b-A189 186 FATIGUE CORROSION IV AMORPHOUS FE(?S-X)CX j(1SI(1) 1/IdIRES(U) NAVAL SURFACE MERPONS CENTER SILVER SPRING MO
UNCLASSIFE T KABACOFF ET AL NOV 86 NSiiC/TR-B6-218
UNCLASSIFIED F/G 11/6.1 NLIIIEEEEEIIIIE
EllhlhlllE
.4
-IZ
NSWC TR 86-218
M C FILE CopyAD-A189 106
FATIGUE CORROSION OF AMORPHOUSFe75-xCrx13 5 1i1 WIRES
BY LAWRENCE T. KABACOFF ANH H. LE -.
RESEARCH AND TECHNOLOGY DEPARTMENT
NOVEMBER 1986
Approved for public release; distribution is unlimited.
... 4.. a¢'-
Ao
NAVAL SURFACE WEAPONS CENTERDahlgren, Virginia 22448-5000 9 Silver Spring, Maryland 20903-5000
S8 1 28 02R
-. CNCLASS I F I ELDSECURITY CLASS.FICATION OF THIS PAGE
REPORT DOCUMENTATION PAGE
'a REPORT SECURITY CLASSIFICATION 1b RESTRCT,vE MARKINGS
LNCILASS1IFiD __'__I)2a SECiRITY CLASSIF CATION AUTHORITY 3 DISTRIBUTION AVAILABILITY OF REPORT
2b DECLASSiFCATION. DOWNGRADING SCHEDULE
4 PERFORMING ORGANIZATION REPORT NUMBER(S) 5 MONITORING ORGANIZATION REPORT NUMBER(S)
NSN,,'(L TR 86-218
6a NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATION(If applicable)
dvi Siu-tce W4eapons Cen1er Code R32
6c ADDRESS (City, State, and ZIPCode) 7b ADDRESS (City. State, and ZIP Code)
"* J1090 1 "ev Hanmpsh Ire Avenue
Si! ,er Spring, .1D 20903-3000
Ba NAME OF FUNDING SPONSORING 8b OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATONj (If applicable)
Bc ADDRESS (City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERS
PROGRAM IPROJECT ITASK WORK UNITELEMENT NO NO NO ACCESSION NO
62234N RI 3 14
I TTLE (Include Security Classificationl
t--- .o,-o-i'l! (IF F" 7 5 _~rXB 1 5 silo
12 PERSONAL AUTHOR(S)Kri : cof-ft Lawrence T. -nd I e Anh3a TYPE OF REPORT 1 3b TIME COVERED 14 DATE OF REPORT (Year, Month Day) 15 PAGE COUNT
FROM TO 1986 November'6 SUPPLEMENTARY NOTATION
COSATI CODES 18 SUBJ P7RMS (Continue on reverse if necessary and identify by block number)FIELD GROUP SUB-GROuP m,,r A I , ,v ) . ' , , ,
r. I( -; i n)I Ob U) -- t 411
1ABSTRACT (Continue on reverse if necessary and identify by block number)
,, .'.tj 11h21e ir'dt-';IH I " the andtiL't 2111- corrosion propiurtieS ol [Ilectllli l' - wi _L ith til,, L',p ,it -t Xn rXCr 5 Si 1 0 (X =, 8, and 10) . Masurements were in121 i ir witt'lr,
-., 9. 2 ./( X~1(1 , ini I "N }l)H.-1 The observed it icl 1Kiits arc in\'lrec-! Iv Ir tipI,) i 1 1 Loi
tii,, ,,rroio, 1 1 riLc tUi (L (latilg : corrosion I;t ii glt e (,i lure fIwch)ifll0i. 21: t i '£ue 11- '.i-' I-tle wires are vatlv superior ti melt s;pun metoil ic las ribbons )I ,i iilir
i onii od stpcerior to, 30, s ta. inless .usteel Y iId L trac litlls uLp to 1()O) KS i!. v,,-.'-" l..m h ( r I I i/o :Cr). Under certain cirlum t,',.nc s, tie wires Arc 1 I. I;
i i' I .' w i c'l icl, -t-rgin 'ppl ied to te,-t .))Li'iYL1 s j interriipted pri to I ir-. IpiI ieU , tile sire behL ves l n cI : -l:nl r iilt- t i(' I ti < l tW, tl11 t iIutl C4'i I1,.
r r ,I ur. cI tL t i e r sIII t, .0 Si , ' l i 'II- Lt0 t 1 )-, (2 0 1 ri o I : p I ril ,,11 Crit i m . P ilt t inc' ,1nd ' ,ivi cc (orrI' i, n ill ('Ill - i ( 11Vir 1 nt 11r1. V01-%
D2 U S'P'B , ON AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION'" [ CLASS"- ;E D UNLIMITED E SAME AS RPT D3 OTIC uSERS '' !\ I 1
2a %A%, O RESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL
DO FORM 1473, 84 MAR 83 APR ed,t,on may be .,sed uil,l exhausted SECURITY CLASSIFICATION OF THIS PAGEAll other editions are obsolete S URITY C S ICA INOF THIS-PAGE
* U.S. Goflni Pi~rln.l Oifio 9IN:l--131-4
0102-LF-014-6602 '"1 I \;. 1ill
V %
UNCLASSIFIED
SECU-ITY CLASSIl"IOATO' OF THIS P 'E
..
.4.
UNCLASSI FlED
SECURITY CLASSIFICATION OF THIS PAGE
0%
NSWC TR 86-218
FOREWORD
This report describes the research performed during FY86 at the Naval
Surface Weapons Center on high strength, corrosion resistant metallic glasswires. These wires, which have only recently become available due to thedevelopment of a new manufacturing process, hold considerable promise as a
a- replacement for steel and titanium in many applications such as tow cables,mooring cables, and sonar domes. This work should be of interest toscientists and engineers who require very high-strength, ductile fibers whichcan tolerate a chloride environment with little or no corrosion (includingpitting and crevice corrosion). This work was supported by NSWC's IndependentExploratory Development Program (IED-105) and by I. Caplen of the David TaylorNaval Ship Research and Development Center, Materials Ship and Submarine BlockProgram. Questions may be referred to Dr. Lawrence T. Kabacoff[(202)394-2645].
Approved by:
Acoeqsion For
2 :. p c d ACK R. DIXON,:. t'fic:tion.. Materials Division
N','I i ty 9/,
O%
~~1~* i/ r)i c\i
NSWC TR 86-218
CONTENTS
Page
INTRODUCTION .......... ............................ . 1AVAILABILITY AND POTENTIAL COST OF METALLIC GLASS WIRE ... ....... 2EXPERIMENT .............. ............................. 3RESULTS AND DISCUSSION ........... ....................... 5
REFERENCES .............. ............................. 15
ILLUSTRATIONS
Figure
1 SCHEMATIC DIAGRAM OF THE APPARATUS USED TO MEASURE FATIGUE- DUE TO CYCLICAL BENDING STRAIN ........ ............... 4
, 2 POTENTIODYNAMIC CURVES OF AMORPHOUS Fe75_CrXBI5SilOIN 3.5% NaCI ............ ........................ 6
3 POTENTIODYNAMIC CURVES OF AMORPHOUS Fe75_xCrXBI5SilOIN 1.0 N H 2SO 4 . . . . . . . . . . . . . . . . . . .. . ... . . . . . . . . . . 7
4 PITTING SCAN FOR AMORPHOUS Fe70 Cr5 BI5Silo
IN 3.5% NaCI ............ ........................ 8
5 PITTING SCAN OF AMORPHOUS Fe 6 7Cr 8BI 5SiloIN 3.5% NaC1 ............ ........................ 9
6 PITTING SCAN ON AMORPHOUS Fe6 4 Cr1iB 1 5SiloIN 3.5% NaCI ...... .... . ......................... 10
K. 7 NTMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRAINFOR THE COMPOSITION Fe7 oCr5 Bl5Silo . . . . . . . . . .. . . . . . 11
p, NUMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRAINFOR THE COMPOSITION Fe67Cr8Bl5SI10 ... ............. . 12
* 9 NUMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRAINFOR THE COMPOSITION Fe6 4 CrlIBI 5SI10. . . . . . . . . . . . . . .. . . . . . 13
v/vi
04
NSWC TR 86-218
INTRODUCTION
Metallic glasses are non-crystalline alloys which are produced by rapidsolidification etther frqm the liquid or vapor phase. 1 Typical quench ratesrange between 103 and 10 C/sec. There are a large number of classes ofmetallic glasses. This report deals only with corrosion resistant Fe-basedmetallic glasses produced by quenching from the melt. These amorphous alloystypically contain about 25 a/o metalloid (usually B, Si, C, and/or P), with
•'V the balance comprised of various transition metals (Fe, Ni, Cr, and Mo).. .Because of the similarity in composition between these materials and 300Iii series stainless steels, thev will be referred to as "amorphous stainless
steels. "
The corrosion resistance of amorphous stainless steels results from twoeffects.2 Since the materials are truly non-crystalline and, therefore,lack grain boundaries and other crystalline defects, they form very highquality homogeneous passive films which especially resist pitting corrosion.Also, since the surface of an unpassivated metallic glass is in a higherenergy state than a corresponding crystalline material, greater surfacedissolution occurs prior to formation of the passive film. Thus, morebeneficial ions (such as Cr) are present during passivation resulting in anenriched passive film. For example, a metallic glass with 10 a/o Cr may havea passive film with more than 50 a/o Cr. Passive films were 98 a/o of themetallic species is Cr have been observed. Such alloys can be boiled in 12 NHCI with no measurable corrosion.
Traditionally, amorphous stainless steels have been produced in the formof continuous thin ribbons by continuous chill block casting onto a copperwheel. These ribbons have very high yield strength (300-625 KSI) andexcell nt ductility in sheer. Tensile failure occurs in very narrow sheer
* hands. Thus, the stress-strain curves resemble brltt'o, materials evenA M.though the fracture surfaces show that this is not the case. The fatigue
properties of ribbons are very poor, largely because the ribbons contiansurface irregularities and large local variations in ribbon thickness.Cenerally, metallic glasses do not work harden. Because of the problems with
[&-'7 fatigue, interest In using these materials for structural applications such asr bles failed to develop.
A recent advance in fabrication technology has changed the situation.Unitil'a, Ltd of Japan has developed a process by which a s ream of melt isinjected directly into a moving layer of a water solution. The entry angle
r, .in! the relative speeds of the melt and water are adjusted to give a smooth,n,n-tur)ulent flow, resulting In high quality wires with round cross.(,ctinn-. Unitika engineers reported that the fatigue properties of the
wir(.,, unlike those of ribbons, are outstanding, and exceed those of, for
V ,'
NSWC TP 86-21P
example, 304 stanless steel7 These tests were carried out in air anddeionized water. Tn all other respects, the properties of the wires were
similar to those of ribbons with similar compositions. This development isextremely important to the Navy because of the strong need for high strength,corrosion resistant materials with outstanding fatigue properties for cableapplications. However, before this technology can be exploited, severalquestions must be answered. First, the data presented by, Unitika must beverified and extended to a seawater environment. This involves developing newcompositions which represent the best compromise between the needs ofcorrosion resistance, strength, fatigue properties, and ease of fabricationinto amorphous wires. Secondly, these wires must be tested under the extreme
conditions experienced by the various tow and mooring cables used by theNavy. Finally, there must be a reasonable prospect that amorphous wire willhe commercially available at a reasonable price.
4-?.
These questions are being addressed by the current program, whichconsists of three phases: (1) developing compositions suitable forfabricating and testing; (2) developing reliable sources of test samples,including an in-house capability for producing research quantities of
,* . amorphous wires; and (3) determining the prospects for the availability ofamorphous wires on a commerical basis. This report summarizes progress forFY86.
U''. AVAILABILITY AND) POTENTIAL COST OF METALLIC GLASS WIRE
At present, the only form of commercially available metallic glass is thecontinuous thin ribbon formed by continuous chill block casting. Typicaldimensions are up to 7" wide by 0.0001" thick. The main applications areb raising foils (Ni based alloys) and power distribution transformer cores (Febased). The cost of the Fe based alloys is currently less than t2.00/lb andis expected to go much lower due to the competition of Si steels. Thecompositions currently being marketed are not suitable for chlorideenvironments. Special compositions can be produced, but the low price wouldlonly be available to a high volume customer (e.g., ordering 1000 lbs or more).
Currently, the round cross section wires are produced commercially only;. by 'nitika, Ltd of Japan, and production is limited to a pilot plant. Wire is
O ,vailable to American customers in amounts up to twenty pounds and only inthree compositions. Research quantities (10 grams) of other compositions can
"tained, thus far at no cost. Unitika participates in an internationalventure company, NAMCO, whose two major partners are the Matsui Group andA Ilied Forn. Under this arrangement, Allied has given Unitika a license toIk .:,,n f'ct tire the wires (Allied holds key patents) in exchange for exclusive. stribhtinn rightc outside Asia.
eAt ip at t be Naval Surface Weapons Center (NSWC)
rpr, n ,,nt1tivrs of U'nitika and Allied discussed the supply situation with.. ,,- ein rnm iSWC and the David Taylor Naval Ship Research and Development
--rt, r (DT N RDC). It was disclosed that the new pilot plant is producing
(' per month at 60/lb. A production facilitv which could produce-',v, rnl tons per month is being considpred. Tt was estimated that potential
' .n~qmpt !yP cnruld exceed og, 000 lbs per year. The manufacturing
0 r4I"' k_
NSWC TR 86-218
situation changed very quickly because of recent interest In u~i !' metallicglass wires in tire cord. if successful, perhaps this will drive the cost ofstandard composition wire below t1.O0/lb. Therefore, a plentitul, Inexpensive
supply of amorphous wire is very likely. Note that the compositions producedfor the tire market will not be suitable for a chloride environment. However,
the price of these wires is controlled almost entirely by the cost of buildingmanufacturing facilities. The cost of raw materials and plant operation is
somewhat less than those for producing conventional drawn wires. If start upcosts are paid by the tire industry, the cost of "special order compositions,"
if orders are large enough, will be quite reasonable.
In the short term, small quantities of metallic glass wire (e.g., 10grams) are sufficient to carry out corrosion, fatigue, and mechanical testingof monofilaments as a function of wire composition. Thus far, samples have
been provided free of cost by Unitika. However, because of the delaysinvolved and the numbers of compositions which will be needed, an in-house
source is necessary. The parts for such an instrument have been obtained andare currently being assembled. It is expected that the facility will beoperational soon. Quantities of wire suitable for fabrication of cables andtire cord can now be purchased from Unitika. They also have the expertise and
are willing to fabricate the wires into a finished product.
EXPERIMENT
Following a review of the literature, the compositionsFe75_xCrxBl5Silo (x = 5, 8, and 11) were selected for initialtesting. This selection was based on the fact that, in the Fe-B-Si system,
Fe75Bl5Silo has the maximum glass forming ability and a relatively highyield strength. The addition of Cr was expected to increase the strength
while only moderately reducing the glass forming ability (as measured by themaximum diameter wire that can be made with no measurable crystallinity). The
Cr, of course, is necessary for corrosion resistance. Carbon and phosphoruswere not used as metalloids at this time because, while they enhance corrosion
resistance (especially P), they reduce yield strength. Another importantconsideration was that these compositions had been produced previously byUnitika, and it was certain that samples could be obtained.
The apparatus for performing bending fatigue testing is illustrated in
Figure 1. This device, which is similar to that used by Hagiwara et al.,
produces strain by passing a wire specimen over a pulley which is immersed inthe medium of interest. The maximum strain is controlled by the diameter ofthe pulley and determined by the relation:
A d/(d + D)
wliere d is the diameter of the wire (,in thi case .005") and D Is the diameterof the pulley. Wire specimens were poxved to two nylon leads, one of which
was t i~d t-n a wheel rotatinc at 7 Hz and the other weighted with a 100 gramoad. Failure of the(- wire cau;sd thil weiht to activate a switch connected totin er. The nuImbcr 1)f cycles to fiilurt, was measured as a function of
m-x.imuin etrain. The faitigue li it was defined :s the largest cyclical strainwhich would not rosult in failure within rl cycles. Fatigue measurements- rk, perfor.ed r iir ((57' RU), d,.i.,ni d water, 1.57 NaCI solution, and I N
% 4%S- - - -
OZ N 0
I'.NSWC TR 86-218
-
LOAD - MOTOR
TIMER
T
SWITCH PULLEY
FIGURE 1. SCHEMATIC DIAGRAM OF THE APPARATUS USED TO MEASURE FATIGUE DUETO CYCLICAL BENDING STRAIN
".44
0
'0,
NSWC TR 86-218
Potentiodynamic and pitting scans were obtained using an EG&G CorrosionMeasurement Console (Model 350A). The three-electrode-cell consisted of aworking electrode, a graphite counter electrode, and a SCE (saturated calomelelectrode) reference electrode. Solutions of 3.5 w/o NaCl and 1.0 N H2SO4were prepared from analytical grade chemicals and deionized water.
Measurements were made at room temperature in aerated solutions. The initial
potential for potentiodynamic curves was set at 250 mV below the open circuitpotential. The scan rate was 1 mV/sec.
Finally, an approximate value for the tensile strength was obtained for
each composition by hanging weights from the looped end of each wire. Severaltrials were made for each composition and the results averaged. The intentionwas only to get a rough estimate of the strength.
RESULTS AND DISCUSSION
Figure 2 shows polarization curves for amorphous wires in 3.5 w/o NaClsolution. The wires containing 5 a/o Cr display an active/passive behavior,wliile the 8 and 11 a/o Cr specimens exhibit complete passivity. As expected,
the corrosion current density (and, therefore, the corrosion rate) decreaseswith increasing Cr content. All three compositions exhibited active/passivebehavior in 1.0 N H2S0 4 (Figure 3). As expected, the corrosion currentdecreased with increasing a/o Cr.
Pitting potentials varied only slightly as a function of composition in
the NaCl solution (1.20 V/SCE for 5 and 8 a/o Cr, and 1.170 V/SCE for 11 a/oCr). This is illustrated in Figures 4, 5, and 6 which show pitting scans for
5, 8, and 11 a/o Cr in 3.5 w/o NaCl, respectively. It is also apparent fromthese figures that the hysteresis is extremely small, indicating a very lowrate of crevice corrosion. In 1.0 N H2SO4, all of the pitting potentialswere the same (0.966 V/SCE).
Figures 7, 8, and 9 illustrate the number of cycles to failure as afunction of maximum bending strain for 5, 8, and 11 a/o Cr, respectively.
Fatigue limits were observed for each composition in air and deionized water,and for 8 and 11 a/o Cr in 3.5 w/o NaCl. Our data on moist air and delonizedwater agree well with published data by Hagiwara et al.6 Comparison of
Figutres 2 and 3 with Figures 7, 8, and 9 clearly show the relationship betweencorrosion rate and fatigue limit (or number of cycles to failure), and that
Failure occurs through a corrosion fatigue mechanism.
In order to investigate the role of hydrogen in the corrosion fatigue.filure of these alloys, wire specimens were immersed in 3.5 w/o NaC], and
subjected to cyclical bending strain for 15 minutes, then baked for 15minutes. Without interruption, for the strain chosen, failure would occurafter 40 minutes. This was repeated (15 minutes on, 15 off for 8 hourswithout failure (a total of 4 hours of strain application). This iconsistent with the hypothesis that failure is due to accumulation ofhydrogen. What was surprising was that the control specimens, which were notbaked but were kept in the chloride solution during the "rest" period, alsoexperienced no fatigue failures. In both cases, after the 8 hours of testing,
.. '
"; 5
V
L- 'U. ...
NSWC TR 86-218
2.500 -55% Cr
- 8% Cr, " 11% Cr
1.500-LA.I
-=" Ii- _ 0.500
101 102 103 104 105 106 107 108
CURRENT DENSITY (nAicm2)
FIGURE 2. POTENTIODYNAMIC CURVES OF AMORPHOUS Fe 7 5 .xCrxB1 5 Si1 0 IN 3.5% NaCI
-6'.=
==6
I,
'a .P' "a?' '-"a.,"?"-" a-?" "-"-"-;--.".-''
NSWC TR 86-218
2.500r-- _ _ _ _ _ _ _ __ _ _
- -11% Cr
1.500LLU
~0.500
-0.500, _;::------
CURRENT DENSITY (nAlcm)
FIGURE 3. POTENTIODYNAMIC CURVES OF AMORPHOUS Fe 7 5 .XCrXB15 S'1 0 IN 1.0 N H2 S0 4
* 7
%6
*- NSWC TR 86-218
1.430
'.
1.290CI
A. 01.150
,K /1.010 - t/
,.,' " 0.8701~7 I l I
100 101 102 103 104 105 106
CURRENT DENSITY (nAJcm2)
6 FIGURE 4. PITTING SCAN OF AMORPHOUS Fe 7 OCr 5 B1 5 Si 1 0 IN 3.5% NaCI
..
8
6
NSWC TR 86-218
1.610
1.430
CAI. 1.250
1.070
0.890100 101 102 103 104 105 106
CURRENT DENSITY (nA/cm2)
FIGURE 5. PITTING SCAN OF AMORPHOUS Fe6 7 Cr8 B 15 Si1 0 IN 3.5% NaCI
-U 9
NSWC TR 86-218
.4
"l 1.610
: 1.430
S1.250-
4,"
1.070
0.890,
- 100 101 102 103 104 105 106CURRENT DENSITY (nAlcm2)
.,V. FIGURE 6. PITTING SCAN OF AMORPHOUS Fe6 4 Cr11 B1 5 Si 10 IN 3.5% NaCI
-..
10
V ,, ,% .% %,% , .& -.,-. L
NSWC TR 86-218
0.015 Fe70 Cr5 Silo B15065% RELATIVE HUMIDITY0 3.5% NaCIz, 1.0N H2S04
_0.010
Z 0.005
0103 104 105 106 107
NUMBER OF CYCLE TO FAILURE N!
FIGURE 7. NUMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRAIN FOR
THE COMPOSITION Fe 7 0 Cr 5 B15 Si1 0
I111- ~ ~ A **. -~ J '.s.
NSWC TR 86-218
-4
0.015 Fe67 Cr8 Si, B15
L65% RELATIVE HUMIDITY
',..5 ".- .0 0
; L ~ON H0 S0 0 04 0
22
Z0.010,
'4,n
S0.005
03 104 105 106 107
NUMBER OF CYCLE TO FAILURE Nf
FIGURE 8. NUMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRAIN FORTHE COMPOSITION Fe6 7Cr 8 B 15 Si 1 0
[12
.
2'
•4
-. 1"
NSWC TR 86-218
Fe-- Cr1 Si B15 13.5% NaCl0.015'JlONHS 4 .
Z
0.010 1
z
0.005 --
0
103 104 105 106 107 ]
NUMBER OF CYCLE TO FAILURE NF
FIGURE 9. NUMBER OF CYCLES TO FAILURE AS A FUNCTION OF BENDING STRA.N FORTHE COMPOSITION Fe6 4 Cr1 1 B1 5 Si 1 0
L
13
Wk.1
r@
%:., NS!C Ti2 86q-21
cyclical strain was applied wi thout interrupt ion on t ii faIl ur,. The t i me t (;
failure in both cases was approximately the same as would be obtained for a"virgin" specimen. Since the control specimens remained in the solution,
there was no removal of hydrogen. Thus, the bulk accumulation of hydrogencould not have been the cause of the fatigue failure in this case.
Extensive work has been done by Hashimoto et al. 6,7,8 on the mechanica]failure of metallic glass ribbons in an acidic chloride environment. Thiswork included tensile testing in a corrosive environment (with interruption,
an! bakinz similar to NSWC's tests cliscussed in previous paragraphs) and, examination of the fracture surfaces. The evidence for brittle failure in an
didlc environment due to bulk accumulation of hydrogen is convincing.touever, it is interesting to note tmat fracture surfaces prducdby failurin a neutral chloride environment differ from those produced in an acidifiedenvironment. It should also be noted that fatigue limits were observed onlIyin cases of spontaneous passivation. in the case of an acidic environment orlow chromium content in a neutral chloride environment, the general corrosionrite 's relatively high, with h,'dropen formed uniformly over the surface. Inthe case of spontaneous passivation, corrosion occurs primarily in cracks,-ceated in the passive film by the application of strain. Thus, it isreasonable to assume that the failure mechanism in the two cases will be; ubstantially different.
A reasonable hypothesis for fat ue failure in the case of thespontaneously passivating specimens is that cracks in the passive film admit
NSWC TR 86-218
REFERENCE S
1. Masumoto, T. and Maddin, R., Mat. Sci. En. , Vol. 1l , No. 1, 197,.
. Hashimoto, K. and Masumoto, T., Cla"sv Metals: Magnetic, Chemical ind
St ructural Properties, ed ited by R. llasaawa (CRC Press, Boca Raton, 1983),p. 23 .
. Davis L. , Metallic Glasses, edited b J. ClIlman and H. Leamv (American
Society for Metals, Metals Park, 197P), p. 190.
. Inroue, A.; Hag1iwara, M. ; and Maumoto, T. , Met. Trans. Vol. 13, No. 373, 1982.
P Hagiwara, M.; Inoue, A.; and Masmoto, T., Re Aidlyuenched Metals, edited by
S. Steeb and H. Warlmont (Elsevier Science Publishers, New York, 1985),
p. 1779.
F. Kawashima, A.; Sato T.; and Hashimoto K., J. Non-Cryst. Sol., Vol. 70,
No. 55, 1985.
Kawashima, A.; Hashimoto, K.; and Masumoto, T., Corr. Sci., Vol. 16, No. 935,
197A.
* Kawashima, A. ; Hashimoto, K. ; and Masumoto, T., Corrosion, Vol. 3A, No. 577,
1980.
9. Brown, B.; Fuji!, C.; and Dahlberg, F., J. Electrochem. Soc., Vol. 116,No. 21P, 19 9.
] /]
NSWC TR P6-?18
DISTRIBUTION
Copies
'fnse Technical Inform;itionr nt er
rceron Station 12
Iexandria, VA 2 .
Librarv of Congress
t tn: Gift and Exchange Divisionhington, DC 20540
ivid Taylor Naval Ship
Research and Development Center,in: Code 2803 (R. Hardy) 1
,polis, MD 21402
, nternal Distribution:
13 (C. Winkler) 1a' c 1
K -I. Kahacoff) 25(N. Mirtlnez) 1
(A. ~') 10"P I 2
1 1I ((
rflnnflTL . .- ~.-'.-.*'.- - .. r- ~,-~r' rr,
.4,.
4.
9 Kb.4
.4
4,.
'N
4
4-
.5-
'4-
*'S i/c?5-
-4
J.
'5.
44
4-44544
-' 5". * -. * ,'. * -4.. -. 4 -. 4 ~45. ~ ~ *~5. -. ~ - - - S - - A. -- 5- *. - .-
