0% found this document useful (0 votes)
297 viewsWeld Integrity and Performance
Weld Integrity and Performance.pdf
Uploaded by
Sandeep SundriyalCopyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
297 viewsWeld Integrity and Performance
Weld Integrity and Performance.pdf
Uploaded by
Sandeep SundriyalCopyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
/ 65
Copyrig® 957
Asa inecational®
All rights reserved
Nopatf this book may be reproduced, stored ina retievalsystem or tanseited, in ony form oy any
means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of
the copyright owner.
First printing, July 1997
Great care is taken in the compilation and production of this Volume, butit should be made clear that NO
WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION
WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot
guarantee that favorable results will be obtained from the use of this publication alone. This publication is
intended for use by persans having technical skill, at their sole discretion and risk. Since the conditions of
product or material use are outside of ASM's control, ASM assumes no liability or obligation in connection
with any use of this information, No claim of any kind, whether as to products or information in this
publication, and whether or not based on negligence, shall be greater in amount than the purchase price
of this product or publication in respect of which damages are daimed. THE REMEDY HEREBY
PROVIDED SHALL BE THE EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT
SHALL EITHER PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES
WHETHER OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As
‘with any material, evaluation of the material under enduse conditions prior to specification is essential,
‘Therefore, specific testing under actual conditions is recommended.
Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or
reproduction, in connection with any method, process, apparatus, product, composition, or system,
whether or not covered by letters patent, copyright, or trademark, and nothing contained in this book shall
be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or
as a defense against liability for such infringement.
‘Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International.
Library of Congress Cataloging: in-Publication Data
‘Weld integrity and performance, technical editor Steven R. Lampman
Pom
Includes bibliographical references and index.
1, Welding—Quality control. 2. Weldability. 3. Welded joints.
L Lampman, Steve. I. ASM International.
‘TS227.2.V38 1997 97-10740
6715'7'0685—de21
ISBN 0-87170-600-8
‘SAN; 204-7586
ASM International®
Materials Park, OH 44073-0002
Printed in the United States of America Table of Contents
Introduction
1: Weld Solidification .
Chapter 2: Weldability Testing .
Weld Imperfections and Cracking
Bu
\perfections:
Chapter 5: GStuing Phenomena Asoogted wit Welding,
Chapter 6: Inspection and Evaluation of Weldments ..... .
Weld Failure Analysis and Prevention
Chapter7: Fatigue and Fracute Control of Wels
Chapter 8: Factor Influencing Weldment Fatigue
Chapter 9: Fracture Mechanics and Service-Fitness of Welds -
Chapter 10: Repair Welding . . 179
Chapter 11: Corrosion of Weldments. . 189
Chapter 12: Failure Analysis of Weldments 197
Engineering Properties of Welds
Chapter 13: Properties of Carbon and Low-Alloy Steel Welds . 27
Chapter 14: Properties of Stainless Steel Welds . . 249
Chapter 15: Properties of Aluminum-Alloy Welds
Chapter 16: Properties of Titantum-Alloy Welds
Chapter 17: Properties of Nickel-Alloy Welds
Chapter 18: Properties of Refractory Metal Wel
Appentix: Reference Tables
Table Al: Weldability of Cast Irons. .
Table A2: Weldability of Steels. .
Table A5: Recommended Preheat and Interpass Temperatures
for Selected Thicknesses of Structural Carbon Steels
Table A6: Recommended Preheat and Postweld Heat Treatment
of Steel Pressure Vessel and ‘Welds
Table A7: Recommended Preheat and Interpass Temperatures
for Heat Treatable Steels Introduction Chapter 1
Weld Solidification
JOINING METALLURGY has assumed an
‘even greater role in the fabrication of metals.
‘within the last few decades, largely because ofthe
derdlopemt of new alloys with tremendously
increased strength and toughness. Therefore, a
working knowledge of metallurgy is essential to
structures:
joining
mnetallargy are the microstructures of a weld
joint, which determine the mechanical proper-
ties, and welding variables such as weld thermal
cycle, chemical reactions in the molten pool, al-
laying, ‘ux and contaminants,
which significantly affect the weld and heat-af-
fected zone (HAZ) microstructures.
Metallurgically, a fusion weld consists of three
the weld pool consisting of a
bution of inclusions and defects such as porosity
and hot cracks. Since the properties and integrity
of the weld metal depend on the solidification.
behavior and the resulting microstructural char-
acteristics, understanding weld pool solidifica-
tion behavior is essential.
In this article, a general introduction is pro-
vided on key welding variables including solidi-
fication, microstructure, and causes and reme-
dies of common welding flaws.
Solidification Behavior
‘The integrity and performance of a weld is, toa
large extent, controlled by the solidification be-
havior of the weld metal or fusion zone. It con-
trols weld-metal microstructure, grain structure,
inclusion distribution, porosity, ‘Heteracking be-
havior, and, ultimately, weld-metal pro}
‘Current knowledge of weld pool solidification
is an active area of research, as fundamental ad-
vances in solidification mechanics for castings
and ingots bave been applied to weldments (Ref
14). Fundamental solidification mechanics di
veloped primarily forcast metal have been sue-
cessfully applied to the solidification of welds.
However, major differences do exist between
casting and welding including:
growth
ariations in temperature gradient and
growth rates within the pool
In solidification mechanics, the important pa-
rameters that influence microstructure are tem-
perature gradient G, growth rate (R), undercoal-
ing (AT), and alloy composition (Co). In the weld
pool, G, R,and AT vary considerably from region
‘This Chapter has been ad:
als Handbook, Volume 6, 1
ted from “I of Joining Metalh Jack Devietian and William Wood in Met-
p22-49and and "nc of Wald ‘Solidiaba and Microstructure" by 8: Devid an.
‘ink in international Trends Welding Science and Technology, ASM, 1953, pia7-157 4 / Introduction
Hear.
sounce
Fig. 1 Schematic diagram showing the three zones within wekdment
(LARGER Fy
‘SMALL Gj)
(oma my,
LARGE Gi}
‘BOND
FUSION ZONE
WELD CENTER
EQUIAXED DENORITE
BASE METAL
Fig. 2 Schematic drawing of sucturl vration of weld microstncture eres fasion rane
to region and process to process, Therefore, the
microstructure that develops in the weld metal
varies noticeably from region to region. Figure 2
shows schematically variations in weld-metal
microstructure influenced by temperature gradi-
ent and growth rate.
Weld Grain Structure
The final weld-metal grain structure has con-
siderable practical significance. It has been
shown that a coarse columnar structure is sus-
ceptible to hot cracking, whereas a finer grain
structure, a distorted columnar structure, or an
equiaxed structure is more resistant to hot crack-
ing. In addition, the fusion zone grain structure
has a strong influence on the mechanical proper-
ties of the weld. To improve both the mechanical
behavior and the hot cracking resistance of weld
metal, efforts have concentrated on refining the
fusion zone grain structure. Several means are
available for refining the weld-metal grain struc-
ture, Since grain growth is epitaxial in nature, the
size of the grains in the base metal controls the
final grain size to a limited extent.
The weld pool shape also influences the fusion
zone grain structure. For example, in an ellipti-
cally or circularly shaped weld puddle, not only
does the magnitude of the maximum thermal
gradient change continuously from the fusion
line to the weld centerline but also the direction
of this gradient changes. Since the average
growth direction during solidification of a weld
pool is approximately normal to the solid /liquid
interface, which is also along the maximum tem-
perature gradient, a given grain will not be favor-
ably oriented during the entire solidification
process since the direction of this gradient
changes. Therefore, many of the grains at the
fusion line thatare initially of unfavorable orien-
tations may become more favorably orientated ‘Temperature gradient (@), Kim
Fig. 3 G-Rdhneram or dierent sition proce
represent orders of magnitude Far
the solicification cyle. Source: Re 4
high energy beam welds or arc welds made at
low welding speeds. For a tear drop shaped pud-
‘die, there is almost an invariant direction of maxi-
K/m and
‘growth velocities in the range of 10° to 10"! m/s.
With the introduction of laser welding tech-
dient. As willbe shown later, this trend may ead
to problems in steel welding ifthe samealloysare
used (Ref 4).
Sot eerie oases
grain-growth selection process. Growth crystal-
Segepy wiliafuence pin growth by vorng
growth along the
biemetala the oy gaa dbecieee mc sits,
The conditions are even more favorable when
with columnar
‘welding and lier Peatment only one process
Weld Solidification / 5
growth structures. The G and values are indicative only and
represented with earditians prevailing ding,
‘one of the easy growth directions coincides with
‘the heat flow direction in welds
Figure 4 shows both epitaxial growth and a
‘Unlike: in castings, the natural occurrence of a
‘columnar-to-equiaxed transition in the grain
‘structure of the weld is not very common. How-
‘ever, depending on the alloy system, such a tran-
‘sition has been observed and is described by
HLW. Kerr (Ref 5).
‘An understanding of the development of grain
structure in the fusion zone of
‘between
a'weld (aside fom the relative sing an tao
rates) is the phenomenon of epitaxial growth in
welds. In castings, formation of solld crystals
from the melt requires
of solid, iyo the mold wal
principally on wal
ialowed by grain rome
For weld solidification, there is no nucleation:
barrier, and solidification occurs spontaneously
by epitaxial growth. The nucleation event in
welds is eliminated during the initial stages of
solidification because of the mechanism of epi- 6 / Introduction
Fig cotsil account ei ion nein ii lo elecon bea we Source: Ret
taxial growth wherein atoms from the molten
weld pool are rapidly deposited on preexisting
lattice sites in the adjacent solid base metal.
‘As a result, the structure and crystallographic
orientation of the HAZ grains at the weld inter-
face continue into the weld fusion zone. In fact,
the exact location of the weld interface is very
difficult to. determine in any weld deposited on
pure metals using matching filler metal. Even
microstructural features, such as annealing twins
located in the HAZ weld joints, will continue to
grow epitaxially into the weld during solidifica-
tion,
Epitaxial growth is always the case during
autogenous welding. For nonautogenous proc:
esses in which a filler metal is used, epi
th may still occur. Nonmatching filler met-
als will also solidify epitaxially, particularly ifthe
filler metal and base metal have the same crystal
structure upon solidification, eg. welding
Monel (fee) with nickel (fec) filler metal.
However, the more classical case of heteroge-
neous nucleation also applies in this case. In
welds, growth of the solid progresses from the
uunmelted grains in the base metal, and the stabil-
ity of the solid /liquid interface is critical in deter-
mining the microstructural characteristics of the
weld metal, The thermal conditions in the imme-
diate vicinity of the interface determine whether
the growth occurs by planar, cellular, or dendritic
growth, Composition gradients and thermal gra-
dients ahead of the interface are of primary im-
portance. The effect of solute and thermal gradi-
ents on the solidification front can be described
by the concept of constitutional supercooling cri-
terion (Ref 6), which can be mathematically
stated as:
GUR> ze Jor plane front satidification (Eq 1)
or
ATy se ses og
GyR< 55" or cellular or dendritic solidification
(Eq2)
where Cr. the thermal gradient in theliquid at the
solidification front, ATi is the equilibrium solidifi-
cation range at the alloy composition (Co), Dj, is the
solute diffusion coefficient in liquid, and R is the
solidification front growth rate.
Equations 1 and 2 define the conditions for the
development of various solidification substruc-
tures in welds. This concept of constitutional su-
percooling is sufficient to understand the devel-
opment of microstructures and the influence of
parameters on these microstructures in
welds.
Weld Pool Shape
‘The macroscopic shape of the weld is an impor-
tant factor that influences not only weld integrity
but also grain size and the microstructure of the
fusion zone, Therefore, itis critical to understand
the dynamics of the weld pool development and
its geometry,
The weld-paol geometry is determined mainly
by the thermal conditions in and near the weld
pool and the nature of the fluid flow,
In general, the weld pool volume is controlled
toa large extent by the welding parameters. The
volume is directly proportional to the are current
(heat input) and inversely proportional to the
welding speed. In addition, the speed of themov-
ing heat source has an influence on the averall
bead shape. For arc welding processes, the pud-
dle shape changes from elliptical to tear drop
shaped as the welding speed increases, Corre-
spondingly, the isotherms also change in shape.
For high energy density processes such as elec >
A
Weld Solidification / 7
Location
of arc
= Welding direction
Location
of are
+ Welding direction
ote
= Wotaing sirection
A
Fig. 5 Comparison of weld pool shapes. Travel speeds: (a) stow, th) intermediate, (c) fast
tron beam or laser welding, the thermal gradients
are steeper and as.a result the puddles are circu-
lar at lower speeds, becoming more elongated
and elliptical in shape as the welding speed in-
‘creases before eventually becoming tear drop
shaped at high speeds.
For example, if a single-phase metal
‘tungsten arc welded at a low velocity, the weld
poolis elliptical (nearly circular), as shown in Fig
‘5{a). The: ‘in the direction of
the
temperature
constantly changing from approximately 90° to
the weld interface at position A to nearly parallel
to the weld axis at position B, the grains must
grow from Aand continuously tun to-
ard the postion ofthe moving azc, The process
of “compelitine growth” provides
‘rains ess, favorably oriented for
grow hee pinched off cr cowed out by geting
better oriented for continued growth. The <001>
and are the generally favored directions
for crystal growth in cubic (fee and bec) and hex-
agonal (hep) metals, respectively. In foc metals,
for example, the <001> most favored direction
Jeads each solidifying grain because the four
close-packed {111} planes symmetrically located
around the <001> axes require the greatest time
to solidify and, therefore, serve bot to drag and
‘guide the growth of solidifying grai
aw shape of ‘of the weld pool. fends to become
with increasing welding speed.
In Fig S04 the diction of maaan tenors
‘ture gradient is perpendicular to the weld inter-
face at positions A and B, but because the weld
‘pool is trailing a greater distance behind the are,
the temperature gradient at position B is no
longer strongly directed toward the electrode.
‘Therefore, the columnar grains do not turn as
much as in the case ofa nearly circular weld pool.
Finally, the weld takes on a teardrop shape at
the fast welding speeds that are usually encoun- '8/ Introduction
Volume
L lament
Solid a
te)
1200}
8
Temperature, °F
5
sool_t
Arc, 518 20 “cu
(al Copper, wi
(ch Distance
ta) Distance
Fig. 6 Solfication of deena wel Slatin of 39 Cu-Aalloy by he poo ees, Sle ibuton
‘occurring ahead of the solldfiquid interface. (d) Constitutional supereaoling develops when the actual temperature of quid in
the copper-sich zone Is greater than the liquidus lemperabre.
tered in commercial welding practice. The weld
pool is elongated so far behind the welding arc
that the directions of the maximum temperature
gradient at position A and B in Fig. 5(c) have
changed only slightly. As a result, the gales
grow from the base metal and converge abruy
the cenertine of te weld witht change
direction. Welds that solidify in a teardrop shape
have the poorest resistance to centerline hot
cracking because low-melting impurities and.
other low-melting constituents tend to segregate
at the centerline. Unfortunately, this solidifica-
tion geometry occurs most frequently in com:
mercial welding applications, because high heat
input and fast travel speeds produce the most
cost-effective method of
‘Quantitative Modeling, Computational mod-
cling is a powerful tool for understanding the
development of weld-pool geometry. In recent
years, significant advances have been made to
Tnodel and understand the development of pool
shape in welds (Ref 7-10). Some of these models
can address coupled conduction and convection
heatstransfer problems to predict weld-pool ge-
ometry and weld penetration. Of the various
modes of hest transfer, convection plays critical
role in determining the weld penetration. Con-
version in the weld pool is driven mainly by
buoyancy, electromagnetic, and surface tension
forces. Significant work has concentrated on the
convective heat transfer and, in particular, the
effect of the spatial variation of surface tension
Se ge ew en weld
penetration (Ref 1114).
‘A fundamental limitation to a realistic
tion of the weld-pool shape is the lack
sary ysical data. Unless a
sive data base on thermophysical properties and
mechanical behavior of solids at very high tem-
peratures (including the mushy zone) during so-
neces-
prediction of microstruc-
ture, defects, and stresses will be stifled. In addi-
tion, verification of the models using, critical ex-
periments is essential.
Solute Segregation
Solute redistribution during weld-pool solidifi-
cation isan it phenomenon that can sig-
nificantly affect weldability and, in particular,
hot- ‘behavior, weld microstructure, and
properties. “As alloys solidify, extensive solute
redistribution occurs, in segregation of
the alloying elements that constitute the alloy. Fegation on a smalll scale (few microns) is
referred to as microsegregation, and large-scale
segregation (hundreds of microns) is known as
macrosegregation.
‘Macrosegregation can occur in welds, but it is
nat very common. In welds, itis primarily due to
a sudden change in the welding parameters and
the resultant change in growth velocity. A com-
mon manifestation of macrosegregation in welds
isbanding. Another form of macrosegregation in
welds is the formation of a solute-enriched re
along the centerline of the weld, which could
promote the formation of centerline hot-crack-
ing. However, evidence or proof of this phe-
nomenon is lacking or not entirely clear.
Microsegregation is characterized by a compo-
sitional difference between the cores and periph-
eries of individual cells and cellular dendrites.
Cells are swicroscopic pencil-shaped protrusions
of solid metal that freeze ahead of the solid-liquid
interface in the weld. Cellular dendrites are more
developed than cells and appear to have a “tree-
like" shape; the main stalk is called the “primary
dendrite arm,” and the orthogonal branches are
called the “secondary dendrite arms.” The cores
of the cells and dendrite arms have a higher soli-
dus temperature and contain less solute than the
intercellular and interdendritie regions. In actual
welding practice, cellular or dendritic mi-
gation is virtually impossible to avoid
unless the metal being welded is a pure element
Generally, the important parameters control-
ling the cellular or cellular dendritic substruc-
tures in welds are:
+ The equilibrium partition ratio, K, which isan
index of the segregation potential of an alloy:
R= CC (Eq 3)
where Cj is the solute content of the solid at the
solid-liquid interface and Ci is the solute content of
the liquid at the solid-liquid interface;
+ Thealloy composition itself, Co
+ The temperature gradient, G, in the liquid at
the weld interface, and
+ The growth rate, R, oF velocity of the interface
‘Considering the solute redistribution in the in-
terdendritic regions, also known as the mushy
zone, it may be adequate to apply the solidifica~
tion models for microsegregation originally for-
mulated for castings. The solute profile in the
solid is given by the Scheil equation (Ref 15):
=pyer!
C= kc
(Eq4)
where C; is the solid composition, Co is the i
alloy composition, F5 is the volume fraction solid,
and 5 the equilibrium partition coefficient.
Weld Solidification / 9
‘This equation was modified by Brody and
‘Flemings (Ref 16) by allowing for diffusion in the
solid during solidification. Kurz. and Clyne (Ref
17) have further refined Brody and Fleming's
model to account for extensive solid diffusion,
These models can be easily adapted to analyze
microsegregation in welds
Example: Solute Redistribution in the Den-
drites of a3%Cu-97%Al Composition. Consider
an alloy of composition Co equal to 3%Cu-97%Al
in Fig, 6(a) so that per Eq 3:
K=CUCL = 1.7%/10% =0.17
‘The first metal to solidify will contain only:
CoK = (3) 0.17 = 051% Cu
while the last liquid to solidify between cells or
cellular dendrites is rich in copper:
Co/K = 3/01
= 17.6% Cu
These values represent the short transients at
the start and finish of solidification of a cell or
cellular dendrite. As the cell or dendrite grows in
the weld, a dynamic equilibrium is achieved be-
tween the newly forming solid of composition, Co
3% Cu, and the copper-rich liquid containing a
maximum of Co/ K 17.6% Cu, at the solid /liquid
interface as shown in Fig. 6(b)and (c). Ifthe actual
temperature distribution ahead of the solid /liq-
uid interface is less than the liquidus tempera-
ture, constitutional supercooling occurs (Fig: 6d).
‘Supercooling means that the solute-enriched liq-
uid ahead of the solid-liquid interface has been
cooled below its equilibrium freezing tempera-
ture, and constitutional indicates that the super-
cooling originated from an enrichment in compo-
sition rather than temperature,
‘Microsegregation results when the copper-rich
liquid at the solid /liquid interface solidifies be-
tween the cellular dendrites. The interdendritic
regions are so segregated with copper (solute)
thata small amount of eutectic (6) is frequently
observed. Eutectic structures can only occur
when the composition of solidifying metal ex-
ceeds the maximum solid solubility
Effect of Temperature Gradient. Whether or
not a planar, cellular, or dendritic substructure
occurs upon solidification is largely determined
by G and R (Fig. 7) which control the amount of
constitutional supercoling. If a weld is depas-
ited at a constant travel speed, R becomes fixed.
By inducing an extremely steep temperature gra-
dient G1 (Fig. 8a), no constitutional supercooling
occurs and the solidified weld-metal grain strac-
ture is planar.
When the gradient is decreased slightly to G2
(Fig. 8b), any protuberance of solid metal on the
interface will grow faster than the remaining flat 10/ Introduction
‘Temperature gradient in mait(@), Kirren
‘Temperature gradlent, G@ —=
Planar Cobar = endeic
Variation of weld microstructure 2s function oftemperature gradient, growth rate. and combinations of these variables GR, C/R,
Fie? Soorcernal and?
Interface because the solids growing into super-
cooled liquid; that is, the solid protuberance éx-
ists at a temperature below that of the liquidus
for that alloy. As a result, a cellular substructure.
Hal
uit
TEE
]
Bie
z
while no measurable segregation is encountered.
in planar growth. Whether planar, cellular, or
cellular dendritic, growth is always anisotropic.
Investigators have found that these solidifica-
tion substructures can be characterized by the .
combined parameter G/R. Figure 9 shows that a
large value of G/R combined with a very dilute
alloy will result in a planar solidification struc
ture, while a low G/R and high salute concentra-
tion will produce a heavily segregated columnar
dendritic structure. Both columnar dendritic and
equiaxed dendritic structures, although com-
‘mon in large castings, are not frequently encoun:
fesed fn src fa. pectin cela oad coal
welding As a general rule, a fast weld-
ing (R) will produce a steep G. The relative
values of G and R, however, determine the solidi-
fcation moxphelogy fora given alloy of fixed Co
and K,
Solidification Rate. While G/R controls the
mode of solidification, the weld cooling rate, in
terms of the ‘GR (solidification rate in
units of *F /s), determines both the size and spac-
ing of cells and dendrites. and others
have demonstrated that the effect of solidifica-
tion rate on the dendrite arm spacing (f) is:
d= a(GR)" 45) Woe
interface
Haz
hte ot
a
E
‘a Distance
fect of themal radeon moti of sa
Fig, 8 eelular pow e)smallC, clr
wheres is a constant and 1 isapproximately 14 for
primary arms and between ¥% and 44 for secon-
dary dendrite arms. The dendrite arm spacing of
stainless steel in an electroslag weld is often several
hundred times greater than that found in a rapidly
cooled laser weld.
Solute Banding. The phenomenon of solute
banding occurs to some degree in all alloy welds.
The formation of ripples on the weld surface and
solute banding within the weld are both caused
by the discontinuous nature of weld-metal solidi-
fication and occurs in manual as well as in auto-
matic welds where the travel speed is mechani-
cally constant. During weld-metal solidification,
however, K fluctuates cyclically above and below
a mean value of growth rate that is determined by
the weld travel speed, Fluctuations in K result in
not only ripple formation, but also solute band-
ing. Because an abrupt increase in R causes a
reduction in theamount of solute that can be held
in the solute-enriched liquid (Fig, 15c), excess
solute is dumped and appears as a solute-rich
band, Similarly, a sudden decrease in R produces
a solute-poor band. Solute banding lines are very
helpful in welding research because they always
outline the weld-pool shape at a given instant
during solidification. For example, the form fac-
tor (ratio of width to depth of weld pool), which
is so important in clectroslag welding (ESW), can.
be easily measured metallographically using sol-
ute band lines.
a
fication in welds for eonstant growth
tic prone.) Solid ican of
Weld Solidification / 11
t) Ts,
Liquids
temperature
(a) Steep G, planar growth. (Interred ine
Equisxed
dendritic, Columnar
dendritic,
Celluiae
dendeitic
content, C,
Solidification parameter, G/VR
Dependence of made of solicfication on C/K parameter
for dflerent solute cancentrations (C3)
Fig.9
rostructure of the Weld and
Heat-Affected Zone
‘Through the process of epitaxial growth, the:
initial columnar grain width of the fusion zone is
determined by the size of the base-metal grains
adjacent to the weld interface, Because the peak
HAZ temperature increases with decreasing dis-
tance from the weld interface and grain growth is
4 function of temperature, the maximum grain 12 / Introduction
size in the HAZ always occurs along the weld
interface. It is this maximum grain size that is
transmitted into the weld fusion zone.
Grain Size. The relationship used to calculate
the grain size in the HAZ is:
D= Dy = be OF (Eq6)
where D is final grain diameter; Dp is the original
grain diameter; eis the natural base for logarithms;
Tp is the peak temperature, which would be the
solidus temperature at the fusion line; fis the time
at temperature; () is the activation energy for grain
growth of thealloy; is the universal gas canstant;
and and n are constants determined by the mate-
rials. Both temperature and time at temperature
produce grain growth in the HAZ, and as stated
previously, the maximum grain size always occurs
immediately adjacent to the weld interface. Be-
cause all welds experience the same spectrum of
peak temperature from Ty to the solidus tempera:
ture, the only significant variable in Eq 6 is the
residence time, ¥. As the cooling rate decreases,
residence time increases, substantially coarsening
the maximum HAZ grain sie
rocess of competitive grain growth may
ther lateral growth of the weld-metal
grain size, Columnar grains emanating from the
HIAZ continue to widen as they grow into the
weld fusion zone,
‘The maximum columnar grain width in the
weld metal is limited only by the physical size of
the weld bead and the arc energy input. For ex-
ample, tis virtually impossible for a gas tungsten
arc weld deposited on a coarse-grained copper
casting to exhibit any distinguishable HAZ be-
cause the residence time is insufficient to cause
noticeable grain growth; for example, D in Eq 6is
insignificantly greater than Dp. Furthermore, be-
cause the large base-metal grains that grow epi-
taxially into the weld must squeeze into a bead of
limited volume, lateral growth of the columnar
grains in the weld is not possible.
When cold-worked allaysare welded, the HAZ.
experiences both recrystallization and grain
growth reactions. The hardness and strength
roperties of the recrystallized HAZ lose the
nefits derived by cold working, and joint
strength approaches that of an annealed alloy.
Although weld-metal properties can always be
controlled by judicious alloying, HAZ properties
canonly becontrolled by regulating heat input or
by changing the base-metal composition,
In steels and other metals that undergo allo-
tropic phase transformation, the HAZ is conven-
tly divided inta two regions:
+ The grain growth region, which lies adjacent
to the weld interface:
+ The grain-refined region, which is farther
away from the weld interface
Because the grain growth region of the HAZ has
experienced peak temperatures approaching the
solidus of the base metal, coarse grains develop in
accordance with Eq 6. The grain-refined region of
the HAZ has been thermally cycled only briefly
into the low-temperature portion of the austenite
region, resulting in significant grain refinement.
This grain-refining reaction occurs by the nuclea-
tion of new grains each time the Ay.and As lines are
crossed, either upon heating or weld cooling. The
general structure of a steel weld will alwaysappear
fine-grained when compared to similar welds de-
posited on single-phase metals, such as. pure
nickel, copper, te brass, and ferritic stainless steel.
The grain size distribution in precipitation-
hardening alloys—which include maraging
steels; precipitation-hardening stainless steels;
eax, 6x, and 7xxx series aluminum alloys; co-
balt- or nickel-based superalloys; copper, tita-
nium, and magnesium alloys; and many others—
is generally similar to that of the single-phase
alloys. The majority of precipitation-hardening,
alloys develop coarse grain structures in both the
weld and HAZ, and the small amount of second-
phase transformation is insufficient to produce
any grain refinement. For example, welling and
slow cooling a typical nickel-based superalloy
containing small additions of titanium and alu-
minum result in a coarse-grained weld and HAZ
structure with small amounts of Nix(AL Ti) phase
along the 7 (nickel solid solution) grain bounda-
ries. If the weld cooling rate is fast, as in an
electron beam weld, the Ni(AI,Ti) does not form
at all on cooling, but remains in.a supersaturated
solid solution. Subsequent aging treatments only
precipitate Nii(Al,Ti) as microscopic particles
throughout the weld and the HAZ.
Multiple-Pass Welds. Grains of single-phase
metals continue to grow without obstruction
through each succeeding weld pass of multiple-
pass welds, until all of the required passes are
complete. Such interpass epitaxial growth leads
to coarse columnar grain structures and extreme
anisotropy af mechanical properties. Peening or
cold working each weld pass prior to-deposition
of the subsequent pass helps mitigate the prob-
em. The pening action sufficiently cold works
the columnar grains ofa freshly deposited pass to
cause development of a refined or recrystallized
grain structure in the new HAZ of this weld pass.
‘Through epitaxial growth, these refined grains
grow into the weld. By interpass pening, colum-
nar grains are restricted to growth only within
each weld pass, thus greatly reducing the overall
grain size and anisotropy of multiple-pass welds
Of single-phase alloys. Peening is not_recom-
mended by most welding codes for the first and
lst (surface) passes because of the likelihood of
fracturing the first pass and heavily distorting the
surface of the last pass. In multiple-pass welds deposited on allotropic
metals, such as steel, substantial interpass grain
refinement occurs in the weld fusion zone. Bach
interpass HAZ recrystallizes into a fine-grained
structure that effectively prohibits the uninter-
rupted growth of large columnar grains from one
weld pass to the next, as occurs in multiple-pass
welding of single-phase alloys. Grain ref
is achieved each time a portion of the steel in the
weld transforms to austenite upon heating fol-
lowed by a transformation back to ferrite plus
bainite or pearlite upon cooling. A thick band of
fine, equiaxed grains separates the coarse colum-
toanother
is companied by the nucletion and growth of
very fine grains which completely replace the
coarse columnar grain structure from the pre-
vious weld pass. When the HAZ is heated into
the austenite region, growth of the newly formed
‘grains may take place rapidly if the residence
time at the peak temperature is long, or the weld
cooling rate is slow. Fast cooling of a weld pro-
duces fine grain structures. For example, a weld
deposited on 2-in.-thick steel plate in 50 passes
‘with the SMAW process will produce a much
finer grain size in both the weld and HAZ than a
similar weld deposited in 1 pass with the ESW
process, which is an extremely high heat input
‘welding process.
‘Influence of Solidification Structure on Solid-
State Transformations. The microstructure al-
Ways remains single phase in ingle phase alloys
te the presence of microsegregation in cells
or cellular dendrites, because no transformations:
‘occur in the solid state. For allays that experience
‘solid-state transformations, such as the eutectoid
transformation in steels, the solidification struc-
ture of cells or dendrites affects to some extent
any subsequent solid-state transformation as the
‘weld cools to room temperature: For example, in.
‘steel weld metal, the interdendritic or intercellu-
lar regions are more hardenable than the cores
Decause these regions contain substantially
nee Percentages of alloying elements such as
nickel, manganese, and molybdenum. Generally,
as the alloy content increases, the compositional
differences that may be developed between the
‘cores and interdendritic regions are more signifi-
cant.
In plain-carbon or low-alloy steel weldments,
‘the cellular dendritic solidification that normally
‘cosurs under most welding condone cannot be
seen metallographically unless specially pre-
pared solute-sensitive etchants are used, because
‘the strong etching characteristics of proeutectoid
ferriteand the eutectoid transformation products.
effects
dence of any solidification structure is apparent
Weld Solidification / 13.
unless the weld is etched with a solute-sensitive
reagent such as Stead's reagent. Stead’s selec-
Fun sopregaticn Booms phaspharts gre
ation. Because
Fates intercellularly or inerdepdritically in
;, the solidification structure of the weld
aera can be revealed. In low-carbon steels, the
jegree of alloy segregation, although apprecia-
ble, is not sufficient to significantly change any
solid-state transformation.
In high-alloy steels, the solidification of 304
stainless steel is an excellent example of the influ-
ence of the solidification structure on the result-
jing solid-state transformation. When wrought
204 stainless plate is examined metallographi-
cally and magnetically (with the ferrite gage), the
srlrostracture i fully usteitic When thisaloy
is aut us! 5 tungsten arc (wil
bear saniee 4 Bow rere the troll mena me
fenger fully austenitic, but now contains about
te. During solidification of the weld, the
firs solid fo form B ferrite dendrites contaizing
an enriched chromium content at the core of the
dendrite arms. Upon further cooling, the 8 trans-
forms almost entirely to austenite, except in the
dendrite core where the concentration of chro-
mium is sufficiently to stabilize approxi-
mately 5% of the 8 ferrite. If this weld metal is
chemically homogenized by hot working and an-
nealing, the 6 ferrite disappears, and the weld
microstructure becomes fully austenitic again.
In weld metal of high-alloy steels, such as 18%
Ni maraging steel, the segregated solidification
structure is directly responsible for formation of
small pools of unwanted retained austenite,
When 18% Ni maraging steel is gas metal are
welded with nearly matching filler metal, the
resulting dendritic structure is so segregated that
the nickel-entiched interdendritic regions do not
transform fully to martensite upon cooling. The
resulting weld-metal microstructure is not 100%
martensite as intended, but is a duplex structure
of martensite with about 8% interdendritic re-
tained austenite that is detrimental to mechanical
properties.
Effects of Welding on Microstructure
‘The solidification structure generally has little
or no effect on subsequent solid-state transforma-
tions when the degree of segregation is small, as
in dilute alloys which have undergone little con-
stitutional supercooling, because the parameter
G/R was high. Heavily alloyed metals, however,
tend tosolidify with a segregated dendritic struc-
ture where the dendrite core is rich in the high
melting point elements(a) and the interdendritic
areas are ich inthe low melting point element)
Because the dendrite core and interdendritiec 14/ Introduction
6 1
ia
2 No.
3 cracking
2 49] i
}
= 30 —
3 Hot
3 cracking
é
20
10
ooo
Carbon, wi%
Fig. 10. Relatonshipof manganese sui rato, carbon eon
tent, and hot eracking susceptibility in welds
gions are virtually two entirely different alloys,
substantial deviations from the normal solid-
state transformations can be expected.
In welding more complex metals that undergo
solid-state phase transformations, alloy compasi-
tion is a major factor in determining the final
microstructure and, hence, mechanical proper-
ties of a weld joint. For example, when welding
plain-carbon and low-alloy steels, a weld metal
carbon content below 0.1%, or as low as possible,
should be maintained to achieve excellent weld
toughness and maximum resistance to hot crack-
cracking section in this article). To counteract this
hot cracking tendency, manganese is usually
added to steel; as the carbon content of the weld
metal increases, the manganese-to-sulfur ratio
booomes larger, a shown in Fig 10, Fortunately,
most commercial filler metals contain minimal
carbon and sulfur levels, while alloy content is
raised to achieve specified strength require-
‘ments. Most structural steels, such as ASTM A36
‘and A588 or AISI-SAE 1018, 1020, and 1045, how-
ever, contain well over 0.10% C primarily be-
cause carbon isa far less expensive strengthener
than alloying additions suchas manganese, chro-
mium, molybdenum, and nickel.
Effects of Cooling Rates. Consider SMAW of
0.18% C, low-alloy steel (ASTM A588) with an
E7018 electrode. Temporarily neglecting the
small amounts of alloying elements in the weld,
development of the microstructure of this weld
can be followed using the iron-carbon diagram or
a CCT diagram if the cooling rate is known. As-
suming the weld metal admixture contains 0.12%
C, the iron-carbon diagram shows that the first
solid to freeze is ferrite cells or dendrites, de-
pending on G/R.and Co. At 1496 °C (2725 °F), the
8 ferrite begins to transform by a peritectic reac-
tion to austenite through epitaxial growth from
the HAZ following the solidification front. Al-
though some nucleation of austenite occurs at &
ferrite columnar grain boundaries, the majority
of austenitic growth occurs epitaxially in a direc-
tion parallel fo the temperature gradi-
ent. The growing austenite grains replace the pri-
inary 3 ferritic structure virtually grain-upon-
grain with little change in the columnar nature of
the weld structure. Despite the transformation of
8 to austenite, austenite grain boundaries can
original 3
occupy the same positions that the
ferrite grains occupied during solidification with
litle deviation. Therefore, the original 8 ferrite
solidification structure, although replaced by
austenite, remains essentially intact.
‘The austenitic weld metal may transform toa
variety of proeutectoid and eutectoid products,
depending upon weld cooling rate and alloy
composition. At a normal or intermediate cooling
rate, as in a typical shielded metal are weld,
proeutectoid ferrite nucleates at the austenite
grain boundaries as the weld cools below the As
temperature. This continuous networkor veining
of ferrite grains outlines the prior austenite grain
boundaries and is often called “grain-boundary
ferrite.” Within each prior austenite grain, the
structure is normally “acicular ferrite,” which is
characterized by fine elongated platelets ar-
ranged in an interlocking path. At high magnifi-
cation, the acicular ferrite 40 to
80 pin, wide, are separated by local regions of
carbon-enriched austenite which transform to
carbide in the form of bainite. This microstruc-
ture is characterized by excellent strength and
fracture toughness. At slow weld cooling rates
associated with high preheating temperatures or
heat input, the amount of acicular ferrite de-
creases substantially and is replaced by a coarse
structure of additional grain-boundary ferrite
and a second form of proeutectoid ferrite called
“polygonal” or blocky erie
‘At an optimum Cooling rate, a maximum
amount of acicular ferrite (as much as 90% of the
weld) is obtained and any further increase in
cooling rate tends to decrease its presence. For
example, ata fast weld cooling rate acieular fer-
rite is replaced by various forms of bainitic
tures, which usually exhibit ‘Widmenstatten oF or
side-plate ferrite. Ferrite side plates grow from
‘the prior austenite grain into the grain
interior. At these fast cooling rates, the bainitic structure may also be accompanied by regions of
martensiteand someretained austenite. The side-
plate ferrite, bainitic, and martensitic structures
are generally undesirable because of their low
ductility and poor fracture toughness.
Weld-Metal Composition. The alloy composi-
tion of the weld metal also influences the result-
ing microstructure. The weld-metal and HAZ mi-
crostructures can be predicted for numerous steel
compositions if appropriate CCT diagrams are
available. This method of predicting microstruc-
tures can be applied effectively to any fusion
welding process and follows a three-step process
depending on (1) calculation of the maximum
weld cooling rate, (2) procurement or availability
of the appropriate CCT diagrams for the weld
admixture and base plate compositions, and (3)
comparison of the calculated cooling rate with
that given in the CCT diagrams to obtain the
microstructural phases present.
For example, the weld-metal and HAZ micro:
structures may be predicted for a shielded metal
arc weld that is deposited on 12-mm (0.5-in.)
thick 1080 steel using E7018 electrodes and a net
arc energy input (Hq) of 800 J/mm (20,320 J/in.)
without preheating, and with a preheating tem-
perature of 325 °C (615 °F). Knowing the groove
geometry and chemical compositions of both
electrode and base metal, an approximate com-
position of the composite zone admixture can be
calculated. In this example, assume the weld-
metal composition was found to be essentially
equivalent to a 1035 steel. Weld cooling rate for
thin plate without preheating is:
S = 2nKpCttyHe)"(Ti ~ To)
$= 2x (0.028)(0.0044)(13/ 800)" (700~ 25)"
$= 63 °C/s = 3800 °C/min
where T; is the temperature of interest, tis the plate
thickness, is density, and K is thermal conductiv-
ity. In English units, the calculation is:
S = 2n{0.395)(40)(0.5/20,320)? (1292 — 7)"
$=
08 "F/s = 6473 °F/min
xt, taking the effect of 325 °C (615 °F) pre-
heating intoaccount and recalculating using S for
‘To equal to325 °C (615 °F}
$= 108 °C/s = 647°C/min
$= 19.4°F/s = 1165 °F/min
With preheating, the HAZ microstructure contains
pearlite plus bainite, while the microstructure of
the weld metal contains grain-boundary ferrite
plus bainite, Unfortunately, many CCT diagrams
do not presently differentiate among the various
‘Weld Solidification / 15
forms of ferrite (ie., acicular or polygonal). In ad-
dition to eliminating martensite from the weld
joint, the slower cooling rates attained with pre-
hheating provide additional benefits, such as
greatly decreased susceptibility to hydrogen-as-
sisted cold cracking and reduced residual stresses:
and distortion.
Welding Flaws
Porosity, or fine holes or pores within the weld
metal, can occur by absorption of evolved gases
and chemical reaction, Metals susceptible to po-
rosity are those which can dissolve large qua
ties of gas contaminants (hydrogen, oxygen, nk
trogen, etc) in the molten weld pool and
subsequently reject most of the gas during solis
fication. Aluminum alloys are more susceptible
to porosity than any other structural material
‘The mechanism by which porosity forms in
welds is inextricably related to the solidification
kinetics and morphologies of each alloy, The
most common types of porosity are either inter:
dendritic or spherical. Variations in solidification
mode (planar, cellular, or cellular dendritic) will
affect not only the resulting size, shape, and dis-
tribution of hydrogen pores, but also the ability
of pores to detach from the interdendritic regions
to be free floating. The process by which gas
bubbles grow in the weld pool is coalescence of
smaller bubbles in accordance with Stokes law,
and a strong convective fluid flow.
‘Weld cooling rate substantially affects the vol-
ume of porosity retained in a gas-contaminated
weld. At fast cooling rates, the level of porosity is,
low, as the nucleation and growth of bubbles in
the liquid are severely suppressed. Similarly, at
very slow cooling rates, porosity is minimal be-
cause bubbles have ample time to coalesce, float,
and escape from the weld pool. At intermediate
cooling rates, the greatest volume of porosity ina
weld is observed, as conditions are optimum for
both formation and entrapment of virtually all of
the evolved gases in the weld.
‘The sources of porosity are contaminants, in-
dluding moisture, oils, paints, rust, mill scale, and
oxygen and nitrogen in the air. The heat from the
welding arc decomposes moisture, oils, and
paints into hydrogen and other gaseous prod-
ucts. Hydrogen may produce severe porosity in
aluminum alloy welds, as well as welds depos-
ited on copper, magnesium, niobium, and tita-
nium alloys. Because hydrogen is not easily nu-
dleated in steels during solidification, the
majority of hydrogen contamination remains in
supersaturated solid solution after welding is
complete.
The base metal itself may contain gas in appre-
ciable amounts, as in rimmed steels or volatile 16 / Introduction
coatings, such as zinc (from galvanizing) and
cadmium. Volatile ingredients, including sulfur,
lead, and selenium, are required for free-machin-
ing applications. However, during welding,
these additions vaporize and produce excessive
poresity, In addition, the evolution of gases in a
weld may cause and oxide inelu-
sions, as well as porosity.
is also caused to a lesser extent by
chemical’ reactions between absorbed active
gases and easily reducible ingredients in the weld.
pool. For example, in steel, absorption of oxygen
gas causes decarburization:
C+O4 CO gas
In copper alloys that almost always contain some
‘oxygen impurity, absorbed hydrogen reacts.as fol-
lows:
2H + O- FLO gas
Hydrogen accidentally introduced into a welding,
atmosphere tends to reduce most metal oxides:
4H + 2Cu20 -9 4Cu + 2H20 gas
TH+ FeO + Fe +Hz0 gas
Reduction reactions occurring.in the weld pool can
be approximated by using simple thermodynamic
calculations or functional charts, such as the Elling-
ham diagram.
Because the production of porous-free welds is
not cost-effective in commercial welding applica-
tions, limited amounts and sizes of porosity are
permitted by all welding codes. The best preven-
tion for porosity in welds is utilization of accept-
able standards of workmanship, as specified in
various structural welding codes. If parts to be
welded and consumables are cleaned and dried,
great percentage of gaseous porosity is elimi-
nated. If the rust and light mull scale on steel
reducing porosity to acceptable levels. Only hy-
drogen contamination cannot be removed by de-
oxidizers.
Hydrogen induced Cold Cracking, A major
source of underbead cracking in welds.
‘on low-alloy and other hardenable steels is cod
cracking (also called delayed cracking). Cold
cracks may form within minutes, hours, or days
after welding and can result in catastrophic fail-
sures of welded structures, Factors required for
cold cracking to occur are (1) a crack-sensitive
microstructure, usually martensitic; (2) sufficient
‘hydrogen concentration in the weld; (3) rigid ten-
sile restraint; and (4) a temperature between ap-
proximately 300 to -150 °F Elimination ofone or
mere of re ofthese factors gray reduces eack sue
“Fydrogen, when entering asteel weld, diffuses
and segregates at pores, discontinuities, inchi-
sions, and other microscopic flaws. These flaws
are effective traps and can severely reduce the
diffusion coefficient of hydrogen (Fig. 11). Gen-
eral belief, however, is that the remaining diffus-
ible hydrogens responsible forthe end eracking
problems in welds. The mechanism of hydrogen-
Induced cold cracking hasbeen widely studied.
This is known to be diffusion-
controlled, time-dependent, and either trans-
granular or intergranular, ‘Several theories ex-
plain why cold cracking is time-dependent.
Generally, a preexisting microcrack ar disconti-
nuity acts as a stress-concentration site. When a
tensile stress is applied, hydrogen diffuses at
fom temperature to the regions of greatest ten-
sile strain. After the concentration of hydrogen at
or near the tip of the discontinuity has accumu-
lated to a critical value, which depends on the
magnitude of externally applied tensile stresses
or residual stresses, the hydrogen is believed to
cause severe reduction in the cohesive bonding
energy between iron atoms ahead of the discon-
tinuity, and racking initiates, Propagation of the
crack occurs in discrete bursts or steps, which are
repeated as fresh hydrogen diffuses ahead of the
crack tip. At low stress intensity values, cracking
is likely to follow an intergranular path between
prior austenite grains which have transformed to
martensite, while at high stress intensities, the
fracture could be transgranular.
In welding, the combination of tensile shrink-
age stresses and hydrogen contamination may
‘cause microcracks to occur in both the weld fu-
sion zone and HAZ. In fact, cold cracking occurs
more commonly in the HAZ because the hydro-
gen contamination entering the molten weld pool
diffuses rapidly into the HAZ and most steel filler
metals have less carbon than the base metal for
good weldability, making the HAZ microstruc-
ture more susceptible.
‘The cold cracking susceptibility ofa given com-
position of steel is related to the Dearden and
‘ONeill equation for CE:
ce- nc eas a
%oCr+ %V + Yio
oe
‘This formula was derived for plain-carbon and
low-alloy steels containing 0.12% C or more.
Weight percentages are used in the calculation. For Weld Solidification / 17
Temperature, °F
Diffusion coetticient, tag 0, mts
Temperature, °C
Fig. 11. Ditision coeicient of hydrogen in iron alloys. Source: Ref 18
low-carbon steels in the range from (0.07 to 0.22%
C, the Ito and Bessyo equation can be used:
Ni Mo
5
ceac+ $i, MneCu+ ce,
v
ett
Generally, most users of the CE equation agree that
a value of CE > 035 to 040 (depending on plate
thickness and the degree of restraint) indicates that
a given steel composition will be susceptible to
cold cracking in the HAZ unless steps are taken to
reduce the amount of hydrogen contamination en-
tering the molten weld
For example, consider the SMAW of ASTM
A572 grade 65 with a CE of Od to 0.48%. Al-
though susceptible to cold cracking, this grade of
steel is welded routinely without cracking if
proper welding procedure is maintained. Strue-
tural welding codes, such as AWS D1.1, a
that A572 grade 65 be welded with low-hydro-
a, sacri EBOIS, OIG, or BOLE; adequate
preheating temperature, depending on
ness; and a suitable welding procedure to ensure
low-h welding conditions.
‘Because hydrogen enters the weld pool by the
dissociation of moisture and hydrocarbons in the
arc, consumables that are inherently high in
moisture should not be used to weld steels hav-
ing greater than either a 300 HB hardness value
or 0.40% CE. In SMAW, cellulosic and rutile elec-
trodes, such as E6010, E6011, E6012, E6013,
E7014, and E7024, contain organic substances
which evolve abundant amounts of hydrogen
during welding: Similarly, in FCAW, an E70T-1
filler metal contains a rutile base flux and may
evolve substantial hy particularly if not
dried by baking prior to welding, In SAW, fused. 18 / introduction
fluxes contain no water, but agglomerated fluxes
may have substantial amounts of residual mois-
ture because these fluxes are hygroscopic and.
should be baked at a prescribed temperature im-
mediately prior to welding.
Te eorure crocefoe welling of banenable
steels, low-hydragen covered electrodes, such as
E7015, E7016, and E7018 for SMAW, must be
baked at 450 to 500 °F for at least 2 h prior to
welding, Higher strength electrodes, such a3,
A016, E9018; and E1OOIB, must be dried at 700t0
800°F for crack prevention. Fused fluxes for SAW
may be dried 9825) TF 1g eliminate moisture.
Preheating is essential to occur
moderne ina, ome
content of the wald & weld is substantially eluced by aby
preheating and maintaining an interpass tem
perature such as 250 °C (480 °F), At ay
mately 250 *C, no hydrogen-induced Te.
ing is possible because hydrogen diffuses so
rapidly that it will not segregate at the tips of
discontinuities or stress concentrations. Eventu-
ally, the rapidly diffusing hydrogen atoms es-
cape from the weld surface.
‘Thelevel of diffusible hydrogen ina given weld
can be measured by standard techniques that are
i military specification MIL-E-
and several other specifications. Once the hydro-
gen content is determined, an upper limit on the
Permited hydrogen level in the weld joint can be
fet based on actualexpertonce The
Institute of Welding’s general yor
Mine Hydroneneonenentons ine is is:
Hysiogen
concentration
Tem ‘mLroe gia
Very low os
Lew 50
Mediom 105
High 6
2) Miliitersof hydrogen per 100g of weld metal deposited
Ferrite Vein Cracking, Another form of hydro-
.gen-induced cracking has been observed in elec-
troslag weld metal. Unlike cold cracking of high-
cracking such 0s ASE and BBS. Although the
of cracking is unknown, the cause
fas boen deSnelyattcbened tothe presence of
hy ing. contaminants in the weld
pool. The cure is simply to ensure that all fluxes
and consumables are sufficiently free of mais-
ture, grease, oil, and excessive oxide scale.
‘Lamellar Tearing. This type of cracking, which.
occurs in the base metal or HAZ of restrained
weld joints, results from inadequate ductility in
the through-thickness direction of steel plate.
to lamellar tearing on the
metry, oxide and sulfide inclusion con-
tent and ine ocent io which these inclaions ae
of area values from tensile tests on steels taken in
the through-thickness direction indicates suscep-
tibility to lamellar tearing. Steel plates exhibiting,
reduction of area values less than 10% will be
sensitive to cracking, Hj is known to be
preerevaly topper nc sions and tends 0
occurrence of lamellar tearing.
Therefor, wrids on sels deposted by low-hy.
plates can be buttered with crack-resistant weld
metal. When joining two buttered plates, the ma-
jor tensile stresses. act on the relatively immune
buttered regions ofthe joint Thnmost somomi
‘cal preventative hen possible, is to
assemble the plates to be welded with the
direction of the plate perpendicular to the
axis.
Reheat Cracking. In welding high-strength
Jow-alloy steels that have been alloyed with va-
‘nium to promote the formation of acicular ferrite
for maximum. stress relief is
though not dethitely established, the mechanism
is believed to be a result of precipitation harden-
ing within the HAZ grains while the grain.
boundaries are left in a weakened condition. If
pensity for a steel to be subject to reheat cracking
can be qualitatively measured by the Ito equa-
tion, which is calculated using weight percent-
ages:
C= 10V + 7Nb+5Ti+Cr+Cu-2
If carbon (C) is equal to or greater than zero, the
steel may be susceptible to reheat cracking. Al-
though boron ls not canskered in the in equation
steels containing, boron,
thon with 12% Mo, ave been suspeced of contri
uting indirectly to reheat cracking, In addition, the presence of nitrogen, sulfur, phosphorus, arsenic,
antimony, and aluminum is suspected of contrib-
luting indirectly to the reheat cracking problem.
The best prevention for reheat cracking is to use a
nonsusceptible alloy. If this is not possible, the
stress relief anneal must not be performed, at least
not within the temperature range of maximum.
susceptibility, which is usually between 550 to 650°
°C (1020 to 1200 °F), Finally, reducing external
restraint and minimizing stress concentrations
through proper design and good welding practice
are helpful in moderating susceptibility to reheat
cracking.
Hot Cracking in the Weld Metal and Heat-Af-
fected Zone. Another type of cracking, which
results from internal stress developed on cooling
following solidification, is hot cracking. This de-
fect occurs at a temperature above the solidus of
an allay. As discussed earlier, dendritically so-
lidified weld metal is characterized by composi-
tional differences between the dendrite core and
the solute-rich, lower melting interdendritic re-
gions. During the last stages of solidification, a
semirigid network of mechanically interlocked
dendrites forms with a small amount of low-
melting liquid interdispersed. As the weld
shrinks during solidification, tensile stresses tend.
to pull the loosely bonded dendrites apart. This
separation is always interdendritic or intercellu-
lar and results in hot cracked weld metal.
‘The amount of interdispersed liquid is a critical
factor in determining the susceptibility of an al-
loy to hat cracking. In the eutectic system shown
in Fig. 12, analloy weld of composition C} solid
fies in the portion of the phase diagram that
most susceptible to hot cracking, because the
greatest solid and liquid temperature ranges oc-
cur at C1, and because the amount of liquid re-
maining after an interdendlritic separation or
crack has taken place is not sufficient to fill ar
heal the crack as a riser would in a casting. If an
alloy of composition C2 is welded, the salidifica-
tion temperature range decreases and, more im-
portantly, a plentiful amount of liquid of eutectic
composition is available to heal any interden-
dritic crack that may form. Finally, the alloy of
composition C3 is virtually immune to hot crack-
ing under normal welding practice because each
dendrite is surrounded by eutectic liquid that
solidifies at one temperature in a planar mode
and provides ample eutectic healing of any den-
dritic separation.
Very offen in complex alloys, hot cracking is
caused by impurity liquid films such as the low-
melting sulfide film in steels, stainless steels, and
nickel alloys. In welding low-alloy steels, for ex-
ample, weld-metal carbon content should be kept
below 0.1% so the weld solidifies as 5 ferrite and
avoids peritectic reactions. Sulfur dissolves read-
ily in 6 ferrite but is relatively insoluble in
Weld Solidification / 19
Temperature =
Solute, % =
ig, 12. ect of comparition on bot cack
Fig. 12 hs na ouch
cepts A, no cracing Bhd es
rack sansitwe sure Bat 19
austenite, As a result, austenite rejects sulfur to
cell or dendrite boundaries and forms an inter-
granular low-melting film, which is partially re-
sponsible for hot cracking during solidification of
many iron- and nickel-based alloys. Fortunately,
increasing the manganese content, as shown in
Fig, 10, helps reduce hot cracking sensitivity by
raising the melting temperature of the sulfides.
Other elements, such as phosphorus and nio-
1bium, can be identified as promoters of hot erack-
ing in steels using this formula developed by
Bailey and Jones:
UCS = 230% C" + 190% $+ 75%P + 45% Nb
—123% Si-5.4%Mn_
IFC < 0.08%, Cis taken to be 0.08% C. If units of
crack susceptibility, ICS, exceed 19 for T-fillet
welds and 25 for butt welds, the chances for hot
cracking become an important consideration in
material, design, and production choices. Methods
tominimize hot cracking in weld metal include: (1)
maintenance of adequate manganese-to-sulfur ra-
tio; (2) reduction of sulfur, phosphorus, carbon,
and niobium to minimal amounts; (3) production
of weld bead convexes, such as slightly overfilling
the weld groove or fillet; andl (4) reduction of the
tensile restraint exerted on the weld.
Hot cracking may also occur in the partially
melted zone by the liquation of low-melting in- 20 f introduction
clusions or second-pha: tes and grain-
melting. Tine wilh at the partially
melted zone depends on the nature of the segre-
and the thermal gradient perpendicular to
the weld interface. The liquation of large manga-
nese sulfide inclusions was described in the sec-
tion on definition of a weld in this article. If the
grain boundary in the HAZ is segregated with a
low meleng phase, cracking crn Be exenalve. For
example, welding of 304 stainless steel with 308,
filler metal is occasionally contaminated with
copper from various types of tooling or other
sources. The accidental addition of copper to the
weld poo! causes not only severe hot cracking of
the weld metal, but also extensive grain-bound-
ary melking or hot cracking in the partially
melted zone by a liquid-metal embrittlement
mechanism.
Chevron cracking, also referred to as staircase
racking, can occur in the fusion zones of both
submerged arc welds using sgplomerated fluxes
and shielded metal arc welds deposited on me-
dium-strength steels. This type of cracking is
characterized by many ‘cracks angled ap-
proximately 45° from the plane of the plates in a
buttjoint. The multiple eracks are best viewed by
cutting a longitudinal section of a weld joint. A
chevron pattern is produced by the 90° intersec-
tions of these cracks, which are partially interco-
lumnar and partially transcolumnar within the
sped feign zone Caused by the presence of hy-
chevron cracking is apparently a form of
hydogerinduead cold’ cracking. The exact
mechanism of cracking is unexplained, although
low-hydrogen welding practice, particularly in
SAW and SMAW, is known to eliminate chevron
“Duct. -Dip Cracking. Many alloys, includ-
ing cupronickels, austenitic chromium-nickel
se some nike-based slays, and aluminum
temperature range over which
duculty suet strength drop sharply. These
alloys are susceptible to HAZ and weld-metal
cracking when welded under excessive restraint.
‘Aboreand below the ductility cipembritlement
range, fracture occurs by microvoid
resulting in high wale 9f serge and duceliy.
Within the embrittlement range, fractures inter-
and brittle.
‘The actual embrittling mechanism is not well
understood, but is believed in most instances to
be caused by the grain-boundary segregation of
and sulfur in austenitic chromium-
nickel steels, nickel in cupronickels, and alumi-
numand nickel in aluminum bronzes. Crack sus-
ceptibility is extremely composition-dependent.
For example, increasing nickel content in cupro-
nickel alloys to above approximately 18% tends
to decrease ductility at 700 °C (1290°F). Similarly,
‘bronzes containing less than approxi-
mately 11% Al and an increased amount of nickel
tend io exhibit reduced ductility in the tempera-
ture range of approximately 500 to 600 °C (930 to
1110 °F), In some cases, cracking may be initiated
by hot cracking followed by ductility-dip crack-
ing when the segregating constituents not only
embrittle grain-boundary regions in the ductil-
ity-dip temperature range, but also substantially
depress the melting temperature at the grain
boundaries.
Fortunately, ductility-
