SSC-296

REVIEW OF FILLET WELD

STRENGTH PARAMETERS
FOR SHIPBUILDING

This document has been approved

for public release and sale; its
distribution is unlimited.

SHIP STRUCTURE COMMITTEE

1980
 SHIP STRUCTURE COMMITI'EE

The SHIP STRUCTURE COMMITTEE is constituted to prosecute a research

program to Improve the hull structures of ships and other marine structures
by an extension of knowledge pertaining to design, materials and methods of
construction.
RADM H. H. BELL (Chairman) Mr. M. PITYJN
C1ief, Office of Merchant Marine Assistant Administrator for
Safety Coercial Development
U. S. Coast Guard Maritime Administration
Mr. P. M. PALERMO Mr. R. B. KR4RL
Deputy Director, Chief, Branch of Marine Oil
Hull Group and Gas Operations
Naval Sea Systems Cor,v'nand U. S. Geological Survey
Mr. W. N. HANNA1' Mr. C. J. WHITESTONE
Vice President Chief Engineer
American Bureau of Shipping Military Sealift Convnand
LCDR T. R. ROBINSON, U. S. Coas t Guard (Secretary)
SHIP STRUCTURE SUBCOMMITTEE

The SHIP STRUCTURE SUBCOMMITFEE acts for the Ship Structure

Committee on technical matters by providing technical coordination for the
determination of goals and objectives of the program, and by evaluating
and interpreting the results in terms of structural design, construction and
operation.
U.S. COAST GUARD MILITARY SEALIFT COMMAND

CAPT R. L. BROWN MR. G. ASR

CDR J. C. CARD M)?. T. W. CRAP/IAN
CDR J. A. SANIAL, JR. MR. A. B. ST4VOVY (Chairman)
CDR W. M. SIMRSON, JR. MR. D. STEIN
NAVAL SEA SYSTEMS COMMAND
U. S. GEOLOGICAL SURVEY
Mr. R. CRIU MR. R. J. GIANGEPELLI
Mr. R. JOHNSON MR. J. GREGORY
Mr. J. B. O'BRIEN
MARITIME ADMINI STRATTON
AMERICAN BUREAU OF SHIPPING
MR. N. O. HAM)R
DR. D. LIII DR. W. MCLEAE
MR. I. L. STERN Mr. F. SE'IBOLD
Mr. M. TOUM4

NATIONAL ACA])EMY OF SCIENCES INTERNATIONAL SHIP STRUCTURES CONGRESS

SHIP RESEARCH COMMITTEE
Mr. S. G. STL4ZVSEN - LiaIson
Mr. O. H. OAKLEY - Liaison
Mr. R. W. RUMKE - Liaison AMERICAN IRON & STEEL INSTITUTE
ThE SOCIETY OF NAVAL ARCHITECTS Mr. R. H. STERNE - Liaison
& MARINE ENGINEERS
STATE UNIVERSITY OF NEW YORK MARITIME COLLEGE
Mr. N. O. HAM'ER - Liaison
Dr. W. R. PORTER - Liaison
WELDING RESEARCH COUNCIL
U. S. COAST GUARD ACADEMY
Mr. K. H. KOOPMAN - Liaison
CAPT W. C. NOLAN - Liaison
U. S. RCRANT MARINE ACADEMY
U. S. NAVAL ACADEMY
Dr. C. -B. KIM - Liaison
Dr. R. BRATTACHARÏYA - Liaison
 's _296

Member Agencies: r " Address Correspondence to:

United States Coast Guard Secretary, Ship Structure Committee
Naval Sea Systems Command U.S. Coast Guard Headquarters,(G-M/TP 13)
Military Sea/ift Command Ship Washington, D.C. 20593
Maritime Administration
United Stetes Geological Survey
Structure
American Bureau of 1ipping Committee
An Interagency Advisory Committee
Dedicated to Improving the Structure of Ships

APRIL 1980

S R-12 48

Fabrication methods are being closely examined by

the shipbuilding industry in an attempt to hold down or
reduce shipbuilding costs. An examination of the fabri-
cating procedures disclosed that as much as 30 percent of
the vessel construction man-hours are devoted to welding.
A further analysis indicated that 75 percent of the welded
joints were fillet welded.

Inasmuch as the requirements of fillet weld sizes

have not been revised for many years, the Ship Structure
Committee considered a review and analysis of current
marine fillet weld requirements might provide an oppor-
tunity to reduce the sizes. This report reviews the fillet
weld requirements of the various ship classification
societies, presents a developmental procedure for a rational
analysis of required weld strength and makes recommendations
for further research.

HeiTFT1-i. Bell
Rear Admiral , U.S. Coast Guard
Chairman, Ship Structure Committee
 Technical Report Documentation Page
1. Report No. 2. C.overnrsent Accesson N .. 3. Recrp en? s Catalog No

SSC- 296

4. T.tle and Sub?, tIe - 5. Report Dote

FFRRIJARY 19.80
6. PrrrfOrrrrng Orgarrzton Code
REVIEW OF FILLET WELD STRENGTH
PARAMETERS FOR SHIPBUILDING
8. Perfornrrg Orgorr zohorr Report No
7. Author's)
C-L Tsai , K. Itoga, A. P. Malliris,
W. C. McCabe and K. Masubuchi
9. Perform,ng Organi zOtion Nome and Addcss 10 Work Unit No. (TRAIS)

Massachusetts Institute of Technology

11. Contract or Grant No
Department of Ocean Engineering
Cambridge, MA 02139 DUT-Cí-714F5-A
13. Type ) Report and Period Covered
12. Sponsoring Agency No,rre and Address
U. S. Coast Guard FINAL
Office of Merchant Marine Safety
14. Sponsrlrr Agency Code
Washington, D.C. 20593

15. Supplementary Notes

Performed under CG contract for the interagency Ship Structure Committee

16. Abstract

This report presents the results of a review of the current

fillet weld specifications of the various classification so-
cieties and a developmental procedure for analyzing these rules
by a rational method for estabilishing required weld size, and
recommendations for further research.
The results indicated large deviations among rules which
relate the fillet weld size to the thickness of the base plate.
The required fillet weld size by the most conservative rule may
be two times that required by the most liberal rule. There appears
to be a need for reviewing these rules more closely by analytical
means.
A computer program, named "Automatic Dynamic Incremental
Nonlinear Analysis (ADINA)', was used to determine the stress
distributions in the welds of tee-joints under simple tension
acting along the edges of the flange. This proqram could be used
in determining the minimum weld sizes. A detailed explanation of
the rationale of using "ADINA" program or similar FEM programs
is presented in this report.

17. Key Words 18. Distribution Statement

STATIC STRENGTH WELDING- Document is available to the U.S. Public
FATIGUE STRENGTH CRACKING through the National Technical
RESIDUAL STRESS WELDING - Information Service, Springfield,
WELD DEFECT PROCESS VA 22161.
INSPECTION
19. Security Classif. (of this report( 20. Security Clasif. (of this poge( 21. N0. of Pages 22. Price

UNCLASSIFIED 57
UNCLASSIFIED
Form DOT F 7OO.7 (B-72) Reproduction 0f completed poge authorized
111-
 METRIC CONVERSION FACTORS

Approximate Conversions to Metric Measures

Approximate Conversions 1mm Metric Measures
rl
Symbol When Yes Know Multiply by T. Find Symbil
Symbol When You Know Multiply by To Find Symbef

LENGTH
e
ru
LENGTH
rrrittrnnselcrs 0.04 erctl,,s IA
cnl centimeters 0.4 rirctrr,s 5
iii inches 2.5 Ceitininler! nIelen 3.3 leer Il
ft feet 30 centitraiters cm meters 1.1 yards yd
yd yards 0.9 meters re hrn kitrersetees 0.6 011(05 mr
n'i m,le 1.6 kilometers km

AREA w AREA

cmi' square centimeters 0.16 square inches

in2 cullare inches 6.5 square cent,mnlers cm2 m2 square nieterS 1.2 square y4rrls
h2 square leer 0.09 square meters yd
ni7
km2 squalo kilniritters 0.4 squali, trilles mr2
yd2 square yards 0.8 squalO litions n' Ira bailaren (10,000 in2( 2.6 acres
lili2 square miles 2.6 squire kilconetors km2
acres 0.4 InicIares ha
n
MASS (weight) MASS (weigh!)
el
o: ounces 28 graIns g g grIsas 0.035 ouricen o:
IIi poruids 0.45 kiluijranus kg kg kilograms 2.2 poiinils lb
short torts 0,9 10011es I t tonnes (1000 kg) 1.1 lie.r toits
2000 lb)
D
VOLUME
VOLUMI
tsp teaspoOns 5 t,illitrlers rit nil cr11) ilitørl 0.03 fIord rriirrcos ft 0!
lbsp tablespoons 16 milliliters nil w t hIers 2.1 purIn pl
Il or huid ounces 30 ertlititern ml liters 1,06 quarts ql
c Cups 0.24 liters I t, titeos 0.26 gallons Oat
pl pints 0.41 liters ini3 Cubic meters 36 cubrc feet tt
ql quarts 0,95 liters m3 cubic meters 1.3 cabre yards yd1
gal gatls 3.8 liters
lt cubic feet 0.03 ciibi renters or3
5d3 cubic yards 0.76 cubic nielare n,3 TEMPERATURE (exact)
TEMPERATURE (exact)
nc Cetsiun 9/S liben Fahrenheit
len.taenature add 32) temperature
Op fahrenheit 5/9 (allen Celsius n
temperature subtracting telsifleralura
321
r. 32 906 212
-40 0 (40 00 I tOO tOO 200
1 r. n
I J
2.51 r'n.il.Iint. In nie. .,.ao, r'orin..r%nn,n oni n.m deI,,Pod Illi,, sao NOS M, P.4,1. 2.k. e..0 y j t ? cr.0 r nj .r
UtS ,il vensirru Ciel 5$l.,i,.iirn5. Pria 57.25, 5V Catatan No, Cr310796. -40 '20 0 20 f40 00 00 tOO
 CONTENTS

i
Introduction
i
1.1 Background
1.2 Outline of the Study 2

Literature Survey
2.1 Statistics on Ship Hull Damage Related to
Weld Defects
Review of Static Strength of Fillet Welds 5
2.2 . . .

2.3 Review of Fatigue Strength of Fillet Welds . . 10

Review of Standards 14

3.1 Review of Fillet Weld Specifications 14

3.2 Corrosion Considerations 16

3.3 Fabrication Limits 22

3.4 Review of U.S. Navy Welding Specifications . . 22

Development of Analytical Method 29

4.1 Analytical Method 29

41.2 Method for Determining Minimum Fillet

Weld Sizes 29

4.3 Mathematical Modeling and General Yielding

Criterion 32

4.4 Numerical Example of the Effect of In-Plane

Tensile Stress in Bottom Shell Plating on Fillet
Weld Strength 38

Conclusions and Recommendations for Further Research 42

5.1 Conclusions 42

5.2 Recommended Further Research 42

46
References
Appendix I List of Literature on Fillet Welds, 1943-1977 48

-V-
- NOTES -

-vi-
 -1-

1. INTRODUCTION

1.1 Background

Fillet welds are extensively used in ship structures. In

a typical ship hull construction, about 75% of the welds are
fillet welds. (1) This is because a ship hull is essentially
composed of a number of panel structures. A typical panel
structure is composed of a plate and transverse and longitudinal
stiffeners. These stiffeners are usually fillet welded to the
panel. For example, a 50,000 deadweight ton cargo ship, of
which the hull weight is approximately 15,OUu tons, has approxi-
mately 7 x l0 feet of fillet welds, of which the weld metal
weighs approximately 60 tons.
An overriding concern by ship designers and fabricators
over the years has been to make sure these fillet welds are
strong enough. Although many efforts have been made to reduce
the weight of the ship structure by reducing thickness and
dimensions of structural members, little attention has been placed
on reducing the size of fillet welds, weight of which represents
only a fraction of the structural weight. Rules on the size of
fillet welds in ship structures have remained virtually unchanged
for many years. It is quite possible that current specifications
on fillet welds are too conservative.

The reduction of fillet sizes can have a significant impact

on construction cost by reducing construction time, labor cost,
the weight of welding consumables, etc. For example, 20% red-
uction in the fillet leg size will result in 36% reduction in
the amount of the weld metal. The welding and assembly of ship
hulls requires approximately the same number of manhours, and
these two functions combined amount to about 60% of the total
manhours for the completion of the hull structure. (2) This
indicates that the welding operation accounts for about 30% of
the labor cost in planning and constructing ship hulls. If we
look at the total linear measure of the welded fillet joints
employed in ship construction (75%), the labor cost in fillet
welding is about a quarter of the total labor cost for constructing
a ship's hull.

Reduction of the fillet size will also result in reduction

of weld distortion. The reduction of out-of-plane distortion
may result in an increase in buckling strength when the panel
is subjected to compressive loading."3)

This project was initiated with an ultimate goal of finding

whether sizes of fillet welds could be reduced without affecting
the structural integrity of a ship. More specifically, the
 -2-

original objective was to recommend updated fillet weld re-

quirements for domestic ship application by reviewing the
development of current marine fillet weld requirements and
available test data.

1.2 Outline of the Study

The one-year study included the following tasks:

Literature survey,
Review of welding standards,
Contact with experts,
Analysis, and
Recommendations.
There have been many publications on various aspects of
ship structural analysis,studies on the overall strength of a
ship hull, and studies to determine stress distributions in various
structural members. In fact, large-scale finite-element
methods (FEM) have been developed by various research groups
including ship classification societies in various countries
for computing stress distributions in various structural members
of a ship hull. However, no published articles specifically
discuss stress distributions in fillet welds.

After searching for suitable techniques for analyzing

the fillet weld strength, the finite-element method was found
to be a reliable tool and probably one of few techniques which
could fulfill the needs of this project. Therefore, efforts
were made to develop a computer application of an existing
program, named "ADINA" (Automatic Dynamic Incremental Nonlinear
Analysis) which was developed by Professor K.J. Bathe in the
Department of Mechanical Engineering at M.I.T.
 -3-

2. LITERATURE SURVEY
A literature search used the following key words:
Static Strength
Fatigue Strength
Residual Stress
Weld Defect
Inspection
Welding Cracking
Welding Process
to generate the 81 papers that were surveyed in Appendix 1.
A review of assumptions and conclusions of the major past
contributions (9 papers) in improving the understanding of fillet
weld strength is summarized in Table 2.1
The survey showed:
The requirements for fillet weld size used in the codes of
various classification societies were based on equivalent shear
loads between a riveted and a welded structure.
The first attempt to compe experiment with theory was done
by Vreedenburgh in l954. (11)
The most recent attempt was byKato in 1974, using a finite
element method (FEM), but again with some simplifications.
An accurate analysis has not yet been done.
Fillet welds are very strong when the current requirements are
applied.
The statistics indicate no fillet failures and the weld size
relates only weakly to cracking at the toe of the fillet.
More papers discuss fatìgue strength than static strength.
Fatigue strength is more critical than static strength in fillet
welds.
Contact angle between the base plate and the weld surface, welding
defects such as undercut or cracks near the fillet toes in the
base metal, and root gap are factors contributing to reduction
of fatigue resistance and a fillet weld failure.
2.1 Statistics on Ship Hull Damage Related to Weld Defects
A study on hull damage related to weld defects has been
carried out by Nippon Kaiji Kyokai.(14) The study dealt with
general structure damages of four types of ships: tankers,
ore carriers, containers and general cargo ships. Out of 1200
surveyed ships, cracks in shell or strength deck plating were
 TABLE 2. 1 SUMMARY OF LITERATURE SURVEY

NAME YEAR SUBJECT ASSUMPTIONS CONCLUSIONS

Vreedenburgh 1954 Static strength (Experiments) Design shouldbe based onan experimentally
derived envelope of weldstrength. Reject
theoretical approaches since they didn' t agree
with experiments. Introduce empirical
coefficients to modify theoretical results.
Macfarlane Fatigue of (Experiments)
65 The fatigue strength of the transverse fillet
Harrjsrn transverse welds is influenced by the relative sizes of
(5) fillet welds the main and cover plates.
Swanell 1968 Static strength Uniform shear Effect of joint stiffness and load application
of longitudinal
UJ on the shearing intensity - Toe displacement
fillet welds relationships.
Report of the Fatigue (Experiments)
Welding Inst. 1968 Considerable increase in fatigue strength of
Research fillet welded joints is reported when theyhave
Laboratories been either ground or hammer peened. For low
(7)
cycle fatigue use grinding and for high cycle
fatigue use peening.
Solumsmoen 1969 Fatigue (Experiments) Welded specimens in mild arid high tensile steel
(8) can be represented, approximately, by the same
S-N curve.
Butler, Kulak 1971 Static strength (Experiments) Transverse welds show about 44% strength
(9)
increase over longitudinal welds but show a
decrease in deformation capacity.
Clark 1971 Static strength (Experiments)
(10) Strength of long joints and groups of fillet
welds under eccentric loading is reported.
Kat o Static strength 1)Direct stress on From elastic solution, transverse fillet
of transverse tle ensi welds
race are 46% stronger than lonqitudinal filletwelds
fillet welds O 4ne îe is
(11) of the same size and lenqth. Failure always
(F.E.M. analysis) occurred at the root of Ehe fillet.
2)Breaking will
occur when the
shear stress ata
point of the fillet
weld is:
Tmax =
e re
= tensj.le
strength of the
weicleQmetal.
Maddox 1975 Fatigue Theory and
(12) experiments Agreement between theory and experiments
 -5-

found in 101 ships. Almost all of these were fatigue-crack

initiated from the toe of fillet joint connecting internals to
and
shell or deck plating, transverse members to shell plating
horizontal girders to bulkhead plating.

Other statistical studies made recently in the general area

following
of the hull structural damages were reviewed and the
seven critical joints were identified:
Internals (longitudinal members) to shell plating.
Internals (longitudinal members) to strength-deck plating.
Primary transverse members to shell plating.
Horizontal girders to bulkhead plating.
Double bottom floor to inner bottom.
Double bottom girders to shell and inner bottom.
Face plates on deep web haunches.

These critical joints are also sensitive to fillet weld

defects according to the statitica1 studies conducted by (15)
Nippon Kaiji Kyokai (Japan) (l4).f
(16)
Newport News Shipyard (USA)
and Prof. Antoniou (Greece)

2.2 Review of Static Strength of Fillet Welds

In order to study the effect of the direction of applied
and Kulak
load on the strength of fillet welded joints, Butler
conducted tests and analyzed resulting data.
the weld
The tests were conducted in four groups, each with 30,
axis being inclined at angles of O (longitudinally loaded), the
60, and 90 (transversely loaded) degrees, respectively, to The
direction of the applied load, (as shown in Figure 2.1)
material of the test specimens was CSA G40.12 which has a speci-
fied yield stress of 44 ksi and a minimum tensile strength of
62 ksi. AWS E6OXX electrodes were used for welding the speci-
mens.

Butler and Kulak chose to analyze their experimental data

employing a load-deformation response for mechanical fasteners
of the following form.

R=Ruit (1- e_)X

 -6-

top view

Weld

P = applied force
e = inclined angle

side view

Weld

FIGURE 2.1 SCHEMATIC REPRESENTATION OF THE TEST

SPECIMEN
 -7-

Where

R = fastener load at any given deformation

R = ultimate load attainable by fastener

uit
A = shearing, bending, and bearing deformation
of fastener and local bearing deformation
of the connected plates
= regression coefficients
e = base of natural logarithms

Trial-and-error curvefitting of the experimental results

was used to obtain the following expressions for the dependent
variables in the equation. The inclined angle,0, is the only
independent variable to be given.
in -- e
R1t = 0.92 + 0.06030
A -0.47
max = 0.225 (0 + 5)

= 75
0.01460
X = 0.4 e

Where O is the weld inclined angle to the direction of the

applied load.
Readers are cautioned that these expressions were developed
specifically for ¼ inch (leg size) fillet welds made with E6OXX
electrodes; and, therefore, care should be used before applying
these to other size welds or welds using different electrodes.
Table 2.2 compares test results and predicted values
for the ultimate load and the maximum deformation.
Figure 2.2 summarizes the results of load vs. deformation
with respect to different inclined angles. The strength of
the fillet welds tested increased approximately 44% as the
angle of loads changed from zero degree (longitudinally
loaded weld) to 90 degrees (transversly loaded weld); however,
there was a substantial decrease in deformation capacity as the
strength increased.
Kato and Moritatudied the strength of fillet welded
joints theoretically by employing an approximate solution based
 -8-

Ultimate load Maximum deformations Pr.dict.d va1u.

kip8/in. in.

Ultimate Maximum
Group Std. Std. load deform.ition.
8, deg tean deviation Mean deviation Kips/in. in.
0 10.9 0.67 0.101 0.008 10.9 0.105
30 14.6 0.03 0.049 0.011 14.6 0.042
60 14.1 0.51 0.031 0.004 15.4 0.031
90 15.5 0.95 0.026 0.002 15.7 0.026

TABLE 2.2 TEST RESULTS AND PREDICTED VALUES9

g - 00

- point of weld failure

0.0' 0.06 0.08 0.10 0.1.

oupj*r., (r,cFos)

FIGURE 2.2 LOAD VS. DEFORMATION FIGURE 2.3 TRANSVERSE LV LOADED

0= 00 to 90° 1/4 in. FILLET WELD (7)
FILLET WELDS (13)
 -9-

on the theory of elasticity and supplemented this by an

elastic-plastic strain-hardening analysis performed numerically
using the finite-element technique. The approximate solution
is based upon the following assumptions:
The direct stress (q) on the tensile face of the
weld is uniformly distributed.

The pattern of the of the elastic stress distribution

remains unchanged until the braking of the weld.

Breaking will occur when the shear stress at a point

of the fillet weld reaches

Tmax =

where
= the tensile strength of the weld
metal
The fillet weld has legs of equal size.

The model used for this study is shown in Figure 2.3.

The maximum strength of a transversly loaded fillet weld was
found to be:

Tt max = 1.46 A Tmax = Aw

in Figure 2.3 is the
The oblique plan RP (e = u/e)
fracture plane of a transversly loaded fillet weld and the
throat RQ is the critical section of a longitudinally loaded
fillet weld.
This indicates that transversly loaded fillet welds are 467e
stronger than longitudinally loaded fillet welds of the same
size and length.

2.3 Review of Fatigue Strength of Fillet Welds

It was reported (17) that the fatigue strength of a 5/16-

inch fillet welded Tee-joint was reduced tremendously from
plain-plate strength under certain types of loading and stress
level. Table 2.3 shows the experimental data of such strength
reduction.

Since fatigue strength is a major factor in fillet welds,

TABLE 2.3 FATIGUE STRENGTH OF FILLET WELDED TEE-JOINT
(17)
UNDER CYCLIC LOADING

O to Tension Reversed
FlOO,000 F2,000,000 FlOO,000 F2,000,000
Plain Plate (A-7 steel) 47.8 ksi 31.7 ksi 26.8 ksi 17.5 ksi
Tee-Joint - 5/16" Fillet
Welds, Failure in Welds 19.1 ksi 9.6 ksi 13.3 ksi 6.2 ksi
 -11-

and it is difficult to alter the design to either avoid fillet

welds or place fillet welds in areas of low stress, there is
much interest in methods that may improve the fatigue strength
of joints. The Welding Institute conducted experiments to
determine the effect of peening and grinding on the fatigue
strength of fillet welded joints.(l The test pieces have
non-load-carrying attachments fillet welded either parallel to
or transverse to the direction of the applied stress. These
specimens were fabricated in such a manner that the direction
of stressing was parallel to the rolling direction of the material.
Details of the test pieces appear in Figure 2.4.

To study the effect of peening, the samples were peened

with a pneumatic hammer, fitted with a solid tool having a
rounded end of approximately 1/2 inch diameter, that was
moved along the toe of the weld at a speed of approximately
18 inches per minute. Usually, three runs of peening were
required on each specimen to ensure that the whole length of
the weld toe was subjected to the peening treatment.

Two types of local machining were also studied. The first

consisted of grinding only at the weld toe. This grinding was
carried out so as to ensure that the grinding marks were
parallel to the direction of the stress. The second type of
machining involved machining the whole weld to yield a concave
profile and a smooth blend of the weld into the plate surface.
The goal of this treatment was to obtain the maximum possible
increase in strength that could result from machining.
During the testing, all specimens were axially loaded
with one of the following stress cycles. Either the test
piece was loaded under pulsating tension with a lower limit
of zero or an alternating load causing minimum and maximum
stresses equal in magnitude but opposite in sign. The cri-
terion of failure was the complete rupture of the test piece.

Some of the samples were fabricated with welds around

the ends of the gussets, while others were left with the ends
unwelded. It was found that the fatigue strength of these
two types of samples was the same for the non-load-carrying
longitudinal fillet welds, and increased with both the peening
treatment and local machining. The increase in strength grew
larger as the life increased in the case of peening; whereas for
the local machining operation, the increase was about the same for
the whole range examined. The test results on the effects of
grinding and peening for mild steel specimens with longitudinal
and transverse gussets are shown in Figures 2.5 and 2.6, respect-
ively.
 16
A. SPECLMEHS WITH TRA1SYEJSE GUSSETS

14

i-1 %
12
z
o
vi_ 10
H
° 8

r ScArrERBAHD
EPL'2JS IN
F01/t
\- 1JLLY caiìo

THE AS-WELDED -.._.

CONDITION
4
B. 'ECJiIEI4S WITH LONGITUOLNAL GUSSETS
2

- iiIIiiUUiiUiUII!ass OL I

'rnuiuiuiniuiui 10' 2 3 4 5 io6 2 4

3 5
IDUHASCE. CYCS

FIGURE 2.14 DETAILS FOR TEST SPECIMENS(18) FIGURE 2.5 FATIGUE TEST RESULTS FOR MILD
STEEL SPECIMENS WITH LONGITtpTNAL
20 GUSSETS SHOWING THE EFFÇTS 0F
i

GRINDING AND PEENING.(1)

16

14
H
12
o FIGURE 2.6 FATIGUE TEST RESULTS FOR
v 10
MILD STEEL SPECIMENS WITH
.....
TRANSVERSE GUSSETS SHOWLMG
EMID FJH THE EFFECTS OF GRINDING
STES X THE
6 AS-WLLGED C0DITOE AND PEENING.(18)
4

o
2 3 4 5 2 3 5
E$DURAJWE. CXCl2S
 -13-

In the tests employing pulsating tension, it was found

that peening increased the fatigue strength by about 75%. With
both pulsating tension and alternate loading, the full local grinding
operation increased the fatigue strength by about 50% in all cases
except that of mild steel specimens with transverse fillet welds
which yielded nearly 100% improvement over the as-welded condition.
Even though this is less of an increase than that obtained from
peening, the difference in the slope of the S/N curves for peened
and ground specimens accounts for the fact that grinding was found
to be more effective than peening for tests in which the number of
cycles was less than about 50,000. Full grinding of the test
pieces with longitudinal fillet welds normally failed as a result
of initiation at the root of the weld. In the case of light grinding
at the weld toe, the improvement varied. This is assumed to be
related to the fact that it is very difficult to control the
degree of grinding. This technique, considered to be unreliable,
is, therefore, not recommended. It is interesting to note that in
tests performed on samples with transverse gussets, fully ground
and also peened, fatigue strengths as high as the parent material
could be obtained (Figure 2.6).
 -14-

3. REVIEW OF STANDARDS
3.1 Review of Fillet Weld Specifications

For convenience, the following abbreviations have been used

to represent various classification societies
are reviewed in this chapter: whose specifications
L.R. Lloyd Register
ABS = American Bureau of Shipping
GER.L.L. = Germanischer Lloyd
AWS = American Welding Society
AISC = American Institute of Steel Construction
D.N.V. = Det Norske Ventas
B.V. = Bureau Ventas
NKK Nippon Kaiji Kyokai
USN = The United States Navy
USSR = Russian Classification Society
There are two measures of fillet weld strength
used in the various codes. One of them is based on the extensively
tive throat thickness (t), defined as the shortest effect-
from the root (A) to the face of the weld distance
(Figure 3.1). Another
one is based on the fillet leg (W) which, for
fillet weld ¡ is equated to the throat thicknessanbyequal leg
W = t
All the rules of the various classification
societies give the
nUnimum required weld size by fillet leg (W) or throat thick-
ness (t)

The fillet leg or throat thickness is given as a function

of the plate thickness of the attached members as well as its
position in the ship structure. The latter reflects the
different
loading conditions to which the attached members are subject
to their position. due

Some of the codes put limits on weld leg size as well as

some allowances (for example, corrosion allowance) or restictions
( for example, maximum permissible gap). As will be discussed
later, ABS incorpoated a corrosion margin of 1.5
mm (0.059 inch)
in throat thickness when they changed their requirements on fillet
weld size from intermittent to continuous weld.
ABS rules have also
incorporated the corrosion margin in the base plate
requirements. thickness
 -15.-

Fillet weld

Throat thickness

A
w
Root weld leg Toe

FIGURE 3.1 DEFINITION OF BASIC PARAMETERS IN FILLET

WELDS
 -16-

The fillet size requirements specified by

ten major clas-
sification societies, the US Navy and the
Handbook are summarize in Table 3.1. Structural Steel Designer's
The 1977, or earlier
editions, of the classification societies were used for
comparisons. these

In order to compare the various rules, the required

thickness for double continuous welds is plotted against throat
the
plate thickness (thinner of the two plates joined by the fillet)
with respect to the location of the most
structure members up to 24 mm thickness incommon applications of
Figures 3.2a through
3.2j. It is seen that the highest value is more than twice
that of the lowest. The plots of fillet sizes also
show
dramatically the variation among the various classification
society rules and suggest opportunity for rational
improvement.
Investigation of the reasons for the differences in
sizes among the major ship classification societies fillet
successful. Many of the welding specifications was not too
were developed
many years ago and the history of their development was not well
documented.

3.2 Corrosion Considerations

In designing welded joints of ships, for general
a method commonly used to ensure a proper design is thecorrosion,
corrosion margin. The U.S. Navy specifications use of
corrosion margin; however, ABS specifications have a corrosion a
do not require
margin built into the requirnents for plate thickness and auto-
matically provide one for the fillet welds because the
fillet welds are based on plate thickness. sizes of

Special protective coatings have been used as an alter-

native for corrosion margin by some ship classification
societies. However, this corrosion margin
can not be taken
as the allowable reduction in either plate thickness
or fillet size.
Plate materials tend to corrode more than weld materials
as far as usual combinations of plate materials and weld
materials are concerned. The corrosion rate of ordinary ship
hull steels in sea water is, according to Professors
H.H. Uhlig
and R.M. Latanission of M.I.T., roughly 0.005 inch
Since this rate decreases due to a build-up per year.
of oxides on the
surface, normally a rate of 0.003 or 0.004 inch
used for a period of ten years. If a,corrosion per year is
margin or 1.5 mm
is imposed on a welded plate, a service life
of approximately
20 years can be obtained. Since most surfaces on a ship that
are exposed to sea water or bilge water are maintained
some form of protective coating is kept intact most so thát
of the time,
the corrosion margin of 1.5 mm is considered sufficient
 -17-

TABLE 3.1 SUMMARY OF FILLET WELD SPECIFICATIONS

(mm) Increase due O

W gap allow
WELD SIZE min max to gap
\ Ir ABS allows reduction in
ABS í(ti 2 No
ti
2'gap<mm fil let size for deep
minimur)4.s (i" 1 mention penetration weld.
\ï 3)
increase leg
size by opening

In the cargo & ballast

Det tanks some connections
= f 5 No ment. No ment. "o
Norske can be reduced by 5 mm
Ventas (tplate)
in excess of the minimum
required (3 + l.t)mm or 6
min which even is less if
an effective corrosion
protection is applied

Bureau = f(tPlate No No Where deep penetration is

5 No
Ventas )j used,reduce throat depth
minimu
up to 15%

Ge riman f(tplt \ No No No No
Lloyd
\ minimum)
t . >.4t max
min
With autotnaticdeep pene-
Lloyd ...(tiate \ tration welds may be re-
Register ininimumiNo No No No duced by 15%
3 jnhl(

t>. 2ltor<hand
3mm
au tom.
deen penetr.
Not excessive deposit of
Taiwan w = fïtplate )No No No No weld is permitted
Register \minimurr (oversize)
Undersized fillet welds
N.K.K. No No
W= can be considered if weld
1N0 length is less than 10%
minimu
f(tplate
Registro w f(t 4 4.5mm
plate No No
Italiano fort< 6mm No
) 5mm
for t>6mni

A.W.S. w tplate 'No No Fillet welds in any single

maximu5 continuous weld should be
permitted to he undersize.
by 1/16' without correctil
provided that the undersij
eld doesn't exceed lfl oil
.the length of the weld.
L
 - 18-
TABLE 3.1 SUMMARY OF FILLET WELD SPECIFICATIONS (CONTINUED)

WELD SIZE MINIMUM LEG gap INCREASE DUE

LENGTH, w max TO GAP °a11ow
USSR
Register
f Shipping :
w=ctT1
weld strength factor
(depends on structur-
3
no no no
al item)

$: factor determined
from the type of
weld
e T1 R1
J.S. Navy W =
1.414 R2 no no no no
e weld efficiency
T thickness of weaker
member
R UTS of weaker member
shear strength of
weld deposit t
Structural 1.411 R1 T1
Steel Desig- w = for 1/8 inch for rio no
-er's Hand- R2
no
tension building, 3/16
ook inch for bridge
1.414 R3 T1
w=
R2

R shear strength of
base metal

10.0
9.0 <- US S R
GER.L.L.
3.0
- 7.0
121 SC
- NKK
U2
6.0
o
5.0
E-'

4.0
o
3.0

8 12 16 20 24
PLATE THICKNESS (mm)
FIGURE 3.2a CURRENT WELDING REQUIREMENTS FOR JOINTS BETWEEN
DOUBLE BOTTOM FLOOR AND SHELL PLATING (LONG'L FRAMING)
 -19-

10. 0

9.0
3.0
7.0
6.0
5.0
4.0
3.0

16 20 24
12
PLATE THICKNESS (mm)

FIGURE 3.2b CURRENT WELDING REQUIRENENTS FOR JOINTS

BETWEEN DOUBLE BOTTOM SIDE GIRDERS AND
INNER BOTTOM

10.0
9.0 ..-...GtR .0. .0..

8.0 ____. - -
7.0 N)(K
iO
u'
hi

V,
4.0
3.0
- 0.N.V.

z
X
3.6 20 24
$ 12
FUITE THICKNESS (mi")

FIGURE 3.2c CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN WEB FRAMES AND SHELL PLATING

10.0
9.0
8.0
NJ<K
7.0
6.0
M1S
5.0
4.0 S-. f
lUn.
3.0 ID.N.V.

16 20 24
8 12
PLATE TH1CP3ESS (mr')

FIGURE 3.2d CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN STIFFENERS AND NON-TIGHT STRUCTURAL
BULKHEADS
 -20-

10.0
9.0
3.0
o, 7.0
-..-- __1SC
:o K
6.0
U5Sk

s 12 16 20 34
PLATE TI!ICKCSS (XX)

FIGURE 3.2e CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN BEAMS AND DECK

10.0
9.0
8.0
7.0

-k.-
4.0
X.

3.0
----

12
¡ I J
1.6 20 24
PLATE THICKJESS (nrn)

FIGURE 3.2f CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN HATCH COVER STIFFENERS AND WEBS,
END ATTACHMENT

10.0
9.0 APS
LSSR
1.0
7:0

-
o,
u,
C
6.0

-0.v.
I.
50
4.0
_-
- -

-
S-
i 3.0

X
12 16 20
PLATE TUICKI000 (r)
FIGURE 3.2g CURRENT WELDING REQUIREMENTS FOR JOINTS
BETWEEN CENTER GIRDER AND KEEL
 -21-

20 24
12 16
B
Pt.ATE THIC(1tSS (Trc)

FIGURE 3.2h CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN DECK AND SHELL PLATING

..GER.I,.L.
:: - No specification S. V -

L
9.0

5.1 vS0sC
L

7.0
USSR
6.0

5.0
4.0
3.0

16 20 24
12
PL?Tf THICKNESS (r

FIGURE 3.21 CURRENT WELDING REQUIREMENTS FOR JOINTS

SHELL,
BETWEEN FOUNDATIONS OF MAIN ENGINE AND
TOP PLATE OR INNER BOTTOM
D.N.V. - No ppeification
n. V.
0.0
I. 1.
9.0
8.0
E
7.0
6.0
u
5.0

4.0
øt
3.0

16 20 24
8 12
PI.ATE T!IICXNESS tmzt)

FIGURE 3.2j CURRENT WELDING REQUIREMENTS FOR JOINTS

BETWEEN RUDDER JOINTS ON THE MAINSTOCK
 -.22-

for steel weld joints in a sea water environment.

3.3 Fabrication Limits

Metallurgical restraints impose a minimum size of fillet

welds. To decrease a weld specification below 3/16 inch (leg
size) would be unrealistic because it is too small for welding
practice.

There are occasions in which fillet welds made under op-

timum welding conditions tend to be slightly undersized, as
shown in Figure 3.3. It is a common attitude for an inspector
to reject these slightly undersized fillet welds. In many cases,
two or more passes of weld metal must be added to satisfy the
requirements. It is usually, in this case, impractical (if not
impossible) to add only a small amount of weld deposit. The weld
is always overwelded, as shown in Figure 3.3b. This not only
wastes time, labor, and material, but also creates more distortion
which causes more fabrication errors in other joints. This kind
of waste can be reduced by allowing some undersized welds, if
the structural integrity of a ship is not impaired. Distortion
can also he reduced by not adding more weld material to the joints.

The maximum gap requirements and the allowable convexity for

fillet welds specified by the U.S. Navy is discussed in Section 3.4.
ABS rules have the same maximum gap requirement but do not require
the maximum allowable convexity. Other classification societies
do not mention requirements for either gap or convexity.
3.4 ReView of U.S. Navy Welding Specifications
The U.S. Navy welding specifications are presented
in a
different format than the other standards discussed in the previous
section. The required weld sizes
are presented in graphical form of
a plot of plate thickness versus joint efficiency.*
Different plots
are presented for each different combination of
and electrodes used, and types of welded joints. Rather than materials
onstruction
use
a different graph for various joint locations in the ship, the
Navy specifications for joint efficiency include the factor of
joint locations. A partial listing of the required
joint efficiency
is given in Table 3.2. For a complete listing, see reference 19.

*
Joint efficiency is defined as the ratio of ultimate strength
of weld deposit to ultimate strength of base material.
 -23-

Required fillet
-size by specification

Slightly Undersized Fillet Weld

Required fillet
size by specification

b. Overwelded Fillet Joint as a Result of

Additional Welding Two Passes

FIGURE 3.3 UNDERSIZED AND OVERWELDED FILLET JOINT

 -24-

TABLE 3.2 REQUIRED USN JOINT EFFICIENCIES FOR VARIOUS FILLET

t1ELDED JOINTS (l

Item Connection Joint Efficiency

(Per Cent)
Bilge Keels Connections to shell 75
Bulkheads, Longi- Main subdivision bulkheads
tudinal and Trans- 100
verse

Decks and Platforms Longitudinal 75

Transverse
With deck on änly one side 75
With deck on both sides 100
Shell and interbottom 75

Foundations Gun Foundations 100

Framing, Longitudi- Connections to flanges or
nal and Transverse faceplates around lightening
holes End connections to 75
intersecting members
Ordinary frames (less than
24-inches in depth) 100

Masts and Booms All joints 100

Piping Penetrations Shell plating and supports 100
Vertical Keel Connections to flat keel and
rider plate 75
 -25-

One example of the Navy welding specifications is shown in

Figure 3.4. The graph is plotted for a continuous, double-
f illet welded tee-joint made of medium steel (U.T.S. 60,000
PSI) using MIL-6011 electrodes.
Comparisons of the U.S. Navy specifications to other
welding standards for required weld size versus plate thickness
for joints between double bottom floors and shell plating, between
web frames and shell plating, and between decks and shell plating
can be seen in Figures 3.5, 3.6, and 3.7, respectively. While
Germanischer Lloyd is the most conservative, followed by the
American Bureau of Shipping and the U.S. Navy ,Bureau Ventas,
Lloyd Register and Det Norske Ventas are the most liberal.
Also, it is very apparent that there is a wide range between the
most conservative rules and the most liberal rules. In fact, there
is over a factor of two difference in some cases. This difference
may not be as large as it seems because the specifications may be
based upon slightly different models or include or exclude dif-
ferent considerations. For instance, one may include a corrosion
allowance and another may tell the designer to add on a margin in
addition to what is required by the chart.

The U.S. Navy specifications have the same maximum gap re-
quirements as that required by ABS specifications. The maximum
gap that is allowed without increasing the weld size is 1/16 inch.
If the gap is greater than 1/16 inch, the required weld size is
equal to the normal required size plus the gap. The maximum per-
mitted gap even with increasing the weld size is 3/16 inch.

The U.S. Navy specifications also limit the maximum

allowable convexity for fillet welds which varies with the weld
size as shown in Figure 3.8.
(19)
The tolerance on fillet weld size is as follows Fillet
welds up to and including 3/8-inch size shall not vary below the
specified size by more than 1/16 inch, and any such variance shall
not extend for a total distance greater than 1/4 of the joint
length nor for more than 6 inches at any one location. Fillet
welds, 7/16-inch size and larger, shall not be less than the gage
limits for their respective sizes."

A-
 -26-

_
_-'-
40.8
40
_I 1

35.7 - 7/8
3

30.6
_____
L
3/4
c1

z
' 25.5 25 5/8

o
20.4 i_L/
FI)

-4
FI)
20
z
14 wa Q

o E-4

15.3 1
3/8

-
-
10.2 10 1/4
-s'
Wi__
5.1
ra ______________
60 70 80 1/3
90 100
EFFICIENCY - PCENT
FIGtJRE 3.4 EFFICIENCY CHART FOR C0NTflU0US
D0U-FIJJFT WELDED T JOINTS NADE BETWEEN IUN STEEL WITH
MIL-6011 ELECTRODES9 BASED ON THE THINNEST 0F THE TWO PLATES JOINED
 -27-

lo o

9.0

A1ICAÎ1 BUREAU 0F SHIPPING (ABS)

GE(!tANISCHER LLOYD
Q
6.0 (GE R.
E-.
BUREAU YERITAS (B.v.)
5.0 U.S. NAVY (usN)
LLOYD RISTER (L R.)
(D. N. V.)
E-.

3.0

PLATE THICKNESS (Mr.)

BETWEEN
FIGURE 3.5 CURRENT WELDING REQUIREMENTS FOR JOINTS
DOUBLE BOTTOM FLOORS AND SHELL PLATING

10.0

9.0

8.0
U)
" 7.0
L)
0.0

20 22 2
1L 16 18
8 10 12
PLATE THICKNESS (PiJ.)

FIGURE 3.6 CURRENT WELDING REQUIREMENTS FOR JOINTS BETWEEN

WEB FRAMES AND SHELL PLATING
 -28-

lo o

9.0

.. 8.0
U)
ti]
14
7.0

6.0
E.

5.0

3.0
X

I i i i i I I I. i
8 10 12 l 16 18 20 22 24
PLATE ThICKNESS (MÌq)

FIGURE 3.7 CURRENT WELDING REQUIREMENTS FOR JOINTS BETWEEN

DECKS AND SHELL PLATING

3/32

5/61+

CONV

1/16

weld leg
t I I t
1/8 3/16 l/ 5/16 3/8 & LARGER
FILIZP WELD LEG SIZE (INCHES)

FIGURE 3.8 MAXIMUM ALLOWABLE CONVEXITY FOR FILLET WELD SIZES (19)
 -29--

4. DEVELOPMENT OF ANALYTICAL METHOD

To study the stress details in the welds of a fillet joint,

aither a photoelastic analysis or mathematical stress-strain
analysis, can be used. With complex structures such as ships,
thousands of combinations of different types of joints and their
applied loads may exist. The photoelasticanalysis is not practical
to apply to all joint-load combinations in a ship. For this reason,
mathematical stress-strain analysis is considered a more useful
and effective means for calculating stresses in fillet welds.

4.1 Analytical Method

A finite-element computer program, named "Automatic Dynamic
Incremental Nonlinear Analysis (ADINA)Tt, was used to develop the
tool for reviewing the currently existing fillet weld specifications.

Program ADINA, a general purpose finite-element computer

program for linear and non-linear, static and dynamic, three
dimensional analysis, is an out-of-core solver, i.e., the
equilibrium equations are processed in blocks, and very large
finite-element systems can be considered. Also, all structure
matrices are stored in compact form, i.e., only non-zero elements
are processed, resulting in maximum system capacity and solution
efficiency.
Inputs in the program are the joint dimensions and geometries,
coordinates of each node, applied loads, boundary restraints and
material properties, such as Young's modulus E, tangent modulus
E (for the case of strain hardening), yield stress a and
Pisson's ratio ij, of the base material and weld depoit.
The outputs from the computer analysis are the stress
distribution over the fillet weld, displacements in every nodal
point and strain conditions (elastic or plastic) of the stressed
areas under certain external loads.
Figure 4.1 shows the general solution procedure of the ADINA
program.

4.2 Method for Determining Miflirtrnnì Fillet Weld Sizes

To determine the minimum fillet weld sizes, either allowable
design stress intensities or the strain conditions in weld have to
be integrated in the analytical steps of ADINA programming. The
fillet joint dimensions and geometry can be obtained from actual
drawings of ships. Figure 4.2 shows the working process of how the
computer program determines the minimum fillet weld sizes if all
required information is known.
 30-

FILLET WELD JOINT

UNDER LOADING

RATIONAL
ASSUMPTIONS
y
STRESS MODEL WITH A
FINITE ELEMENT MESH

Y
SOLUTION BY FINITE ELEMENT
METHOD

STRESS DISTRIBUTION AND

STRAIN CONDITIONS
OVER THE WELD

a g ree EXPERIMENTAL CHECK not

BY PHOTOELASTICITY ANALYSIS a gre e

Criteria for
determi ni ng
minimum
fillet weld
sizes _

UPDATING FILLET WELD

SPECIFICATIONS

FIGURE 4.1 GENERAL SOLUTION PROCEDURE OF

THE ADINA PROGRAM
 -31-

Experience Ac tuai
drawings

Fillet joints Overall stress

sensitive to analysis of the
failure ship structure
L
Fillet joint dimensions Local loads
and geometry

Weld size from rules

Deduction of
corrosion margin
from weld size Applied loads and
boundary conditions
Material properties
of weld and plates

FEM stress analysis

Criterion quantity, X

Required

W=W± ZifX>X.. X=XR equired is

Req.
W=W- L9ifX<X. no satisfied?
+ Reo

ye s

Corrosion

Manufacturing
limits

Final weld size

Stop

FIGURE 4'.2 PROCESS FOR DETERMINING MINIMUM FILLET WELD SIZE

 -32-

Overall stress analysis of the ship structure using either

theory of structures or finite-element stress analysis
must be
performed to obtain the local loads acting on the joint. The
boundary conditions supporting the edges of the cut-off joints
are also essential factors in the analysis and have to be rationally
assumed.

A criterion is required in the analysis for determining the

minimum weld sizes. Let X be designated as a criterion quantity
such as yielding, and X
X is not equal to X require1 as the critical quantity. If
tne fillet weld size may be reduced
or increased by an g 3f'i and a new weld dimension is
formed. The iterative process then begins until the situation
of X=X is reached. The size of fillet welds resulting
d
U±rative
from -

process is the theoretical minimum size

required under certain applied loading conditions. Corrosion margin
may be added to this theoretical weld size. The final fillet weld
size is then checked by the
manufacturing limits caused by metal-
lurgical or operational considerations.
welds that can be made by the available For example, the smallest
welding process or the
minimum required weld sizes for preventing
cooling. cracking due to rapid

4.3 Mathematical Modeling and General Yielding Criterion

Due to the overall geometry of ships, the local details

of a tee-joint may be analyzed in accordance with the types
of loading which are either longitudinally effective or trans-
versely effective. In some cases the longitudinal and
transverse structures interact, such as the corner welds of a
panel stucture,so that a three-dimensional model must be used.
In other cases it is possible to isolate the longitudinal effects
from a transverse structure, treating them as boundary conditions,
and deal only with the transverse joints. In such a case, a two-
dimensional analysis may be applied.

A two-dimensional analysis for a tee-joint under simple tension

acting on the flange was used to check the validity of the
ADINA program and to modify the program for general applications
of the fillet weld strength analysis. Figure 4.3 shows the mathe-
matical model of a tee-joint under simplified loading condition.

The joint with tension on the flange shown in Figure 4.3 may
represent a tee-joint in the floor of midship section, midway
between the stiffeners but with only ship hull girder stress
which is uniformly distributed (approximately) across the flange
thickness.
 -33-

y
£

- t
2

h1

A
I
k 2--.4 p
t 1
i V
X

e = gap between the web and the flange

= angle of fillet weld with x axis
e
= fillet leg size
h1 = web height
= length of flange
t1 = flange thickness
t2 = web thickness
FIGURE 4.3 MATHEMATICAL MODEL OF A FILLET WELDED TEE-JOINT
UNDER TENSILE LOAD ACTING ON THE FLANGE
 -34-

A criterion X is needed for determining the minimum

weld sizes in In this research, a general
yielding concept was developed. This is to use the general
yielding condition along the weld leg as a determining factor.

The criterion of general yielding is defined by

X=Xrequired
Length of yield plastic zone along the
where weld leg
= -r
X = Weld leg size

The required yielding criterion is assumed to be 1.

To illustrate the concept of general yielding criterion

numerically, a calculation was conducted for a tee-joint shown
in Figure 4.3 with rough mesh sizes.* Since the joint is
symmetrical, only half of the joint was analyzed.

Figure 4.4 shows half of the tee-joint with finite-element

meshes. The dimensions of the joint are assumed as follows:
Length (2 /2) = 800 mm
Plate 1 thickness (t1) = 18 mm
Plate 2 thickness (t2) = 10.5 mm
Web height (h1) = 300mm
Fillet weld leg (2) = 5 mm
Root gap (e) = 1.59 mm
The applied boundary conditions are as follows:
Clamped, across face A
Simply supported, at point B
Since the stress-strain relationship above the yield
point depends on the history of loading, the loading function
of applied tension on the joint flange is assumed t have 36 step
increments. The load increases from O to 50 Kgr/mm , as shown in
Figure 4.5.

*
Rough mesh size was used in this test run because the
accuracy of results was not important and the purpose of this
run was merely to demonstrate the concept of general yielding.
 -35-

z
** Not all the elements are
Plate 2 Shown in the figure
Face A ** Gap = 1.59 mm (l/16')
Y

gap

plate i

I t t I
0 10 20 30 (mm)

FIGURE !LL FINITE ELEMENT MESH FOR TEST RUN

50

30

20

10

5 10 15 20 25 30 35
Time

FIGURE LL5 LOADING FUNCTION FOR TEST RUN

 -36-

The variable finite-elennt mesh consists of 33 elnts and 48 nodes.

The mesh is finer near the root and toe due to possible stress concentrations
in these areas.
Assume that the material properties of base plates and weld
metal are the same; They are:
E = 21,000 tgr/rnm2
y = 21 Kgr/mm 2
'J =0.3
Et =
40

The analysis is an elastic-plastic, 2-D, plain-strain

analysis, using the Von-Mises yield criterion, with a linear
strain hardening. As the load starts to increase, all elements
are in the elastic region. At some load level, some elements
into the plastic region. The results of the analysis are showngo
in Figure 4.6. The shaded lines give the elements that become
plastic at a given time step.
Hence, at time0step 2 (corresponding to a load of 15 Kgr/mm2)
the small element n 24 at the toe of the fillet is plastic
(given by vertical lines). Then at time step 5, the next element
becomes plastic and so on.
So, for each time step (corresponding to a given applied
load), the portion of the fillet weld leg which is in the
plastic region can be found.At time step 2, for example, 10%
of the fillet weld leg is plastic, or using the definition of
X, X=lO%. Similarly, at time step 5, X = 20%, at time step 9,
X = 40%, and so on, For a given load, as long as X is less than
piecntis
X
d the fillet size can be reduced and the iteration
until the condition of X = 1 is reached.
In the above example, the state of X = 1 happened a time
step 14, with an applied load of approximately 25 kgr/rnm
it is interesting, physically, to examine how the plastic
zone progresses. First, plasticity appears in the toe element
due to high stress concentration (of the order of 1.5). The
next element which goes to plastic region is the one next to the
toe element. Then, plasticity appears in the root of the
fillet. This observation may be a good explanation of the
statistical results which indicated that the fatigue crack
always initiated from a fillet toe.
 fl time step 2 0 element in past1 region
time step 1?
time step 5
Lj time step 18
D time step 9
D time step 19
time step 12 time step 27

iip

Hi 'i.:i .\T T1YI' STEP 12 EI.EPENT AT Ti E STEP 27

u element in niastic region

element in plastic re.in
time step 13
time step 30
time step 14 time ste' 35
time step 36
D time step 15
time step 16

ELEi[NT AT TWE STEP 16 iElENT AT TIME STEP 36

FIGURE )1.6 SPREAD OF PLASTIC REGION IN THE TEST ELEMENT

 -38-

4.4 Numerical Example of the Effect of In-Plane Tensile Stress

in Bottom Shell Plating on Fillet Weld Strength

The example shown in this section is simply to demonstrate

how the fillet weld sizes can be reduced using the ADINA program.
The effect of root gap on the stress concentration at the toe Of
a fillet weld under in-plane tensile load applied on the flange
was also analyzed.

The joint analyzed is shown in Figure 4.3. The finite-

element model consists of 72 plain-strain elements* and 219
nodes1as can be seen from Figure 4.7. Dimensions of the joint
shown in the Figure are the same as that between a transverse
floor and the shell plating of an AD-37 class ship.

Assuming that the uniform tensile load is caused by ship

hull girder bending on the bottom shell plate in the midship
section of an AD-37 class ship, t1)1oad may be determined by
simple beam thry and the load is
= 34,538 psi
= 24.29 Kgr
nun

This load is applied in the line connecting the nodes 1,

2, and 3, and in a negative Y direction.

The analysis performed was an elastic-plastic analysis

using the Von-Mises yield criterion, with a linear strain
hardening. The tangent modulus was assumed to be
Et=mE= E
40
The joint was assumed to be made of mild steel, with the
following material properties:

E = 21,000 Kgr/mm2
2
24.5 Kgr/Inm
= 0.3

Since an elastic-plastic analysis is performed, it is

important to make small load step increments in the plastic
region. If the step increments are large, no equilibrium
stiffness matrix can be reformed in the program. So, the total
load was applied in three static steps of 50%, 70% and 100%
(Figure 4.8)
*
Since the joint is long in the direction of welding compared
to other two directions, a plain-strain condition is assumed
on the weld cross sections.
 .39...

c tire step i
J tLstep2
Length Unit: rr

Tensile Load: 24.29

605.24
GA? 1. 59rrr

zt
4 76 y

elnt r. 38
15.87

- 2032.48 4.76 4.76

f.
20 32 . 48
9.5,

FIGURE 14.7 TWO DIMENSIONAL FINITE ELEMENT MESH

t OF LOAD

no

70

5f)

7.10 24.29
APPLIED LOAD

FIGURE 14.8 THE STATIC STEPS OF APPLIED LOAD ON A TJOINT

 -40-

This computer model is suitable to perform various kinds

of parametric analysis, provided that only one parameter varies
per time holding all the others constant.

The first calculation was performed by varying the weld

size of the joint under simple tension (Figure 4.3). The
iterative process started with an arbitrary value, = 4.76 mm*
,

and gradually reduced the fillet sizes by 10%, (4.284 mm) 20%
(3.808 mm) and 30% (3.332 mm). The results are shown in Figure
4.9. This figure shows that since the slope of the curve X vs.
% of reduction is still far from zero at the point of 30% reduction,
a 30% reduction is, therefore, feasible in this case.

Ship structures, like any other structures, are never com-

pletely free of imperfections and defects due to design or
fabrication limitations. These imperfections may cause the
real structure to depart from the ideal model which is used in
the strength calculations. Knowledge of the extent of the
departure may provide reasonable insight into the safety con-
siderations which can be integrated into the structural design
procedures.

There are several ways in which a gap can be formed between

the web of the joint and the base plate. The most common way is
due to the fact that plates are never straight. They always
have some initial deflections, so that in a micro scale, the
plate shape is like the one shown in Figure 4.10. At point A
of the figure, the maximum gap occurs. The cross section of
the joint at point A is the one shown in Figure 4.7.

The second calculation was performed by varying the root gap

of the fillet weld from the maximum allowable value of 1.59 mm,
allowed by ABS rules, to as much as 30% more.

The computer calculations were conducted for the following

cases:

Case #1: Maximum allowable gap (1.59 mm)

Case #2: Gap increase of 10% (1.749 ram)

Case #3: Gap increase of 20% (1.908 mm)

Case #4: Gap increase of 30% (2.067 mm)

*
This value according to ABS rule, is a required fillet weld
size for a joint in the transverse floor of midship section in
the double bottom.
 -41-

40
30

20

10

.2 .3 .4 .5 .6 .7 .8 .9 i
Length of yield plastic zone
Length o weld leg
FIGURE 24.9 REDUCTION IN FILLET WELD SIZE VS. QUANTITY X, FOR A
TRANSVERSELY LOADED TEE-JOINT UNDER SIMPLE TENSION
ON ITS FLANGE

FIGURE 14.10 GAP

FORMATION IN A FILLET
WELD

During all these calculations, the weld leg, the applied

load, the geoneterv of the joint and the material properties
the
were held constant. Figure 4.11 prezents the results of
computer calculations. This figure shows how the stress
concentration varies with the percent of increase in root gap,
for the clement number 38. The conclusionfor the loading con-
dition tested, is that the stress cpnccj-itration near the toe of
a fillet weld decreases slightly - th q increases.

Stress at the toe elent r. 38

27.80
27.70

27.60

27.50

27.40 FIGURE LI.11 STRESS

27.30
CONCENTRATION AT THE
TOE ELEMENT
27.20
27.10

i r
I

0 10 20 30
G12 fl')JEASE %
 -42-

5. CONCLUSION AND RECOMMENDATIONS FOR FURTHER RESEARCH

5.1 Conclusions

Some experimental and simple mathematical studies on fillet

weld strength (mostly fatigue strength) have been conducted by
various researchers, but very little analytical work on detailed
stress analysis in welds has been done.

Comparisons of fillet weld requirements of various classifica-

tion societies indicate that the most conservative rule may re-
quire twice the size than that required by the most liberal rule.
Many failures in ship structures were fatigue cracks initiated
from the toe of fillet welds.
Slightly undersized welds are sometimes inevitable due to the
welding process limitations in actual practice. Overwelding
may arise if corrections are made to satisfy the requirements by
specifications. More distortions as well as waste of time,
labor, and materials may cause many other adverse effects.
Up-
dated rules should then be determined through analysis to
accept such slightly undersized welds.

A general yielding criterion which requires a full plastic

zone along the weld legs as the indication of failure is pro-
posed to determine minimum fillet weld sizes.
The "ADINA" program or a similar FEM program with modifications,
can be used for analyzing fillet weld joints under complicated
loading conditions.
5.2 Recommended Further Research
it is recommended that further research be made on the follow-
ing tasks:

Task 1. Determination of Stress Distributions in Ship Structures

to Assist the Analysis in Welds. To analyze the stress
and strain conditions in welds of various fillet joints
of ship structures using "ADINA" computer propraln, it is
first necessary to determine the stress distributions in
the ship structures. The stress found in the cross-
sections beside a particular joint are used as the stress
boundary condition (local loads) acting on the cut-off
edges of that joint in the computer analysis. Many
analyses and measurements have been conducted to determine
the stresses in the ship structures caused by various
combinations of loads to which the ship is subjected in
the open sea. Theories of structures are usually used
 -43-

for determining the stresses caused by simple loads such

as hull girder. bending and plate bending due to lateral
water pressure and stiffener restraints. Stresses caused
by some special type of loads, such as liquid sloshing
loads, and stresses in the areas with more complicated
combination of joint geometries are often studied using
numerical techiques. Recently, a finite-element method
has been used to determine the stress distributions over
the entire ship structure of an oil tanker by ABS.
Although the efforts have been made for determining the
stresses in the ship structures, these stresses have not
yet been characterized for reviewing the rules. It is
therefore recommended that stresses in the ship structures
under various combinations of loads be characterized in
terms of joint geometry and joint location for a parti-
cular ship.
Task 2. Review of Fillet Weld Strength of Various Joints in
Ships by Computer Analysis. Ship Structure Committee
sponsored research has developed a computer method, using
the "ADINA" computer program, for analyzing the strength
of fillet welded tee-joints. A simple tensile load
acting on the two edges of the flange was analyed in a
numerical example to demonstrate the program. It is
recommended that analysis of weld strength of various
joints in shipsusing "ADINA" or similar computer programs
be conducted. The expected results will relate the
minimum allowable fillet weld sizes (where the critical
yield criterion is just met) to the plate thickness of the
joints at a particular location in ships. Photoelastic,
or similar stress analysis, experiments for determining
the fillet weld strength of several tee-joints under
simple loading should also be conducted to check the
validity of the mathematical modeling and the computer
results. Any modifications in the mathematical modeling
or in the computer programs should then be made before
going on to the analysis of the joints under more com-
plicated loading conditions.
Task 3. Development.of a.Rational Procedure for Updating the
Fillet Weld Specifications. A ship structure Committee
sponsored study has indicated the possibility of reducing
fillet weld size requirements. One way to achieve a solution
is to develop an algorithm approach. In this approach,
the required weld sizes would be the sum of the increments
in weld size which are required for each of the factors
that might affect the strength of a welded joint. With-
in each category, the value of the increments could vary
from zero to some maximum value depending on the conditions
of the particular joint in question.
The joints would be classified by type to take into account
 -44-

the required joint efficiency, possible different

re-
quirements of different classes of ships and the location
of the joint in a given ship. A matrix would be set
up
to give the value of each increment for different joint
classifications.
Table 5.1 represents the elements of the algorithm system.
With information previously developed, along with
a simple computer program can be developed to take some experiments,
account all the
factors in the algorithm chart and to perform the optimization
the iterative procedures in the computer. The algorithm charts through
can then be reviewed and incorporated in the specifications.
Task 4. Study of the Significant Benefits from the Reduction of
Fillet Weld Sizes. The reduction of fillet weld size
requirements can benefit ship construction by allowing
smaller welds, reducing the cost of construction,
accepting slightly undersized welds if the itegrity of
the joint is not impaired, and reducing weld distortion
by depositing less weld metal to the joint. Among the
anticipated benefits, the cost saving may be the most
important concern by the shipbuilders.
A preliminary study on the cost saving due to the ;eduction
of fillet weld size has been conducted by Malliris 2])
Based on a possible 30% reduction of fillet weld size
the total costsaving including welding consumables,
welding time and labor cost in the construction of a
50;000 DWT tanker can reach $102,900 and 22 tons of
welding consumables.
It is, therefore, recòmmended that an economic study
be undertaken that would include the reduction of welding
consumables, welding time and labor cost.
It is also recommended that the beneficial effect of
reduction of fillet weld size requirements on the
weld distortion be studied as it may be one of many
important consideràtions in determining the acceptable
undersized weld in actual welding practice. Either
experimental approach or computer analysis, such as
using the computer progns developed at M.I.T., may
be used for this study. ''
 -45-

TABLE 5.1 FILLET WELD ANALYSIS SYSTEM ELEMENTS

Category d. Method of Determination

i

Static strength Coxruter FEM

Fatigue margin a2 Experimental results

Fabrication or workmanship FEM

Welding Method a4 Experimental results

Conditions of welding as Industrial data

Environmental:
Corrosion (general) d Calculation-experimental
6
results

Corrosion (local) a7 Experimental results

Quality control:
bility to detect
defects d8 Industrial data

Test procedure
required d9 Specifications

Design method dl0 Ju dg eme n t

Strengths of the weld metal Experimental results

Metallurgical restraints Experimental results

 -46-

REFERENCES

Nasubuchi, K., Monroe, R.E., and Martin, D.C., "Interpretive

Report on Weld Metal Toughness," SSC-169, Ship Structure
Committee, July 1965.

Burner, W.K., "Accurately Specified and Controlled Fillet

Weld Size in Ship Hull Construction," Coast Guard Engineers
Digest, Jan., Feb., Mar., 1975.

Masubuchi, K., et al., "Analysis of Thermal Stresses and

Metal Movements of Weldment," A paper presented at the Annual
Meeting of SNAME, New York, N.Y., November 13-15, 1975.

Vreedenburgh, C., "New Principles for the Calculation of

Welded Joints, Welding Journal, 33 (8) ,1954.

Macfarland, D.S., Harrison, J.D., "Some Fatigue Tests

of Load Carrying Tranverse Fillet Welds", British
Welding Journal, Vol. 12, No.12, Dec. 1965, pp. 613-623.

Swannel, P., "Deformation of Longitudinal Fillet Welds

Subjected to a Uniform shearing Intensity", British
Welding Journal, Vol. 15, No. 3, March 1968.

Report of the Welding Institute Research Laboratories,

"Effect of Peening and Grinding on the Fatigue Strength
of Fillet Welded Joints", British Welding Journal, Vol.
15, No. 12, December 1968, p. 601-609.

Solumsmoen, O.H., "Fatigue Tests on Specimens with Holes,

Butt and Fillet Welds in Mild and High Tensile Structural
Steels", Metal Construction, Vol. 1, No. 3, March 1969,
p. 138-142.

Butler, L.J., Kulak, G.L., "Strength of Fillet Welds as

a Function of Direction of Load", Welding Journal, Vol.
50, No. 5, May 1971, p. 231-s-234-s.

Clark, P.J., "Basis of Design for Fillet Welded Joints

Under Static Loading", The Welding Institute, Conference
on Improving Welded Product Design, Paper No. 10, 1971.

Kato, B., Monta, K., "Strength of Transverse Fillet

Welded Joint", Welding Journal, Vol. 53, No. 2,
February 1974, p. 59-s-64-s.
Maddox, .j,"An Analysis of Fatigue Cracks in Fillet
Welded Joints,", International Journal of Fracture, Vol.
11, No. 2, April 1975, p. 221-243.
 -47-

Blodgett, O.W., "New AISI-AWS Allowables", Welding Journal,

Vol. 49, No. 8, Sept. 1970.
Nippon Kaiji Kyokai, "Study on Hull Damage Related to Hull
Defects", Report of the 109th Shipbuilding Research
Committee, Japan.
Cochran, C.S., and Jordan, C.R., "In Service Performance
of Structural Details," SSC-272, Ship Structure Committee, 1978.

Antoniou, A., "A Survey on Cracks in Tankers Under Repairs",

PRDS Symposium, Tokyo, 1977.
Masubuchi, K., Materials for Ocean Engineering, pp. 369,
M.I.T. Press, May 1970.
"Effect of Peening and Grinding on the Fatigue Strength
of Fillet Welded Joints, British Welding Journal, Vol.
15, No. 12, December, 1968. pp. 601-609.
"Fabrication, Welding and Inspection of Non-Combatant Ship
Hulls," Department of the Navy. Naval Ships Engineering
Center. December, 1966. NAVSHIPS 0900-014-5010.

McCabe, W., "A Comparison of Fillet Weld Strength and U.S.

Navy Design Specifications for Non-combatant Ships and
The Economic Implications", MIT Engineers thesis, Dept.
of Ocean Engineering, May 1978.
Malliris, A.P., "Static Strength of Fillet Welds Using the
Finite Element Method," MIT M.S. Thesis, Dept. of Ocean
Engineering, Sept. 1978.
 r'

-48--

APPENDIX I

LIST OF LITERATURE ON FILLET WELDS, 1943 - 1977

Appendix I is a list of literature on fillet welds from

Northeast Acaemjc Science Information Center (NASIC) which is
available at M.I.T.

Example 76

0.1 Subject

Serial Number of the

Year Published (1976)

Subjects are classified as follows:
S: Static Strength
F: Fatigue Strength
R: Residual Stress
D: Weld Defect
I: Inspection
C: Welding Cracking
P: Welding Process
7700lD Antoniou, A.C. A Survey on Cracks in Tankers Proceeding of the PRADS-Inter-
Under Repairs national Symposium, Tokyo,
Japan, Oct. 1977

77002D Kaku, S. Recent Tendency of Hull Struc- Proceeding of the PRADS-Inter-

tural Damages and Their Counter- national Symposium, Tokyo,
measures Japan, Oct. 1977

76001S Gurney, T.R. Finite-Element Analyses of Some Weld. Res. Vol. 6, No. 4
Joints with the Welds Transverse mt.
to the Direction of Stress

76002F Maddox, S.J. Fracture-rIechanics Analysis of Weld Res. Vol. 6, No. 5

the Fatigue Behavior of Fillet mt.
Welded Joints
Can. Mach. Vol. 83, No. 3 L
760035 Design of Welded Joints, Part 2
Metal-Work

76 00 4FD Japan Ship- The Research on the Brittle- Jap. Ship- SR-153
Building Fracture and Fatigue Strength building Res.
Research of Thick Welded-Plate with Inst.
Institute High Heat-Input Welding Pro-
cesses (in Japanese)
Schweisstech- Vol. 30, No. 2
76005R Malisius, R. Shrinkage and Residual Stresses
in Welded T and Cruciform nik
Joints (in German)

760061 Webber, D. Problems in the Detection and Welding Institute Conf. on the
Maddox, S.J. Monitoring of Fatigue Cracks Detection and Measurement of
in Al-Zn-Mg Alloy Fillet Welds Cracks, Abington, Cambridge

76007D Lamba, H.S. The Effects of Clustered Poro- Construction CERL-TR-M-196

Cox, E.P. sity on the Shear Strength of Engineering
a 514F Transverse Fillet Welds Research Lab.
(Army)
76008D1 Antoniou, A. Fabrication Factors Affecting Report of Committee 111.3
et al. Structural Capability of Ships ISSC, 1976
and Other Marine Structures
75001F Maddox, S.J. Analysis of Fatigue Cracks in Internat. J. Vol. li, No. 2
Fillet-Welded Joints Fracture
75002F Mathers, E. Fatigue of Stainless Steel Conf. on W1ding Low Tempera-
Fillet Welds at Low Temperatures ture Containment Plant, Weld-
ing Institute, Abington, Cam-
bridge
75003F Fatigue Strength of Transverse Engineering ESDU-75016
Fillet and Cruciform Butt Sciences Data
Welds in Steels Unit Ltd.,
London
75004C Chew, B. Diffusion of H in Fillet Welds Metals Tech- Vol. 2, No. 2
no log y

75005C Araki, M Cracking in Multipass Fillet Nippon Kokan No. 18

Nagae, T. Welds Tech. Rep.
(over sea)
75006P Brayton, W.C. Shipyard Welding--Applicability Maritime Ad- MA-RD-920--
Evance, R.M. of Firecracker Welding to Ship ministration, 77040
Naister, R.P. Production Washington, DC
75007S Burner, W.K. Accurately Specified and Con- Coast Guard Jan-Feb-Mar,
trolled Fillet Weld Size in Engineer' s 1975
Ship Hull Construction Digest
74001S Vidmir, M. Static and Dynamic Load-Bearing Vanna Technica Vol. 23, No. 3
Capacity of Al Welds
740025 Hot Tap Connection on Gas Pipe- Gas Engineer- Vol. 14, No. 10
lines Operating at High Pressure ing Management
 Strength of Transverse Fillet Welding Jour- Vol. 53, No. 2
74003S Kato, B.
Monta, K. Welded Joints na i

74004F Heins, C.P. Design Procedure for Fatigue Federal High- FHWA/MD/R-76/12
Yamada, K. Due to Daily Traffic way Admin.,
Maryland Div.

74005F Nunn, D.E. An Investigation into the Fa- Transport and TRRL-LR- 629
tigue of Welds in an Experi- Road Research
mental Orthotropic Bridge Deck Lab., Crow-
Panel thorn, England

74006D Sato, K. Study on Treatment of Fillet J. Soc. Naval Vol. 136

Ueda, Y. Weld with Root Gap (in Japanese) Arch. of Japan
Tanaka, T.
Seo, K.
Tunenani, T.

73001S Beichuck, G.A. Influence of Depositing Fillet Welding Pro- Vol. 20, No. 2
Naletov, V.S. Welds on the Shape, Microstruc- duction
Orekhov, V.P. ture and Fatigue Strength of
Welded Joints

73002S Sager, R.J. Designing for Welding Aluminum Welding Seminar Tech-
nical Papers, The Aluminum
Assoc., New York

73003C Fukui, T. Lap Joint Fillet-Weld Cracking Sumitomo Light Vol. 14, No. 4
Sugiyama, Y. Tests of Al Alloys (in Jap.) Metal Tech. Rep.

73004C Araki, M. Cracking in Multi-Run Welds Nippon Kokan No. 61

Magano, T. (in Japanese) Tech. Rep.
Harasawa, H.
Flange-to-Web Connection Re- Weld. J. Voi. 52, No. 3
73005C Sherman, D.R.
Fisher, J.M. quirernent on Beams with Corru-
gated Webs
730061 Method for the Radiography of Schweissen Vol. 25, No. 5
Fillet Welds in Welded-in and Schneiden
Welded-on (Pipework) Connections
730071 Dubreson, J. Non-Destructive Testing of Fil- Sound. Tech. Vol. 27, No.
Evrard, M. let Welds Connexes 1 and 2
Lepenven, Y.
72001F Mori, M. Application of Program Fatigue Mitsubishi Jul., 1972
Matoba, M. Test to Member Joints of Hulls Tech. Bulletin
Kawasaki, T.
Nakajima, M.
Hirose, M
72002C Clark. P.J. Basis of Design for Fillet- Improving Welded Product De-
Welded Joints Under Static Load- sign Conf., Welding Institute
ing
'J'
720031 Francois, C. Detection of Cracks Using Acous- t\)
Sci. Tech. Vol. 46, No. 4 i

Nouvet, A. tic Techniques in Fillet-Welded Armement

Tanaka, T. Steel Sheets During Strain Tests
71001S Nishida, Y. Strength of Fillet Welds in V- Sumitomo Search May, 1971
Shapes
71002S Butler, L.J. Strength of Fillet Welds as a Weld. J. Vol. 50, No. 5
Function of Direction of Load
71003F How Explosive Peening Affects Metal Con-Vol. 3, No. 10
Fatigue Properties in Maraging struction and
Steel, Fortiweld and an Al-Zn- British Weld.
Mg Alloy. Pt. 1, Experimental J.
Details and Results
71004C Boniszewski, T. Association Between Service Metal Construe- Vol. 3, No. 12
Failure and Fillet Profiles of tion and Bri-
Tube Stub-Header Weldmerïts tish Weld. J.
 Krakovyak, M.F. Ultrasonic Inspection Technique Weld. Prod. Voi. 18, No. 9
710051
Grebennik, I.L. for Fillet Welds in Smali-
Diameter Tubes

Ishiyama, K. Blowholes of Fillet Weld Joint J. Mech. Eng. Vol. 25, No. 5
71006D
Nakamura, Y. (in Japanese) Lab.

7l007D Fujita, Y. The Strength of Fillet Welded J. Soc. Naval Vol. 130
Hagiwara, K. Structures with Misaligned Architects of
Fujino, H. Members Japan
Hashimoto, H.

7 00 0 iS Strength of Welded Joints of Schweisstech- Vol. 20, No. 12

Wrought Mg Alloys n 1k

70 002S Investigations of the Geometri- Schweissen Vol. 22, No. 5

cal Shape of Butt and Fillet Schneiden
Welds

What Designers Should Know About Weld. Des. Vol. 43, No. 4
70003S
Welding Aluminum Fabr.

70004F Archer, G.L. Fatigue Strength of Mild Steel Metal Constr. Vol. 2, No. 5
Fillet Welded Tube to Plate Brit. Weld J.
Joints

70005S Kassov, D.S. Selecting Safe Limiting Stresses Weld. Prod. Vol. 17, No. 2
Reiderman, Y.I. in the Calculation of Fillet
Luvshits, M.G. Welds
Kutepov, Y.N.

Dikum, V.N. Automatic HV Fillet Welding With Weld. Proc. Vol. 17, No. 3
70006P
Chernov, Y.A. Powder-Filled Wire
Pelevin, Y.P.
Duben, L.V.

70007P Persson, H.A. Utilization of the Penetration Acta Polytech- APS-ME-51

Gredborn, K.E. for Fillet Welds nica Scandina-
via, Stockholm,
Sweden
70008P Persson, S.A. Vertically Welded Fillet Welds Acta Polytech- APS-ME-5l
nica Scandina-
via, Stockholm,
Sweden
70009S Blodgett, O.W. New AISI-AWS Allowables Weld. J. Vol. 49, No. 8
6 9 00 iS Influence of Joint Thickness on ZIS MITT Vol. 11, No. 9
the Static Strength of --Steel--
Fillet Welds
69002F Fatigue Tests on Specimens with Metal Constr. Vol. 1, No. 3
Holes, Butt and Fillet Welds Brit. Weld. J.
in Mild and High-Tensile Struc-
tural Steels
69 00 3C Lamellar Tearing in Multi-Pass Weld. J. Vol. 48, No. 9
Fillet Joints Ui

69004C A Fractographical Examination Metal Constr. Vol. 1, No. 2

of Lamellar Tearing in Multirun Brit. Weld. J.
Fillet Welds
69005P Archer, G.L. Fatigue Strength of Mild Steel Metal Constr. Vol. 2, No. 5
Fillet Welded Tube to Plate Brit. Weld. J.
Joints
69006P Recommendations for Arc-Welded Soud. Tech. Vol. 23, Nos.
Joints in Clad Steel Construc- Connexes 9 and 10
tion (in French)
68001F Effects of Peening and Grinding Brit. Weld. J. Vol. 15, No. 12
on the Fatigue Strength of Fil-
let Welded Joint
68002R S. Wanell Deformation of Longitudinal Fil- Brit. Weld. J. Vol. 15, No. 3
let-Welds Subjected to a Uni-
form Shearing Intensity
 Proposed Working Stresses for Weld. J. Vol. 47, No.
68003S Higgins, T.R. 10
Preece, F.R. Fillet Welds in Building
Construction
Influence of Weld Thickness on Schneiden Vol. 19, No.
67001S Feder, D. 7
the Static Strength of Side
Fillet Welds
Ultrasonic Inspection of Fillet Prezegl Vol. 19, No.
670021 9
Welds in Lap Joints Spawalnictwa

Deformation and Strength of End Tokyo Univ. Vol. 28, No. 7

66001S Naka, T.
Fillet Welds Fac. Engineering

The Load Testing of Double Strap Schweissen Vol. 18, No. 7

660023 Bornscheuer, F.W.
Joints Made from Grade ST 37 Schneiden
Steel Having Flank and Frontal
Fillet Welds
Eng. J. of AISC Vol. 3, No. 3
66003S Abolitz, A.L. Plastic Design of Eccentrically
Loaded Fasteners
An Investigation of the Relia- Oregon State Sept., 1966
660041 Berger, D.S.
Herrala, J. bility of Prequalified Fillet Highway Dept.
Paasche, O.G. Welding Procedures for Welded
Steel Bridges
Proposed Modification of the ISO Lastechniek Vol. 32, No. 6
66005S Van Douwen, A.A.
Witteveen, J. Formula for the Calculation of
Welded Joints
Some Fatigue Tests of Load Brit. Weid. J. Vol. 12, No.
65001F MacFarlane, D.S. 12
Carrying Transverse Fillet
Welds

Fatigue of Structural Elements Naval Applied 13 Sept., 1965

65002FR Cordiano, H.V.
Werchniak, W. Development of Theory and Mea- Science Lab.,
Silverman, B.S. surement of Residual Stresses Brooklyn, NY
at the Fillet Welds in 1-1/2
Inch HY-80 Steel
65003F Reemsnyder, H.S. Fatigue Strength of Longitudi- Weld. J. Vol. 44, No.
nal Fillet Weidments in Con- 10
structional Alloy Steel
6400lS Partial Projections Caused by Nippon Kokan No. 31
Fillet Welding on the Surface Tech. Rep.
of Hull Plating
63001S Lord, O.S. Tensile Strength of Steel Con- Georgia Inst. of Tech. Atlan-
Schutz, Jr., F.W. nections Having Transverse and ta Engineering Experiment
Longitudinal Fillet Welds Station, 22 Jan., 1963
6300 2P Development of Techniques for Naval Applied Science Lab.,
Manual Fillet Welding on Thick Brooklyn, NY, Tech. Memo
Plate Titanium Alloy 1924H2 USGRDR6516
60001S Blank, G.F. Ultimate Strength of Butt and General Dy- GDA-7E--2 367
Fillet Welds of ASTM 203-D namics, Astro-
and USS T-1 Structural Steels nautics, San
Diego, CA
5600lS Lois, H. Stiffeners in Welded Plate Weld. J. Vol. 35, No. 1
Campus, F. Girders
54001S Vreedenburgh, New Principles for the Calcu- Weld. J. Vol. 33, No.
C.G.J. lation of Welded Joints 8

53001S Koenigsbeger, F. Load-Carrying Capacity of Fil- Weld. J. Vol. 39, No.

Green, H.W. let Welded Connections 9
490015 MacCutcheon, E.M. Rivetted vs. Welded Ship Weld. J. Vol. 28, No.
Structure 2
46001S Leavitt, C.M. Proportioning of Weld Sizes Weld. J. Vol. 35, No.
2
45001S Boardman, H.C. Stresses in Welded Structures Weld J. Vol. 24, No. i
 Weld. J. Vol. 23, No. 8
44 00 iS Spotts, M.F. Stresses in Fillet Welds with
Eccentric Loads
Weld. J. Vol. 22, No. 2
430015 Fergusson, H.B. Strength of Welded T-Joint
for Shipst Bulkhead Plates
Weld. J. Vol. 22, No.
43002S Mackusick, H.H. Fillet Welding Applied to Ship
12
Hiatt, J.B. Construction
Anderson, R.V.

z
o
o
M
M
r)
M

'a
o