.,SERIAL NO. SSC-61

!.’”1

,.

.,. .

II

1

,-;!. .

‘~-,,-’

>.>

,,,

“’ :.

..

0z

Fifth

~~PROGRESS REPORT

(Proiect SR-99)

,,’ on

,,,.

THE FUNDAMENTAL FACTORS INFLUENCINGTHE

I

1

,.

,-,

I

BEHAVIOR OF WELDED STRUCTURES:

The Effect of Subcritical Heat Treatment-

on the Transition Temperature of a

Low ‘Carbon Ship Plate Steel I,,

,,

by

E. B. Evahs and D. J. Garibotti

CASE INSTITUTE OF TECHNOLOGY

I

Transmitted through

NA710NAL ‘RESEARCHC10UNCIL’5

COMMITTEEON SHIP STEEL

Advisory to

SHIP STRUCTURE COMMITTEE

,!,,

l“Division of Engineering and lnd”stri~l Re~earCh

Na!ional‘Academy of Sciences -I National Research Council.1

Washington, D. C. ,,

I ,,:

October 30, 1953

I

-..

SHIP STRUCTURE COMMITTEE

MEMBER AGENCIES: ADDRESS CORRESPONDENCE TO:

BUREAU OF SHIPS, D.CPT. DP NAVY SECRETARY

MILITARY SEA TRANSPORTATION SERVICE. Dcrr. gF NAVY SHIP STRUCTURE COMMI’fTEE

UNITED STATES COAST GUARD. TUEASURY DEPT. U. S. COAST GUARD HEADQUARTERS

MARITIME ADMINISTRATION. DEPT. OF COMMCRCK WASHINGTON 2s. D. C.

AMERICAN BUREAU OF SHIPmNa

October 30, 1953

Dear Sir:

As part of its research program rej.atedto theimprovement of hull structures of suips, the Skip Struc-ture Committee is sponsoring an investigation on %PheFundamental Factors Influencing the Behavior of WeldedStructures under Conditions of Multiaxial Stress andVariations of Temperature” at the Case Institute ofTechnology. Herewith is a copy of the Fifth ProgressReport, SSC-61, of the investigation, entitled ‘TheFundamental Factors Influencing the Behavior of Weldedstructures: The Effect of Subcritical Heat Treatmenton the TransitIon Temperature of a Low Carbon Ship PlateSteel” by E. B. Evans and D. J. Garibotti.

The project is being conducted with the advisoryassistance of the Committee on Ship Steel of the NationalAcademy of Sciences-National Research Council.

Any questions, comments, criticism or other matterspertaining to the Report should be addressed to the Secre-tary, Ship Structure Committee.

This Report is being distributed to those individualsand agencies associated with and interested in the work ofthe Ship Structure Committee.

Yours sincerely,

—

K“.K. C~WART 1Rear Admiral, U. S. Coast GuardChairman, Ship Structure Committee

FIFTHProgress Report(Froject HL99]

The Fmdament,al Factors Influencingthe Behavior of Welded Structures:

The Effect of Subcritical Eeat Treatmenton the Transition Temperature of a Low Carbon Ship Plate Steel

-by

E. B. lW~KWD. J. (+aribotti

CASE INSTITUTE OF TECHNOLOGY

under

Department of the NazBureau of Sips HQBS-4 70

BuShips Project No. NS-011-078’

for

SHIP STRUCTUFU3COMMITTEE——

—. .- -.

r

J

u

o

0

0

c

0

0

0

0

0

.

0

.

Impact Tests ,

Iiicmstru.elmres

Discussion, . . . . ,

Conclusions . , . . .

Future work o . . 0 0

BibliOgi-aphy . . , .

‘TABLE

o

0

0

0

0

0

0

m

o

.

0

0

Cl

0

0

b

o

e

.

3

0

0

0

0

0

0

0

d

o

0

0

0

0

0

0

u

o

LIST OF FIGURES

20

3.

4-0

5-.

6.

7.

80

110

Effect of Various Aging Times and Temperatu~es ontileHardness of ‘UCqlSteel after liaterQuenchingfromljcO~F. . . . . . 0 . 0 . 0 . . 0 , . a . . ~

CkkarpyV-Notch Transition Curves for As-Recsived‘~C’iSteelo . . . . . . . . . . . .0 . , . . .0 9

C&LarpyV-Notch Transition Curves for ~’Cf~SteelSubcritically Eeatet!at 1300~F Tor 15Minutes and Air Cooled. . . . . . . . . . . . . . 9

C$harpyV-Notched Transition Curves for ‘“G’fSteelSubcritically Heated at 1300°F for 15]linu-tesand W-aterQuenched. Aged at RocqjTemperature for Times Indicated . . . . . . . . . 11

Charpy V-Notcl~ed.Transition Curves for ltCtiSteelSubcritically Heated at 1300°F for 15Minutes and Water Quenched. Aged at 125QFfor Times Inciica,ted. . . . . , . . . . . . . . . ,12

Charpy V-Notched Transition Curves for ‘~C~OSteelSubcritically E&ted. at 1300°F for 15Minutes tii~dWater Quenched. . . . . . . , . . . . 1.2

Charpy V-Notched Transition Curves for 9’C”SteelSubcritically Heated at l~OO°F for 15Flinutesand WatEr Quenched. Aged at 600QFfor Times Indicated . 0 . . . 0 . ~ . . , . . . . 13

Sumaiyj Curves Siilowingfiff~ctof Various &inSTemperatures antiTimes on the TransitionTe~p~rature and Eardness of ‘UC*USteel.Specimens Mater Quenched from 1300°F beforeAging.. . . . . . .0 . .0 . . . .0 . . .0 . 13

Microstructure of As-Received and Various&uench-Ages Conditions. . . , . . . . . , . . . . 16

Effect of Room Temperature Agin~ Time on theTransition Temperature and Hardness or ‘iCT’Steel af’terWater Quenching from TemperaturesIndicated . . . . . . . . , . . . . . . . 0 . . . 18

Solution and Room Temperature Aging Effects onthe Hardness and Impact Properties of ‘t(l’iSteel. . .0 . .0 . . . . . . . . . . .OO . . 20

—

ABSTRACT

An investigation was made to determine the impact transi-

“iim temperature and hardness changes att~ndant to the quench-

aging of Project Steel ~lC~ta semi-killed ship plate steel.

Aging temperatures;~xtendedover the range{:from35~ to llOO°F

after water quenching from 1300°F.

Both impact and hardness tests revealed that this steel

can be severely embrittled by the quench-aging ~ech~dsmi. with

aging temperatures ~p to.. -

Were obtained? i.e.a the

3500F9 characteristic aging curves

peak embrittlement and the time to

attain this peak degreased with increasi~~ ~ging

Far mom temperature aging this peak amounted to

in transition temperature and 25 points increase

temperature.

a 90°F increase

in Rockwell )5

hardness above that of a series air cooled from 1300~ (unembrit-

tied condition). Specimens aged above 350°F ‘overagedv so rap-

idly that no peak in the aging curve could be detected.

Metallographic

2000X showed that a

examination of quench-aged specinens at

two-stage precipitation reaction was

operative. At low aging temperatures the precipitate was

detected as a mottling of the ferrite grains; at higher aging

temperatures where an ‘overaged~ condition was rapidly reached$

the precipitate had grown so as to be resolvable.

It is believed that the quench-aging phenomenon is re-

sponsible for the ~$ittle zone previo-usl~m$oundin the sub-

critically heated region in weldrnentsof this and similar ship

iii

plate steels. This study suggest&that a law temperature post-

heat at 650*F (the solution temperature below which quench-

aging effects are absent) would lead to ra’pidoveraging in the

brittle zone of ship plate weldments and thus largely eliminate

the embrittlement.

IJSTRODUCTION

**weldments

temperature

investigate

of steel by

-2-

in a region not heated above the lower critical

(115 It was consider~d advisable toat any time

the maximum embrittlement possible in this grade

quench-aging and how to minimize or eliminate it.

TO this end, Charpy V-notch impact specimens wex=esolution heat

treated at the maximum subcritical temperature (1300°F)Y water

quenched7 and then aged for various periods of time in the ~~”

to llOO°F range. T& embrittlement was ey~luatei by transi=~

tim temperature and hardness changes7 supplemented by

microscopic examination

MATERIAL

The semi-killed? ship plate steel (C St~el) used in tkw

present work was the same 3/4-in. thick7 as-rolled plate

(Plate II) which supplied specimens for the earlier work on

subcritical heat treatment. The properties reported for this

(!5)steel are as fallows:

TABLE I

PROPERTIES OF C STEEL PLATE—-— —

Chemical AnalysisCarbon 0.24 copperManganese 0.&8 ChromiumPhosphorous 00012 MolybdenumSulfur 0.026 TinSilicon 0.05 NitrogenAluminum 0,016 VanadiumNickel 0002 Arsenic

Mechanical Properties

0.030.03(3.0050.0030.0090.020.01

**At least for weldmmts made or A and C steels$ the twoproject steels investigated.

>,,.

“3-

mommlus

blanks (C).&207tsquare) were cut from the as=

after quenehingo In following the progress of agingy specimens

were removed from the aging baths for a time only long enough

for the hardness tests to be made.

With the hardness checks as a guideq series of Charpy

V-notch impact blanks were taken from thE plate in the same

manner% quenched from 1~00%9 and

room temperature E’5°9 ~~Oo9 and

series was air cooled from 13000F

col-lditiono**

aged for various times at

600°F. In addition$ one

to give the unembrittled

*The composition of the neutral chloride salt was asfollowss13MI12 55$ mm 22$KCI 22$ l?Di@$ l?

**TIM Transition Temperature of this se;ies was approximatelythe same as that of the as-received material. In addition~hardness checks of broken specimens showed NO change withtime at room temperature.

-4--

After aging, the specimens were immediately grQund to

size and the notch cut perpendicular to the surface of the

plate9 with each series being machined and tested within

seven hours after aging. The impact testing procedure was

the same as given previously(Il. JLS pointed out in the

earlier work, the possibility exists that accelerat~d aging

can accur in the testing bath. In establishing the transi-

tion curves, it was necessary to test as high as 350°F

(specimens were held 10 minutes in the bath to assure tempera-

ture uniformity)? and thus7 accelerated aging in the test

bath may occur in those specimens prmioudy aged at room

temperature and 125°F. Hardness checks made on specimens

from these series before and after breaking indicated that the

hardness was unaffected up to testing temperatures of 17.5°F0

In view of the fact that the 15 ft~lb transition temperatur~

is to be used as the criterion of’em’brittlementand in no case

did this exceed 175°F5 the results WOU~U appear to be uIIaffecte~o

RESULTS

Hardness Tests

The Rockwell B harnesses obtained after aging in the 35°

to llOO°F range are shown as a function of the aging time in

Fig. 10

From an as-quenched hardness

specimen aged at roon temperature

for about one hour then gradually

of RB877 the hardness of the

(800Fl remained unchanged

increased7 reaching a maximum

:m: 85

c?o= 75

70

65

x

~——— 900 “F— Iioo”r- ALL SPECIMENS AIR COOLED

T\ — FROM AGING TEMPERATURE —

I ‘AS- RECEIVED

‘~ AIR COOLED

FROM 1300° F

3’ 0.1 I 10. 100 1000 10,000LOG AGING TIME-HOURS

FIG. I : EFFECT OF VARIOUS AGING TIMES AND TEMPERATURES ON THE

HARDNESS OF “C” STEEL AFTER WATER QUENCHING FROM 1300° F.

level of RB96 after five days. Aging at a lower temperatur~

(35QF) resulted in a slower rate of increase and a longer

tiu.e(about 50 days) to reach a slightly higher peak level

%9of “ 70

With increasing aging temperatures above room temperature

to 250°F$ the hardness increased at an ever increasing rate;

but the peak hardness reached and the time to reach this peak

decreased. Once the peak had been attained,

time caused a decrease in hardness at a rate

with aging temperature.

further aging

which Increased

At 350~F and higher the hardness dtd not rise above that

of the as-quenched specimen but fell off at a rate which again

increased with temperature.

A summary of the maximum hardness reached and the time

to reach this maximum for the various aging temperatures

employed are given in Table 11.

EFFECT GF AGING TEMPERATUXL ON TliEPEAKHARDNESS REACHED AND THE TIME TCIATTAIN

THI~ PEAK

$gingemnerature. oF

Peak HardnessReacheda RR

Time to ReachPeak Hardness

23 97 50 days96 5 days

125 95 23 hours200 91 2 hours250 88 15 minutes350 Not higher than as-quenched hardness of 11~87600 !1

900 #l

1100 11

Note: All specimens water quenched.aging. Hardness of specimenwas RE 70..

from 13000F prior toair cooled from 1300°F

“7-

These resultsshow that by aging the hardness can be raised

a maximxm Qf about 10 points above that of the as-quenched hard-

ness @#). In turn the as-quenched hardness was 17 points

Rockwell B above that of the air cooled hardness (RB70). The

maximum cumulative hardness increase due to solution and aging

then amounts to about 27 points RB.

h~act Tests

The impact transition temperatures by three criteria

obtained of the (1) as-received*g (2) air ccded~ and (3] the

various quench-aged conditions are swarized in Table 111.

The individual transition curves for each condition are plotted

in Figs. 2--7 with both the energy absorbed and the per cent

fibrous fracture plotted as the ordinate.

In the following sections of the report? the effects of

the various subcritical heat treatments are evaluated with the

15 ft-lb transition temperature as the criterion of enbrittle-

ment and with the properties of the air cooled series reflecting

the unembrittled state. The choice of either of the other two

criteria listed would revea,lthe same general effects.

In comparing the impact properties of the as-received

plate$ Fig. 2$ with those of the air cooled series, Fig. 3?

it is evident that there is little difference in properties

other than about a five ft-lb higher upper level for the air

cooled series.

The transition curves after aging for various times at

*Previous~y reported(l)

-ti -

?7

‘wCwLFm Y-Mm C%

.-_-;

m

.

,,

FIBROUS FRACTURE XPERCENT

n —FIBROUS FRACTURE XPERCENT

I

N. .

o

zc)+

+

-+ 07+mITr0H--

0

0

EE-P000

-P

I-vo -P

o00

IMPACT ENZRGY-FT.-LBS, IMPAGT E114ERGY~FT. -L8S.

___

Io l---~amam I I I I I I I 50

20

L

AGED 7 HOU

40 JJ

+—AL +-++1 ‘o

ii ‘Od-+l+4i+o.TO+-i~o!

30

I!i=ii‘-’J+k-

,0-●4* I ‘ I\; I I i3

& 100m. o 0

L(C)

‘wI* 0*0

d

ji~-~ 50 =20AGED 5 DAYS ‘~ —m 0\. : &-** I i 40

40

.,&”i I ! I 3060

-EIO o 00

TEST

FIG. 4.( A-E ) : CHARPY

FOR *’C” STEEL

1300” F FOR 15

AGED AT ROOM

INDICATED.

160 240 320 400

TEMPERATURE = ‘F

V-NOTCHED TRANSITION CURVES

SLIBCRITiCALLY HEATED AT

MINUTES AND WATER QUENCHED

TEMPERATURE FOR TIMES

‘m-”r”-!-ohl. ‘ ‘ ‘ ‘ I 150

t

r I .- 1 I!- 20 AGED 32 DAYS Iz

\i

KEIH3EPk-

1 “%U-.

●=p*40“o 1 #*1

30;,

20

mmm++’20

-eo o 80 !60 240 360 400

TEST TEhlPERATUREN*F

FIG. 4: 130 NT[MUED ,

Lt-’I

–80 o 80 160 240 320 400

TEST TEMPERATURE x“F

or ‘‘ ‘-y

0+ 50

(A]

20 AGEO 15 MIN. $Q ---- ---- ~-.iJO,.&!

I40 -– _“;- :_________ ;;

60 –– ?L:- q$:i

%$i{ ---/--,0

11 .,1

‘?

TEST TEMPERATURE +-F

FIG. 6: (A-C) CHARPY V-NOTCHED TRANSITION CIJRVES

FOR “C” STEEL SUBCRITICALLY HEATED AT 1300” F

FOR 15 MINUTES AND WATER QUENCHED, AGEDAT 350” F FOR TIMES INDICATED,

FIG. 5: (A-D) CHARPY V-NOTCHED TRANSITION CURVES

FOR “C” 5TEEL SUBCRITICALLY HEATED AT 1300°F

FOR 15 MINUTES AND WATER QUENCHED. AGED

AT 125°F FOR TIMES INDICATED ,

d3-

--..—.. .,

ILo

<’L –- -------J—---—-T--——--—

11

O-J-0co

I

“-iEo—1-

wILILI&l

(n

1

mu

1

0m—

ss3r4aw4

Ln IL

ou

oIn

om

.#,.

_>

---–~ -—I

.34.

The transition temperature--aging time relationship for

125°F aging--shows that the transition temperature starts to

increase at much shorter times? reaching a peak (168%) atter

about 20 hours and renaining at this value at least up to

three days aging time.

For 350°1?aging it appears that a slight peak may be

present at an aging time less than 15 minutesq but it probably

would not be much greater than the 115°F transition temP~ra-

ture obtained after 1~ minutes aging because the hardness was

unchanged during this time interval. Aging beyond 15 minutes

results in a deerwasing embrittlement with tiime.

Aging at 600°F for two and ten minufiesare~pectivelyz

effects a considerable improvement in t“heimpact properties.

For both cases the transition temperature indicated an embrit-

tlement of but 10°F.

From the general trend of These aging curves? it is to

be expected that at aging temperatures higher than 6000F the

impact properties would approach those of the unembrittled

state very rapidly. However~ as will be pointed out in the

Discussion? this conclusion only holds for specimens cooled

relatively slowly from the aging temperature; a fast cool from

aging temperatures above about 650° can introduce another

quench-aging cycle.

Microstructure~

In order to detect any microstructural changes due to

quench-aging, a number os specimens representing various quench-

aged conditions were examined at 2000X.

‘Thestructure of the as-received condition is shown in Fig.

9(a). No difference in structure was noted in the air cooled

condition; howeverl immediately after quenching~ a mottled

ferrite was evident which did not appear to change with time

at room temperature. Fig. 9(b) shows the structure after one

month at room temperature. No change in this mottled structure

was revealed after aging at 125°F$ even after 23 days at tkcls

temperature, Fig. 9(c).

After aging 15 minutes at 350°Fq the mottled ferrite was

still in evidence; after 21 hours at 3500FY Fig. 9(d), the

mottled structure appeared to be more intense.

Upon aging at 600°F for two minutes, a precipitate, evenly

distributed throughout the mottled ferrite grains, was

resolvable. An increase in aging time to ten minutes, Fig. 9(e)~

resulted in better definition QT the precipitated particles

which appeared to be platelets and an increase in their size.

After aging ten minutes at llOO°F~ Fig. 9(f), the precip-

itate is no longer evenly distributed throughout the ferrite

grains but appears to have coalesced along the grain boundaries

as spheroids.

DISCUSSION

On the evidence of the hardness properties~ Fig. 19 aging

curves were obtained which were characteristic of quench-aging

-.—

-,.

-. >--,

(b) Aged30Daysat Fkmm (c) Aged2? Daysat 125*F.Temperature

L

(d) Aged21 Hoursat 3S0°F (e) Aged10

u Fig. ~: MICROSTRUCTH OF AS-RECEIVEDANDVARIOUSby a waterquenchfrom1300°F. 2000X. 3%

Minutesat 600*F (f)

QUENCH-AGEDCONDITIONS. AllNITAL●

Aged10 Mimutes atllOO°F

agingtreatmentspreceded

systems$ i.eo~ the maximum hardness attained a,ndthe time to

reach this maximum decreased with increasing aging temperature

The similar changes observed with the impact transition

temperature? Fig. 8? are believed to be new experimental data.

The only information found on the impact transition temperature

changes due to quench-aging in a similar grade of steel was

‘6] and was restricted to the room tempera-in the work by Low

ture aging effects. In this paper it was shown that after

quenching from 700°C <1290°F] the transition temperature in-

creased continually over the aging interval of three years.

The total increase amounted to a 110~ rise in Charpy keytiole

transition temperature (10 ft-lb value)$ with the greatest in-

crease occurring in the first ten days aging. This work was

done with a l/2-inch hot rolled semi-killed plate containing

0.17ficarbon.

In conjunction with the earlier work under this project,

it is now possible to show, Fig. 107 the effect of solution

temperature on the magnitude of the peak transition tempera-

ture and hardness reached by room temperature aging. As is

evident, increasing the solution temperature in the 1100° to

13000F range results in a greater initial (fas-quenchedt]

hardness and, apparently a greater initial transition tempera-

ture. Upon subsequent room temperature aging, the magnitude

of the peak also increases with increasing solution temperature.

This is to be expected in line with other quench-aging systems

—!

200

190

180

I70

160

150

140

I30

120

110

IOc

90

80

i’-

H

9-

-

-L8-

GHARPY

+rLL-

‘--- AS RECEIVED

100

90

0

‘ “--”””-l”-”----fi--+----”

t

80 --------------- —. ..

75x~.-– AS RECEIVED

I95 - 1300°F--

>i

‘o~o, o0-

1

❑ u~

1200” F’ –--””-u-”--–-—-

r’

.D-u ❑ .-”~-

m--<

70

k“-- ‘“

—. —————

65

600 “v 0.I I

....____. –-\---- _>.. , ,~o OF

-—— -——---— --- --- —..-

10 100 1000 10,000

LOG AGING TIME “-’ HOURS

FIG. 10: EFFECT OF ROOM TEMPERATURE AGING TIME

ON THE TRANSITION TEMPERATURE AND HARDNESS

OF “C” STEEL AFTER WATER QUENCHING FROM

TEMPERATURES INDICATED.

—

which show that a greater change in properties can be expected

with i.nereasingsolution temperature (greater degree of super-

saturation].

To obtain a quantitative measure of the solid solution

and the aging effects as a function of solution temperature

from room temperature to l~OOoF, Fig. 11 has been prepared.

Here for each property three heat treated conditions are under

mnsideration: (1) air cooled~ (2) as-quenched, and (3) quenched

and aged one month at room temperature. The air cooled data

sets the base line for evaluating the two effects because in this

condition it is assumed that the solid solution and aging effects

are absent*. The difference between thetas-quenched~ proper’~ies

and the base line then yields the effect due to solid solution~

whereas the difference between the ‘as-q-uenched9and ~quench-

agedi properties gives the effect due to aging. The aging ef--

fect may be altered for aging times greater than one vonth~

but it is believed that the change would be slight? if any~ and

therefore the effect shown may be considered a maximum.

This figure also shows that below a solutian temperature

of about 650°F the solid solution and aging effects are nil but

that above this temperature both effects increase with increasing

solution temperature. It is interesting to note that the hard-

ness is affected to a greater extent than the transition tempera-

ture by the solid solution effect, whereas the reverse is true

*Hardness checks showed no change with time at roomtemperature.

-

100

95

90

85

80

75

70

65

190

180

170

160

150

140

130

I20

110

100

90

80

AS RELEIVED

L_+_o—

——

——— .—

—o—

HARD

I x AIR COOLED I

<:( tAs QUENCHED (WATER)

QUENCHED AND AGED AT

ROOM TEMPERATURE

--IFI I

MONTH.

1

AS RECEIVED

/

I I I I I

o 200

——

400 600

:CT [

4.

7

/—

a

I I I I I

800 1000 1200

SOLUTION TEMPERATURE x “F

1400

FIG. 11: SOLUTION AND ROOM TEMPERATURE AGING

EFFECTS ON THE HARDNESS AND IMPACT

PROPERTIES OF “C” STEEL.

--

-21-

with aging. In any case the two effects are interrelated and

governed by the solution temperature which in turn controls

the amount of carbon available for the precipitation reaction.

With cooling rates intermediate to an air cool and water

quench? it is to be expected that these two effects will be

minimized due to the lower degree of supersaturatian~ i0@05 less

carbon is retained in solution on the quench.

From the rnetallographicchanges occurring during aging$

it would appear that a two-stage precipitation reaction is

operative. The first stage (evident as a mottling of the fer-

rite) was responsible for the greater degree of embrittlement~

while the second stage, in which a resolvable precipitate was

detected$ was associated with an improvement in properties

corresponding to a rapid ‘overaged~conditiong Although 110

attempt was made at identification? the sequence of micro-

structural changes observed appear to be the same as reported

recently (7)0 In this paper the metallographic changes occurring

during the quench-aging of an ingot iron {0.026~ carbon} were

followed by means of the optical and electron microscopes and

the identification of the two-stage precipitate made with

electron diffraction technique. The first stage was identified

as the hexagonal &-iron carbide after 15 hours at 200°C (390°F).

The second

lattice to

30 minutes

stage was associated with a change in the crystal

cementite~ (orthorhombicFe3C]9 observed after aging

at 300°C (570°F).

=22-

It can now be speculated that in the aging of C steel the

formation of the transition lattice (E-iron carbide) is associ-

ated with a high degree of ernbrittlement,while subsequent

formation of the equilibrium phase

a large improvement in properties

~overaged~ condition.

(cementite) brings about

corresponding to the rapid

Insofar as microstructural changes are concerueda the

brittle zone previously found in the subcritically heated

region in ship plate weldments has not been identified with a

precipitation reaction*O Howeve~~ due ho the cc]mplexityof

the temperature time at ternperatureyand cooling rate condi-

tions in multiple pass weldmentst it is difficult to determine

the exact quench-aging cycle experienced. It has been shown

previously (1] that the cooling rate in the subcritlcally heated

region** is such that the ferrite is not supersaturated to the

maximum possible as in water quenehed test specimns. Conse-

quently? less carbon is available fer the subsequent precipita-

tion reactions and the resultant precipitate may not be detected***.

In regard to the changes in hardness and transition tem,pera-

ture~ the results of this investigation givs the maximm possible——

*It shoL1ldbe recalled that the de~onstrated beneficial ef.fects of increasing preheat (decreasing ccoling rate) andpostheating (Qoveragingf certainly point to the quench-aging mechanism.

**The cooling rate at the critical zone was found to beintermediate to water quenched and air cocled impact testspecimens.

***It is not expect~d that the electron microscope would beable to detect any precipitation. In the reference citedin the Discussion (7)9 the only advantage of the electronover the light microscope was inprecipitate once it had formed.

better-resolution of the

-23-

embrittlement

ment designed

be applicable

of

to

to

C steel by quench-aging, Fi&. 11. Any treat-

remove these effects in the base plate should

weldments made of this and.similar grades of

steel. As was shown in Figs. 1 and 8B ‘overagingfitreatments

served to minimize the embrittlement, with the degree of em-

brittlement decreasing with increasing aging temperature

However, this finding must be qualified.in view of the fact

that the results were based on specimens aircooled from th~

aging tenperaturesa A consideration of Fig. 11 shaws that

if aging temperatures above 6500F were used, then the pos-

sibility exists that another quench-aging cycle could be

initiated

different

fore, the

should be

if a fast cool were employed. (Lowf6J~ using a

approacha arrived at the same condusion]o There.-.

postheating of weldments susceptible to quench-aging

restricted to a maximum of about 650°F when this

danger exists. As this investigation showed? aging in this

neighborhood (6000F) effected a rapid and considerable improve-

ment in the ductility by ‘overaging~.

CONCLUSIONS

In conjunction with previous work(11 on the subcritical

heat treatment of a low carbon ship plate steel (C steel)q the

following conclusions seem justified%

1. The quench-aging mechanisn was responsible for the loss inductility and-the increase in hardness.

-.

2.

3.

b o

5.

pa>g severity Of the embrittlement imer~a~ed “~’~ith~n~reasingsolution temperature from 6J00 to 1300 F; m embrittlementwas present when employing solution ter,peratvresless than650°F’O

On the evidence of impact transition temperature and hardn-ess changes? characteristic aging curves were obtained$i.e.$ the peak transition temperature and hardness reachedand the time to attain this peak decreased with increasingaging temperature.

Metallographic examination showed that a two-stage precipita-tion reaction was operative.

A low temperature postheat at about 650°F would do rrILIch “k

eliminate by overaging the zone of minimum ductilitypreviously fm.md in.the subcriticallyheated region of weld-ments Of this and similar grades of steel. ;Si,gherpostheattemperatures rufithe danger of introduci+~ another agingcycle if a fast cool is employed.

FUTURE WORK

This report concludes the experimental work done under this

contract. A final report will be prepared summarizing all the

work done on (1] the distributim of relative ductility in ship

plate weldments and (2) the subcritical beat treatment of base

‘plate o

.—

-25-

BIBLIOGRAPHY

1. E. B. Evans$ and L. J. JQinglerj ~iTheEffect of fibcritica~Heat Treatment on the Transition Temperature of a Low CarbonShip Plate Steel’~a Fourth Progress Report, Ship StructureCommittee Serial No. SSC-603 October 30, 1953.

2. G. Sachs5 L. J. Eberta and A. W. Da~7 Jrg? ‘tTheFundarLenta~Factors Influencing the Behavior of Welded Structures UnderConditions of Multiaxial Stress and Variations of Tempera-ture, Stress Concentration~ and Rates of Strain”9 NavyDepartmenta Bureau of Ships9 Contract NObs-45k70~ SerialNoe SSC-2k, May 10~ 1949*

30 Lo J, KLinglera L. J. Ebert and W. M. Baldwin, Jr.~ Ibid.$2, Serial NO. SSC-3k$ November 28? 1(3490

~. E. B. Evans and.L. J. Klingler~ Ibid 27 Serial No. SSC-~3October 143 19520

~. Technical Frogress Report of the ShiP St~u~ture Cammi~t~e~Welding Jouma15 Vol. 13 (July 1948)9 pPo 37?s-38~s0

6. Jo R. LOW, Jroa ‘tTheEffect of Quenc’h-Agingon the NotchSensitivity of SteelU’qWelding Research Supplement$ Vol.17 (my 1952)9 PO 253s-256s0

7. Anna L. T’soutJ. Nutting, and J. “W.Menterl ‘tThe@ench-Aging of Iron:f~Journal of the Iron and.Steel Institute$ Vol.1722 part 2, october, 1952$ pp. I63-I71O