Iß CO l-H 9 D6- 2: i )78 June 1969 The Effect of Heat Treatment on the Fracture Toughness and Subcritical Crack Growth Characteristics of a 350-Grade Maraging Steel C. S. Carter D > BoeingCommcrcial Airplane Group Renton, Washington Sponsored in Fart by Advanced Research Projects Agency ARPA Order No. 878 TMl ducument has been approved for public release and sale; its distribution is unlimited. CLEARINGHOUSE for Federal Scientific & Tec. Informstian Springfield Va. 22151
Transcript
Iß CO l-H
9
D6- 2:i)78
June 1969
The Effect of Heat Treatment on the
Fracture Toughness and Subcritical Crack Growth
Characteristics of a 350-Grade Maraging Steel
C. S. Carter
D >
BoeingCommcrcial Airplane Group Renton, Washington
Sponsored in Fart by Advanced Research Projects Agency
ARPA Order No. 878
TMl ducument has been approved for public release and sale; its distribution is unlimited.
CLEARINGHOUSE for Federal Scientific & Tec. Informstian Springfield Va. 22151
,,,
THIS DOCUMENT IS BEST QUALITY AVAILABLE. THE COPY
FURNISHED TO DTIC CONTAINED
A SIGNIFICANT NUMBER OF PAGES WHICH . DO NOT
REPRODUCE LEGIBLY.
I III. EFFECT OF- III AT TREATMENT ON THE FRACTURE TOUCHINESS AND SUBCRITICALCRACK CJROWTII
CHARACTERISTICS OF A 3S0<tRADE MARA(ilNC, STEEL
by
C. S. Carter
ABSTRACT
The heat treatment response of a 350-grade maraginf steel, with the nominal composition 18.5 Ni, 12 Co. 4.6 Mo, 1.4 Ti, balance le. has been determined in billet and bar form. When ■ged at temperatures below 90(rF, the material was very susceptible to subcritica! crack growth, and premature brittle Iracture occurred in unnotched tension specimens loaded at a slow strain rate in laboratory air. Fracture mechanics was used to interpret this behavior. The introduction of reverted austenite signiticantly decreased the strength level but had little effect on Iracture toughness. The resistance to brittle Iracture of this material is contrasted with that ot high-strength steels currently used by the airplane industry.
INTRODUCTION
F'he continuing demand lor ultra-high-strength steels prompted the recent development ol two 3S0-grade maraging steels. The development and properties of one alloy nominal composition P.5 Ni. i:.5Co. .VK Mo. 1.7 Ti. 0.15 Al. balance le were recently described ( I ). Vasco Metals, a subsidiary of I eledyne. Inc.. introduced a second alloy that is lower in titanium than the first to avoid segregation problems and has slightly higher
nickel and molybdenum contents. Fn this paper, the effect of heat treatment on the fracture toughness and subcritical crack growth resistance of this second alloy is described.
MATERIAL
Samples from two heats of consumable-electrode, vacuum-remelted. 35()-grade maraging steel were supplied by Vasco. The form and chemical analysis of these samples are shown in Table F. Most of this study was conducted on the 4-in.-square billet.
EXPERIMENTAL PROCEDURE
All specimens were machined in the longitudinal direction unless otherwise noted. To determine tensiFe properties. (J.25-in.-diameK r tension specimens were loaded in an Instron machine at a strain rate of 0.005 in.'in./min. fatigue precracked, standard-si/e Charpy specimens were loaded in three-point bending to determine the plane strain Iracture toughness K|c; the data were analyzed using the recommended \.S IM procedure (2t.
Similar tpecimeni were used to establish the subcritical crack growth characteristics. These were deadweight loaded in cantilever bending
uMiii! ihe tcclinii{uc üescrihcd In Urown(3).The environmcnl was either laboratorj ail «T 3.5 aqucoiiN sodium ehloridc solution, with the latter eontinuousl) Jrippcd mii> the specimen notch. Stress intensity was calculated Irom the relatiDiiship ^iven hj Drown and Srawiey(4)b) assuming thai cantilever bending corresponded to three-point loadinti with .1 span-to-width ratio ol eight, Ilk' results ol the u-Ms were plotted as the initial applied stress intensit) level k^ versus time tu i.nluu'. and the plane strain4 threshold stress intensity level K|v. . beiovt vslnch crack arowth il KI not occur, u.is determined. Fatigue preeracked single-edge-notched specimens 7.S in. long, 1.5 in. wide, and 0.48 in. thick were used to determine the crack velocit) characteristics. I he crack length was measured optically, usnii; low-power magnification, at selected tune intervals. From th se observations, a curve of crack length versus tune was constructed, I he slope ol the curve, winch represents the instantaneous crack growth rate, was determined graphkaH) for various crack length values; from this, the relation between stress intensity and crack velocity was determined.
lo determine the amount ol austemte. the X-ray technique of Lindgren was used (5). Ilns technk)ue compares the integrated intensities of the 1 2JO» auslernte line and the (200) maftensite line and is claimed to give an accuracv o| |'/f.
RESUL TS AND DISCUSSIUIM
s 3b0 r
■■n 1/)
Z 300 -
TENSILE
YIELD"
•o
S^
z o - F "*, Q < LU u. ir o
^ 60
0
-.40
RA
X -L. -L.
Z 1 ~
or u
Fig-
IbOO U'.OO 1700 1800 1900
FIRST ANNEALING TEMPERATURE (0F)
/. Effect of first .innealnig temperature on mechaiiical properties (second jnnnal at 1500eF 1 hr and aged 900^/8 fir).
350 r I 8 g? 300 Ui ■
250
z- * 2<50
Q < HI» r. CC O 0 ^- Um _L.
50^
40 50
- 30
i;i50 1400 1450 1500 1550 I6O0
SECOND ANNEALING TEMPERATURE '0Fl
Fig. 2. Effect of second annealing temperaturn on mechanical properties (first aniwal ai
/ 700° F/1 hr and aged 900° F/8 hr I
< c:
EFFECT OF SOLUTION ANNEALING TEMPERATURE
Ihe effect ol double solution annealing on the properties of the J-m-square billet was examined, and the results are shown in I igs, I and 2, Ihe
I ii si annealing temperature hail no significant
eltecl on tensile properties except that t he
reduction ol area was less when the annealing
temperature exceeded I TOO" I dig. 1 I. This
could be associated with grain growth, i racUire
toughness Kj.. however, was independenl ol annealing temperature \s indicated m I ig, 2, the second annealing temperature should exceed I 1NI°I Vnnealing below this temperature gave lowei tensile properties since some austemte W.J-
present. Ihereloic. lor the remamdci ol tins
Study, the specimens were douMe solution
annealed at 1700 I loi 1 hr followed by IM)t)0l toi 1 hr.
Ihe specimens were sufficiently thick that plane strain conditions existed ( 2).
EFFECT OF AGING
The luirdncss behavior of the 4-in.-square billet ■fed at various temperatures for times up to 100 hr is shown in Fig. 3. The curves are typical of precipitation hardening alloys, with maximum hardness being achieved alter aging at the lower temperature for a longer time.
The elTect of aging temperature on the tensile properties and Iracture toughness ol tlie same material is shown in Fig. 4. Significant dilterences in behavior were observed between specimens aged at less than l>0()0 I and those aged at higher temperatures. Because of this, the two temperature regimes are discussed separately below.
60 i-
1 LU z a < i
100
AGING TIME (HR)
Fig. 3. Effect of aging on hardness (double solution annealed }700'F/1500'F).
Aging Ik-low ()Ü0° F
I'remature brittle fracture occurred during loading of the tension specimens aged at XOü* F lor periods of 3. 6, and 8 hr. Similar behavior was exhibited by the specimens aged at BSflT F for 3 and 6 hr but not alter aging lor B hr (Fig. 4). Failure usually occurred in the gage section, normal to the loading axis, but in a few tests a threaded end of the specimen broke. Those that failed in the gage section exhibited a bright, thumbnail-shaped region of fracture at the specimen circumference (Fig. 5). The chevron pattern in the remainder of the Iracture face indicated that rapid brittle Iracture initiated from this site. The dimensions of this region were measured from enlarged photographs, and a K|c
value was determined for the initiation of brittle Iracture using the stress intensity solution lor a surlace-cracKed plate ((>!. Values for individual specimens are shown in Table II. A similar region was also observed at the thread roots in specimens that failed at the ends; k|c values were not determined for these specimens. This behavior did not depend on the position from which the specimens were removed from the billet and was also observed in similarly heat treated specimens taken from the l-.V4-in. by 3/4-in. bar.
350 r
325
I LU X I- V)
TENSILE 2=
YIELD 300 fpREMATURE 275;BRITTLE
'FRACTURE dp 350"
325 ^
^ 300 -PREMATURE
275 BRITTLE a<Ic ,FRACTURE ?b0^
TENSILE
225P",/
350-
275
300PPREMATURE 'BRITTLE ;FRACTURE
2b0^ B- 225
YIELD
.□-O
800 850 900 950 1000 1050 1100
AGING TEMPERATURE (»F)
Fig. 4. Effect of aging on mechanical properties (double solution annealed l/OO'F/IBOO'FJ.
Fig. 5. h'iision spi\ imtn lh.it pn'tti.iturrly KIIIIHI in
Lihurdtory dir iJuedSSO'F 3 hrj.
Iiul the thumhn i i was due
I ension specimens i1 ii jithei
tested in twu addilionul V,.I\N lit one U'si scru spceiiiKMi! « :w rupidij lensioi IOJ Ii ü to
fuilurc m laborutor) aii al ,1 crosshead travel speed ol 2 in min us compared wit\
min in the tests discussed above. Mi th< specimens tested in this wa) in .1 ductiK mannei withy cup-and-eone-typt fracture (Table III 1. tiMis cont'inning thai the ra» k * er« nol preexisting but only developed altei loading foi .. di 1 iinu period ol lime. In 1 »nd series ol k-stv i'4 in ol the 1.25-iii-IOICJ pane section of the tension specimens were surrounded witli Seattl tap water and tin »pecimens tension loaded ii> failure al .1 crosshead travel speed ol 0.005 in. min. Premature brittle fracture occurred in the Kclion expoied to lap watei ;ii .1 limilai stress level and in an Identical manner to the specimens tested m air. I hese results, included in
J.ihli' II fnctun datM ><>r prtntttuniy f.iilt'd tension specinwm,.
\ ting treatment
I nvironmenl 1 racturc siicsv
(ksil
Critkil etat k lizc ( HI 1 ki ( ksiv III. 1
Depth 1 CIlL'll,
Slid0 l (.In t aburatui \ JH 215.1 0.(147 (1.107 64.'
MMl't Si,,
1 iburati 11 > .in
232.1
24 7.0
::.vs
0.030
(i.()4s
0.068
0.073
O.I20
0.1 SO
S8.< 1
79^1 80-' 70.1
1 ip watci 2M.6 0.034 0X162 (.:.. |
sin r 1 16 hi 1 iiburatur) 111 238.7 0.010 0.(1^1 43.7
SMTf 3 III
I aboiator) air 291.0
:i5.o
0X130
0.05!
0.07(1
0.102
7l.iil „,, (mean
64.5)
1 .ip watci 25M (MCI (I.Oss (>().(! 1
Table III. Mechanical properties of tension specimens tested at a fast strain rate.
Aging trc.it meiit
Yield strength (ksi)
I Itimate tensile strength (ksi)
hlongation {'%) Reduction of
area {%)
«Ü0oh7« hr 850*1 3hr
299.0 2(>7.4
302.3 290.7
9.6 11.1
45.7 52.9
Pabk II. siiii^csk'il that water (or water vapor in
the airl was involved in the crack extension process.
hraneli on a microscopic scale dig. 7). Possibly
these microscopic branch cracks were sufTicient
to dissipate the applied load attd ellectively blunt
M icroexamination of sections taken Irom
specimens exposed to both environments (air and
waten revealed that the thumbnail regions had
extended m i predominantly intergranular manner. In some specimens, seondary
intergranular cracks extended parallel to the
loading axis along dark etching bands, indicating that these discontinuities had very low crack
growth resistance. Previous studies have shown
that these bands are associated with nickel,
molybdenum, and titanium segregation although the cause of the dark etching appearance is not
load dig. 6). I his was eliminated by increasing the loading rate (erosshead speed increased trom
0.05 in./min to 2 in.'mint. In specimens
exhibiting crack growth. K|c was computed from
the maximum load and the distance Irom the
surlace containing the notch to the position
where subcritical crack growth terminated. Values
determined in this way were sigmlicantly higher
than the K|e, obtained by rapid loading (Table
IV). Ihis was unlikely to be a strain rate effect
because other investigations have shown that the
tracture toughness of maraging steel is essentially
independent of strain rate (S).
Metallographic examination of the subcritical
crack growth region revealed an intergranular
mode of cracking with the cracks tending to
FRACTURE ' .■.■'^f, ' ■^>' . ■>? . ' . , i'jTOV
■
X6
Fig. 6. Subcritical crack growth in a notched bend specimen tested under rising load in laboratory air (aged 800" F/3hr).
'tA*
X4J0
Fig. 7. Microscopic crack branching in a notched bend specimen tested under rising load in laboratory air (aged 800* F/8hr).
1 ■
SI,.«'1 ' (mean \9 v \ SL2)
( rack i'i iwth
Fa,tb 30 B No crack jr< •* 11
Mllfl |h 1,1 Slow3 42.6 (rack growth
asfff \ hi Slow1
4l>.s |
4N.4 ( (mem
72.0 I (,-,,'H>
B4.8 1
Crack growth
ft** 30.5 No crack gmwtli
'( rotthead ^p^.'^.•d 0.05 in. nun Crouhead spoi-J 2 in, mm
ilk' crack tip. I ins would be most apparenl in hriltie materials \Mtii sin.iii plastic zones. In loughet materials the lanier plastic zone would override the influence <>i the crack morphology, \ comparison ol Pahles II and IV indicates that ihc Ki values obtained from tin; tension .nul
K NIUVSK loaded notched bend specimens were similar. I here was KMIM icatter in both test leries, which is attributed to variations in crack morpholog)
fo i|uantitativel) measure the resistance to crack growth, precracked Charpj specimens were cantilevei loaded to selected stress intensity levels in laborator) air and the lime to failure established, fhe results, shown in Fig, 8, indicate thai crack growth did not occur below .i stress mtciisiu level ol ISksisin, fhis information can be used to explain the mechanics of fracture in tin; tension specimen*.
lo estimate the depth ol tl.iw required foi cracking during tension loading»<> typical fracture siiess value of 230 ksi (lai^le Hi and a threshold
value of IS ksiNin. were substituted into the stress Intensity equation tor ■ surlaee crack(6). Uns led to a critical crack depth of 0.001 in., for
a crack with a lengtlhtoKlepth ratio ol 10 However, the fracture stress was used rather than the lower (unknown) stress at which crack growth initiated, so the actual critical depth would he less than 0.001 in.
As discussed previously, the dark etching bands were very susceptible to crack growth. Since these bands were oriented parallel to the loading axis, and then thickness was typically 0.OO0S to 0.002 in., they probably provided the initiation site. The short failure tunes of the bend specimens loaded lo stress intensity levels above 20 ksKin. indicate a negligible incubation period and rapid crack propagation, thus explaining the extent of crack growth observed in the tension specimens (Table II).
Resistance I" crack growth was similarly determined in }-\,'2''' aqueous sodium chloride solution (ITg. S), Threshold stress intensity was
reduced to 5 ksiv/in. in this onvironment altliougli the tunes to failure at stress intensities exceeding; 20 ksivin. were similar to those recorded in air. ("rack growth measurements revealed that crack velocity was directly proportional to applied stress intensity (liü (').
40 r
I
J
a.
30 -
20 -
U)
^0
o ̂ * X X _L J. -^ -i
0.01 0.02 0.03 0.04 0.05 0.06
CRACK VELOCITY (IN./MIN)
Fig. 9. Relationship between stress intensity and crack velocity (aged 800'F/8 hr).
The response to variations in strain rate and the mtergranular mode of fracture suggest that a hydrogen emhrittlement mechanism was responsihle for the suheritical crack growth. Hydrogen could have entered the steel during processing or as a result of crack tip corrosion
reactions with water (e.g.. water vapor in the air). The actual source has not heen definitely estah- hshed. hut three observations indicate a corrosion reaction. I irst. the cracks invariably initiated at the tension specimen surface, suggesting that BCCeH to the environment was required. Second, the tension specimens fractured in the limited region exposed to tap water, linally. aging at 800° to BS(f F might be expected to "bake out"" hydrogen remaining from earlier processing.
The detrimental effect of aging at temperatures less than lH)()0 F also is reflected in other grades of maraging steel, lirishane et |L (9) reported that the delayed failure resistance of precracked sheets of 3üü-gradc maraging steel, tension loaded in distilled water, was reduced by lowering the aging temperature Irom ('(H)0 to 750° or 800° F. Schapiro(lü) reported that the threshold value of 2ü()-grade maraging steel aged at 840° F for 4 hr was 48 ksKrfn. in sea water, whereas, values exceeding 100 ksisin. have been obtained for similar material aged at 1>00U I ( 1 I ).
This indicates that some critical change in the precipitation hardening reaction in maraging steels occurs at 850° F. Peters (I 2) has suggested that below 850° F there are two forms of precipitate, a dislocation nucleated precipitate
and .1 matrix precipitate. \i higher temptfratures, precipitation occur!« onl) al Jislucations. Such changes cuuId si^niticantl) atteel the deformation processcü 01 the availabilit) <>i sites for crack mull.'.it ion uiilim the crack lip plastic zone.
temperature and would also contribute to IIK-
OIIM-TU'II strength reduction. Unfortunately, the reverted austenite onlj slightl) improved the fracture toughness (Fig. III. I hh may be
\iim;j \hove S50 I
IV,ik N1K'1):JI1I was achieved In aging .11 925 I for N In However, the 350-ksi ultimau tensile strength level ^.IN nol achieved; typical ^.iliu-s ohtained In agini! .it ^OO0tol)250F forM In were within the range J35 to 347 ksi il I:JN I and 4».
I he fracture luughness K|^ for this strength range W.IN 35 to 4(1 ksi s in
Vging .it temperatures ibove lootri led to .i marked reduction in tensile properties (Fig 4». Phe amount ol reverted austenite retained ;ilK"r aging .it various temperatures lor 3 In IN shown in Kig. 10, and comparison with Fig. 4 shows the decrease in strength could be associated with reverted austenite; however, some ol the austenite formed i uring aging may have transformed to virgin (unaged) martensite upon cooling to room
HOr
5 3 _j O >
< a Z <
60
40
20
800 -o-o
1000 1200 1400
AGING TEMPERATURE (0FI
K.OO
Fig. 10. Austenite retained at room temperature after aging for 3 hr (bar material).
b 10 15
RETAINED AUSTENITE (VOLUME %l
Fig. 11. Influence of retained austenite on fracture
toughness.
contrasted with niartensitic stainless steel in which retained austenite significantly improves fracture toughness (13). This steel is austeiiiti/cd and incompletely transformed to martensite. so that retained austenite is dispersed between the martensite plates. On the other hand, manfing steel reversion commences at the martensite platelet boundaries, and a lamellar-typt' structure is subsequently formed (14). It would appear that this structure is ;i much less effective crack arrestor than the stainless steel with its intimate dispersion of austenite.
fracture toughneü tests were conducted on specimens aged at 9O0^F tor S hr either before or alter an aging treatment at SOOT I tor S hr. In both cases, there was no evidence of suhcntical
crack growth, and the fracture toughness was typical of a single 90u F age. Therefore, crack- urowth-susceptible microstructures, which could be present in the heat-affected /ones of welds in solution-annealed material, will be removed upon aging to peak strength.
IIK- crack powth resistance of 900* and 95(fF
aged specimens, statically loaded in 3.5'^ squeoitl sodium chloride solution, is shown in lig. I 2. I lie threshold stress intensity was 10 ksi^Tii. and the
crack velocity was constant (independent of stress
intensity) at approximately ().()()()() in./min. tor
both heat treatment conditions. Both the
threshold level and the time to failure at a given
initial stress intensity were increased relative to
the .S()()ur aged condition, which is shown for
comparison in lig. 12. It was not apparent why
the crack velocity changed from a stress intensity
dependancy to a constant, and much slower, rate
when the aging temperature was increased from
800° to mnf or (>5()0i-; the Iracture mode was
intergrantllar lor all heat treatments.
The tune to failure at a given initial stress intensity level was increased by raising the aging
temperature from 90(f to 95(fF. This can be
attributed to an increase in the incubation time
lor crack initiation since the crack velocity and
fracture toughness { hence the critical crack length
lor ligament rupture) were similar for both heat
treatment conditions. It has been proposed that
the incubation period c be explained as either
the time Tor the environment to permeate a crack
tip film or diffusion of the active species to a
critical site in the crack tip region (15). The latter
seems to be unlikely because the crack velocities
were identical lor both heat treatment conditions
and. therefore, suggests the crack-tip film was
more impervious in the 950^ F aged condition.
I he cracks were extensively branched on a
macroscopic scale in both heat treatment
conditions (I-ig. 13). This can be contrasted with
XOO0 I aged specimens in which there was no
evidence of macroscopic brandling and the crack
velocity was directly proportional to stress
intensity. As discussed elsewhere, branching only occurs when the crack velocity is essentially
constant and I critical stress intensity level is
attained ( 16).
PROPERTIES OF BAR MATERIAL
A similar study on 1-3/4- by 3/4-in. bar material
yielded almost identical results to those disaissed
above tor the 4-in. billet. In particular, stable
crack growth was observed alter aging below l'i)0ol;, and the 3S0-ksi strength level could not
t?
V
>
i I z
I- cn
<
40 <5_
B 30
^0
10
\
Hlr
800° F,'8 HFt 900« F ,'8 HR QBO8 F '3 HFt
O A D SPECIMEN FAILURE
• ▲ ■ TEST INTERRUPTED NO CRACK GROWTH
X V 0 KIc
900« F ,8 HR ■\ 950°F/3 HR
10 100
TEST DURATION (MIN)
1,000 10,000
Fig. 12. Effect of aging treatment on suhcritical crack growth resistance in 3.5% aqueous sodium chloride solution.
x w Fig. 13. Macroscopic crack branching in notched bend
specimen loaded to KV = 27ksi^in. in 3.5% aqueous sodium chloride solution (aged 900'F/8hr).
be achieved. I lure I ore. the KUIIVKIUJI results arc
not reconled here, hut the relalionsmp between
Itrcngtll and K|c lor material aiied at temperatures m the range 90Cf to HfXTF is sliown in Fij:. 14 lor both billet and bar material.
I he poor tradeott between decreasini; itltngth
aru' toughness is apparent and emphasizes the
inellectiveness of reverted austenite as a crack
stopper.
EFFECT OF SPECIMEN ORIENTATION
midsection ol the billet are compared with typical
longitudinal properties m Table V. Low
elongation and reduction ot area characterize the
transverse properties. I xamination ol the broken
specimens revealed that the Iracture path was
associated with the dark etching bands described
earlier dig. 15). However, the Kj^. values were
similar in the longitudinal (RT) and transverse
( I \\ and IK) directions. Apparently the
m'tJi/precrack plane was not aligned with the
arrays ot dark etching bands. consei|iuntly. the
toughness of the matrix was measured. These
observations may be i »ntraaled with those made by Salmon Cox et al. (7), which showed thai
banding in 250-grade maraging steel plate could
be detected by fracture toughness tests.
|S 60 /
o 4-IN.SQUARE BILLET
• 1-3/4-BY 3/4-IN.BAR
w Mi z I a o
40 -
< CC u- 200
.'0
.'-.
-I_ X 2b0 300
ULTIMATE TENSILE STRENGTH (KSli
350
I he results obtained trom tension specimens
machined in the transverse direction Irom the
fig. 14. Relation between strength and toughness for bar and billet (aged at various temperatures
above 850" F).
Table V. Influence of specimen orientation .
Orientation Yield Mrength
(ksi)
Ultimate tensile ■tfMfth (ksi) Elongation ('.) Reduction of
area {''<) K^. (ksiv'in.)
Transverse 337.'» 346.9 1.« 5.5 36.9 (TW)
3<v9(TR)
I ongitudinal 334.4 342.3 0.8 33J6 30.1 (RT)
aData arc nican of duplicate tests on specimens aged at ^OO'F for 8 lir.
10
350
■'^•v<ti^ 100
Fig. 15. Section through fracture face of transverse tension specimen showing fracture along dark etching bands.
Became of the heivier (eduction m the plate, the bands would be moie closely placed than In the billel and, hence, J peater probahUity oi notch alignment existed. Also, the stress lielii .it the tip ol ■ piecruck in the matrix could he about 10 limes more extensive in the tOUghei 2SO-frade material than in the 350 grade. 1 hi^ MfouM in- more likely to eause Iraeture along neifhbortil|| discontinuities. These observations point OUl thai the conventional tension lest has a diNlinct advantage over notch-type tests in the qualit) control testing (i.e., detection of discontinuitiesl ol thick sections of relatively brittle materials.
EFFECT OF TITANIUM CONTENT IN MARAGING STEELS
In the tensile strength range 180 to 350 ksi. the strength level of maiagmg steels is primarily increased by raising the titanium content. The effect of this element is shown in l-ig. 16 using typical values of strength and titanium content lor each grade. It is apparent thai titanium is much less effective in strengthening when the content exceeds 1%, particularly when it is remembered that the cobalt content of the 350 grade is approximately 4'.' higher than in the other grades.
■s.
0 1
I
1 I
-I 3
300
?50
?00
150
-200 GRADE
0 0.6 1.0 1.5
TITANIUM l%l
Fig. 16. influence of titanium content on ultimate tensile strength.
The relationship between strength and fracture toughness (l ig. 17) indicates that K|c rapulK decreases with strength level in the range 300 to 350 ksi, I his can be associated with the high titanium contents ( 1.()'. to 1.591) in this strength range because, as shown m 1 ig. 18, the Iraeture toughness is inversely related to the titanium content.
STRUCTURAL APPLICATION CONSIDERATIONS
lint He Iraeture resistance is a primary considera- tion in the application of ultra-high-strength steels. Using fracture mechanics analyses, an
11
150 r
I?
I Ui 7. X O
(T D K U <
100 -
50
,200ÜHADE
,250 GRAPE
»O 300 GRADE
350 GRADE
X X 200 250 J00
TENSILE STRENGTH IKSII
350
Fig. 17. Relationship between tensile strength (typical) and fracture toughness (optimum).
Z - I Ji
inz
3 ir F t- o to <
i50r
100
n
O 200 GRADE
\ 250 GRADE
^O300GRAnE
\ O 350 GRADE
-l_ o 1.0 io
TITANIUM (%l
Fig. 18. Influence of titanium content (typical) on fracture toughness (optimum) values.
estimate (.'.in be imde Of the critical crack M/C as .i function of applied stress. Such ■ plot is shown in Fif. l1' for 351)-^' ulc manging steel (at maximum ■ttength) ami compared with similar relationsliips
500 r
8
Q UJ
3 ■-
100
50
10
YS--
(3501 MARAGING (K. - 35 KSI . IN
YS - YIELD STRENGTH
^ 2 _ "Ic 1 21 "'i) WHERE 'T APPLIED STRESS
a - CRACK DEPTH
0 - CRACK SHAPE PARAMETER (REF. 171
1300 IMARAGING (K, MKSlJlN.I Ic s-
300M LOW ALLOY STEEL
(K, 70 ;si JTN.I Ic N
0.001 0.01 0.1
CRITICAL CRACK DEPTH, i (IN.)
10
Fig. 19. Comparison between critical crack sizes in 350 grade maraging steel and other high strength steels.
12
li'i ; »O.vl low-ullo) steel! : ■ ultimate lensil
M,\ i M ■mi u.mtK i ; ii th;
lowei strengt! steel Kurthermiire. it must b€ ; I design stresses ften u
Imiv • tensile *(i Mlowini such .in Increast in applied stress Indicates thai tht criti« crack MA in 35i)-gradi steel will be upproximatel) five times less than ii i lüM (a widel) used landing gear ,ill"\ i, I in ii ol course mean iii.it brittle Fracturt will occur in
»mponents mad< Irom IIH^ maragint! steel, In.t this significanl decrease in critical crack si/t miiNt he carefull) conside ed in rebtion to
structural weigh! saving >>lU •;!
ACKNOWLEDGEMENTS
I IK .miiio: i-> grateful to \. M. Kii>s lor luv assistance \MIII ihc sustained IO.H! tests
V •
CONCLUSIONS
I IK- SSO-gradc maragini iteel is particuiarl) Hticeptibk to Mibcritical crack growth in .nr when iiiioil ;ii temperaturea beknt ('()() I . Uns has been attributed to .1 hydrogen embrittlemenl mechaniam. Aging al 90Cr and 950^F improved thieshold streu intensity and changed the crack velocity from ;i siivss intensity dependency t<> ;i constant, and much slower, rate. Reduced crack'growth re ai stance in tin.- low-temperature>aged condition appears to be a characteristic <>r the l8'/i Ni maraging sy stem.
I he plane strain 'i.ietuie toughness ol 3S0-grade maraging steel, and hence the critical crack si/e For the initiation of brittle Fracture, is significantly lowei than the hij:h-slien;j|li steels emrenlK used In the aerospace industry,
Reverted austenite significantly decreased tensile strength Imi had little influence on fracture loughneaa.
REFERENCES
I. (.. U I uffnell and R, I . (amis, '-iv Ni 350>Maraging Steel.' \SM Tnuu Qtrl}'..b\, |96t i
"Proposed Recommended Practice for Plane Strain Fracture foughneas festing ol Higl Strength Metallk Materials l sing a Fatigue t i.ieked Bend Specimen.*1 ISTMStandai 1 3 I 1968, 1 . 1018
3, 15. I Brown, "A New Stress Corrosion Cracking lest for High Strength AHoys," t\M/ \ijii\ Res, and Stds., <)(>. 1966, p, 12''.
4, W. I. Brown and .1. 1. Srawley, "Plane strain Crack foughneas Testing of High Strength Metallic Materials." ISTM, STP 410. 1966,
5, R. I indgren, "Measuring Retained \ustenite by \-Ka\ I eehniques." HetaiProgress, \pril 1965, p. 102,
(>. G. R, Irwin, "Crack Extension Force fm < Part I hrough c rack in 1 Plate,*1 Trans. •l.VI//.. 29,(4), ll»(>:.p. 651.
7. I'. II. Salmon Cox, 15. G, Reisdorf, and G, I . IVIIissiei. "The Origin and Significance of Banding in 18 Ni (250) Maraging Steel." rnns. We/, Sttc. UME, 239, p. 1809,
N. \. k. Shoemaker, "Static and Dynamic K|c
Values for I hree Steels." Presentation to \SIM \-lA committee, Subcommittee Ml meeting, NRI. Washington D.C., September 1967.
13
A. W. Brisbane. J. M. Haun. and R. 1. Ault. "Fracture Tottghnea and Delayed I-allure Behavior of IH Per Canl Nickel Maragint; Steel.,M.S/.\/.\/(7//v fits, and Stds.. 5. 1^5, p. 395.
L Schapiro, "Steels lor Deep Quest and DSRV-I." Melä Progress. March l%S, p. 75.
J. 11. Ciross, "1 he New Development of Steel Weldments," 7. oj WeUtHg, 47. 1968, p. 241a
12. D. T. Paten, Trans. Met Soc. 1967. p. 1981.
AIMh. 239,
13. D. Webster. ' Increasinj; the roughness of the Martensitic Stainless .• -el AFC 77 by Control ol Retained Austenite Control.
Auslonninn. and Strain Aging." ASM Trans. (Jtrlv.. M, l%8.p. XI6.
14. S. Floraan, 'The Physical Metallurgy of Maraging Steels," Metals and Materials. September l%S. p. 115.
15. W. I). Benjamin and L. A. Steigerwaid. "An Incubation Time lor the Initiation of Stress Corrosion ("racking in Precracked 4340 Steel." ASM Trans. Qtrly. 60, 1^67, p. 547.
I(>. (". S. Carter, "Stress Corrosion Crack Branching in High-Strength Steels." Boeing Document D6-23871, 1969.
17. C. F. Tillany and J. N. Masters. "Applied Fracture Mechanics." I'racture Toughness lesting and Its Applications, ASTM, SIP 381, 1965, p. 249.
14
Unclassified i> i :.••.■ i<..
DOCUMENT CONTROL DATA R&D v,.,,„,v /.. it' .,'i'.r, •( n f/, I...,I-, al mbsitm I and mdfMmt mtnolalian mu ■!■ mfefvrf wfnn A< »♦•»•II fB^o« I '■■ ili»cl
i »i GROUP
•..■.■ A - . • . i . 'r/",f tumor)
I Ml hol |N(ICOMPAN> ( OIIIIIKI.MI airplane <iroup Kenton Washington .ii .. • ■ - (
I he Effect ol Heat Preatment on the Fracture roughness and SuhcriticaiCivckCtrowth ( haracieristics ol a 35CKGrade Maraging Steel
.■.. 1.1 pot« T Sf r . i. ' ■ r L Ass. i i c * ■
Unclassified
Research Report /■ ■ .. i. . / :, ' nun ' umldli inm.ii I - I nan i
( S. Cartel
A ' 1
June 1969 ■ i. A '■ i. a . r NO
\()()()l4-(.(.-( 0365 ( XKI'A Order No. 878) i ■ ■.
- i . ..
14
■i ■.
17 R .. ■■(*
boeing DoemnenI D6-2 n,V^
.' . ■■! ■. '
I Ins ducumenl h.is been approved ti>i public release and "».lie, its distribution is unlimited
■ 4 l A • . '
advanced Research Projects Agency. Departmenl i>i Defense
13 A (■
the heal treatment response ol a 3S0-grade maraging steel, with the nominal composition IK.5 Ni, 12 ( o. 4.ft vlo. 1.4 11, balance I e. IKIS been determined in hillet and bar form. When aged .it temperatures below "oo I . the material «.is very susceptible to subcritical crack growth, and premature brittle fracture occurred in unnotched tension specimens loaded ;it .i slow strain rate in laborator) air. Fracture mechanics «.is used to interpret this behavior. I he introduction ot reverted austenite significantl) decreased lite strength level bul had little effeel on fracture toughness, fhe resistance to brittle fracture i>t this material is contrasted with that of high-strength steels current!) used in the airplane industry.
DD :ORM.1473 Unclassified ■ ■ . t ! .. *. s ! ( , , ■
