Download Heat treatment of duralumin and more Study notes Thermal Analysis in PDF only on Docsity!
HEAT TREATMENT OF DURALUMIN
By P. D. Merica, R. G. Waltenberg, and H. Scott
CONTENTS (^) Page
I. Introduction 271
II. GDmposition and preparation of alloys (^273) III. Heat treatment and aging (^273)
- Effect^ of^ quenching^ temperature^278
- Effect of aging temperature 281
- Effect^ of^ temperature^ of^ quenching^ bath^285
- Effect^ of^ prior^ heating^ at^ quenching^ temperature^293
- Effect^ of^ preheating^ to^ 515°^ C^ before^ quenching^ from^ lower^ tem-
peratures 293
IV. Miscellaneous^ tests^293
- Density^ and^ dilatation^296
2. Electrical resistivity 297
V. Mechanism^ of^ hardening^ during^ aging^ after^ rapid^ cooling^299
- Structure of duralumin (^306)
- Analogy between the hardening of duralumin and that of steel ... (^310)
- Eutectic^ structure^ and^ influence^ of^ magnesium^311 VI. Conclusions relative to the manufacture and heat treatment of duralumin (^3 ) VII. Summary and conclusions (^315)
I. INTRODUCTION
The remarkable phenomena exhibited by the aluminum alloy
known as duralumin were discovered during the years 1 903-1 911
by A. Wilm^' ^^ and have been described by him and others. ^'^ ^'^ ^'^ ^
The unusual feature of this alloy is the fact, as was shown by
Wilm, that it can be hardened quite appreciably by quenching
from temperattu-es below its melting^ point followed by aging at
ordinary temperatures, which consists merely of allowing the
material (^) to stand at these temperatures. The hardness (^) is not produced by the quenching alone, but increases during the (^) period
* A. Wilm, Physical-Metallurgical investigations of aluminum alloys containing magnesium, Metallurgie
8, p. 225; 1911.
* A. Wilm, The hardening of light alimiinum alloys, Metallurgie, 8, p. 650.
» I,. M. Cohn, Duralumin, Verh. Z. Befordering des Gewerbefleisses, 89, p. 643; 1910.
* L. M. Cohn, Changes in the physical properties of aliuninimi and its alloys, with special reference to
duralumin, Elektrotechnik u. Maschinenbau, 31, p. 430; 1913.
6 L. M. Cohn, Duralumin, Elektrotechnik u. Maschinenbau, 30, pp. 809, 829; 1912.
- (^) P. D. Merica, Aluminum and its light alloys. Circular 76 of the Bureau of Standards, (^) 1918, also Chem.
and Met. Eng., 19, pp. 135, 200, 329, 587, 635, 729, 780; 1918.
272 Scientific Papers^ of the Bureau (^) of Standards [Voi.
of aging, which (^) may be (^) from one to three days. Cohn (see notes
3 and^ 5, p.^ 271) gives^ data^ showing^ the^ increase^ of^ hardness^ of
dm-alumin dm-ing aging, after quenching in water from about
450° C. Upon annealing the alloy so hardened by aging, (^) it is
softened exactly as is hardened steel.
The composition of this alloy usually varies within the following limits:
Per cent
Copper (^) 3-4-
Magnesium o. 4— i. o
Manganese (^) o — o. (^7) Aluminum (^) Balance
Iron (as impurities) o. 4— i
Silicon (as impurities)^ o. 3 — o. 6
Its density is about 2.8. It is used only in the forged or rolled
condition.
This alloy has been produced for some years commercially and
is in demand for the fabrication of parts for which both lightness
and (^) strength are required, such as for aircraft. (^) Its tensile strength
will average 50 000 to 60 000 pounds per square inch after appro-
priate heat treatment, such as that described by Wilm.
With the pmpose of ascertaining whether the heat treatment
described by him actually developed the best mechanical proper-
ties possible for dm-alumin, the authors undertook a study of the
effect of variation in heat-treatment conditions, that is, quenching
temperatin-e, aging^ temperatiu-e,^ etc.,^ upon^ these^ properties^ and,
in connection with another investigation,^ a study of the effect of
chemical composition upon them.
E. Blough^ had^ already^ called^ the^ attention^ of^ one^ of^ the^ authors
to the fact that the amount of hardening produced by heat treat-
ment was influenced quite markedly by the temperature from
which the material was quenched, a most interesting fact which
was not brought out by Wilm's published investigations, which
mentioned merely the effect of aging after quenching from one
temperature, in the neighborhood of 450° C.
An explanation was sought also for the mechanism of hardening
during aging of this alloy, and additional data were obtained
bearing upon this phase of the matter.
The experiments^ here^ described^ were^ carried^ out^ partly^ in^ the
laboratories of^ the^ Bureau^ of^ Standards^ and^ partly^ in^ cooperation
7 p. D. Merica, R. G. Waltenberg, and A. N. Finn, The tensile properties and resistance to corrosion of
rolled light alloys of aluminum and magnesitim with copper, with nickel, and with^ manganese,^ Techno-
logic Paper No. 132 of the Bureau of Standards, 1919.
274 Scientific^ Papers^ of the^ Bureau^ of Standards^ [Vol.^ IS
electric furnace,^ quenching^ in^ water,^ and^ aging^ at room or other
temperatures for^ different^ periods.^ The^ results of these tests are
given in Table 2.
TABLE 2,—^The Tensile Properties and Scleroscope Hardness of Rolled, of Annealed,
and of Heat-treated Aluminum-Copper-Magnesium Alloys
Number
As rolled
Sclero-
hardness
magni-
fying er
Tensile strength
Elonga-
tion in 2 inches
After annealing at 422" C
Sclero- scope
hardness
magni-
fying hammer
Tensile strength
Elonga-
tion in 2 inches
CI
C2.
C3.
C
CS.
C6.
C
C8,
C
CIO.
Cll.
C12.
Lbs./in. 49 000 48 400 48 600 49 600 25 800 23 600 23 600 34 900 35 700 34 000 38 400 38 600 (^37 ) 35 900 37 500 37 700 35 300 38 500 38 100 44 200 45 500 45 300 38 100 38 100 41 200 43 200 41 200 44 800 44 600 47 500 56 700 52 900 58 400 38 900 38 600
Per cent
Lbs./in. 33 000 33 100 32 700
Per cent
Merica, WcdtenbergA (^) Heat Treatment (^) of Duralumin (^275)
TABLE 2.—The Tensile^ Properties^ and^ Scleroscope^ Hardness^ of^ Rolled,^ of^ Annealed,
and of Heat-treated^ Aluminum-^ Copper-Magnesium^ Alloys—Continued
Hmnber
CI
01
cs
C
C
C
C
CIO
cu
C
After heat treatment consisting of quenching in water and aging
Quenched from 478° C
Aging
Aged
at 110°
Days
Aged
at 20°
Days
11 11 11
Scleroscope
hardness
magnifying hammer
Tensile strength
Lbs./in.a 36 870 37 080 36 260
16 830 16 510 16 510
28 020 25 810 25 440
29 300 28 910 30 280
33 220 32 580 31 930
31 050 33 790 31 640
42 350 42 530 42 350
46 400 47 030 48 900 47 650 31 790 30 450 30 070
38 030 37 630 38 430
50 450 48 950 51 740 50 880 38 330 38 730 35 910
Elonga- tion in 2 inches
P.ct.
Quenched from 510° C
Aging
Aged
at 110°
Days
Aged
at 20°
Days
1
Scleroscope
hardness Tensile
magnifying
hammer
strength
Lbs./in. 17 38 030 37 220 27 48 120 47 210 16 670
8
Elonga- tion in 2 inches
P.ct.
T5.
Merica, Wallenberg,! Scott (^) J Heat^ Treatment^ of^ Duralumin^277
All of the^ alloys^ except^ those^ containing^ no^ copper (Nos. C
C4, and^ C9)^ show^ an^ increase^ of^ hardness^ of^ the^ hest-treated^ speci-
mens over that of the annealed samples. The increase of hardness
in those alloys containing copper, but no magnesitmi, is smaller
than that in those containing both, but is quite definite. This is
shown in the following table
Number of alloy
Increase of ten- sile strength of heat-treated
alloy (510° C)
ovei annealed
alloy
Alloys containing no copper:
C ...
Per cent
C4 -f 3
C9 —^4
Alloys containing copper but no magnesium:
C3 +
C5.- +
Alloya containing both copper and magnesium:
CI 45
Cll 56
C12. 110
It is noticed that the best mechanical^ properties are produced
by quenching from the higher^ temperatures^ (500 to 525° C). This is shown in^ Table^ 3, giving^ further^ data^ on^ two^ alloys,^ C8 and Cii, and^ will^ be shown^ more^ clearly^ below. 121040°— 19 2
278 Scientific^ Papers^ of the^ Bureau^ of Standards^ [Voi. 15
TABLE 3.—Effect of Quenching Temperature on Tensile Properties
Alloy C8 Alloy Cll
Quenching
Aging Mechanical properties Aging Mechanical properties
temperature
20° C 110° C
Sclero- scope
hard-
ness
Ultimate tensile strength
Elonga-
tion in 2 inches
20° (^) C 110° (^) C
Sclero- scope
hard-
ness
Ultimate tensile strength
Elonga-
tion in 2 inches
37r (^) C...
Days
7 7 13 13 7 7 13 13 7 7 11 11 [ ^^ 11 8 [ 8 [ 11 11 { 11 I 11
( 1 '' 11 8 [ 8
I ^^ 1 13
Days
11
12
}"
22
....
Lbs./in. 27 260 1 26 220 J^29 1 29 130 J^39 1 40 160 J
J^46
J^48
J 47 230
f 44 130 1 44 910
J
I 44
J^46
J^53
I 53 000
J^34
Per ct.
Days
7 7 13 13 7 7 13 13 7 7 11 11 11 11 8 8
Days
I 25
1 ^^
\ 28
1 ^^
Lbs./in. { 35 900 I 36 550 J^35 1 35 020 J
J
j 50 450 1 48 950 J^51 1 50 880 J^52 1 52 590 J^51 1 50 870 J
Per ct.
422° C...
478° C
500° (^) C
510° C...
1 520° C .
525° (^) C
533«»C
r 48 370 t 47 060
Not only does the hardness increase after heat treatment, but so also does the ductility, as evidenced by the elongation in the tensile test. This is shown in Tables 2 and (^) 3.
- EFFECT OF QUENCHING TEMPERATURE
In Fig. I are shown the scleroscope hardness values of Cii quenched in water (20° C) from different temperatures and aged
at room^ temperature^ for^ periods^ of^ time^ from a^ few^ hours^ to^30 days. The form of these aging curves is similar to that shown by Cohn (see notes 3 and 5 on (^) p. 271) ; that is, the hardness increases
after quenching,^ at^ first^ rapidly^ and^ then^ more^ slowly.^ It^ is^ fur-
28o Scientific Papers (^) of the Bureau (^) of Standards' [Vol. IS
ther evident that the maximum hardness attained increases with
the temperatm*e up to approximately 520° C.
The effect of quenching temperature is also shown very nicely
in an experiment of which the results are shown in Fig. 2. Two
strips of 0.087-inch sheet of alloy N34 were used. The strip (^) was
placed in the furnace for heating in such a manner that a nearly
linear temperature gradient^ existed^ between^ the^ two ends,^ as
Ift ^ "O ^ Ifj "^ (^) ^
shown by thermocouples placed along the strip. Upon attaining
the desired range of temperattires, the strip was quenched in boiling water and aged 20 hours at 110°^ C. The scleroscope hardness was then determined along the axis of the strip, and is shown in Fig. 2 as a function of the distance from one end of the sample. The distance may be regarded as a rough temperature
M^ka.waitenberg.-j //^^^ Treatment
of Duralumin^281
scale, the outside temperature^ limits^ having^ been^ determined^ and
marked on^ the^ curve.^ One^ strip^ was^ quenched^ when^ the^ two
ends were^ at 520 and^ 280°^ C,^ respectively;^ the^ other,^ when^ the
ends were at 490 and 210°^ C, respectively. Beginning at about
300° C, the effect of increased quenching temperature, other fac-
tors remaining alike, is to increase the hardness after aging until
a temperatiu'e of about 520° C is reached. Beyond that temper-
ature the hardness again decreases; the material becomes covered
with a dark gray oxide coating and generally also with blisters,
marking the temperature of eutectic melting. The effect of heat-
ing to temperatures around 300°^ C is chiefly to anneal the specimen
and to give lower values of the hardness (minimum^ on the curve)
than is given by heating^ at lower^ temperatures.
- EFFECT OF AGING TEMPERATURE
In Table 4 are given^ results of tests showing the effect of tem-
perature of quenching^ bath and^ of aging^ carried out in the bath.
The samples used were strips of (^) A1-12 quenched from 520° (^) C.
The increase of strength witli time of aging (^) is evident.
Merica, Waltenberg,! ScoU J Heat^ Treatment^ of^ Duralumin^283
A more complete^ picture^ of^ the^ phenomenon^ of^ hardening^ by
aging at different^ temperatures^ is^ obtained^ from^ Figs.^ 3, 4, and^ 5,
based upon^ data^ obtained^ on^ specimens^ of^ N34.^ The^ sclerocsope
values of^ Fig.^3 were^ obtained^ upon^ samples^ quenched^ in^ boiling
water from^ two^ temperatures,^515 and^ 525°^ C,^ and^ aged^ at^ differ-
ent temperatiu-es.^ The^ same^ figures^ are^ replotted^ in^ Fig.^4 in
different form.
T/fr>f af /Jyin^ //> JJayj.
Fig. 3. Effect of aging at different temperatures on the^ scleroscope^ hardness^ of samples
quenched from 51 5°^ C and 52 5° C. {Alloy N-34)
It is noted (i) that the rate of hardening increases as the tem-
perature of aging (^) increases, (2) that the maximum hardness (^) is obtained by aging at temperattu-es above 100° (^) C, and (^) (3) that at aging temperatures above 140° C the hardness eventually drops after reaching its maximum.
284 Scientific Papers (^) of the Bureau (^) of Standards [Vol. (^) IS
'^ifcuudc?^ Bu/Aj/u^£:?/^ -^ S'^^c/^^i:^//^ »^<?:?s^o^^/:?^
286 Scientific Papers (^) of the Bureau (^) of Standards [Voi.is
effect of the temperature of the quenching bath indicated in these
results. Those samples quenched to 150° C gave practically the
same results as those quenched to 230° C, although there is a
slight improvement in the tensile properties of those quenched to
150° (^) C over those quenched to 100°^ C.
In Table^6 are shown results^ of^ tests^ to^ determine^ the^ effect^ of
aging at room temperature after aging^ at the temperature of the
quenching bath. It will be noted^ that^ there^ is^ only^ a slight
increase in the strength of the alloy produced by aging at 20°^ C
after aging at the temperature of the quenching bath.
Merica, Waltenberg, Scott (^) ] Heat^ Treatment^ of^ Duralumin^287
g Q Pi , « , «iO ^g >ao U <J »o .0^ t>k^ m
1 1 Jr. tS| a
•« a
n .S
. o
i
)
1
^ ^ KJ {Q 1 S (^) eg : ^ ^
II
Per
cent 23.5 26.0 21.5 26.0 22.5 (^) 22. 53 • 53 S
•St
il 52 rt^ CM^ to^ TJ-^ Tf^ to
3 ^^ ^^ ^^ *^ *^ *
^CO st^ •<1- CO
il
2
I
Is
II
C3\ t«. CO CO CO (^) t^ CO CO CO
II
S
§ « to^10 to
CS? to ^ .0 to
OtotOOOiOOO ;o;HvgvgvocMvo^
Ig
Si
42
950 43
600 49
300 50
450 49
750 48
(^550) §1111 |S|
(^1) 6"
g
Ij
J3 53 2 JQ
o J5 (^53) CO CO
«
II
4§§ ^55?^ i^ 1 Z^ s?^ S
.a
1
Is il g^ S^ {Q^ S ^^ S^ S^ s^ ?5^ s^ J?^ ^^ s
II (^) goo to cd 5^ 2 53 53
to to 10 ;5 S 53 5^ 53 5^
1 —^
45
750 45
025 44
900 44
450 44
550 44
700 47
100 45
950 47
000 47
000 48
250
i3«
::i^ 2 ^ ^ ^ CO CO 0^ CJ^ ? 00 s s
Merda, Waltenberg,
Stott ]^ Heat^ Treatment^ of^ Duralumin^289
5S 55 S ^
irt a> o^ 06
47
600
b
47
100 50
400 50
750
:? ^
CO CO CO CO a a
n 10 Tl^ CM COCM (^) CM-^ CMCM a
51
450 48
500
a
54
600
a
53
850
S ^
^ (^) :
00 c^
CM CM (^) S s
: 5?
^ ^ « a
«MCM COcq
10
a
CM a
CM CM (^) s s
290 Scientific Papers (^) of the^ Bureau^ of Standards^ \V0L
Ir, JS !^ S> iTi Q"
