- Email: [email protected]
The effect of homogenization and precipitation treatments on the extrudability and ambient-temperature mechanical properties of aluminum alloy AA2024
Materials Science and Engineering, A 169 (1993) 9-17
9
The effect of homogenization and precipitation treatments on the extrudability and ambient-temperature mechanical properties of aluminum alloy AA2024 M. E. Kassner and X. Li Department of Mechanical Engineering and Center for Advanced Materials Research, Oregon State University, Corvallis, OR 97331 (USA)
H. J. McQueen Department of Mechanical Engineering, Concordia University, Montreal, Que. (Canada) (Received September 17, 1992; in revised form February 1, 1993)
Abstract This study determined the independent effects of various homogenization cycles and precipitation treatments on the elevated temperature workability and the final ambient temperature mechanical properties of aluminum alloy AA2024, The elevated-temperature tensile and extrusion tests indicate that the workability of AA2024 improves with elevatedtemperature precipitation treatments as suggested by earlier investigations. The precipitation treatments do not appear to degrade the ambient-temperature T3 and T8 tensile properties. The time at the precipitation temperature appears to affect the T3 and T8 tensile properties in homogenized ingot, longer times especially providing both relatively high ambient-temperature strength and ductility. The time at the standard homogenization temperature and the heat-up and cool-down rates do not dramatically affect the T3 tensile properties of homogenized ingot. However, long soak times (24 h) at the homogenization temperature and more rapid cooling rates (120 °C h-~) may improve the properties somewhat.
I. Introduction
In a previous study [1, 2] which included one of the coauthors of this article, the elevated-temperature torsional behavior of aluminum alloy AA2024 [3] was determined at relatively high strain rates. These tests were performed subsequent to both homogenization and precipitation treatments at various temperatures. The purpose of the work was to determine thermal pretreatments that led to optimal extrudability which was presumed to occur where the elevated-temperature stress vs. strain behavior indicated lower peak stresses and higher ductility, the former implying lower break-through pressures. This would increase the limiting extrusion ratio of a given press and increase the permissible extrusion velocity [2, 4-7]. These investigators [1, 2] performed torsion tests at 310 and 370 °C at a strain-rate g of 1 s ~ on A A 2 0 2 4 alloy, homogenized between 420 and 520 °C, and either water quenched or air cooled and subsequently aged at 290 or 350 °C. The stress vs. strain curves of the alloy in the quenched condition exhibited relatively large peak stresses Op and work softening with relatively 0921-5093/93/$6.00
low strain to failure. The peak stress and work softening were reduced and the ductility improved as either the aging or precipitation temperature or the deformation temperature was increased; the behavior became closer to that of pure aluminum in which a high level of dynamic recovery provides excellent workability [6]. The homogenization temperature had a large effect on Op but little effect on the ductility for deformation at 310 °C, whereas at 370 °C there was little effect on op but a large effect on ductility which was much higher. These results suggest that suitable homogenization and precipitation treatments can lower the elevatedtemperature flow stress, notably decreasing the initial high peak stress. The authors concluded that the decrease in flow stress and increase in ductility subsequent to precipitation treatments arise from the relative amounts of solute and dynamic precipitation (of presumably CuAI2 and (Cu, Mg)A12 [2]). such a reduction in peak stress and increase in ductility would seem to be beneficial for direct extrusion of the alloy. In review papers McQueen and Evangelista [8, 9] showed that similar effects to the above had been found in other age-hardenable aluminum alloys. © 1993 - Elsevier Sequoia. All rights reserved
10
M. E. Kassner et al. /
Extrudability and mechanical properties of Al alloy
The present study intended to verify the previous investigation [1, 2] and expand the experimental scope. Basically, the intent of this study was also to determine the effect of different precipitation treatments on the hot workability of aluminum alloy AA2024. Instead of torsion tests, the present study emphasized elevatedtemperature tension which is important at the periphery as metal passes over the die land in the extrusion process. This study also considered (1) additional precipitation temperatures, (2) the effect of the heating rate to the homogenization temperature, (3) the time at homogenization temperature, (4) the rate of cooling from the homogenization temperature, and (5) the time at the precipitation temperature, to find their effects on the T8 and T3 (ambient-temperature) tensile properties of AA2024. Furthermore, (6) the "laboratory" conditions that predict favorable extrudability were tested by extruding AA2024 directly, and (7) by determining whether adequate ambient-temperature properties are evident subsequent to extrusion. Therefore, this work may be considered as a logical extension of the earlier work presented in this journal.
2. Experimental procedure This study utilized AA2024 aluminum provided in the form of direct chill cast ingots of 83 mm and 178 mm diameters. The compositions of the ingots tested in this study were within the specification given in Table 1. Tensile tests were performed on an Instron 4505 screw driven tensile machine with computerized data acquisition. Specimen geometries varied, and the typical gage dimensions for ingot characterization were 5.1 mm diameter and 25.4 mm length. Specimens from extrusions had 2.8 mm diameter and 20 mm or 11.4 mm gage length. Specimens were evaluated from random positions within the ingots; it was determined that the mechanical properties were independent of the position. Elevated-temperature tests were performed utilizing a three-zone furnace with Eurotherm controllers and power supplies. Strain rates varied between 0.67 × 1 0 - 3 and 3.0 × 1 0 -3 s -1. The solution annealing temperature was always 493 °C. The temperature of the specimens was controlled to within + 2 °C of the set temperature during solution annealing. A 10 min heat-up was required to achieve the solution anneal temperature once specimens were inserted into the
furnace. Generally, solution annealed tensile specimens were maintained at temperature for 1 h. The ductility was measured as the engineering strain to failure (El) equal to A L / L o where L 0 is the initial length and, for high-temperature tests, also as a reduction in area (RA) equal to (A 0 - A f ) / A o where A 0 is the initial area and A f is the final area. The yield and ultimate tensile stresses are reported as engineering stresses. The T3 treatment consists of a solution anneal followed by a water quench, refrigeration for 1-3 h, "stretch" of 1%-2.5%, and an ambient temperature age. The "stretch" increases the density of dislocations which act as heterogeneous nucleation sites, presumably for CuAI 2 [1, 2], and increases the strength over the undeformed and aged (T4) alloy. More dramatic (T8) strengthening can be achieved if the alloy is aged artificially at an elevated temperature subsequent to the stretching to attain an optimum precipitate size. The reported yield stress is based on a 0.002 plastic strain offset. The plastic "stretch" varied between 1% and 2.5% plastic strain for T3 and T8 treatments. The ambient temperature age for T3 treatments varied from 2.5 to 19 days. The difference in T3 tensile properties over this range was not substantial: only 4 MPa (roughly 1%) for the yield stress, 6 MPa for the ultimate tensile strength (also 1%), and 1.5% El (factor of 1.1) [10], consistent with other work [11]. The T8 temperature and time-at-temperature were 190 °C and 12 h respectively. Specimens were kept in the freezing compartment of a refrigerator subsequent to the solution anneals (1-3 h) and prior to the stretch to suppress precipitation. The time from the refrigerator to the T8 temperature was always less than 1/4 h.
3. Results and discussion 3.1. Elevated-temperature tests of precipitation treated alloy A set of tensile experiments was performed to compare with the earlier work [2, 3] by McQueen and coworkers. Specimens were extracted from 83 mm diameter homogenized ingots. Homogenized specimens received precipitation treatments at 270, 330, and 350 °C for 1, 6 and 12 h. This resulted in precipitation of CuAI 2 and (CuMg)A12, based on the microscopy performed by the earlier work that was performed with one of our coauthors (HJMcQ). This
TABLE 1. Specificationfor the chemicalcompositionof alloy AA2024 used in this study Limit
Si
Fe
Cu
Mn
Mg
Cr
Zn
Ti
V
Be
Minimum(wt.%) Maximum (wt.%)
0.15
0.30
4.15 4.35
0.50 0.70
1.30 1.50
0.05
0.18
0.05
0.01
0.001 0.005
M. E. Kassner et al.
/
Extrudability and mechanical properties of Al alloy
1l
TABLE 2. The elevated temperature mechanical properties of"precipitation-treated" AA2024 specimens from homogenized 83 mm diameter ingot (g = 10 -3 s- E) Test temperature (°C)
Precipitation temperature (°C)
Precipitation time (h)
Oy (MPa)
UTS (MPa)
El (%)
RA (%)
350 350 350 350 350 350 350 350
270 330 330 330 350 350 350 0
12 1 6 12 1 6 12 --
69.6 75.8 79.6 66.2 76.5 69.6 62.1 76.5
84.1 89.6 91.7 82.0 89.6 84.1 79.3 89.6
29 33.3 36.3 35.3 35.4 36.6 33.4 30.3
86 89 88 79 89 88 88 84
100"
90"
80"
~
70"
~60-
-----
350 o C - l h r - 350 °C-6hrs ....... 350 °C-12hrs . . . . . . . . . . 270 °C- 12hrs . . . . . . . . . untreated
50"
4(J
-
0.00
i
,
0.02
=
•
0.04
i
-
0.06
i
0.08
0.10
Strain
Fig. 1. The engineering tensile stress vs. engineering strain behavior of AA2024 aluminum alloy at 350 °C subsequent to various elevated-temperature precipitation treatments. Specimens were extracted from 83 mm diameter ingots (g= 10 -3 s- ~).
- 40 a=
' 20 0
270
330
330
330
Precipitation Temperature
350
350
350
°C
Fig. 2. Bar-graph illustration of the data of Table 2.
was followed by elevated-temperature testing at 350 °C (a common extrusion temperature). From Table 2 and Figs. 1 and 2 it appears, based on the 350 °C tests on specimens that were not precipitation treated, that prolonged precipitation treatments decrease the tensile
strength and increase the ductility. More specifically, the table and figures indicate that precipitation treatments between 330 and 350 °C for 12 h reduce the yield stress and ultimate tensile strength up to 14% and increase the tensile ductility by a factor of up to 1.3. These tensile results are qualitatively consistent with the earlier work [2, 3] performed by torsion testing. Some scatter in the data may arise from the procedure of heating to the testing temperature. This heating was done as quickly as possible (30 min for our furnace) to "simulate" (rapid) induction heating which is frequently utilized in extrusion processes. As a consequence, some non-reproducible temperature gradients existed along the specimen axis (e.g. 20 °C variation). The error in stress measurement is about + 3 MPa. This may explain some of the anomalous results (e.g. 6 h 330 °C). 3.2. Elevated-temperature extrusions of the precipitation treated alloys Although the precipitation treatments appear to increase workability (specifically extrudability) based on simple mechanical tests, a more important test is extrusion testing. The extrusion cross-sections are shown in Figs. 3 and 4. In the first series of tests, 178 mm diameter ingots were extruded into L-shaped beams (air cooled) of 5.1 mm thickness, and 63.5 mm leg width (R = 38.4) as shown in Fig. 3. Ingots were extruded after various commercially relevant thermal processes, some of which were believed to cause precipitation. The treatments included the following: (1) a fan-cool (FC) with a cooling rate of 3 1 5 - 3 7 0 °C h -j which followed a 6 h homogenization (482499 °C with 8-10 h linear heating to the homogenization temperature); (2) oven-cool (OC), rather than FC, following an identical homogenization cycle as in (1) except a reheat to 504 °C followed by a 12 h (40 °C h -1) cool to ambient temperature; (3) an oven-stepcool (OSC), where subsequent to the homogenization
12
M. E. Kassner et aL
/
Extrudability and mechanical properties orAl alloy
in (2) the ingot is cooled to 349 °C (a "precipitation" temperature) in 4 h and maintained for 6 h before an 8 h linear cooling to ambient temperature. The heat treatments are shown schematically in Fig. 3. As expected, it appeared that the oven-step-cool ingots could be most easily extruded (billet tempera30
6O0/ J • yieldStrength
550 D %EL 20 -15 •
o~
-lO
Te I
T3
I
(a)
Ac
"~3
~8
T3 t
FC
Tj8
T~
oc
T8I OSC
"L" specimen
8 hr 8-10 h
~
/ 2 0 °C
I 20 °C
FC 1 hr
FC
+
8-10
hr~f-~
TABLE 3. T3 and T8 properties of specimensextracted from 178 mm diameter ingots, extruded after various thermal treatments
OC 1 hr 504 oC
FC
+
8-10
~
4 hr 6
"~ hr
349 °C'
20oc
(b)
ture at 316-343 °C after 10-15 min induction heating) in terms of the maximum possible extrusion speed of the ingot with acceptable surface quality. With this criterion, it is expected that softer, perhaps more ductile alloy, would be most extrudable. The extrusions of fan-cooled ingots showed tearing at the surface at 7.3 m min-1 while oven-step-cool extrusion did not give evidence of tearing until about 8.8 m min -1. Extrusions of oven-cooled ingots (probably some precipitation since the slow cool allowed for a period, albeit small, within the precipitation range) did not show tearing at 8.5 m min- 1 but showed tearing at 9.1 m min- 1. This is in agreement With improved extrudability of billets with prior precipitation treatment observed in 2024 [7] and in AI-Mg-Si alloys [6, 12]. Table 3 and the associated bar-graph, Fig. 3, list the tensile test results of T3 (2.5% stretch, 19 days) and T8 specimens extracted parallel to the axis of the extrusion at random locations. The mechanical properties listed are an average of five tests. Table '3 and Fig. 3 show that the "final" T3 and T8 properties are essentially identical for the three "pretreatments". Therefore, there is encouraging evidence that the oven-step-cool pretreatment can increase the extrudability and yet not compromise the critical T3 and T8 mechanical properties (at least for this starting ingot dimension, extrusion temperature, and final configuration). Experiments analogous to the above were performed on precipitation-treated ingots of a smaller diameter (83 mm) and extruded into a simple crosssectional configuration of 7.9 mm by 32.5 mm with "rounded" corners (R=21.1) shown in Fig. 4 as S specimens. The oven-cooled (OC) ingots were heated linearly to the homogenization temperature of 504 °C within 12 h, maintained at 504 °C for 4 h, and cooled linearly to ambient temperature within 12 h (40 °C h-l). The oven-step-cooled (OSC) ingots were also maintained at 504 °C for 4 h (after a 12 h heating) and cooled linearly to 349 °C (the same "precipitation" temperature as for the L-shaped extrusions) within
8 hr 20o0
OSC
Fig. 3. (a) Bar-graph illustration of the data of Table 3 and (b) schematicdiagramsof the three heat treatments.
Cooling method
Treatment
ay (MPa)
UTS (MPa)
El (%)
FC OC OSC FC OC OSC
T3 T3 T3 T8 T8 T8
359 340 341 474 473 472
445 420 418 495 483 482
13.6 15.1 15.5 6.3 4.4 4.6
FC, furnace cool; OC, oven cool; OSC, oven-step-cool.
/
M. E. Kassner et al.
Extrudability and mechanical properties oral alloy
1.5 h, maintained at 349 °C for 4 h, and cooled to 93 °C in 4 h before removal from the furnace. The ingots were induction heated to a temperature between 370 and 425 °C immediately before extrusion (about 50 °C higher than the previous (L) extrusion temperature). The induction heating of the extrusion required about 15 min. The heat treatments are shown schematically in Fig. 4. It was found that the OSC billets could be processed at a greater extrusion speed without surface defects than conventionally homogenized 2024 (although the quantitative rates were not reported). Table 4 and Fig. 4 list averages of three ambient temperature tensile tests for each cooling cycle on extrusions and homogenized as-cast ingots which provide a reference. First, the T3 (2.5% stretch, 19
3O
• YieldSlrength El Lrrs El %EL
25
~ 450"~
5 ~o~
4o0.
o
350. .
,o
7
(a)
T8 I
7
AC
,.
.
.
oc
.
t
.
J
osc
"S" Specimen 32.5 mm ~ __~.7.9mm
C
~
)T
13
days) and T8 mechanical properties of homogenized, unextruded ("as cast") ingots were determined. The homogenization included a 9 h linear heating to the homogenization temperature of 504 °C, maintenance at temperature for 6 h, and cooling at 150 °C h- ~ to ambient temperature. The variation of the test results was very small, _+ 1%. Since these ingots were not extruded, the tests do not reflect production properties. The strength values (average of six tests) are slightly higher than the typical AA2024 T3 (2.25% stretch) and T8 values and the ductility is slightly lower [3, 13]. These provide "baseline" or reference values with which the T3 and T8 properties of extrusions can be compared. The T3 and T8 properties of the OC and OSC extrusions were well above average. We note that the previous L extrusions have lower strength than both the baseline (non-extruded or homogenized) material and extrusions from 83 mm ingots, indicating the variability of properties resulting from different ingot sizes and extrusion processes. The correlation between high ductility in tension and in extrusion seems reasonable. Previous work on 2024 [5, 7] does not confirm this; material cooled from the homogenization temperature to the billet insert temperature had higher torsional strength, lower ductility, and higher breakout pressure, but higher maximum extrusion velocity than furnace cooled material. Part of the problem lies in the lack of precise knowledge of the distribution and concentration of constituent particles and of precipitates in both the present and previous materials. Another part lies in the extrusion process in which there is a rise of 50-100 °C in average temperature in the deformation zone, being larger for higher billet stresses (lower insert temperature, higher alloying or particle content) and larger extrusion ratios and ram speeds [4, 5]. However, there
4 hr
TABLE 4. T3 and T8 properties of specimens extracted from 83 mm diameter AA2024 ingots,extruded subsequent to various thermal treatments
2 hr 20C
Cooling Treatment method
oy UTS El (%) RT age time (MPa) (MPa) (Days)
OC OC OSC OSC AC OC OC OSC OSC AC
431 432 409 408 380 492 498 490 496 484
OC
4 hr
/
C--~ .5 hr
\,.r
12/
349oc 90 ~oC~hr
/ 20°(3 (b)
\
20°C
OSC
Fig. 4. (a) Bar-graph illustration of the data of Table 4 and (b) schematic diagrams of the three heat treatments.
T3 T3 T3 T3 T3 T8 T8 T8 T8 T8
543 541 508 513 464 523 529 518 525 512
13.7 13.0 13.8 15.5 10.8 9.9 9.8 8.8 9.7 4.7
12.5 19 12.5 19 19 ------
OC, oven cool; OSC, oven-step-cool; AC, as-cast; RT, roomtemperature.
14
M. E. Kassner et aL / Extrudability and mechanical properties ofAl alloy
is an additional temperature increase from the average to that of the periphery as it passes near the die land. A detailed study of AI-Mg-Si alloys was performed by Reiso [14, 15]. In slowly cooled billets large Mg2Si precipitates form on constituent particles and do not dissolve on preheating or in the deformation zone but form a eutectic with the constituent particles which melt at the die land, causing tearing. In billets cooled rapidly to an intermediate temperature, precipitates are smaller and more homgeneous with the result that they redissolve in preheating and deformation so that no eutectic is formed and the tearing temperature is close to the true solidus of the alloy at a higher ram speed. This theory may be applicable to the present 2024 alloys but the distribution and nature of constituent particles and location or size of precipitates in different cooling cycles have not been determined. A comparison between A I - C u - M g and AI-Mg-Si alloys is relevant. In both alloys (although one can take the composition, as defined by the principal elements and determine the solidus temperature) melting occurs below the solidus owing to the presence of Fe-Al silicates. The presence of these silicates with the other elements can lead to melting below the solidus. All research suggests that proper control of the precipitation produces lower flow stresses at 3 0 0 - 4 0 0 °C and lower extrusion breakout pressures [1, 2, 5-7, 9, 12, 14-17].
3.3. Effect of duration of the precipitation treatment on the ambient-temperature mechanical properties Another part of this study consisted in determining the effect of precipitation time on the T3 mechanical properties. The homogenization cycle consisted of linear heating from ambient temperature to 504 °C within about 8 h. The specimens (of about 12.7 mm x 12.7 mm x 152 mm cut from an 83 mm diameter ingot prior to homogenization) were homogenized for 6 h at 504 °C. The samples were then step cooled by a 3 h linear cooling to 350 °C, the precipitation temperature, and maintained for various times up to 72 h. Subsequently the samples were removed from the furnace and air cooled. The T3 properties were then assessed after an age of 12.5 days. The results illustrated in Fig. 5 show that, for all precipitation treatment times up to 72 h (limit considered commercially practical), higher yield stress and ultimate tensile strength are associated with higher ductility (all specimens were solution annealed for 1 h subsequent to the precipitation treatments). Relatively favorable T3 properties are observed with very prolonged precipitation times (e.g. greater than 40 h), as well as at times of about 20 h, and times up to about 6 h. Lower (but not unacceptable) T3 properties are observed for precipitation times between 8 and 12 h.
480
18
'~...~
~!'.'.~.~._..
~6
450 .g
14
.
420
12
"~ 39o .
.
.
.
.
.
.
.
.
-'=~ 3SO0 [-~ 30' '
.
-
.................
~
_- . . . . . .
Yield Stress
t 8
/
f~
t--~l
I'2 300
, 0
, I0
. . . . 20
30
, 40
•
, 50
•
, 60
-
, 70
•
tO 80
Precipitation Time (hrs) at 350°C
Fig. 5. The relationship between the precipitation time at 350 °C and the T3 (ambient temperature) ultimate tensile strength, the yield strength, and the tensile strain in AA2024 specimens extracted from as-cast 83 mm diameter ingots (each data point is an average of two tests).
The microstructural (e.g. precipitation) features that are associated with these changes were examined by optical metallography but the observations were not conclusive. It must, of course, be emphasized that the details of Fig. 5 may change at the earlier times (e.g. less than several hours) as the result of any changes in dimensions, since longer times are required to heat the interior of thicker specimens to the desired temperature. Also, the results may be different as a result of subsequent mechanical processing, as shown with the earlier extrusion tests. Just as the data of Fig. 5 suggest that up to 6 h and over 40 h precipitation treatments will optimize T3 properties, Table 2 suggests that these same times (at least up to 12 h) at 350 °C will increase workability. These observations are consistent with the extrusion experiments (4-6 h "precipitation" treatments), where favorable extrudability and T3 properties were observed. This indicates that the large particles formed during prolonged aging dissolve during the standard solution treatment. The long-range rise in strength and ductility with time cannot be related directly to precipitation because of the subsequent solution anneal. It can perhaps be related to continued homogenization. Paterson and Sheppard [18] have noted progress of dissolutions at 400 °C; large precipitate and constituent particles become smaller and regions of eutectic became fewer up to 24h. The earlier work of McQueen and coworkers [1, 2] showed that ductility rose as a result of 460 °C annealing as well as a result of 500 °C annealing, but that in 24 h the hot strength and segregated particle density (CuA12, (FeMn)3Si2AI6) at 4 6 0 ° C decreased only half as much as at 500 °C [2]. For 8 h
M. E. Kassner et al. /
Extrudability and mechanical properties orAl alloy
homogenization, the ductility at 350 °C rose as the temperature increased from 440 °C to 520 °C [1]. While the diffusion rate is lower, there is still a tendency for useful alloying elements to redistribute from constituent particles or eutectic into the uniform distribution of large precipitates; those atoms would then be available for hardening upon heat treatment. It is possible that at 300 °C there would be some coalescence of the dispersed particles which would decrease the amount of heterogeneous precipitation, thus enhancing homogeneous formation and improved hardening.
3.4. The effects of various homogenization cycles on the ambient-temperature mechanical properties Analogously, we determined the effect of various homogenization times (at 504 °C) as well as heating rates (to the homogenization temperature) and various cooling rates to ambient temperature on the T3 (2.5% stretch, 4 days) tensile properties. The selected times for nearly linear heating to the homogenization temperature were 4, 8, and 12 h; the durations of homogenization times were 6, 12, and 24 h; the times for linear cooling to ambient temperature were 4 h ( - 1 2 0 ° C h ~), 8 h ( - 6 0 ° C h-1),and 12h(-4°C
b-I).
The results are listed in Table 5; each value is an average of three tests. This table indicates that the T3 properties are not dramatically sensitive to our implemented changes in heating-rate, homogenization or "soak" time, or to cool-down times for cast ingots. Table 6 and Fig. 6 provide some insight into subtle independent effects of the heat-up times (rates), the homogenization times, and the cool-down times (rates). The average Oy, ultimate tensile strength, and percentage E1 (nine average values as derived from Table 5) are listed for each heating time, saturation time, and cooling time. The data suggest that the 8 h heating followed by the m a x i m u m homogenization time of 24 h followed by a relatively rapid cooling within 4 h is expected to produce superior T3 properties. (However, 8 h cooling could be better for improved extrudability, Section 3.2, with negligible change in properties.) In fact, Table 5 reports that this combination produced T3 properties of 374 MPa yield stress, 640 MPa ultimate tensile stress, and 13.3% elongation. These values are somewhat better than the average values reported in Table 6. The explanation for these trends is not readily apparent. Homogenizing treatments to shrink and spheroidize constituent particles (mainly Fe-AI silicates) and to remove coring, and particles that result from composition gradients, are usually more complete for longer times and higher temperature [19]. The observed behavior is in agreement with the work of McQueen
15
TABLE 5. T3 tensile properties of AA2024 extracted from 83 mm diameter ingot heated to the homogenization temperature 504 °C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Heat-up time (h)
Soak time (h)
Cool-down time (h)
Yield stress (MPa)
UTS (MPa)
El (%)
4 4 4 4 4 4 4 4 4 8 8 8 8 8 8 8 8 8 12 12 12 12 12 12 12 12 12
6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24
4 4 4 8 8 8 12 12 12 4 4 4 8 8 8 12 12 12 4 4 4 8 8 8 12 12 12
374 374 371 377 373 372 369 373 369 374 377 374 374 375 374 372 371 369 376 375 373 374 372 373 374 371 570
461 458 456 460 461 457 452 447 454 461 463 461 461 462 462 456 456 454 463 461 458 461 461 458 456 456 455
13.4 13.2 14.1 12.0 12.9 13.9 11.2 11.5 14.1 13.7 14.1 13.3 12.4 13.4 13.6 13.4 12.9 13.7 13.2 13.3 13.4 11.8 13.1 12.9 12.3 13.3 13.3
and coworkers [1, 2] that the strength and the density of particles at the grain boundaries decrease as the time increases from 8 to 24 h at the homogenization temperature [2] and the ductility at 350 °C rises as the homogenization temperature increases from 440 to 520 °C [1] for a fixed time. Homogenization treatments of 24 h have been shown to reduce the flow stress of 2014 by 20% [19]. Disappearance of the CuAI 2 particles and reduction of the Mg2Si particles was observed. Constituent particles of Cu2MgsSi6AI 5 and (FeMn)A16 became smaller in size as leveling or homogenization of soluble elements took place (Mg2Si particles diminished to 1.5 p m in 8 h and to 0.6 p m for the same or fewer particles). In AI-Mg-Si alloys, Precht and Pickens [20] showed that 4 h homogenization at 550 °C produced the maximum improvement in ductility. At 550 °C, ductility improved up to 6 h and remained stationary from there to 10 h; however, the strength was hardly altered at all. It is also well known that faceted fl(A1FeSi) gradually changes over long times to spheroidal a(A1FeSi) which is much less deleterious to ductility [19], which would seem to be present in 2024.
16
M. E. Kassner et al.
/
Extrudability and mechanical properties orAl alloy 15
5O0
TABLE 6. Effect of heat-up, soak, and cool-down times on the T3 mechanical properties of specimens (from Table 5)
• Y ~ Strength r~urs []
%EL
Time (h) A
O'y (MPa)
UTS (MPa)
E1 (%)
372 374 373
457 460 459
12.9 13.4 13.0
Athomogen~anon wmperature 6 374 12 374 24 373
459 459 458
12.6 13.1 13.6
Cool-down 4 8 12
374 374 371
460 460 455
13.5 12.9 12.9
373
458
13.1
450 ¸ 13
,,,,,,% ,,,,,,,.
425 ¸
....%%
,;.;.;,:
.,.,.,% ,,.,,,%
P
400,
;,;,;,;,
,,.,.,% ,,.,.,%
,.,,;,;,
%,,.,% ,.%%% ,,%.,% ,.,..,%
;,;-7,
%-,-.%
;,;,;,;, 4
8
12
Heat-upTime
(hrs)
1
Heat-up 4 8 12
-15
500
tarns []
475
%EL
-14 ?.,.,-,
;:.:.:.: .
Over-aOaverage
45o -13
;:::: 425
:7";-';';: ,,,.,.%,
4oo
:+-,:,:
d
.%,,,, •
.,,,%, ,,%,,, .,%,,,
-12
!?-?;?;-?
40 h) providing both relatively high strength and 375
ductility.
:;:,,,-;: r..,
350
4
12
8
Homogenization
Time
(hrs)
15
500
•
~
[]
% EL
[]ors 475
"
Strength
-:,::::] ":,J::,2
13
..,.,.% %,,,,%
,,,,%..
,12
.,,,,,%
4
8 Cool-down Time
(4) The heating and cooling rates to and from the homogenization temperature and the time at the homogenization temperature do not dramatically affect the T3 tensile properties of unextruded ingots. However, longer soak times and more rapid cooling rates may slightly improve the properties.
Acknowledgments This work was performed as part of the Oregon Metals Initiative Program which is funded through partnership with the Oregon Economic Development Department and the US Bureau of Mines.
12
(hrs)
Fig. 6. Bar-graph illustration of the data of Table 6.
4. Conclusions (1) The elevated-temperature tensile and extrusion tests indicate that the hot workability of aluminum alloy AA2024 appears to improve with some hightemperature precipitation treatments. (2) The precipitation treatments before extrusion do not appear to degrade the ambient-temperature T3 and T8 tensile properties of the alloy. (3) The time at the precipitation temperature appears to affect the T3 and T8 tensile properties in unextruded ingots with longer times (e.g. more than
References 1 P. Wouters, B. Verlinden, H. J. McQueen, E. Aernoudt, L. Delaey and S. Cauwenberg, Mater. Sci. Eng., A123 (1990) 239-245. 2 B. Verlinden, E Wouters, H. J. McQueen, E. Aernoudt, L. Delaey and S. Cauwenberg, Mater. Sci. Eng., A123 (1990) 229-237. 3 Aluminum Standards 1988, Aluminum Society of America, Washington, DC, 1988. 4 H.J. McQueen and D. C. Celliers, Mater. Forum, (1993) in press. 5 T. Sheppard, in J. Jeglitsch et al. (eds.), Proc. 8th Light Metals Congress, The University, Leoben, 1987, pp. 301-311. 6 H.J. McQueen, in T. G. Langdon and H. D. Merchant (eds.), Hot Deformation of Al Alloys, TMS, Warrendale, PA, 1991, pp. 105-120. 7 T. Sheppard and E P. Vierod, Mater. Sci. Technol., 1 (1985) 321-324.
M. E. Kassner et al.
/
Extrudability and mechanical properties orAl alloy
8 M.J. McQueen, in T. G. Langdon and H. D. Merchant (eds.), Hot Deformation orAl Alloys, TMS, Warrendale, PA, 1991, pp. 31-54. 9 E. Evangelista and H. J. McQueen, Modelling of Plastic Deformation and Its Engineering Applications, Riso National Laboratory, Roskilde, 1992. 10 M. E. Kassner and X. Li, unpublished research, Oregon State University, Corvallis, OR, 1992. 11 J. E. Hatch (ed.), Aluminum Properties and Physical Metallurgy, ASM, Metals Park, OH, 1984. 12 J. Baumgarten, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washington, DC, 1984, pp. 40-44. 13 M. E. Kassner, unpublished research, Oregon State University, Corvallis, OR, 1991. 14 O. Reiso, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1984, pp. 31-40.
17
15 O. Reiso, Proc. 4th Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washingtdon, DC, 1988, pp. 287-295. 16 J. Langerweger, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1984, pp. 41-45. 17 J. Langerweger, Proc. 4th Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washington, DC, 1988, pp. 381-384. 18 S. J. Paterson and T. Sheppard, Met. Tech., 9 (1982) 389-390. 19 E. C. Beatty, Proc. 2nd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1977, pp. 225-228. 20 W. Precht and J. R. Pickens, Metall. Trans, A, 18 (1987) 1603-1611.
9
The effect of homogenization and precipitation treatments on the extrudability and ambient-temperature mechanical properties of aluminum alloy AA2024 M. E. Kassner and X. Li Department of Mechanical Engineering and Center for Advanced Materials Research, Oregon State University, Corvallis, OR 97331 (USA)
H. J. McQueen Department of Mechanical Engineering, Concordia University, Montreal, Que. (Canada) (Received September 17, 1992; in revised form February 1, 1993)
Abstract This study determined the independent effects of various homogenization cycles and precipitation treatments on the elevated temperature workability and the final ambient temperature mechanical properties of aluminum alloy AA2024, The elevated-temperature tensile and extrusion tests indicate that the workability of AA2024 improves with elevatedtemperature precipitation treatments as suggested by earlier investigations. The precipitation treatments do not appear to degrade the ambient-temperature T3 and T8 tensile properties. The time at the precipitation temperature appears to affect the T3 and T8 tensile properties in homogenized ingot, longer times especially providing both relatively high ambient-temperature strength and ductility. The time at the standard homogenization temperature and the heat-up and cool-down rates do not dramatically affect the T3 tensile properties of homogenized ingot. However, long soak times (24 h) at the homogenization temperature and more rapid cooling rates (120 °C h-~) may improve the properties somewhat.
I. Introduction
In a previous study [1, 2] which included one of the coauthors of this article, the elevated-temperature torsional behavior of aluminum alloy AA2024 [3] was determined at relatively high strain rates. These tests were performed subsequent to both homogenization and precipitation treatments at various temperatures. The purpose of the work was to determine thermal pretreatments that led to optimal extrudability which was presumed to occur where the elevated-temperature stress vs. strain behavior indicated lower peak stresses and higher ductility, the former implying lower break-through pressures. This would increase the limiting extrusion ratio of a given press and increase the permissible extrusion velocity [2, 4-7]. These investigators [1, 2] performed torsion tests at 310 and 370 °C at a strain-rate g of 1 s ~ on A A 2 0 2 4 alloy, homogenized between 420 and 520 °C, and either water quenched or air cooled and subsequently aged at 290 or 350 °C. The stress vs. strain curves of the alloy in the quenched condition exhibited relatively large peak stresses Op and work softening with relatively 0921-5093/93/$6.00
low strain to failure. The peak stress and work softening were reduced and the ductility improved as either the aging or precipitation temperature or the deformation temperature was increased; the behavior became closer to that of pure aluminum in which a high level of dynamic recovery provides excellent workability [6]. The homogenization temperature had a large effect on Op but little effect on the ductility for deformation at 310 °C, whereas at 370 °C there was little effect on op but a large effect on ductility which was much higher. These results suggest that suitable homogenization and precipitation treatments can lower the elevatedtemperature flow stress, notably decreasing the initial high peak stress. The authors concluded that the decrease in flow stress and increase in ductility subsequent to precipitation treatments arise from the relative amounts of solute and dynamic precipitation (of presumably CuAI2 and (Cu, Mg)A12 [2]). such a reduction in peak stress and increase in ductility would seem to be beneficial for direct extrusion of the alloy. In review papers McQueen and Evangelista [8, 9] showed that similar effects to the above had been found in other age-hardenable aluminum alloys. © 1993 - Elsevier Sequoia. All rights reserved
10
M. E. Kassner et al. /
Extrudability and mechanical properties of Al alloy
The present study intended to verify the previous investigation [1, 2] and expand the experimental scope. Basically, the intent of this study was also to determine the effect of different precipitation treatments on the hot workability of aluminum alloy AA2024. Instead of torsion tests, the present study emphasized elevatedtemperature tension which is important at the periphery as metal passes over the die land in the extrusion process. This study also considered (1) additional precipitation temperatures, (2) the effect of the heating rate to the homogenization temperature, (3) the time at homogenization temperature, (4) the rate of cooling from the homogenization temperature, and (5) the time at the precipitation temperature, to find their effects on the T8 and T3 (ambient-temperature) tensile properties of AA2024. Furthermore, (6) the "laboratory" conditions that predict favorable extrudability were tested by extruding AA2024 directly, and (7) by determining whether adequate ambient-temperature properties are evident subsequent to extrusion. Therefore, this work may be considered as a logical extension of the earlier work presented in this journal.
2. Experimental procedure This study utilized AA2024 aluminum provided in the form of direct chill cast ingots of 83 mm and 178 mm diameters. The compositions of the ingots tested in this study were within the specification given in Table 1. Tensile tests were performed on an Instron 4505 screw driven tensile machine with computerized data acquisition. Specimen geometries varied, and the typical gage dimensions for ingot characterization were 5.1 mm diameter and 25.4 mm length. Specimens from extrusions had 2.8 mm diameter and 20 mm or 11.4 mm gage length. Specimens were evaluated from random positions within the ingots; it was determined that the mechanical properties were independent of the position. Elevated-temperature tests were performed utilizing a three-zone furnace with Eurotherm controllers and power supplies. Strain rates varied between 0.67 × 1 0 - 3 and 3.0 × 1 0 -3 s -1. The solution annealing temperature was always 493 °C. The temperature of the specimens was controlled to within + 2 °C of the set temperature during solution annealing. A 10 min heat-up was required to achieve the solution anneal temperature once specimens were inserted into the
furnace. Generally, solution annealed tensile specimens were maintained at temperature for 1 h. The ductility was measured as the engineering strain to failure (El) equal to A L / L o where L 0 is the initial length and, for high-temperature tests, also as a reduction in area (RA) equal to (A 0 - A f ) / A o where A 0 is the initial area and A f is the final area. The yield and ultimate tensile stresses are reported as engineering stresses. The T3 treatment consists of a solution anneal followed by a water quench, refrigeration for 1-3 h, "stretch" of 1%-2.5%, and an ambient temperature age. The "stretch" increases the density of dislocations which act as heterogeneous nucleation sites, presumably for CuAI 2 [1, 2], and increases the strength over the undeformed and aged (T4) alloy. More dramatic (T8) strengthening can be achieved if the alloy is aged artificially at an elevated temperature subsequent to the stretching to attain an optimum precipitate size. The reported yield stress is based on a 0.002 plastic strain offset. The plastic "stretch" varied between 1% and 2.5% plastic strain for T3 and T8 treatments. The ambient temperature age for T3 treatments varied from 2.5 to 19 days. The difference in T3 tensile properties over this range was not substantial: only 4 MPa (roughly 1%) for the yield stress, 6 MPa for the ultimate tensile strength (also 1%), and 1.5% El (factor of 1.1) [10], consistent with other work [11]. The T8 temperature and time-at-temperature were 190 °C and 12 h respectively. Specimens were kept in the freezing compartment of a refrigerator subsequent to the solution anneals (1-3 h) and prior to the stretch to suppress precipitation. The time from the refrigerator to the T8 temperature was always less than 1/4 h.
3. Results and discussion 3.1. Elevated-temperature tests of precipitation treated alloy A set of tensile experiments was performed to compare with the earlier work [2, 3] by McQueen and coworkers. Specimens were extracted from 83 mm diameter homogenized ingots. Homogenized specimens received precipitation treatments at 270, 330, and 350 °C for 1, 6 and 12 h. This resulted in precipitation of CuAI 2 and (CuMg)A12, based on the microscopy performed by the earlier work that was performed with one of our coauthors (HJMcQ). This
TABLE 1. Specificationfor the chemicalcompositionof alloy AA2024 used in this study Limit
Si
Fe
Cu
Mn
Mg
Cr
Zn
Ti
V
Be
Minimum(wt.%) Maximum (wt.%)
0.15
0.30
4.15 4.35
0.50 0.70
1.30 1.50
0.05
0.18
0.05
0.01
0.001 0.005
M. E. Kassner et al.
/
Extrudability and mechanical properties of Al alloy
1l
TABLE 2. The elevated temperature mechanical properties of"precipitation-treated" AA2024 specimens from homogenized 83 mm diameter ingot (g = 10 -3 s- E) Test temperature (°C)
Precipitation temperature (°C)
Precipitation time (h)
Oy (MPa)
UTS (MPa)
El (%)
RA (%)
350 350 350 350 350 350 350 350
270 330 330 330 350 350 350 0
12 1 6 12 1 6 12 --
69.6 75.8 79.6 66.2 76.5 69.6 62.1 76.5
84.1 89.6 91.7 82.0 89.6 84.1 79.3 89.6
29 33.3 36.3 35.3 35.4 36.6 33.4 30.3
86 89 88 79 89 88 88 84
100"
90"
80"
~
70"
~60-
-----
350 o C - l h r - 350 °C-6hrs ....... 350 °C-12hrs . . . . . . . . . . 270 °C- 12hrs . . . . . . . . . untreated
50"
4(J
-
0.00
i
,
0.02
=
•
0.04
i
-
0.06
i
0.08
0.10
Strain
Fig. 1. The engineering tensile stress vs. engineering strain behavior of AA2024 aluminum alloy at 350 °C subsequent to various elevated-temperature precipitation treatments. Specimens were extracted from 83 mm diameter ingots (g= 10 -3 s- ~).
- 40 a=
' 20 0
270
330
330
330
Precipitation Temperature
350
350
350
°C
Fig. 2. Bar-graph illustration of the data of Table 2.
was followed by elevated-temperature testing at 350 °C (a common extrusion temperature). From Table 2 and Figs. 1 and 2 it appears, based on the 350 °C tests on specimens that were not precipitation treated, that prolonged precipitation treatments decrease the tensile
strength and increase the ductility. More specifically, the table and figures indicate that precipitation treatments between 330 and 350 °C for 12 h reduce the yield stress and ultimate tensile strength up to 14% and increase the tensile ductility by a factor of up to 1.3. These tensile results are qualitatively consistent with the earlier work [2, 3] performed by torsion testing. Some scatter in the data may arise from the procedure of heating to the testing temperature. This heating was done as quickly as possible (30 min for our furnace) to "simulate" (rapid) induction heating which is frequently utilized in extrusion processes. As a consequence, some non-reproducible temperature gradients existed along the specimen axis (e.g. 20 °C variation). The error in stress measurement is about + 3 MPa. This may explain some of the anomalous results (e.g. 6 h 330 °C). 3.2. Elevated-temperature extrusions of the precipitation treated alloys Although the precipitation treatments appear to increase workability (specifically extrudability) based on simple mechanical tests, a more important test is extrusion testing. The extrusion cross-sections are shown in Figs. 3 and 4. In the first series of tests, 178 mm diameter ingots were extruded into L-shaped beams (air cooled) of 5.1 mm thickness, and 63.5 mm leg width (R = 38.4) as shown in Fig. 3. Ingots were extruded after various commercially relevant thermal processes, some of which were believed to cause precipitation. The treatments included the following: (1) a fan-cool (FC) with a cooling rate of 3 1 5 - 3 7 0 °C h -j which followed a 6 h homogenization (482499 °C with 8-10 h linear heating to the homogenization temperature); (2) oven-cool (OC), rather than FC, following an identical homogenization cycle as in (1) except a reheat to 504 °C followed by a 12 h (40 °C h -1) cool to ambient temperature; (3) an oven-stepcool (OSC), where subsequent to the homogenization
12
M. E. Kassner et aL
/
Extrudability and mechanical properties orAl alloy
in (2) the ingot is cooled to 349 °C (a "precipitation" temperature) in 4 h and maintained for 6 h before an 8 h linear cooling to ambient temperature. The heat treatments are shown schematically in Fig. 3. As expected, it appeared that the oven-step-cool ingots could be most easily extruded (billet tempera30
6O0/ J • yieldStrength
550 D %EL 20 -15 •
o~
-lO
Te I
T3
I
(a)
Ac
"~3
~8
T3 t
FC
Tj8
T~
oc
T8I OSC
"L" specimen
8 hr 8-10 h
~
/ 2 0 °C
I 20 °C
FC 1 hr
FC
+
8-10
hr~f-~
TABLE 3. T3 and T8 properties of specimensextracted from 178 mm diameter ingots, extruded after various thermal treatments
OC 1 hr 504 oC
FC
+
8-10
~
4 hr 6
"~ hr
349 °C'
20oc
(b)
ture at 316-343 °C after 10-15 min induction heating) in terms of the maximum possible extrusion speed of the ingot with acceptable surface quality. With this criterion, it is expected that softer, perhaps more ductile alloy, would be most extrudable. The extrusions of fan-cooled ingots showed tearing at the surface at 7.3 m min-1 while oven-step-cool extrusion did not give evidence of tearing until about 8.8 m min -1. Extrusions of oven-cooled ingots (probably some precipitation since the slow cool allowed for a period, albeit small, within the precipitation range) did not show tearing at 8.5 m min- 1 but showed tearing at 9.1 m min- 1. This is in agreement With improved extrudability of billets with prior precipitation treatment observed in 2024 [7] and in AI-Mg-Si alloys [6, 12]. Table 3 and the associated bar-graph, Fig. 3, list the tensile test results of T3 (2.5% stretch, 19 days) and T8 specimens extracted parallel to the axis of the extrusion at random locations. The mechanical properties listed are an average of five tests. Table '3 and Fig. 3 show that the "final" T3 and T8 properties are essentially identical for the three "pretreatments". Therefore, there is encouraging evidence that the oven-step-cool pretreatment can increase the extrudability and yet not compromise the critical T3 and T8 mechanical properties (at least for this starting ingot dimension, extrusion temperature, and final configuration). Experiments analogous to the above were performed on precipitation-treated ingots of a smaller diameter (83 mm) and extruded into a simple crosssectional configuration of 7.9 mm by 32.5 mm with "rounded" corners (R=21.1) shown in Fig. 4 as S specimens. The oven-cooled (OC) ingots were heated linearly to the homogenization temperature of 504 °C within 12 h, maintained at 504 °C for 4 h, and cooled linearly to ambient temperature within 12 h (40 °C h-l). The oven-step-cooled (OSC) ingots were also maintained at 504 °C for 4 h (after a 12 h heating) and cooled linearly to 349 °C (the same "precipitation" temperature as for the L-shaped extrusions) within
8 hr 20o0
OSC
Fig. 3. (a) Bar-graph illustration of the data of Table 3 and (b) schematicdiagramsof the three heat treatments.
Cooling method
Treatment
ay (MPa)
UTS (MPa)
El (%)
FC OC OSC FC OC OSC
T3 T3 T3 T8 T8 T8
359 340 341 474 473 472
445 420 418 495 483 482
13.6 15.1 15.5 6.3 4.4 4.6
FC, furnace cool; OC, oven cool; OSC, oven-step-cool.
/
M. E. Kassner et al.
Extrudability and mechanical properties oral alloy
1.5 h, maintained at 349 °C for 4 h, and cooled to 93 °C in 4 h before removal from the furnace. The ingots were induction heated to a temperature between 370 and 425 °C immediately before extrusion (about 50 °C higher than the previous (L) extrusion temperature). The induction heating of the extrusion required about 15 min. The heat treatments are shown schematically in Fig. 4. It was found that the OSC billets could be processed at a greater extrusion speed without surface defects than conventionally homogenized 2024 (although the quantitative rates were not reported). Table 4 and Fig. 4 list averages of three ambient temperature tensile tests for each cooling cycle on extrusions and homogenized as-cast ingots which provide a reference. First, the T3 (2.5% stretch, 19
3O
• YieldSlrength El Lrrs El %EL
25
~ 450"~
5 ~o~
4o0.
o
350. .
,o
7
(a)
T8 I
7
AC
,.
.
.
oc
.
t
.
J
osc
"S" Specimen 32.5 mm ~ __~.7.9mm
C
~
)T
13
days) and T8 mechanical properties of homogenized, unextruded ("as cast") ingots were determined. The homogenization included a 9 h linear heating to the homogenization temperature of 504 °C, maintenance at temperature for 6 h, and cooling at 150 °C h- ~ to ambient temperature. The variation of the test results was very small, _+ 1%. Since these ingots were not extruded, the tests do not reflect production properties. The strength values (average of six tests) are slightly higher than the typical AA2024 T3 (2.25% stretch) and T8 values and the ductility is slightly lower [3, 13]. These provide "baseline" or reference values with which the T3 and T8 properties of extrusions can be compared. The T3 and T8 properties of the OC and OSC extrusions were well above average. We note that the previous L extrusions have lower strength than both the baseline (non-extruded or homogenized) material and extrusions from 83 mm ingots, indicating the variability of properties resulting from different ingot sizes and extrusion processes. The correlation between high ductility in tension and in extrusion seems reasonable. Previous work on 2024 [5, 7] does not confirm this; material cooled from the homogenization temperature to the billet insert temperature had higher torsional strength, lower ductility, and higher breakout pressure, but higher maximum extrusion velocity than furnace cooled material. Part of the problem lies in the lack of precise knowledge of the distribution and concentration of constituent particles and of precipitates in both the present and previous materials. Another part lies in the extrusion process in which there is a rise of 50-100 °C in average temperature in the deformation zone, being larger for higher billet stresses (lower insert temperature, higher alloying or particle content) and larger extrusion ratios and ram speeds [4, 5]. However, there
4 hr
TABLE 4. T3 and T8 properties of specimens extracted from 83 mm diameter AA2024 ingots,extruded subsequent to various thermal treatments
2 hr 20C
Cooling Treatment method
oy UTS El (%) RT age time (MPa) (MPa) (Days)
OC OC OSC OSC AC OC OC OSC OSC AC
431 432 409 408 380 492 498 490 496 484
OC
4 hr
/
C--~ .5 hr
\,.r
12/
349oc 90 ~oC~hr
/ 20°(3 (b)
\
20°C
OSC
Fig. 4. (a) Bar-graph illustration of the data of Table 4 and (b) schematic diagrams of the three heat treatments.
T3 T3 T3 T3 T3 T8 T8 T8 T8 T8
543 541 508 513 464 523 529 518 525 512
13.7 13.0 13.8 15.5 10.8 9.9 9.8 8.8 9.7 4.7
12.5 19 12.5 19 19 ------
OC, oven cool; OSC, oven-step-cool; AC, as-cast; RT, roomtemperature.
14
M. E. Kassner et aL / Extrudability and mechanical properties ofAl alloy
is an additional temperature increase from the average to that of the periphery as it passes near the die land. A detailed study of AI-Mg-Si alloys was performed by Reiso [14, 15]. In slowly cooled billets large Mg2Si precipitates form on constituent particles and do not dissolve on preheating or in the deformation zone but form a eutectic with the constituent particles which melt at the die land, causing tearing. In billets cooled rapidly to an intermediate temperature, precipitates are smaller and more homgeneous with the result that they redissolve in preheating and deformation so that no eutectic is formed and the tearing temperature is close to the true solidus of the alloy at a higher ram speed. This theory may be applicable to the present 2024 alloys but the distribution and nature of constituent particles and location or size of precipitates in different cooling cycles have not been determined. A comparison between A I - C u - M g and AI-Mg-Si alloys is relevant. In both alloys (although one can take the composition, as defined by the principal elements and determine the solidus temperature) melting occurs below the solidus owing to the presence of Fe-Al silicates. The presence of these silicates with the other elements can lead to melting below the solidus. All research suggests that proper control of the precipitation produces lower flow stresses at 3 0 0 - 4 0 0 °C and lower extrusion breakout pressures [1, 2, 5-7, 9, 12, 14-17].
3.3. Effect of duration of the precipitation treatment on the ambient-temperature mechanical properties Another part of this study consisted in determining the effect of precipitation time on the T3 mechanical properties. The homogenization cycle consisted of linear heating from ambient temperature to 504 °C within about 8 h. The specimens (of about 12.7 mm x 12.7 mm x 152 mm cut from an 83 mm diameter ingot prior to homogenization) were homogenized for 6 h at 504 °C. The samples were then step cooled by a 3 h linear cooling to 350 °C, the precipitation temperature, and maintained for various times up to 72 h. Subsequently the samples were removed from the furnace and air cooled. The T3 properties were then assessed after an age of 12.5 days. The results illustrated in Fig. 5 show that, for all precipitation treatment times up to 72 h (limit considered commercially practical), higher yield stress and ultimate tensile strength are associated with higher ductility (all specimens were solution annealed for 1 h subsequent to the precipitation treatments). Relatively favorable T3 properties are observed with very prolonged precipitation times (e.g. greater than 40 h), as well as at times of about 20 h, and times up to about 6 h. Lower (but not unacceptable) T3 properties are observed for precipitation times between 8 and 12 h.
480
18
'~...~
~!'.'.~.~._..
~6
450 .g
14
.
420
12
"~ 39o .
.
.
.
.
.
.
.
.
-'=~ 3SO0 [-~ 30' '
.
-
.................
~
_- . . . . . .
Yield Stress
t 8
/
f~
t--~l
I'2 300
, 0
, I0
. . . . 20
30
, 40
•
, 50
•
, 60
-
, 70
•
tO 80
Precipitation Time (hrs) at 350°C
Fig. 5. The relationship between the precipitation time at 350 °C and the T3 (ambient temperature) ultimate tensile strength, the yield strength, and the tensile strain in AA2024 specimens extracted from as-cast 83 mm diameter ingots (each data point is an average of two tests).
The microstructural (e.g. precipitation) features that are associated with these changes were examined by optical metallography but the observations were not conclusive. It must, of course, be emphasized that the details of Fig. 5 may change at the earlier times (e.g. less than several hours) as the result of any changes in dimensions, since longer times are required to heat the interior of thicker specimens to the desired temperature. Also, the results may be different as a result of subsequent mechanical processing, as shown with the earlier extrusion tests. Just as the data of Fig. 5 suggest that up to 6 h and over 40 h precipitation treatments will optimize T3 properties, Table 2 suggests that these same times (at least up to 12 h) at 350 °C will increase workability. These observations are consistent with the extrusion experiments (4-6 h "precipitation" treatments), where favorable extrudability and T3 properties were observed. This indicates that the large particles formed during prolonged aging dissolve during the standard solution treatment. The long-range rise in strength and ductility with time cannot be related directly to precipitation because of the subsequent solution anneal. It can perhaps be related to continued homogenization. Paterson and Sheppard [18] have noted progress of dissolutions at 400 °C; large precipitate and constituent particles become smaller and regions of eutectic became fewer up to 24h. The earlier work of McQueen and coworkers [1, 2] showed that ductility rose as a result of 460 °C annealing as well as a result of 500 °C annealing, but that in 24 h the hot strength and segregated particle density (CuA12, (FeMn)3Si2AI6) at 4 6 0 ° C decreased only half as much as at 500 °C [2]. For 8 h
M. E. Kassner et al. /
Extrudability and mechanical properties orAl alloy
homogenization, the ductility at 350 °C rose as the temperature increased from 440 °C to 520 °C [1]. While the diffusion rate is lower, there is still a tendency for useful alloying elements to redistribute from constituent particles or eutectic into the uniform distribution of large precipitates; those atoms would then be available for hardening upon heat treatment. It is possible that at 300 °C there would be some coalescence of the dispersed particles which would decrease the amount of heterogeneous precipitation, thus enhancing homogeneous formation and improved hardening.
3.4. The effects of various homogenization cycles on the ambient-temperature mechanical properties Analogously, we determined the effect of various homogenization times (at 504 °C) as well as heating rates (to the homogenization temperature) and various cooling rates to ambient temperature on the T3 (2.5% stretch, 4 days) tensile properties. The selected times for nearly linear heating to the homogenization temperature were 4, 8, and 12 h; the durations of homogenization times were 6, 12, and 24 h; the times for linear cooling to ambient temperature were 4 h ( - 1 2 0 ° C h ~), 8 h ( - 6 0 ° C h-1),and 12h(-4°C
b-I).
The results are listed in Table 5; each value is an average of three tests. This table indicates that the T3 properties are not dramatically sensitive to our implemented changes in heating-rate, homogenization or "soak" time, or to cool-down times for cast ingots. Table 6 and Fig. 6 provide some insight into subtle independent effects of the heat-up times (rates), the homogenization times, and the cool-down times (rates). The average Oy, ultimate tensile strength, and percentage E1 (nine average values as derived from Table 5) are listed for each heating time, saturation time, and cooling time. The data suggest that the 8 h heating followed by the m a x i m u m homogenization time of 24 h followed by a relatively rapid cooling within 4 h is expected to produce superior T3 properties. (However, 8 h cooling could be better for improved extrudability, Section 3.2, with negligible change in properties.) In fact, Table 5 reports that this combination produced T3 properties of 374 MPa yield stress, 640 MPa ultimate tensile stress, and 13.3% elongation. These values are somewhat better than the average values reported in Table 6. The explanation for these trends is not readily apparent. Homogenizing treatments to shrink and spheroidize constituent particles (mainly Fe-AI silicates) and to remove coring, and particles that result from composition gradients, are usually more complete for longer times and higher temperature [19]. The observed behavior is in agreement with the work of McQueen
15
TABLE 5. T3 tensile properties of AA2024 extracted from 83 mm diameter ingot heated to the homogenization temperature 504 °C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Heat-up time (h)
Soak time (h)
Cool-down time (h)
Yield stress (MPa)
UTS (MPa)
El (%)
4 4 4 4 4 4 4 4 4 8 8 8 8 8 8 8 8 8 12 12 12 12 12 12 12 12 12
6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24 6 12 24
4 4 4 8 8 8 12 12 12 4 4 4 8 8 8 12 12 12 4 4 4 8 8 8 12 12 12
374 374 371 377 373 372 369 373 369 374 377 374 374 375 374 372 371 369 376 375 373 374 372 373 374 371 570
461 458 456 460 461 457 452 447 454 461 463 461 461 462 462 456 456 454 463 461 458 461 461 458 456 456 455
13.4 13.2 14.1 12.0 12.9 13.9 11.2 11.5 14.1 13.7 14.1 13.3 12.4 13.4 13.6 13.4 12.9 13.7 13.2 13.3 13.4 11.8 13.1 12.9 12.3 13.3 13.3
and coworkers [1, 2] that the strength and the density of particles at the grain boundaries decrease as the time increases from 8 to 24 h at the homogenization temperature [2] and the ductility at 350 °C rises as the homogenization temperature increases from 440 to 520 °C [1] for a fixed time. Homogenization treatments of 24 h have been shown to reduce the flow stress of 2014 by 20% [19]. Disappearance of the CuAI 2 particles and reduction of the Mg2Si particles was observed. Constituent particles of Cu2MgsSi6AI 5 and (FeMn)A16 became smaller in size as leveling or homogenization of soluble elements took place (Mg2Si particles diminished to 1.5 p m in 8 h and to 0.6 p m for the same or fewer particles). In AI-Mg-Si alloys, Precht and Pickens [20] showed that 4 h homogenization at 550 °C produced the maximum improvement in ductility. At 550 °C, ductility improved up to 6 h and remained stationary from there to 10 h; however, the strength was hardly altered at all. It is also well known that faceted fl(A1FeSi) gradually changes over long times to spheroidal a(A1FeSi) which is much less deleterious to ductility [19], which would seem to be present in 2024.
16
M. E. Kassner et al.
/
Extrudability and mechanical properties orAl alloy 15
5O0
TABLE 6. Effect of heat-up, soak, and cool-down times on the T3 mechanical properties of specimens (from Table 5)
• Y ~ Strength r~urs []
%EL
Time (h) A
O'y (MPa)
UTS (MPa)
E1 (%)
372 374 373
457 460 459
12.9 13.4 13.0
Athomogen~anon wmperature 6 374 12 374 24 373
459 459 458
12.6 13.1 13.6
Cool-down 4 8 12
374 374 371
460 460 455
13.5 12.9 12.9
373
458
13.1
450 ¸ 13
,,,,,,% ,,,,,,,.
425 ¸
....%%
,;.;.;,:
.,.,.,% ,,.,,,%
P
400,
;,;,;,;,
,,.,.,% ,,.,.,%
,.,,;,;,
%,,.,% ,.%%% ,,%.,% ,.,..,%
;,;-7,
%-,-.%
;,;,;,;, 4
8
12
Heat-upTime
(hrs)
1
Heat-up 4 8 12
-15
500
tarns []
475
%EL
-14 ?.,.,-,
;:.:.:.: .
Over-aOaverage
45o -13
;:::: 425
:7";-';';: ,,,.,.%,
4oo
:+-,:,:
d
.%,,,, •
.,,,%, ,,%,,, .,%,,,
-12
!?-?;?;-?
40 h) providing both relatively high strength and 375
ductility.
:;:,,,-;: r..,
350
4
12
8
Homogenization
Time
(hrs)
15
500
•
~
[]
% EL
[]ors 475
"
Strength
-:,::::] ":,J::,2
13
..,.,.% %,,,,%
,,,,%..
,12
.,,,,,%
4
8 Cool-down Time
(4) The heating and cooling rates to and from the homogenization temperature and the time at the homogenization temperature do not dramatically affect the T3 tensile properties of unextruded ingots. However, longer soak times and more rapid cooling rates may slightly improve the properties.
Acknowledgments This work was performed as part of the Oregon Metals Initiative Program which is funded through partnership with the Oregon Economic Development Department and the US Bureau of Mines.
12
(hrs)
Fig. 6. Bar-graph illustration of the data of Table 6.
4. Conclusions (1) The elevated-temperature tensile and extrusion tests indicate that the hot workability of aluminum alloy AA2024 appears to improve with some hightemperature precipitation treatments. (2) The precipitation treatments before extrusion do not appear to degrade the ambient-temperature T3 and T8 tensile properties of the alloy. (3) The time at the precipitation temperature appears to affect the T3 and T8 tensile properties in unextruded ingots with longer times (e.g. more than
References 1 P. Wouters, B. Verlinden, H. J. McQueen, E. Aernoudt, L. Delaey and S. Cauwenberg, Mater. Sci. Eng., A123 (1990) 239-245. 2 B. Verlinden, E Wouters, H. J. McQueen, E. Aernoudt, L. Delaey and S. Cauwenberg, Mater. Sci. Eng., A123 (1990) 229-237. 3 Aluminum Standards 1988, Aluminum Society of America, Washington, DC, 1988. 4 H.J. McQueen and D. C. Celliers, Mater. Forum, (1993) in press. 5 T. Sheppard, in J. Jeglitsch et al. (eds.), Proc. 8th Light Metals Congress, The University, Leoben, 1987, pp. 301-311. 6 H.J. McQueen, in T. G. Langdon and H. D. Merchant (eds.), Hot Deformation of Al Alloys, TMS, Warrendale, PA, 1991, pp. 105-120. 7 T. Sheppard and E P. Vierod, Mater. Sci. Technol., 1 (1985) 321-324.
M. E. Kassner et al.
/
Extrudability and mechanical properties orAl alloy
8 M.J. McQueen, in T. G. Langdon and H. D. Merchant (eds.), Hot Deformation orAl Alloys, TMS, Warrendale, PA, 1991, pp. 31-54. 9 E. Evangelista and H. J. McQueen, Modelling of Plastic Deformation and Its Engineering Applications, Riso National Laboratory, Roskilde, 1992. 10 M. E. Kassner and X. Li, unpublished research, Oregon State University, Corvallis, OR, 1992. 11 J. E. Hatch (ed.), Aluminum Properties and Physical Metallurgy, ASM, Metals Park, OH, 1984. 12 J. Baumgarten, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washington, DC, 1984, pp. 40-44. 13 M. E. Kassner, unpublished research, Oregon State University, Corvallis, OR, 1991. 14 O. Reiso, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1984, pp. 31-40.
17
15 O. Reiso, Proc. 4th Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washingtdon, DC, 1988, pp. 287-295. 16 J. Langerweger, Proc. 3rd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1984, pp. 41-45. 17 J. Langerweger, Proc. 4th Int. Al Extrusion Technology Seminar, Vol. 2, Aluminum Association, Washington, DC, 1988, pp. 381-384. 18 S. J. Paterson and T. Sheppard, Met. Tech., 9 (1982) 389-390. 19 E. C. Beatty, Proc. 2nd Int. Al Extrusion Technology Seminar, Vol. 1, Aluminum Association, Washington, DC, 1977, pp. 225-228. 20 W. Precht and J. R. Pickens, Metall. Trans, A, 18 (1987) 1603-1611.