TECHNICAL DIGEST 8

WELDING OF OPEN-WEB
STEEL JOISTS AND
JOIST GIRDERS

OCTOBER 2008
The information presented in this publication has been developed by an
Ad-Hoc Group of the Steel Joist Institute’s Engineering Practice
Committee. The group was comprised of the following members: John
Grayson, PE, Patrice Harvey, Tim Holtermann, PE, Jason McCool, PE,
Carl Pugh, Jr., PE, and Michael Winarta, PE and Perry S. Green, PhD,
Technical Director and is produced in accordance with recognized
engineering principles and is for general information only. The SJI and its
committees have made a concerted effort to present accurate, reliable,
and useful information on the welding of steel joists and Joist Girders.
The publication of the material contained in this Technical Digest is not
intended as a representation or warranty on the part of the Steel Joist
Institute. Any person making use of this information does so at one’s own
risk and assumes all liability arising from such use.

Copyright © 2008

by

Steel Joist Institute

All rights reserved. This Technical Digest or any part thereof must not be
reproduced in any form without the written permission of the Steel Joist
Institute.

Printed in the United States of America

Second Edition
First Printing – October 2008
TECHNICAL DIGEST 8

WELDING OF OPEN-WEB STEEL

JOISTS AND JOIST GIRDERS

John Grayson, Engineering Manager, Nucor-Vulcraft Group – Indiana Division,

St. Joe, IN
Patrice Harvey, Engineering Manager, Canam Steel Corp.,
Jacksonville, FL
Tim Holtermann, Corporate Engineering Manager, Canam Steel Corp.,
Washington, MO
Jason McCool, Welding Engineer, CMC Joist & Deck, Co.,
Fallon, NV
Carl Pugh, Jr., Engineering Manager, New Millennium Building Systems,
Salem, VA
Michael Winarta, Engineering Manager, New Millennium Building Systems,
Butler, IN

Perry S. Green
Technical Director, Steel Joist Institute
Myrtle Beach, SC

Steel Joist Institute

Administrative Offices
1173B London Links Drive
Forest, VA 24551
Technology, Engineering, and Education Center
3127 Mr. Joe White Avenue
Myrtle Beach, SC 29577-6760

www.steeljoist.org
 ACKNOWLEDGEMENT
First Edition
Technical Digest No. 8 was prepared under the auspices of the Institute’s
Research Committee. The Digest’s primary authors are Robert E. Somers and
Theodore V. Galambos. Over 6 years of concerted effort by the authors and the
members of the Research Committee are represented in this Digest, and this
dedicated effort is hereby acknowledged.

Second Edition
The authors would like to thank the members of the Engineering Practice
Committee and the Research Committee of the Steel Joist Institute for their
review and contributions to the writing of the second edition of this digest.

FOREWORD
This second edition is an extensive update of the previously issued Technical
Digest No. 8 on welding of Open-Web Steel Joists and Joist Girders. The
Technical Digest is another addition to the series of Steel Joist Institute
publications designed to give the reader information regarding the application
and usage of steel joists and Joist Girders.

Technical Digest No. 8 provides the reader information on: (1) the various types
of welding processes currently and previously used by SJI member companies in
the manufacturing of Open-Web Steel Joists and Joist Girders; (2) the pertinent
research and testing procedures utilized in gathering data for this Digest; (3) the
results of the weld testing conducted by the SJI; (4) the analysis of the data; and
(5) the conservative conclusions drawn from such analysis.

This and other SJI Technical Digests serve to highlight specific areas of design
and/or application for the benefit of architects, building inspectors, building
officials, designers, engineers, erectors, students and others.

i
 TABLE OF CONTENTS
ACKNOWLEDGEMENT ………………………………………………………… i

FOREWORD …………………………………………………………………...... i

TABLE OF CONTENTS …………………………………………………………. ii

BACKGROUND AND PURPOSE………………………………………………… iii

GLOSSARY ………………………………………………………………………. v

Chapter 1 INTRODUCTION …………………………………………....….. 1

Chapter 2 INDUSTRY PRACTICES ……………….……………..…......... 3

Chapter 3 BASE METAL AND WELD METAL MATERIALS …..……….. 11

Chapter 4 SJI STANDARD WELDING PROCEDURES ….……...…....... 14

Chapter 5 SJI OPERATOR QUALIFICATION ……………………………. 20

Chapter 6 SEISMIC WELDING REQUIREMENTS ….…………………... 22

Chapter 7 WELD DESIGNS …………………………………………….….. 26

Chapter 8 PLANT INSPECTION CRITERIA ………………………….….. 43

Chapter 9 FIELD WELDING AND FIELD INSPECTION ...……………… 50

REFERENCES …………………………………………………………………… 56

SJI STANDARD SPECIFICATIONS - WELDING

Appendix A 59
REQUIREMENTS ...……………………….……..……………...

SJI WELDING REQUIREMENTS: H-, LH-, DLH-SERIES

Appendix B 70
JOISTS AND JOIST GIRDERS – HISTORICAL DATA ……..

ii
 BACKGROUND AND PURPOSE
Open Web Steel Joists have been manufactured in the United States since 1923.
The original joist was a Warren truss, with a single continuous round bar bent
web, plus two straight round bars outside the web for both the top and bottom
chords. The joists were welded by the manual shielded metal arc method (stick
welding).

Various joist designs with different load-carrying capacities were provided by

individual manufacturers, which made it difficult for architects, engineers, and
contractors to compare the load capacity and economics of various steel joist
configurations. To simplify this problem, producers in the industry organized the
Steel Joist Institute (SJI) in 1928, and adopted the first standard specifications for
Open Web Steel Joists. In 1929 the first standard load table for joists was
promulgated.

In the years since the first Open Web Steel Joist was arc welded, many
fabrication and manufacturing innovations have been introduced. The use of
resistance welding permitted the making of joints without the addition of filler
metal, the weld area being controlled by projections on the parts being joined.

In the 1950’s semi-automatic gas-shielded welding of steel was introduced. This

method used a continuous wire electrode fed automatically through a hand-held
gun, which also introduced a shielding gas to the arc. The process of feeding the
electrode through about 15 feet of flexible electric cable required that the
electrode wire be small—about 1/16 inch diameter. This small electrode
diameter and the freedom from the slag in shielded metal arc (stick) welding
provided improved access and weld penetration for the flare bevel, flare vee, and
fillet welds encountered in open web joists.

Based on the number of individual welds made per year, the steel joist industry is
truly a mass production operation. It is estimated that, in 2007, SJI member
companies in Canada, Mexico and the United States made approximately 900
million welds on steel joists and Joist Girders! The quality control for this
incredible number of welds has to be on a sampling basis.

The design standard for Open Web Joists is based on an ultimate load factor, or
factor of safety, of 1.65 times the design load, verified by laboratory qualification
tests of a particular design. As with any structural steel member designed
primarily for vertical in-plane loads, open web joists require lateral support. The
welded joints are required to develop twice the calculated design load in weld
shear. Production welds may be: (1) inspected visually; (2) proof tested by non-
destructive pry loading; (3) proof tested by loading of a complete joist; (4) or
tested to destruction.

iii
The steel joist industry has literally billions of welds in service, in tens of
thousands of buildings, an enviable record for economical lightweight structures.
All of these joists were manufactured to a performance specification.

iv
 GLOSSARY

NOTES:

Terms in Bold and their definitions come from the AISC AND AISI STANDARD,
“Standard Definitions for Use in the Design of Steel Structures,” 2004 Edition,
First Printing - April 2005.

* These terms are usually qualified by the type of load effect, e.g., nominal tensile
strength, available compressive strength, design flexural strength.

ASD (Allowable Strength Design). Method of proportioning structural

components such that the allowable strength equals or exceeds the required
strength of the component under the action of the ASD load combinations.

Allowable Strength*. Nominal strength divided by the safety factor, Rn/Ω.

Available Strength*. Design strength or allowable strength as appropriate.

Bearing. The distance that the bearing shoe or seat of a joist or Joist Girder
extends over its masonry, concrete or steel support.

Bridging. In general, a member connected to a joist to brace it from lateral

movement.

Chords. The top and bottom members of a joist or Joist Girder. When a chord is
comprised of two angles there is usually a gap between the members.

Connection. Combination of structural elements and joints used to transmit

forces between two or more members. See also splice.

Deck. A floor or roof covering made out of galvanized, painted or unpainted cold-
formed steel attached by welding or mechanical means to joists, beams, purlins,
or other structural members.

Design Load. Applied load determined in accordance with either LRFD load
combinations or ASD load combinations, whichever is applicable.

Design Strength*. Resistance factor multiplied by the nominal strength, φRn.

v
End Diagonal or Web. The first web member on either end of a joist or Joist
Girder which begins at the top chord at the seat and ends at the first bottom
chord panel point.

End Welds. Welds at the ends of an existing member or the reinforcing member.

Expected Yield Strength. Yield strength in tension of a member, equal to the

expected yield stress multiplied by Ag.

Expected Tensile Strength. Tensile strength of a member, equal to the specified

minimum tensile strength, Fu, multiplied by Rt.

Expected Yield Stress. Yield stress of the material, equal to the specified
minimum yield stress, Fy, multiplied by Ry.

Filler. A short piece of round bar, plate or angle welded between the two angles
of a chord member or the two angles of a web member to tie them together,
usually located at the midpoint of the member between joints.

Joint. Area where two or more ends, surfaces or edges are attached.
Categorized by type of fastener or weld used and the method of force transfer.

Joist. A structural load-carrying member with an open web system which

supports floors and roofs utilizing hot-rolled or cold-formed steel and is designed
as a simple span member. Currently, the SJI has the following joist designations:
K-Series including KCS, LH-Series and DLH-Series, and CJ-Series.

Joist Girder. A primary structural load-carrying member with an open web

system designed as a simple span usually supporting equally spaced
concentrated loads of a floor or roof system acting at the panel points of the
girder and utilizing hot-rolled or cold-formed steel.

Load. Force or other action that results from the weight of building materials,
occupants and their possessions, environmental effects, differential movement,
or restrained dimensional changes.

LRFD (Load and Resistance Factor Design). Method of proportioning

structural components such that the design strength equals or exceeds the
required strength of the component under the action of the LRFD load
combinations.

Nominal Strength*. Strength of a structure or component (without the

resistance factor or safety factor applied) to resist the load effects, as determined
in accordance with the Standard Specifications.

vi
Required Strength*. Forces, stresses, and deformations produced in a
structural component, determined by either structural analysis, for the LRFD or
ASD load combinations, as appropriate, or as specified by the Standard
Specifications.

Resistance Factor, φ. Factor that accounts for deviations of the actual strength
from the nominal strength, deviations of the actual load from the nominal load,
uncertainties in the analysis that transforms the load into a load effect and for the
manner and consequences of failure.

Safety Factor, Ω. Factor that accounts for deviations of the actual strength from
the nominal strength, deviations of the actual load from the nominal load,
uncertainties in the analysis that transforms the load into a load effect and for the
manner and consequences of failure.

Serviceability Limit State. Limiting condition affecting the ability of a structure to

preserve its appearance, maintainability, durability, or the comfort of its
occupants or function of machinery, under normal usage.

Specified Minimum Yield Stress. Lower limit of yield stress specified for a
material as defined by ASTM.

Specifying Professional. The licensed professional who is responsible for sealing

the building Contract Documents, which indicates that he or she has performed
or supervised the analysis, design and document preparation for the structure
and has knowledge of the load-carrying structural system.

Splice. Connection between two structural members joined at their ends by

either bolting or welding to form a single, longer member.

Stability. Condition reached in the loading of a structural component, frame or

structure in which a slight disturbance in the loads or geometry does not produce
large displacements.

Standard Specifications. Documents developed and maintained by the Steel

Joist Institute for the design and manufacture of open web steel joists and Joist
Girders. The term “SJI Standard Specifications” encompasses by reference the
following:
ANSI/SJI-K1.1 Standard Specifications for Open Web Steel Joists, K-Series;
ANSI/SJI-LH/DLH-1.1 Standard Specifications for Longspan Steel Joists,
LH-Series and Deep Longspan Steel Joists, DLH-Series;
ANSI/SJI-JG-1.1 Standard Specifications for Joist Girders; and
ANSI/SJI-CJ1.0 Standard Specifications for Composite Steel Joists,
CJ-Series.

vii
Webs. The vertical or diagonal members joined at the top and bottom chords of
a joist or Joist Girder to form triangular patterns.

Yield Point. First stress in a material at which an increase in strain occurs

without an increase in stress as defined by ASTM.

Yield Strength. Stress at which a material exhibits a specified limiting deviation

from the proportionality of stress to strain as defined by ASTM.

Yield Stress. Generic term to denote either yield point or yield strength, as
appropriate for the material.

viii
 CHAPTER 1
INTRODUCTION
Open Web Steel Joists and Joist Girders are economical, versatile, and efficient
products used to support floors and roofs in such diverse types of structures as
commercial, educational, industrial, office, and residential buildings. They are
available in a wide variety of depths, spans, configurations, profiles, and load-
carrying capacities. Joists and Joist Girders are designed using a wide variety of
customized software that aids the joist design engineer in optimizing the material
size of each component while checking against a large array of design criteria.
Yet it is the welds that hold the joist together, and without adequate welds all the
care in selecting optimum and adequate materials is pointless. Steel joists have
a proven track record with billions of in-service welds.

There are several decisive factors that must be considered with regard to welded
connections. Chapter 2, Industry Practices, defines the processes typically
used by SJI member companies in the manufacture of joists and Joist Girders.
Next, the materials to be used must be determined, both for the steel members
being joined and the welding consumables that are used. This is discussed in
Chapter 3, Base Metal and Weld Metal Materials. Chapter 4, SJI Standard
Welding Procedures, describes how the type of weld is determined, and for
each type of weld, a documented welding procedure needs to be established.
The person performing the welding must demonstrate their ability to follow the
written weld procedure for that type of weld as described in Chapter 5, SJI
Operator Qualification. The design strength of the type of weld specified must
be calculated, and this is discussed in Chapter 7, Weld Designs. The completed
weld is then subject to inspection whereby certain acceptance criteria must be
met as described in Chapter 8, Plant Inspection Criteria.

While the primary focus of this document is the welding done in the joist
manufacturer’s plant, proper field welding during the installation of steel joists is
critical to their overall performance, and this topic is covered in Chapter 9, Field
Welding and Field Inspection. There are additional welding considerations,
both in the plant and the field, when joists or Joist Girders are used to resist
seismic loads and those are discussed in Chapter 6, Seismic Welding
Requirements.

Frequent reference will be made throughout this digest to the American Welding
Society (AWS) Welding Specifications. Compliance with AWS criteria will satisfy
the SJI welding requirements, and in many cases the SJI requirement is in
compliance with a particular AWS provision. However, there are some
differences. The reasons for the SJI to maintain their own specification language
for welding will be discussed throughout this document.

1
The performance of steel joists and Joist Girders welded in accordance with the
SJI Standard Specifications (SJI 2005) has been verified through extensive
testing both by the Steel Joist Institute as well as by many of its member
companies. The testing requirements relative to current SJI membership are
described in Chapter 2.

Appendix A, SJI Standard Specifications – Welding Requirements provides a

quick reference to the applicable sections of the K-Series, LH/DLH-Series, CJ-
Series and Joist Girders Standard Specifications that pertain to welding.

Appendix B, SJI Welding Requirements: H-, LH-, DLH-Series Joists and

Joist Girders – Historical Data reviews the results obtained from tests
conducted over the years by the SJI that were used to develop past and present
welding specifications. Most recently in 2007, a series of full-scale joist tests
were conducted at Villanova University. Part of the study was to assess the
performance of joists containing crimped angle web members. These joists were
manufactured by three different member companies using their respective
welding procedures to produce the joints between the ends of the crimped angle
web members and the joist top and bottom chord angles. A more detailed
description of this research project and its conclusions with respect to welding of
the crimped angle web members can be found in Chapter 7.

2
 CHAPTER 2
INDUSTRY PRACTICES
The purpose of this chapter is to familiarize the reader with the current joist
industry practices related to welding. It should be noted however, that the
industry practices, as well as welding types and procedures of this chapter, are
not meant to be all inclusive, or to restrict a manufacturer to only the types and
procedures listed here.

This chapter will also outline the differences between the Steel Joist Institute
(SJI) welding criteria and American Welding Society (AWS) welding criteria.
While there are considerable similarities between the two, the differences and the
reasons for these differences, will be described.

Joist manufacturing plants are typically set up in more of an assembly line

process than a piece work or fabrication shop process. Within a joist
manufacturing plant, the welding functions can be categorized by three main
purposes: Chord splices, temporary welds, finish/joint welds.

Chord splices
The current industry practice is to use angle sections for joist top and bottom
chords. Prior to assembly, the joist chords are prepared by cutting and/or butt
splicing angles to obtain the required chord lengths. For efficient use of material,
the chord angles are essentially butt spliced to form a continuous piece, and then
sheared to create the desired chord length. As a result, chord splices will occur
even in chord lengths that are shorter than the standard inventory length. Chord
splices are also referred to as butt welds within the industry. The range of angles
that are butt welded in joist shops extends from 1 x 1 x 0.109 angles to 6 x 6 x
3/4, and for some manufacturers 8 x 8 x 1 1/8 angles. Obviously, the welding
procedures may vary significantly depending on the size of the angle being butt
welded. The SJI Standard Specifications (SJI 2005) only govern the required
strength of the butt welds, as discussed in Chapter 7, and provide acceptance
criteria, as described in Chapter 8. It should be noted that there are no SJI
criteria for the location(s) or quantity of butt welds allowed to be placed to create
a single member.

Temporary welds
Industry practice is to assemble the joists in a fixture or table, commonly referred
to as rigging or jigging, and then to move the joists out as quickly as possible to
allow the next piece to progress. The welded joints are not completed at the
rigging table, but instead are completed at a separate finish welding station or
portion of the assembly line. The assembly done at the rigging table need only

3
be strong enough to allow the joist to be moved to the finish weld area. When
assembling the joists some manufacturers employ clamps, which are then
removed after finish welding. However, the clamps have a limited capacity, in
both size of pieces being assembled, and in strength. Alternately, the joists can
be assembled using a minimal weld, typically referred to as a temporary or tack
weld, to hold the pieces together. The profile of this weld may or may not equal
the thickness of the material being joined, and may be only 3/8 or 1/2 inch long.
There are no strength or serviceability requirements imposed by SJI on tack
welds. The acceptance criteria for tack welds are discussed in Chapter 8.

Finish/joint welds
Finish welds are the web to chord joint welds, placed in the finish weld area, to
the thickness and length required by the shop details, or manufacturer’s
standards. The type of welding being used can vary depending on the type of
product. Most joist manufacturers have several production lines, each of which
cater to a specific dimensional range of product, and also to the type of web
construction. There are three types of joist web members that are typically used
by most joist manufacturers: Round bar, Single member (used inside the chord
gap), and Double members (used on the outside of each chord angle vertical
leg).

Round Bars
Round bars can be bent in a continuous pattern of “V” or “W” shapes for
assembly into joists. Typically, because the continuous nature of this type of web
requires the same size for multiple panels, i.e. for compression and tension
members, the design is less efficient with regard to the weight of the webs.
Hence, a continuous round bar web is typically used only for smaller joists – the
lower half of the range of K-Series joists, and small LH-Series joists. The weld
between a round bar and an angle chord is a flare bevel groove weld. The
strength of this type of weld is determined as described in Chapter 7. It is
important to note that the required weld strength is significantly different for a
round bar that is continuous through the joint, and one which is discontinuous,
such as where two “W” web sections adjoin. In the case where there is
continuity, the weld design is only for the horizontal components of the web
forces, plus a small shear component in the panel. Where the round bar web is
discontinuous, the weld must be designed for the full axial force in the web.
Round bars are also commonly used for the first diagonal, or end bar, in all K-
Series joists, and the lower half of the range of LH-Series joists. It should be
noted that a weld must be made to each chord angle, but is typically made on
only one side of the round, to each chord angle. As proven by full-scale tests,
the placement of the flare bevel groove weld on only one side of the round bar
does not create a structurally significant eccentricity.

4
Single member, inside web
Where a continuous round web is not utilized, and the strength requirements and
panel size will allow, the industry preference is to use individual, single web
members placed in the gap between the chord angles. These single web
members may consist of angles crimped at the ends; angles crimped or formed
along the full length to create a “U” or “C” shape; a single, rotated un-crimped
angle, or a cold-formed “U” or “C” channel section. The weld from single web
members to the chord angles is typically a fillet weld, equal to the thickness of
the web member. Certain shapes, such as cold-formed “C” channels could
actually utilize a flare groove weld along the heel of the web. However, the
placement and welding position for these web members is typically such that at
one chord the weld is being placed along the web member end and “toe”, while
at the opposite chord, the weld is placed along the web member end and “heel”.
For simplicity, the same weld requirement is normally stipulated for both ends of
the web member, and is based on a fillet weld for both ends, rather than a flare
bevel weld. See Chapter 7 for discussion of the design of single web welds, and
Chapter 8 for acceptance criteria. Single web members are typically found in
larger K-Series joists, many LH-/DLH-Series joists, and some Joist Girders.

The nature of the positioning of a single web member, placed in the chord gap, is
such that it is not typically possible to place the weld concentrically at the end of
the member. The weld profile is typically an “L” shape, along one side of the web
member, and across the end, to the extent possible. The SJI Standard
Specifications (SJI 2005) allow this un-centered weld placement. The slightly
eccentric placement of the weld on these types of joints has been shown not to
be detrimental to the connection capacity of the joints, as demonstrated by the
full-scale load testing required to verify a member company’s designs, as
described at the end of this chapter.

Welds similar to those used at web to chord joints are also used for fillers. Fillers
are small pieces of web material used to join the two chord angles, in the gap,
where required, usually at the mid-point between panel points.

Double member, outside webs

Where the strength or stiffness of a single web is not adequate, industry practice
is to use double angles, where one angle is placed on the outside of each chord
angle. The welding is similar to that used for single angle webs. The member
will typically be welded to each chord angle, at both toe and heel, and across the
end. The SJI Standard Specifications (SJI 2005) do not require that the weld
lengths be balanced between toe and heel. Once again, see Chapter 7 for weld
design and Chapter 8 for acceptance criteria. The thickness of double angle web
members can vary from as little as 0.109 inches to as much as 3/4 inches. SJI
does not regulate the thickness of a fillet weld size relative to the thickness of the
member. Because of the controlled factory conditions, a fillet weld equal to the
thickness of the web member may be used for thickness of 1/4 inch or even
greater. For very thick web members, most joist manufacturers will opt for a

5
longer, thinner weld profile, rather than a using a multi-pass weld that needs to
be built up to the member thickness less 1/16 inch. To increase the compressive
buckling resistance, double angle web members are often tied at the center with
a tie clip. The welding for the tie clip is similar to the web to chord weld.

SJI Member Company Welding Survey

In 2006, a survey of each active SJI member plant was conducted to determine
the welding processes currently in use. The survey confirmed that there are four
main welding types in use and each of these will be described in some detail in
the following sections.

A. Shielded Metal Arc Welding (SMAW)

Shielded Metal Arc Welding, also commonly known as “Stick” welding has
shielding provided by the welding electrode coating. The electrodes for
SMAW consist of a solid-core wire typically ranging in size from 1/8 in. to
1/4 in. in diameter, surrounded by material called “flux” which functions as
the shield for the weld pool from the atmosphere.

B. Flux Cored Arc Welding (FCAW)

Flux Cored Arc Welding has shielding provided by a flux-cored tubular
electrode. The flux material is located inside the tubular metal core of the
electrodes, and functions to shield the weld pool during the welding process.
The electrodes range from 0.030 in. to 1/8 in. in diameter and are spooled
into different packaging needs.

There are two different categories of flux-cored arc welding: self-shielded

flux core welding (FCAW-S), and gas-shielded flux core welding (FCAW-G).
Self-shielded flux core welding depends entirely on the flux from the core of
the electrodes to shield the weld pool. Gas-shielded flux core requires an
external supply of shielding gas for its shielding purpose.

C. Gas Metal Arc Welding (GMAW)

Gas Metal Arc Welding, also commonly known as “MIG” (Metal Inert Gas)
welding has shielding provided by shielding gas that surrounds the weld
pool. The process is much similar to the FCAW-G, where GMAW can be
done using solid electrodes or metal-cored electrodes that contain no flux
material, which resulted in “slag free” weld pool. The solid electrodes
typically range from 0.035 in. to 0.052 in., while the metal-cored electrodes
range from 0.045 in. to 1/16 in. in diameter.

D. Resistance Welding
Resistance Welding is a welding process where work pieces are welded
together from a combination of pressure and localized heat generated by
passing electrical current through the work pieces. The contact surface of

6
 the work pieces creates a high resistance to the electric current, thus
generating enough thermal energy to melt the metal at the point of contact.
The solidified molten metal under a given pressure will join the work pieces.

There are numerous Resistance Welding processes being used in many

different industries. Some of the most common processes found in the steel
joist industry include:
Resistance Spot Welding (RSW)
This is a Resistance Welding process that produces weld at a certain spot
of the contact surface of the work pieces. The size and shape of the formed
welds are largely dependent on the shape and contour of the electrodes
used.

Flash Welding (FW)

This is a Resistance Welding process that can produce a weld
simultaneously over the entire area of the abutting surfaces. The process
will produce a flashing arc and heat over the abutting surfaces. After a
predetermined time, the arc will bring the two abutting surfaces into a
melting point, then the two pieces are pressed together with welds
connecting the abutting surfaces. The process is commonly used in butt
splicing of chord or web materials used in the steel joist.

Upset Welding (UW)

This is a Resistance Welding process that is similar to the Flash Welding
process. This process is started by pressing the two abutting surfaces of
the work pieces together. A high amperage current is then passed through
the pieces, creating enough heat to bring the surfaces to a forging
temperature. An upsetting force is then applied and the current stopped,
allowing the pieces to be welded together.

A Performance-Based Welding Specification

The SJI approach to welding is a performance based specification. The AWS
acceptance criteria are literal and empirical, and a given weld either passes or
fails, with no middle ground. The essence of the SJI Performance-Based
Specification with regard to welding is that a third possibility exists – a weld, or
portion of a weld, that neither contributes to the overall structural performance,
nor takes away from the overall structural performance. The SJI Specification is
based on welding in a controlled environment, in a repeatable and predictable
manner, validated by full-scale physical testing. Unlike AWS acceptance criteria,
a weld or portion of a weld that does not meet all of the SJI acceptance criteria
does not necessarily require removal or repair. Instead two questions are raised
– is there an adequate length of defect free weld to meet the required strength
(length), and is the defect detrimental to the structural integrity of the member
being joined? As an example, excessive porosity or an apparent lack of fusion
are cause to reject and discount the length of weld affected by these defects, but
these types of defects effect only the integrity of the weld and not the connected

7
materials, and hence, may not need correction if an adequate length of defect
free weld is observed. This is not to suggest that these types of defects are not a
cause for concern, with regard to the ability of the welding operator. Conversely,
a defect such as excessive undercut is cause for rejection and repair, as it has
affected the strength and capacity of the undercut member. The application of
the acceptance criteria in a performance based manner is the key difference
between the SJI and AWS welding criteria.

This is particularly significant for the tack welds applied at the rigging table during
joist assembly. Often, these tack welds are made on the back edges or surfaces
of the web members, where they will not be incorporated in the final, finish weld.
As a result, the position and angle are often poor when the tack weld is applied,
and in the assembly line type of setting, the tack welds are made very quickly.
As described above, the tack weld has no structural purpose beyond temporarily
holding the pieces together from rigging to finish welding. Since an independent
tack weld is not counted as contributing to the overall required length of weld, it
need not be free of defects in the weld itself. The only concern is that the tack
weld has not undercut the base metal. AWS, however, applies the same
acceptance criteria to all welds, including tack welds.

Another difference between SJI and AWS welding is that the same SJI criteria
apply to all the welds. Unfortunately, the thickness of welds being made on
joists, from approximately 7/64 inch to 3/4 inch fillets, straddles two different
American Welding Society provisions – D1.1 (AWS 2008) and D1.3 (AWS 2008).
Table 2-1 gives a brief summary of the differences between AWS D1.1 and D1.3.

Table 2-1 Use of AWS D1.1 or D1.3 Based on Material Thickness being Joined
AWS D1.1 applies to structural steel • 1/8” AWS D1.3 applies to sheet steel up to 3/16”
thick. thick.
Comment: When base materials < 1/8” are used, D1.3 should be specified in addition to D1.1
(D1.1,1.2). Note use of the word “should”. Use of D1.3 is recommended, but not required (D1.1,
1.3.6.2).

As discussed in Chapter 7, the SJI has developed criteria for the strength of flare
bevel groove welds which differ from AWS. On the basis of the controlled factory
environment, the small weld wire diameters, and the testing, SJI has in the past
allowed a more liberal weld capacity than AWS. However, the 2005 AWS
Specification and subsequently the AISC Specification for Structural Steel
Buildings (AISC 2005) substantially increased the allowable weld capacity of the
flare bevel groove welds. Without new testing, and on the basis of the track
record of millions of in service welds, SJI has elected not to pursue what is now
for most round bar diameters the more liberal AWS weld strength provision for
flare bevel groove welds.

8
While the last few paragraphs have briefly outlined differences between SJI and
AWS, there are many similarities and a lot of overlap. As will be discussed in
Chapter 3, the welding consumables for SJI products should meet an appropriate
AWS standard. For joint types, SJI recognizes the AWS pre-qualified joints, and
requires a qualification process for welding procedures that are not AWS pre-
qualified. And like AWS, SJI requires Weld Procedure Specifications (WPS’s)
with the same basic information required as AWS.

For the qualification of individual welding operators, SJI requires a program of

testing and certification, and recognizes AWS certification as one means to
accomplish this, as described in Chapter 5. In fact, many joist manufacturers use
the AWS certification as their sole method of operator qualification.

In all cases, it should be understood that a consumable, procedure, operator, or

acceptance criteria that meets the AWS Specification will also meet the SJI
Specification.

SJI Required Inspection and Testing

While the responsibility for the integrity of any individual weld on a project lies
solely with the joist manufacturer and not the Steel Joist Institute, SJI has a
strong history of testing. Selected test results are presented in Appendix B of
this digest. In addition, all SJI member companies have undergone and continue
to undergo inspection and testing at three levels: Full-scale load tests for initial
membership, joint testing with initial plant inspection, and biennial plant
inspections.

Full-Scale Load Tests for Verification of Design for Initial Membership or New
Joist Designs
As a requirement for membership and design approval of K-Series joists, the
manufacturer must conduct independently audited full scale load tests. These
tests are to failure, rather than a simple proof test, and after testing, tension
coupons are cut to determine the actual member yield strengths. The joists are
required to withstand a design load of 1.65 times the predicted design load of an
ASD design for the actual member sizes and strengths used in the manufacture
of the joist tested. These tests are required for each type of construction the joist
manufacturer employs within the range of K-Series joists. For example, a new
manufacturer that intends to build both rod web and crimped angle web K-Series
joists is required to undergo the full-scale tests for both types of construction.
After testing as part of initial membership, a manufacturer is also required to
undergo the same full-scale load tests if a new design type is submitted for SJI
approval, or in the event of a significant design change that warrants testing
(such as the introduction of a new joist series).

9
Joint Testing with Initial Plant Inspection
As part of the certification process for a new joist manufacturing facility, a
rigorous initial plant inspection is conducted by an independent agency hired by
the SJI. As part of the inspection, joists are randomly selected and partial span
samples are prepared for shipment to an independent lab. The key members are
reinforced, but not the joints, and the idea of the test is to be able to proof test the
adequacy of the joints to a safety factor of at least 2.0 without failure of the joints.

Biennial Plant Inspections

Each member company’s manufacturing facilities are subject to an independent
biennial plant inspection. Welding processes and procedures are key
components of these inspections.

10
 CHAPTER 3
BASE METAL AND WELD METAL MATERIALS
Base Metal Requirements

The steel used in the manufacture of chord and web sections of open web steel
joists and Joist Girders shall conform to one of the following ASTM Specifications
of latest adoption:
• Carbon Structural Steel, ASTM A36/A36M.
• High-Strength, Low-Alloy Structural Steel, ASTM A242/A242M.
• High-Strength Carbon-Manganese Steel of Structural Quality ASTM
A529/A529M, Grade 50.
• High-Strength Low-Alloy Columbium-Vanadium Structural Steel, ASTM
A572/A572M Grade 42 or 50.
• High-Strength Low-Alloy Structural Steel with 50 ksi (345 MPa) Minimum
Yield Point to 4 inches (100 millimeters) thick, ASTM A588/A588M.
• Steel, Sheet and Strip, High-Strength, Low-Alloy, Hot-Rolled and Cold-
Rolled, with Improved Atmospheric Corrosion Resistance, ASTM A606.
• Steel, Sheet, Cold-Rolled, Carbon, Structural, High-Strength Low-Alloy
and High-Strength Low-Alloy with Improved Formability, ASTM
A1008/A1008M.
• Steel, Sheet and Strip, Hot-Rolled, Carbon, Structural, High-Strength Low-
Alloy and High-Strength Low-Alloy with Improved Formability, ASTM
A1011/A1011M.

Steels other than those listed in the Standard Specifications (SJI 2005) are
permissible provided that such material in the state used for final assembly and
manufacture are weldable and have been proven by test to exhibit a yield
strength equal to or exceeding the design yield strength and an elongation of not
less than (a) 20 percent in 2 inches (51 millimeters) for sheet and strip, or (b) 18
percent in 8 inches (203 millimeters) for plates, shapes and bars with
adjustments for thickness for plates, shapes and bars as prescribed in ASTM
A36/A36M, A242/A242M, A529/A529M, A572/A572M, or A588/A588M,
whichever specification is applicable on the basis of design yield strength. The
yield strength used for the basis for the design stress shall be at least 36 ksi (250
MPa), but not greater than 50 ksi (345 MPa).

11
Welding Consumables Requirements

Arc welding is the most common welding process found in a steel joist
manufacturing facility. The characteristic of arc welding is the heating of steel
pieces to be connected with an electric arc. During the process of melting and
solidifying of the metal, it is important to shield the weld pool from the normal
atmospheric environment. Our atmosphere consists of roughly 80 percent
Nitrogen and 19 percent Oxygen. Weld pool unshielded from Nitrogen and
Oxygen could cause porosity, inclusions, and embrittlement which can greatly
reduce the quality and mechanical properties of the weld and therefore all joist
arc welding should be shielded. The different shielding methods used during arc
welding further clarify the different processes of how an arc weld is classified:

A. Shielded Metal Arc Welding (SMAW)

The filler metal specifications for the SMAW process are governed by AWS
A5.1 for carbon steel electrodes, while AWS A5.5 governs the low-alloy
steel electrodes. Electrodes covered by AWS A5.1 will have either 60 or 70
ksi minimum tensile strength. These electrodes are most commonly used
within a steel joist manufacturing facility using SMAW process due to the
fact that the design of the steel involves materials with a nominal yield
strength up to 50 ksi. Electrodes covered by AWS A5.5 are used when
weathering steel is used (due to better atmospheric corrosion resistance of
the weld deposit), and when higher strength steels and welds are a design
requirement.

Electrodes governed by AWS A5.1 typically will be identified with E70XX.

The “E” stands for Electrode, the 70 indicates the minimum specified tensile
strength of 70 ksi, the first “X” indicates the position in which the electrodes
may be used, and the second “X” indicates the covering type, welding
current, and CVN properties.

In addition, electrodes governed by AWS A5.5 typically will be identified with

E70XX-X. The difference between this identification and that given in AWS
A5.1 is the inclusion of an additional suffix letter “X” that indicates the alloy
type of the electrodes.

B. Flux Cored Arc Welding (FCAW)

The filler metal specifications for FCAW process are governed by AWS
A5.20 for carbon steel electrodes, while AWS A5.29 governs the low-alloy
steel electrodes. Electrodes covered by AWS A5.20 will have either 60 or
70 ksi minimum tensile strength. These electrodes are most commonly
used within a steel joist manufacturing facility using FCAW process due to
the fact that the design of the steel involves materials with a nominal yield
strength up to 50 ksi. Electrodes covered by AWS A5.29 are used when
low alloy steel electrodes are required.

12
 Electrodes governed by AWS A5.20 typically will be identified with E7XT-
XM. The “E” stands for Electrode, the 7 indicates the minimum specified
tensile strength of 70 ksi, the first “X” indicates the position in which the
electrodes may be used, the letter “T” indicates a tubular electrode, the
second “X” designates the shielding type (gas or self shielding), welding
polarity, single versus multiple pass, and CVN properties, and the letter “M”
designates the type of shielding gas used.

In addition, electrodes governed by AWS A5.29 typically will be identified

with E7XTX-XM. The difference between this identification and that given in
AWS A5.20 is the inclusion of an additional suffix letter “X” followed by a
number (after the letter “T”), which indicates the alloy type of the electrodes.

C. Gas Metal Arc Welding (GMAW)

The filler metal specifications for GMAW process are governed by AWS
A5.18 for carbon steel electrodes, while AWS A5.28 governs the low-alloy
steel electrodes.

Electrodes governed by AWS A5.18 typically will be identified with ER70S-

X, E70C-XC, or E70C-XM. The “E” stands for Electrode, the 70 indicates
the minimum specified tensile strength of 70 ksi, the letter “S” indicates a
solid wire while the letter “C” indicates a composite (metal-cored) electrode,
”X” indicates the electrode, or depository chemistry, and CVN properties,
the letters “C” or “M” indicate the type of shielding gas used.

In addition, electrodes governed by AWS A5.28 typically will be identified

with ER70S-XXX or E70C-XXX. The difference between this identification
and that given in AWS A5.18 is the inclusion of an additional suffix letter “X”
followed by a number which indicates the alloy type of the electrodes.

Extreme Temperature Conditions

The Lowest Anticipated Service Temperature (LAST) is a critical variable.
When joists and Joist Girders are being manufactured for use in extreme
temperature conditions, it may be required that Charpy V-Notch testing be
conducted to determine weld ductility at specific temperatures. Typical
requirements are 20 ft.-lbs at -20°F, but may be lower based on LAST (e.g.
20 ft.-lbs at -100°F in some areas such as Alaska, Canada, as well as some
Northern States.

13
 CHAPTER 4
SJI STANDARD WELDING PROCEDURES
Manufacturers shall have a written program establishing weld procedures for all
variations of welding performed at their facility. Welding procedures shall
document all essential variables for the process used. An essential variable is
any welding parameter that will impact the quality of the product if changed
beyond prescribed limits. Common essential variables may include, but are not
limited to, process, amperage, voltage, arc transfer mode, electrode size, travel
speed, and shielding gas mixture and flow rate (e.g. see AWS D1.1 Chapter 4,
Table 4.5 for typical essential variables and ranges). These are typical variables
affecting weld quality; if specific scenarios impose additional variables that
impact weld quality, such as joint access restriction, high levels of restraint, or
unusual material thicknesses, these should be addressed in the weld procedure.

Each weld procedure shall be based on a documented Procedure Qualification

Test (PQT) to verify the adequacy of the weld procedure for producing
acceptable welds, unless prequalified per AWS D1.1 Chapter 3. The PQT shall
give in detail all variables used to produce an acceptable weld with a given
process for a given joint configuration. It shall state the welder’s name and date
the test was administered. The results of all nondestructive or destructive testing
performed to verify weld soundness shall be described on the PQT. Tests used
may include bend testing, tensile testing, macro-etching, radiographic, ultrasonic,
or other methods suitable for the joint being tested. Every weld procedure that is
not pre-qualified shall have a PQT, and all welding performed at the SJI
manufacturing facility shall be based on written weld procedures. Any changes
made to essential variables outside of prescribed limits shall require a new
procedure and/or PQT. The welder responsible for a successful PQT is
automatically qualified for procedures based on that PQT. An AWS PQR
(Procedure Qualification Record) or a CWB WPQR (Welding Procedure
Qualification Record) shall be considered equivalent to a PQT in all respects.

The following pages give examples of both a prequalified welding procedure, and
a procedure qualified by testing. Many of the joints and processes used in joist
manufacturing can use AWS prequalified procedures and avoid the added work
of qualification testing. However, qualification by testing is often beneficial in that
it can verify sound welding over a wider range of parameters or the realization of
consistently larger welds for a given set of parameters. Also, some processes,
like Flash Welding (sometimes used for chord splices) do not have prequalified
procedures and would require a PQT. The sample forms shown are samples
only. Any form generated by a member company with a different format is
acceptable as long as it provides the critical information such as essential
variables, test results if qualified by testing, joint information, base metal and filler

14
metal information, and is signed by a company representative authorized to do
so. The person signing the weld procedures and PQT’s should be a Certified
Welding Inspector (CWI) per AWS QC1.

Each manufacturer shall have an established program for minimum weld

sampling and testing. Weld testing shall use a destructive or non-destructive
method applicable to the joint being tested, and sufficiently capable of
determining weld integrity. The testing method may be done by the joist
manufacturer’s personnel or by third-party personnel.

15
Figure 4.1 Sample Prequalified Welding Procedure

16
Figure 4.2 Sample Welding Procedure qualified by testing

17
Figure 4.3 Sample Procedure Qualification Test (PQT) Form

18
Figure 4.4 Sample Test Results Form for PQT

19
 CHAPTER 5
SJI OPERATOR QUALIFICATION

Welders and welding operators may be qualified per SJI standards as described
below. Qualification methods complying with the American Welding Society
(AWS) and the Canadian Welding Bureau (CWB) may be used for welder or
welding operator qualification in lieu of SJI qualification described below.
Welders and welding operators shall be qualified per SJI as follows:
• A welder shall be tested by welding each type of joint he or she will be
expected to weld in the production environment. Successfully passing a
test with a flare bevel groove weld would qualify the welder to weld that
joint, while a fillet test qualifies the production of fillet welds, etc.
• The required weld throat and length for a weld typical of that joint type
shall be specified.
• The welder or welding operator shall weld a representative specimen such
as shown in Figure 5.1, in the position to be required in production. If the
welder is to produce welds in the overhead position, for example, the test
specimen would need to be set up for an overhead weld. Flash welder
operators shall weld three specimens representing the minimum, middle,
and maximum cross-sections for which they are to be qualified.
• Inspect the weld for conformance to SJI visual inspection criteria. If
applicable, the specimen may be broken by sledge hammer or press for
break tests (Figure 5.1), and the fractured weld throat and length
recorded. This would apply to fillet, flare bevel groove, and flare vee
groove welds. Joints such as complete joint penetration groove welds or
other joints not suitable for break tests shall be tested as appropriate to
the joint. Complete joint penetration welds (such as flash-welded
components) would typically be tested by tensile testing (full section or
reduced section) or bend testing, unless specimen geometry was not
suitable for tensile or bend testing. Pry tests and torsion tests may be
used for resistance welds. Plug welds may be tested by loading in shear
or torsion. Non-destructive examination methods such as Ultrasonic
Testing (UT) and Radiographic Testing (RT) may be used in lieu of
destructive tests depending on joint geometry. In all cases, macro-etching
multiple cross-sections through the weld to verify completeness of
penetration, acceptable weld throat, and thorough fusion between layers
would be acceptable.
Welding personnel shall be tested on a manufacturer determined schedule such
that each welder is re-qualified, at a minimum, at least once every 2 years on
each process and position for which he was previously qualified per SJI

20
standards, or in accordance with AWS or CWB standards if tested to those
standards. Welders and welding operators may be qualified by in-house
personnel provided they are Certified Welding Inspectors per AWS QC1.

Figure 5.1 Operator Qualification Specimens for Open Web Steel Joists

21
 CHAPTER 6
SEISMIC WELDING REQUIREMENTS
The basic premise of a Seismic Load Resisting System (SLRS) is to create a
mechanism that allows a ductile, plastic hinge to form and dissipate seismic
energy. In a typical lateral load resisting frame it is the beam member that is
designed to perform inelastically, whereas in a Joist Girder frame it would be the
column member. In either case the welded beam or Joist Girder to column
connection needs to behave in a ductile manner in order for the appropriate
hinge mechanism to form; therefore, this joint requires special attention within an
SLRS. Requirements for welds in a SLRS are provided in ANSI/AISC 341-05,
Seismic Provisions for Structural Steel Buildings (AISC 2005c), and they vary
depending on the type of system. This chapter will review the specific
requirements of this document which may affect the design and manufacture of
open web steel joists or Joist Girders.

Where steel joists or Joist Girders are used as part of a SLRS moment frame,
they should only be used as part of an Ordinary Moment Frame (OMF), as
discussed in SJI Technical Digest 11, Design of Lateral Load Resisting Frames
using Steel Joists and Joist Girders (SJI 2007). In an OMF consisting of wide
flange beams and columns, the theory in use is that of “strong column – weak
beam”, whereby the seismic hinge would form in the beam. An OMF consisting
of columns supporting steel joists or Joist Girders is exactly opposite. The theory
instead is that of “strong beam – weak column”, where the seismic hinge will form
in the column. Fortunately then, the many welds of the steel joist or Joist Girder
are part of the “strong beam”, away from the hinge mechanism, and largely
unaffected by the special seismic welding provisions.

Ordinary Moment Frame Welding Requirements

The welding requirements from ANSI/AISC 341-05 cited below are required for
ordinary moment frames (OMF) when the structure is located in Seismic Design
Category D, E, or F, or in Seismic Design Category A, B, or C, if values greater
than 3 are used for the Response Modification Factor, R.

1. Section 7 of the AISC Seismic Provisions is to be followed for chord splices in

the seismic load resisting system (SLRS), and for the connections between
the joists in the SLRS and the columns.

2. For the fillet welded connections between the joist chords and web members
in the SLRS, Section 7.3 Welded Joints shall be followed with the exception
that the welding is performed per SJI requirements as defined in this chapter
of the technical digest. In addition, Section 7.3a General Requirements
shall be followed in its entirety.

22
Citations from the AISC Seismic Provisions:
From Section 7. “Connections, Joints, and Fasteners”
7.3. Welded Joints
Welding shall be performed in accordance with Appendix W.
Welding shall be performed in accordance with a welding procedure
specification (WPS) as required in AWS D1.1 and approved by the
engineer of record. The WPS variables shall be within the
parameters established by the filler metal manufacturer.

From Appendix W2. “Structural Design Drawings and Specifications,

Shop Drawings and Erection Drawings”

W2.1. Structural Design Drawings and Specifications

Structural design drawings and specifications shall include, as a
minimum, the following information:
(1) locations where backup bars are required to be removed
(2) locations where supplemental fillet welds are required when
backing is permitted to remain
(3) locations where fillet welds are used to reinforce groove welds
or to improve connection geometry
(4) locations where weld tabs are required to be removed
(5) splice locations where tapered transitions are required

W5.4. Maximum Interpass Temperatures

Maximum interpass temperatures shall not exceed 550°F (290°C),
measured at a distance not exceeding 3 in. (75 mm) from the start of
the weld pass. The maximum interpass temperature may be
increased by qualification testing that includes weld metal and base
metal CVN testing using AWS D1.1 Annex III. The steel used for the
qualification testing shall be of the same type and grade as will be
used in production.

The maximum heat input to be used in production shall be used in

the qualification testing. The qualified maximum interpass
temperature shall be the lowest interpass temperature used for any
pass during qualification testing. Both weld metal and HAZ shall be
tested. The weld metal shall meet all the mechanical properties
required by Section 7.3a, or those for demand critical welds of
Section 7.3b, as applicable. The heat affected zone CVN toughness
shall meet a minimum requirement of 20 ft.-lbs (27 J) at 70°F (21°C)
with specimens taken at both 1 and 5 mm from the fusion line.

23
 7.3a. General Requirements
All welds used in members and connections in the SLRS shall be
made with a filler metal that can produce welds that have a minimum
Charpy V-Notch toughness of 20 ft.-lbs (27 J) at 0°F (minus 18°C),
as determined by the appropriate AWS A5 classification test method
or manufacturer certification. This requirement for notch toughness
shall also apply in other cases as required in these Provisions.

Special Design Criteria related to Seismic Detailing of Ordinary Moment

Frames
In order for the joist manufacturer to properly design the joists for the seismic
requirements, certain facts about the SLRS as designed by the Specifying
Professional must be provided. The Specifying Professional is always required
to designate on the Structural Design Drawings, and/or in the Project
Specifications, the items listed in Section 5.1 of the AISC Seismic Provisions for
Structural Steel Buildings (AISC 2005c), regardless if the Seismic Design
Category is other than A, B, or C, or if an R value greater than 3.0 is used for the
design. No special requirements are imposed for Seismic Design Category A, B,
or C, if an R value of 3 or less is used for the design.

From the following list of items taken from Section 5.1 of the AISC Seismic
Provisions, several may potentially be important to the joist manufacturer:

(1) Designation of the seismic load resisting system (SLRS)

(2) Designation of the members and connections that are a part of the
SLRS
(3) Configuration of the connections
(4) Connection material specifications and sizes
(5) Locations of demand critical welds
(6) Locations and dimensions of protected zones
(7) Welding requirements as specified in Appendix W, Section W2.1

However, Items (3) and (4) above are the responsibility of the Specifying
Professional, and complete details should be provided on the contract drawings.
Also, it should be noted that for an OMF there are no protected zones. The only
demand critical weld in an OMF would be a full penetration weld directly to the
column, which is unlikely, or would be made in the field, if required. If the
applicable building code requires the use of the AISC Seismic Provisions for
Structural Steel Buildings (AISC 2005c) and the joist is a part of the SLRS, the
requirements stated at the beginning of this chapter under “Welding
Requirements” must be followed.

24
Section 7.1 of the Provisions requires a ductile limit state to govern design.
Therefore, it is implied that weld strengths should not govern. For welds, the
requirements of Section 7.3 of the Provisions must be followed. Specifically, this
requires welding to be performed in accordance with the American Welding
Society, AWS D1.8 Structural Welding Code – Seismic Supplement (AWS 2005),
and electrodes must meet certain minimum Charpy V-Notch (CVN) toughness
requirements [Note: Pending the code adoption of AWS D1.8, these
requirements can be found on an interim basis in Appendix W of the AISC
Seismic Provisions]. Unless the weld is designated as demand critical, the CVN
requirement is as per Section 7.3a, as noted above. The level of CVN as
required in Section 7.3a is readily available in many electrodes. The joist
manufacturer may elect to either ensure that all welding wire meets the CVN for
use in a SLRS, or may elect to switch to special welding wire only when a SLRS
is designated.

If the joists are a part of an OMF, the requirements of Section 11 of the AISC
Seismic Provisions for Structural Steel Buildings (AISC 2005c) must be followed.
For the fully restrained (FR) moment connections the requirements of Section
11.2a must be followed.

The joist manufacturer must list on the Erection Drawings (Joist Placement
Plans) the following items related to welding from Section 5.3 of the AISC
Seismic Provisions for Structural Steel Buildings (AISC 2005c):
(1) Designation of the members and connections that are part of the SLRS
(2) Field connection material specifications and sizes, if applicable
(3) Field welding requirements as specified in Appendix W, Section W2.3, if
applicable

Welding Requirements for Open Web Steel Joists used in Diaphragm

Systems
It is the opinion of the Steel Joist Institute that joists incorporated into horizontal
diaphragms as collectors or chords, need only to adhere to the requirements of
Section 7 and Appendix W for end connections and any splices in the chords if
the seismic forces do not go through the web members. Welded connections of
web members to the top chord and any bottom chord welds would therefore not
need to meet these criteria.

Welding Requirements for Other Seismic Load Resisting Systems

Additional requirements apply to other SLRS, such as Intermediate Moment
Frames (IMF), Special Moment Frames (SMF), and Special Truss Moment
Frames (STMF), which are beyond the scope of this document, or the SJI
Standard Specifications.

25
 CHAPTER 7
WELD DESIGNS
Advances in technology have changed the way open web steel joists and Joist
Girders are being designed and manufactured. What was standard to the joist
industry 20 or more years ago may no longer be the best design option. The
development of single angle and channel web members in lieu of round rod webs
has greatly increased the strength to weight ratio of steel joists providing a more
efficient and economical product for the end user. The efficiency of these type
web systems increases with size and span. While the round rod web system is
still being used in the smaller, lighter sections, the larger sections are single web,
double web or a combination of both. The weld of the joist’s web members to the
chords can be a combination of flare bevel groove or fillet welds depending on
the type of web member being utilized. Figure 7.1 shows the manufacture of a
single Joist Girder while Figure 7.2 illustrates the multiple process many joist
manufacturers use in building the products.

Figure 7.1 Single Joist Girder being Welded

26
 Figure 7.2 Multiple K-Series Joists being Welded

The Standard Specifications for K-Series, LH- and DLH-Series joists, and Joist
Girders (SJI 2005) and CJ-Series joists (SJI 2007) provide only the broadest
language on welding and connections and the applicable specification sections
for each of these joist series can be found in Appendix A of this digest. The K-
Series Specification, Section 4.5(b) addresses the required design weld
strengths in general terms whether that weld is for a joint connection or a shop
splice. No specific weld joint configurations are currently written into the
specification (i.e. flare bevel, flare vee or butt type groove welds, fillet welds, or
plug and slot welds). When a joint connection is between a round bar web and
an angle chord, the SJI has determined the appropriate weld strength to be used
in designing the weld based on research as described in the following section.

The LH- and DLH-Series Specification, contains slightly more weld design
information as Section 103.2(d) addresses the design and allowable weld
strength for fillet welds used in joist manufacturing. This section also addresses
groove or butt welds relating their strength to that of the connected material.
Again, Section 103.5(b) addresses the required design weld strengths in general
terms whether that weld is for a joint connection or a shop splice. As stated
above, when a joint connection is between a round bar web and an angle chord,
the SJI has determined the appropriate weld strength to be used in designing the
weld based on research as described in the following section. The Joist Girder
Specification and the CJ-Series Specification have the same language as the
LH- and DLH-Series Specification regarding weld strengths.

27
Since the SJI Standard Specifications do not address the strengths of all weld
types or joint configurations, it is incumbent on each joist manufacturer to design
these joist member connections using a recognized or verified design approach
that achieves the required weld strength. Two resources that may be used in
developing weld design procedures are: The 2005 AISC Specification for
Structural Steel Buildings (AISC 2005b) and the 2007 Edition North American
Specification for the Design of Cold-Formed Steel Structural Members (AISI
2007). The appropriate portions of these specifications related to weld joint
configurations used in joist manufacturing will be described in this chapter.

SJI Welding

Weld strength for round bar flare groove welds

Research conducted in the 1970’s at Washington University in St. Louis, MO was
done exclusively for the welding of the round rod joist web system. Lamprecht
and Galambos (1977) conducted an extensive series of joist tests to provide
verification of the strength of flare V and flare bevel groove welds commonly
used for the manufacture of round rod web joists. It was determined that a prime
variable in these welds is the web bar diameter and that a relationship exists
between the web bar diameter and the throat of a flare bevel weld. The details of
this research and its findings can be found in Appendix B.

Based on the research, a relationship was established that the flare groove weld
joint strength be determined by the web diameter and the throat of the flare bevel
weld as follows:

T = 0.12D + 0.11 (Equation of minimum weld throat) Eq. 7-1

S = 21T (Based on 0.3 x nominal tensile strength of E70 electrode) Eq. 7-2

Where D = web diameter, inches

T = weld throat, inches
S = weld strength, kips/in.

Table 7-1 provides a summary of the SJI criteria that was developed for flare
groove welds using Eqs. 7-1 and 7-2.

28
 Table 7-1 SJI Flare Bevel Weld Throat Thickness and Weld Strength

Web Size SJI Throat Thickness SJI Weld Strength

(Eq. 7-1) (Eq. 7-2)
Diameter Radius
in. k/in.
in. in.
1/2 0.250 0.170 3.57
9/16 0.28125 0.178 3.73
5/8 0.3125 0.185 3.88
11/16 0.34375 0.192 4.04
3/4 0.375 0.200 4.20
13/16 0.40625 0.208 4.36
7/8 0.4375 0.215 4.52
15/16 0.46875 0.222 4.67
1 0.500 0.230 4.83

The above research was conducted when the prevalent shortspan joist type was
the H-Series and the longspan joist types were the LH- and DLH-Series. Since
then, the SJI has adopted the K-Series (in 1986), and most recently the CJ-
Series (in 2006) to its Standard Specifications (SJI 2005, 2007). The many years
of successful application of the research results to the H-Series, in conjunction
with other, independent testing by different joist manufacturers has allowed the
SJI member companies to use the same weld strength results where flare groove
welds are used in joint connections when designing K-Series, LH- and DLH-
Series, and CJ-Series joists.

Chord and web splice design and requirements

Chord butt weld splices are common in joist manufacturing. Joist chord material
goes through a procedure that limits the amount of drops (i.e. scrap material) by
butt welding consecutive pieces and shearing to the required lengths.

The Steel Joist Institute’s Standard Specification for Longspan Steel Joists, LH-
Series and Deep Longspan Steel Joists, DLH-Series (SJI 2005) Section
103.5(b)(2) Shop Splices states that,
“Shop splices may occur at any point in chord or web members.
Splices shall be designed for the member force but not less than 50
percent of the member strength. Members containing a butt weld
splice shall develop an ultimate tensile force of at least 57 ksi (393
MPa) times the full design area of the chord or web. The term
“member” shall be defined as all component parts comprising the
chord or web, at the point of the splice.”

29
Obviously, if a chord or web member or component is spliced with a complete
joint penetration weld, the full capacity of the pieces being joined is developed
and all specification and performance requirements are met. However, since the
SJI Standard Specifications are performance based, the full capacity may not
necessarily need to be developed in order to perform adequately, with an
appropriate safety factor. In addition, the intent of the specifications is to allow
the use of added materials or reinforcement as part of the overall design of the
spliced joint. Figure 7.3 shows an unreinforced chord butt weld splice and Figure
7.4 shows a reinforced chord butt weld splice.

Figure 7.3 Typical Chord Angle Butt Splice Weld

On a performance basis, each chord or web has a minimum required cross-

sectional area in order to limit the required stress to less than the design stress
(LRFD) or allowable stress (ASD). This minimum area is what is meant by the
expression “full design area”. The chord or web cross section is often made up
of multiple components, such as two angles, or possibly two angles plus a
reinforcing piece. The performance of the chord or web should be based on the
components acting together, which is the intent of the definition for the term
“member”.

The 57 ksi value has been a part of the SJI Specifications for many years, but
does not take into account allowed variations in yield strengths, or variations in
ultimate strengths, depending on the ASTM Specification for the base material

30
being used. The authors suggest that the acceptance criteria for a spliced joint
be based on developing the overall required capacity of the entire joist, with a
safety factor. Where steel with a yield strength of 50 ksi is utilized, a safety factor
of two on allowable stress results in an acceptance criteria of 60 ksi, or 1.2 times
the yield strength. This criteria, in lieu of the 57 ksi value, is being proposed for
the next editions of the SJI Standard Specifications.

Figure 7.4 Typical Reinforced Chord Angle Butt Splice Weld

Fillet welds
Fillet welds are the most common weld used in joist manufacturing. For single
hot-rolled crimped angles and cold-formed channels, double angle webs and
base plates, a fillet weld may be used in the manufacturing process. Crimped
angle webs and cold-formed angle and channel webs may use flare-bevel groove
welds as well as fillet welds. For simplicity, the flare-bevel groove welds may be
treated as though they are fillet welds, for design purposes. Appendix A contains
the applicable sections of the Standard Specifications for fillet weld strength:
Section 103.2 of the LH-and DLH-Series Specification, Section 1003.2 of the
Joist Girder Specification, and Section 103.2 of the CJ-Series Specification.

The fillet welding of two double-angle web members to a chord are shown in
Figure 7.5 and the fillet welding of a top chord angle to a seat angle is shown in
Figure 7.6.

31
Figure 7.5 Fillet Welded Double-Angle Web Members to Bottom Chord

Figure 7.6 Typical Lapped Seat Angle to Top Chord Fillet Welds

32
Weld placement on single angles
A minimum of one-half of the total required design weld length should be placed
on each chord member for web to chord connections. It is okay to have more
weld on one chord member than the other, so long as each is at least one-half of
the total required design weld length. Research recently concluded at Villanova
University (Dinehart et al. 2008) on single angle crimped and uncrimped joist web
members concluded the following:
1) Web members with the largest amount of weld parallel to the length of
the web member sustained the highest load capacities in compression
(see Figure 7.7). Column buckling of the web member was the
controlling limit state, not the strength of the welded joint.
2) The welds perpendicular to the length of the member, or ‘wrap around
welds’ (see Figure 7.7) do not play a significant role in increasing
column buckling strength of the web member, though they likely would
increase strength of the welded joint, itself, as indicated by AWS and
AISC weld design criteria.

Figure 7.7 Web Weld Measurements

3) For uncrimped, rotated, single angle joist web members in the gap
between the chords, the effects of eccentricity reduce the load capacity
of the member. A bending moment exists in the member because the
load is not applied through the member’s centroid (see Figure 7.8).
However, the location and placement of the welds on these members
was not a factor in, and did not control, the performance of these
members.

33
Figure 7.8 Centroid Alignment for Crimped and Uncrimped Angle Web Members

In total, the full-scale experimental testing program at Villanova University

involved the load testing of over 100 joists with a total number of welded joints in
excess of 1800. All of the joists were taken to failure, and in not one of these
tests, did the weld control or limit the joist capacity, nor did any welds tear or fail
prematurely. The experimental program validated the industry practices for the
location and placement of the welds on single angle web members.

AISC WELDS, Chapter J (AISC 2005b)

The design strength, φRn and the allowable strength, Rn/Ω, of welds shall be the
lower value of the base material and the weld metal strength determined
according to the limit states of tensile rupture, shear rupture or yielding as
follows:
For the base metal
Rn = FBMABM (AISC J2-2)

For the weld metal

Rn = FwAw (AISC J2-3)

Where,
FBM = nominal strength of the base metal per unit area, ksi (MPa)
Fw = nominal strength of the weld metal per unit area, ksi (MPa)
ABM = cross-sectional area of the base metal, in.2 (mm2)
Aw = effective area of the weld, in.2 (mm2)

Fillet welds
The effective area of a fillet weld shall be the effective length multiplied by the
effective throat. The effective throat of a fillet weld shall be the shortest distance
from the root to the face of the diagrammatic weld.

34
AISC Table J2.5 Available Strength of Welded Joints, kips (N) gives the weld
requirements for Fillet Welds Including Fillets in Holes and Slots and Skewed T-
Joints in shear is:
Fw = 0.6 FEXX

Where,
FEXX = electrode classification number, i.e., minimum specified
tensile strength, ksi

The design strength, φRn and the allowable strength, Rn/Ω, of fillet welds is as
follows:
φ = 0.75 (LRFD) Ω = 2.00 (ASD)

Rn = FwAw

For an E70xx electrode the weld stress is,

Fw = 0.60 FEXX = 0.60 (70) = 42 ksi

And the nominal weld strength for an effective throat thickness tw is,
Rn = 42 Aw = (42)(tw)(weld length)

The design weld strength (LRFD) is,

φRn = 0.75(42) Aw = 22.271(t)(weld length)

The allowable weld strength (ASD) is,

Rn/Ω = (42/2.0) Aw = 14.847(t)(weld length)

Where,
t = nominal fillet weld thickness
tw = effective fillet weld throat thickness = 0.707t

Groove welds
The effective area of groove welds shall be considered as the length of the weld
times the effective throat thickness. The effective throat thickness of a complete-
joint-penetration groove weld shall be the thickness of the thinner part joined.
The effective throat thickness of a partial-joint-penetration groove weld shall be
as shown in AISC Table J2.1. Effective Throat of Partial-Joint-Penetration
Groove Welds.

35
The effective weld size for flare groove welds shall be as shown in Table J2.2
Effective Weld Sizes of Flare Groove Welds. The applicable portions of this table
related to joists are given in Table 7-2.

Table 7-2 Effective Weld Sizes of Flare Groove Welds

Welding Process Flare Bevel Groove Weld Flare V Groove Weld
Size* Size
SMAW and FCAW-S 5/16R 5/8R
GMAW and FCAW-G 5/8R 3/4R
*For Flare Bevel Groove Welds with R < 3/8 in. (10 mm) use only reinforcing fillet weld on filled
flush joint.
General Note: R = radius of joint surface, in. (mm)

Table J2.5 Available Strength of Welded Joints, kips (N) gives the weld
requirements for Complete-Joint-Penetration Groove Welds as well as Partial-
Joint-Penetration Groove Welds Including Flare Vee Groove and Flare Bevel
Groove Welds. For complete-joint-penetration groove welds the strength of the
joint is controlled by the base metal regardless of whether the direction of loading
is normal to the weld axis, either in tension or compression, or the weld is in
shear.

The strength of partial joint penetration groove welds in shear is:

Fw = 0.6 FEXX

The design strength, φRn and the allowable strength, Rn/Ω, of partial joint
penetration groove welds is as follows:
φ = 0.75 (LRFD) Ω = 2.00 (ASD)

Rn = FwAw

Plug and slot welds

Plug and slot welds can be used to effectively transmit shear parallel to the
faying surface in lap joints and to prevent buckling of lapped members. A plug
weld is made in a circular hole of a member in a lap joint, while a slot weld is
made in an elongated hole of a member in a lap joint. Table J2.5 Available
Strength of Welded Joints, kips (N) gives the weld requirements for Plug And Slot
welds in shear parallel to the faying surface on the effective area as follows:
Fw = 0.6 FEXX

36
The design strength, φRn and the allowable strength, Rn/Ω, of plug and slot weld
is as follows:
φ = 0.75 (LRFD) Ω = 2.00 (ASD)

Rn = FwAw

Where,
Aw = effective area of plug and slot weld (the nominal cross-
sectional area of the hole or slot in the plane of the faying
surface)

The effective shearing area of plug and slot welds shall be considered as the
nominal cross-sectional area of the hole or slot in the plane of the faying surface.
The base metal strength is governed by Equation J2-2.

There are geometrical limitations when using plug and slot welds. Some of these
limitations include:
• The minimum hole diameter or slot width shall not be less than the
thickness of the containing member plus 5/16 inch, rounded to the nearest
odd 1/16 inch.
• The maximum hole diameter or slot width shall be the minimum diameter
(slot width) plus 1/8” or 2¼” times the thickness of the containing member,
whichever is greater.
• The length of the slot for slot weld is not to exceed 10 times the thickness
of the containing member.
• The ends of the slot for slot weld shall be semicircular or shall have
corners rounded to a radius no less than the thickness of the containing
member, except those ends which extend to the edge of the member.
• The minimum center to center spacing of slot welds in the direction
transverse to their length shall be four times the width of the slots. The
minimum center to center spacing of slot welds in the longitudinal direction
shall be two times the length of the slot.
• The thickness of plug and slot welds in material 5/8 inch or less in
thickness shall be the thickness of the material. For material over 5/8 inch
thick, the thickness of the weld shall be at least one half of the thickness of
the material, but not less than 5/8 inch.

The intent of the above limitation is to provide geometrical conditions of plug and
slot welds that are conducive to a good fusion of the weld material and base
material.

37
AISI CONNECTIONS AND JOINTS, Chapter E (AISI 2007)
The following design criteria shall apply to welded connections used for cold-
formed steel structural members in which the thickness of the thinnest part is
3/16 in. (4.76 mm) or less. For the design of welded connections in which the
thickness of the thinnest connected part is greater than 3/16 in. (4.76 mm), refer
to the AISC Specification for Structural Steel Buildings (AISC 2005b).

Fillet welds
The requirements of Chapter E, Connections and Joints, E2.4 Fillet Welds shall
apply when welding cold-formed steel members in accordance with the North
American Specification for the Design of Cold-Formed Steel Structural Members
(AISI 2007),

The nominal shear strength, Pn, of a fillet weld shall be determined as follows:

(1) For longitudinal loading:

For L/t < 25:
§ 0.01L ·
Pn = ¨ 1 − ¸ LtFu (AISI Eq. E2.4-1)
© t ¹
φ = 0.60 (LRFD) Ω = 2.55 (ASD)

For L/t • 25:

Pn = 0.75 tLFu (AISI Eq. E2.4-2)
φ = 0.50 (LRFD) Ω = 3.05 (ASD)

(2) For transverse loading:

Pn = tLFu (AISI Eq. E2.4-3)
φ = 0.65 (LRFD) Ω = 2.35 (ASD)

In addition, for t > 0.10 in., the nominal strength determined in (1) or (2) shall
not exceed the following:
Pn = 0.75 t wLFxx (AISI Eq. E2.4-4)
φ = 0.60 (LRFD) Ω = 2.55 (ASD)

Where,
Pn = Nominal strength of fillet weld
L = Length of fillet weld
Fu = Tensile strength
Fxx = Tensile strength of electrode classification

38
 tw = Effective throat
= 0.707w1 or 0.707w2, whichever is smaller. A larger effective
throat is permitted if measurement shows that the welding
procedure to be used consistently yields a larger value of tw
Where
w1 and w2 = leg of weld (see Figures 7.9 and 7.10) and w1 ” t1 in
lap joints
t = Least value of t1 or t2 , as shown in Figures 7.9 and 7.10

Figure 7.9 Fillet Weld Lap Joint t1 and Figure 7.10 Fillet Weld T Joint t1 and t2
t2 definitions (AISI 2007) definitions (AISI 2007)

Flare groove welds

The AWS weld criteria for flare bevel groove welds was revised in 2006 (Packer
and Frater 2005) and is now incorporated into the 2005 AISC Specification in
Table J2.2 and is proposed to be incorporated into the next edition of the AISI
NAS as Table E2.5-1. Table 7-2 provides the current AISC and the proposed
AISI requirements for the effective throat sizes for flare bevel and flare V groove
welds.

The nominal shear strength, Pn, of a flare groove weld shall be determined as
follows:
(a) For transverse loading (see Figure 7.11):
Pn = 0.833tLFu (AISI Eq. 2.5-1)
φ = 0.60 (LRFD) Ω = 2.55 (ASD)

(b) For longitudinal loading (see Figures 7.12 and 7.13):

(1) For t ” tw < 2t or if the lip height, h, is less than weld length, L:
Pn = 0.75tLFu (AISI Eq. 2.5-2)
φ = 0.55 (LRFD) Ω = 2.80 (ASD) (ASD)

(2) For tw • 2t with the lip height, h, equal to or greater than weld length, L:
Pn = 1.50tLFu (AISI Eq. 2.5-3)
φ = 0.55 (LRFD) Ω = 2.80 (ASD) (ASD)

39
(c) For t > 0.10 in. (2.54 mm), the nominal strength determined in (a) or (b)
shall not exceed the value of Pn calculated in accordance with,
Pn = 0.75twLFxx (AISI Eq. 2.5-4)
φ = 0.60 (LRFD) Ω = 2.55 (ASD) (ASD)

Where,
Pn = Nominal strength of flare groove weld
t = Thickness of welded member as defined in Figures 7.11 to
7.17
L = Length of weld
Fu = Tensile strength
Fxx = Tensile strength of electrode classification
tw = Effective throat of flare groove weld filled flush to surface (see
Figures 7.14 and 7.15
= (5/16) R for flare bevel groove weld
= (1/2) R when R ” 1/2 in. (12.7 mm) for flare V-groove weld
= (3/8) R when R > 1/2 in. (12.7 mm) for flare V-groove weld
= Effective throat of flare groove weld not filled flush to surface:
= 0.707w1 or 0.707w2, whichever is smaller (see Figures 7.16
and 7.17)
= A larger effective throat than those above is permitted if
measurement shows that the welding procedure to be used
consistently yields a larger value of tw
Where
R = Radius of outside bend surface
w1 and w2 = Leg of weld (see Figures 7.16 and 7.17)

Figure 7.11 Flare Bevel Groove Weld – Transverse Loading (AISI 2007)

40
Figure 7.12 Flare Bevel Groove Weld Figure 7.13 Flare V Groove Weld in
in Shear (AISI 2007) Shear (AISI 2007)

Figure 7.14 Flare Bevel Groove Weld, Figure 7.15 Flare Bevel Groove Weld,
filled flush to surface, w1 = R (AISI filled flush to surface, w1 = R (AISI
2007) 2007)

Figure 7.16 Flare Bevel Groove Weld, Figure 7.17 Flare Bevel Groove Weld,
not filled flush to surface, w1 > R (AISI not filled flush to surface, w1 < R (AISI
2007) 2007)

Typical examples of fillet and flare groove weld configurations are shown in
Figures 7.18 and 7.19, respectively.

41
 Figure 7.18 Cold-Rolled Channel Webs Welded to Top Chord

Figure 7.19 Cold-Rolled Channel Webs Welded to Bottom Chord

42
 CHAPTER 8
PLANT INSPECTION CRITERIA
The SJI Standard Specifications for K-Series, LH- and DLH-Series joists and
Joist Girders (SJI 2005) and CJ-Series joists (SJI 2007) list seven specific
acceptance criteria for the quality of welded connections. These are listed below,
with the exact same language as they appear in each of the Standard
Specifications, followed by some commentary language, to further clarify and
explain each item.

ACCEPTABLE WELD QUALITY

1. Inspection method—selected welds shall be inspected visually by the

manufacturer. Prior to this inspection, weld slag shall be removed.
Commentary: To be able to adequately see the weld for visual inspection,
weld slag shall be removed. However, the GMAW process
used prevalently in the joist industry provides an
essentially “slag-free” weld.

2. Cracks are not acceptable and shall be repaired.

Commentary: Cracks are not acceptable in any portion of the structural
weld. They shall be repaired by grinding down to sound
metal and filling. This does not include temporary welds.

3. Thorough fusion shall exist between layers of weld metal and between
weld metal and base metal for the required design length of the weld; such
fusion shall be verified by visual inspection.
Commentary: SJI does not require regular, ongoing, or daily physical
testing to ensure fusion. Once the joist is constructed, its
configuration renders many types of physical testing
impractical, as discussed later in this chapter. Visual
inspection for fusion is adequate, although sampling and
testing is discussed later in this chapter. The types of
conditions that may suggest a concern or potential for lack
of fusion include: “cold-lapping”, an uneven or erratic weld
profile, or a poor welding angle as evidenced by a
significant difference in the leg lengths of a fillet weld.

43
4. Unfilled weld craters shall not be included in the design length of the weld.
Commentary: Design length is taken as length of weld required to resist
axial force in members connected. Actual weld length
during manufacture may be longer than required design
length, in which case the length of a crater or inadequate
weld profile can be deducted from the overall weld length
and then the new, reduced length can be compared with
the required length.

5. Undercut shall not exceed 1/16 inch (2 millimeters) for welds oriented
parallel to the principal stress.
Commentary: While undercutting the base metal, particularly the chord
material, is always a concern, obviously undercut that is
transverse to the direction of the member axial load is of
more concern than undercut parallel to the axial load. If
undercut occurs in a chord member, it is normally not in a
clearly transverse or parallel position, because the web
members and their welds typically intersect the chords
diagonally. These “diagonal welds” are not typically
considered as fully parallel to the principal stress. It should
be noted that for almost all types of joist construction, and
particularly for double angle chords, the welded cross
section is limited to roughly one third of the member cross
section. For angles, the outstanding leg, as well as the
heel and part of the distance of the vertical leg, are not
typically welded to, and therefore, are not exposed to the
possibility of undercut.

6. The sum of surface (piping) porosity diameters shall not exceed 1/16 inch
(2 millimeters) in any 1 inch (25 millimeters) of design weld length.
Commentary: SJI limits the sum of the piping porosity diameters to 1/16”
in any 1” of design weld whereas AWS uses a more explicit
criteria in that the sum of diameters of visible piping
porosity 1/32” or greater in fillet welds shall not exceed 3/8”
in any linear inch of weld and shall not exceed 3/4” in any
12” length of fillet weld. As discussed in Item 4 above,
porosity beyond the SJI criteria would require the affected
portion of the weld to be excluded for comparison with the
required weld length.

7. Weld spatter that does not interfere with paint coverage is acceptable.
Commentary: Weld spatter is not considered a structural concern, unless
indicative of other issues with the welding process. Thus,
the primary concern with spatter is that it not interfere with
the application or appearance of a painted surface. Please

44
 note that the shop coat of paint in accordance with the SJI
Standard Specifications is a temporary and impermanent
coating, may not be uniform, and may include drips, runs,
and sags.

In addition to the specific items listed above, Table 8-1 provides a more detailed
comparison between SJI Welding (SJI 2005, 2007) and AWS D1.1 (AWS 2008).

WELD INSPECTION BY OUTSIDE AGENCIES

Each of the SJI Standard Specifications also include instructions for projects
which require weld inspection by outside agencies.

Joists shall be inspected by the manufacturer before shipment to verify

compliance of materials and workmanship with the requirements of the SJI
Standard Specifications. If the purchaser wishes an inspection of the steel joists
by someone other than the manufacturer’s own inspectors, they may reserve the
right to do so in their “Invitation to Bid” or the accompanying “Job Specifications”.

Arrangements shall be made with the manufacturer for such shop inspection of
the joists at the manufacturing shop by the purchaser’s inspectors at the
purchaser’s expense.

The agency shall arrange for visual inspection to determine that welds meet the
standards of acceptable weld quality as defined above. X-ray, ultrasonic and
magnetic particle testing are inappropriate for steel joists due to the configuration
of the components and welds.

WELDING PROGRAM

All of the SJI Standard Specifications include the following requirement regarding
the joist manufacturer’s welding program:
“Manufacturers shall have a program for establishing weld procedures and
operator qualification, and for weld sampling and testing.”

Commentary: The development of weld procedures and operator

qualification has been discussed in Chapters 4 and 5 of this digest. As stated
above, visual inspection of completed joists and welds is adequate, and the
constructed joist configuration makes many test methods impractical. The
“sampling and testing” mentioned in this specification section is not intended
to be part of the inspection of joists, but rather as validation of the processes.
The controlled environment of the manufacturing shop, and the repetitive
nature of the typical welds, allow for predictable and repeatable welds, and
minimizes the need for sampling and testing. However, care must be taken to

45
 avoid, and periodically check for, conditions that could create a general lack
of fusion. One method of sampling and testing for joist welds to rule out the
possibility of a complete lack of fusion is a pry-bar or wedge test. Another
concern is the chord splice welds and the question of complete fusion through
the complete thickness of the weld and member, particularly for thicker
members.

While fully constructed joists are not conducive to many testing methods, a
chord splice, before assembly of the chord into the joist, can be tested a
number of ways, such as ultrasonic testing. The manufacturer can also
readily use sampling for chord splices, and simply replace the chord splices
taken for sampling with a new, identical chord splice. One common method
of testing chord splice samples is a bend test. Another possibility is tension
testing of either full spliced chord sections or of coupons cut from the spliced
cross section.

As an alternative or supplement to the testing and sampling mentioned

above, some joist shops are equipped with full-scale load testing setups,
which allow “proof testing” of random joists to ensure a load-carrying capacity
of 1.65 times the ASD Load Table total load.

SJI does not set or establish specific parameters for the amount of sampling and
testing and leaves this to the joist manufacturer. The volume of joists, and
welding, can vary a great deal from manufacturer to manufacturer, or even
product line to product line, on a daily basis. As a minimum, the sampling and
testing program shall include a method of checking joint fusion and a method of
checking chord splice fusion, and/or full-scale proof testing in accordance with
the manufacturer’s quality control manual.

46
 Table 8-1 SJI Welding vs AWS D1.1
SJI
Applicable Methods
Joint connections and splices shall be made by attaching the members to one another by
arc or resistance welding or other approved method.
SJI requires selected welds be inspected
visually by the joist manufacturer. Prior to
this inspection, weld slag shall be removed.
SJI states that cracks are not acceptable
and shall be repaired.
SJI states that thorough fusion shall exist
between weld and base metal for the
required design length of the weld; such
fusion shall be verified by visual inspection.
SJI states that unfilled weld craters shall
not be included in the design length of the
weld. Craters outside of the effective weld
length are acceptable.
SJI states that undercut shall not exceed
1/16” for welds oriented parallel to the
principal stress.
SJI limits the sum of the piping porosity
diameters to 1/16” in any 1” of design weld.
SJI does not regulate this.
Weld spatter that does not interfere with
paint coverage is acceptable.
Often joist welds with defects outside of
their effective length are not repaired but
simply discounted.
SJI does not address preheat or ambient
temperature.
SJI does not address base metal condition.
Maximum spacing of intermittent welds is
not addressed by SJI.
47
AWS
AWS requires all welds be visually
inspected (D1.1,6.9) and shall be acceptable
if the following conditions are satisfied.
AWS requires determining extent of crack
prior to removal and rewelding. Welds shall
have no cracks.
AWS requires that thorough fusion shall
exist between adjacent layers of weld metal
and between weld metal and base metal.
AWS requires that all craters shall be filled
to the full cross section of the weld, except
for ends of intermittent fillet welds outside
their effective length.
AWS requires that for material less than 1”
thick, undercut shall not exceed 1/32”,
except that a maximum 1/16” is permitted
for an accumulated length of 2” in any 12”
For material equal to or greater than 1”
thick, undercut shall not exceed 1/16” for
any length of weld.
The sum of diameters of visible piping
porosity 1/32” or greater in fillet welds shall
not exceed 3/8” in any linear inch of weld
and shall not exceed 3/4” in any 12” length
of fillet weld per AWS. No visible piping
porosity for transverse butt welds under
tension.
No arc strikes per AWS.
Tightly adherent spatter remaining after the
cleaning operation is acceptable (AWS D1.1
5.30.2).
AWS requires defective welds to be
repaired.
AWS requires 50°F preheat for A529 steel
up to 1½” thick. Also, no welding below
0°F.
Base metal condition (rust, paint, mill scale)
per D1.1, 5.15
Maximum spacing of intermittent welds is
addressed per AWS (D1.1 2.11.2).
 Fillet Welds
SJI does not specify allowable undersizing
of fillet welds.
SJI does not regulate this.
SJI allows 1/16” undercut for welds parallel
to principal stress regardless of thickness.
SJI does not address minimum weld size
thickness.
SJI does not specify allowable root gap.
SJI does not regulate this.
SJI does not regulate the minimum length
for a given fillet weld size.
SJI does not address reduced throat at
corners.
SJI does not regulate this.
Flare Bevel Groove Welds
SJI has an equation for the effective throat
as a function of the round bar diameter (see
Chapter 7).
48
AWS states that fillet welds are allowed to
be slightly undersized for up to 10% of the
weld length.
Minimum length of 1½” for intermittent fillet
welds per AWS.
AWS limits undercut to 1/32” for material
under 1” thick (1/16” undercut for up to 2”
in 12” of weld), or 1/16” for 1” thick and
greater.
Minimum weld size is based on thickness in
addition to stress requirements per AWS.
AWS limits root openings for fillet welds to
3/16” maximum. If over 1/16” gap, weld size
must be increased accordingly.
AWS has a series of acceptable weld
profiles for fillet and groove welds.
Minimum length of 4 x size for fillet welds
per AWS (e.g. 1” min. length for 1/4” fillet).
Fillet welds on opposite sides of a common
plane cannot be tied together per AWS (i.e.
backside welds on double angle webs).
Fillet welds must stop short on lapped parts
where one part extends over the edge of a
tension member per AWS D1.1, 2.8.3.2 (e.g.
double angle webs).
The effective throat for FBG welds is
increased from 5/16R to 5/8R for GMAW.
However, this is dependent on being filled
flush. If not, the amount of underfill below
flush level now must be subtracted from the
5/8R value per AWS.
May be larger per AWS if qualified by
testing per Section 4.10.5.
 Tack (Temporary) Welds
SJI does not regulate tack or temporary
welds.
SJI does not address this.
SJI does not regulate this.
AWS requires tack welds to meet the same
quality requirements as primary welds
(D1.1, 5.18). Tack welds shall be subject to
the same welding procedure requirements
as final welds. They shall be removed when
required by the Engineer of Record. When
they are removed, the surface shall be
made flush with the original surface.
Incorporated tack welds must be cleaned
first.
Tack welders should be qualified

Butt Welds
SJI does not address this. Butt joint alignment to the smaller of 10% of
thinner material or 1/8” per D1.1, 5.22.3.
SJI does not address this. Butt welds within 1/8” of flush (D1.1 5.24.4).
SJI does not address this. Welds to be cleaned between passes and
after finish per D1.1, 5.30. Technically,
GMAW not exempted (i.e. silica islands).
SJI does not address this. Wind in vicinity of GMAW < 5 mph per AWS.

49
 CHAPTER 9
FIELD WELDING AND FIELD INSPECTION
The performance of open web steel joists and the safety of the steel erection
workers are greatly influenced by the welding executed in the field. Proper
welding technique and inspection are essential to the satisfactory performance of
the joists during the construction phase and the life of a given structure.

Field welding shall be in accordance with AWS D1.1, Structural Welding Code-
Steel and AWS D1.3, Structural Welding Code-Sheet Steel. The welding must
be performed with electrodes conforming to AWS A5.1, Specification for Carbon
Steel Electrodes for Shielded Metal Arc Welding or AWS A5.20, Specification for
Carbon Steel Electrodes for Flux Cored Arc Welding. Electrodes must be
properly stored to maintain flux quality. Inspection shall be limited to the welding
executed in the field and shall be performed by qualified personnel.

The welding executed in the field is essential to the stability and serviceability of
a structure. Some key elements will be reviewed in more detail in the following
sections.

Joist end anchorage

In most instances the minimum end anchorage welds specified by SJI are
adequate for Joist Girders, K-, LH-, or DLH-Series joists. However, special
attention from the Specifying Professional is required when joists are subjected
to uplift forces, when axial forces are transferred from the joist’s top chords to the
supporting structure or when joists are utilized in rigid frames. Larger and/or
longer welds may be required to transfer these reaction forces to the supporting
structure.

It is recommended that the Specifying Professional call out 1/8 inch fillet welds
for K-Series joists and 1/4 inch fillet welds for LH- and DLH-Series joists and
Joist Girders with additional length rather than specifying larger welds whenever
possible, due to the fact that the thickness of the joist or girder seat angles may
not accommodate larger weld sizes.

Joist bridging
Joist bridging is employed to stabilize the joists during the construction phase
and therefore is essential to the safety of the field workers. The bridging is also
essential to the performance of the joists particularly in the presence of uplift
forces or when the joist top chords are not laterally supported by decking (e.g. a
standing seam roof). The attachment of the bridging to the joists chord members
must be sufficient to resist anticipated bracing forces and must be in

50
conformance with the Standard Specifications of the relevant joist series. Welds
connecting bridging angles to chord members shall be fillet welds (see Figure
9.1). Four point spot/ tack welds are not acceptable (see Figures 9.2 and 9.3).

The minimum weld requirements for the various SJI Standard joist series are
provided in Tables 8-1 through 8-4 of SJI Technical Digest No. 9 (SJI 2008) to
assist the erector in determining the length of the bridging weld needed for a
particular joist series. Those tables are reproduced as Tables 9-1 through 9-4
below.

Figure 9.1 Acceptable Weld of Bridging Angle Horizontal Leg to Vertical Leg of
Bottom Chord Angle

51
Figure 9.2 Unacceptable Weld of Bridging Angle Horizontal Leg to Vertical Legs
of Both Bottom Chord Angles

Figure 9.3 Unacceptable Welds of Bridging Angle Vertical Leg to the Vertical
Legs of Both Bottom Chord Angles

52
 Table 9-1 Weld to Provide Minimum 700 lb Nominal Strength for K-Series
Joists
Weld Type Weld Leg Size (in.) Minimum Weld Length (in.)
Fillet 7/64 7/16
Fillet 1/8 3/8
Fillet 5/32 5/16

Table 9-2* Weld to Provide Minimum 1000 lb Nominal Strength for LH- and
DLH-Series Joists, Section Numbers 02 through 12, and CJ-Series Joists
Weld Type Weld Leg Size (in.) Minimum Weld Length (in.)
Fillet 7/64 5/8
Fillet 1/8 9/16
Fillet 5/32 7/16

Table 9-3* Weld to Provide Minimum 2000 lb Nominal Strength for LH- and
DLH-Series Joists, Section Numbers 13 through 19, and CJ-Series Joists
Weld Type Weld Leg Size (in.) Minimum Weld Length (in.)
Fillet 7/64 1 1/4
Fillet 1/8 1 1/8
Fillet 5/32 7/8

Table 9-4* Weld to Provide Minimum 3500 lb Nominal Strength for

CJ-Series Joists
Weld Type Weld Leg Size (in.) Minimum Weld Length (in.)
Fillet 7/64 2 1/4
Fillet 1/8 1 7/8
Fillet 5/32 1 1/2

*Tables 9-2, 9-3 and 9-4 have been created to help determine the length of the
bridging weld needed for a given joist series. It is possible that different
combinations of nominal force and weld length can be used depending on the
joist series requirements. Contact the joist manufacturer for different weld
combination possibilities.

53
Metal decking
Welding of metal decking shall be in accordance with AWS D1.3, Structural
Welding Code - Sheet Steel and the Steel Deck Institute requirements. Welding
washers shall be used on all deck units with metal thickness less than 0.028
inches (7 mm). Where welding washers are not used, a minimum visible 5/8 inch
(15 mm) diameter arc puddle weld shall be used. Weld metal shall penetrate all
layers of deck material at end laps and shall have good fusion to the supporting
members. When used, fillet welds shall be at least 1½ inch (38 mm) long.
Special care should be given to the welding of the deck units to short span joists
as the members forming the top chords are generally relatively thin (1/8” thick).
The potential for melt through is therefore greatly increased and, if melt through
occurs, it must be reported to the joist manufacturer for evaluation and
recommendation.

Composite joist construction

Shear connection between Composite Steel Joists, CJ-Series, and the concrete
slab is typically made by welding shear stud connectors ranging in diameter from
3/8 inch to 3/4 inch through the metal deck. Stud welding shall be in accordance
with AWS D1.1, Structural Welding Code-Steel, Section 7 (AWS 2008).

Field modifications
All approved field modifications shall be executed in accordance with AWS D1.1,
Structural Welding Code-Steel. The welding must be performed with the
appropriate E60 or E70 series low hydrogen electrodes, as prescribed by the
joist manufacturer, conforming to AWS A5.1 or A5.20. Electrodes must be
properly stored to maintain flux quality. Since the thickness of the joist members
is generally thin, 1/8 in. welding electrodes are recommended for most repairs.

Field welding through paint

In accordance with the latest SJI Specifications and Code of Standard Practice
(2005), Section 3.2 Paint, it states in part, “The paint is intended to be an
impermanent and provisional coating which will protect the steel for only a short
period of exposure in ordinary atmospheric conditions.” This coating is typically
1.0 mil in thickness, but frequently varies, with ranges between 0.8 mils and 2.0
mils being very common, and often thicker on some horizontal members. The
minimum thickness condition comes from SSPC-Paint 15, Steel Joist Shop
Primer Section 4.4 which states, “The dry film thickness of the primer shall be as
recommended by the manufacturer. If no recommendation is given, the dry film
thickness shall be a minimum of 20 micrometers (0.8 mils) unless indicated
otherwise.” Even at the upper thickness this coating is thin as compared to any
field applied paint system that might be utilized for Open Web Steel Joists.

In AWS D1.1 Section 5.15 Welding to Base Metal, it states that one of the
exceptions allowing welding is, “Mill scale that can withstand vigorous wire

54
brushing, a thin rust-inhibitive coating, or antispatter compound may remain …”
This thin rust-inhibitive coating is essentially the same description that the SJI
uses in their description of the paint and should be interpreted as such.

Therefore, according to the current provisions in AWS, welding through shop

applied joist paint is an acceptable practice by the Steel Joist Institute. If there
are concerns regarding the ability to weld through the shop applied paint, the joist
manufacturer can supply a copy of the specifications for the paint they are using
to show what, if any, limitations there may be for welding through the paint.

55
 REFERENCES
AISC (2005a), Steel Construction Manual, Thirteenth Edition, American
Institute of Steel Construction, Chicago, IL.
AISC (2005b), ANSI/AISC 360-05, Specification for Structural Steel Buildings,
March 9, 2005, American Institute of Steel Construction, Chicago, IL.
AISC (2005c), ANSI/AISC 341-05, Seismic Provisions for Structural Steel
Buildings, March 9, 2005, Including Supplement No. 1, November 16, 2005,
American Institute of Steel Construction, Chicago, IL.
AISC (2005d), ANSI/AISC 358-05, Prequalified Connections for Special and
Intermediate Steel Moment Frames for Seismic Applications, December 13,
2005, Including Supplement No. 1, November 16, 2005, American Institute
of Steel Construction, Chicago, IL.
ASCE (2005), ASCE/SEI 7-05, Minimum Design Loads for Buildings and other
Structures, Including Supplement No. 1, American Society of Civil
Engineers, Reston, VA.
ASTM (2005), ASTM A36/A36M-05, Standard Specification for Carbon Structural
Steel, ASTM International, West Conshohocken, PA.
ASTM (2004), ASTM A242/A242M-04, Standard Specification for High-Strength
Low-Alloy Structural Steel, ASTM International, West Conshohocken, PA.
ASTM (2005), ASTM A529/A529M-05, Standard Specification for High-Strength
Carbon-Manganese Steel of Structural Quality, ASTM International, West
Conshohocken, PA.
ASTM (2006), ASTM A572/A572M-06, Standard Specification for High-Strength
Low-Alloy Columbium-Vanadium Structural Steel, ASTM International, West
Conshohocken, PA.
ASTM (2005), ASTM A588/A588M-05, Standard Specification for High-Strength
Low-Alloy Structural Steel with 50 ksi [345 MPa] Minimum Yield Point to 4-
in. [100-mm] Thick, ASTM International, West Conshohocken, PA.
ASTM (2004), ASTM A606-04, Standard Specification for Steel, Sheet and Strip,
High-Strength, Low-Alloy, Hot-Rolled and Cold-Rolled, with Improved
Atmospheric Corrosion Resistance, ASTM International, West
Conshohocken, PA.
ASTM (2007), ASTM A1008/A1008M-07a, Standard Specification for Steel,
Sheet, Cold-Rolled, Carbon, Structural, High-Strength Low-Alloy, High-
Strength Low-Alloy with Improved Formability, Solution Hardened, and Bake
Hardenable, ASTM International, West Conshohocken, PA.

56
ASTM (2007), ASTM A1011/A1011M-07, Standard Specification for Steel, Sheet
and Strip, Hot-Rolled, Carbon, Structural, High-Strength Low-Alloy with
Improved Formability, and Ultra-High Strength, ASTM International, West
Conshohocken, PA.
AWS (2004), AWS A5.1-2004, Specification for Carbon Steel Electrodes for
Shielded Metal Arc Welding, American Welding Society, Miami, FL.
AWS (1996), AWS A5.5-96, Specification for Low Alloy Steel Electrodes for
Shielded Metal Arc Welding, American Welding Society, Miami, FL.
AWS (1997), AWS A5.17-97, Specification for Carbon Steel Electrodes and
Fluxes for Submerged Arc Welding, American Welding Society, Miami, FL.
AWS (2001), AWS A5.18-2001, Specification for Carbon Steel Electrodes and
Rods for Gas Shielded Arc Welding, American Welding Society, Miami, FL.
AWS (1995), AWS A5.20-95, Specification for Carbon Steel Electrodes for Flux
Cored Arc Welding, American Welding Society, Miami, FL.
AWS (1997), AWS A5.23-97, Specification for Low Alloy Steel Electrodes and
Fluxes for Submerged Arc Welding, American Welding Society, Miami, FL.
AWS (1996), AWS A5.28-96, Specification for Low Alloy Steel Filler Metals for
Gas Shielded Arc Welding, American Welding Society, Miami, FL.
AWS (1998), AWS A5.29-98, Specification for Low Alloy Steel Electrodes for
Flux Cored Arc Welding, American Welding Society, Miami, FL.
AWS (1998), Welding Handbook Eighth Edition, Volume 4, Materials and
Applications – Part 2, American Welding Society, Miami, FL.
AWS (2008), Structural Welding Code – Steel, ANSI AWS D1.1/D1.1M-2008,
American Welding Society, Miami, FL.
AWS (2008), Structural Welding Code – Sheet Steel, ANSI AWS D1.3/D1.3M-
2008, American Welding Society, Miami, FL.
AWS (2005), Structural Welding Code–Seismic Supplement, ANSI AWS
D1.8/D1.8M, American Welding Society, Miami, FL.
Galambos, T.V. and Somers, R.E. (1988), “Report of Tensile Tests of Web to
Lower Chord Weld Connections in Longspan Joist,” Conducted for the
Research Committee of the Steel Joist Institute, September 30. 1988.
Green, William S. Hama, Mahmoud F. and McCauley, Roy B. (1958) “The Effect
of Porosity on Mild Steel Welds,” American Welding Society Journal
Supplement, May, pp 266s-209s.
Note: This paper demonstrated with tensile tests of welded structural steel
that static tensile strength could be maintained with up to 7% volume
of X-ray porosity.
Key to Metals Steel, (2008), website http://www.key-to-steel.com.

57
Lamprecht, Richard M. (1977) “Investigation of Flare Vee and Flare Bevel
Welded Connections,” Master’s Thesis, Structural Division, Washington
University, St. Louis, MO, September 1977.
Note: This work indicated that undercut beyond the limits established for
heavy structural members did not significantly reduce the load
carrying ability of Open Web Joists, provided it was oriented parallel
to the principal stress.
Lamprecht, R.H. and Galambos, T.V. (1977), “Investigation of Flare Vee and
Flare Bevel Welded Connections,” Progress Report to the Steel Joist
Institute, Research Report No. 47, Structural Division, Civil Engineering
Department, Washington University, St. Louis, MO.
Miller, D.K., (2006), Welded Connections – A Primer for Engineers, Design
Guide 21, American Institute of Steel Construction, Inc., Chicago, IL.
Packer, J.A. and Frater, G.S. (2005), “Recommended Effective Throat Sizes for
Flare Groove Welds to HSS,” Engineering Journal, American Institute of
Steel Construction, Inc., First Quarter, Vol. 42, No. 1.
SJI (2005), 42nd Edition Catalog containing Standard Specifications, Load
Tables and Weight Tables for Steel Joists and Joist Girders: K-Series, LH-
Series, DLH-Series, Joist Girders, Steel Joist Institute, Myrtle Beach, SC.
SJI (2007), 1st Edition Composite Steel Joist Catalog containing Standard
Specifications, Weight Tables and Bridging Tables for Composite Steel
Joists, CJ-Series, Steel Joist Institute, Myrtle Beach, SC.
SJI (2008), Technical Digest No. 9, Handling and Erection of Steel Joists and
Joist Girders, Steel Joist Institute, Myrtle Beach, SC.

58
 APPENDIX A

SJI STANDARD SPECIFICATIONS -

WELDING REQUIREMENTS
(Applicable sections only are reprinted below; Text shown in Times Roman Italics
is not currently part of Specification)

STANDARD SPECIFICATION
FOR OPEN WEB STEEL JOISTS, K-SERIES
Adopted by the Steel Joist Institute November 4, 1985
Revised to November 10, 2003 - Effective March 01, 2005

SECTION 4.
DESIGN AND
MANUFACTURE
4.5 CONNECTIONS

(a) Methods

Joist connections and splices shall be made by attaching the members to one
another by arc or resistance welding or other accredited methods.

(1) Welded Connections

a) Selected welds shall be inspected visually by the manufacturer. Prior
to this inspection, weld slag shall be removed.
b) Cracks are not acceptable and shall be repaired.
c) Thorough fusion shall exist between weld and base metal for the
required design length of the weld; such fusion shall be verified by
visual inspection.
d) Unfilled weld craters shall not be included in the design length of the
weld.
e) Undercut shall not exceed 1/16 inch (2 millimeter) for welds oriented
parallel to the principal stress.

59
 f) The sum of surface (piping) porosity diameters shall not exceed 1/16
inch (2 millimeter) in any 1 inch (25 millimeter) of design weld length.
g) Weld spatter that does not interfere with paint coverage is acceptable.

(2) Welding Program

Manufacturers shall have a program for establishing weld procedures and
operator qualification, and for weld sampling and testing. (See Technical
Digest #8 - Welding of Open Web Steel Joists and Joist Girders.)

(3) Weld Inspection by Outside Agencies (See Section 5.12 of these

specifications)
The agency shall arrange for visual inspection to determine that welds
meet the acceptance standards of Section 4.5(a)(1) above. Ultrasonic,
X-Ray, and magnetic particle testing are inappropriate for joists due to the
configurations of the components and welds.

(b) Strength

(1) Joint Connections - Joint connections shall be capable of withstanding

forces due to an ultimate load equal to at least 1.35 times the LRFD, or
2.0 times the ASD load shown in the applicable Standard Load Table.

(2) Shop Splices – Splices may occur at any point in chord or web members.
Members containing a butt weld splice shall develop an ultimate tensile
force of at least 57 ksi (393 MPa) times the full design area of the chord or
web. The term “member” shall be defined as all component parts
comprising the chord or web, at the point of the splice.

(c) Eccentricity

Members connected at a joint shall have their centroidal axes meet at a point
if practical. Otherwise, due consideration shall be given to the effect of
eccentricity. In no case shall eccentricity of any web member at a joint
exceed ¾ of the over-all dimension, measured in the plane of the web, of the
largest member connected. The eccentricity of any web member shall be the
perpendicular distance from the centroidal axis of that web member to the
point on the centroidal axis of the chord which is vertically above or below the
intersection of the centroidal axes of the web members forming the joint.
Ends of joists shall be proportioned to resist bending produced by eccentricity
at the support.

60
(Applicable sections only are reprinted below; Text shown in Times Roman Italics
is not currently part of Specification)

STANDARD SPECIFICATION
FOR LONGSPAN STEEL JOISTS, LH-SERIES AND
DEEP LONGSPAN STEEL JOISTS, DLH-SERIES
Adopted by the Steel Joist Institute February 15, 1978
Revised to November 10, 2003 - Effective March 01, 2005

SECTION 103.
DESIGN AND
MANUFACTURE
103.2 DESIGN AND ALLOWABLE STRESSES

(d) Weld Strength:

Shear at throat of fillet welds:
Nominal Shear Stress = Fnw = 0.6Fexx (103.2-14)

LRFD: φw = 0.75
Design Shear Strength =
φRn = φwFnw A = 0.45Fexx A (103.2-15)

ASD: Ωw = 2.0
Allowable Shear Strength =
Rn/Ωw = FnwA/Ωw = 0.3Fexx A (103.2-16)
A = effective throat area

Made with E70 series electrodes or F7XX-EXXX flux-electrode

combinations.
Fexx = 70 ksi (483 MPa)

Made with E60 series electrodes or F6XX-EXXX flux-electrode

combinations.
Fexx = 60 ksi (414 MPa)

61
 Tension or compression on groove or butt welds shall be the same as those
specified for the connected material.

103.5 CONNECTIONS

(a) Methods

Joist connections and splices shall be made by attaching the members to one
another by arc or resistance welding or other accredited methods.

(1) Welded Connections

a) Selected welds shall be inspected visually by the manufacturer. Prior
to this inspection, weld slag shall be removed.
b) Cracks are not acceptable and shall be repaired.
c) Thorough fusion shall exist between weld and base metal for the
required design length of the weld; such fusion shall be verified by
visual inspection.
d) Unfilled weld craters shall not be included in the design length of the
weld.
e) Undercut shall not exceed 1/16 inch (2 millimeter) for welds oriented
parallel to the principal stress.
f) The sum of surface (piping) porosity diameters shall not exceed 1/16
inch (2 millimeter) in any 1 inch (25 millimeter) of design weld length.
g) Weld spatter that does not interfere with paint coverage is acceptable.

(2) Welding Program

Manufacturers shall have a program for establishing weld procedures and
operator qualification, and for weld sampling and testing. (See Technical
Digest #8 - Welding of Open Web Steel Joists and Joist Girders.)

(3) Weld Inspection by Outside Agencies (See Section 104.13 of this

specification).
The agency shall arrange for visual inspection to determine that welds
meet the acceptance standards of Section 103.5(a)(1). Ultrasonic, X-Ray,
and magnetic particle testing are inappropriate for joists due to the
configurations of the components and welds.

(b) Strength

(1) Joint Connections - Joint connections shall develop the maximum force
due to any of the design loads, but not less than 50 percent of the strength
of the member in tension or compression, whichever force is the
controlling factor in the selection of the member.

62
 (2) Shop Splices – Shop splices may occur at any point in chord or web
members. Splices shall be designed for the member force but not less
than 50 percent of the member strength. Members containing a butt weld
splice shall develop an ultimate tensile force of at least 57 ksi (393 MPa)
times the full design area of the chord or web. The term "member" shall
be defined as all component parts comprising the chord or web, at the
point of splice.

(c) Field Splices

Field Splices shall be designed by the manufacturer and may be either bolted
or welded. Splices shall be designed for the member force, but not less than
50 percent of the member strength.

63
(Applicable sections only are reprinted below; Text shown in Times Roman Italics
is not currently part of Specification)

STANDARD SPECIFICATION
FOR JOIST GIRDERS
Adopted by the Steel Joist Institute November 4, 1985
Revised to November 10, 2003 - Effective March 01, 2005

SECTION 1003.
DESIGN AND
MANUFACTURE
1003.2 DESIGN AND ALLOWABLE STRESSES

(d) Weld Strength:

Shear at throat of fillet welds:
Nominal Shear Stress = Fnw = 0.6Fexx (1003.2-14)

LRFD: φw = 0.75
Design Shear Strength =
φRn = φwFnw A = 0.45Fexx A (1003.2-15)

ASD: Ωw = 2.0
Allowable Shear Strength =
Rn/Ωw = FnwA/Ωw = 0.3Fexx A (1003.2-16)
A = effective throat area

Made with E70 series electrodes or F7XX-EXXX flux-electrode

combinations.
Fexx = 70 ksi (483 MPa)

Made with E60 series electrodes or F6XX-EXXX flux-electrode

combinations.
Fexx = 60 ksi (414 MPa)

Tension or compression on groove or butt welds shall be the same as those

specified for the connected material.

64
1003.5 CONNECTIONS

(a) Methods

Joist connections and splices shall be made by attaching the members to one
another by arc or resistance welding or other accredited methods.

(1) Welded Connections

a) Selected welds shall be inspected visually by the manufacturer. Prior

to this inspection, weld slag shall be removed.
b) Cracks are not acceptable and shall be repaired.
c) Thorough fusion shall exist between weld and base metal for the
required design length of the weld; such fusion shall be verified by
visual inspection.
d) Unfilled weld craters shall not be included in the design length of the
weld.
e) Undercut shall not exceed 1/16 inch (2 millimeter) for welds oriented
parallel to the principal stress.
f) The sum of surface (piping) porosity diameters shall not exceed 1/16
inch (2 millimeter) in any 1 inch (25 millimeter) of design weld length.
g) Weld spatter that does not interfere with paint coverage is acceptable.

(2) Welding Program

Manufacturers shall have a program for establishing weld procedures and

operator qualification, and for weld sampling and testing. (See Technical
Digest #8 - Welding of Open Web Steel Joists and Joist Girders.)

(3) Weld Inspection by Outside Agencies (See Section 1004.10 of this

specification).
The agency shall arrange for visual inspection to determine that welds
meet the acceptance standards of Section 1003.5(a)(1). Ultrasonic,
X-Ray, and magnetic particle testing are inappropriate for joists due to the
configurations of the components and welds.

(b) Strength

(1) Joint Connections - Joint connections shall develop the maximum force
due to any of the design loads, but not less than 50 percent of the strength
of the member in tension or compression, whichever force is the
controlling factor in the selection of the member.

(2) Shop Splices – Shop splices may occur at any point in chord or web
members. Splices shall be designed for the member force but not less

65
 than 50 percent of the member strength. Members containing a butt weld
splice shall develop an ultimate tensile force of at least 57 ksi (393 MPa)
times the full design area of the chord or web. The term "member" shall
be defined as all component parts comprising the chord or web, at the
point of splice.

(c) Field Splices

Field Splices shall be designed by the manufacturer and may be either bolted
or welded. Splices shall be designed for the member force, but not less than
50 percent of the member strength.

66
(Applicable sections only are reprinted below; Text shown in Times Roman Italics
is not currently part of Specification)

STANDARD SPECIFICATION
FOR COMPOSITE STEEL JOISTS, CJ-SERIES
Adopted by the Steel Joist Institute May 10, 2006
- Effective May 10, 2006

SECTION 103.
DESIGN AND
MANUFACTURE
103.2 DESIGN AND ALLOWABLE STRESSES

(d) Weld Design Strength: φ w = 0.75

Fn = 0.6 Fexx (103.3-14)

φFn = φw (0.6 Fexx ) (103.3-15)

Design Shear Strength = φR n = φFn A = φ w (0.6 Fexx ) A (103.3-16)

Where, A = effective throat area

Made with E70 series electrodes or F7XX-EXXX flux-electrode

combinations
Fexx = 70 ksi (483 MPa)

Made with E60 series electrodes or F6XX-EXXX flux-electrode

combinations
Fexx = 60 ksi (414 MPa)

Tension or compression on groove or butt welds shall be the same as those

specified for the connected material.

103.6 CONNECTIONS

(a) Methods
Joint connections and splices shall be made by attaching the members to
one another by arc or resistance welding or other accredited methods.

67
 (1) Welded Connections
a) Selected welds shall be inspected visually by the manufacturer.
Prior to this inspection, weld slag shall be removed.
b) Cracks are not acceptable and shall be repaired.
c) Thorough fusion shall exist between layers of weld metal and
base metal for the required design length of the weld; such
fusion shall be verified by visual inspection.
d) Unfilled weld craters shall not be included in the design length of
the weld.
e) Undercut shall not exceed 1/16 inch (2 millimeters) for welds
oriented parallel to the principal stress.
f) The sum of surface (piping) porosity diameters shall not exceed
1/16 inch (2 millimeters) in any 1 inch (25 millimeters) of design
weld length.
g) Weld splatter is acceptable.

(2) Welding Program

Manufacturers shall have a program for establishing weld
procedures and operator qualification and for weld sampling and
testing (See Technical Digest #8 - Welding of Open Web Steel Joists and
Joist Girders.)

(3) Weld inspection by Outside Agencies (See Section 104.13 of this

specification).
The agency shall arrange for visual inspection to determine that
welds meet the acceptance standards of Section 103.6(a)(1)
above. Ultrasonic, X-Ray, and magnetic particle testing are
inappropriate for joists due to the configurations of the components
and welds.

(b) Strength
(1) Joint Connections - Joint connections shall develop the maximum
force due to any of the design loads, but not less than 50 percent of
the nominal strength of the member in tension or compression,
whichever force is the controlling factor in the selection of the
member.
(2) Shop Splices - Shop splices may occur at any point in chord or web
members. Splices shall be designed for the member force but not
less than 50 percent of the nominal member strength. Members
containing a butt weld splice shall develop an ultimate tensile force
of at least 57 ksi (390 MPa) times the full design area of the chord
or web. The term “member” shall be defined as all component
parts, comprising the chord or web, at the point of splice.

68
(c) Field Splices
Field splices shall be designed by the manufacturer and may be either
bolted or welded. Splices shall be designed for the member force, but not
less than 50 percent of the nominal member strength.

(d) Shear Studs

Shear studs, after installation, shall extend not less than 1 1/2 in. (38 mm)
above the top of the steel deck and there shall be at least 1/2 in. (13 mm) of
concrete cover above the top of the installed studs.

69
 APPENDIX B

SJI WELDING REQUIREMENTS: H-,

LH-, DLH-SERIES JOISTS AND JOIST
GIRDERS – HISTORICAL DATA
This Appendix on Welding of Open Web Steel Joists is in two parts:
PART 1: A description of load tests of production joists welded by each of the
five different welding processes used by SJI members, plus the tests on the
welded joints. These joists were loaded to the 1.65 times standard SJI design
load, after which individual welded joints were cut out and retested to destruction
in punch-through shear tests (see Fig. B.1).

PART 2: A determination of the relationship between flare bevel weld throat

dimensions and web bar diameters from the examination of joint weld
specimens. These joints were made by three arc-welding processes: flux core,
CO2 shielded, and Argon-CO2 shielded, using the small diameter electrodes
(0.045 in. to 0.094 in.) characteristic of these semi-automatic welding processes.
These relatively small diameters facilitate weld penetration into flare weld
grooves. Figure B.2 illustrates the advantage in access and penetration of small
diameter electrodes.

Almost all open web steel joist welds are made by arc welding in one pass, at
high speed, and for short lengths. Figure B.2 shows schematically the
configuration of the joist parts and welds and the arc weld nomenclature.
Typically, joists are welded by short flare welds deposited at approximately 3 ft.
per minute. The welders develop specialized skills for this operation, which are
far different from conventional “heavy” structural welding.

Joist manufacturers are sometimes asked why open web steel joists aren’t
manufactured to “heavy” structural steel specifications. The reasons for not
using such specifications are:
(a) Through many years of experience, an acceptable welding procedure
has been developed which is suitable and efficient for steel joists.
(b) The cost of welding and inspection would be greatly increased.
(c) Steel joists are intended to be used only under static loading and thus
many of the criteria for heavy steel welding, which have their basis in
fatigue design, do not apply

70
 Figure B.1 Schematic of Punch Through Shear Test

Therefore, the SJI decided to develop welding criteria appropriate to the

manufacture of open web steel joists, based on the long established “fitness for
service” SJI performance specification.

A perspective on the relative importance of weld quality among the various

problems that might be encountered in welded joist structures can be obtained
from service experience. Most of the reported failure problems were caused by
one or more of the following:
Field Construction:
• Overloads in shipping and handling, e.g., unloading by dropping off the
truck.
• Erection overloads – failure to install bridging; or overloading bare joists
with construction material.

In service:
• Clogged drains, with consequent water-ponding overloads.
• Snow and ice overloads e.g., the Chicago blizzards in the winter of
1978-79, in which many ice and snow loads far exceeded the Building
Code design values.
• Adding concentrated loads (equipment, monorails, etc.) not considered
in the original design.

71
 Figure B.2 Flare Groove Welds

Weld imperfections, while important, are only one of many possible problems.
Nevertheless, every effort is being made to reduce weld imperfections that
jeopardize serviceability—that is one of the objectives of this report.

A survey of SJI members revealed that five different welding processes were
being used to manufacture open web steel joists. To study the relationship
among welding conditions, weld quality, and load carrying ability, a test program
was organized to sample routine shop production joists for each welding process
employed in joist manufacture. These joists were subjected to visual inspection,
loading test, and metallographic examination (cross sectioning), as outlined in
what follows. The program included the examination of 13 joists, ranging in size
from 10H3 at a span of 15 feet to a 30H11 at a span of 53 feet. The numbers
preceding the letter H in the SJI joist designations indicate joist depth in inches,
while those following the H indicate the relative sizes of joist members (larger
numbers, larger members).

72
The welding conditions and weld sizes were observed and recorded in the joist
plants for correlation with load test results. The data obtained in this way
provided a basis for the proposed “Recommended Code of Standard Practice for
Flare Bevel, Flare Vee and Fillet Welds in Open Web Steel Joists” at the end of
this digest.

SCOPE OF TEST WORK

Joist Series Manufactured

There are four series (depth and span types) of joists covered by Steel Joist
Institute Standard Specifications (SJI 1982):
1. Open Web, H-Series
2. Longspan, LH-Series
3. Deep Longspan, DLH-Series
4. Joist Girders

The letter H denotes high strength; for open web steel joists, this indicates that a
nominal 50 ksi minimum yield strength steel is used in the chords and either a 36
ksi or 50 ksi minimum yield strength steel is used in the webs.

The tests reported here involved only the open web steel joists, H-Series; the
chords were either rolled angles or cold-formed sections, and the webs were
solid rounds in a Warren truss configuration. The nominal depth and span size
range for open web steel joists, H-Series is listed in Table C-1.

Table B-1 Size Range for H-Series Joists

Depth Maximum Span
From 8 inches 16 feet
To 30 inches 60 feet

73
 PART 1
A. STRUCTURAL STRENGTH TESTS

1. Joists Tested
Table B-2 lists the depth and span of the H-Series Open Web Joists tested.
The span for Specimens 3a, b, and c was determined by selecting, for a given
depth, the longest span for which shear force governs the design. The other
specimens were selected to represent different welding processes. All joists
were sampled from shop production.

Table B-2 Joists Tested—Description

Number of Dimensions Members
SJI Joist Joist
Specimen Joists Depth, Span,
Designation Mfg. Code Chord Web*
Tested inches feet
1 14H4 1 A 14 21 Cold-formed
2 10H4 1 B 10 16’-8” Angles
3a 10H3 3 C 10 15 Angles
3b 18H6 3 C 18 26 Angles
3c 30H11 3 C 30 53 Angles
4 14H3 1 D 14 22 Angles
5 22H9 1 E 22 25 Angles
* All webs were round bars

Specimens 1, 2, 4 and 5 were load tested at Washington University under the

supervision of Dr. T. V. Galambos. A loading frame on the lab floor placed
the joist in a horizontal position, where it was loaded by hydraulic jacks as
illustrated in Figure B.3. The joist load tests for Specimens 3a, b, and c were
performed in the manufacturer’s plant in a special multiple panel-point
hydraulic jack loading machine which could accommodate the longest span
tested (53 feet), see Figure B.4. All weld joint tests were performed at
Washington University under Dr. Galambos’ supervision.

Each of the Specimens 3a, b, and c in Table B-2 were tested in triplicate in
the first phase of this work, for a total of 9 joists.

Because of the observed consistency in properties and results, the remaining

tests were conducted on a single joist for each of the four remaining weld
processes.

74
 Figure B.3 Full Scale Joist Test at Washington University

2. Welding Processes Used

The various member companies of the Steel Joist Institute employ one of
more of five welding processes in the manufacture of open web steel joists:
(A) Arc processes:
(1) Flux shielded
(a) Shielded metal arc (SMAW). Manual “stick”.
(b) Flux cored arc (FCAW). Hand held gun, no external gas
shield.
(2) Gas shielded metal arc (GMAW). Hand held gun, external gas
shield surrounds arc.
(a) Shielding gas 100% CO2.
(b) Shielding gas 75% argon, 25% CO2.
(B) Electric resistance process:
Resistance projection. A projection on the chord angles, pressed
against the web, localizes a high electric current; the resistance heats
the parts to make the welds.

Table B-3 contains a summary of the weld processes used and the number of
test made. The weld processes are those used on the entire joist, except for

75
 Specimens 4 and 5, which used a flux core welding process (FCAW) for the
end web.

Figure B.4 Test of 30H11 Joist at 1.65 Times Design Load

Table B-3 Joist Welding Processes and Number Tested

Number Tested
SJI Joist Electrode
Specimen Mfg. Weld Process Welded
designation Dia., Inch Joists
Code Joints
1 14H4 A SMAW “stick” 5/32 1 7
2 10H4 B Flux core 3/32 1 8
3a 10H3 {GMAW: 75% 3 26
3b 18H6 C Argon, 25% 0.052 3 25
3c 30H11 CO2} 3 25
Gas metal arc
4 14H3 D 0.045 1 12
CO2
Resistance
5 22H9 E 0.052 1 9
projection
13 112

76
 Section Polished and Etched for Measurement of Weld Throat

Nomenclature for Measurement of Throat and Face

Figure B.5 Flare Bevel Weld

3. Design of Test Program

One or more open web steel joists welded by each of the five welding
processes in commercial use by SJI members were sampled from shop
production for testing. Each joist was first:
• Tested to 1.65 times design load by means of concentrated loads at each
panel point.
Then:
• Welded joints were cut out for (a) punch-through shear tests (see Fig.
B.1), and (b) cross-sectioning for metallographic soundness and hardness
tests (see Figs. B.5 and B.6).

In this way, the welds were tested under the non-destructive forces created
by the joist overload of 1.65 times the design load, plus the higher forces from
the punch-through shear test, to failure. This permitted evaluation of weld
quality based on load carrying ability. Put another way, the amount of weld
quality needed to provide “fitness for service.”

77
B. TEST RESULTS

1. Test Loading of Joists

Test loading at 1.65 times design load resulted in no failures, no distress of
members, nor any excessive permanent deflection.

2. Punch-Through Shear Tests of Welded Joints

Table B-4 presents a summary of the punch-through shear tests of welded
web-to-chord joints. Specimen 1 had cold-formed chords which presented a
testing problem. It was difficult to support and grip the light gage chord in
proper alignment while simultaneously gripping the round web. It was
decided to base the evaluation upon the joist load test plus sectioning.

Specimens 2, 4 and 5 in Table B-4 show individual weld joint tests. For
Specimens 3a, b, and c, the average of 3 corresponding weld joints from 3
separate joists is given.

For simplicity, punch-through shear safety factors for all welded joints were
calculated using the SJI Maximum End Reaction (SJI 1982), not the usual
calculated shear which decreases along the joist from the ends to the center
of the span.

Using this more severe criterion, the safety factors for all 53 weld punch-
through tests averaged 8, with a range from 3 (only two tests) to 13. The SJI
standard requires a minimum of 2; thus, the average of all 53 welds exceeded
the requirement by 300%. The lowest two tests exceeded the standard
requirement by 50%.

There was little difference in strength between the various weld processes for
comparable depth joists, which supports the diversified choice of weld
processes by different manufacturers.

There was an effect of joist size on weld joint safety factor, shown by
Specimens 3a, b, and c in Table B-4, all with Argon—CO2 gas shielded
welding. These tests covered the range of open web steel joist sizes. As
joist sizes increased, the average weld joint safety factor decreased from 11
for the 10H3 joist to 6 for the 18H6, and 5 for the 30H11. The reason for this
decrease is that the throat of the single pass welds does not increase as
rapidly as the size of the web and chord members. Of course, all safety
factors exceeded the minimum required weld strength.

78
 Table B-4 Weld Joint Punch-Through Shear Results
SJI Maximum.
Joist Mfg. Punch-Through
Specimen Joist End Reaction, Joint No.
Code Shear, lbs
lbs
Cold-Formed Chord – No Punch-
1 A 14H4 3500 Through Test Feasible
J3 20,300
J4 31,000
J5 29,000
2 B 10H4 2800 J12 30,000
J13 30,400
J14 37,400
3 29,100*
8 29,200*
3a C 10H3 2500 12 23,000*
16 24,600*
3 33,100*
8 23,100*
3b C 18H6 4800 12 36,500*
15 23,600*
3 44,600*
9 51,200*
3c C 30H11 8700 13 45,500*
16 41,900*
S2 20,500
S3 30,400
S4 32,400
4 D 14H3 3200 S13 29,800
S14 29,400
S15 29,200
K3 30,200
K4 30,200
5 E 22H9 6700 K13 30,300
K14 23,000**
K15 31,000
* Test in triplicate, from three joists. All others from one joist.
** Chord angle splice at this joint may have affected weld.

3. Joist Material Properties

Table B-5 shows a chemical analysis check of the chords and webs made
with a Jarrell-Ash vacuum spectrograph on a randomly selected weld cross-
section from each of the five manufacturers. The steels have small residual
quantities of Vanadium (V), Columbium (Cb) and Titanium (Ti) which could
benefit yield strength. The compositions are normal except for a high carbon
content of 0.35% in the chord of Specimen 1, and 0.31% in the web of
Specimen 4. Although this high carbon could contribute to weld cracking, no
cracks were found on the cross sections.

79
 Table B-5 Chemical Composition and Tensile Strength of Joist Members
Sample from Cross Section of Welded Joint Chemical Analysis (Vacuum Spectrograph)
Joist
Specimen Mfg. Carbon Manganese Phosphorous Sulphur Silicon
No. Part
1 A C19 Chord 0.35 1.07 0.011 0.028 0.21
Web 0.28 0.78 0.037 0.046 0.14
Tensile (ksi) from hardness: chord 76; web 87
Chord 0.23 0.88 0.010 0.025 0.08
2 B J7
Web 0.28 0.82 0.022 0.038 0.15
Tensile (ksi) from hardness: chord 68; web 79
Chord 0.21 0.83 0.020 0.038 0.19
3 C 12
Web 0.22 1.02 0.017 0.039 0.33
Tensile (ksi) from hardness: chord 65; web 60
Chord 0.22 0.70 0.017 0.041 0.31
4 D S18
Web 0.31 0.92 0.025 0.037 0.09
Tensile (ksi) from hardness: chord 66; web 85
Chord 0.23 0.76 0.023 0.056 0.11
5 E K12
Web 0.25 0.81 0.038 0.040 0.08
Tensile (ksi) from hardness: chord 73; web 87
Joist
Specimen Mfg. Vanadium Columbium Copper Aluminum Titanium
No. Part
1 A C19 Chord 0.004 0.005 0.03 0.093 0.002
Web 0.002 0.004 0.32 0.12 0.002
Tensile (ksi) from hardness (cont’d): weld 99; chord HAZ 121*
Chord 0.003 0.005 0.38 0.004 0.001
2 B J7
Web 0.009 0.005 0.43 0.004 0.002
Tensile (ksi) from hardness (cont’d): weld 100; chord HAZ 98*
Chord 0.003 0.004 0.31 0.004 0.001
3 C 12
Web 0.004 0.005 0.27 0.011 0.003
Tensile (ksi) from hardness (cont’d): weld 93; chord HAZ 88*
Chord 0.003 0.004 0.22 0.004 0.002
4 D S18
Web 0.003 0.005 0.51 0.004 0.001
Tensile (ksi) from hardness (cont’d): weld 83; chord HAZ 92*
Chord 0.014 0.008 0.34 0.004 0.001
5 E K12
Web 0.018 0.006 0.21 0.013 0.001
Tensile (ksi) from hardness (cont’d): weld 116; chord HAZ 120*

* Only hardest of web or chord weld heat affected zones (HAZ) shown.

80
 Figure B.6 Schematic of Hardness Readings
x indicates Rockwell “B” hardness measurement

Hardness measurements were made on the weld cross-sections (as shown in

Fig. B.6) for the chord, the web, the weld, and the weld heat affected zones
(HAZ). The hardness of steel is proportional to its tensile strength, and a
standard conversion chart was used to convert Rockwell “B” hardness
measurements to tensile strength. These converted tensile strengths are
shown on the third (bottom) line for each item in Table B-5. The tensile
strength of the weld is higher than the chord and the web in all but Specimen
4; there the weld strength is only slightly lower. The overall trend of the
tensile strength of joist members is shown in Figure B.7; chords averaged 74
ksi, webs 81 ksi, and welds 88 ksi.

4. Welding Conditions
The welding variables in production for the different arc welding processes
remained in a relatively narrow range that produced acceptable weld shear
strength. There was not a good correlation observed between weld heat
input per linear inch of weld and weld strength in the range of production
conditions tested.

5. Weld Cross Sections – Metallographic

Samplings of the weld cross-sections prepared for the metallographic
examination are shown in Figures B.8 through B.16, inclusive. Note the
ample throat dimensions. An important point to note is that all of the welds
pictured here withstood joist loadings up to 1.65 times the design value before
the sectioning or punch-through shear test. Any weld imperfections shown
did not prevent the joints from meeting the specified joist load capacity and
exceeding the weld punch-through tests by a wide margin.

81
 Figure B.7 Tensile Strength of Joist Materials
Converted from hardness measurements on cross-sections of welded joints

82
6. Weld Cross Sections – Throat Dimension
The “stick” welds in Figure B.8 had a throat dimension of 5/32 inch, slightly
greater than the thickness of the cold-formed “hat” section chord member.
The smallest throat dimension of the flux-cored welds in Figure B.9 was 1/4
inch, far exceeding the nominal 1/8 inch thickness of the angle leg. Figure
B.11 exhibits the characteristic ripple-marked surface of the CO2 welds; these
welds had excellent penetration and strength (Table B-4, Specimen 4). In
Figure B.12, the resistance projection weld shows fusion approximately 4
times the angle leg thickness. Welds made with small diameter electrodes
resulted in exceptionally good throat dimensions throughout the tests; this
contributed to the high weld strength.

C. SUMMARY AND CONCLUSIONS

Thirteen (13) open web steel joists, H-Series, representing the range of various
joist designations, were taken from production lines at joist plants using five
different weld processes. These joists were subjected to extensive tests
including:
(1) Loading all the joists to 1.65 times the design load. No distress, failure,
or permanent deflection resulted.
(2) After loading in (1) above, weld joints were cut out for punch-through
tests from these same joists. All of these exceeded by at least 50%
(average 300%) the SJI joint safety factor of 2 times the design load.
(3) Cross sections of welds were polished and etched; a few revealed
imperfections which did not affect load-carrying ability (see Figs. B.13
through B.16).

Figure B.8 Shielded Metal Arc “Stick” Welded Joint

14H4 Joist

83
 Figure B.9 Flux Core Welded Joint
10H4 Joist

Figure B.10 Argon-CO2 Welded Joint

30H11 Joist

84
 Figure B.11 CO2 Welded Joint
14H3 Joist

Figure B.12 Resistance Projection Welded Joint

22H9 Joist

85
Figure B.13 Argon-CO2 Welded Joint Showing Imperfections

Figure B.14 Argon-CO2 Welded Joint Showing Imperfections

86
Figure B.15 Punch-Through Test Showing Porosity and Poor Fusion.
This joist tested to 5.9 times the design end reaction.

Figure B.16 Punch-Through Test Showing Porosity and Poor Fusion.

87
 PART 2
A. RELATIONSHIP BETWEEN WELD STRENGTH AND WEB BAR DIAMETER

These tests involved only weld joints cut from production joists, which were
sectioned to measure the throat. Similar joints were punch-through tested to
determine shear strength (see Fig. B.1).

A prime variable in flare bevel welds is the web bar diameter. Experience
indicated that a relationship did exist between the bar diameter and the throat of
a flare bevel weld. Were a consistent relationship to exist, a minimum flare vee
welded joint strength could then be determined from any given web bar size.

The following materials were employed in the program for testing flare bevel joist
welds.

Table B-6 Materials for Flare Bevel Joist Weld Testing

Weld Electrode Web No. of Joist
Process Dia., inches Dia., inches Joists Designation
1 12H4
1 20H5
1/2, 5/8, 3/4, 7/8, 1 26H8
Flux cored arc 0.094 (3/32)
1 1 28H9
1 30H11
1 10H3
Argon-CO2 1/2, 5/8, 3/4, 7/8, 2 20H5
0.052 1 24H10
shielded arc 1
1 30H11
0.045
CO2 shielded arc 9/16 5 16H6
0.052

All of the joists used “H” grade 50,000 psi yield strength steel and nominal 70,000
psi electrodes. The joists were manufactured under common shop conditions in
three different plants regularly using the particular weld process. Summarizing,
the test included:
(a) The three arc welding processes which are used to manufacture the
great majority of open web steel joists in the United States.
(b) Fifteen open web steel joists representing a range of six web bar
diameters from 1/2 inch to 1 inch, manufactured under ordinary shop
conditions.

The weld throat dimension was measured on polished and etched cross-sections
of each of 120 welds, shown schematically in Figure B.5. Figure B.2 shows the
similar configuration of flare bevel and flare vee weld grooves, and the advantage
of small diameter electrodes in producing a larger (and therefore stronger) weld

88
throat. The tests were made on flare bevel welds, but similar conclusions would
also apply to flare vee welds.

The relationship between web diameter and weld throat is given in Figure B.17.
The average throat dimension for the three weld processes trends upward with
increase in web bar diameter. Moreover, the envelope line for the minimum weld
throat encountered in the 120 welds sectioned provides a basis for establishing a
calculated minimum weld strength to be expected in production.

These values are listed in Table B-7, in which weld shear strength is based on
the minimum throat dimension and allowable shear strength of 0.30 times the
nominal tensile strength (70,000 psi) of the AWS E70 class welding electrodes
used throughout the tests. Additionally, the stress in the members being joined
should not exceed 0.40 of the nominal yield of the base metal (1978 AISC
Specification 1.5.1.2); however, in the case of the single pass joist weld (and the
resistance weld), the shear paths provide greater areas than the weld throat and
therefore the 0.4Fy does not govern the joint strength (see Figure B.18).

Figure B.17

89
 Figure B.18

Table B-7 Web Diameter vs. Weld Throat vs. Weld Strength
Web Diameter Minimum Weld Throat Minimum Weld Strength*
inches inches kips/in.
1/2 0.170 3.57
9/16 0.178 3.73
5/8 0.185 3.88
11/16 0.192 4.04
3/4 0.200 4.20
13/16 0.208 4.36
7/8 0.215 4.52
15/16 0.222 4.67
1 0.230 4.83

T = 0.12 D + 0.11 (Equation of minimum weld throat line in Figure C.17)

S = 21T
Where,
D = Web Diameter, inches
T = Weld Throat, inches
S = Weld Strength, kips/in.

* Based on:
0.3 x Nominal Tensile of E70 Electrodes = 21 ksi

For a discussion on the “0.4Fy of the base metal” criteria, see last paragraph in
Part 2, Section A.

90
B. CONCLUSIONS

1. The weld strength of single pass flare bevel welds made under shop
production conditions with small diameter electrodes (0.045 inch to 0.094
inch) by the flux cored, CO2 shielded, and Argon-CO2 shielded welding
processes will equal or exceed the strength per unit length shown in Table
B-7 for joist web bar diameters from 1/2 to 1 inch.
2. The above values would also apply to flare vee welds (e.g., two round
bars in contact longitudinally; see Figs. 5.1 and B.2).
3. Load tests of joists and punch-through shear tests of resistance projection
welded joints showed strengths in excess of design requirements,
comparable to those of arc welded joints (see Table B-4, Specimen 5).

91