View
5
Download
0
Category
Preview:
Citation preview
Lehigh UniversityLehigh Preserve
Fritz Laboratory Reports Civil and Environmental Engineering
1970
Fatigue strength of welded A514 steel beams,Preliminary Publications, Fatigue of WeldedStructures Conference, Brighton, England Vol.1,Paper 9, 1971John W. Fisher
Follow this and additional works at: http://preserve.lehigh.edu/engr-civil-environmental-fritz-lab-reports
This Technical Report is brought to you for free and open access by the Civil and Environmental Engineering at Lehigh Preserve. It has been acceptedfor inclusion in Fritz Laboratory Reports by an authorized administrator of Lehigh Preserve. For more information, please contactpreserve@lehigh.edu.
Recommended CitationFisher, John W., "Fatigue strength of welded A514 steel beams, Preliminary Publications, Fatigue of Welded Structures Conference,Brighton, England Vol.1, Paper 9, 1971" (1970). Fritz Laboratory Reports. Paper 408.http://preserve.lehigh.edu/engr-civil-environmental-fritz-lab-reports/408
· Low-Cycle Fatigue
FATIGUE STRENGTH OF WELDED
A514 STEEL BEAMS
by
John W. Fisher·
r:'j'\JiI1t-f_
I .. , ,j\.:,!j\!Lt,F!!f'J'G
Li81l:f\[-?1"
Th{s paper was prepared as part ·of a study of LOW-CycleFatigue, sponsored by the Office of Naval Research,'Department of Defense, und~r Contract N 00014-68-A-514i.NR 064-509. Reproduction in whole or part is permittedfor any purpose of the united States Government.
Fritz Engineering Labo.ratoryLehigh University
Bethlehem, Pennsylvania
---April, 1970
Fritz Eng~neering L~boratory Report No. 358.16
•
J
..
TABLE OF CONTENTS
SUMMARY
INTRODUCTION
FATIGUE STRENGTH OF COVER-PLATED BEAMS
F~TIGUE STRENGTH OF WELDED BEAMSWITHOUT ATTACHMENTS
STRESS ANALYSIS OF CRACK PROPAGATION
SUMMARY AND CONCLUSIONS
ACKNOWLEDGEMENTS
REFERENCES
. TABLES AND FIGURES
.' Rage
1,
'2
6
8
.10
13
14
./
SUMMARY
This' paper summarizes the findings of a comprehensive
study on the fatigue strength of rolled and welded built-up beams
without attachments, rolled and welded beams with cover plates~, and
welded beams with flange splices. Altogether 374 steel beams with
two or more details were fabricated and tested.
Only two of the beam series are discussed herein. Em-
phasis is placed on ASTM AS14 steel although comparisons' are also
provided for other grades of steei. T~e beam types discussed herein
represent the upper and lower boundaries of fatigue behavior of "
welded beams. The lower bound is provided by:beams with partial, ..
le~gth cover plates - a severe notch producing detail. The upper
bound is provid~d by the plain-welded' beam - a minimum notch pro
ducing detail.
For purposes of design, this study has shown that the fa
tigue strength of the up~er and lower bound details is independent
of ~he strength"of steel. A36, A441 and AS14 steel beams provided
the same fatigue strength for a given detail, and stress range wa~
~bserved to ~ccount for ,nearly all the variation in cycle life~
2.' ·
In previous studies failure to properly control and measure
the variables influencing the f?tigue strength was usually the major
rea~on for the app~rently conflicting and contradictory claims on
the significance of stress variables and material characteristics.
The empirical exponential model relating stress range to
cycle life was obs~rved .to provide the best fit to the test data
for all beam series. A theoretical stress analysis based on the
fracture mechanics of crack growth substantiated the empirical ex
ponential model that provided the best fit to the test data. In
a~dition, the theoretical analysis was used to rationally'explain ,
the observed behavior of the experimental results.
I-NTRODUCTION
In the present design (1970) of steel bridges for fatigue,
pr~visions have been specified in many instances on the basis of
limited test data. Equally im'portant to the fatigue life ~f highway
bridges is the, significance of such factors as the loading history
to, which the structures are subjected,- the type of materials used,
'the" design details',. and the quality of fabrication.
Recognition of "these facts lead to th~ bridge studies 'of
the AASHO Road Test (1) and the development of a comprehensive study
of rolled and'welded built-up beams with a variety of welded details.
" "This program was sponsored by the American Association of State High
~ay Officials in cooperation with the Bureau of PublibRoads, u.s. '
Depar-tment of Transportation" under the National Cooperative Highway
•
3.
Research Program which is administered by the Highway Research Board
of the National Academy of Sciences. ~'(
Objectives and Design Variables
The major objective of this study was to develop suitable
mathematical design relationships for between 50,000 and 10 million
cycles of loading. Altogether, 374 beams with one or more details
were studied. Previous work had not adequately investigated the be-
havior of beams in terms of str~ss, detail and type of steel. The
effects of variables $llch as stress, stress ratio, cover-plate geom-
e~ry, details and type of steel were ~ot clearly defined (2,3,4).
The principal design variables for this study were grouped into three
categories: (i) type of weld detail, (ii) stress condition, and
(iii) type of steel.
Although three grades of steel were examined - ASTM A36,
A441 and AS14 - the" primary emphisis in this paper i~ on AS14 steel.
Comparisons- are made with the other steels where appropriate.
Four different types of beams ~were -examined including cover-
plated beams, plain-welded beams withqut 'any attachments, welded
"beams with groove-welded flange splices and plain-rolle4 beams (5).
Only two details are discussed in this repor~. One detail' is a rolled
or welded steel I beam with cover plates'which provided a severe
notch producing detail and a lower-bound to fatigue behavior. The
* This· study was conducted under National Cooperative Highway ResearchProgra~ Project, 12-7. The opinions and findings expressed or > implied in this paper are those of the author. They. are not necessarily those of the Highway Research 'Board, the National Academy ofSciences, the Bureau of Public Roads, the American Association ofState Highway Officials, nor of the 'individual. states participating~in the National .C~operative Highway Research Program.
4.
second detail discussed is the plain welded beam without attachments,
which provided a minimum notch producing detail and an upper bound
for the behavior of welded beams.
Details are given in the Final Report of the study on the
effect of weldments on the fatigue strength of steel beams (5).
Figure 1 shows schematically the basic details considered
in the cover-plated beam study. Cover plates were attached to both
rolled and welded beams~ All A514 steel beams, had'11.4 em (4~ in)
wide cover plates which were 1.5 times the flange thickness. Other
cover-plated beams were examined with thicker, wider, or multiple
cover plates, but all these were fabricated from A36 steel.
The plain-welded beams were identical in cross section to
the cover-plated beams and were fabricated using the same technique.
All longitudinal fillet welds were made by t~e automatic
. submerged-arc proc~ss. Tack welds and the transverse end welds on
the cover plates were manual welds. Al~ beams in this study were
320 em (10 ft.-6 in.) long and were tested on a 304.8 em (10 ft.)
span. The ends of the cover plate details were positloned in· the
shear spans ·30.5· em (12 in) from each load poi~t. The plain-welded
beams were loadect so that a 106.7 em· (42. i·n.) constant moment region
resulted in the center.
Minimum stress and stress range were selected as the con
trolled stress variables. This permitted v~riation in one va~iable
while the other'was maintained at a constant level. Had stress ratio
(ratio of minimum to maximum stress) been selected as the independent
5.,
variable, both minimum stress and maximum s,tress would ha've to be
changed simultaneously to maintain the ratio at a constant level.
~he controlled stress levels were the 'nominal flexural
stresses in the base metal of the tension flange at the end of the
cover plate, and at points of maximum moment for the plain-welded
peams.
pxperiment Design
Each beam series was arranged into factorial experiments.
The basic factorial for each steel wi~hin ea~h beam· type was defined.by the stress variables as illustrated in Table 1 for the cover-plated
and plain-welded AS14 steel beams. The same factorial existed for
e~ch type of steel and comparable. factorials existed for the other
types of beams and details.
All factorials with cover-plated beams had at least three
specimens assigned to each cell. This permitted the variance of each
cell to be estimated., Each cell of the 'plain-~elded beam factorial
had two specimens or replicates assigned, since it was considered that
more than one fatig,ue crack was probable between the load points.
None of .the experimental factorial~were complete. Partial
. factorials for each series were developed because of known boundary
conditions. The maximum values of stress had to be l·imited to the
yield stress of ·the material~ The lower .values of stress range were
not examined at all values of minimum stress because of excessive
lives. At least 10 million cycles---'\vere applied before'.- te-sting---w-as---~--'--'-"'-·"-
discontinued and a fat~gue'limit was assumed'to be reached..
6.
The fact that all beams were fabricated symmetrically
about their neutral axis provided information on the behavior of
bOth -the tension and compression flanges..
The order of testing of the specimens for each beam type
was.randomized so that the effect of uncontrolled variables such as
temperature, humidity, laboratory and testing personnel would be
random. The randomization of the uncontrolled variables allows the
e~fect of 'these variables to be included as a random error in the
analysis and prevents any systematic bias due to these effects on
the controlled variables.
FATIGUE STRENGTH OF COVER-PLATED BEAMS
. .The results of tests on cover-plated beams provided data
for a severe notch producing detail so that a lower bound to fatigue
strength could be examined. When a transverse end wel~' existed, the
crack initiated at the toe of the transverse weld as illustrated in
Fig. 2. During the first 'stage of growth, the crack grew through the
flange ~n an elliptical shape. Thereafter, it grew toward the flange
tips and into the web. 'Cracks initiated at the ,toe of the longitud
.inal fillet welds connect,ing the cover pl~te to the beam at, ends where
no transverse end weld was preser:tt. These c~acks also grew through
the flange in a·n elliptical shape and were·"similar to the first· stage,
of growth exhibited by cracks at the end with, a transverse end weld.
Th~ effects of the controlled variables of minimum stress,
stress raIfge and type of steel were analyzed using statistical methods •
. The domin~nt variable was stre.ss range for all cover- plate geometrie,s,
,end details and steels ,te"sted."
7.
Figure 3 summarizes the test data for AS14 steel beams with
an end weld. Cycle life is plotted against stress range for differ
ent ·levels of minimum stress. Also shown are the mean regres~ion
line and the limits of dispersion as given by two standard errors of
estimate. It is apparent that the variation due to minimum stress is
insignificant and that stress ~ange accounted for the variation in
c~cle life. The mathematical relationship between the applied stress
,range and cycles to failure.for each series and geometry was deter
mined using regression analysis. The analysis showed that the loga
rithmic transformation of stress range and cycl~ life proyided the
best fit to. the data.
·The effect of typ~ of steel was also evaluated since the
experiment design provided equal factorials and sample sizes of A36,
A441 and AS14 steel beams with cover plates. The test data for all
three types of steel cove~-plated beams are compared in Fig. 4. Th~
AS14 steel beams yielded only a slightly longer life. The" variation
in life 'due to type of steel is too small for consideration in the
" design of structures.'
. The results of thi~ study have. been c~mpared with the earlier
work of Wilson (6), Lea and Whitman (7),.and Munse and Stallmeyer (8)
for both end details. The limits of dispersion provided by two stand
ard errors of, e~timate included almost all the data. Most of the
test- points falling below the lower limit of dispersion were from the
early 5tu~ies of Wilson (6).
This study has confirmed that no great differences exist in
o
8.
the fatigue strength of square. ended cover plates; that cover
plates affect rolled and, welded beams similarly; that welded cover~
·plated beams yield about the same fatigue streng~h for A36, A44l and
AS14 steels, and that only stress range is the critical stress var-
iable. Greater detail is given in Refs. 5 -and 9.
FATIGUE STRENGTH OF WELDED BEAMS WITHOUT ATTACHMENTS
The results of the tests on plain-welded beams provided a
minimum notch producing detail and an upp~r bound to the fatigue
'strength'of welded beams. Nearly all cracks initiated at a flaw in
the fillet weld ,at the flange-to-web connection as illustrated in
Fig. 5. The fillet weld flaw was usually a gas pocket or worm hole
in. the fillet weld caused by gas trapped in the weldment. Cracks
were initially'inside the weld, but grew out to the fillet weld sur
face and were analogous to. stage 1 of crack growth at. the end of a
,cover plate. After penetrating the outside fibers of 'the flange, .
the. crack grew on two fronts toward the flange tips which was com
parable to stage 2 ,in the 6over-plated beams.
Most of'the AS14 steel beams showed a clear tendency to
form multiple cracks in the tension flange. ~any cracks also oc
curred in'the compression flange·with more cracks formed with a de- .
creasing minimum stress.
The significance of the design factors, minimum stress and
stress range are illustrated in Fig. 6. S'tress range is observed to
account for the variation in cycle 1i£e4 Multiple regression an~lysis-
9.
also indicated that the logarithmic transformation of both stress-
range and cycle life provided the best fit to the data.
Residual stresses were measured in several of the welded
shapes and all indicated the presence of large tensile residual stresses
in the vicinity of the flange-to-web fillet welds. As the strength
of the steel increased, there wa~ greater probability of th~ compres-
sian flange being subjected to the full tensile stress range in the
vicinity of the weld since the residual stresses were about equal
to the yield stress.
As was the case"with cover-plated beams the experiment de-
"s~gn prov~ded equal s~mple s~zes ~f plain w~lded A36, A441 and A514
steel beams. The test data for all three types are identified in Fig. 7.'!
There was no statistically signif~cant difference due to type of steel.
All the variation was due to stress range.
The test results were" compared wit~ the previous studie~
, .by Gurney (10) and Reemsnyder (11) on we"lded beams fabricated by auto-
matic welding procedures. "The ,test data indicating failure at acci-
dental start/stop positions fall between the lower limit of dispers~on
and the mean. Beams with cracks initiating at flaws in the fillet
~eld fall between the mean and th~ upper liffift of dispers"ion. Beams
not failing at w~il defined weld flaws were similar to plain rolled
beams and tended to provide the longest life. The AS14'steel beams
and ~-specimens tested by Reemsnyder fall near the upper limit ,of .dis
persion as might be expected considering. their careful fabrication
and the fact that they were only subjected to constant stress over a .
o
21.3 em (8 .in.) length which reduced the. probability of a large
flaw withing the maximum stress region.
Details of the experimental work and comparisons with the
earlier work are given"in Refs. 5 and 12. Although not discussed in
this summary, work was also done on plain-rolled beams and welded
beams with the reinforcement removed from groove-welded flange splices ..
STRESS ANALYSIS OF CRACK PROPAGATION
. For the prediction of macrocrack propagation the relation
ship proposed by Paris (13) was used. The relationship
C LlKn = da/dN (1)
expresses the change in crack length, 8, for a sinusoidal stress
cycle N as a function of the change in K during that cycle and a
constant of proportionality. The K value for a crack can be ex
pressed in terms of the tfcorrected crack en as
K = cr /TTC. (2)
Since K is determined both by crack geometry and the nominal stress
and can be derived analytically, it provides a means of determining
the influence of geometry and nominal stress upon the stresses at the"
"crack tip. Sig~es et al. (14) have shown the fatigue cracks at the
toe of- fillet welds start from small cracks at the weld toe •. These
cracks exist before cyclic loading is applied to the joint. The
'to
. 11.,
applicability of Equation 1 has been illustrated for many types of
materials and geometric configurations (13,15,16).
Equation 2 can be substituted into Eq. 1 and integrated be-
tween the limits of applied .cycles Ni and Nf corresponding to the
values of the corrected crack size at initiation (ci
) and failure (cf ).
Neglecting dc/da this yields
(3)
where ~cr = AS is assumed constant and equal to the product of ther .
stress concentration factor, A, and the applied stress range, S ,r
n/2 n nAI = Aa.n A, liN = Nf-Ni
'. and a. = 2' - 1. The value of the corrected
failure cra9k, cf
' is large and can be neglected since a >0. The
relationship between life ~N and the applied stress range is expo-
,nential in form which -agrees with the results of the' regression
analysis.
The exponent of stress range, n, was observed to vary be-
tween-2.80 and 2.87 when cover-plates were attache~ to A36, A441
and AS14 steel beams. Hence, a variation in yield stress from 25'. 2
. kg/mm2 (36 ksi) to' 70 kg/mm2 (100 ksi) caused a negligible change
in exponent. The variation in mean exponent between the cover-plat~d. .
beams and the plain-welded beams was from"2.80 to 3.33. A value of
n = 3' was selected since it provided a reasonable fit "to the experi
mental data. No plasticity correction was used for computation of
the corrected crack size c.
\.< - 12.
Equation 3 could then be expressed as
(4)
T~e crack growth in the cover-plat~d beams was analyzed by Frank and
Fisher (5,9). They observed the cracks to be semi-elliptical in
shape with a ratio of a to b (see Fig. 2) that remained constant and
equal to 2/3. The crack size at different numbers of cycles for the
unwelded end of the cover- plate was determined by measuring the size
at several stages. The stress intensity factor K for a semi-ellip-
tical surface crack as developed by Irwin (17) was used along with
the more accurate secant correction ~or finite width (l~). Equation 3
was used to evaluate the parameter AT and the corrected initial crack
· -1/2 Th· · Id d -1/2 rv 130/' rr:; and A· f 2 -7/ 3 ~s~ze c. • 18 Yle e c. = v 1n . = 1.0 xlO ksi vin.1 1
Substituting these values into Eq. 4 gives the number Qf :cycle_s"'fQr'.~.. .
the crack ,to propagate through' the flange ~s
9 -3!JaN·= 1.28xlO Sr(5)
EquationS is compared' with the mean regression line for
beams with end welded cover plates in Fig. 8. The predicted crack
.growth is in good agreement with the experimental-results. Also
shown in· Fig. 8 is the mean regression curve for plain-welded beams
together with the exponential model using t-he exponent n~3.· The us,e
of n=3.is seen to give good correlation for both the upper and lower
abounds of the fatigue. strength of welded beams.
-_.. _- ~ - --'-- -'- __..- ~ ~.- __ ._- -_ .. ..__~.__- ,~ _ _~ - ._. . . .. . .. e:t . .. . '.. 0
~ .The-··quant'ity 6f (171\ t) (c: ) ·'i-s-a-- iiie-asure '-of"-t1ie--hbtt~h o-erf-e"ct-'1. .
of each type' of specimen and detail. This quantity may be eX,pressedas
..13.
(6)
It is therefore inversely dependent upon the initial corrected crack
size, c., the constant of crack growth, A, and the stress concentra1.
tion factor, X. The cover-plated beams represented the most severe
condition of these parameters.
SUMMARY AND CONCLUSIONS
This report summarizes the results of a study on the fa-
tigue streng~h of welded beams. The study has determined the signif-;
of several design factors manner for firsticance in a rational the
time. -The conclus ions are based on the analysis and evaluation of
the experimental data, a study and correlation with earlier work, and
on theoretical stMdies based on the application of contin~m mechanics
to macro,crack propagation. Details of the study are given in Refs. 5,
9 and 12 .
. 1.. 'Stress range was the dominant stress variable, for all welded de-
tails and beams tested.
2. Other stress variables were not significant for design purposes •. '.
'3. Structural steels with 25.2 kg/mm2 (36 ksi) to 70 kg/mm2 (100 ksi)
'did not ,exhibit significantly different fati~~e strength for a
given ~elded detail.
14.
4. The logarithmic tran~formation of cycle life resulted in a normal
distribution of the test data at nearly all levels of s~ress range
for ,each welded beam detail.
5~ A theoretical stress analysis based on fracture mechanics for
macrocrack propagation substantiated the experimental model that
provided the 'best fit to, the "test data.
ACKNOWLEDGEMENT,S
The research summarized in this paper was performed under.
NCHRP Project 12-7 by the Fritz ,Engineering Laboratory, Department
of Civil Engineering, Lehigh University, Bethlehe~, Pa. and the De
partment of Civil Engineering, Drexel University, Philadelphia,- Pa.
Lehigh 'University, was the contractor for this study which commenced
in 1966.
. .
The preparation of this paper wa~ supported by the Offi'ce
of 'Naval Reseanch, Department of Defense, under contract N00014-68-A-OS14;
-NR064-509.
The author is indebted to his colleagues Karl H. Frank and
Manfred A. Hirt, Research Assistants and Ben.T. Yen, Asso~iate Pro
fessor at Fritz Engineering Laboratory for their ~ssistance and work
throughout this study. Thanks are also due Fred Schmitt, Research
Assistant and,Bernard M. McNamee, Professor of Civil Engineering at
Drexel University.
.'
REFERENCES
1. FISHER, J. W. and VIEST, I. M.Fatigue -Life of Br~dge Beams Subjected to Controlled TruckTraffic, Preliminary Publications, 7th Congress, lABBE,pp •. 497-510, 1964
2._ GURNEY,T. R.Fa~igue of Welded Structures, Cambridge University Press, 1968
3. MUNSE, W. H. 'and GROVER, L. M.Fatigue of Welded Steel Structures, Welding Research Council,New York, N. Y., 1964
4. ASCE TASK COMMITTEE ON FLEXURAL MEMBERS,-Commentary on Welded Cover-plated Beams, Journal of the StructuralDivision, ASC~, Vol~ 93, No. ST4 August 1967
5. FISHER, J. W., FRANK, K. H., HIRT, M. A. and McNAMEE, B. M.Effect of Weldments on the Fatigue Strength of Steel Beams, FinalReport, NCHRP Project 12-7, Fritz Laboratory Report No. 334.2Lehigh University, Bethlehem, Pa., September 1969
6. WI,LSON, W. M.Flexural Fatigue Strength of Steel Beams, Bulletin No. 377, Univ.of Ill., ,Urbana, Ill., '1943
7. LEA, F. C. and WHITMAN, J. G., The Failure of Girders Under Repeat,ed Stresses, Welding JournalVol. 18, January 1939
8. MUNSE, W. H. and STALLMEYER, J. 'E.Fatigue in Welded Beams and Girders, Bulletin No. 315, HighwayResearch Board, 1962
·9. FRANK, K. H. and FISHER, J. W.The Fatigue Strength of Welded Coverplated Beams, Fritz Lab Report334.1, Lehigh University', Bethlehem, Pa.; March 1969
10., G,URNEY, T. R.Investigation into the Fatigue Strength of Welded. Beams, Part II:High Tensile Steel B~ams Without Stiffeners, British Welding
, Journal, Vql.' 7, 1962
, - 11. ~EMSNYDER, H. S.Fatigue Strength of Longitudinal Fillet Weldments in Constructi~nal Alloy Steel, Welding Journal, Vol. 44, October, 19Q~ .__
12. HIRT, M. A. and FISHER, J. W.The Fatigue Strength of Rolled and Welded Steel I-Beams,Fritz Laboratory Report No. 334.3, Lehigh Univepsity,Bethlehem, Pa., March 1970
13. < PARIS, P. C•The Fracture Mechanics Approach to Fatigue, Proc. lOth SagamoreConf., Syracuse Univ. Press, p. 107, 1965
14. SIGNES, E. G., BAKER, R. G., HARRISON, J. D., and BURDEKIN, F. M.'Factors Affecting the Fatigue Strength of Welded High StrengthSteels, British Weldin~ Journal, Vol. 14, M~rch 1967
15. JOHNSON, H. H. and PARIS, P. C.Subc,ritical Flaw Growth, Engineering -Fracture Mechanics', Vol. 1,No.1, June 1968
16. GURNEY, T. R.The ~ffect of Mean Stress and Material Yield Stress on FatigueCrack Propagation in Steels" Metal Construction and BritishWelding Journal, Vol. 1, No.2, February 1969
17. ,PARIS, ,P. C. and 8IH, G. C.Stress Analysis of Cracks, STP No. 381, ASTM, 1965
'18. IRWIN, G. R., LIEBOWITZ,. H. and PARIS, P. c.A Mystery of Fracture Mech<:tnics, Engineering Fracture Mechanics,Vol. 1, No.1, June 1968
I ":. -.,;
TABLE 1
'a. EXPERIMENT DESIGN FOR COVER-PLATED BEAMS
. kg/rnm2S . Stress Range S. mln r (ksi)
kg/rom2 5.6 8.4 11.2 14.1 16.9(ksi) (8) (12) (16) (20) (2~)
--4.2 CRC131 CRC141 CRC151(-6) CWC132 CWC142 CWC152
CWC133 CWC143 --CRC144 --
1.4 CRC221 CRC231 CRC241 --'( 2) CWC222 CWC232 CWC242 CWC251
CWC223 CWC233 CWC243 --CRC234
7.0 CRC311"': '1 CRC321 CRC331 CRC341I
(.10) CWC312 CWC322 CWC332 CWC342CWC313 CWC323 CWC333 CWC343
CRC324 CRC344 .
* CR = Cover Plate on Rolled BeamCW = Cover Plate on Welded Beam
b. EXPERIMENT DESIGN FOR PLAIN WELDED BEAMS
kg/rnm2'8 • Stress Range S. m1n r (k:;;i)
kg/nun2 12.7 16.9 21.1 ... 25.3 29.5
(ksi) (18) (24) (3q) (36) (42).-
- ':"7.0 . PWC131 FWC14.1 PWClSl(-10) PWC132 PWC142' PWC1S2.----1.4 PWC221 PWC231 PWC241,( 2) 'PWC222 PWC232 PWC242
...
9.8 PWC311 PWC321 PWC331 PW·C341(14) PWC321 PWC322 PWC332· PWC342
;; .
·a
. ,.-.
-----------
---------
'No \'teLd~ or- cae\::\\,fe,\.c\s. on tn~s. e.nc\
,s.erle~
C.B Ser~es
Gmm (V4-1n,)) I 1
----------..---------------~_tl''''·j-----_tl~- - ...... - - - - - - --)rll----------------------...
- f'-"! = -= ::'-==-
\4-.mm (e:JAG \'0,) 10\ C1Z)C.~ ~ cw w~es
i l'Jmm (~/+ ill.) 1tr C.T ~e,\~e..~
__ \4 V'F ~O -1-0\ CI2 }e.g -$ CT Ser\e~
We\.de.d ~hd.pe'£. ~r C\'-! S<2r\es.
CJ~( mrn (~c;j tn.) - \2 \~ mo. ( 43 t'n. ) '3'3 \ mm (~'.:l in.) _
'3'200 ITP1\ (\Lb )'n.)
Fig. 1 Details of beams with cover.platesattached to each'flange
~
.Stqge -I
2b'TrO'rfsvcrseFillet \Veld
•
.:/'/' '.' 'llf-~' ." '. .
i 4
1;;· "." ; ~
: " ~~" ~ ."
~ ",' •••• ~.-.-....-...,;.-....~..-:...........,_:--_'.~."'.:u:.:-'~~''''''''''~"~':'';'''::'''~':~_''' o';.'j h2·. ,~.,., ..~~,: •. ;,,',.~;, :,~.. ,;,.:....,..;.:>;' .~.~:
:. .. "Fig·. 2 Crack growth at the transversely welded end ofcover-plated beams
It,"j J 'IIi.. .
CKs\). .
't:~/mm'Z, .'
~o
''lS
, ",
" 0 •
#
.0
%~Q: Coo"dence. ,.um~\:
. . .4-0\ 15ro' ~Ltr'''\''~\s
. '...
·0·0' L~netzeCfre~~\.Mean ~oJ'. .
A .... ,. '" :
.......... ""'A....o~ '0-~ '. . .'.......... """"- ' a.... ~J)~ ..............
' '. ~ ..........co C1~~. ~, .. '.. ':. · --.................... o..aol9~6.~ ........
. , ~~~ .~.~, ,
"".~ ......... c ·~.......o. OO~ ............' ~ ,""",C" ............
' '~-.. , ~ .......... 0,y----.......... , -.........::/ "-0
, ..........
\0
5
30.'7.5
. \~
\
, 'lO'
gO
40
7.0
o \0
~l
. 0 0
Sll2.~SS
~ '1ZAN&~ . \S
£:.: SM:M'· 'Netd'(:!d ~9\.\.ed:.
'.-'G' . '6 ,,'. ~
e.~2·
\0 ..
o
.. "'0. '.
o
s. l ...
t
:j;lO·~s}\.0O,S
'\' , I , 't t I I I I I f I I I I I I I I J• f •
o 0\01· 0,', - \ .. C\'CLE~ TO f"A\ LU2'1;: " (\06)
,# ..• Fig. 3. Effect' 'of str~5S range and minimum stress' on the cycle:: :life for the 'welded end of AS14.steel cover-plated beams
., j ~
t . • ~~ _ ~ L ~ ....
~o
~~mm'2. .
... ' ~
; .
"
.
\ .
, I'
Scee\.'., -'
.6 .~.~ A~G .,',
0" ~A44\
o . A5l4'
\0 ~,
•Z5~"
I . ''20, . ........0 .. '.
A~6~Oq,-0"""""0 O. , , " ,........ C>' -OO~Cb~~'. ' .. ,~t-. -"""'1\"'" 6./~{JiJp·_.':::,):J"" Q n-o.,........... .~._ / . .. .'
' ~ ~ ~-tJ ~\~ )'nd:] CCJ 0"",,-- _ , .
. . "'-....... &eP;.{i~@~ d .~.~~'~~, ~~~oo 0 00--........ , ,
--........ '-v """" ........ .. ~. \~........ , ,
............:: '~" 0"'?~ 0~'L\6. a~, .' L\mic. / "'" ",-0 00 . ~"'-.......Co f1dence -.........:.. ...........,<15% n '. ~\...s ....................... 0'-{or, '7so;o ~ur\J\ v -......--....· 5
\0
s
1
,
(\0:.\ )I
· .sO
4()I
0
!>o.-
-, ~s:
srQ.S:SS rzo~Nt,;e.
\5
-..1
,
\, ,':'U-LLL.O.~CJ 0.\, 0::; \.0 _ '. ' 5 ' \0
-- .\ " C'\CLES TO FA\L\),R~ (\0<0)'Fig. 4 Effect of grade, of steel on the fatigue strength
of 'cover-plated beams,
'Stage I
.•. -:,'{
-~-~~; .": - '", - ~
. ~', ~.
t.' :;.:Z? >;~;, '.: :.
;.
1.·}~~::<T\~~.. :<:' ,:.
/ ,I\.,
. I ~:~\~, ,:~.:~;~~_,... ;
:;.:. ~ I - ~:-.1!. 1.~ t~~,. +
".
'i
~;.• ~::'" M"~;:"-;:.J
• +: t".~ ~ ...... ~11.""~ tlt,.....!~ - ,~, ~1
S'
"Stage 2
:,_':.{/~.
~t~~·1ji:~fi%.?~~t~~i0~iyj·j~jj
• ~ r-
Fig.' 5~ Crack growth in welded beams
."
!
L~Y)e "Mean ~€l(\re~~on
\
..,. t ~ ... : ~~, .... ~.:
, 1S,~Co'"~~~ce 'l,Y'I\ ; t-tor' e;5()~ 'St.trv'v C),\~
I '
. &~ .............••.. ........A~ &6. ,/ 0--~ --........-..
':~. c.............. .1:1. 06b.co::tLi~~08 .1J 0 ~ ~~ '""""'--- A 0, ~o-....~ rJ:) O:?O .
. --... --..... , <9 raret" ........... '. e- '
------- --........-.. . . --..;~ . '.............. ~. 0 a~ .. / '. ----..o~
, ., ,~
, \0\S
f
•;
(\::~~ ) .
. . • ISO
'40,•• O.
<;.TQESS··· .3.°r .. rzO"~S
~NGE·. ,7.0
\0
s
. '?
", .
... "
. , \, ~
. ,0' ., . '
CJ
...
", S"M\N' ,
,-\0"',,,
rz :"\4 ·
',~, .
."
I
...·1 .
' ..
\050- .. ,::::.
\ I I I I , I I I I I
C\ \ I IV. I !., I I 1 I
t,O
c.'{c.'-1C:~ '\0 F'A..\.LU12~ (\0&)
Fig. 6' 'Effect ·of'.~tress· rartge and min,imum stress on the fatigue: strength' of ~elded beams ..
.... "• 'j
...,
Mean . l2e~t"es.s;:\on· .Une
-,.
\0
~~/mtr\~
'lo,.
~ " ..........1\. 0 @ ~/~... ~~ 0........................... \~O ~~ --.y'-cf---... . Cb O~r;g_o.. <Q 0 """--..... """--.....~ . .
. """--.....~ 6.0 ~~nS)O 6'20 """--..... ~ ,
---... -......... -ct. --..........~ At:::r-""""--- " ~ ~ .'-......~ "'0 O~ ./4 "'-.....O~, 't'c\~ ~ce L\fn\
%1C ~1-' ~rvl"a\S~O~ cqs ~'O'
z;
! .
'Z£
flOI ''.. ·lS
sO
. \S
I,
i
(~\),50
40r' •-,
. ST12£S~. ~N.c;E
. '.\
, \0~I '_. "
.t';',
5.J-'
..·t
s .. ,"
..~\
. ,. \ -, " I I I , If' , . I I I , I , t I
" (}'\ . O,S to . 5 ' \0
.' .
C'(ClES 10 1=A\ lUe. E (\06 )Fig.'7 Effect of grade o~ steel on the fatigue.
strength ~f welde~ beams f .
_-:J________________________________________ l ~ --,__ "'------"_._. ._I__~.__~~ ........"""""""""""'.",,,,,,,."',,
'J. _ '1!"1,
(ks;) , I . 'Z .~~fmrn
~,~
n=-3
E~. ~
/
•
~ °0 ft~ 0°0••• •O t"':"'.-~ ..,o-~,u oco
C) d-'~'<";--() """f,tll~'"'O0000 (';"J (;..,,) '.,lo~.....~Q J'Q., 0 oe '.
~ .'.. e·co 0 ••' ~ ~0 o~<N,Q;}~a 0 ....,. .~~~O~.. .. .d
'Nek\ed' . 0~ ............... ~'... •~n .. d Deam~ ...................... • _"~~.o~oo~..Dl 'c.~ D . ~ 0 0 o./'r'ller - 'I 0' .....O~ ...~ .
L.v" '" ,~o~.......................... -......;:
~ - OJ, ,., .--~"':'--'. IP ° 'lr--~ .. ',cb'aCClJa:I~.,~ Cbo.c,"""-............... °° ~~., ~
, ............... ~-"""'f"'['[J:;l l....U°-......... -......... ~~tJ'~ , .
. ............. --... / -d~CJ:::-..........--"'-o.g,..., ~
'Ne.\c:jed ~e.am-s
s
7.0, ~o.__ 0 :. '
~~ooo •-~~ ~ '0.o v.:;, ~ "0"... .., 0"--00_0.~ ~ "'-;)~~o""'~-~iO\LI- ~. oO<::QVi\~o''';''
IJ ~"oo... [;g.. ~,
.\0
"30
rz.s
\0
'20
40.
· 30
25s,-,z.~s~
~K&E:
b
. '\
Mean '£e~ teg,,~, on \.\ne
,s% Con~c\ence. LIm;\:~\ c:l~"lo ~ur-v\Vd.\s
..
1_-"Q-:\ 0,5. \\0 5 \0 .
(.
F'ig. "8
C,(C\.:~.~ TO ~A\ LU R..E: (\oG)Mean fatigue strength and 95% tonfidencelimits for welded and cover-plated beams i
iJ-
Recommended