyieldstress
on which
mostdesigns
arebased.
(c) inmostweldedjointsthereis diffusion
of theserviceloads
through
a greater
area- thuseffectively
reducing
thestress;
and
(d) theprobability
of al1 factorsactingadversely
is extremely
remote.

In specialcritical
caseswhereveryhighstresses
areinvolved,
special
consideration
mayneedto begiventodesign
ormaterial.

w(J

Negligible
Risk

x : 20
Ci
IV-P

SmallRisk

Eg15
m
N .10
5
n
u

Low

Fig B-3

Medi
urn
JointRestraint

Approximate
RiskofLamellax
Teaxing
forC andC-MnSteels.
Specified
minimum
yield
stress
<40.8kg/mm2

B-9

..

High

LISTOF REFERENCES
FORAPPENDIX
B
B1 Farrar,
J.C.M.,InclusionsandSusceptibility
toLamellar
Tearing
of Welded
Structural
Steels,
Welding
Journal,
August,
1974
B2 Croll,
Factors
Affecting
Lamellar
Tearing,
AWRA
J.E., Material
synlpos
urn
onLamellar
Tearing,
April,
1976
B3 Croll,
Properties
ofStructural
Steels,
J.E., Through-Thickness
Journa
ofAustralian
Institute
ofMetals,
September,
1975
B4 Wilson.
W.G..Minimizing
Lamellar
Tearing,
WeldingJournal,
November
1974
Teststo
B5 Farrar,
J.C.M., etal,UseofSmallScaleDestructive
Assess
theSusceptibility
to Lamellar
Tearing,
Proceedings
of
International
Conference
on WeldinginOffshore
Constructions,
February,
1974
SteelPlatewithImproved
ThroughSteelCorporation,
B6 British
Thickness
Ductility,
October,
1973
Tearing
andMarine
Structures,
Welding
and
B7 Takeski,
Y:,Lamellar
MetalFabrication,
December,
1975
B8 Schonherr,
as Criterion
forJudgingLamellar
W., ? th-V?IIUe
Tearing
Tendency
ofSteel
Structures,
IIWDOC.IX-948-76.
Rules
forDesign,
Construction,
andInspecB9 DetNorske
Veritas,
tionofFixed
Offshore
Structures,
1974
B1O

Comments
on Through-Thickness
IIWDOC. IX-872-74: Japanese
Tensile
TestforEvaluating
Lamellar
Tearing
Susceptibility,
1974

B-10

SHIP RESEARCH

COMMITTEE

Maritime

Transportation Research Board

NationalAcademyof Sciences-National
ResearchCouncil
********

The Ship
interagency
Ship

Research Committee has technical cognizance

Structure Committees research program:

of the

Mr. O. H. Oakley, Chairman, Co?;sultant,McLean, VA

Division, Dep.zrtientof the N~Y,
Mr. M. O. Bur khart, Nava L Ocecmogaphu

Washington, b.C_.Dr.
Mr.
Mr.
Mr.
Dr.

J.
D.
W.
A.
W.

N. Corciea, Senior Staff Metallurgist, ARMCO INC., MiddLe tow, OH

P. Court Sal , Vice president, DPJVO Corporation, Pittsburgh, PA
J. Lane, Consultant, Baltimorw, MD
C. McCI ure, Akz C. McClure Associates, Inc., Houston, TX
Ft. Porter, Vice ties. for Academ<c Affairs, State Univ. of N. Y.

Maritime Co1loge
Prof. S. T. RO 1fe, Civil Engineering Dept., University of Xuwas
Mr. R. W. Rumke, Ezecutive Secretary, ship Research Committee

********

The.ShipMaterials, Fabrication, & Inspection Advisory Group

prepared the project prospectus, evaluated the proposals for this project,
provided the liaison technical guidance, and reviewed the project reports
with the investigator:
Dr.
Mr.
Mr.
Dr.
Dr.
Mr.

J.
W.
T.
W.
H.
P.

.sti~ff

Metallurgist, ARMCO INC., Middletown, OH

N. Cordea, Chairman, Senior
C. Brayton , Asst. to General Manager, Bethl.shamSteel corp., Spari-otis
Point,;,~
E. Koster, Naval Architect, AMOCO International Oil Co., Chicago, IL
C. Leslie, Dept. of Materia2s & Met. Eng?g., Univ. of Michigan, Ann Arbor, Ml
1. McHenry , cryogenics Division, Nationa L Bweau
of Standards, Boulder, CO
W. Narshal 1, Senior
Eng?., SHELL OIL CO., .Youston,TX
Pt_Of.P. F. Packman, Cman, Dept. of Civil 4 Mech. hg?g. , Southern MethodistUniv.

sti~ff
civil

Dallas, TX
PrOf. G. C. Sih, Inst. of Fracture & Solid Mechanics, Lehigh Univ., Bethlahern,PA

SHIP STRUCTURE

COMMITTEE

PUBLICATIONS

These documents me distributed by the National Technical

Infomnation Service, Springfield, Va. 22151.
Those ohwnents have been announced in the Clearinghouse journal
U. S. Govezwnent Research & Development Reports (USGRDR)
under the indicated AL nwnbers.
SSC-275,

The Effect of Strain Rate on the Toughness of Ship Steels by P. H. Franci S,

T. S. Cook, and A. Nagy.

1978.

AO-A059453.

SSC-276 , Fracture Be??wviorCharacterization of Ship Stee1s and Weldments by

P. H. Francis, T. S. Cook, and A. Nagy.
1978. AO-A058939.
SSC-277 , @iginal

Radar and Standard Tucker Wavt?meterSL-7 Containarship Data

Reduction and Correlation Sample by J. F. Dal zel 1.
1978.
AD-A062394.

SSC-278,

Wavemeter Data Reduction Method and Initial Data for the SL-7
Contairwrship by J. F. Dal zel 1. 1978. AO-A062391.

SSC-279. Modified Radar and Standard Tucker WavemetezSL-7 Containership Data

by J. F. Dalzel 1. 1978. AD-A062393.
SSC-280, Results and Evaluation of the SL-7 Containership Radar and Tucker WaveAD-A062392.
meter Data by J. F. Oalzell . 1978.
SSC-281 , Bibliography for the Study of I+opelZer-Induced Vibration in Hull
Structural Elements by O. H. 8urnside and D. O. Kana. 1978. AO-A062996.
SSC-282 , Comparison

of Stresses Calculated Using the DAISY System to Those

Measured on the SL-7 Containership Frogrm by H. Y. Jan, K. T. Chan9, and
M. E. Wojnarowski .

SSC-283 , ; L~~ture

1979.

AD-A069031

Survey on the ColZieion and Grounding Protection of Ships by

AD-A069032

1979.

SSC-284 , Critical Evaluation of Lou-Energy Ship Collision-Damage Theories and

Design Methodologies - Volume I - Evaluation and Roconnnendationsby
P. R. Van Mater, Jr. , and J. G. Giannotti.
1979. AD-A070567

of Low-Energy Ship Collision-Damage Theories and

SSC-285, Critical Sval?.zztion
Design Methodologies - Vol%nw?II - Literaiur%?Search and Review by
P. R. Van Mater,

Jr. , and J.

G. Giannotti.

1979.

AD-A070568

SSC-286 , Results of the First Five Data Years of ErctremeStress Scratch Gauge
Data Colbaction Aboard Sea-Lands SL-7 .sby R. A. Fain and E. T. Booth.
1979.
SSiC-287, Ezanination of Service and Stress Data of Three Ships for DeveLopment
of HULL Girder Load Criteria by J. F. Oalzell , N. M. Maniar, and
1979.
M. W. tisu.
SSC-288,

The Effeet.sof Varging Ship Hull Proportions and Hull Materials on Hull
Flexibility Banding and Vibratory Stresses by P. Y. Chang.
1979.

SSC -289, A Method for Econotie Trade-Offs of Alternate Ship Structural Materials
1979.
by C.: R. Jordan, J. 8. Montgomery, R. P. Krumoen, and D. J. Woodley.