CIDECT - DG9

11.1.3 Welding

This is suitable for all hollow sections and typical details are shown in figure 11.5. The
welds can be full or partial penetration butt (groove) welds and should be designed to carry

Figure 11.5 – Plane column splices – site welding with a weld backing piece

the required loads specified by the design code or specification being used. Weld backing
pieces are normally required for full depth welds.

The welds should be made to their full size around the whole periphery of the hollow section.

11.1.4 Welded column splices in seismic areas

This section describes design requirements and practices for column splices to be used
for moment resisting and braced frames in seismic areas. Column splices are important
elements to maintain integrity of structures because the splices are, like columns them-
selves, critical gravity load bearing elements and therefore should be designed to allow a
significant safety margin.

The AISC Seismic Provisions (1997a, 2000) specify that column splices in special moment
frames should not be located within 1200 mm of the beam-to-column connections. If the
column clear height is less than 2400 mm, the splice can be located at the mid-height of
the column. These requirements are intended not only to reduce flexural demand on the
splice but also to simplify field erection and construction due to increased accessibility.
Additionally, the AISC provisions require to use CJP groove welded joints for the column
splices of special moment frames. Then, the column splices have strength closely com-
parable to the columns and no further strength calculation is required.

The AIJ Design Recommendations (2001) allow the use of PJP groove welded joints for the
column splices in moment resisting and braced frames, provided that stresses in the col-
umn splices are kept within an elastic regime. Note that welded joints should be continu-
ous throughout the periphery of the column as shown in figure 11.6.

176
16.1–114 WELDS [Sect. J2.
AISC 360-10

TABLE J2.5
Available Strength of Welded Joints,
ksi (MPa)
Nominal Effective
Stress Area
Load Type and (FnBM or (ABM or Required Filler
Direction Relative Pertinent Fnw) Awe) Metal Strength
to Weld Axis Metal φ and Ω ksi (MPa) in.2 (mm2) Level [a][b]
COMPLETE-JOINT-PENETRATION GROOVE WELDS
Matching filler metal shall
be used. For T- and
Tension Strength of the joint is controlled corner joints with backing
Normal to weld axis by the base metal left in place, notch tough
filler metal is required.
See Section J2.6.
Filler metal with a
strength level equal to
Compression Strength of the joint is controlled
or one strength level
Normal to weld axis by the base metal
less than matching
filler metal is permitted.
Filler metal with a
Tension or Tension or compression in parts joined parallel
strength level equal to
compression to a weld need not be considered in design
or less than matching
Parallel to weld axis of welds joining the parts.
filler metal is permitted.
Shear Strength of the joint is controlled Matching filler metal
by the base metal shall be used.[c]
PARTIAL-JOINT-PENETRATION GROOVE WELDS INCLUDING FLARE V-GROOVE
AND FLARE BEVEL GROOVE WELDS
φ = 0.75
Base Fu See J4
Tension Ω = 2.00
Normal to weld axis φ = 0.80
Weld 0.60FEXX See J2.1a
Ω = 1.88
Compression
Column to base plate
Compressive stress need not be considered
and column splices
in design of welds joining the parts.
designed per
Section J1.4(1)
Compression φ = 0.90
Base Fy See J4
Connections of Ω = 1.67
Filler metal with a
members designed
strength level equal
to bear other than φ = 0.80
Weld 0.60FEXX See J2.1a to or less than
columns as described Ω = 1.88 matching filler metal
in Section J1.4(2)
is permitted.
φ = 0.90
Compression Base Fy See J4
Ω = 1.67
Connections not
finished-to-bear φ = 0.80
Weld 0.90FEXX See J2.1a
Ω = 1.88
Tension or Tension or compression in parts joined parallel
compression to a weld need not be considered in design
Parallel to weld axis of welds joining the parts.
Base Governed by J4
Shear φ = 0.75
Weld 0.60FEXX See J2.1a
Ω = 2.00

Specification for Structural Steel Building, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
16–106 GENERAL PROVISIONS [Sect. J1.

3. Moment Connections
End connections of restrained beams, girders and trusses shall be designed for the com-
bined effect of forces resulting from moment and shear induced by the rigidity of the
connections. Response criteria for moment connections are provided in Section B3.6b.

User Note: See Chapter C and Appendix 7 for analysis requirements to establish
the required strength for the design of connections.

4. Compression Members With Bearing Joints

Compression members relying on bearing for load transfer shall meet the following
requirements:
(1) When columns bear on bearing plates or are finished to bear at splices, there shall
be sufficient connectors to hold all parts securely in place.
(2) When compression members other than columns are finished to bear, the splice
material and its connectors shall be arranged to hold all parts in line and their
required strength shall be the lesser of:
(i) An axial tensile force of 50% of the required compressive strength of the
member; or
(ii) The moment and shear resulting from a transverse load equal to 2% of the
required compressive strength of the member. The transverse load shall be
applied at the location of the splice exclusive of other loads that act on the
member. The member shall be taken as pinned for the determination of the
shears and moments at the splice.

User Note: All compression joints should also be proportioned to resist any ten-
sion developed by the load combinations stipulated in Section B2.

5. Splices in Heavy Sections

When tensile forces due to applied tension or flexure are to be transmitted through
splices in heavy sections, as defined in Sections A3.1c and A3.1d, by complete-joint-
penetration groove (CJP) welds, the following provisions apply: (1) material
notch-toughness requirements as given in Sections A3.1c and A3.1d; (2) weld access
hole details as given in Section J1.6; (3) filler metal requirements as given in Section
J2.6; and (4) thermal cut surface preparation and inspection requirements as given in
Section M2.2. The foregoing provision is not applicable to splices of elements of
built-up shapes that are welded prior to assembling the shape.

User Note: CJP groove welded splices of heavy sections can exhibit detrimental
effects of weld shrinkage. Members that are sized for compression that are also
subject to tensile forces may be less susceptible to damage from shrinkage if they
are spliced using partial-joint-penetration PJP groove welds on the flanges and fil-
let-welded web plates, or using bolts for some or all of the splice.

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
Sect. A3.] MATERIAL 16.1–7

(6) Sheets
ASTM A606/A606M
ASTM A1011/A1011M SS, HSLAS, AND HSLAS-F
1b. Unidentified Steel
Unidentified steel, free of injurious defects, is permitted to be used only for members
or details whose failure will not reduce the strength of the structure, either locally or
overall. Such use shall be subject to the approval of the engineer of record.

User Note: Unidentified steel may be used for details where the precise mechan-
ical properties and weldability are not of concern. These are commonly curb
plates, shims and other similar pieces.

1c. Rolled Heavy Shapes

ASTM A6/A6M hot-rolled shapes with a flange thickness exceeding 2 in. (50 mm)
are considered to be rolled heavy shapes. Rolled heavy shapes used as members
subject to primary (computed) tensile forces due to tension or flexure and spliced
or connected using complete-joint-penetration groove welds that fuse through the
thickness of the flange or the flange and the web, shall be specified as follows.
The structural design documents shall require that such shapes be supplied with
Charpy V-notch (CVN) impact test results in accordance with ASTM A6/A6M,
Supplementary Requirement S30, Charpy V-Notch Impact Test for Structural Shapes
– Alternate Core Location. The impact test shall meet a minimum average value of
20 ft-lb (27 J) absorbed energy at a maximum temperature of +70 °F (+21 °C).

The above requirements do not apply if the splices and connections are made by
bolting. Where a rolled heavy shape is welded to the surface of another shape using
groove welds, the requirement above applies only to the shape that has weld metal
fused through the cross section.

User Note: Additional requirements for joints in heavy rolled members are given
in Sections J1.5, J1.6, J2.6 and M2.2.

1d. Built-Up Heavy Shapes

Built-up cross sections consisting of plates with a thickness exceeding 2 in. (50 mm)
are considered built-up heavy shapes. Built-up heavy shapes used as members sub-
ject to primary (computed) tensile forces due to tension or flexure and spliced or
connected to other members using complete-joint-penetration groove welds that
fuse through the thickness of the plates, shall be specified as follows. The structural
design documents shall require that the steel be supplied with Charpy V-notch
impact test results in accordance with ASTM A6/A6M, Supplementary
Requirement S5, Charpy V-Notch Impact Test. The impact test shall be conducted
in accordance with ASTM A673/A673M, Frequency P, and shall meet a minimum
average value of 20 ft-lb (27 J) absorbed energy at a maximum temperature of
+70 °F (+21 °C).

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
Sect. J1.] GENERAL PROVISIONS 16.1–107

6. Weld Access Holes

All weld access holes required to facilitate welding operations shall be detailed to
provide room for weld backing as needed. The access hole shall have a length from
the toe of the weld preparation not less than 11/2 times the thickness of the material
in which the hole is made, nor less than 11/2 in. (38 mm). The access hole shall have
a height not less than the thickness of the material with the access hole, nor less than
3
/4 in. (19 mm), nor does it need to exceed 2 in. (50 mm).

For sections that are rolled or welded prior to cutting, the edge of the web shall be
sloped or curved from the surface of the flange to the reentrant surface of the access
hole. In hot-rolled shapes, and built-up shapes with CJP groove welds that join the
web-to-flange, weld access holes shall be free of notches and sharp reentrant corners.
No arc of the weld access hole shall have a radius less than 3/8 in. (10 mm).

In built-up shapes with fillet or partial-joint-penetration groove welds that join the
web-to-flange, weld access holes shall be free of notches and sharp reentrant cor-
ners. The access hole shall be permitted to terminate perpendicular to the flange,
providing the weld is terminated at least a distance equal to the weld size away from
the access hole.

For heavy sections as defined in Sections A3.1c and A3.1d, the thermally cut surfaces
of weld access holes shall be ground to bright metal and inspected by either magnetic
particle or dye penetrant methods prior to deposition of splice welds. If the curved
transition portion of weld access holes is formed by predrilled or sawed holes, that
portion of the access hole need not be ground. Weld access holes in other shapes need
not be ground nor inspected by dye penetrant or magnetic particle methods.
7. Placement of Welds and Bolts
Groups of welds or bolts at the ends of any member which transmit axial force into
that member shall be sized so that the center of gravity of the group coincides with
the center of gravity of the member, unless provision is made for the eccentricity. The
foregoing provision is not applicable to end connections of single angle, double
angle and similar members.
8. Bolts in Combination With Welds
Bolts shall not be considered as sharing the load in combination with welds, except
that shear connections with any grade of bolts permitted by Section A3.3, installed
in standard holes or short slots transverse to the direction of the load, are permitted
to be considered to share the load with longitudinally loaded fillet welds. In such con-
nections the available strength of the bolts shall not be taken as greater than 50% of
the available strength of bearing-type bolts in the connection.

In making welded alterations to structures, existing rivets and high-strength bolts

tightened to the requirements for slip-critical connections are permitted to be utilized
for carrying loads present at the time of alteration and the welding need only provide
the additional required strength.

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
16.1–118 WELDS [Sect. J2.

are designed using the nominal strength and resistance factor or safety factor as
applicable for a partial-joint-penetration groove weld
(2) Complete-joint-penetration groove welded splices subject to tension normal to
the effective area in heavy sections as defined in Sections A3.1c and A3.1d

The manufacturer’s Certificate of Conformance shall be sufficient evidence of com-

pliance.
7. Mixed Weld Metal
When Charpy V-notch toughness is specified, the process consumables for all weld
metal, tack welds, root pass and subsequent passes deposited in a joint shall be com-
patible to ensure notch-tough composite weld metal.

J3. BOLTS AND THREADED PARTS

1. High-Strength Bolts
Use of high-strength bolts shall conform to the provisions of the Specification for
Structural Joints Using High-Strength Bolts, hereafter referred to as the RCSC
Specification, as approved by the Research Council on Structural Connections,
except as otherwise provided in this Specification. High-strength bolts in this
Specification are grouped according to material strength as follows:
Group A—ASTM A325, A325M, F1852, A354 Grade BC, and A449

Group B—ASTM A490, A490M, F2280, and A354 Grade BD

When assembled, all joint surfaces, including those adjacent to the washers, shall be
free of scale, except tight mill scale.

Bolts are permitted to be installed to the snug-tight condition when used in:
(a) bearing-type connections except as noted in Section E6 or Section J1.10
(b) tension or combined shear and tension applications, for Group A bolts only,
where loosening or fatigue due to vibration or load fluctuations are not design
considerations

The snug-tight condition is defined as the tightness required to bring the connected
plies into firm contact. Bolts to be tightened to a condition other than snug tight shall
be clearly identified on the design drawings.

All high-strength bolts specified on the design drawings to be used in pretensioned

or slip-critical joints shall be tightened to a bolt tension not less than that given in
Table J3.1 or J3.1M. Installation shall be by any of the following methods: turn-of-
nut method, a direct-tension-indicator, twist-off-type tension-control bolt, calibrated
wrench, or alternative design bolt.

User Note: There are no specific minimum or maximum tension requirements for
snug-tight bolts. Fully pretensioned bolts such as ASTM F1852 or F2280 are per-
mitted unless specifically prohibited on design drawings.

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
16.1–384 GENERAL PROVISIONS [Comm. J1.

connection must have sufficient strength, stiffness and deformation capacity to sat-
isfy the design assumptions.
4. Compression Members with Bearing Joints
The provisions for “compression members other than columns finished to bear” are
intended to account for member out-of-straightness and also to provide a degree of
robustness in the structure to resist unintended or accidental lateral loadings that may
not have been considered explicitly in the design.

A provision analogous to that in Section J1.4(2)(i), requiring that splice materials and
connectors have an available strength of at least 50% of the required compressive
strength, has been in the AISC Specifications since 1946. The current Specification
clarifies this requirement by stating that the force for proportioning the splice mate-
rials and connectors is a tensile force. This avoids uncertainty as to how to handle
situations where compression on the connection imposes no force on the connectors.

Proportioning the splice materials and connectors for 50% of the required member
strength is simple, but can be very conservative. In Section J1.4(2)(ii), the
Specification offers an alternative that addresses directly the design intent of these
provisions. The lateral load of 2% of the required compressive strength of the mem-
ber simulates the effect of a kink at the splice, caused by an end finished slightly
out-of-square or other construction condition. Proportioning the connection for the
resulting moment and shear also provides a degree of robustness in the structure.
5. Splices in Heavy Sections
Solidified but still hot weld metal contracts significantly as it cools to ambient tem-
perature. Shrinkage of large groove welds between elements that are not free to move
so as to accommodate the shrinkage causes strains in the material adjacent to the
weld that can exceed the yield point strain. In thick material the weld shrinkage is
restrained in the thickness direction, as well as in the width and length directions,
causing triaxial stresses to develop that may inhibit the ability to deform in a ductile
manner. Under these conditions, the possibility of brittle fracture increases.

When splicing hot-rolled shapes with flange thickness exceeding 2 in. (50 mm) or
heavy welded built-up members, these potentially harmful weld shrinkage strains
can be avoided by using bolted splices, fillet-welded lap splices, or splices that com-
bine a welded and bolted detail (see Figure C-J1.1). Details and techniques that
perform well for materials of modest thickness usually must be changed or supple-
mented by more demanding requirements when welding thick material.

The provisions of AWS D1.1/D1.1M (AWS, 2010) are minimum requirements that
apply to most structural welding situations. However, when designing and fabricat-
ing welded splices of hot-rolled shapes with flange thicknesses exceeding 2 in. (50
mm) and similar built-up cross sections, special consideration must be given to all
aspects of the welded splice detail:
(1) Notch-toughness requirements are required to be specified for tension members;
see Commentary Section A3.

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
Comm. J1.] GENERAL PROVISIONS 16.1–385

(2) Generously sized weld access holes (see Section J1.6) are required to provide
increased relief from concentrated weld shrinkage strains, to avoid close juncture
of welds in orthogonal directions, and to provide adequate clearance for the exer-
cise of high quality workmanship in hole preparation, welding, and for ease of
inspection.
(3) Preheating for thermal cutting is required to minimize the formation of a hard
surface layer. (See Section M2.2.)
(4) Grinding of copes and weld access holes to bright metal to remove the hard sur-
face layer is required, along with inspection using magnetic particle or
dye-penetrant methods, to verify that transitions are free of notches and cracks.
In addition to tension splices of truss chord members and tension flanges of flexural
members, other joints fabricated from heavy sections subject to tension should be
given special consideration during design and fabrication.

Alternative details that do not generate shrinkage strains can be used. In connections
where the forces transferred approach the member strength, direct welded groove
joints may still be the most effective choice.

Earlier editions of this Specification mandated that backing bars and weld tabs be
removed from all splices of heavy sections. These requirements were deliberately
removed, being judged unnecessary and, in some situations, potentially resulting in
more harm than good. The Specification still permits the engineer of record to spec-
ify their removal when this is judged appropriate.

The previous requirement for the removal of backing bars necessitated, in some sit-
uations, that such operations be performed out-of-position; that is, the welding
required to restore the backgouged area had to be applied in the overhead position.
This may necessitate difficult equipment for gaining access, different welding
equipment, processes and/or procedures, and other practical constraints. When box
sections made of plate are spliced, access to the interior side (necessary for backing
removal) is typically impossible.

(a) Shear plate welded (b) Shear plate welded (c) Bolted splice plates
to web to flange tips

Fig. C-J1.1. Alternative splices that minimize weld restraint tensile stresses.

Specification for Structural Steel Buildings, June 22, 2010

AMERICAN INSTITUTE OF STEEL CONSTRUCTION
the cracking tendency. Vertical-down welding also has the hydrogen can cause cracking as soon as the steel cools to
tendency to generate these crack-sensitive, concave surfaces. room temperature. The more common and the more prob-
Vertical-up welding can remedy this situation by providing lematic situation is when there is some delay.
a more convex bead. For gas-shielded processes, such as The natural implication of the delayed nature of the crack-
GMAW and FCAW-G, a change in shielding gas can change ing is that weldments can be inspected and accepted before
the surface profile. cracking has occurred. For this reason, AWS D1.1 requires
a delay of 48 hours after completing the welds before NDT
Weld Pool Length Cracking is preformed on quenched and tempered ASTM A514 and
A517, 100 ksi yield strength steels, which are particularly
There is a fourth cause of centerline cracking that is caused
sensitive to hydrogen cracking.
when the weld pool (puddle) becomes long and tear-drop
Many theories explaining the mechanics of hydrogen
shaped. Typically, this is associated with higher welding
cracking have been proffered over the years, and no single
travel speeds that are not commonly used for structural steel
theory appears to adequately explain all observed hydrogen-
applications. Thus, this topic will not be addressed in this
related phenomenon (Gibala and Hehemann, 1984). Current
Guide.
thinking is that hydrogen interferes with the movement of
dislocations, imperfections in the atomic lattice that permit
5.3.2 Heat-Affected Zone Cracking
inelastic deformations (Bailey, 1994). The lack of a definitive
Heat-affected zone (HAZ) cracking is characterized by explanation, however, has not precluded the development of
separation that occurs in the region immediately adjacent to good practices that mitigate the occurrence of cracking.
the weld bead (Figure 5–9). The cracking occurs in the base
material, and although it is certainly related to the welding A Sufficient Level of Hydrogen
process, the crack does not occur in the weld material. This
Hydrogen cracking cannot occur without a sufficient level
type of cracking is also known as “underbead cracking,” “toe
of hydrogen. Hydrogen can be introduced into the weld pool
cracking,” or “delayed cracking.”
from the arc. The arc does not create hydrogen, but rather
Heat-affected zone cracking is one form of cold cracking;
breaks down hydrogen-bearing compounds introduced into
it cannot occur when the steel is hot. For structural steels, it
the arc region. Water and oils are the chief sources of such
is unlikely to occur if the temperature is above 300 °F (Bai-
compounds. Water may be introduced through shielding
ley, 1994).
fluxes, whether they are coatings on the outside of SMAW
Heat-affected zone cracking occurs due to three factors,
electrodes or granular fluxes used for SAW. Water may be
each of sufficient magnitude to cause this phenomenon, as
present on the steel in the form of condensation (including
follows:
condensation of the combustion byproducts from preheating
• A sufficient level of hydrogen. torches) or moistened scale and rust. Humid atmospheres
are also sources of hydrogen. Oils, grease, paint, and other
• A susceptible HAZ microstructure. materials generally considered contaminants on the steel
surface can introduce hydrogen-bearing compounds. Exces-
• Applied or residual stresses. sive lubricants on the filler metal, whether residual from the
Some authors will add one or two additional factors, in-
cluding the aforementioned temperature (e.g., a low temper-
ature must be reached) and time (e.g., a sufficient length of
time after the steel cools must transpire in order for this type
of cracking to occur). These two factors are inevitable (i.e.,
the steel must cool, and time will elapse) and are thus not
included here as primary factors. As will be discussed below,
there are ways that time and temperature can be manipulated
to reduce cracking tendencies.
The fact that HAZ cracking is delayed, however, warrants
some discussion. First, the time element gives rise to the de-
scription of “delayed cracking.” The reason for the delay is
that it takes time for the diffusible hydrogen to move in the
steel to form concentrations that will cause cracking. This
is sometimes called an incubation period, and estimates for
the required time range from 16 to 72 hours (Bailey, 1994).
Some authors suggest that, if present in excessive quantities, Figure 5–9. Heat-affected zone cracking.

72 / DESIGN GUIDE 21 / WELDED CONNECTIONS—A PRIMER FOR ENGINEERS

manufacturing process, or deliberately added, may add Applied or Residual Stress
hydrogen-bearing compounds into the arc. Finally, although
The final component required for HAZ cracking is stress,
it is rarely the case, hydrogen can be contained within the
either applied or residual. As was discussed in Section 5.2.1,
steel itself, either in the base metal, or in the metallic com-
in as-welded assemblies, yield-point residual tensile stresses
ponents of the filler metal.
are always present. While residual stresses immediately after
When these compounds are introduced into the arc, the
welding cannot be eliminated, they can be reduced. These
compounds are broken down into their elemental forms,
measures are discussed in Sections 5.6, 5.7 and 5.8 of this
including the release of atomic hydrogen. This hydrogen is
chapter.
readily absorbed into the molten weld pool. During solidi-
fication and cooling, much of this hydrogen is released. As
Limiting Hydrogen
the steel cools further, however, the rate of release decreas-
es, and some hydrogen will be retained in the weld when it To limit HAZ cracking, all three contributing factors are
reaches room temperature. typically controlled. The level of retained hydrogen should
be minimized. This is best accomplished by not introducing
A Susceptible HAZ Microstructure hydrogen-bearing compounds into the arc in the first place.
Thus, base metal surfaces should be clean and filler metals
In order for HAZ cracking to occur, the HAZ must have a
should be dry. AWS D1.1 requires that when SMAW is used
susceptible microstructure. The HAZ is the narrow region 345N/mm 2

on steel with a minimum specified yield strength of 50 ksi or

next to the weld metal, the part of the base metal that is
greater, electrodes with low-hydrogen coverings are required
heated by welding to a temperature lower than that required
for prequalified WPSs. This includes ASTM A992, which is
to melt the steel. While the chemistry in this region is un-
the primary material used for W-shapes today. Also, AWS
changed, the mechanical properties of the material may be
D1.1 prescribes exposure limits for filler metals to ensure
significantly affected, depending on the composition of the
that excessive moisture pickup from the atmosphere does not
steel and the cooling rate experienced by the HAZ. A sus-
occur.
ceptible microstructure is typically hard and brittle.
Should excessive hydrogen be suspected, a post-heat
To avoid the development of a susceptible microstructure,
hydrogen treatment can mitigate such concerns. Post-heat
two basic approaches can be employed. The first method is
involves heating the weld area to a temperature of 400
to select base metals that are low in carbon and low in al- 200C
to 450 °F and holding the assembly at that temperature an
loy content. The carbon content determines the maximum
hour for each inch of weld thickness. At this temperature, the
hardness that can be achieved by quenching, and alloys de-
mobility of hydrogen increases so that most of the hydrogen
termine the material’s hardenability, which is defined as that
will diffuse out of the region. It is essential, however, that the
property of a ferrous alloy that determines the depth and dis-
welded region not be allowed to cool to room temperature
tribution of hardness induced by quenching (Stout, 1987).
before the post-heat is applied. Recall that hydrogen crack-
The second means used to avoid the development of a sus-
ing cannot occur until the steel becomes cool. Accordingly,
ceptible microstructure is to control the cooling rate of the
if a hot, just-welded region is immediately subject to post-
HAZ. The HAZ will become hot in arc welding, but it need
heat, cracking cannot occur. During the post-heat, most of
not automatically cool rapidly. Cooling rates are controlled
the hydrogen will be released, and when the assembly cools
by temperature differentials; the greater the difference, the
to room temperature, cracking will not occur since one of the
faster the cooling rate, and correspondingly, the greater the
three requirements for this type of cracking (e.g., sufficient
possibility of the development of a sensitive HAZ. Cooling
hydrogen) is no longer present.
rates increase under these conditions:
• Low ambient temperatures. Limiting Sensitive HAZs

• Low steel temperature (i.e., low preheat temperature). Many measures are contained within AWS D1.1 to limit the
development of susceptible microstructures, beginning with
• Low interpass temperature. the list of steels, whether prequalified or code approved.
Ambient welding conditions are controlled, and minimum
• Thicker steel sections. preheat and interpass temperatures are specified for prequal-
ified WPSs, as a function of the steel grade and thickness
• Lower welding heat input. used. The minimum prequalified fillet weld size, discussed
in Section 3.2.2 of this Guide, is a means of controlling the
When a steel with enough carbon and alloy is cooled rapidly, heat input.
a susceptible microstructure can be formed. A variety of carbon equivalency (CE) formulas have been
empirically derived over the years, each developed with the

DESIGN GUIDE 21 / WELDED CONNECTIONS—A PRIMER FOR ENGINEERS/ 73

strength, overmatching the base material. Transverse crack- erally necessary to develop sufficient longitudinal stress to
ing is another form of cold cracking, and like HAZ cracking, cause this type of cracking. As the weld bead shrinks lon-
is driven by the same three factors: excessive hydrogen, a gitudinally, the surrounding base material resists this force
susceptible microstructure, and applied or residual stress. by going into longitudinal compression. The strength of the
surrounding steel in compression restricts the ability of the
Hydrogen weld material to shrink. When the weld material is high in
strength, it has a reduced capacity to plastically deform. Due
The sources of hydrogen for this type of cracking are no dif-
to the restraint of the surrounding base material, the weld
ferent than those identified in Section 5.3.2 for HAZ crack-
metal may crack in the transverse direction.
ing. A feature associated with transverse cracking is that
multiple-pass welds are often involved, and the thicknesses
Solutions
of steel involved are typically over 2 in.. As has been men-
tioned, hydrogen begins to diffuse out of welds as soon as When transverse cracking is encountered, the strength of the
weld metal is deposited. In the case of multiple-pass welds, actual weld metal deposit should also be reviewed. Empha-
if the time between weld passes is short, the hydrogen that sis is placed upon the actual weld metal deposit because the
is trying to leave the previous deposit finds a new layer of filler metal may deposit lower-strength, more ductile metal
metal (the next weld pass) through which it must diffuse. Ac- under normal conditions. However, with the influence of
cordingly, hydrogen contents can increase when individual alloy pickup and welding procedures that encourage rapid
weld passes in multiple-pass welds are rapidly deposited on cooling rates (e.g., small weld passes made with low heat-
top of each other. This may occur in applications like beam- input levels), it is possible for the weld metal to exhibit higher
flange to column-flange welds that are short in length. strengths with reduced ductility. Using lower-strength weld
Reducing the hydrogen content in the deposited weld met- metal is an effective solution, but caution should be taken to
al is a key method to overcome this type of cracking. This ensure that the required joint strength is attained.
can be accomplished by selecting the proper filler metals and For fillet welds and PJP groove welds, undermatching is
controlling exposure of the filler metals, interpass tempera- possible and is a frequently applied solution for situations
ture, and time between weld passes. The weld pass size is where transverse cracking is expected. The common appli-
another factor. Normally, high heat-input levels are good, in cation of this concept is the web-to-flange weld of ASTM
that they encourage slow cooling. In this case, those benefits A514 steel. The 70 ksi filler metals are routinely used to
are offset by the greater distance through which the hydro- make this connection. When a higher-strength filler material
gen must diffuse. Thus, within limits, thinner weld beads that matches the base material is used, the tendencies toward
are helpful. However, extremely small weld beads are as- cracking, including transverse cracking, increase.
sociated with low heat-input welding and, correspondingly, Increased preheat may alleviate transverse cracking by as-
faster cooling rates, which may offset the benefits of smaller sisting in diffusing the excessive hydrogen. When preheat is
hydrogen diffusion paths. Extra preheat and higher interpass applied, it may expand the length of the weld joint, allowing
temperatures are also helpful. the weld metal and the joint to contract simultaneously and
reducing the strains applied to the shrinking weld. This is
Susceptible Material particularly important when making circumferential welds.
When the circumference of the materials being welded is ex-
In the case of transverse cracking, the susceptible material is
panded, the weld metal is free to contract along with the sur-
the weld metal itself. For most steel applications, the harden-
rounding base material, reducing the longitudinal shrinkage
ability of the base metal will be greater than the weld metal,
stress. Finally, post-weld hydrogen release treatments that
as the base metal typically has higher carbon contents. Thus,
involve holding the steel at 400 to 450 °F for extended times
the more common form of hydrogen cracking is HAZ crack-
will assist in diffusing residual hydrogen.
ing, as discussed in Section 5.3.2. This includes most steels
with actual yield strengths (versus minimum specified yield
5.4 LAMELLAR TEARING
strengths) of less than about 85 ksi (Bailey, 1994). When
steels require low or no preheat, often driven by low CE
Lamellar tearing is a welding-related type of cracking that
computations, the weld metal may be the more susceptible
can occur in the base metal. It is caused by welding shrink-
material. Typically, when transverse cracking occurs, the ac-
age strains acting perpendicular to planes of weakness in
tual deposited weld metal will overmatch the base metal.
the steel and the tearing occurs slightly outside the HAZ.
These areas of weakness are the result of inclusions in the
Residual Stress
base metal that have been flattened into very thin disconti-
The residual stress is due to the longitudinal shrinkage of nuities that are roughly parallel to the surface of the steel.
the weld. A long weld (greater than about a foot) is gen- When strained perpendicular to the direction of rolling,

DESIGN GUIDE 21 / WELDED CONNECTIONS—A PRIMER FOR ENGINEERS/ 75

AWS D1.1/D1.1M:2015 CLAUSE 5. FABRICATION

[485 MPa], tack welds outside the final weld and construc- 5.18.2 Correction. Corrections of errors in camber of
tion aid welds shall require the approval of the Engineer. quenched and tempered steel shall require approval by
the Engineer.
5.17.3 Removal. At locations other than 5.17.2, tack
welds and construction aid welds not incorporated into
final welds shall be removed when required by the
Engineer.
5.19 Splices
5.19.1 Subassembly Splices. All welded subassembly
5.17.4 Additional Tack Weld Requirements splices in each component part of a cover-plated beam or
(1) Tack welds incorporated into final welds shall built-up member shall be made before the component
be made with electrodes meeting the requirements of part is welded to other component parts of the member.
the final welds. These welds shall be cleaned prior to 5.19.1.1 Shop Splice Location. Shop splices of webs
incorporation. and flanges in built-up girders may be located in a single
(2) Multipass tack welds shall have cascaded ends or transverse plane or multiple transverse planes.
be otherwise prepared for incorporation into the final 5.19.1.2 Cyclically Loaded Splices. For cyclically
weld. loaded members, the fatigue stress provisions of the gen-
(3) Tack welds incorporated into final welds that are eral specifications shall apply.
qualified with notch toughness or are required to be 5.19.2 Member Splices. Long girders or girder sections
made with filler metal classified with notch toughness may be made by welding subassemblies. Splices be-
shall be made with compatible filler metals. tween sections of rolled beams or built-up girders shall
5.17.5 Additional Requirements for Tack Welds In- be made in a single transverse plane, when practicable.
corporated into SAW Welds. The following shall apply
in addition to 5.17.4 requirements.
5.20 Control of Distortion and
(1) Preheat is not required for single pass tack welds
remelted by continuous SAW welds. This is an exception
Shrinkage
to the qualification requirements of 5.17.1. 5.20.1 Procedure and Sequence. In assembling and
joining parts of a structure or of built-up members and in
(2) Fillet tack welds shall not exceed 3/8 in [10 mm]
welding reinforcing parts to members, the procedure and
in size and shall not produce objectionable changes in the
sequence shall be such as will minimize distortion and
appearance of the weld surface.
shrinkage.
(3) Tack welds in the roots of joints requiring spe-
5.20.2 Sequencing. Insofar as practicable, all welds
cific root penetration shall not result in decreased
shall be made in a sequence that will balance the applied
penetration.
heat of welding while the welding progresses.
(4) Tack welds not conforming to the requirements of
5.20.3 Contractor Responsibility. On members or
(2) and (3) shall be removed or reduced in size by any
structures where excessive shrinkage or distortion could
suitable means before welding.
be expected, the Contractor shall prepare a written weld-
(5) Tack welds in the root of a joint with steel back- ing sequence for that member or structure which meets
ing less than 5/16 in [8 mm] thick shall be removed the quality requirements specified. The welding se-
or made continuous for the full length of the joint quence and distortion control program shall be submitted
using SMAW with low-hydrogen electrodes, GMAW, to the Engineer, for information and comment, before the
or FCAW-G. start of welding on the member or structure in which
shrinkage or distortion is likely to affect the adequacy of
the member or structure.
5.18 Camber in Built-Up Members 5.20.4 Weld Progression. The direction of the general
progression in welding on a member shall be from points
5.18.1 Camber. Edges of built-up beam and girder
where the parts are relatively fixed in position with re-
webs shall be cut to the prescribed camber with suitable
spect to each other toward points having a greater rela-
allowance for shrinkage due to cutting and welding.
tive freedom of movement.
However, moderate variation from the specified camber
tolerance may be corrected by a careful application of 5.20.5 Minimized Restraint. In assemblies, joints ex-
heat. pected to have significant shrinkage should usually be

173