Embed Size (px)
Citation preview
11
--’,
SERIAL NO. SSC-60
Fourth
PROGRESSREPORT‘)
(Proiect SR-99)
on
THE FUNDAMENTAL FACTORS INFLUENCINGTHE BEHAVIOR OF WELDED STRUCTURES:
The Effect of Subcritical Heat Treatment
on the Transition’ Temperature of a Low Carbon Ship Plate Steel
!,
.’
an-d
Supplement on
Ernbrittlemen+ of “C” Steel by ,Nitrogen
)’
by!,
E. B. Evans and L: J. KlinglerICASE INSTITUTE OF TECHNOLOGY
1
TransmifPed Through
NATlONAL RESEARCH COUNCIL’S, , 1
COMMITTE5 ON SHIP “STEEL
Advisory to
SHIP STRUCTURECOMMITTEE
Division of Engineering and Industrial Research
National Academy of Sciences National Resmrch Council
Washington, D, C.
October 30, 1953
..-——-.-.,:..__..... ,.,. . ...... .. -.... ——-—
SHIP STRUCTURE COMMITTEE
MEMBER AGENCIES: ADDRESS CORRESPONDENCE TO:
BUREAU OF SHIPS. DEPT. OF NAVY SECRETARY
MILITARY SEA TRANSPORTATION SERVICE, DEPT. OF NAVY SHIP STRUCTURE COmmittee
UN ITEU STATES COAST GUARD. TREASURY DEPT. U. S. COAST GUARD HEADQUARTERS
MARITIME ADMINISTRATION, DEPT, OF CO MMCRCE WASHINGTON 2s, D. c.
AMERICAN BUREAU OF SHIPPINGOctober 30, 1953
Dear Sir:
As part of its research program related to theimprovement of hull structures of ships, the Ship Struc-ture Committee is sponsoring an investigation on “TheFundamental Factors Influencing the Behavio~ of WeldedStructures under Conditions of Multiaxial Stress andVariations of Temperature” at the Case Institute of Tech-nology ● Herewith is a copy of the Fourth Progress Report,SSC-60, of the investigation, entitled “The .FundamentalFactors Influencing the Behavior of’Welded Structures:The Effect of Subcritical Heat Treatment on the TransitionTemperature of a Low Carbm Ship Plate Steel~’by E, B.Evans and L. J. Klingler.
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 t6 the Report should-be addressed to the Secret-ary, Ship Structure Committee.
This Report is being distributed toand agencies associated with and interestedthe Ship Structure Committee.
those individualsin the work of
Yours sticerely,
K. K. COWARTRear Admiral, U. S. Coast GuardChairman, Snip Structure Committee
..
—.
on
The Fundamen,taJ Factors Influencingthe Eehaviur of Welded Structures:
The hffect of Subcritical E-eatTreatr(lenton the Transition Te~Iper&ture of a Low Carbon W-,ipFiate Steel
andSupplement cn
Ewbrittlement of Wlr ite~l by fiTitrogen
by
E, B. LvansL. J. Klin~ier
CASE INSTITUTE OF TECHNOLOGY
Imder
~e~J&Lr”hI!&Iltd ~h@ ~av1llureau C@ Ships I!C~~s-~E570EuShi~;s Froject No. HS-C)ll-G78
fox’
SHIP STRUCTURE COMMITTEE
‘1’A.bu!iPage
iContents . . . .
List of Figures
Abstract 0 . . 0
Introduction . .
Material . . . 0
Procedure . . .
.● O*
.30.2
● *O
● .*
*OO
aoo
-d
on
● *
**
at
*e
O*
●
●
●
●
0
0
.
44a
●
●
●
☛
●
●
4
●
0
●
e ● 0 e .*
*
.
0
9
●
o
.
.
●
*
*
0
.
0
9
0
●
‘ii● . ● ● .●
vi● ● ● m 0 e
14 ● 0.
● ● * ● e 0 0
0* ● ● ● a
4Specimen Preparation .Subcritical Heat TreatmentCooling Curves n . 0 ~
8●
✎
●
D*o
0
●
●
✎
0
●
●
✎
4● O
O*
and
Testing Procedure . . .
Resuits
Transition Temperatures 11Harnesses ● . . w
17Base Plate . .Air Cooled . .Furnace CooledWater Quenched
Microstructure . .
Cooling Curves . .
Discussion . . 0 . . . .
Conclusions . . . . . .
Future Work . . . . . .
Acknowledgments . . . .
Bibliography . . . . . .
Appendix
A. Eccentric Notch
0a
●
●
●
●
●
●
a
4
e
●
Q
0
●
0●
●
●
●
o
n
.
●
e
●
●
●
✎
e
o
.
a
*
●
●
0
●
●
●
●
●
0
*
.
0
●
*
4*o
●
9
m
0
0
●
.
0
m
.
0●
0
a
a
9
0
0
a
19232b
34
36
38
45
46
46
k7
0 0 0 0
Q0 0
04 0
e 0 * ●
0 e
9 a
● e . ●
Tensile Transition CurvesMethod of Determination of Eccentric NotchTensile Transition Temperature . . 0 . s ● 8
B. Charpy V-Notch Transition Curves Plate II . .
Supplement
Embrittlement of ‘tC1~Steel by Nitrogen . . . . .
4
67
&
1°0
2*
2>*
4*
.5*
6.
7*
8.
9.
10 ●
11 ●
LIST OF FIGURES
Title
T~st Specimens . . . . . . . . . . . . . . . .
Preparation of Gharpy V-Notch and Notch TensileSpeciuens fi”C)llr!cilSteel Plate. . . . .Y@
Method of Loading to Obtain l/4-IncP. Eccentric-ity. (Eccentricity and the Position ofFixtures are Exaggerated.). . . . . . . .
Comparison of Eccentric Notck Tensiie Transi-tion Behavior for Two Different Plates ofAs-Received “C” Steel . . . . . . . . . .
liceentric Notch Tensile Transition Tempera-tures of lICI1Steel as a Function of Timeat Various tiubcritical Temperatures.AirCooieda . . . . . . . . . . . . . 0 .
Charpy V-lJotch Transition Temperature as aFunction of Time at Subcritical T~mpera-tures. Air Cooled. Plate II. . . . . .
Transiti~n Tempcra~ures of ‘rC’!Steel,as aFunction of Time at Subcritical Terrrpera-tllres. Water Querlched and Aged OneMonth at Room Temperature. Plate 11. . .
Elfect of Room Temperature AginS Time onChiarpyV-NotcliTransition Curves of ltC’iSteel. . . . . . . . . . .0 . . . . . .
Effect of Room Temperature A~in~ Time on theChdrpy V-Notch Transition Temperatureand Ha3x5ness of “G’~Steel . . . . . . . .
Effect of Room Temp~ra.ture Aging Time onCharpy V-l;otch,Transition Curves of “C’rSteel. . , . . . . . . . . . . . . . . .
Comparison offor RoomASin~ of
Charpy V-Notch Transition CurvesTemperature and Acceleratedrlcl~Steel. . . . . . . . . . 0 .
PaEe
5
?
9
18
20
22
26
29
29
32
32
ii
.NIicrestmcture,sof ‘?CiBSteel in the As-ReceivedCondition and aTter Varicus SubcriticalEeatTreatments . . . . . . . . . . 0 0 . .
Comparison of Cooling Curves in the Region GfLowest Ductility far Two Welding Comlitimswith those Gbtained w~.thHeat Treated TestSpecim.~nso . . . . . . . . . . . . . . .0
Heating and Cooling Curves in the Region ofLowest Ductility for TWCI6 Pass Weldnents .
Eccentric Notch Tensile Transition Curve ofAs-Received ~FC1;St@el. Plate 110 . . . . .
Eccentric Not& Tensile Transition Curves ofIIc!l~~-e~o Subcritically Eeated at 700%Ffor Tires kiieated and Air Cooled. . . . .
Eccentric Notch Tensile Transition Curves of “C1!Steel. Subcritically Heated at 8CC°F forTi~es Indicated and Air Cooled. . . . . . .
Eccentric fiTotchTensile Transition Curves for‘fcff Steel. Subcritically Heated at ~~O°Ffor Times Indicated and Air Cooled. . . . .
Eccentric Notch Tensile Transition Curves for‘~C’9Steel. Subcritically Heated at lloo°Ffor Times Indicated and Air Cooled.Pla.teII . . . . . . . . . . . . . . . . .
Eccentric Notch Tensile CurUvesof ‘tC~~Steel.Subcritically Heated at 1200°F for TimesIndicated and Ai~ Cooled. Plate 11 . . . .
Eccentric Notch Tensile Transition Curves forl’CtnSteel. Subcritically Heated at ll.OOQFfor Times Indicated and Water Queneh~d.Aged 1 Month at Room Ter~perature.pktell . 0 0 0 . 0 . 00 0 . . . 0 . .0
Eccentric Notch Tensile Transition Curves of‘~CtuStee,lo Submitically heated at 1200°Ffor Times Indicated and Water ~ucnched.k~sd. 1 Month at Room Temperature.FlateII . . . . 0 0 0 . 0 . , . . . 0 0 0
m
— —
Pap,@
l.Bo
2B.
3B.
‘+Be
7B0
Charpy V-liotch Transition Curves for “~+’Steel. Subcritically He:it,edat llOO°Ffor Times Indicated and.Air Cooled. . . . 60
Ckrpj’ lT-NGtckiTransition Curves for ‘fC~fstEelo Subcritically Heated at 12CJOoFfoi-Times Indicatd and Air Ccoled . . . ~~
Charpy V-NOtCli Transition Curves for “C”steel. StibcriticallyHeated at 1200°Ffor l\2 Minute and Furnace Cooled. . . . 52
Charpy V-Notch Transition Curv~s fcm ‘lC~’Steel. Subcriti.tallyHeated at 12000Ffor ‘li~nesIndicatd and,Water Quenched.Aged at RoorI~Temperature.. . . . . . . . 64
SUPPLEMENT
1. Charpy V-Nctch Transition curves of ‘~Ct~St@elAfter Various Subcritical Feat Treatmentsinl~itrate Salt . . . . . . . . . . . . 69
2. Eccentric Notch Tensile Transition Curves of~’C?~Steel After Various SidbcriticalHeatT~%eat]ilel~t~h-lNitrate Salt . . . . ~ . - 71
3. Transition Temperatures of ‘IC’?Steel Resultingfrom Various Subcritical Heat Treatmentsin Nitrate Salt Bath.and in Airt EKJploy-inganAirCool . . . . . . . . s . . . 73
iv
No o Title
v
ABSTRACT
The dependence of transition temperature upon subcritical
heat treatment has been investigated in a low carbon ship plate
steel (Project Steel ‘~C1l).The effect of time at temperature
in the 700°--1200°F range has been determined, employing three
different cooling rates--air cool, furnace cool, water quench.
In additionP a limited study was made of the room temperature
and the accelerated aging effects after water quenching from
1200°Fo The degree of embrittlement was evaluated by means of
eccentric notch tensile and Charpy V-notch impact tests, with
the as-received plate as a basis of comparison.
For the air cooled seriesj the transition temperature-
isothermal time relationship obtained with impact specimens
indicated no embrittlement at the shorter times, but a slight
embrittlement was present after long times at either 1100° or
1200°F. In contrast, notch tensile specimens revealed a
slight, constant embrittlement at the s’hortertimes which was
decreased or eliminated at the longer times in the 700°--12000F
range.
Witlfia furnace Cools spot checks made with impact speci-
mens showed the same transition behavior as specimens heat
treated and air cooled.
For tineseries quenched from llOO°F {aged one month at
room temperature)9 the plot of transition temperature vs.
isot?oermal time indicated a severe embrittlement at all iso-
thermal times~ i.e.$ the entire curve was displaced consider-
ably abbve And approximately parallel to the curve for the
air cooled series for each specimen type. The same findings
were
merit
evident after quenching from 1200°F3 with the embrittle-
being more pronounced.
Impact specimens water quenched from 1200QF and aged for
various periods or time up to two months at room temperature
~’as-quenched~+$however~showed no embrittlement the transition
temperature increased with aging time$ reaching a maximum
level after about two weeks. Accelerated aging at ~OO”F for
one hour immediately after quenching from 1200°F resulted in
a marked improvement in impact properties, approaching that
of the as-received plate.
No change in microstructure could be noted between the
subcritically treated specimens and the as-received plate with
these exceptions: (1) a general precipitation was evident
after accelerated aging, and (2) slight spheroidization was
apparent at the long isothermal times. Rockwell B hardness
tests showed that, in general, appreciable hardening occurred
when specimens were embrittled.
Previous work on welded plate at this laboratory showed
the necessary conditions for quench-aging are present inthat
the welded material and this phenomenon appears to be the
only possible explanation for the zone of minimum ductility
vii
located in a region adjacent to the weld which was not heated
above the lower critical temperature. Cooling rate curves
are presented for the critical zone in weldments made with
various preheat temperatures to show that not only does the
embrittlement increase with increasing cooling rates but
that the degree of embrittlement is about the same for the
critical zone in weldments as for subcritically heat treated
base plate cooled at the same rate from the same temperature.
INTRODUCTION
This report summarizes the work completed on a project
sponsored by t’heShip Structure Committee under Department
of the Navy, Bureau of Ships contract NObs-4~470 and under
tileguidance of the Committee on Ship Steel of the National
Academy of Sciences-National ~esearch Council and covers
the period from January 1, 19~OP to September 1$ 19~2. Tmee
Technical Progress Reports on commercial ship plate weldments,
SSC-24(1), SSC-34~2)~ and SSC-g4.(3) covered the progress of
the investigation from July 1, 19)+7, tO January 1, 19s0.
In tineprevious work on the exploration of the relative
ductility across weldments of A and C ship plate steels+, it
was found that a zone of minimum ductility occurred just out-
side the so-called heat affected zone in weldments made with
IOO”F preheat$
metallographic
was evident at
in a region which appeared to have the same
structure as the base plate. This behavior
both the rnidthickness and surface levels of
3fh--inch plate. The ductility was evaluated by means of the
eccentric notch tensile test, which poasesse:l the advantage
of measuring the properties of a very small volume of metal
from any position in the weldment.
++The designations A and C refer to steels “A’tand ‘~Crtin the series of Ship Structure Committee ‘tProject’?Steels.
- z-
Additional tests revealed that a higher welding prsheat
(@O°F) minimized the embrittlement in the critical zone,
while a llOO°F postheat almost eliminated it. The magnitude
of the improve-meritis shown, Table 1$ in the cor~parison of
the transition temperatures.
the critical zone after the
of the unaffected base plate and
various welding conditions.
TABLE I
Transition Temperatures% of Unaffected Base Plateand Zone of Minimum Ductility in A and C Steel WeILments
Location and Welding Conditions Trar~sition Temperature~O~
Unaffected Base Plate(2n or more from weld centerline)
Steel C Steel A
zone(0.3
100°l?pre~eat afidinte;pass iempe~ature -~y;+: -80koo ‘F ~;g
llOOQF postheat with 100DF preheat.
of Minimum Ductilityinch from weld centerline)%++
-40100QF pre:eat aid inte~pass temperature -20400”F it “4.5
11009F postheat with 100QF preheat -70
l~Atthe midthiclmess of 3/4~inch plate.+%At the surface level the transition temperature was -600F.
+-++%-Atthe sur~ace level the zone of minimum ductility wasshifted to 0.4 inch from lineweld centerline due to the~eometry of the double-V weld used~0
Temperature measurements made during welding showed that
the ~one of minimum ductility was not heated above the lower
critical temperature; consequently~ the embrittlement (and the
improvement brought about by preheat and postheat treatments)
.
-3-
.
..
:.
occurred at subcritical temperatures suggesting a quench=aging
mechanism resulting Trom the solution and precipitation of
carbides ~rom the alpha phaseo
No weld ~ailures have beefifound to originate in this
critical zo’ne. It was felt however, that further work was
desirable on the embrittling characteristics of ship plate
steel under more closely controlled subcritical conditions
than exist in meldmen-ts.. Specifically, the program consisted
of suhjectin.gbase plate of llC’lsteel to temperatures in the
700°--HOOoF range for various times, employing three dif-
Terent cooling rates--air COOIY furnace CO019 and water quench.
In addition, an aging study was carried out a.l?terwater q-uench-
ing from 1200*F. The ern.brittlementwas evaluated by means of
eccentric notch tensile and Charpy V-notch impact transition
temperatures supplemented by hardness tests and microscopic
examinati OKLO
The present investigation was complicated at the start
by embrittlement of test specimens with nitrogen, introduced
by a scaling reaction in the nitrate sal-tbath used as the
heating medium. Although interesting these results are not
pertinent to the present investigation and are presented and
discussed in.a Supplement to this report.
MATERIAL
The llC1lsteel selected Tor the present investigation was
11Poject steel“~hesale P ‘rwhich had been used i:nthe earlier
work at this laboratory and in other Ship Structure Committee
investigations. It was a semi-killed steel in the form of
3/4-inch material ~rom two large plates from the same heat~
The program started with material remaining from the weldment
studies~ hereafter referred to as Plate I; however, the
majority of tinework was done with Plate 11. The plates
have been identified becausej as will be shown later, a
significant difference in transition temperature was found
between plates. The properties reported for this steel are
TABLE II
Properties of ~~flllSteel Plate
Chemical Analysis
Carbonk
Copper 0.03Manganese ::f Chromium O*O3Phosphorous o e012 Molybdenum 0.005Sulphur O.O26 Tin 0.003Silicon 0.05 Nitrogen 0.009Aluminum 0.016 Vanadium <0.02Nickel O*O2 Arsenic <0.01
Mechanical Properties
Yield Point$ psi 39,000Tensile Strength$ psi 67,1+00Elongation per cent 2~.~ (8’1Gauge)
PROCEDURE
Specimen Preparation
Notch tensile and standard Charpy V-notch specimens3 Fig. 19
were prepared from the plate as follows:
Specimen blanks were taken from the midthiclmess so that
-5”
LINE OF
7/16°
ECCENTRIC NOTCH TENSILE SPECIMEN
\/
4V ElT–T-- ,~g~”’,010 R. .315” ~.ool”
F
1.050”3.005”—- +-Tl.08011t.oo511 -~ ,::;j:ll’
“–- 2,160 f.010
●CHARPY
FIG. I : TEST
V-NOTCH IMPACT SPECIMEN
SPEGIMENS.
“6-
the long axis’ in each case was perpendicular to the rolling
direotion. The surface of the plate was located by an iden-
tifying mark on each specimen. The Charpy blanks were then
rough machined 0.020 inch oversizej and the notch tensile
blanks 0.01s inch oversize. After
were machined to size and notched.
speciw.enswas cut perp~ndicular to
heat treating, the blanks
The notch for th8 Gharpy
the plane of the plate.
The steps involved in the preparation of the test speci.
mens are illustrated h Fig. 2.
Subcritical Heat Treatment
The specj.menblanks were heated at a temperature wibhin
the 700°--12000F range for periods of isothermal time ran~ing
from a Tew seconds at temperature to as lang as one week. For
times greater than five minutes~ a Lindberg forced air con-
vection furnace was used: for times five minutes or less~ a
neutral chloride salt bath. A chromel-alumel thermocouple “
at the center of a specimen blank was used to obtain an
accurate measure of the time at temperature for the shorter
isothermal times---
Three different cooling rates--air cool, furnace COOI-X-++F
%-For impact specimens the time required to reach a tempera.ture of llf)O”Fwas 1 1/2 minutes$ and to reach 1200°F,2 1/4 rnimtes; for notch tensile specimens, the times were3/’4and 1 minute, r@spectivelyO
%+-Furnace cooled at an average rate of 1.8°F per minutedown to ~OOQF9 then air cooled to room temperature.
-7-
ROLLING DIRECTION
FIG. 2; PREPARATION OF CHARPY V-NOTCH ANDNOTCH TENSILE SPECIMENS FROM “c”STEEL PLATE.
-8-
water quenc’h--were employed from the subcritical temperatures.
Normally$ spectiens were tested one month after heat treating
in order to approximate the elapsed time between welding and
testing in the previous work; however~ the room tem-perature
aging e~fects were also evaluated by testing Charpy specimens
at different times after water quenching from 1200°F.
The heat
notch tensile
specimens.
treatments employed are given in Table 111 for
specimens and in Tables IV and V for impact
Cooling Curves
The cooling curves for test specimens which were air
cooled and also water quenched from 1200°F were determined
with a cb.rornel-alumelthermocouple and a portable potentiom-
eter. The couple was positioned in a saw cut extending to
the center of a specimens with the cut then being peened shut.
The cooling curve for furnace cooled specimens was obtained
by taking temperature-time readings i%om a furnace controller.
Testing Procedure
The test equipment and procedure for the eccentric notch
tensile tests were essentially the same as those used previously
(1,2,3)0 The specimens were positioned in the Tixtures, Fig. 3~
so that the fiber in line with the identifying mark received
the maximum tensile load. The initial eccentricity was set
at l/~.inchj that is~ the centerline of the specimen
placed 1/4.inch from the loading axis of the tensile
was dis-
machine
—.
-9-
A TENSION
FIG. 3: METHOD OF LOADING TO OBTAIN 114 INCHECCENTRICITY. (ECCENTRICITY AND THEPOSITION OF FIXTURES ARE EXAGGERATED.)
“le-
as shown in Fig. 3.
In testing below room temperature the specimen was
cooled to about ~°F below the desired testing temperatures
allowed to warm up to the testing temperature and then tested.
In testing above room temperature, the specimen was heated
to about ~°F above the desired testing temperature, allowed
to cool down and then tested. The tests were performed at
constant temperature since the testing time was about 30
seconds, whereas the warming-up or cooling-down rate was about
l°F per minute. TIIe specimens were brought to temperature by
means of an appropriate bath of isopentane-dry ice r.ixture
or hot water. Temperatures were measured by a copper-con-
stantan thermocouple wrapped around the specimen. All of the
tests were carried out at a low strain rate; the crosshead
speed of the tensile machine was approximately 0.1 inch per
minute. The property that was measured was the eccentric
notch strength, maximum load divided by the original area
at the notch bottom.
The impact specimens were cooled or heated in an ap-
propriate bath of isopentane-dry ice or a high temperature
oil* As before, a copper-constantan thermocouple was used
to measure temperatures. The specimen was held in the bath
ten minutes to assure temperature uniformity, and then
transferred and tested in a standard impact machine in less
than $ seconds. Both the energy absorbed in fracture and the
per cent fibrous fracture values were obtained.
RESULTS
Transition Temperatures and Hardness
Transition curves obtained”’from the eccentric notch tensile
tests have been assembled in Fig. 1A--8A in Appendix A; and
from Charpy V-notch tests$ in Figs. lB--7B in Appendix B. From
these curves the transition temperatures were determined using
the following criteria:
[A) Eccentric Notch Tensile:
The temperature corresponding to a notch strength
midway between maximum and minimum on the average
notch strength
See Appendix A.
(B] Charpy V-Notch:
curve (dashed line
(1)
(2)
(31
The temperature
The temperature
between maximum
curve.
The temperature
in the figmes)
at which 15 ft-lbs were absorbed.
corresponding to an energy midway
and minimum on the average energy
at which the fracture was ~0 per
cent fibrous.
Transition temperatures determined from the notch tensile tests
are summarized in Table 111; and from Charpy V-notch tests~
in Tables IV and V. Rochwell B hardness values are also given
in Tables IV and V.
-12-
TABLE III
Subcritical Heat Treatments and Transition Temperatures(Eccentric Notch Tensile) of “C” Steel
The atTemperature Transition Temperature ‘F Plate No.
As-Received~tAs-Received
-65-40 1;
700°F$ Air cool
10 minutes1 hour8 hoursI.s hours30 hours1 week
10 minutes10 minutesI hour24 hours2& hours1 week1 week
10 minutes20 minutes1 hour1 hour12 hours1 week1 week1 week
s minutes15 minutesI hour12 hours112 hours1 week
-34-40
8000F, Air COOl—
+4
9505F, Air Cool
=42“35-38-32-36-58-38-36
-38-30-26“30-48-50
11I
11I
1
III
11I
II
III
II11I
11II
IIIIII
(1)+$Previously reported in weldment study .
-13-
~ minutes1 hour10 hours20 Ijgtis72 hO~s
s minutes1 hour10 hours96 hours
0-10 seconds1/2 minute~ minutes1 hour10 hours72 hOur~
TABLE III (Continued)
Transition Temperature ‘F
1200°F, Air Cool
-38==38-36-28-40
llOODF, Water Quench++-
+1.2
+6+8“10
1200°F, Water ~ench+:~
+46+~o+68+68+86+74
Plate No.
IIIIIIIIII
II11II11
IIII
IIIIII
}$+$A1lwater quenched series aged one month at roomtemperature before testing.
“l\”
Tmr..zlIv
Subcritical Heat Treatments Transition Temperatures (CharpyV-i[otch) and Harnesses of liC~~St~el
Transition Temperature ‘F—. —
Time Midpoint 50% Fibrousat Temperature Fracture
As-Received 118 13a
10 minutesT hour20 hoursLO hours72 hours112 hours120 hours
1100’F, Air Cool.—
108 142110 142U5 138115 14513.5 140125 158133 252.
llOO°F~ Furnace .@ol
10 minutes 110 lq.g120 hours 137 152
1200”F, Furnace COOT
1/2 minute 110 135
11009F, Water Quench%-
10 minutes 135 1)81 hour 135 105i4 hours 152 165112 hours 157 1681 week 157 180
%-Aged one month at room temperature
15 Ft. Rockwell BLbs. Hardness
87 74”7~
82929892
110115
as’102
85
110116116122118
747371,696’86767
7466
74
85868682
..
NOTE : All results based on Plate 11
-15-
TABLE IV (Continued)
Time Midpoint SO% Fibrous 15 Ft. Rockwell Bat Temperature Fracture Lbs. Hardness.
0-10 seconds1/2 minute2 minutes10 minutes1 hour~ hours10 hours20 hours72 hO~s
1200W, Water Quench+$
110”155 140-195165 190160 190158 200160 198177 205’170 ZLz184 210182 212
90”140L46142lko140148l~o162158
77-869190899089888682
+Aged one month at room temperature
NOTE : All results based on Plate II
-16-
TA.BLEV
Aging Treatments$ Transition Temperatures (Charpy V-Notch],and Hartiesses of IIc~lSteel, Mater Quenched from 1200°F.
Aging TimeIsothermal at RoomTime at 1200°F Temperature
l/2 minute As quenched1/2 minute ~ hours1/2 minuto 26 hours1/2 minute 3 days1/2 minute 7 days1/2 minute 14 days1/2 minute 30 days1/2 minute 42 days1/2 minute 65 days
10 hours 10 days10 hours 30 days
20 hours ~ da~s~2X)hours 30 days
Midpoint
.-.
180. . ..-.------
165
150178
137182
Transition Temperature~°FSO% Fibrous 15 l?t.Rockwell BFracture Lbs. Hardness— —.
-..---
177212
165212
+@ed at 400°F for one hour immediatelyand prior to room temperature aging.
NOTE% All results based on Plate II
13015’8
after quenching
7282
—
“l7-
In the following sections of the report, the ef’fectsof
the various subcritical heat treatments are evaluated with A
and B(l)++above as the criterion of embrittlement for the
notch tensile and the impact testy respectively.
Base Plate
As a check on
in the as-received
any variation between the two large plates
condition, the notch tensile transition
behavior was determined for Plate 11. The results are shown
in Fig. 1A%+. In Fig. 4 a comparison of the results with
those previously reported(1) for Plate I shows that the
distribution of values for Plate II was shifted to higher
testing temperatures, the transition temperature being -40”F
as compared with -659F for Plate I. In addition, the upper
, level of the notch strength values was slightly lower for
Plate II, indicating a lower tensile strength. Due to this
difference in
subcritically
plate number.
transition behaviors the results obtained with
heat treated plate have been separated as to
The transition curve obtained with impact specimens from
Plate II is shown in Fig. 13. No check was necessary because
+PTheuse of either B(2) or B(3) as the criterion ofembrittlement for the impact tests would reveal thesame general effects as B(l).
+H+The letter following the figure number refers to thecorresponding Appendix.
-18-
—
..
1100 ----------
I I1--- I 1
K --- t- --I I 1
1- 1
75
t--
PLATEIt
II-+... -----------
I I I
:140 -1oo -60 -20 20 60 100 14
TEST TEMPERATURE ~ “F
FIG, 4: COMPARISONTRANSITIONPLATES OF
OF ECCENTRIC NOTCH TENSILEBEHAVIOR FOR TWO DIFFERENTAS-RECEIVED “C” STEEL.
.-. —
this plate supplied all of the impact specimens for this
investigation.
Air Cooled
The individual transition curves for the various sub-
critical heat treatments employing an air cool are shown
in Figs. 2A--6A for notch tensile tests~ and in Figs. 2B
and 3B for impact tests.
The relationship between notch tensile transition
temperature and time at various temperatures is plotted in
Fig. ~. For each of the five temperatures investigated--
7oo”, 800”, 950°, 1100°$ and 1200BF--there is a slight
embrittlement at the shorter times. With increasing time
at 7C)O°F,the transition temperature appears to be unchanged$
at least up to times of one week~ At each of the higher
temperatures the transition temperature remains constant
with time after the initial embrittlement and then approaches~
or drops below the base plate value at longer times.
It should be noted that at 700°P 800~, and 9~00F both
Plates I and II were used and yielded transition temperature-
time curves which were similar$ but with the highest transi-
tion temperature of Plate I (-@°F) being about 10°F lower
than Plate II (-30°F) over the time interval considered. This
is in agreement w~th the base plate values which showed that
Plate I had a liner transition temperature.
With the as-received plate as a basis of’comparison it
IACL
-20-
-10
K “
I I
● PLATE It700° F
. . . - ., .-— .—..— ... .. . . . ..————
:~:,- :,D ~“”:. ‘.* +
J-70 ‘ _-lo
r
-3~
●
-50 ---
-70° _-lo
r
tt~ _~o ..:
w
zt
~_c@ . .
-50
-70-lo
“30I●
P’y - ;950° F
-* ----
0 0●
o
““]-T
~ I-.
il
—.
~
-.
t
~. . . . i ---- . .
7 i
I 1 1 1200° F I I
-50
1) I
III i
-70 ~~; I
o 0. I
lSOTHERiAAL TIM#”~ HOUR’;(
ba%0 .
W*
\ ‘-0
.—
O*,_
—
—
. . .
—
—
.
—
FIG. 5: ECCENTRIC NOTCH TENSIL15 TRANSITION TENl-PERATURES OF “C” STEEL AS A FUNCTION OFTIME AT VARIOUS SUBCRITICAL TEMPERATURES.AIR COOLED.
.—
-21”
-
is seen that the maximum embrittlement amounted to a 10°F
increase in transition temperature for Plate II, and-a
2~”l? increase for Plate 1.
An examination of the individual notch tensile transition
curves in Figs. 2A--.4Areveals thatflalthough the slope and
the amount of scatter in the transition range are about the
same for the two plates$ the upper level of notch strength
for Plate I is consistently higher (about 10~000 psi] than
Plate 11. Again$ this is in agreement with the findings for
the as-received base plate and denotes a slightly higher
tensile strength for Plate 1.
In Fig. 6, the transition temperature-isothermal time
relationship is shown for impact specimens heated at 11009F
and 1200°F. The transition temperature remains essentially
the same as the as-received plate until~ after about ten hours
at 1200°F and 72 hours at IIOO°F~ the transition temperature
slowly increases with tile1200°F curve showing a somewhat
faster rate of increase. It should be noted, however$ that
the upper level of energy absorbed slowly increases not only
with time at temperature but also with temperatures thus
indicating an improvement in impact properties at the higher
testing temperatures. This improvement is not reflected in
the transition temperature using any of the three criteria
employed.
From the hardness values listed in Table IV$ it is
“22-
160
L‘;- :‘“‘1\
140 — O 1200”F ...}.
● IIOOOF
120
1---”----1
~: “: I
i--l “’“- - ‘“--O(-1100 .+ ....
IAS RECEIVED ‘“ I ,/
-o- ““- “+o ●;% -
● ●
80 * ---L---- ---––- --- --- ---- j- -- J--------- ----- -- t . .. . .. .. . .. ,.,,.
600 0. I 100 1000
ISOTHERMAL TIME Lo HOtJRS
FIG. 6: GHARPY V-NOTCH TRANSITION TEMPERATUREAS A FUNCTION OF TIME AT SUBCRITICALTEMPERATURES. AIR COOLED. PLATE It.
-23-
appaeent that with increasing time at llOO°F there is a
gradual softening; at 1200°F$ the decrease in hardness with
time is even more pronounced. A.swill be sllownlater this
is an exception to the increased hardness usual~y associated
with increased Charpy transition temperature.
A comparison of the notch tensile and impact results for
the llOO°F and 1200”F heat treatments points out two interest-
ing differences: (1) At the shorter times, a slight embrit-
tlem,ent is evident with notch tensile but not with impact
specimens. (2) The transition temperature then remains con-
stant with time, until at the longer times, softening sets
in and.the notch tensile transition temperature decreases as
contrasted to the impact transition temperature which increases.
Furnace Cooled
In order to investigate the effect of a slower cooling
rate$ three spot checks were made with impact specir.ens
furnace cooled at the rate of 1.8°F per minute after: (1)
10 minutes at llOOQF, (2) 120 hours at 11OO”F, and (3) l\2
minute at 1200°Fo The transition curves for these treatments
are shown in Figs. @ and ~B. For each of these three cases
there was no significant difference in transition temperature
or hardness from specimens heat treated and air cooled. On
the basis of these results, it would appear that furnace cooled
impact specimens should exhibit the same behavior as the air
cooled specimens over the ranges of timleand temperature under
study.
-24-
W’ater Quenched
The notch tensile transition curves for the various water
quenched series are assembled in Figs. 7A and 8A and the impact
transition curves in Figs. 6B and 7B. All results are based
m Plate 116
Aged One Month at Room Temperature
In order to maintain approxhnately the same aging interval
as previously used in the weldment studies~-gboth notch tensike
and impact specimens were aged one month at room temperature
after water quenching.
The transition temperature-isothermal time curves for
notch tensile specimens ’’quenchedfrom llOO°F and 1200°F are
shown in Fig. 7. For the lZOO°F series, the transition
temperature is raised to +lO°F at the sho~ter times$ amount-
ing to an increase of sO°F above that of the as-received plate.
With
this
This
increasing times the transition temperature remains at
level and then decreases slightly at the longer times~
general shape of the curve is in agreement with the
llOO°F air cooled series+
For the 1200”F series, the entire
still higher transition temperatures.
ture is about +455F at times less than
curve is displaced to
The transition tempera-
five minutes$ rising
$+Inthe work with weldments, about one month elapsedbetween the time of welding and testing.
.25.
to a maximum of about +8~°F after 10 hours and then decreasing
slightly at 72 hours. Thus, the transition temperalmre has
been raised from ~~” to 12s°F above that of the as-received
plate. Again, the general trend of the curve is in agreement
with the corresponding air cooled series.
It should be noted that two points were obtained at O to
1.0seconds and 1/2 minute at temperatures respectively~ in
order to investigate times at temperature which would ap-
proxtiate those actually existing in weldments. The transi-
tion curves for these times, Fig. 8A (a and b) show a some-
what greater scatter band than those for the longer ttii~es.
At isothermal times greater than 1/2 minute, the high values
of the scatter band have been lowered$ which? in effect$
raises the transition temperature.
In comparing the individual transition curves, the
water quenched series treated at both 1100° and 1200°F
evidenced a higher upper level than the comparably heat treated
and air cooled series. This indicates that water quenching
served to increase the tensile strength.
The transition ternperature=isothermal time relationship
for impact specimens quenched fron.llOO°F and 1200°F are also
shown in Fig. 7. The ssme general shape of curve is evident
for both the llOO°F and 1200°F water quenched series as for
the comparable air cooled seriesj i.e.~ a constant transition
temperature with j.sothermal time~ increasing slightly at
m“5E1-
FIG.
-26-
110~ fl I I I I I I I
II ECCENTRIC NOTCH TENSILE
90 I ) 1 1 1 J’-4--- 1
F~~~-~~~~.0
\2Q0°F
70 -–– ‘ -“““-”--–~----------9— --------------.
~!
50- o—’ –.....+..----~....--_+_.._..__.r
..L.....
1,!
~30’’’ \’-- \-it~
I.. . .... —.+ ——
““”--l”-–-----”; -- ““-u-
1
----.---..l_.__..1“o~,-_ j IIOO”F I 110 .
“o
~
~~
-lo ---’”------- l----- \ -~
!
~ ‘. ‘%~~~t--- ----+-----– c
I1 1 I
L I
-30 ..— -j... .... ...1 -. ... , “:. ----- .; . . .. .. .I --’- ‘--- ‘---
~AS RECEIVED t
-5011
2207t I ‘ ‘- I I I I
CHARPY V-NOTCH
200 + ~~
I
III
I~. 1---:
J . . . ..-. +..-.. ----- -i----------- ----
I
180.---.,; I ~ - ~......-
lG~ ... J
●“--”-’~”””Y”jY”””’””’-”-”-’---1200e F .
140 -*—- ~ I~
.--+ --—..-1 I ~ ,I
,20__.. :_._..--+_ . .--.-d..l[@F”~. ................-oo_.,.,..>___—“~ “~o
100-’- ; ‘ !1
1 i-AS RECEIVED
I80 d< I I I
0 0. I 1000lSOTHkRMAL TllliE*HOU’i?8
7: TRANSITION TEMPERATURES OF “C” STEEL ASA FUNCTION OF TIME AT SUBCRITICAL TEMPER-ATURES. WATER QUENCHED AND AGED ONEMONTH AT ROOM TEMPERATURE. PLATE II,
-.
“27”
longer times and with the 1200PF curve showing a slightly
fast~r rate of increase. A comparison with the air cooled
series shows that the quench-aging treatment served to
embrittle the steel at all isothemml times~ as noted by
the displacement of both the 1100° amd 3200°F curves to
higher transition temperatures and with the embrittlement be-
ing of a higher magnitude for the 1200°F heat treatment.
The llOO°F water quenched curve is displaced about 2~°F and
the 1200aF water quenched curve about SS~F above their respec-
tive air cooled series (which at the shorter times was the
same as the as-received plate~]
For the 1200°F heat treatment, a series of specimens
tested for isothermal time of O to 10 seconds$ Figo 7B (a)
showed considerable scatter in the test results. The
minimum and maximum transition temperatures to ‘be expected
are ~iven in Table IV; however, this point is o~itted in the
transition temp~rature-isothermal time plot because of the
apparent difficulty of reproducing structures from specir.en
to specimen at such short solution times.
An examination of the individual impact transition curves$
Figs. 7B and 8B, reveals that for the shorter times at llOO°F
and 1200°F the maximm energy level is about 7 ft-lb less than
t“heas-received plate. With increasing times the upper level
is gradually raised above that of the as-received plate with
the 1200QF series showing a slightly faster rate of increase.#j
-.
-28”
As noted previously this effect of time and temperature on
the upper level was also evident in the air cooled s~ries~
A review bf the hardness values in Table IV shows that
for the llOO°F series$ the hardness, after an initial in-
crease of”ten points Rockwell B~ gradually decreases at the
longe~ times. The L200DF series shows a similar trend ex-
cept that the initial embrittlement amounted to a hardness
increase of 15 points. It is interesting to note that the
hardness and transition temperature do not parallel one
another with increasing isothermal tirnem Although both
indicate an initial embrittlement, the impact transition
temperature increases at long isothermal times; whereas the
hardness decreases. This behavior at long isothermal times
was also evident in the air cooled series.
Room Temperature Aging Study
In order to check the room temperature aging effects,
impact specimens were tested after treating at 1200’F for 1/2
minute and aging at room temperature f’ortimes ranging from
five hours to 65 days. The transition temperatures and
hardness values are summarized in Table V, while the indi-
vidual transition curves are shown in Fig. 7B (b--j).
In Fig. 8 the energy transition curves for aging times
of five hours and 30 days are compared with the as-received~.&plate. It can be seen that f’orfive hours$ aging time, the
40
3C
20
10
.29.
I I IHEATED AT 1200°F
— FOR 1/2 MINUTE
AND WATER QUENCHED.
..q______
/
.——
/1-0 0 50 100 150 200 250 300 350
TEST TEMPERATURE G=”F
FIG. 8: EFFECT OF ROOM TEMPERATURE AGING TIME ON CHARPYV-NOTCH TRANSITION CURVES OF “C” STEEL.
zoo~– 1 I 1
IL IHEATED AT 1200”F I
I 80 FOR 1/2 MINUTE --- ‘ -; 1
AND WATER QUENCHED I I ,.I
160 ------’-- I ‘ 1 I <---, - ,.
~“ 0140 ./O”: ’10
1/-~120 -
00
.—
Ioolf : j ..1.
/ IVSI RECEIVED
mI
/-As pcE’vED ~ ; I ~I I
o 10 20 30 40 50 60 70
G. 9:
AGING TIME -DAYS
EFFECT OF ROOMGHARPY V-NOTCHHARDNESS OF “C”
TEMPERATURE AGING TIME ON THETRANSITION TEMPERATURE ANDSTEEL.
=30-
energy values at the lower testing temperatures are about the
same as the as-received plate but lower at the higher testing
temperatures-- the upper level “being decreased from 42 to 36
ft-lb. For 30 days! aging timeg the entire energy curve is
shifted to higher testing temperatures. This shift amounts
to about a 60°F increase in 15 ft-lb transition temperature.
To obtain a measure of the rate of embrittlement~ ad=
ditional impact data were obtained for intermediate and
longer aging times. Complete transition curves were not
obtained$ but a rough indication of the 15 ft-lb transition
temperature was found from spot tests at temperatures which
would contaiii the 15 ft-lb value.
W*S then drawn through the points
taken. These additional resultsg
va~ues~ are plotted as a function
The best straight line
and the Is ft-lb value
along with the hardness
of aging time in FigO 90
Both the transition temperature and hardness show a fairly
rapid increase in the aging time interval up to about two
weeks$ then leveling of’f~ From these,results~ it can be
expected that all the water quenched series which were aged
one month at room temperature experienced the maximum embrit-
tlement by room temperature aging.
A check of the aging effect after heat treating at
another time (10 hm.ms) at ~200°F was made with two ~eries
of specimens aged 10 days and 30 days$ respectively. The
energy transition curves are compared with that for the
.,
.~~~
as-received plate in I’igo100 Prom this figure9 t’he
dependence CII’the degree of embrittleinent on room temperat-
ure a~ing time is again evidento
Accelerated A~ing
In establishing the impact energy transition Curvesg
it was necessary to test above room temperature--at times
as high as j~ODFO As a check on,accelerated aging in the
Testing bath$ hardu.essmeasurements were made on the series
or specimens water quenched after heating for mae-=halfminute
a-bL?OO°F and ag~d at room temperature for five hours and for
30 days. The Rockwell B hardness values before heating in
the testing bath and immediately after breaking3 are given
in Table KC.
The series of specimens aged for five hours indicates
that the hardness does not change more tlnanone point from
the as-quenched hardness of R~ 84-85 after ten minutes in
the testiingbath at temperatures up to 300°F] haweverfl those
specimens ag~clfor ~()days show thab the hardness is pro-
gressively decreased from R~ 91 to RB 83 in the temperature
interval from 175°--3~OoF0 The latter series of tests
indicates that accelerated aging9 as measured by hardness
tests9 can take place in the testing bath
above l~~9F in ten minutesO Howeverj the
critical heat t~eatment on the transition
at temperatures
ef~sc%s of sub-
temperature would
-32-
.- <I
HEATED AT 1200° FFOR 10 HOUR13AND WATER QUENCHED
40 —
30 -–I
20 --
10
0“50 Q 50 100 150 200 250 300 35
TEST TEMPERATURE *“FFIG. 10: EFFECT OF ROOM TEMPERATURE AGING TIME ON
CHARPY V-NOTCH TRANSITION CURVES OF “C” STEEL.
50 \
HEATED AT 1200”FFOR 20 HOURS. A.
UJ
q 40t
AND WATER QUENCHED
uzLLl
30
20 – ., . . ..——.
10
FIG.
I———... .
1
..———. . . .
1
0-50 0 50 100 150 200 250 3
TEST TEMPERATURE ~°F
&
. ...—
.. .
) 350
11: COMPARISON OF CHARPY V-hiOTGH TRANSITION CURVES~OR ROOM TEMPERATURE AND ACCELERATED AGING OF“C” STEEL.
-33-
appear ticbe unaffected in view-or the fact that the tempera=
iure at the 15 ft-lb en~rgy value was taken as the transition
temperature, and in all series this temperature was less than
1,7$‘F.
TABLE VI
Hardness Cheek on Accelerated Agin~ of Charp~ Specimens
SubcriticalHeat Treatment
Heated at 1200°Ffor 1/2 minuteand water quenched.Aged ~ hours atroom temperaturebefore placing intest bath.
Heated at 1200°Ffor 1/2 minuteand water quenched.Aged 1 month atroom temperaturebefore placing intest bath.
..-
Rockwell B HardnessAfter
Before Heating., Breaki=
275250225202174 91 92125 91 92100 91 91.
82 (RT] 91 91
A spot check on the effeck of accelerated aging on the
transition temperature was made by heat treating a series of
impact specimens at 1200°F for 20 hours$ water quenching$ and
then immediately aging at 4000F for one hour. The resulting
ermrgy transition curve is compared~ Fig. llj with a comparably
heat treated series naturally aged at room temperature for one
month . The pronounced improvement brought about by the
-3L
accelerated aging treatment is at once evident in the shift
or the entire curve to lower testing temperatures approach-
ing that of the base plate. The transition temperature of
the accelerated aged plate was only
the series treated at 1200QF for 20
Hardness checks on broken specimens
slightly higher tlhan
hours and air cool~d.
showed no change with
t h.es indicating that this accelerated aging treatment
resulted in virtually complete stability. Further work on
the effects of accelerated aging on the transition tempera-
ture is now under way. The general program to be followed
is outlined in the section on Future Work.
Microstructure
An examination of the microstructure was made of the
base plate and after all conditions of heat treatlnent to
aff’ord a possible explanation of the transition behavior.
Representative photomicrographs at 2000X are shown in
Fig. 12.
The microstructure for the as-received condition$ Figm
12(a)~ showed the same structure for both Plates I and 11.
No apparent explanation can be given to account for the
superior properties of Plate Is other than a difference
in composition or rolling practice which is not evident
in the microstructure.
The photomicrographs in Figs. 12{b) and 12(c) were
made after heat treating at 1200QF for 1/2 minute and
—. .
(n)&.yoeewd
.–,-..
(’d)1200~, 72 hours, ~aCecooled,not tested
Fist* 12: MICROSTRUCTURESOF ‘Cw
{c) 1200=, l/2mlnute, waterquenohed and aged onemonth at room tempernture~TT = +~6w
~e] L200V, 72 hours, waterquenohed and aged onemonth at mm temp.m = +138°F
(r) 1200TS 20 hours, waterquenched.Aged one hour●t 400V, TT = +11O*F
STEEL IN THE AS-RECEIVEDCONDITIONMD AFIWR VARIOUSWBCRITICALWAT TREATMENTS.
Wharpy Is f’t-lbTransltlonTemperature.
-36-
employing a furnace cool and a water quench~ respectively.
No difference in structure from that of tho as-received
plate could be noted for either of these two heat treatments.
At lon~ isothermal times at 1200°F, spheroidization starts
to set ins as noted in the f’umace cooled stru~ture in Fig.
12(d) and the water quenched structure in Fig. 12(e); how-
ever~ no difference between these latter two microstructure
could be seen due to the difference in cooling rate.
In Fig- 12(f), the microstructure is shown I’orthe water
quenched series heat treated at 1200°F for 20 hours followed
by accelerated aging at 400QF for one hour. It appears that
this accelerated aging treatment has resulted in a general
precipitation throughout the ferrite grains~ and as the
transition temperature and hardness checks showed, was ac-
companied by a considerable improvement in impact properties
and a pronounced decrease in hardness.
Cooling Curves
The cooling curves for the various cooling rates
employed from 1200UF are given in Figti13 and compared with
the cooling history for the first weld pass at the region
of low ductility for the two welding conditions under study
(1,2)previously ~ It can be seen that although the 100°F
preheat weldment has a faster cooling rate than the 4009F
preheat weldment$ the cooling rate for both these welding
conditions is intermediate to the air cooled and water
—-
140
120
100(
I&
sId mm
?aIXw
~ 60(
U+
4a
20(
c
FIG. 13:
-37-
LEGENDTEST SPECIMENS WELDMENT
(0.3 INCH FROM WELD ~)A..,. FURNACE COOLED—B..., AIR COOLED (IMPACT) E...,4OO”F PREHEAT 8 lNTER-c....NR COOLED (NOTGH TEWILE) PASS TEMP., I ST WELD PASS.D.. ..WATER QUENCHED F,... IOO”F PREHEAT B INTER- I
I
I
\
—
PASS TEMP., 1ST WELD PASi
IA
1
I 11 I
I I
I
. “~-”-”””-”““”-”-:”----J__L
30 45 60 75 90
COMPARISONOF LOWESTTIONS WITH
TIME- SECONDS
OF COOLING CURVES IN THE REGIONDUCTILITY FOR TWO WELDING CONDl-THOSE OBTAINED WITH HEAT TREATED
TEST SPECIMENS.
-38-
quenched specimens. Also, it should be noted that with an
air cools the impact specimens cool at a slightly slower
rate than the notch tensile specimens due to the larger
mass of metal.
DISCUSSION
Previous investigations(5,6)
have shown there are two
factors which increase the tensile strength and hardness
and lower the ductility of low carbon steel when it is cooled
from subcritical temperatures, namely, the solid solution
and the aging e~f’ects. The solution effect is the formation
of a supersaturated solution of carbon in Territe and has
the naximmm effect after fast quenching. The second.factors
aging, is the precipitation of carbides from the supersat”
mated solution which has the maximum effect at some
critical point of time and temperature, beyond which the
effect is decreased.
The results of the work to date can be discussed in
terms of the combined solid solution and aging phenomenons
commonly referred to as ~*quench-aging~f.
Considering the air cooled series first, the notch
tensile test detected a slight embrittlernent after heat
treating in the 700D--1200”F range, as contrasted to the
impact test which indicated no change in properties after
treatment at llOOPF and 1200°F at the shorter times. From
-39-
these results it can be reasoned that precipitation oc-
curred largely during cooling
in properties was realized on
slight ern.b~ittlementrevealed
and no appreciable change
subsequent aging. The
by notch ten~i~e tests
could be attributed to a slight solution and subsequent
aging effect due to the slightly faster rate of cooling
with notch tensile specimens. At the longer isothermal
timesO softening set in which resulted in a divergence
in transition behavior between the two specimen typesg
i.e.$ the notch tensile transition temperature decreased
while the impact transition temperature increased~ At
the same time~ howevers the energy absorbed by Charpy
specimens at the higher testing temperatures was increased.
The eccentric notch tensile transition and Charpy energy
behavior atilonger isothemal times might be expected as
a result of the slightly spheroidized structure. The
reasons for these differences in behavior evidenced by
spheroidized structures are not completely understioodo
It is believed that this occurrence is not the result of
any solution or aging effects$ but is due’to different
reactions of the two types of specimens to a slightly
spheroidized structure.
With a decrease in cooling rate--furnace cool-=the
spot checks with impact specimens showed that the transi-
tion temperature and hardness were unchanged from the
-k3-
ccmparably heat treated and air cooled series. Al,though
no notch tensile tests were conducted with furnace cooled
specimens$ it is believed that this slower cooling rate will
result in no ernbrittlement because it is to be expected that
the solid solution and aging effects will then be nil.
A consideration OT the notch tensile and impact
results obtained by water quenching reveals a pronounced
embrittlernent which is influenced by the following:
1. Isothermal time
20 Subcritical temperature
3. Aging tine
4. Aging temperature
The isothermal time at a particular temperature must
be long enough to allow the complete solution of the soluble
phase (carbon). With incomplete, solution, the lower is
the degree of supersaturation and$ consequently, the slower
is the rate and amount of precipitations resulting in smaller
changes in properties as shown “~ythe tests after heat
treating .at1200°F for short tires. The divergent behavior
between the two specimen types at long isothermal times
can agair~be attributed to the spheroidization effect men-
tioned above.
If time at temperature is long enough to permit complete
solution$ the solubillty increases with increasing temperat-
ure . The degree of supersaturation correspondingly
inoreases on fast quenching exerting a more pronounced
effect on the properties after subsequent aging. Evidence
of this was obtained in the series water quenched from 1200°F
and aged one month at room temperature which showed a higher
transition temperature and hardness than specimens similarly
treated at llOO°F~
The aging study carried out at 1200*F with impact spec5.-
mens indicated that although the solution effect served to
harden the steel, the impact properties for the series aged
at room temperature for five hours were about the same as
ttm as-received plate. It can be reasoned then that no
significant precipitation occurred in this short aging
time; however$ an increase in aging time was accompanied
by an increase in transition temperature and hardness
indicating precipitation from the super-saturated solid “
solution.
Aging at a higher temperature (400°F) after water
quenching from 1200°F effected a large improvement in
the impact properties and a pronounced decrease in hard-
ness. This occurrence can be attributed to l’overagingl~~
i.e.~ the aging effect was carried past the critical
point of time and temperature.
Although all these factors have pronounced effec”ts
on transition temperature and hardnessg no visible ef’f-ect
is apparent in the microstructure with the exception .
-h2-
of the overaged series. Here, a general precipitation is
evident in the ferrite grains, Fig. 12(f) . It can be
speculated t’hatin all the other water quenched series the
precipitate was retained coherent with the matrix and of
SUCIIa size so as not to be microscopically visibleo
Upon exceeding a critical time and temperature, as in the
accelerated a~ed series$ the precipitate broke free or the
matrix and grew in size so as to be visible~
A review of the work on weldments shows that tile
necessary conditions for quench--aging are present and
appear to be the on137possible explanation for the zone
of minimum ductility located outside the so-called heat-
affected area, i.e.~ in a region which was not heat~d
above the lower critical temperature at any time.
From Table I it can be seen that maximum embrittle-
rnent (4g°F increase in notch tensile transition tempera-
ture) occurred in a weldment made with 100°F preheat. With
a 400°F preheat, the embrittlement amounted to a .20DFin-
crease in transition temperature. In comparison, the air
cooled series of subcritically heat treated specimens
showed an increase of 2~°F in notch tensile transition
temperature for the same plate (Plate 1) and a 10°F increase
for Plate 11. From tlneseresults$ it would appear that the
cooling rate for the 100QF preheat weldment is somewhat
greater than that obtained by air cooling notch tensile
speeimensz while for the 400°F preheat weldment the cooling
rate is about the samea Confirmation of ~~is is seen in
-43-
the cooling curves in Fig. 13. Only the cooling history
of the first weld pass of the six-pass weld is shown at the
zone OT minimum ductility for the two weldments. Consider-
ing each weldment separately, subsequent passes resulted
in a lowering of the peak temperature reached at the
critical zone and a decrease in the cooling rate$ Fig. 14.
It can be expected that each weld pass would contribute
to the solid solution and aging effects~ but it is believed
that the maximum temperature reached, time at this tempera-
tures and the subsequent cooling rate of the first few
weld passes govern the amount of carbon initially retained’
in solution$ while the ’following passes serve mainly as
short accelerated aging treatments.
The almost complete elimination of the critical
region by a postheat treatment at llOO°F (see Table I) could
be attributed to overaging.
Due to the complexity of tiletimeg temperature and
cooling rate factors in a multiple pass weld--factors which
have been shown to be interrelated in the quench-aging ,
mechanism-- it is not possible to make more than this general
comparison with the present investigation
“44”
)
-. .-
-45.
CONCLUS1ONS
The quench aging mechanism appears responsible for
the loss in ductility and the increase in hardness of a
low carbon ship plate steel when subcritically heat
treated. The severity of the embrittlement increased
with increase in (1) solution temperature, (2) severity
of quenchJ and (3) aging time at room temperature.
Direct evidence of precipitation was obtained in the
microstructure of an !overaged~ specimen (Figure 12f).
Isothermal time at temperature had relatively little
effect on the properties other than a softening due to
spheroidization.
The results
firm the earlier
of this investigation appear to con-
supposition that the quench-aging phenom-
enon was responsible for the zone of maximum embrittle-
ment beiqg located outside the so-called heat affected .
zone in ship plate weldments.
.-
-46-
FUTUR~WORK
To determine the maximum effects involved in the
quench-aging phenomenon, it is planned to supersaturate
‘lC”steel to the maximum, i.e., water quench from 1300°F
(just below the lower critical temperature), and age for
various periods of time at room temperature and at
selected elevated temperatures. The change in properties
will be followed by impact and hardness tests~ supple-
mented by microscopic examination. It is hoped that
this work will also suggest possible methods for
eliminating the quench-aging effects responsible
the embrittlement of’steel when welded.
ACKNOWLEDGMENT S
for
The authors wish to aclmowledge the aid of Messrs.
C. A. Beiser and D. J* Garibotti in performing the tests
and also for their many helpful suggestions~
-’
BIBLIOGRAPHY
1. G. Sachs, L. J. Ebert, and A. W. Dana, ~tTheFundamentalFactdrS: Influencing the Behavior of Welded StructuresUnder Conditions of Multiaxial Stress and Variations ofTemperature~ Stress Concentration and Rates of Strain”,First Progress Report, Ship Structure Co.mmittee$ SerialNo. SSC-24, JYaY m, 1949.
2. L. J. Klingler$ L. J. Ebert9 and W. M. Baldwin, Jr.,l~TheFundamental Factors Influencing the Behavior ofWelded Structures Under Conditions of Multiaxial Stressand Variations of Temperature Stress Concentration andRates of Strainl~~ Second Progress Report, Ship StructureCommittee, Serial No..SSC-34, November 28, 19490
30 Lo J. Klingler and E. B: Evans, ~tTheFundamental FactorsInfluencing tineBehavior of”Welded Structures underConditions of Multiaxial Stress, and Variations ofTemperature!t, Third Progress Report, Ship StructureCommittee, Serial No. SSC-~4, October 14, 19S2.
q. Technical Progress Report of the Ship Structure Committee,The Welding Journal, Research Supplementj Vol. 13 (July1948)S PPt 377s-384s.
~. G. Sachs, L. J. Ebert, G. B. Kasik, and J. F. Nejedlik,“Fundamentals of Annealing Low Carbon Steelt?,Part I,Iron and Steel Engineer, Vol. .23 (November 1946),pp. 88-98.
6* R. D. Pollard, “Metallurgical Aspects of Annealing!’,J. Iron and Steel Inst., Part 1, VQ1.’162 (May 1949),PP” 79-85=
-48-
APPENDIX A
ECCENTRIC NOTCH TENSILE TRLNSITION CURVES
and
METHOD OF DETERMINATION OF ECCENTRIC
NOTCH TENSILE TRANSITION TEMPERATuRE
.-
Method of’ Determination of dccentrie—. — ~.—- — —..._Notch Ten.s~le Traas~tlon Termeratures
In establishing the chan~e Trom ductile to brittle
behavior with the eccentric notch tensile test, 30 specimens
or iTOl~ewere used Tor each series. }.scan be seen from tke
test dLa.ta in ~7igs. ~~ to 8A, considerable scatter occ-urred.
in the ductile-brit’cle transition zone, where=s at hi:her
or lower testin~ te~:lperaturesthe resvlts were Y“!oipemlil-~rr~,
For each seri~~ the majmity of the tests (about 20) were
conducted at tem:psratures wl-khirlthe transition ran~e. ./L
sCaJG~e~ banti was obtained by”drawing an upper and.a lower
limiting curve (solid lines i-nthe figures) whit’h coiitained
all the test points, Tnese limitiq curves were ccmstmc-ted
parallel to each other deviatj.ng only at the lTkneel~and the
~[toe’tof the curves, An average notch st~ength curve
(dashed lines in the fi~ures ) was then dra,wnparallel to
the limiting curves$ and blsectinq ti2ehorizontal clistancc
i“fltiilE i~~m,sition rar.?:e,1 TMe transition temperature was....
taken as the temperature at the vertical midpoint of the
avers.g~notch stren~th curve.
The averaqe notch strength curve was also determined
by averaging tinenotch strength values at e~ch test temper.
ature for each series and fittin~ the best curve throu:”u
these points. Using the midpoint criterions the resulting
trmsition temperatures agreed very well witlnthose ohtainecl
by the m::!thodoutlined above .
See also pages 9 and 13 and Figures 10 and 11 of Ref’~ 1.
1251 I I I I I 1 I I 1 1 1 1 I
100
75
w
I80- .~
259 .!- ~. “w
o I-140 -1oo -60 -20 100 140 180
TEST TEMPERATURE ~°F
5~ FIG. 1A: ECCENTRIC NOTCH TENSILE TRANSITION CURVE OF
AS-RECEIVED “C” STEEL. PLATE II.
o0T
c,
I
. .
r
c, I
,00 (01 ‘— 10 MINUTES
PLATE I
75
W ,
25
mo
m 125(b)
g 100–10 MINUTES
PLATE 1 l-- -—” -—- “ — w
7 75 ‘ -]
z~0zw 25cc+(n 12:”
(c)~ 100‘1 HOUR
1- PLATE Ioz 75u~E*z -0— Iu
25
- 1-
+ ——
000I&l 125
,00 (d)
— 24 HOURS
PLATE I
75 , I ]
50A. .m
25 i
c
t25
k
k)100-24 HOI-IRS
PLATE ~
m75L
o 500(7 .. tlmlHw7TJT--ry 11‘-u I PLATE Iw
v1- 75 —-t- --t- I 4-
‘nIll
01 I I I 1 I I-140 -100 -60 -20 100 140 Ico
TEST TEMP&ATURE6~0F
FIG, 3A (CONT.)
-140 -1oo -60 -20TEST TEMP?RATUR;O— OF
100 140 180
FIG. 3A(0-9): ECCENTRIC NOTCH TENSILE TRANSITION CURVES OF“C” STEEL. SUBCRITICALLY HEATED AT 800*F FORTIMES INDICATED AND AIR COOLED,
-53-~ 0
I I m
1 —--—
I—.”.,-., ..—. ~ .—L—. — o~— -;‘I ..______________________.. .—–—+.—.--4.-...-./.;.,..-..-.._ . . –~!’—----1---------—8-8 8,. ~
I I Im II‘ ~~
{,. l“-”-”.... Owt-”,i 1 ‘u—.—.. -.—... .1,I I ?d
FllII~\am, o— ----- -— —-—-
1,,
/~L . . ... -.
i
lSd 0001 -= H19N3H1S H91ON 91M1N3093
..
-54-
– -—~
1’.,.-,-.,..}.....‘-i-+—
: “[r+.—L
I’llf+
T-*-\K~f-~,\&‘0k ‘8<
_ ‘-$\\w?
Io-l
30x
~~--
1U-)goocd Wkl——
lSd 0001 H
. .
—.
—
1
+t&1’-
---
+-~t..
8CJ
wE
Io
I I ~!’
., . . ..–...--i..—. -. ——..–—.–.–-——.–-—..- :,-–--—-+--. .L . . ———I ,,{..;
----1-------;--“---’ ~ - ‘------“”
......“-”:1
o.,.. i — ~
—= .... 1..-.: --------
:_l;-
;- ..– . ...-. .–------ ‘ +--- -
.“ . -. ~ -.L+-- ->-- ----- ----+-- —--- -“- --— g
% -- ~----
j>
—.- -.—
i“ ~ ~ ‘~~-
0w
I..— 1II
11
: ; :: - ~~~ T ; ,: ~~~ ~ ,
~ $ --- ---jail..”””””””””’. “ ! !T :- , _ ;-.— ---- - -x- ~’
\\
—T .,.. . . . ..-\*
--- -.
r
.+y
r “< “\
00%. - —-—-\-
41
—q a +----– –- --J-
\+
--*“NW - ._..T_ -–- L+\
~- ,4 :--- ““””‘- ;-”: y-- ‘— o
.. . . . ..—\
y
L-–[\
J ig{mw :t-. . 1-
-1 :1* _
(n b3 3
zm
z a— ‘g
3
5 z I3 a
. .--?” —-”— “
x-.. . —.. — . ..— —.
urn ZS? ~ _- ~~1
:
moLrJomo~ ~yobmm tio z%%%~z$ltioslg 2R:~7—_ ._
am
00
00T
0WT
lSd OOOl~H19N3HlS I-I91ON W11N3393
d(D
tiii
mixrrw
..
125
10(
7!
x
25
c12:
Ioc
75
5C
25
c125
100
75
50
25
c125
100
75
5C
25
0
—+----I1
~ -“----l---~”---~l”--~”-----~’---/”-–
(b)—I
—r
—.. .
(c)
I I1 I I I1
HOUR
I-.—-—— -.._—— ——. - ,.. P < ~
425 * 1
-----J ---i--’”–.”.-. - -“----
——-—— ..;____ —-.. . .–— —
(
I I I 1 I I10 HOURS -“”+*T
T++- “ : ‘-p “ :
1 .–.—–.~!””- x
#0°
e
.-__.p-”“ -$”’
r
-..
(d I 1 I I I I96 HOURS
-140 -1oo -60TEST
-20TEMPER&JRE
.-.
---- .+++-
1 [.:,.—....--..——.-—.——-—..
#t
—. —● ✍
✍ ❞✝
W-—.-——...
“—.—-
60 100 140— ‘F
. .—
—=
-—
FIG. 7A(a-d): ECCENTRIC NOTCH TENSILE TRANSITION CURVES
180
FOR “C” STEEL, SUBCRITICALLY HEATED AT lIOOoFFOR TIMES INDICATED AND WATER QUENCHED.AGED I MONTH AT ROOM TEMPERATURE.PLATE II.
-57-U
\I
I
lSd 0001 —H1S)N3L11S HC)1ON 51EllNX)Xi
C-5ii
.—
-y3-
APPENDIX B
CHARPY V-NOTCH TRANSITION CURVES
PLATE II
“59-
0° T “;=-”%..40 1● “50
++
0 4$ ;-~ ~ ----
+●%* 1
●-ga _ &. i 402 . .. .. -...00 0
● .*#
● I
I
@ ● - ~__ ,-’40 ------- ------ ----
T
30
●%60 o-* l-- \o -- ,..-...—. ..
/
/ 20
4
‘N.1.. -.... ..!-— ,0w ------ ----- ---- ●Oe~”– - .-L -– -–%– ----
I ●#* o[mo O* o
‘Wm& Okoooo ● Q
-80 0 80 160 240 320 400
TEST TEMPERATURE ~ ‘F
FIG. IB: CHARPY V-NOTCH TRANSITION CURVES FORAS-RECEIVED “C” STEEL.
-80 0 ‘- 326 400TES;OTEM Pg;ATURg4~ ‘F
FIG. 26 (a-g): CHARPY V-NOTCH TRANSITION CURVESFOR “C” STEEL, SUBCRITICALLY HEATEDAT tlOO°F FOR TIMES INDICATED ANDAIR COOLED.
,: ,
1-
IL
0(.] ‘“-”* , 50
20 – 1/2 MINUTE ‘%:%0 .-*’ ** 40
#
40
1-”4”--2~ ~:”
%~ 30
60 y 0, 20
●480 !0
~o~100 ~- “- “ . 0\.
0(b)
40d
60! 9*
●480
0 (c)o
20 --10 HOURS
40 —o
60
80
2C —20 HOURS
40
60- —— —-–- ---- ; I ‘~-‘
w/
t0
100 “~ ~-80 0
TES? TEM&RATU#CCOF
FIG, 3B (o-f): CHARPY V-NOTCH TRANSITION CURVESFOR “C” STEEL. SUBCRITICALLY HEATEDAT 1200°F FOR TIMES INDICATED ANDAIR COOLED,
-f52-
“S91 ’14 ~A9H3N3 lf3VdWlIL
~“-”r– ! I I
. .mm
wii
O::ooogo :~oom cd— 0:
.-—.- ...——
~m
0 ●
“1#
Y im(
+’QTT4+i?$L1-t t-+<
●:K =Ulmz
0 8IA1-
0
‘-T\. mh
-1
—
o —% ~s’~\
50
Z* _(a) —●
●
-w - ● - -m : 40JO MINUTES P= ●
●
4C L m-/
30
1-= 60woE 80u
● ●
3& 40
a~ GO
20
/ ‘ o~ 80 .*
* to
$ loo _8 . .
m“k 2*- ‘: HouR.-O 0>%
—
●/ &
40 0 30
60-
20
80●
r
Io
* * 0100.
-80 0 80-o-o-o .0
160 240 320 400
vIL
FIG. 6B(a-e):
TEST TEMPERATURE+”F
Ct-JARPY V-NOTCH TRANSITION CURVESFOR “C” STEEL, SU8CRITICALLY HEATEDAT IIOO”F FOR TIMES lNDiCATED ANDWATER QUENCHED, AGED 1 MONTH ATROOM TEMPERATURE.
s
L
002mii
TEST TEMPERATUREQ”F
FIG. 6B (CONT.)
I
0 l-o(0)
50
AGED I MONTH.
40 - - 30
60●
80 J ‘ - 10
l%
●40
AGED 5 HOURS40 –
60 .1m
00, 8/ s10 *
9I 1I00 ●- / b,-m——a OE
0 (c)50 g
Zc ‘1/2 M[NUTE 40w
AGED 26 HOURS 1-40 - 30 ~
n
60 * z9 20 –
so ● * 10●
100
0 (d)~; :0
20 ‘1/2 MINUTE 40AGED 3 DAYS I
40 - - 30
60 20
m 4- 10●
100. 0
FIG. 7B(a-s):CHARPY V-NOTCH TRANSITION CURVES FOR“G” STEEL, SUBGRITIGALLY HEATED AT i200°F
uwa-
!
.
0“~ “-
(e)50
20 – 1/2 MINUTE!
40AGEO 7 OAY3 I
40 —
pp---- -
1-30
60 1 ---– 20.<*’ I
00& ●. --
+— —
,1----
..—~-. ,0
100. II
0 (f):0
ii20 — 1/2 MINUTE —1... ----- :_. . 40 vi
AGED 14 OAYS 1 m
40 ‘- f-A
130
F
60 % - 20 L
!00 A
● C 10 >m
t0: Ow
(a)”~o so z
0w
—o
iNTH
1
t
.=.20 J/2 M~NIJT[
AGED I M(
40Ii ~’ ~
60 ● - 20
d
so0
0 ●T 10
too ●
.-. ‘“‘y.. .-.. +
0 (h)5°0
t20 1/2 MINUTE
AGED 42 DAYS~*,-,,40
4C ~ 30!
60●
20
80 ● 10
100 111111/ I 10-60 0
TES+OTEMPf;ATUR%OF320 400
FIG. 78 (CONT.)
FOR TIMES INDICATED AND WATER QUENCHED,AGED AT ROOM TEMPERATURE.
4,.
‘((j) 1 I 1 1 1 50i+ I,,, !!
~ - t/2 MINUTEt,
-; !1 II 1 40
AGED 65 DAYS ;:
40: ‘:
, ! 30,~
! 1EC ●
1● 20
i:1 ●
eo - 10
I
(k)5%
0~ 20 ‘2 MINUTES
,0%040
z AGED I MONTH
w 4@ , 1
0uId GO :CL
!
L
60 ; *“
Iuu 100
1-
AGED I MONTHL 40
0I 1 r
in
~ 60 ;
o IE —cn~ ;IL / 10,: O\
1000 —8+4=K:*, “% ‘-0-0 —0 :
(m) %. 5:
20 – 1 HOUR~o 40
AGED ! MONTH‘ f
40 j!~ 30
60 -’ ,-— ~– 20.!
i80 ; 1 ! 10
11- >*f100 “ ●- 0
-60 0 60 160 240 320 400
TEST TEMPERATURE~°F
FIG. 76 (CONT.)
-66-
0CE1#
IL
●-l-l o-80 - 0 w 160 240 320
TEST TEMPERATURE G-%
FIG. 7B (CONT)
.
-L.
—.
-67-
INTRODUCTION
In the first stages of the investigation on the effect
of subcritical heat treatment on the transition temperature
of ~?(l~~steels both air and a nitrate salt bath were used as
the heating media. It was noticed that$ generally$ test
specimens heated in the salt had higher transition tempera-
tures and harnesses than specimens comparably heated in air.
This anomalous behavior was revealed in hotihthe Charpy V-
notch impact and the eccentric notch tensile data~ and led to
the supposition t-satthe salt &,asintroduc&ig an embriktling
agents whose effect was superimposed on the effect of sub-
critical heat treatment. Metallographlc examination chemical
analysesfl and x-rays showed the embrittlimg agent to be nitrogen,
Although not pertinent to the present investigations the
results obtained by subcritical heat treatment in a nitrate
salt bath are of sufficient interest to be reported here and
compared with those previously obtained after heat t~eating
in sire
MATERIAL AND PROCEDURE
The procedure for both the Charpy V-notch and eccentric
-68-
TABLE I
Transition Temperature and Harnesses ofCharpy V-Notch Specimens after Subcritical Heat
Treatment in Nitrate Salt Bath
Transition Temperatures ‘F50% Fibrous IS’Time at
Temperature ‘E Hardness Midpoint
W-i%As Received
73737373
92909288
950 “F,, Air Cool
858895
llp~
5 minutes1 hour10 hours122 hews
73737386 24.0
llOO°F, Furnac@ Cool
85 2-10
llOO°F, Water Quench+$
1.12 hours 248 197
4 1/’2hours20 hours112 hours
9499
101
136205295
168225308
13020029.2
t$Aged one month at room temperature before tes’cinga
-.
0~9‘o I
20 kO\
40(d
.30700”F, I HOUR
60 — AIR COOLED d
60 10
Alf? COOLED
60-
80*“
●4
‘% ~ 4 “+ .
20(c) C.9
40 —700°F, 30 HOURS _ ~.
AIR COOLEDz
60 ,2 Ou
80
100
20 ~~J 1 /40
40 —700°F, ! WEEKAIR COOLED
3 0
60 20
80 -i o
0-80
I (a-q):
oTEsT ?EMPE#A?IjRE ~wOF
320 400
CFIARPY V-NOTCH TRANSITION CURVES OF “C”STEEL AFTER VARIOUS SUBCRITICAL HEATTREATMENTS IN NITRATE SALT.
Oo 50
40
40 -950”F, 5 MIN.AIR COOLED
30
60 20
80 / ‘0
10
‘ -1”
\Q-
10:8 L&*
F\ . ●- ●* ,Wm 50●
20 (05 OOF ‘6 &\ 40
40 -AIR iOOLED
30
“
J
I
1 I L 20
~o*
20+ / “u~‘< ‘1*’*’”9
40 - 950°F, 24 I-IRS. 30AIR COOLED
60 1 1 , ● 20
20(h) ‘I I I = I i 40
950”F, I WEEK Y&”= %*40 -
AIR COOLED30
●“o
60#“
1 20
8) I ●- % 10
Iv ~!*A”-K3O* Mo_cL_Uo+30 0 240
TE5T~”TEMP&0ATURE — *F320 400
FIG. I (CONT.):
w
notch tensile specimens were the same as described in the
port proper~ with the exception that the subcritical heat
treatments were carried out in an immersed electrode type
re-
Of
salt batho The salt was a nitrate mixture of the following
composition: 40%, ~03 and 60%, NaN03.
A summary of the various heat treatments used and the
resulting transition temperatures and harnesses are given
in Tables I and II for the impact and the notch tensile
a specimens, respectively. All specimens were obtained from
the same large plate (Plate 11) ofltCl~steel.
RESULTS AND DISCUSSION
Transition Temperatures After Subcritical Heat Treatment
The individual impact transition curves for the various@
heat treatments In the nitrate salt bath are”shown in Fig. 1
(a-q), and the notch tensile transition curves in Figo 2 (a-g).
The same criteria of embrittlement--l~ ft-lb value for
impact specimens and midpoint value for notch tensile speci-
mens--have been employed as in the preceding reporte
Air Cooled--Fig~ 1 (a-m) and Fige 2 (a-g) show the
individual transition curves of impact and notch tensile tests~
respectively, after heating in the 700°--1100~F temperature
range for various times, employing an air COOla An examination
of the curves for both specime~ types shows that not only is
the transition range shifted to higher testing temperatures
m30KmE
o -- * I I 50
●.* I20 %% * 40
a
40 ‘J!OO”F 5 M[N
30
60 - AIR ;OOLEO ‘D’ 20
80 >c
o%0
20 (k) ‘b
I60
80
0>
60
80L
too4
m- , “k .ma ~01J
20 (m)40 IIOO”F, 112 HOURS
AIR COOLED
60 ● 20/ ‘
00●O
●@
100? x < ----- ‘:-60 0 80 160 240 320 400
TEST TEMPERATURE -“F
FIG. 1 (CONT):
o 50,
20 i 40
40 (n)30
llOO°F. 4-1/2 HOURS
T
20I60 WATER” QUENGHE-DAGEO I MONTH ATRT#~
80
R- %% % L , -1 ‘-
T
,r~– ‘- ---
t
10
100 P--0 5;’
‘“Hll---”t+H3Tttl+1401-Z 40 I1w (dIIOOOF, 20 HOURSo -1
K 60 -WATER QUENGf+ED
Id AGED I MONTH AT R.T.
3$ 20
a
E 40 - WATER QUENCHEOAGED I MONTH AT R.T.
m3 60
c)
% 80 “
E
o
20A .
3 I(@llOO°F, 112 HOURS
‘- -~40
40 – FURNACE COOLED~ .* 4 -t-l
●-4**–●>
11111111 -Jr1130
801 I I I I I L#
Y ‘1I yl \ I I I
20
IIIIJJ<100 I %&&j o-80 0 ao 160 240 320 400
TESTING TEMPERATURE%”F
FIG. 1 (CONT.)
‘--
I
1
I
I
I , 1 1(o)
100 – 950-F, 4 mum5
AIR COOLED
75
m0
z& 25
000Q (b)
/75
w
25
0 I
(c) I :
100 – 950”F, 30 HOURS ;AIR COOLED
75
50
25
0 .
(d)!00 – 9+30”F, I wEEK.
AIR COOLEO
75
w
~’ “25
0-140 -1oo -60 -20 100 140 I80
TEST TEMPER~TURE ‘OOF
FIG. 2: ECCENTRIC NOTCH TENSILE TRANSITION CURVES OF “G”
(a-9) STEEL AFTER VARIOUS SUBCRITICAL HEAT TREATMENTSIN NITRATE SALT.
125 ~e) I I I II IOCFF, I HOUR
100 - IAIR COOLED 1
75I
50
. -.i-
~–- 1
25I
,, :.
0
Icm –
A!R COOLED
75
50
2.58.- --- .—
t(g) ’’’’ 11 ii II iilit
100 llOO°F, 96 HOURS I I I 1J 1 F I I I
I +~+--+1AIR COOLED
,!
Ol!il I ~~i I I I 1-1oo -60 –20 20 60 I 00 !40 !s0 220 260 ?nc? ~dn
TEST TEMPERATURE-°F. . . .
,, i 7
-7’3-
with both increasing subcritical temperature and ‘time$but the
upper level is lowered as well. In addition,$ the hardness
correspondirugl-yincreases. This is in contrast to the air
heated specimens noted previously, which show the transition
range is relatively unaffected and the upper level increases
with increasing time and temperature.
This embrittling effect is more evident in Fig. 3, wherein
the transition temperature is plotted as a function of time
at the various temperatures for both specimen types and compared
with the results after comparable ‘neat treatments in air++.
In the impact test the transition temperature is constant with
time at least up to one week at 700°F. At 9~O°F the transition
temperature starts to rise after about 20 hours, and after one
week$ has increased 60°F. At llOODF an increase in transition
temperature is noted after about 20 hours~ increasing 10~°F
after 112 hours.
With the notch tensile test a comparison of the nitrate
salt and air heating media shows that the salt starts to
embrittle the steel after”about 2 hours at 9~O°F and 1/2
hour at 11OO”F. After one week at 9~0°F9 the transition tem-
perature has been raised about IOO”F; after 96 hours at llOO”Ffl
the transition temperature has been increased about l~O°F.
Thus , the two specimen types show that the magnitude of
$:Seepreceding report.
IL n,
o -fy-
200
r
I I I I
CHARPY V-NOTCH ,V180
160
140
I20
100
80
120 -ii I I f I I I I
ECCENTRIC NOTCH TENSILE #,100 {-
80 1/
60/v
A
/ {t
4
40 $v 1
.4’4
#
20 0°,1 .4/r--
./0
/’v
f
,/’
“20●
w950°F AIR
-40 I
\1 I \
AS RECEIVED
I IOO”F AIR-60
I
-800-i 0,1 I 10 100 1000
,
I
,.
I
I
I
I
..\
FIG* 3:
ISOTHERMAL TIME- HOURS
TRANSITION TEMPERATURES OF “C” STEEL RESULTING
FROM VARIOUS SUBCRITICAL HEAT TREATMENTS IN
NITRATE SALT BATH AND IN AIR, EMPLOYING AN
AIR COOL.
.— .—
-75-
the ernbrittlemen-tincrea~:s with both time and temperature.
l?urnace Cooled-- One series of impact specirn~nswas tested—
in the furnace cooled condition after heating at llOO°F for
112 hours. A comparison of the transition curve for this treat-
ment~ Fig~ lq~ with that of the same
an air Cool$ Fig. lmt shows the same
of th~ same magnitude. The hardness
specimens are embrittled to the same
heat treatment employing
embrittling effect and
checks also show that the
degree.
Water Quenched--The individual impact transition curves
obtained after heating at llOOQF for various times~ employing
a water quenc’h+are plotted in Figs. 1 (n-p). These curves
show that the transition range is shifted to higher testing
temperatures and
time at llOO°F.
ness with time.
the upper level lowered with increasing
This is also followed by an increasing hard-
l’ig~4 provides a comparison of the transition temperat-
ures after heat treatment in the nitrate salt with those
after comparable heat treatments in air. As in the air
cooled seriesq it is evident that the nitrate salt has
introduced some embrittling agent whose effect is now notice-
abl~ after about 4 hours at IIOODF. From an embrittlement
amounting to about 30°F rise in transition temperature at
this point, the embrittlement continually increases with
time, amounting to a 200DF increase in transition tempera-
ture after 112 hours.
——-
+Aged one month at room temperature
-m-
. .
lJ-0
Iw(z3
$.
rn-J
FIG.
32(
28(
24(
20(
16C
I20
80
11
CHARPY V-NOTCH
/I
I
IL4 I
s
Iv
I...— 1—.. — .— —
/
I/
v..—
~AS RECEIVED
{1o 0,1 I 10 100 1000
ISOTHERMAL TIME AT llOO°F,— H(XJRs
4: TRANSITION TEMPERATURES OF “C” STEEL RESULT-ING FROM HEATING AT llOO°F FOR VARIOUS TIMESIN NITRATE SALT BATH AND IN AIR, EMPLOYING
A WATER QUENCH. AGED ONE MONTH AT ROOMTEMPERATURE,
. .. .. .. . . . . —
-77-
Fig. ~ is a summary curve, comparing the Charpy transition
temperatures of the air cooled and water quenc’hed series from
llOO°F, using both nitrate salt and air as the heating media.
It is apparent that, for a given medium, the relation of
transition temperature ..isothe~l~al time for ~~e air cooled
series~-is displaced below and approximately parallel to that
of the water quenched series. For either the water quenched
or the air cooled series, there appears to be an incubation
period before the el’1’ectof the nitrate salt on the trafisition
temperature is noticeable, With increasing time the embrit-
tling action of the nitrate salt is evident in a continuous
increase in transition tem~rature at about the same rate
for either the air cooled or water quenched series.
In Fig. 6 the individual transition curves obtained
after heatin~ at llOOQF for 112 hours are compared for the
two heating media, employing both a water quench and an air
cool. For a given cooling rate, the extreme embrittlement
introduced by heat treatment in the nitrate salt is clearly
revealed by a s’bittof the transition range to higher testing
temperatures and a lowering of the maximum level.
Microstrwtures--In order to detect any str’~cturaldif-—
t’ereilceswhich would account for the anomalous results be-
t,weenspecimens heated in nitrate salt and in airs a number,.
-J
CHARPY V-N(
AC-AIR COOLWQ-WATER QUENCH
/1 I/-
f
o“ 0.1 1,0 10 100 1000
ISOTHERMAL TIME —HOURS
FIG, 5: CHARPY TRANSITION TEMP, – ISOTHERMAL TIME CURVES
AFTER HEAT TREATMENT AT llOO°F IN NITRATE
SALT BATH AND IN AIR,
I I I
CHARPY V-NOTCH
AC-AIR COOL
WQ-WATER QUENCH
-80 0 so 160 240 320 400
TEST TEMPERATURE—*F
FIG. 6: CHARPY TRANSITION CURVES AFTER HEAT TREATMENT
AT IIOO”F FOR 112 HOURS IN NITRATE SALT BATH ANDIN AIR.
..
I
I
.— —
-79=
of photomicrographs were made at 2000X~ Fig. 7$ showing the
structures after heating at llOOGF for 112 hours.
The air cooled and water quenched structures after
heat treatment in the salt are shown in Figs. 7b and c re-
spectively. The only difference evident between these
structures $ is the mottled appearance of the ferrite in the
water quenched specimen. A comparison of these structures
with the as-received platez Fig. 7a, indicates that slight
spheroidization is present after the subcritical heat
treatimenta
With a furnace Coql$ an unusual microstructure was
evident after the nitrate salt heat treatmdkt. Fig. 7d
taken at the center of a Charpy bar shows short plates of
a precipitate arranged in a Widmanstatten pattern. At the
edge? Fig. 7e9 a mixed precipitate of short plates and long
needle-like plates apparently nucleated at the grain
bo~n~arfe~ is e~id~nta This precipitate, then seems to be
associated with a critical cooling rate.
The two different forms of tileprecipitate suggest a
concentration gradient from the edge to the core, and
their form and distribution suggest nitride needles. This
structure is not present in the comparable series heat
treated in air and furnace cooled, Fig. 7.f,even though
‘tGt’steel has the highest original nitrogen content of the
project steels; thus, it would appear that t’henitrate salt
n
w
4
-c*4”
‘,,; ,’
,,,,, ,,
!,,,!
!,
,’!,, ,“!!
“’ ,,.!,
,.
!
. .
ZECl-l!+
I
bil%+k
I
I
~
,—
-81-
served to introduce nitrogen during subcritical heat treatment.
Electron micrographs prepared by Drt A. Revere at 6,000
and 12~000 diameters showed substantially the sm”e features
as the optical micrographs.
Nitrogen Analyses--To confirm that the specimens heated
in salt were enriched with nitrogen, selected specimens were
analyzed by both the wet method for combined nitrogen (as
nitrides) and the vacuum fusion method for total nitrogen
(as nitrides, as molecular nitrogen in holes~ and in solution)a
The results are tabulated below in Table III; each value is
the average nitrogen content of the cross section of a finished
Charpy specimen.
TABLE 111
Nitrogen Analyses of liCi?Steel AfterVarious Heat Treatments
Heat Combined TotalTreatment NitroRen++ Nitrogen+++ source
As-Received --- 0.0098 Bureau of Ships0.017 --- Commercial Laboratory
Heated in air at llOO°F o.ol~ --~ Commercial Laboratoryfor 112 hours and air --- 0.0104 Republic Steel Corp.cooled
Heated in nitrate salt at 0.226 --- Commercial LaboratoryllOO°F for 112 hours andfurnace cooled
Heated in nitrate salt --- 0.1120 Republic Steel Corp.at llOOPF for 112 hours 0.106 --- Commercial Laboratoryand water quenched
Wet Method++Vacuw fusion method
.— .— —.
=82.
Both methods of analysis show that the nitrogen content,,
of specimens mbcritically heated in air at llOO°F for 112
hm~ys and air cooled is essentially the same as the as-received
plate; however, with specimens comparably heat treated in the
nitrate salt, the nitrogen content is.greatly increased over
that of the as-received plate. The wet method indicates that
the combined nitrogen content al’the f&nace cool~d series is
increased by a factor of about sevens and of the water q-uenched
series by a factor of about six. The greater combined nitrogen
content of the furnace cooled specimens indicates that an ap-
preciable amount of nitrogen has been retained in solution in
the water quenched series. The total nitrogen content of the
water quenched series was increased by a factor of about eleven
after the nitrate salt treatment.
It should be noted that for both the as-received and air
heated conditions the combined nitrogen value is greater than
the comparable total nitrogen value. This is hardly possible
and$ undoubtedly~ the wet method employed for combined nitrogen
is in error. Therefore,, all values obtained by this method
have no numerical significance other than to show the relative
nitrogen contents.
X-Rays--In order to identify the structure of the nitrogen
compounds~ several x-ray patterns were made of specimens which
had been heat treated in the nitrate salt at llQO”F for 112
hours and furnace cooled. Samples about 0.020 inch square and
-83-
..,,
1/2 inch long were CU-5from the edge and the core of impact
specimens and Debye patterns made from both positions, using
,mcmocl-uromatic~ol$~radiation. Exposure times of 40 hours or
~1.great L],5’KJ?and ~rlawere required to show up the nitrogen
compounds .
The specimen from the core showed a weak pattern with
the presence of Fe N being indicated~ while the specimen fromh
the rim revealed a stronger pattern with both Fe3N and Fe N4
being indicated. The short plates in the microstructure can
now be labeled tileFel,Nphases and the long plates$ the
Fe N3
mens
4=
phasec.
Fatterns made of air cooled and also water quenched speci-
failed to disclose the presence of a new phase.
Source of Nitrogen--The increase in nitrogen content ai’ter
the heat
reasons:
creasing
trea-tment in nitrate salt su~gested two possible
(1) a preferential oxidation of the iron, thus de-
the amount of metal without the loss of any nitrogen,
and (2) an enrichment with nitrogen by a scaling reaction of
the sp~cimeilswith the salts allowing the nitrogen to diffuse
into the metal.
Reason 1 was considered possible in view of the fact that
after long times in the salt baths specimens were badly scaled.
TIM+ scale was m~~lti~layered and of a compressive nature. The
extent of metal loss was determined for impact specimens sub-
critlcally heated in salt and in air at lloooF”for 112 ;aours$.
— —J
-84-
by breaking off the scale, wire brushing and then determining .,
the cross sectional area 10ssO The loss in area after heat
treatment In air was 1/2 per cent, and ~ 1/2 per cent after
treatment in salt. Although the metal loss was high for the 1.’
.salt heat treatments a calculation shows that it ‘could only
account for an increase in nitiog~.~ from 0.0098 per cent to,’
0.010 per cent, if the nitrogen did not diffuse out; whereas,
the analysis showed a total nitrogen content of 00112 per cent.
Although this difference would seem to preclude’ Reason 1$I
attempts were made to duplicate the structure by different
heat treatments in an oxygen atmosphere. In this manner ,,.
nitrogen bearing media such as air and nitrate salt could be I
excluded. The heat treatments employed are listed in Table IV* .!
TABLE IV
Heat Treatments in Oxygen Atmosphere
Time-eratu~e m Cooling Rate
1830°F 21 hours Furnace Cool from llOO°F
18309F 21 hours Air Cool
.1
Cross SectionalArea Loss, Per Cent
18 l/2I
1
18
Al-though the scaling loss was greater for both specimens than
any of the salt heat treatments~ a microscopic examination
failed to reveal the presence of eit”her of the nitride phases,I
Both specimens were then wrapped in copper and soaked in
an air furnace at llOO°F for 112 hours and furnace cooled-
-85-
Again~ the microscopic evidence was negative.
A ~till further check was made by removing
impact specimens by vacuum heating to 2200°F in
nitrogen from
100°F steps
f~om 1600W after outgassing. The pressure at the start was
O.~ x 10-4mm but with each increase in temperature the pres-
sure built up to ~ to 6 times this initial value and then sub-
-4sequently decreased to 0.~--O07 x 10 mm HgO This was carried
out over a two day interval and the nitrogen content was then
assumed to be nil. Subsequent heat treatment at llOO°F for
112 hours in nitrate salt and in airs employing a furnace Coolfi
definitely showed the presence of the nitride structure in the
salt treated specimen~ Fig. 8a and not in the alr heated speci-
m.en$ Fig. 8bm
As a check on the contamination of the salt bath and also
on the possibility that the electrical circuit was contribut-
ing a catalytic effect, one as-received sample of ~tC1’steel
was placed in a small crucible containing used salts and
another in new salt. The crucibles were then placed in a
small hevi-duty furnace at IIOO°F for 112 hours, and furnace
cooled. The resulting microstructure of the two samples
showed the presence of nitride needles in the specimen heated
in the old salt but not in the specimen in the new salt. The
specimen in the new salt was then further heated at llOODF
for 96 more hours and furnace cooled. The microstructure now
revealed the presence of nitrides. Apparently the nitrate.,
,))
‘,,,,
‘)”,,,’, ,“:,,’
(a)
Fig. 8:
“,,I,*
(b)
MICROSTRUCTURES OF ‘°C” STEEL AT 2000X AFTER FOLLOWINGHEAT TREATMENTS:
~a ) 2200”F in vacuum and furnace cooled,followed by subcritical heat treat-ment at llOO°F in nitrate salt for112 hours and furnace cooled.
(b) Sme as (a] except subcritical heattreatment was carried out in air.
-.
-a7-
salt had decomposed to react with the steel and introduce
nitrogen.
All these experhents served to e~tablish that the en-
rielmnen-kwith nitrogen was brought about by a reaction of
the specimens with the nitrate salt, resulting in a diffusion
of nitrogen into the steel and not a nitrogen build-up by
prefere~ntial oxidation of’the iron.
SUMMARY
The obse~~vation of anomalous transition behavior between
‘fCf?steel subcritically heated in air and in nitrate salt led
to an investigation of the cause of enbrittlernent when this
steel is heat treated in nitrate salt.
The embrittlement was followed by both Charpy V-notch
impact and eccentric notch tensile tests after subcritical heat
treatment in the 700°--llOO°F range for various times. Three
different cooling rates were employed--air COO1, f~pn~ee Cool,
and water quenc’h.
The embrittlement was found to increase with subcritical
temperature time at temperature, and cooling rate after an
i~.itialincubation.period.
Netallographic studies, X-rays, and nitrogen analyses show
that the er,brittlin~ agent is nitrosen introduced by a sealing
reaction of the steel ~’ith t’henitrate salt used as tileheat-
ing rfletliumO
~,~eembrittlem,ent due to nitro~en pick up is superimposed on
any embrittlement resulting from subcritical heat treatment carried
out in al~a
—.
