Elevated Temperature Mechanical Properties of Al-Li-Cu-Mg Alloy
When compared to Alloy 2024, Al-Li Alloy 2091 displays favorable strength, superior ductility and less susceptibility to HAZ cracking
BY T. ZACHARIA AND D. K. AIDUN
ABSTRACT. A detailed experimental producers of aluminum and other hot-cracking susceptibility of the alloy investigation was carried out to charac research organizations have stepped up being used are essential. terize the elevated temperature mechan their research activities in developing a A number of investigators (Refs. 5-7) ical properties of aluminum-lithium Alloy new generation of aluminum-lithium have successfully determined the elevat 2091. Gleeble hot tensile tests were con alloys. Al-Li-Cu-Mg Alloy 2091 is one of ed temperature mechanical properties of ducted to evaluate the relative elevated the new-generation aluminum-lithium aluminum alloys with the object of relat temperature strength and ductility of Al- alloys intended to have combinations of ing these properties with crack suscepti Li-Cu-Mg (2091) and Al-Cu-Mg (2024) strength, toughness and damage-tolerant bility. The impact resistance at elevated alloys. In general, Alloy 2091 in the "as- properties comparable with those of temperatures of several commercial alu cast" condition exhibited very low ele Alloy 2024 (Ref. 4). minum alloys was evaluated by Archbutt, vated temperature strength and ductility The purpose of determining the ele et al. (Ref. 5), and found that the resis for all the test conditions. This was attrib vated temperature mechanical properties tance to shock was very low at the point uted to the coarse dendritic structure of of any material is to determine its endur of incipient fusion. It was postulated that the as-cast material. The hot tensile tests ance limits in a number of practical appli the alloys exhibiting rapid recovery of indicated that the tensile strength of the cations to which it may be subjected, at impact resistance, with fall in temperature wrought Alloy 2091 adequately matched temperatures approaching and exceed immediately below the solidus, would the strength of Alloy 2024 throughout the ing the melting point. Among such oper exhibit a lower susceptibility to cracking temperature regime. ations are welding and casting, which than alloys in which the recovery was more gradual. The weld heat-affected zone (HAZ) involve fusion of the metal. In any weld hot-cracking susceptibility of the two alu ment, low-melting minor elements, impu Singer and Cottrell (Ref. 6) determined minum alloys was also evaluated using rities or eutectic phases can ultimately the elevated temperature tensile proper the Gleeble thermal-mechanical simula cause hot-cracking. This is also true for ties of ten Al-Si alloys at temperatures tor. The results of hot ductility tests show aluminum-lithium alloys. Cracking that is both above and below the equilbrium that Alloy 2091 in the wrought form normally intergranular can occur in both solidus. The alloys were heated to a exhibits superior ductility throughout the the weld metal and the associated heat- preselected temperature and held at that range of temperatures investigated. affected zone, especially under condi temperature for sufficient time for attain Based on the recovery of ductility from tions of high restraint. Heat-affected- ing structural equilibrium. The specimens the nil ductility temperature (NDT), Alloy zone cracking in a range of alloys has were then tested at temperature, at a 2091 was judged to be crack resistant in been associated with low ductility during constant strain rate. Their results show the weld metal HAZ. cooling from peak temperatures at and that all the alloys exhibit a sudden loss in above the melting point of the alloy. For strength and complete loss of ductility at proper design of a joining process (weld the point of incipient fusion. Above this Introduction ing), reliable measurements of the elevat temperature, the alloys failed with brittle, In recent years, there has been consid ed temperature mechanical behavior and intercrystalline fractures. The extent of erable interest devoted to aluminum-lithi this temperature range was found to be a um alloys due to their reduced density function of the composition of the alloy. and increased elastic modulus compared Pumphrey and Jennings (Ref. 7) deter mined the high-temperature tensile prop with conventional aluminum alloys. Addi KEY WORDS tion of lithium, as an alloying element, erties of Al-Si alloys during cooling from decreases the density and increases the Al-Li-Cu-Mg Alloy the liquid state and during rapid heating elastic modulus of aluminum alloys (Refs. As-Cast 2091 Alloy from room temperature. Their results 1-3). Given the potential of this light Wrought Alloy 2091 show that the material exhibits lower weight, high-stiffness material for aero Mechanical Properties strength on cooling from the solidus. space application, a number of primary Elevated Temperature Evaluation of hot ductility is another Binary Al-Li Alloy method of determining the heat-affected Hot Cracking zone hot-cracking susceptibility of many T. ZACHARIA is with the Metals and Ceramics Nil Strength Division, Oak Ridge National Laboratory, Oak alloys. The test is based on the premise Hot Ductility Tests that the deformation behavior of a mate Ridge, Tenn. D. K. AIDUN is with the Depart Crack Sensitivity ment of Mechanical Engineering, Clarkson Uni rial during a given thermal cycle should versity, Potsdam, N.Y. reflect its cracking tendency. In principle,
WELDING RESEARCH SUPPLEMENT | 281-S Table 1—Alloy Composition (wt-%) HfiZ THERMAL CYCLE Element 2091-AC 2091-T3 2024-T3
Si 0.04 0.03 0.09 450 Fe 0.05 0.05 0.30 Cu 2.00 1.80 4.00 400 Mn 0.00 0.00 0.60 Mg 1.30 1.40 1.40 (_, 350 Cr 0.00 0.00 0.00 O Ul Zn 0.02 0.01 0.09 o 300 Li 2.30 1.70 0.00
UJ Zr 0.09 0.09 0.00 § 250 Ti 0.01 0.01 0.03 a- CT cr ui 200 OL 3; Ul •- ISO on the top surface of the specimen. The resulting thermal cycle is shown in ioo Fig. 1. Microstructural characterization was so done by optical and scanning electron
IIIII , i microscopy. Longitudinal sections from all Fig 1-Weld three heats were prepared from the metal HAZ 0 5 10 15 20 25 30 35 40 45 EO SS GO TIME. SEC specimens near the fractured tip for thermal cycle 1MflCE" metallographic inspection. All metallo graphic specimens were prepared using standard procedures and then etched a hot ductility test involves making hot duced center section to ensure that frac using Keller's reagent. tensile tests at preselected temperatures ture took place within the gauge during heating and cooling in an attempt length. to reproduce the strength and ductility of Elevated temperature tests were per Results and Discussion the material during welding. The segrega formed using a Duffers Scientific, Gleeble Fusion joining by a conventional weld tion present in the microstructure would Model 1500. The Gleeble is a pro ing process is a key issue in the develop allow the formation of low-melting con grammed thermal-mechanical testing ma ment and commercial exploitation of alu stituents that often wet grain boundaries chine that can be used to simulate the minum-lithium alloys. During welding, the and cause a loss of both mechanical thermal cycle associated with the GTAW weld metal and the adjacent heat-affect strength and ductility. The greater the process. The accuracy of the thermal ed zone are subjected to extremely high temperature range over which this liqua servo is ± 1°C (1.8°F) at equilibrium and temperatures. Figure 2 shows the numer tion phenomenon occurs, the greater the ±2°C (3.6°F) at 150°C/s (300°F/s) rate, ically computed temperature distribution susceptibility to hot cracking. per the manufacturer's specifications. It in the longitudinal and transverse section can also be used to strain specimens in Nippes, et al. (Ref. 8), developed the of a typical GTA weld made on an either tension or compression while technique of using the Gleeble thermal aluminum alloy (Ref. 12). The calculated being heated by resistance heating. A mechanical simulator to evaluate HAZ temperature gradients near the weld are two-wire chromel-alumel thermocouple crack susceptibility. The thermal mechan of the order of 2800°C/cm (12,283°F/ was percussion welded to the specimen ical simulator allows the modeling of in.). The relatively large temperature gra to provide required temperature control. rapid thermal cycles found in the heat- dients can generate thermal stresses and Elevated temperature tensile tests with a affected zone during fusion welding. strains that can ultimately cause hot- stroke rate of 0.635 mm/s (0.025 in./s) Both flat and cylindrical specimens have cracking in the fusion zone and the heat- were performed on cylindrical specimens been used in various studies of weldabili affected zone. Therefore, it is important for both heating and cooling conditions. ty and hot ductility (Refs. 9, 10). Recently, to characterize the elevated temperature Tests at preselected temperatures were Balaguer, et al. (Ref. 11), examined the behavior of the material under conditions performed immediately upon reaching hot ductility behavior of AI-4Mg-2Li and that are similar to those that occur during the temperature. AI-3.5Mg-4Zn alloys to evaluate their rel welding. The present investigation focus ative crack susceptibilities. Their tests The determination of the HAZ es on the elevated temperature behavior show that the lithium-containing alloy has mechanical properties involved produc of the material pertaining to the heat- a relatively lower ductility and, therefore, ing a microstructure, in the Gleeble spec affected zone. higher crack susceptibility. imen, typical of a portion of the HAZ of the weldment. Autogenous bead-on- Nil Strength Temperature plate welds were made on 250- X 64- Experimental Procedure X 12-mm (10- X 2.5- X 0.5-in.) speci For the present investigation, the peak The materials used in this investigation mens in a He atmosphere. Welding was temperature, for the Gleeble thermal comprise two heats of Alloy 2091, one carried out by a Hobart, Cyber Tig 100 cycle, was chosen as the nil strength each, in the as-cast and in the T3 condi series, welding machine at 170 A, 14 V temperature (NST). The NST of the alloy tion and one heat of Alloy 2024 in the T3 and 3.4 mm/s (8 ipm), using 3.18-mm 2% is the temperature at which the material condition. Compositions of these alloys thoriated tungsten electrode. The weld loses all strength and ductility. NST was are given in Table 1. Cylindrical speci HAZ thermal cycles were obtained using determined by placing a small constant mens of 6-mm (0.24-in.) diameter were a Digital Equipment Corporation, MINC load of 10 Ib (4.5 kg) on the specimen machined from all three heats for the (modular instrument computer) and and heating at the predetermined heating Gleeble tests. For test temperatures chromel-alumel thermocouple. Two ther rate (corresponding to measured HAZ below 350°C (660°F), specimens from mocouples were located in the HAZ, 10 thermal cycle) until the specimen frac each heat were machined with a re mm and 9 mm from the weld centerline tured. The temperature at which the
282-s I DECEMBER 1988 specimen fractured was taken as NST. The average values of NST for Alloy 2091 in the as-cast and T3 condition are 550°C and 600°C (1022° and 1112°F), respectively. The eutectic temperature for the aluminum-lithium alloy is reported to be close to 520°C (968°F) (Ref. 13). Differential thermal analysis of Alloy 2091-T3 indicates the solidus tempera ture to be approximately equal to 540°C (1004°F). The experimentally observed
NST for Alloy 2091 is well above these -2.0 O.D 2.0 temperatures and may have been associ r-H0RIZ0NTRL-.iXIS (mm I ated with some partial melting at the temperature. The NST for Alloy 2024 is 7.0 - close to 550°C. S.O
Elevated Temperature Strength S.O
The elevated temperature strength of o the two heats of aluminum-lithium Alloy 4.0 ^y^y o 4? i>> 2091 and Alloy 2024 was determined j 3.0 r> r\ . • SS2 1 o during heating to the peak temperature o (on heating) and during cooling from the 1 2.0 •*. I X i 1 , peak temperature (on cooling). First, the -Z.0 0.0 2.0 relative elevated temperature strength of X-HORIZONTRL-flXIS [mm 1 the three heats "on heating" are present ed in Fig. 3. The results show that, for the Fig. 2 — Typical temperature distribution in the longitudinal and transverse section of an autogenous range of temperatures investigated, the GTA weld ultimate tensile strength of Alloy 2091 compared very well with Alloy 2024. This is particularly significant since Alloy 2091 ture regime. It is probably due to the an understanding of the mechanical is developed for possible replacement of large grain size of the cast material. Figure behavior of the material in this tempera Alloy 2024 in structural applications. In 4 is a typical photomicrograph of the ture regime is particularly important to the T3 condition, the room temperature base metal structure for the alloy in the understand the HAZ crack sensitivity of strength of Alloy 2091 is approximately as-cast condition, indicating a very coarse the material (Ref. 18). A thermal cycle, 425 MPa (61.6 ksi) and that of Alloy 2024 dendritic microstructure. The dark areas with the experimentally determined heat is approximately 475 MPa (68.9 ksi). The in the microstructure indicate the lithium ing and cooling rate and a peak tempera room temperature strength has been rich 5 (AILi) phase. ture equaling the NST, was imposed on measured previously for the Soviet- During welding, the weld HAZ experi the gauge section of the Gleeble speci developed 01420 (AI-5Mg-2Li-0.1Zr) al ences temperatures far in excess of the men. Elevated temperature mechanical loys (Ref. 14) and 8090 (AI-2.5Li-1.2Cu- transformation temperature (approxi properties were determined at the test 0.7Mg-0.12Zr) alloys (Ref. 15). These mately 350°C) of the material. Therefore, temperature for all three heats during the studies reported room temperature strength in the range of 400-425 MPa. 500.0 The elevated temperature mechanical properties have been measured for Al- Li-Mg (Refs. 11, 16) and Al-Li-Co (Ref. 17) alloys. These studies indicate that the o 2091-AC 400.0 • 2091-T3 elevated temperature strength of these • 2024-T3 alloys decrease rapidly, at temperatures exceeding 100°C (212°F), when these alloys contained zirconium. This rapid reduction in strength is attributed to O 300.0 dynamic recovery caused by the move CO £ ment of dislocations to sub-boundaries in the alloy (Ref. 16). The results shown in co Fig. 3 show that the 2091 alloy retains its I- high strength until about 200°C (392°F). •3 200.0 Beyond this temperature, the strength decreases rapidly to a minimum mea sured value of 33 MPa (4.8 ksi) at 550°C. The improved performance of the 2091 100.0 alloy is apparently due to the presence of the T1 (AI2CuLi) and the S' (Al2CuMg) phases at the subgrain boundaries and
therefore, delaying the dynamic recov 0.0 ery. The elevated temperature strength 0.0 100.0 200.0 300.0 400.0 500.0 600.0 of Alloy 2091 in the as-cast condition was Temperature (Deg C) relatively lower throughout the tempera- Fig. 3-On-heating hot tensile test results for Alloys 2091-AC, 2091-T3 and 2024-T3
WELDING RESEARCH SUPPLEMENT I 283-s heat up (on heating) and the cool down (on cooling) phase of the thermal cycle. t V -yr ' \.y/Ayy\ This was done to obtain a complete history of the variation in strength of weld HAZ during welding. a. Figures 5-7 show the influence of the O test temperature on the elevated tern-'- UJ perature strength of the three heats. In > general, for all three heats, the elevated temperature strength measured on cool a ing, as expected, was lower than the sc strength on heating. Alloy 2091 in the < as-cast condition exhibited relatively poor UI CO fig. 4 — Typical strength for all the test conditions —Fig. 5. UJ photomicrograph During the heating cycle, the tensile CC of the as-cast *-». strength dropped rapidly from 100 MPa structure of Alloy V C 2091 • - ' T, . \ *~ 100 fm (14.5 ksi) at 350°C (662 F) to approxi 2 mately 35 MPa (5.1 ksi) at 500°C (932°F). a. The results show that during the cooling o 200.0 cycle the material recovered only about 50% of the on-heating tensile strength. > The elevated temperature strength of Ul • On Heating the wrought Alloy 2091 in the T3 condi Q o On Cooling tion is shown in Fig., 6. During the on- O 150.0 heating cycle, the strength drops rapidly from 150 MPa (21.8 ksi) to about 30 MPa Ui (4.4 ksi) at 550°C. Barring some scatter in ui the data, the results show that between ac CO a. 450°C (842°F) and the NST, the elevated z S- 100.0 temperature strength measured on heat ui CO ing and cooling did not differ significantly. E For the test conditions considered in the a. o present investigation, the maximum dif _l ference in strength of about 20 MPa (2.9 ui ksi) was observed at 350°C. > 50.0 The result of the elevated temperature X strength of wrought Alloy 2024-T3 is o NST CK shown in Fig. 7. In general, Alloy 2024-T3 «c had relatively higher strength for all the Ul to T test conditions. However, the results Ul o.o show that the elevated temperature a: 300.0 350.0 400.0 450.0 500.0 550.0 strength measured on cooling is consider Temperature (Deg C) ably lower than the strength measured Fig. 5-Hot tensile test results for Alloy 2091-AC Ul on heating. The maximum difference in 5 a. strength between the on-heating and the O 200.0 on-cooling test was close to 125 MPa _i ui (18.1 ksi) at 350°C. This is significantly > larger than the difference in strength for Alloy 2091-T3, suggesting that the weld • On Heating X o On Cooling HAZ of Alloy 2024 may be more suscep u tible to hot cracks. 150.0
284-s I DECEMBER 1988 in temperature, the ductility decreased rapidly to about 40°o at 500°C. This is • probably associated with the melting of • the low melting (8) constituent along the • On Heating \ grain boundary. Optical metallography c On Cooling 5 indicated partial melting of the alloy at elevated temperatures near the NST. The o ui Gleeble hot ductility test depends on > producing an isothermal plane, within the Ui gauge length, perpendicular to the load ca ing axis. The heat transfer from the gauge Q- X section is predominantly along the axis. =• 100.0 o co OS Figure 9 is a photomicrograph of the 1- < 3 LU longitudinal section of the fractured spec co UJ imen tested close to the NST. The region £ in the center, indicated by the arrow, shows small dendrites along this axis 50.0 where melting and subsequent solidifica z tion occurred at the grain boundary triple ui NST point. Fracture typically occurred inter- s granularly due to liquation and failure of o. o the grain boundary constituent. _l 650.0 ui The on-cooling hot ductility results for 300.0 350.0 400.0 450.0 500.0 550.0 600.0 > Temperature (Deg C) Alloy 2091-AC show no recovery of Ul Q ductility. For all the test conditions, Alloy Fig. 7-Hot tensile test results for Alloy 2024-T3 O CC 2091-AC shows zero ductility during the < cooling phase of the thermal cycle. Figure Ul 100.0 CO 10 is the photograph of a typical speci Ul men at the fracture location, indicating a oc so.o brittle intercrystalline fracture. Examina • On Heating tion of the surface indicates considerable o On Cooling oxidation and a very porous surface tex s ture. The complete loss of ductility, as the Q. 70.0 o material is cooled from the NST, is prob • _l ss Ul ably due to the deterioration in the sur a) • y,y^**\ 60.0 - > face of the specimen. Figure 11 is the a> Ul longitudinal section, near the surface, of < a the specimen shown in Fig. 10. The _c 50.0 - / • c oc microstructure shows the surface degra o • < dation due to oxidation and depletion of 40.0 Ul o • / • co the lithium-rich (<5) phase. 3 ui TJ oc Figure 12 shows the hot ductility 30.0 response of the wrought Alloy 2091 in • 20.0 Ul the T3 condition. The room temperature 2 ductility of the alloy, as expected, was NST a low (about 10%). This is attributed to the 10.0 o _i coherent &' (AI3L1) precipitates, which ui impede the movement of dislocations. e— 1 > 0.0 9 9 9— 1— i Ul The ductility increased with increasing 300.0 350.0 400.0 450.0 500.0 550.0 600.0 650.0 temperature to a maximum value of 96% Temperature (Deg C) a at 450°C. The ductility did not decrease Fig. 8-Hot ductility test results for Alloy 2091-AC u from this maximum value with further CE increase in test temperatures beyond < Ul D Fig. 9- 450 C. CO Photomicrograph Ul Unlike the as-cast material, the of the longitudinal oc wrought alloy exhibited a recovery of section of the 70% ductility at 550°C. The lower ductili fractured specimen ty in comparison to the on-heating test at (Alloy 2091-AC. s CL this temperature is probably due to tested "on o cooling'• at 500°C) _l liquated grain boundary (5) constituent. Ul The ductility further increased with > decreasing temperature to a maximum of 95% at 400°C. Figure 13 is a photograph X of the fracture surface indicating the o oc tendency of the material to delaminate at < elevated temperature. This is particularly Ul CO evident from Fig. 14. The specimen in this Ul case was subjected to a compressive 0£ stress resulting in delamination and frac ture. Delamination typically occurs as a
WELDING RESEARCH SUPPLEMENT | 285-s i- z ui 2 a. • • ^ O -i >Ui Ul Cl X o " • oc < Ul CO ui sc ui 2 o. o _l >ui Fig. 10— Typical photograph of the fracture surface for Alloy 2091-AC Fig. 11 — Photomicrograph of the longitudinal section, near the surface, (tested on cooling) of the specimen shown in Fig. 10. X o SC result of the rolling operation, which < 100.0 Ul causes an elongated "pancake" structure co in the rolling direction. At elevated tem Ul 90.0 SC peratures, the low-melting phase forms a • On Heating continuous liquid film parallel to the roll \ o On Cooling I- 80.0 ing direction, causing the material to z delaminate. 2LU a. ~s 70.0 Scanning electron microscopy (Fig. 15) o of the test specimen at 450°C for the —I ca ui
286-s I DECEMBER 1988 I- Z ui 2 oa _i U>J Ual X o oc < LU CO
z U2J Fig. 14 — Photograph of Gleeble sample tested in compression indicatingFig. 15 — Photomicrograph of the longitudinal section of the fractured a. the tendency of wrought Alloy 2091 to delaminate specimen (Alloy 2091-T3, test done on heating at 450°C) O _i >LU Ul a : ', .. " " • I T ~-™ *-*. x o oc < Ul CO
0. o
X o oc < :: Ul CO UJ Fig. 16 — Photomicrograph of the longitudinal section of the fractured Fig. 17 — Photomicrograph of the longitudinal section of the fractured OC specimen (Alloy 2091-T3, test done on cooling at 550°C) specimen (Alloy 2091-T3, test done on cooling at 350''Q
LU 2 a. ductility was not as rapid as that exhibited 100.0 O by Alloy 2091 in the T3 condition. Fifty >Ul degrees below the nil strength tempera eo.o LU ture, Alloy 2024 exhibited a ductility of • On Heating a 15% in comparison to 75% for Alloy eo.o o On Cooling x 2091. o oc SEM secondary electron image of the 70.0 < Ul longitudinal section of the fractured spec CO imen shows a coarsening of the grain Ul o 60.0 boundary constituent with increasing test < oc temperatures up to 450°C —Figs. 19 and 50.0 20. Further increase in test temperatures z c LU results in the dissolution of the precipitate o 2 in the bulk —Fig. 21. However, there 40.0 oa. were isolated pockets of large frag T3 mented second-phase constituent. The 30.0 > ui low ductility exhibited by Alloy 2024 at a 500°C on cooling is probably due to the 20.0 liquation of these isolated pockets at the o NST oc higher temperatures. 10.0 < The strength-to-ductility ratio of Alloy Ul CO 2024-T3, at the highest temperatures, is 0.0 Ui considerably larger than that for Alloy 300.0 350.0 400.0 450.0 500.0 550.0 600.0 oc 2091-T3. This together with the signifi Temperature (Deg C) cantly lower ductility of the Alloy 2024- Fig 18-Hot ductility test results for Alloy 2024-T3
WELDING RESEARCH SUPPLEMENT | 287-s Fig. 19 — Photomicrograph of the longitudinal section of the fractured Fig. 20 — Photomicrograph of the longitudinal section of the fractured specimen (Alloy 2024-T3, test done on heating at 350°C) specimen (Alloy 2024-T3, test done on heating at 450CC)
T3 at these temperatures indicates that heating and on-cooling test for Alloy T. H. Sanders, Jr., TMS-AIME, Warrendale, Pa. Alloy 2024 is more susceptible to HAZ 2024 was 125 MPa at 350°C in contrast 1984, p. 206. cracking. Alloy 2091-T3, on the other to 20 MPa for Alloy 2091 at the same 4. Cieslak, S.)., Hart, R. M., Mehr, P. L, and hand, is very ductile at temperatures temperature. Mueller, L. N. 1985. Presented at the 17th international SAMPE technical conference, close to the solidus temperature, suggest 4) The results of hot ductility test show October 22-24, Kiamesha Lake, N. Y. ing that the material is resistant to hot- that Alloy 2091 in the wrought form 5. Archbutt, S. L, Grogan, |. D., and lenkin, cracking in the weld metal HAZ. exhibits superior ductility throughout the ). W. 1928. /. Inst. Metals 40:219. range of temperatures investigated. In 6. Singer, A. R. E., and Cottrell, S. A. 1946. Conclusions particular, 50°C below the NST, Alloy Jour. Inst. Met. 72:33. 2091-T3 exhibits excellent recovery of 7. Pumphrey, W. I., and Jennings, P. H. A detailed experimental study was car ductility (70%) in comparison to Alloy 1948. Jour. Inst. Met. 75:235. ried out to characterize elevated temper 2024-T3 (15%). This would indicate lower 8. Nippes, E. F., Savage, W. F., Bastian, B. J., Mason, H. F., and Curran, R. M. 1955. Welding ature mechanical properties of aluminum- weld metal HAZ crack susceptibility for Journal 34:183-s. lithium Alloy 2091. Results and conclu Alloy 2091. 9. Savage, W. F„ and Krantz, B. 1966. sions are summarized as follows: Welding Journal 45:13-s. 1) In general, Alloy 2091 in the as-cast Acknowledgment 10. Canonico, D. A., Savage, W. F., Weren- condition exhibited very low elevated er, W. |., and Goodwin, G. M. 1969. WRC The authors would like to thank Mr. R. temperature strength and ductility. The publication, luly, p. 68. P. Martukanitz of Alcoa Technical Center, relatively poor strength and ductility is 11. Balaguer, |. P., Nippes, E. F., and Walsh, for the material and technical support apparently due to the coarse dendritic D. W. 1986. Proceedings of an International provided during this investigation. structure of the material. Conference in Welding Research. Gatlinburg, 2) The hot tensile tests indicated that Tenn., p. 723. tensile strength of wrought Alloy 2091 References 12. Zacharia, T., Eraslan, A. H.. and Aidun, D. K. 1988. Welding Journal 67':18-s. adequately matched the strength of Alloy 1. Wald, G. C. 1981. NASA Contractor 13. Martukanitz, R. P. 1987. Aloca Technical 2024 throughout the temperature regime. report 16576, NASA, Washington, D. C. Center, Pittsburgh, Pa. 2. Sankaran, K. K., and Grant, N. ). 1980. This is particularly encouraging since Alloy 14. Pickens, |. R., Langan, T. )., and Barta, E. Proceedings of the 1st International Aluminum- 2091 was developed for replacing Alloy 1986. Proceedings of the Third International Lithium Conference. Stone Mountain, Ga. 2024 in aerospace applications. Aluminum-Lithium Conference. University of Edited by T. H. Sanders, )r. and E. A. Starke, Jr., Oxford, luly 1985. Edited by C. Baker, P. I. 3) The elevated temperature strength TMS-AIME, Warrendale, Pa., 1981, p. 206. measured for Alloy 2024 "on cooling" Gregson, S. J. Harris, and C. J. Peel, The 3. Quist, W. E., Narayanan, G. H., and Institute of Metals, London, England, p. 137. was considerably lower than the strength Wingert, A. L. 1983. Proceedings of the 2nd 15. Pridham, M., Noble, B., and Harris, S. |. measured "on heating". The maximum International Aluminum-Lithium Conference. 1984. Proceedings of the 2nd International difference in strength between the on- Monterey, Calif. Edited by E. A. Starke, )r. and Aluminum-Lithium Conference. Monterey, Calif. Edited by E. A. Starke, lr. and T. H. Sanders, Jr., TMS-AIME, Warrendale, Pa. 1984, Fig21- p. 547. Photomicrograph of 16. Noble, B„ Harlow, K., and Harris, S. ). the longitudinal 1984. Proceedings of the 2nd International section of the Aluminum-Lithium Conference. Monterey, fractured specimen Calif. Edited by E. A. Starke, Jr. and T. H. (Alloy 2024-T3, test Sanders, Jr., TMS-AIME, Warrendale, Pa. 1984, done on heating at p. 65. c 550 'Q 17. Sastry, S. M. L, and O'Neal, J. E. 1984. Proceedings of the 2nd International Alumi num-Lithium Conference. Monterey, Calif. Edited by E. A. Starke, Jr., and T. H. Sanders, Jr., TMS-AIME, Warrendale, Pa. 1984, p. 79. 18. Aidun, D. K., Zacharia, T., and Martu kanitz, R. P. Hot cracking susceptibility of Al-Li-Cu-Mg alloy. Fourth International Confer ence on Aluminum Weldments, INALCO '88, Tokyo, Japan, April 4-12, 1988.
288-s I DECEMBER 1988