''T: -7AIN

STRAIN AGING AND"DELAYED FAILURE IN

0 AHIGH-STRENGTH STEELS

TECHNICAL DOCUMENTARY REPORT No. ASD-TDR-62-968

NOVEMBER 1962

DIRECTORATE OF MATERIALS AND PROCESSES

AERONAUTICAL SYSTEMS DIVISION

AIR FORCE SYSTEMS COMMAND

WRIGHT-PATTERSON AIR FORCE BASE, OHIO

Project 7351, Task 73521

(Prepared under Contract No. AF 33(657)-7512 byThompson Ramo Wooldridge Inc., Cleveland, Ohio;

E. A. Steigerwald and G. L. Hanna, authors.)

¶ _

NOTICES

When Government drawings, specifications, or other data are used for anypurpose other than in connection with a definitely related Government procure-ment operation, the United States Government thereby incurs no responsibilitynor any obligation whatsoever; and the fact that the Government may haveformulated, furnished, or in any way supplied the said drawings, specifications,or other data, is not to be regarded by implication or otherwise as in anymanner licensing the holder or any other person or corporation, or conveyingany rights or permission to manufacture, use, or.pell any patented inventionthat may in any way be related thereto.

Qualified requesters may obtain copies of this report from the Armed

Services Technical Information Agency, (ASTIA), Arlington Hall Station,Arlington 12, Virginia.

This report has been released to The Offi~e of Technical Services, U.S.Department of Commerce, Washington 25, D.C., for sale to the general public.

Copies of this report should not be returned to the Aeronautical SystemsDivision unless return is required by security considerations, contractualobligatiorn, or notice on a specific document.

-4,"

7~< -L. 7- 7~ -77 .""A7-

FOREWORD

This report was prepared by TAPCO, a division of Thompson RamoWooldridge Inc., under USAF Contrast No. AF 33(657)-7512. The contractwas initiated under Project No. 2(8-7351) and Task No. 73521. The projectwas administered under the direction of the Directorate cf Materials andProcesses, Deputy for Technology, Aeronautical Systems Division, withLt. Robert T. Ault acting as project engineer.

This report covers the period from Octobar 15, 1961, to October 14, 1962,

II.•

ABSTRACT

Spontaneous strain aging which occurs during tensile testing wasexamined for several high-strength steels. The results of smooth andnotch tensile tests indicated that significant strain aging effectsoccurred in most high-strength steels in the 300*F to 800*F temperaturerange and this behavior was analogous to "blue brittleness" in mild steels.

Constant load, stress rupture tests were conducted on the steels todetermine the possible relationship between strain-aging embrittlement anddelayed failure. Only the 300 M steel tested at 400"F exhibited anappreciable degree of delayed failure. This embrittlement, was extremelysensitive to test environment and was elimainated when tests were conductedin argon. Although strain aging was not a sufficient condition to initiatedelayed failure, it appeared to increase the severity of the environmentaleffects in the particular range where sufficient interstitial mobilityexisted.

This technical documentary report has been reviewed and is approved.

:- '.4

W. J. TRAPP

Chief, Strength & Dynamics BranchMetals & Ceramics LaboratoryDirectorate of Materials & Processe

iii

.• *= J

TABLE OF CONTENTS

Page

I INTRODUCTION 1

II MATERIALS EVALUATED 3

III TEST PROCEDURE 3

IV RESULTS AND DISCUSSION 8

A, Smooth and Notch Tensile Teszs 8"B. Failure of High-Strength Steels Under

Sustained Loads 20C. Possible Strain-Aging Mechanisms in

High-Strength Steels 30V CONCLUSIONS 41

vI BIBITOGRAPHY 42

Appendix I - Calculations of Ordering Effects DueTo Interstitials 45

-~ iv

21

LIST OF FIGURES

Fig2ure ±age

I SmfLooth Sheet Specimen for Tensile Tests. 6

2 Geometry of Notch Tensile Specimens, 7

3 Apparatus for Mechanical Testing in an Argon

Atmosphere. 9

-4 Tensile Properties of 300 M Steel (600*F Temper) Cross-Head Speed of 0.050 in/miii, 10

5 Notch Tensile Prop;erties of 300 M Steel (600"F Temper). 11

6 Notch Tensile Strength of 300 M Steel (1025*F Temper)

as a Function of Test Temperature. 12

7 Tensile Properties of 4340 Steel (750*F Temper) as a

Function of Test Temperature,Crosshead Speed of0.050 in/mi. 13

8 Notch Tensile Properties of 4340 Steel (750@F Temper). 14

9 Tensile Properties of H-II Steel (IO000F Temper) Cross-Head Speed of 0.050 in/min. 15

10 Tensile Properties of 17-7 PH Steel (RH 950) as a Functionof Test Temperature, Crosshead Speed of 0.050 in/ran. 16

11 Tensile Properties of AM 355 Steel (SCT 850). 17

12 Tensile Properties of !8% Nickel Maraging Steel. 19

13 Influence of Test Temperature on the Delayed FailureCharacteristics of 300 M Steel (60OOF Temper), NotchTests in an Air Environment. 21

14 Influence of Test Temperature on the Stress RuptureCharacteristics of 300 M Steel (1025"F Temper), NotchSpecimens in an Air Environment. 22

15 Influence of Test Temperature on Stress RuptureCharacteristics of 18% Nickel Maraging Steel

(240,000 psi Yield Strength), Notch Specimens in anAir Environment. 23

V

IN

LIST OF FIGURES (Continuzi)

~Pag

16 Macroscopic Apprarance (1OX) of Sustained Load Failuwesin 300 M Steel (600*F Temper), Specimens Te,-ted at 400*Fin an Air Ehvironment 0 24

-17 Kinetics of Crack Growth as Determined by ResistanceMeasurement~s on Sustained Load Tests of 300 M Steel(600*F Tempei), Notch Specimens Tested at 300"F underan Applied Stress of 645OO psi. 26

18 Kinetics of Crack Growth as Determined by ResistanceMeasurements on Sustained Load Tests of 300 M Steel(600*F Temper), Notch Specimens Tested at 400*F underan Applied Stress of 45,750 psi 0 27

19 Stress Rupture and Crack Initiation Characteristics of300 M Steel (6000 F Temper), Notch Specimens Tested at300OF in an Air Atmosphere. 28

20 Influence of Environment on the Delayed FailureCharacteristics of 300 M Steel (600F Temper), NotchSpecimens Tested at 400*F. 29

21 Arrhenius Plot of Crosshead Speed and Temperature at,.which Minima in Notch Tensile Strengthl Occurred. 32

22 Arrhenius Plot of Strain Rate and Temperature at whichMaximum in Smooth Tensile Strength Occurred. 33

23 Comparison of Data Obtained for "Blue Brittleness" ofMild Steel (Ductility Minima) with Notch TensileStrength Mirdima Obtained for High-Strength Steels. 34

24 Caiculated Force on Dislocations Due to InterstitialOrdering as a Fiu~ction of Temperature. 36

Kvi; M

LIST OF FIGJRES (Continued)

_rePag

25 Schematic Illustration Comparing True ActivationEnergy of Interaction Process and ApparentActivation Energy Obtained from Tensile Test Data. 37

26 Electron Micrographs of 300 X Steel (600OF Temper)Mag., -10,O00OX- 39

'27 Influence of Prestraining 1% and Simultaneous Agingat 400OF on the Room Temperature Notch TensileStrength of 300 M Steel (600*F Temper). 40

S211 The Force to move a Dislocation in qteady State as aFunction cf Dislocation Velocity (a , )o 48

:-1

p,'-v Q.1

I INTRODUCTION

The r3ed for airborne ccmponents with very high strength-to-weipnht ratioshas generated many xinique problems for the materials engineer. These problemareas involve not only the selection of meaningful design stren th parameters,•".•." but also the consideration of certain types of embrittlements which may limit

the performance capaci±V of these high-strength materials. The irreversible* embrittlement produced .f-tcr tempering in the 500*F to 600F temperature range

(1)* and the time dependent failures due to hydrogen (2) or mildly corrosiveenvironments (3, 4, 5) are well-known examples where the load-carrying abilityof a structure can be significantl.y reduced by embrittlewents.

The embrittlement due to strain aging in mild steels is well recognizedand has been frequently studied (6, 7, 8). In the case of high-strength steel,however, this phenomenon has only recently received attention and relativelylittle data exist on the decrease in notch strength which is produced by thisfactor. Strain aging is generally defined as a time-dependent change inproperties that results from a sequence of plastic strain and aging. Undercertain conditions of strain rate and test temperature, strain aging canoccur spontaneously during a normal tensile cest and result in ductilityminima and in most cases strength maxima. In mild steels this strain agingprocess, which occurs during conventional tensile testing, is often referredto as "blue brittleness" (8).

In high-strength steels the sequence of tensile straining followed by

aging below the original tempering temperature htas been investigated as amethod for increasing the smooth strength properties of a material. Recentwork with the low alloy martensitic steels (W340) and the hot-work die steels(11-11) has indicated that a significant increase in tensile and yield strengthcan be produced by this mechanical-thermal treatment (9, 10, 11),

The property variations which have been observed as a result of thestraining plus aging sequence have been attributed to a change in the carbidemorphology during the aging treatment. The phenomenon of spontaneous strainaging during a tensile test (blue brittleness) has not been systematicallyevaluated in the high-strength steels. In these materials The major interesthas been in the improvement of the transition tcmperature as measured by asharply-not'hed specimen, and relatively little data. are available on proper-ties in the super-transition temperature regions where strain aging may bepresento In addition, the high-strength steels do not generally exhibit-i•eld point effects and therefore the simple interactions between dislocationsand interstitial atoms (8) which are employed to describe strain aging inmild steels are not believed to play a dominant role in the mechanical proper-ties of the ultra-high-strength steels.

•* Numbers in parentheses pertain to references in the Bibliography.

Manuscript released by authors November 1962 for publication as an ASD___ Technical Report1

Recent daba by Srawley and Beachem (12) and Preston and Morrison (13)have indicated that a milnimum in the notch tensile strength occurs in thelow alloy martensitic high-strength steels when they are tested in thesuper-transition temperztl÷re range. Since the temperature at which thisdecrease in notch properties takes place is comparable to the region where"blue brittleness" occurs in mild steels, the question arises as to whetherspontaneous strain aging effects are present to a significant degree in high-strength steels. There is considerable current interest in the nechanicalproperties of materials in this super-transition temperature region sincethis is the temperature spectrum under consideration for the supersonictransport. Certain pressure vessel applications may also involve temper-atures up to 5OO*F and any decrease in load-carrying capacity may be acritical consideration at these temperatures which are above the conventionalsharp-notch transition temperature.

Since strain aging is a time dependernt phenomenon, there is the pos-sibility that this embrittlement will also be operative in constant loadtests where the test time is a maxirmu. Troiano (2) has recently proposeda mechanism for strain aging embrittlement which predicts this effect andwhich involves the clustering of interstitial aton3 at the tip of a crackor dislocation pile-up. The interstitial grouping is believed to lowerthe cohesive strength of the material and cause crack extension. Althoughhis res'ilts are based on work with hydrogenated steels, the possibilityexists that carbon and/or nitrogen strain aging can also lead to delayedfailure in temperature regions where these interstitials have a mobilitysimilar to hydrogen at room temperature. The high-strength steels withtheir marginal crack propagation resistance represent materials which maybe particularly susceptible to a non-hydrogen induced type of static fatiguee

The purpose of this investigation was to evaluate the magnitude andkinetics of spontaneous strain aging in high-strength steels and determineits possible association with delayed failure under static loads.

•- .

II MATERIALS EVALUATED

Four basic types of high-strength steel (see Table I) were selected forNf• the experimental program. All the materials, with the exception of the Ni-

Co-Mo maraging type, were conventionally air melted. The maraging steel wasreceived from a commercial consumable electrode, vacuum-melted heat. The

'1• 300 M and h3h0 steels were selected as representative of the low alloymartensitic high-strength steels while the H-11 was evaluated as a char-acteristic hot-work die steel. The 17-7 PH and AM 35$ were chosen astypical precipitation-hardening stainless steels. The Ni-Co-Mo alloy wasincluded as representative of the recently developed maraging steels. Theheat treatments employed for these materials are summarized in Table II.All the austenitizing treatments were performed in neutral salt baths whilethe tempering was done in an air furnace. No stock was removed from thesurface of the material. In the as-received condition the 300 M steel hada slight degree of surface decarburization which produced a hardnessdifference of approximately 2 R from the surface to a depth of 0.002".The other steels used in the instigation had no significant decarburization.

III TEST PROCEDURE

Strain aging effects in high-strength steels were sttdied by conductingsmooth and notch tensile tests as a function of loading rate and testtemperature. The maxima in ultimate strength or the minima in notchstrength* were employed to evaluq4e the strain aging phenomenon. The smoothtensile specimen geometry is presented in Figure 1, while the notch tensilespecimen is shown in Figure 2A. The notch specimen corresponded to the re-commendations of the ASTM Committee on Fracture Testing of high-StrengthMaterials (lh) and consisted of a center notch generated by fatigue crackingafter heat treatment.

The correspondence between strain aging and delayed failure was evaluatedby conducting stress-rupture tests under sustained loads in temperatureregions where strain aging occurred in the tensile test. Notch specimenswith the geometry shown in Figure 2A were used for 300 M tempered at 600OFbecause relatively low breaking loads were required. In other -materials, thebreaking loads exceeded the capacity of the stress-rupture apparatus and thenotch specimens shown in Figure 2B were used.

* Notch strength is defined as the load at fracture divided by the initialcross-section area at the plane of the notch.

3

i4 Rm5...

*1d

0-

4.1 0

00 000

bz _:I m mm*H H C; C

0 C) 0 00 Cr-

--I I I, I I .

~ 41

020 0

00H 0

4 1 0~ 00 0 E- IA.2 HI- 04. r 0i-

0 0

H, 0@

HI Ho

a) 00 0

0 ~ A-~ 008

09 .

02 E- 00 0 0

0 - l0 c

V-.

H4

*jNIT1ov V .A; Vj

0 C) C.) 0

U)~ 0 .

-P co C\ f

0r- r=4 4_3 4-D 4- 'f.)

H Hý H- H.Q) 0 1 0 0

E- 0CT E4 ýq E- 0 0 00 0

00 000J H ;-j 0. 00 rI

-P0 CV 00 c 0 0- 0-O (0i *-0 0 0l

Ea $4 0 co -P40 ~0 U)i -1~ U))L

a)-. 0 0 frfr4.i. 0 0 P-4*~'

4-') 0) ) *H 0 - - O) 0wO U) U £ 4 U)

S0 0 % 00~ 0ri 0\0 1 ? co'u ooca

'o H 1) 0) H 0 rI 0)

0 0)

-P 4- 94 -P

CL4 0~ to

-P 0 (nE- c

IIXci)w

0 0

0 -11+1 c

0~l z

- H-

wLw zzLL w

C/) Li.w - z0 w

ci: w

(1)

C/)

It

L-A

4 DIA. W DIA.

SAWCUT A APPROX. 0.600',' EXTENDED BY FATIGUEPRECRACKING TO 0.700"

FIG. 2A

SDIA.

SAWCUT B APPROX. 0.350" EXTENDED BY

FATIGUE PRECRACKING TO 0.400"

FIG. 2B

FIG. 2: GEOMETRY OF NOTCH TENSILE SPECIMENS.

7

In order to study the kinetics of the slow crack propagation undersustained load, a resistance bridge was employed. The details of theresistance technique which have been previously described (15, 16),involve making the specimen one leg of a Kelvin double bridge and measuringthe increase in resistance which accompanies the crack extension. Althoughthe majority of sustained-load tests were conducted in an air environment,in certain instances an argon atmosphere was also employed. Figure 3illustrates the apparatus used for the tests conducted in a dynamic argon

environment.

IV RESULTS AND DISCUSSION

A. Smooth and Notch Tensile Tests

Ti;e results of smooth tensile tests conducted on 300 M (600"F temper)•24 . •as a function of test temperature are presented in Figure 4. These data

which were obtained at a crosshead speed of 0.050t/min. indicate that thereis an anomalous increase in the tensile strength at approximately 3000F anda possible decrease in ductility as measured by percent elongation inessentially the same range of test temperatures. The results obtained with300 M notched specimens presented in Figure 5 indicate a very noticeable,strain rate dependent embrittlement. In accordance with conventional behaviorfor a thermally activated process., such as strain aging., the temperature whereembrittlement occurred decreased as the testing time increased (i.e. at de-creasing strain rates). Figu,•nre 6 illustrates a comparable type of notchembrittlement which occurred in the 300 M tempered at 1025*F. In this casethe region of maximum embrittlement occurred at a slightly higher testtemperature than when the steel was tempered at 600"F.

The results of smooth and notch tensile tests of 4340 steel temperedat 750"F are shown in Figures 7 and 8. Although the results are similarto those obtained with the 300 M, the degree of notch embrittlement in the4340 steel at this strength level was more severe.

The smooth and notch tensile properties of the H-11 die steel arepresented as a function of test temperature in Figure 9. This material,which was double tempered at 1O000F, exhibited no anomalies in propertieswhich could suggest significant strain aging effects. It is interestingto note however, that the notch tensile strength as determined by theprecracked specimens was below the yield strength over the entire range oftest temperatures.

The 17-7 PH stainless steel (see Figure 10) also exhibited no noticeablemaxima or minima in mechanical properties that would be indicative of spon-taneous strain aging during the test. As shown in Figure I1, the AM 355precipitation-hardening stainless steel, in agreement with previouslypablished data (17), exhibited a maxima in tensile strength in the 600*Fto 800OF test temperature range. The peak tensile strength values wereshifted to lower temperatures with decreasing strain rate, indicating thata strain aging phenomenon was operative. Although the notch tensile tests

8

;' • _> ' ' . "•,- "• , .. • "- -.S..',"""'''"2 ''' - ' -- " ----- -' ' "' -- ' " " '-"-'• -;.-.

'4 •,.- -'\ , s-,S. • • .S, - •--:••." ' .'-•.-J'•.""'• ''""''"•,,.' -'- .. -= -,- • -;',- ,

BOLT - RUBBER "O'RINGPRESSURE PLATE

-TSTAINLESS END CAP

GAS OUTLET-'S E L T B-STEEL TUBE

o 0

0 0

0 "-"FURNACE

BUBBLING CHAMBE o0

a INOTCH SPECIMEN

00

0

0

0 0b

C

DIBUTYL PHTHALATEBATH

ARGON- `KHERMOCOUPLE

FIG. 3: APPARATUS FOR MECHANICAL TESTING IN ANARGON ATMOSPHERE.

"9 ~9

2600

-4

0.2TRELDSTENGHH

260

00

2200

CAP4

ý44

%. ELONGATION

16010

0 100 TS 20 0 E 00E400 5E 600*

FIG.4 TENSILE PROPERTIES OF 300M STEEL (600*FTEMPER), CROSSHEAD SPEED O.050"/MIN.

4'~ -. 10

-I?,

-

00 •0

; •:.! e90

F--

II II__ _ _ _ _ _ _ _ _ _ _ _

- CROSSHEAD SPEED70

•_a 0.005 in/rain

!• 0.05 in/min

60

I

5904

0 1o 200 oo0 400 500 / 00

•! ~TEST TEMPERATURE" O• F

•':•FIG.5: NOTCH TENSILE PROPERTIES OF 300M STEEL,(600°F TEMPER).

NO70

N_

180 I_ _ _ _ _ _ _

- - CROSSHEAD SPEED '-0.05'1 MIN

160 __

0

z 14 _ _

Ir-

zA

0

F-

Z I

54I

80)

0 200 400 600 800 :1000

TEST TEMPERATURE '-F

FIG. 6: NOTCH TENSILE STRENGTH OF 300 M STEEL (10250 F TEMPER)p4AS A FUNCTION OF TEST TrEMPERATURE.

II12

I I

220

I TENSILE STRENGTH

200~

0 _ 0

!• ~180 "-

• m ~~160 . . .-S002% YIELD STRENGTH

~09

• .• ~ 120 1

:• % ELONGATION

________

%'' ELNATO 12 z

;j 0'._1

I II 0w

0 200 400 600 800 1000TEST TEMPERATURE" 0 F

I.N FIG.7: TENSILE PROPERTIES OF 4340 STEEL (75 0 OF

TEMPER) AS A FUNCTION OF TEST TEMPERATURE,CROSSHEAD SPEED O.050'TMIN.

13

Z,.,,

200

180i

9/.

t #'° CROSSHEADCROSSHEAD 1 \ E" ,. mSPEED 0".01lin/min I ,SPEED-0..05Min,/i n

1601

0 1

• o I

z I

-I OSHEI CROSSHEAD

~ I SPEED

i ', Io.00 o

0

CROSSHEAD SPEED -0.50 in/min80-

S0 200 400 600 800 1000TEST TEMPERATURE"- OF

FIG.8: NOTCH TENSILE PROPERTIES OF 4340 STEEL1T17550F TEMPER).

14

IRS 4*..

Lw 1.1-

300~TENSILE STRENG (H

•.-o"' 2.50 •

0.2% YIELD STRENGTH

• 200[

0# 15 NOTCH TENSILE STRENGTH•2001-I ..

05

/ TET NC TENILOE SNGATI

0 0

I002040 0 00I0

50ET TELONGATIO E "' °

ITEMPER CROSSHEAD SPEED 0.050''MIN.

15"

__ 00 1-4020 0 8010

210I1 TENSILE STRENGTH

0

-180-

S150 0.2% YIELD-STRENGTH

Uf)Ul)

p.. w

120 - NOTCH TENSILESTRENGTH

-•l 90

20

I z/ _o% ELONGATIO0NI

0 w1 0

• ill 00

0 200 400 600 80C 1000

TEST TEMPERATURE " F

FIG.IO:TENSILE PROPERTIES OF 17-7PH STEEL (RH 950)AS A FUNCTION OF TEST TEMPERATURE. CROSSHEADSPEED O.050'/MIN.

16

,-- - w-..1 - .-,• 4%- ý -• • V - -. -- 7

220 -0 CROSSHEAD SPEEDTENSILE STRENGTH O.O05'ýMIN

200

CROSSHEAD SPEED 0.050'•MIN

180 %- CROSSHEAD SPE0.2% YIELD STRENGTH 0.O5d/MIN

X

160 ,•

140"160 \ CROSSHEAD SPEED

0."- 0.005 in/mi n

0 NOTCH TENSILE STRENGTH • •

'4*

'- u) CROSSHEAD SPEED O.05in/min,,

w/

c-

co

0 N

(I)CRSSEADST EPEEDR0 ATRE " O

100 % ELONGATION (CROSSHEAD SPD D0.050"/MIN)

z0

4 -1w

TEST TEPEA 0R -85OF

suggest that a minimum in notch properties may be present at temperaturesgreater than 800*F, the data are inconclusive. Test temperatures greaterthan 800*F 'kere not employed to avoid the introduction of variations inmixrostructure which may be produced by testing above the aging temperatureuied during heat treatment,

The results of tensile tests conducted on the maraging steel arepresented in Figure 12. The data obtained from smooth tersile specimensindicate that a peak in both tensile and yield strength o, curred atapproximately 650*F when a crosshead speed of 0.050"/min, was used.Definite minima in notch strength values were observed ii, approximatelythe same region of test temperature where the increase in smooth strengthoccurred. The minima in notch tensile strength were shifted to slightlylower temperatures when the strain rate was decreased.

The influence of strain aging during the tensile test was observed invarying degrees in the low alloy martensitic, the maraging, and the AM 355steels. The 17-7 PH and the H-11 steels did not exhibit any strain agingeffects for the particular heat treatments investigated. There was ageneral tendency for the phenomenon to oc.ur at higher test temperaturesin steels which were heat treated at higher tempering temperatures. Themorphology of the carbide may therefore play a dominant role in the strainaging behavior,

The presence of the strain-aging type embrittlement in high-strangthsteels raises two important questions:

1) What influence will this strain aging have in engineering, applications?N 2) What is the mechanism producing the strain aging behavior?

The importance of the strain-aging embrittlement in determining theperformance capabilities of structural components is difficult to generallyevaluate since it is dependent on the particular application. In cases,such as AM 355, Ni-Co-Mo maraging steel, and 300 M tempered at 1025*F,the basic crack propagation resistance of the material as measured byconventional room temperature tests was relatively high. In these steelsthe strain aging phenomenon produced only a slight degradation in notchproperties and therefore would not be expected to significantly influencecomponent reliability. The strain aging behavior present in steels suchas 4340 tempered at 750"F and 300 M tempered at 600"F produced a morepronounced decrease in notch strength and under these conditions theembrittlement could be an important consideration.

Strain-aging effects in tensile tests may be indicative of susceptibilityto delayed failure under constant load. This type of static fatigue can be aparticularly perplexing engineering problem since the embrittlement may notbe apparent in conventional tensile tests and only occur under the propersequence of time, temperature and applied stress. Although the classicalexample of static fatigue of high-strength steels occurs with hydrogen,

1 18

I+

* K.÷ + , + , . .,+ .. > ...

?60 .,,

240

TENSILE STRENGTH

220 0.2% YIELD ' 0S TRE NGT H-'1 ..- •"' -- • \"t-- lI

200 k

0N

80- NOTCH TENSILE

I,'• STRENGTH .4N '_S163 CROSSHEAD SPEED o

0o 0.005in/min oS0.050 in/min

0 o & 0~.500 in/min -

S140 I0

LI-

120 . ... /60U'/ 8 -

ELONGATION 4o

100, 10 200 400 600 800 1000

TEST TEMPERATURE OF

FIG.12" TENSILE PROPERTIES OF 18% NICKEL MARAGING

STEEL.

19

there is a theory which predicts that comparable effects could be producedby other interstitials (2). In fact, there have been instances of non-hydrogen induced delayed failure which have been attributed to carbonand/or nitrogen (18). In addition, extensive sustained load tests con-ducted on steels for bolt applications at temperatures in the 900 - 1200"Frange have indicated a time-temperature dependent notch embrittlement (19).The question arises as to whether the strain aging behavior observed in thisinvestigation can produce a significant embrittlement under conditions ofsustained load, where the time for interstitial movement is maximized.

B. Failure of High-Strength Steels Under Sustained Loads

The results obtained from stress rupture tests conducted at 300'F, 400*F,and 550"F with 300 M high-strength steel tempered at 600*F are shown inFigure 13. In order to compare the results for different temperatures, theapplied stress was normalized by dividing it by the conventional notch tensilestrength obtained at the particular test temperature and with a crossheadspeed of 0.050"/min. (see Figure 5). The specimens tested at 400"F, which

N is in the temperature region where strain aging was prominent in the tensiletest, failed in relatively short times at applied stress values as low as0.6 the normal notch tensile strength. This rather large rarge of delayed

N] failure was not present in tests conducted at 300*F and 550'F. The factthat failures did not occur at 550OF indicates that the fractures at 400"Fcould not be attributed to conventional creep processes. It is also veryimprobable that the failures could be due to hydrogen since the steel wasi- -n unhydrogenated condition, and hydrogen-induced delayed failures arenL present at temperatures above 300*F (20).

Sustained load tests conducted on 300 M specimens tempered at I025*Fdid not show any failure in 200 hours at temperatures below 800*F and atapplied stress levels which were 90% of the normal notch tensile strength.The static fatigue curves obtained from notch tests at 800"F and 900*F areshown in Figure 14. In this series of tests the failures appear to be dueto conventional creep behavior,

Stress rupture tests conducted on notch specimens of the 18% nickel-maraging steel are shown in Figure 15. Once again no appreciable regionof delayed failure was present which could be attributed to non-creep typeprocesses. Sustained load tests on 4340 steel tempered at 750*F, and AM 355,showed no delayed failure in 200 hours at an applied stress of 0.9 times theconventional notch strength and in temperature regions where strain agingwas experienced.

Since the association between strain aging and delayed failure was onlyapparent to any appreciable extent in the 300 M steel tempered at 600*F,this material was selected for further evaluation of the delayed failuremechanism. The macroscopic appearance of the fracture surface of a specimenof 300 M steel which failed under static load is shown in Figure 16. Thefailure pattern consisted of the following segmentst

20

W _NN _ %Y _ ......

LZ

-< m

2 zLO ~W

0 1Z

1W 0

WW0 00

____ __cr 0

U< w

¼ 0 ' 000

I- -

U) 0

WDzoUW L

0 a7) Go( LO~

HIDN381S 3-11SN3i. HOION 0.1 SS3RLS 3uflivd :ounv

21

00-I.-- Wz/ZLW

w0i /0D z

W oz

0 4 NW LL

-ww

ZLLn Z Z

DD N 0w0

0-l~~~~~d7 001 SwL GfdVL.

22Xw a

~ ~ Jm

oz

o(r

W W

0D //L w

0 cn 0

CIDw w* /t

0r Z I 0

Z~H

z

Z L LLW:z

U)LLHw Wz

8~ _ 0

M N N - 0 - _l

I~~0 000 CS3iL lu ld*1 (

23O ý¾, c 0F

.-..

06378-3

CRACK EXTENSION ON INITIAL LOADINGFATIGUE CRACK, //

[SAWCUT RAPID CRACK PROPAGATION

"•" I'-~INITIAL

'NOTCHISLOW CRACK GROWTH

APPLIED STRESS 64,700 PSI ; FAILURE TIME 19.5 HOURS

FIG.16: MACROSCOPIC APPEARANCE (MAG. IOX) OF SUSTAINED

" LOAD FAILURES IN 300M STEEL SPECIMEN , (6000 FTEMPER) TESTED AT 400°F IN AN AIR ENVIRONMENT.

"v- .• 24

`-" l ""

a) Initial notch (sawcut plus fatigue precrack),

b) Crack extension which occurred during the applicationof a sufficiently high initial load,

c) Slow crack growth which occurred under sustained load,

d) Rapid crack propagation.

The heat tinting which occurred during the test allowed each mode of crackpropagation to be readily distinguished. The appearance of the slow crackwas considerably more fibrous than either the crack extension that occurredon initial loading or the fatigue precrack. In an effort to further studythe kinetics of the slow crack growth, resistance measurements were em-ployed daring the sustained load tests on the 300 kl high-strength steel°The variation of resistance with time is presented in Figures 17 and 18for test temperatures of 300"F and 00*F. At 300"F, an incubation timewas observed before slow crack growth was initiated and this incubationtime was extremely sensitive to applied stress (see Figure 19). At thet00*F test temperature, ao incubation was observed at any applied stresslevel where delayed failure occurred and slow crack growth was initiatedas soon as the load was applied.

The lack of an incubation period is typical of environmentally-inducesdelayed failure in high-strength steels (4). To evaluate this possibility,sustained load tests were conducted on notch specimens of 300 M at 400F inan argon atmosphere. The results presented in Figure 20 indicate that thedelayed failure at 400OF was completely eliminated when an argon atmn spherewas used. The static fatigue in 300 M steel was therefore induced oyenvironmental effects and was presumably due to stress corrosion by theair atmosphere. The strain aging effects previously described with referenceto tensile testing were re-'valuated by conducting notch tensile tests on300 Y, (600*F temper) in an ai gon atmosphere. A minimum in notch tensilestrength was present at the same test temperatures in both test environments,indicating that the tensile test data were associated with a material phe-nomenon and not with the external test conditions.

Although the delayed failure is environmentally induced, it is stillsignificant that the temperatu.- region where delayed failure occurredcorresponded to the range where strain aging effects were observed. Asthe crack slowly grows in a constant load test, because of environmentaleffects, the applied stress is increasing due to the decrease in loadcarrying area, On this basis, the sustained load test is comparablc to aslow 3train rate tensile test provided a mechanism, such as stress corrosion,is available to continually produce a slow crack propagation. Strain-agingeffects should therefore influence the failure properties in this temperaturerange just as they decrease the notch tensile strength in conventional tensiletests. The lack of delayed failure in the other materials investigated is ingeneral agreement with their decreased susceptibility for the initiation ofslow crack growth and stress corrosion effects (h)*. Additional work is stillneeded to determine the mechanism whereby mild environments such as air canproduce such a rapid deterioration in load carrying ability of a high-strengthsteel.

* A possible exception to this statement is the H-11 steel.

600 _ FRACTURE

0

500

wz

400

w

300

w

200 _

INITIATION OF SLOW:'°•'ICRACK GROWTH

o I

0 20 40 60 80 100

TIME -'MINUTES

FIG. 17: KINETICS OF CRACK GROWTH AS DETERMINED BY RESISTANCEMEASUREMENTS ON SUSTAINED LOAD TESTS OF NOTCHSPECIMENS OF 300 M STEEL, (6000 F TEMPER) TESTED AT3000 F, UNDER AN APPLIED STRESS OF 64,500 PSI.

26

* N

1400

FRACTURE

1200

I000

-i ,AUDIBLE

800

W, 800 -_ _ _ _ _ _ _ _ _ _ _ __ _ _ _

II

wC.)

• 600 00____0_0_Lo AUDIBLEU)w

w

cl)

200

_ _w. N_ _ _ _ _ _ ._

0 10 20 30 40 50 60TIME- MINUTES

FIG. 18: KINETICS OF CRACK GROWTH AS DETERMINED BY RESISTANCEMEASUREMENTS ON SUSTAINED LOAD TESTS OF NOTCHSPECIMENS OF 300 M STEEL, (6000 F TEMPER) TESTED

AT 400OF UNDER AN APPLIED STRESS OF 45,750 PSI.

27

-41A~, *.* V -~

f - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - -

30

0 z

___ ____ ____LL o0 0

C.)

__( 0

cr w

U))______ _ Q:

wzi~ Zt L

w

a-

ww

w O

0r Z

w -z

0. 0f >

WA w w

a:

00 0 0

I Sd 0001 SS38LLS 031flddV

28

.all

Li:(j)00

H0

1 0

4- W4.-

L<L

w

0

* I- 0

zDM U)

00

z00 0O

o ww

LL U-j U)

w zocnWa

w Li..-J -0

00 0L

iSd 0001 SS381JS Q31-lddV0

29--*.~-- . - .

C. Possible Strain-Aging Mechanisms in High-Strength Steels

Strain aging and "blue brittleness" have been extensively investigatedin mild steels and at piesent the following three factors are believed to be

Vý% . associated with the increase of smooth strength properties which accompanystrain aging phenomena:

1) Short-range ordering about moving dislocations (21, 22),

2) The dragging of a solute atmosphere with a dislocationline (23, 24),

3) The formation of a precipitate about a dislocationline (7).

The short range ordering of solute atoms about dislocations has beendescribed in some detail by Schcedc and Seeger (21) and Wilson and Russell(22). The interstitial atom produces a distortion with a symmetry whichis different from the symmetry of the lattice. On this basis, the inter-stitial will take up preferred positions within the lattice when a non-hydrostatic stress field is present. Under certain time-temperature con-ditions the ordering of the interstitials about the moving dislocation willproduce a restraining force because energy is dissipated during the orderingwhich takes place. This dragging force is not apparent at low dislocationvelocities or high-test temperatures where the ordering always reaches thermo-dynamic equilibrium. At high velocities or low temperatures the time is in-sufficient for redistribution to take place and the restraining force is alsoabsent. An ordering phenomen•on of this type produces the well-known Snoekpeak in internal friction studies and is characterized by 'he relativelyshort times involved, since redistribution only takes place over atomicdimensions. A second type of dragging force which can occui is due to the

S.for!iation of a Cottrell cloud about a moving dislocation (8, 23). Thetimes required for this strengthening to occur should be approximately100 times greater than that involved in ordering since the interstitialshave to move over greater distances to form the cloud (25).

The role of strain-induced solution and precipitation of carbide particleshas been used in both mild (7) and high-strength steels (10, 11) to describestrengthening effects produced by tensile straining, followed by suitableaging. A simple application of precipitation as a mechanism for strain agingwhich occurs during the test (blue brittleness) is difficult to visualizesince it would require the precipitation process to occur on a moving dis-location. Certain internal friction peaks, which involve oscillating dis-locations however, have been attributed to very small coherent carbide pre-cipitation (26). In addition, the flow process may be fairly complex in thehigh-strength steels and may involve intermittent moveme it of specific dis-locations which would allow precipitation effects to become operative. Onthis basis, the possibility of solution and reprecipitation of carbides in"blue brittle" behavior should not be disregarded.

___. 30

L

Before briefly discussing the possible significance of each of thesethree strain aging mechanisms to high-strength steels, it is interesting to

evaluate the activation energy for the strain aging behavior. The embrittle-ment can be expressed in terms of the conventional rate equation:

&Ae

N. ~where: C strain rate

A constant

Q activation energy

R = gas constant

T = absolute temperature

To evaluate the activation energy Q, the logarithm of the crosshead speed wasplotted as a function of l/T. T was taken as the temperature where the tensilebtrength was a maximum, or the notch tensile strength was a minimum. Theresults are presented in Figure 21 for the time-temperature dependency of thenotch strength minima and in Figure 22 for the smooth tensile strength maxima,These data indicate that both parameters conform to the conventional rateequation and have an activation energy of approximately 25,OOO cal/mole.Due to the difficulty in selecting the exact maxima or minima from themechanical property data the activation energies can have a spread ofapproximately +hOOO cal/mole. The data which are available for the "bluebrittleness" of mild steels (27) have been plotted in Figure 23. Theactivation energy for this embrittlement in mild steel is essentially equalto that which occurs in the high-strength steels and strongly suggests thatcomparable mechanisms are operative.

Although stress-induced ordering of carbon or nitrogen and the drag ofCottrell clouds of interstitials with dislocations are generally accepted ascontributing to the spontaneous strain-aging behavior in mild steels, thefollowing results must still be explained before these mechanisms can beapplied to the present data on high-strength steels:

1. The temperature region where strain-aging effects occur ishigher for steels which have been heat treated at relativelyhigh tempering temperatures.

2. The apparent activation energy for the strain aging effects inboth mild steels and high-strength steels is slig1tly higher"than that obtained for the diffusion of carbon or nitrogen inalpha iron.

" - 3. Stress-induced ordering of interstitials (Snoek peak) is notfound in internal friction studies of quenched and temperedsteels (28).r 31

0. _._ 7000 F 6000 F 5000 F 4000F

• --

•"1•''•0.5-- t o• • ..

4340 0

_______ ______ ~ 3O00M 0

",(6000 F TEMPER)

Q 23,900 cal/mole

•v---------

z

Z;. 0.05 --A----O

0w - -

w

0.00

"____NVE SE TEM ERTUR "_ I0-___K-_

aq"0.0031 -

300INVER(S05E TEMPER)TR-I

C--:

7000 F 6000 F 5000 F 4000 F1.0 : "_ _ _ _

~ °'4.1

300 M

(6009 F TEMPER)

0.1 _ _ _ _ _ _ _ _ _ _ _ _ _ _

Q 26,200 cal/mole

w% .01

a::

AM355 i

2 0\o •I

Q 23,500 cl/0mole-•_.-

.001

1.4 1.6 1.8 2.0 2.2 2.4 2.6

-_ INVERSE TEMPERATURE - 10-3, K-I

FIG.22* ARRHENIUS PLOT OF STRAIN RATE AND TEMPERATUREAT WHICH MAXIMUM IN SMOOTH TENSILE STRENGTH

OCCURRED.33

700 690 500 4000 F

10 - _ _ _ _

THIS INVESTIGATIONZ'- __, NOTCH TENSILE STRENGTH

___ I__ MINIMA (4340 STEEL)

(REF. 27•• ,10 • MILD STEEL) "

0

:•;• "% I0-2z10 - -' 0 ---

wa---) THIS INVESTIGATION

300 M, i025-F TEMPERW -3x 10-- _ _

• )(I. (REF. 270,, :MILD STEEL)

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2

INVERSE TEMPERATURE - 10-0 K I

FIG. 23: COMPARISON OF DATA OBTAINED FOR "BLUE-BRITTLENESS"OF MILD STEEL (DUCTILITY MINIMA) WITH NOTCHTENSILE STRENGTH MINIMA OBTAINED FOR HIGH-STRENGTH STEELS.

34

v4

Some insight into the poss' ntribution of stress-induced inter-stitial ordering and the Cottrel± atmosphere effects can be made by employinga method used by Hahn, Reid, and Gilbert to calculate the effect of orderingon the relationship between strain rate and lower yield stress in mild steel(29). Although these techniques and the assumptions involved are probablyover-simplified when applied to high-strength steels, the results can in-dicate the relative temperature regions where these interstitial effectscan be prevalent. The details of the calculations are presented inAppendix I and the results are summarized in Figure 24. The degree ofstress-induced ordering is presented in terms of a dimensionless parameter(F) which yields a relative measure of the increase in strength which canbe expected by the ordering. The calculations are very dependent on thedislocation density (see equation 2, Appendix I). The results presentedin 8 Figure were ob tined for two arbitrarily chosen dislocation densities,10 and lO± lines/cm . The significant point is that at higher dislocationdensities, which would be expected at the lower tempering temperatures, thecalculations predict that the strain aging effects should occur at lower testtemperatures. These predictions are in qualitative agreement with the ob-served results on the high-strength steels. There is sufficient uncertaintyin the proper selection of the dislocation density so that an accurateappraisal cannot be made of whether a stress-induced ordering of inter-stitials or a gross diffusion of interstitial atmospheres is the primarymechanisw. The strengthening due to the drag of interstitial clouds wouldbe approximately two orders of magnitude slower (25) than the stress-inducedlocal ordering pictured in Figure 24. This would raise the temperature atwhich the strengthening peak would occur by approximately 2000F.

The mechanism of spontaneous strain aging which depends on an inter-action between interstitials and dislocations requires that the activationenergy for the strengthening process be equal to the activation energy forthe diffusion of the responsible atomic species. Since carbon and/ornitrogen are believed to be responsible for the strain aging effects in the400*F to 800F temperature range the activation energy for the strengtheningprocess should be approximately 18,000 to 20,000 cal/mole (8). The ex-perimental data on "blue brittleness" in high-strength steels indicatesthat this embrittlement is governed by a slightly higher activation energy(approximately 25,000 cal/mole). One possible explanation for this differencqis that the actual peaks of tensile strength versus test temperature which areused to determine the activation energy for strain aging are dependent notonly on the strengthening process, but are a combined effect of the strength-ening process, superimposed on the normal decrease of strength which accompa-nies an increase of test temperature. This point is illustrated in Figure 25.The sole contribution of the strengthening process is pictured in Figure 25Aand this process, in accord with theory (21, 25), can have an activationenergy of approximately 20,000 cal/mole. The normal decrease in tensilestrength with increasing test temperature is shown in Figure 25B and thecombined effects are shown in Figure 25C. The resulting strength peaks arethen dependent on the two processes of general softening due to the in-creased test temperature and the strengthening due to interaction effects.As shown qualitatively in Figure 25C, these peaks can produce an apparentactivation energy which is greater than that required solely for thestrengthening process. 5

* .- 35

P 10'0 LINE"i/CM2

"" •= I0 LINES/CM

r m 4

z 3

a 0I II.

00

L) 0 0-- -. L 0 C 0 1&--

0 200 300 400 500 600 700TEST TEMPERATURE - F

FIG. 24: CALCULATED FORCE ON DISLOCATION DUE TO

INTERSTITIAL ORDERING AS A FUNCTION OF

TEMPERATURE.

36

1 22

13. C%

-. (A)

STRENGTHENING DUE TO INTERACTIONS

PRODUCES TRUE ACTIVATION ENERGY (Q)

coW NORMAL STRENGTH

I.- *, DECREASE()U) (13

C3

Woo TOTAL (A) +(B)

IIPRODUCES APPARENT2¶ACTIVATION ENERGY d > Q

*PUE TO SHIFT IN PEAKS OF(CISTRENGTH OURVES(4,. 4; f) I___ (C)__

200 400 600 800 10

TEST TEMPERATURE -'F

V FIG.25: SCHEMATIC ILLUSTRATION COMPARING TRUE ACTIVATIONENERGY OF INTERACTION PROCESS AND APPARENTACTIVATION ENERGY OBTAINED FROM TENSILE TESTDATA. 37

No completely satisfactory explanation is available to explain the factthat internal friction peaks of the Snoek type are not found in quenched andtempered martensitic steels, yet the existence of interactions between inter-stitials and dislocations are necessary to explain the observed strain-agingeffects. Perhaps the presence of certain interaction effects are dependenton a specific degree of plastic flow which occurs in the mechanical tests butwhich is not normally present in internal friction studies.

Additional support for the mechanism of spontaneous strain aging whichinvolves an interaction of interstitials with dislocations, rather than theformation of a precipitate, is obtained from electron micrographs and pre-stressing experiments. Figure 26 presents the microstructure of 300 Mbefore and after stressing in the temperature range where strain agingtakes place. In no case did electron micrography, using replica techniques,indicate any difference in carbide morphology between the specimens whichhad been stressed to exhibit strain aging and those which had not.

4 If a precipitate were formed during the tests, in the temperatureregion where strain-aging effects were present, then this precipitateshould also influence room-temperature properties by causing notch em-brittlement and an increase in smooth strength. Tests were conducted byprestraining 1" wide, smooth specimens of 300 M, 1.0% at 400OF and allowingthe load to remain on the specimen for various times. The specimens werethen water quenched, center drilled, and a sharp notch was generated byfatigue precracking. The specimens were tfsted at room temperature and theresults are presented in Figure 27. If an embrittling precipitate wereformed by the combined straining and aging treatment at 4000F, then it shouldbe even more effective towards promoting embrittlement at the lower testtemperatures. As shown in Figure 27, no evidence of any room temperatureembrittlement was present. Similar tests were conducted with smooth specimensof 300 M, where prestraining temperatures of 250F and 350*F were employedwith prestrains of 2.0 and 3.0%. In this series of experiments the specimenswere quenched from the prestressing temperature and immediately tested atroom temperature to determine if any variations in smooth strength occurred.For the aging times investigated (up to 10 minutes), no significant variationin room temperature tensile strength was obtained. These results indicatethat "blue brittleness" in high-strength steels is a unique function of theparticular temperature region where sufficient interstitial mobility ispresent. The strain aging effects are not inherited at other test temperaturesand this factor strongly supports a mechanism involving an interaction betweeninterstitials and moving dislocations.

38

4 , .*..,. • • • ,••. . • • - • . . , -• .- • - - -• - - - - - . -* - * - - .o - - - -- - - -

S-,

'•, 0836

• 300 M SPECIMEN, STRESSED IN STRAIN AGING REGION,•<! MICROSTRUCTURE OF NECKED REGION

i• 0839

.-S

•'J 3 00 M S PECIMEN -UNSTRESSED

¼FIG. 26" ELECTRON MICROGRAPHS OF 300 M STEEL, TEMPERED

•'- "AT 600* F, MAG. 1 0,000 X.

39

-ZN)

01

a- 0w

S160 --o

0w

z 120w

wL-J

0 100

I-

I-

C-)

I--

00

Mz

0 I0 20 30 40 50 60i AGING TIME UNDER STRESS AT 4000 F ""MINUTES

~ F'IG. 27; INFLUENCE OF PRESTRAINING I%, AND SIMULTANEOUS

AGING AT 4000 F ON THE ROOM TEMPERATURE

NOTCH TENSILE STRENGTH, 300 M STEEL, 600°FI ~TEMPER. 4

V CONCLUSIONS

High-strength steels can exhibit significant strain aging effects duringtensile testing, when tested in the temperature range 300*F to 800*F. Thestrain aging phenomena are characterized by an increase in smooth tensilestrength and a decrease in notch tensile strength. The strain-aging behavioris dependent on strain rate in a manner which is analogous to "blue brittle-ness" in mild steels. The activation energy for the phenomena is comparableto that shown by "blue brittleness" and is approximately 25,000 cal/mole.The strain aging effects were observed in the low alloy martensitic steels(300 M and 4340), the precipitation-hardening stainless steel (AM 355), andthe maraging steel (18% nickel, 7% cobalt, 5% molybdenum), No strain agingwas noted in the H-11 hot-work die steel or the 17-7 PH precipitation-hardening stainless steel.

Constant load tests were conducted to determine a possible correlationbetween the strain aging effects noted in the tensile tests and the sus-ceptibility for delayed failure. An appreciable region of delayed failurewas present only in the high-strength 300 M steel tested at hOO"F° Thisdelayed failure phenomenon was extremely sensitive to environmental effectsand was completely eliminated by testing in an argon atmosphere. Althoughthe strain aging did not initiate failure under sustained load, it wasbelieved to occur during slow environmentallye-induced crack propagationand contributes to the decrease in load carrying ability of notch specimenstested at hOO4F.

The spontaneous strain aging in high-strength steels which occurredduring tensile testing was attributed to an interaction between interstitialsand moving dislocations rather than precipitate formation.

~41

LL

VI BIBLIOGRAPHY

1. L. J. Klinger, W. J. Barnett, R. P. Frohmberg, and A. R. Troiano,"The Embrittlement of Alloy Steel at High Strength Levels", Trans.

ASM, 46, 1557, (1954).

2. A. R. Troiano, "The Role of Hydrogen and other Interstitials in the

Mechanical Behavior of Metals", Campbell Memorial Lecture, Trans*ASM, 52, 54, (1960).

3. M. E. Shank, C. E. Spaeth, V. W. Cooke, and J. E. Coyne, "Solid-FuelRocket Chambers for Operation at 240,000 psi and Above", Metal Progress,76, No. 5 and 76, No. 6, (Nov. and Dec., 1959).

4. E. A. Steigerwald, "Delayed Failure of High-Strength Steel in LiquidEnvironments", Trans. ASTM, 60, 750, (1960).

5. E. H. Phelps and A. W. Logii- w, "Stress Corrosion of Steels for Air-craft and Missiles", Corrosion, 16, No. 7, 97, (July, 1960).

6o E. S. Davenport and E. C. Bain, UThe Aging of Steels", Trans. ASM,23, 1047, (1935).

7. B. B. Hundy, "The Strain-Age Hardening of Mild Steel", Metallurgica,53, 203, (1956).

8. F. R. N. Nabarro, "Mechanical Effects of Carbon in Iron", Report onStrength of Solids, Phys. Soc., London, 38, (1938).

9. H. W. Paxton and C. C. Busby, "Strain Aging of AISI 434U0", Trans.AIME, 206P 788, (1956).

10. E. T. Stephenson, M. Cohen, and B. L. Averbach, "Effect of Prestrainand Retempering on Ultra-High Strength Steel", Trans. ASM, 54, 72, (1962)t

U. P. J. Fopiano, S. Das Gupta, D. Kalish, "Effect of Mechanical and ThermalProcessing on High Strength Steels", Final Tech. Rep. WAL 320o 4/4-3 forWatertown Arsenal Laboratories Contract DA-19-020-Ord-5253, (June, 1962).

12, J. E. Srawley and C. D. Beachem, "Crack Propagation Tests of Some High-Strength Sheet Steels", NRL Report 5263, (January 10, 1959).

13. J. D. Morrison and J, B. Preston, "An Investigation for Determining theCrack Propagation Resistance of High Strength Alloys", Southern Re-search Institute, Bimonthly Progress Report 5, Bureau Naval WeaponsConte No. as 60-040 c. (May 31, 1960).

14. ASTM Committee Report, "Fracture Testing of High-Strength SheetMaterials", ASTM Bulletin, No. 243, 18, (January, 1960).

-~ 42

BIBLIOGRAPHY (Continued)

15. W. J. Barnett and A. R. Troiano, "Crack Propagation in the HydrogenInduced Brittle Fracture of Steel", Trans. AIRE, 209, 486, (1959).

16. E. A. Steigerwald and Go L. Hanna, "Initiation of Slow Crack Prop-agation in High-Strength Materials", ASTM, Preprint 74, (1962).

17. Do C. Ludwigson, "Semiaustenitic Precipitation-Hardenable StainlessSteels", DMIC Report 164, (December 6, 1961).

18. T. E. Scott, "Role of Interstitial Solutes in Fracture Mechanics",Ph. D. Thesis, Case Institute of Technology, (1961).

19. G. Sachs and W. F. Brown, "A Survey of Embrittlement and NotchSensitivity of Heat Resisting Steels", Symposium on Strength andDuctility of Metals at Elevated Temperatures, ASTM, STP 128, 6, (1952).

20. J. Destrez, "Delayed Failure in High-Strength Steel", Me S. Thesis,Case Institute of Technology, (1962).

21, G. Schoeck and A. Seeger, "The Flow Stress of Iron and Its Dependenceon Impurities", Acta Mete, 7, No. 7, 409, (July, 1959).

22. D. V. Wilson and B. Russell, "Stress Induced Ordering and Strain-

23. A. H. Cottrell and M. A. Jaswon, "Interactions of Solutes with Movingr Dislocations", Proc. Roy. Soc. (London) A199, 104, (1949),

24. A. H. Cottrell, "Dislocations and Plastic Flow in Crystals", p. 136,Oxford University Press, (1953).

25. G. Schoeck, "Theories of Creep", Mechanical Properties of Metals atElevated Temperatures, edited by J. Dorn, Mc Graw Hill Company,New York, (1961).

26. T. Mura, I. Tamura, and J. 0. Brittain, "On the Internal Friction ofCold Worked and Quenched Martensitic Iron and Steel", ASTIA No. 646700,Northwestern University, (September 6, 1960).

27. H. S. Gurev and W. M. Baldwin, "Research on Strain Aging Effects inTitanium", WADC TR 59-223, (September, 1959).

43

Etl

BIBLIOGRAPHY (Continued)

28. P. Stark, B. L. Averbach, and Morris Cohen, "Influence of Microstructureon the Carbon Damping Peak in Iron Carbon Alloys", Acta Met., 6, 149,(1958)o

29. G. T. Hahn, C. N. Reid, and A. Gilbert, "The Relation Between Delay*. Time, Strain Rate and Strain Aging Phenomena in Mild Steel", Acta

Met., No. 8, 10, 747, (August, 1962).

-," .-.

N

APPENDIX I

Calculation of Ordering Effects Due to Interstitials

Schoeck and Seeger (21) present a functional relationship between theforce exerted on a dislocation by interstitial ordering and a dimensionlessparameter which involves the relaxation frequency for the interstitial andthe speed of the moving dislocation. This relationship which can beexpressed as:

where: F M the force exerted on the moving dislocation expressedin units of U/R,

Li = the line energy of the dislocation,

R = the effective radius of the dislocation,

4 = the speed of the dislocation, and

V = frequency factor.

The relationship given in equation (1) has been solved graphically(see eq. 28, Ref. 21) and is presented in Figure 28. The problem ofdetermining the restraining force produced by interstitial ordering on amoving dislocation then resolves itself into determining the values ofCL/j R and using Figure 28 to obtain the desired result.

The average speed of the moving dislocations (0() can be related tothe strain rpte (i) by using the relationship presented by Hahn, Reid, andGilbert (eq. 1, Ref. 29):

- 0.5 bf f- (2)

where: C = strain rate

b = Burgers' vector

f' = constant (approx. 101)

dislocation density

c( = average dislocation velocity

F - The average relaxation frequency (•) can be determined from therelationship (21):

p = 1.52 V~

where: D - diffusion coefficient and

b - Burgers' vector

Using the following constants:

-8b 2.47 10 cm

D 5 5.2 x I 0 "9000/T cm2/sec. (ref, 8)

R - 1.84 x I0"20

-• Rcm (ref. 21)

the value ofs which have been determined for arbitrarily chosen values offA 10 /cm and O = lolO/cm2 , are sum•arized in Table III,

44

46

N --

0 0 0 0 0 0

19 0H0 00 0~f

0 0 000 0

00 0H~ H

0N 0H

p) c'coZ corC:N 0 H1I- 0 lf

H IN ON ---t m1 . 0 ON0 CZJ

I 0ý l. Hj Hl H- 0- HN H0 H 0 0 0O

~0 0 a0 0E-41 H H HI H H H - A oc

0 Ho aco H Ir XD ' X N

0 0-

zII 00i *- I

.1 C) 00x0o 0)I H- ONCjV c e C

EO \0 co 02 CC I-p * H' H- C 0 r

HOU CjI H H Ho

C.)C. C) E--i crs C H .

cl 0) (1)8 o0M 0M CMI mM ( 4 t~N -

H a) 0 0 0 0 0 0 07

4.' U

0.6

0.5

0.4

"0 0.3

I 0.L-

0.1 _

0 2 3 4

a

FIG. 28: THE FORCE TO MOVE A DISLOCATION IN STEADY STATE

AS A FUNCTION OF DISLOCATION VELOCITY (a).

48

-- 'T 74FT-:

0 * 0

40 co*04

~~ C- 04> CI~C4t--4' 0- 0.- c - Z- , C % '-3&'0 9

C ~ ~ ~ ~ , w44 -- j0~~4C0C

t'04-C40't.404-V

-4 - '-4

@4 2() 4- -f 0 1- , 4 25. n. 0, 0 0> :h 4

...c C) 0 4' C., I. CI

*5. ~ C 4.4 4. .0 441C

140 ( I -4 z~v Q D . i

0 >,oi .00 .0.40.

-j - 2 0-' -; '0 0 00 00 cc

> .0 4- CD 4C . 0 0 4) ~.~ - 0~ 03. C c

Q c 0 co 4.00'-, 4oC '0 , . 40 .4.40~~ o. o i00 -4 4

v 14 -444 .- 4. z C44j o> 2 co0a0.0 '.OE! 4D 0 0~ 0 0.0 ~ 40.4

402 14-4 4 - -00 @4 to)0.4~~ a~~'-' 0~ 0 65 4 a* '0

(D00 -P .)00 0 4-

4- a; 9.1- .40C 4 0 -04 4J O a4

0 @4- 0~L. 0 E! 4 D 0-i- 0 0v-., 00o.6C 0) - T0 43 cz I

Q01. 0 W 01 § 1- 10 aO i9 ;40 2 4...O. ~ C4' 4

4 0 3 ~-- 1.4 cc.4-~ ~ .1 .0 o0 9) 044~4@U)0 .. 4 tO .0 . 4-

C-C

@4 -

< E 30 43 C, 10 00;;.

.f4 4 .4 . .4i

0 o43 ) a~ 10C0 4D44 0.4eC1 (4 0,4 O ->0.0 C C.0 Le- 0X

la) I . I. C7 ý4 0

000 CD0.42-40 4 .0 co 4)>

-I~L to -. @45)153.

''4,4- 000 I4

0 zo 01)

MIMI_ z 4.44) X -1 0 c Ne-1 O, C4 C.4-. .1- 0 0 co.0 Q0* .4

0.0 S..- 440 to

.4 *C0 C 00 0 1- 0 . Ico -`0 . x' -,4 .C)4) " a0>4 4 a)

r- 0 ca v" "- to

-4 0 ' :Z'3 S.- ~0 C6 . 0 0

OD 'a 0 00 0

w 0