Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
B y D . F. S O C I E A N D S . D . A N T O L O V I C H
ABSTRACT. Fracture toughness, a low temperature high toughness consists of austenitizing at 1700 F
fatigue, stress corrosion and cor- steel. These structural steels are used (927 C) for one hour and water spray
rosion fatigue were studied in both in a quenched and tempered condi- quenching. The plates are then
the base metal and HAZ of an ASTM tion so that the necessary balance of tempered at 1100 F (593 C) for one-
A 5 3 7 grade steel w h i c h was welded strength and toughness is obtained. half hour.
w i t h heat inputs of 25, 50 and 75 During welding, the metal next to the In order to obtain reliable measure-
k J / i n . A uniform HAZ w a s main- w e l d joint is affected by the heat re- ments of the properties of the HAZ
tained perpendicular to the side quired to melt the base metal in order the crack must be kept entirely w i t h i n
planes of the specimen and parallel to insure a good weld to base metal it. Most studies using conventional
to the crack by welding a " K - t y p e " bond. This heat can alter the micro- " V - t y p e " weld joints result in a non-
joint of different section sizes. The re- structure of the heat-affected zone uniform HAZ w h e r e the crack is free
sults of this investigation indicate (HAZ) and significantly change both to run through the base metal, HAZ,
that the subcritical f l a w g r o w t h prop- physical and mechanical properties. or filler metal depending on the
erties of this steel tested in 31/2% Welds often contain small f l a w s from geometry and external stress field
aqueous NaCI are not degraded rela- w h i c h cracks originate so that the (Ref. 4). To avoid this, a HAZ that is
tive to those of the base metal over properties of the material near the not only straight but also perpendic-
the range of welding conditions stud- w e l d zone are of considerable i m - ular to the specimen surface was
ied and that this steel can be used for portance. produced using a " K - t y p e " w e l d joint
marine applications in t h e welded The primary causes of crack g r o w t h s h o w n in Fig. 1. The w e l d s w e r e
condition w i t h a high degree of con- are cyclic stresses and aggressive en- made by the submerged arc process.
fidence. vironments. Linear elastic fracture Three groups of weldments using
mechanics concepts have been suc- heat inputs of 25, 50 and 75 k J / i n .
cessfully used to describe the g r o w t h were made w i t h Armco W - 1 8 elec-
Introduction
of cracks (Refs. 1-3). It is of value trodes and Linde 709-5 flux.
Many large engineering structures, then, to determine the crack g r o w t h Fracture toughness tests were con-
such as ships, bridges, and offshore characteristics under both static and ducted on the base metal using 0.5 in.
drilling platforms are fabricated by dynamic loading in an aggressive en- (12.7 mm) and 2.0 in. (50.8 mm) thick
some type of welding. One of the vironment such as sea water. W i t h a compact tension specimens s h o w n in
most c o m m o n modes of failure is the good stress analysis, life calculations Fig. 2. These tests as well as all other
growth of a small flaw or crack to a can be made for a structure contain- mechanical testing were done on an
critical size w h e r e catastrophic fail- ing cracks by using the crack g r o w t h MTS closed loop testing system in
ure occurs. The purpose of t h i s in- rates determined in the laboratory. accordance w i t h the recommended
vestigation was to provide some use- procedure set forth in ASTM specifi-
ful design information and to deter- cation E399-72. The specimens were
mine the effect of welding on the sub- Experimental Procedure
fatigue precracked and fractured in
critical flaw g r o w t h characteristics of The steel used in this investigation monotonic loading.
is covered by ASTM Standard A 5 3 7 - A tapered specimen as s h o w n in
67, "Carbon-Manganese-Silicon Fig. 3 was used for corrosion tests
D. F. SOCIE is associated with Structural Steel Plate, Heat Treated for Pressure and is described in detail elsewhere
Dynamics Research Corp., Cincinnati, Vessels". The chemical composition (Ref. 5). This specimen was used be-
Ohio and S. D. ANTOLOVICH is with the and mechanical properties are listed cause the stress intensity is inde-
University of Cincinnati, Cincinnati, Ohio. in Table 1. The thermal treatment pendent of crack length and only a
2.5(KQ/o-ysp
268-s | J U N E 1 9 7 4
or no effect on the fatigue crack substituting into eq (4) for A K, one ob- evaluated from the toughness at frac-
growth rate of this material. One of tains ture data obtained from this investi-
the unique features of fatigue is the gation by substituting K Q into Equa-
formation of striations although the da tion 3.
exact mechanism is not well defined. = C(Aof> (7ra)2
dN (5) Kr
Figures 7 A and 7B show striations
formed at lower stress intensities integrating 5.1 in.
w h i l e 7C shows branch cracking that
is more typical of higher stress inten- N, a, Now the equation for cycles to failure
M
sities. da can be solved w i t h the result:
(6)
Stress Corrosion and C(A<*) 4 T 2 a 2 N f = 4 . 8 x 10 B cycles
Corrosion Fatigue
In principle the above calculations
Corrosion tests were carried out in can be carried out for any load/crack
solving for cycles to failure N f , one
a 31/2% aqueous NaCI solution. The geometry providing the stress inten-
threshold stress intensity below obtains:
sity parameter is k n o w n and a good
w h i c h there is no crack g r o w t h was
determined by fatigue precracking a N,
1
4 2
va< (7) stress analysis is available.
C(Aff) 7r afa0 Conclusions
tapered specimen and then holding it
at a constant load w h i l e monitoring
the crack length. The threshold stress The fracture toughness, fatigue and
The final crack length ( a f ) can be corrosion fatigue properties of an
intensity w a s found to be 4 0 ksiv'Tn.
(43.6 M P a \ / m ) or about one half of
the toughness.
The results of t h e corrosion fatigue
tests of the base metal are graphical- Table 2 Fracture Toughness Data
ly s h o w n in Fig. 8. There is n o statis-
tical difference between tests con- B, W,
ducted in air and 31/2% aqueous NaCI in. in. lb a/w
at 75 F ( 2 4 C). Similar tests were .500 1.000 3800 .50 73
conducted on the 5 0 k J / i n . w e l d .500 1.000 4200 .48 74
samples w i t h the same result (i.e. no 500 1.000 3100 .61 86
difference for tests in air and 3V2% Average 77 ksivMn.
aqueous NaCI). There is a large differ- (85 MPavrrT)
ence between tests conducted in 2.000 4.000 31,500 .54 85
31/2% aqueous NaCI at 75 F ( 2 4 C) 2.000 4.000 15,000 .71 88
and 3 0 F (-1 C) indicating that the 2.000 4.000 39,000 .49 90
process is thermally activated. Such a 2.000 4.000 24,000 .57 74
process might be dislocation move- Average 84 ksiVin.
ment in the plastic zone ahead of the (92 MPaVm)
crack tip. The 4 t h power model can be
modified to include temperature as
follows:
da 4
.750
T l . l x 1Q--3 e x p / " O O P V | (AK)
dN
(2)
w h e r e R is t h e gas constant and T is
the absolute temperature in degrees
Kelvin.
A p p l i c a t i o n of R e s u l t s
Perhaps the best w a y to demon-
strate the usefulness of this type of re-
search is to illustrate a typical life
calculation. Let us assume that a
large plate such as that on a ship hull 5.875 1.720
has a center crack 0.5 in. long and is
subject to a stress varying f r o m 0 to
20 ksi. The stress intensity factor for
this crack can be expressed as (Ref.
6):
K=o-s/~T~a~ (3)
Dimensions in inches
where
o" = stress 1 in = 25.4mm
2a = crack length
4.875
The crack g r o w t h rate is given by:
5.375
da ,
(4) Fig. 3 Tapered specimen used for corrosion tests
dirC(AK)4
WELDING RESEARCH SUPPLEMENT! 269-s
.500 DIA
Thickness
.150 i n
(3.8mm)
':" 3SL$&tiff
* > ,,'i
J* T
_ 11111 1 1 1 1 I I I I ' -'*>,
Z A Base Metal
O 25 kJ/in.
D 50 kJ/in.
O 75 kJ/in.
' ' 1 V * \ V.'T j * "
^f
MB
'TT fi $ 'y TIT'; \\f'/ ^ - ' \> ^
c|>.
-tlrj AK = 30 KSI ."in
\K = 20 KSI . i n
(22 MPa /m) (33 MPa /Si
11111 J I I I I 1 111
AK KSI /In
_J I I I i I ' 'I
100 ..K 50 KSI . i n
AK MPa /m
(55 t'.Pa /Si
Fig. 6 Comparison of crack growth rate for welded and
unwelded steel Fig. 7 Fatigue striations formed at various cyclic stress intensities
270-s | J U N E 1974
ASTM A 5 3 7 steel welded by the sub-
10 merged arc process at heat inputs of
1_ I I I I
2 5 , 5 0 and 75 k J / i n . has been evalu-
D Air 75F ated. Tests were carried out in the
base metal and HAZ. The main con-
O 3.5% NaCI 7 5 F clusions are:
10 A 3.5% NaCI 30F 1. The fracture toughness w a s rela-
tively independent of specimen thick-
ness over the range in w h i c h this
steel is used.
2. The fatigue crack g r o w t h rate
was independent of heat input or the
location of the crack (i.e. base metal
or HAZ).
3. The crack g r o w t h rate in 3V&%
10 aqueous NaCI was highly temper-
ature dependent, decreasing by al-
most an order of magnitude w h e n the
temperature was reduced f r o m 7 5 F
75F (24 C ) t o 3 0 F(-1 C).
-4 4. This steel can be used for
10 (24C)
marine applications w i t h a high de-
gree of confidence. A w i d e range of
o heat inputs does not appear to signif-
c l >, icantly reduce the fracture properties
u r H| o
of the HAZ compared to the base
metal.
iTJl Z io Z
al -a al -d
A ckno wledgement
10
The Authors would like to thank the
Armco Research Center of Middletown,
Ohio for supplying the material and weld-
ing the plates. We are also pleased to
acknowledge the advice and assistance of
-5 Mr. G. Linnert.
10
References
The American Welding Society publishes over a hundred different Codes, Standards,
Recommended Practices, Guides, Handbooks and other books related to welding. If you
would like a copy of our Publications List, simply circle No. 151 on our Reader Info-Card
and mail it.
WELDING RESEARCH SUPPLEMENT! 271-s
WRC Bulletin
No. 184
June 1973
WRC Bulletin
No. 185
July 1973
The purpose of this project, sponsored by the Pressure Vessel Research Com-
mittee of the Welding Research Council, was to investigate means for obtaining
improved characterization of the size, shape and location of subsurface dis-
continuities in metals. This objective was met by applying computerized data-
processing techniques to the signal obtained in conventional ultrasonic pulse-
echo systems. The principal benefits were improved signal-to-noise ratio and
resolution.
The price of WRC Bulletin 185 is $3.50 per copy. Orders should be sent to the
Welding Research Council, 345 East 47th Street, New York, N.Y. 10017.
272-s | JUNE 1 9 7 4