CASTABILITY OF Al-Li-Mg AND Al-Li-Cu-Mg ALLOYS

Transcription

1 CASTABILITY OF Al-Li-Mg AND Al-Li-Cu-Mg ALLOYS C. Tong, L. Yao, C. Nieh, C. Chang, S. Hsu To cite this version: C. Tong, L. Yao, C. Nieh, C. Chang, S. Hsu. CASTABILITY OF Al-Li-Mg AND Al-Li-Cu-Mg ALLOYS. Journal de Physique Colloques, 1987, 48 (C3), pp.c3-117-c < /jphyscol: >. <jpa > HAL Id: jpa Submitted on 1 Jan 1987 HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

2 JOURNAL DE PHYSIQUE Colloque C3, suppl6men-t au n09, Tome 48, septembre 1987 CASTABILITY OF Al-Li-Mg AND Al-Li-Cu-Mg ALLOYS C.H. TONG, L.G. YAO, C.Y. NIEH, C.P. CHANG and S.E. HSU Chung shan Institute of Science and Technology, P.O. Box , Lung-Tan, Taiwan 32500, Eepubfic of China ABSTRACT The objective of the present work is to study the casting characteristics of various A1-Li alloys, which include fluidity and strengths of the alloys and their interaction with cast molds. Materials investigated are Al-Li-Mg and Al-Li-Cu-Mg alloys with Li content of 2.5 wt%. The results show that sand molds with resin binders are good for A1-Li casting. Ceramic coatings can further reduce the metal-mold interactions. However, the permeability is also reduced by coating. The fluidity of Li-bearing aluminum alloys is inferior to that of A356. Compared with copper, magnesium gives a better combination of fluidity and strength to A1-Li alloys. Al-Li-Cu, together with Al-Li-Mg, are capable of providing a wide range of strength-ductility combination to casting designs. Al-Li-Mg sand cast also possesses around 80% of the strength of its wrought counterpart. INTRODUCTION Although considerable progress has been made in the development of high strength and damage tolerant aluminum lithium alloys to the extent that commercial grade alloys are available in recent years, there are few works reported in the field of cast alloy development, Considering the inherent isotropy in mechanical strength, the potential for good material utilization, and the convinence in making large components, the castability of aluminum lithium alloys is worth studying. A good aluminum cast alloy should, among other things, be compatable with the cast mold, and have enough fluidity and mechanical strength to make useful components. It should also have minimum shrinkage and gas induced porosity and minute solute segregation tendency. Unfortunately, lithium containing aluminum alloys seem to possess none of the above properties. Therefore, it is difficult to make a cast which retains all the merits of aluminum lithium alloys, namely high strength, high stiffness, and low density. In the present work, aluminum lithium alloys with various amount of magnesium and copper are investigated. Melt-mold interaction is studied first, aiming at developing some sand treating techniques which can effectively eliminate the interaction induced defects in casts. A specially designed cast mold, which had surface layers composed of inserts of treated sand in the form of biscuit was used in comparing the relative interaction tendency of the materials with the A1-Li alloy. The least-reactive sand materials, which was chosen based on X-ray radiography and scanning electron microscopy evaluations, was then used to construct molds for spiral fluidity tests and tensile specimen casting. EXPERIMENTAL PROCEDURE Since magnesium and copper are the major alloying elements used in aluminum lithium alloys, it is decided to explore the alloys with 2.5 wt% Li, 0-5 wt% Mg Article published online by EDP Sciences and available at

3 30URNAL DE PHYSIQUE Table 1 Composition and designation of alloys Figure 1. A schematic drawing of the mold used in melt-mold investigated. interaction study. In order to min~mize the other possible sources of reaction with the melt, the melt-mold interaction experiments were conducted in an induction furnace under a protective atmosphere. A schematic drawing of a half of the specially designed test mold is shown in Figure 1. Besides the sand biscuit inserts, part of the mold surface is made of cast iron which serves as a reference check of possible misconducts in melting practices. All mold materials were bakedin a vacuum oven before testing. Biscuits were made in a mold in which sand, after bing added with binder, was pressed into shape. Sodium silicate or resin were used as the binder. Some biscuits were coated with silicon carbide, alumina or zircon (ZrSio,) paste bonded by silica gel to form a protective layer. The interaction was measured by a combination of X-ray radiography, optical microscopic examinations of the cast seructure near surfaces and SEM analyses of possible reaction products on the mold surfaces. Spiral fluidity tests were conducted in air using resin bonded sand mold. After melting and degasing of the aluminum alloy, separately melted pure lithium was injected into the ladle before casting. Pouring temperatures of 710 OC to 730 OC were used in the experiment. The ladle and mold were lavishly flushed with argon to reduce the oxidation of the melt. Ingot for mechanical property evaluations were prepared in the same furnace used for interaction studies, except that resin bond sand were used as the liner in the entire mold. ASTM subsize plate specimens with an 1" gauge length were used. Each specimen was subjected to a T6 treatment. Peak aging condition based on hardness test was used. Tensile tests were conducted on an Instron machine with a 2 mm/min. strain rate.

4 Figure 2. Cross-sectional view of the ingot cast into (a) an iron mold, (b) a resin bonded sand mold. 0.lm - RESULTS AND DISCUSSION Melt-Mold Interaction Visual inspections of the as-cast ingot surfaces indicete that there is almost no interaction of the alloy with iron mold. On the contrary, interactions with sodium silicate bonded sand are severe. Underneath the surface layer which was in contact with the silicate bonded sand, large size gas bubbles were detected. However, not even tiny bubbles could be found on the part of the ingot contacted with iron. The melt-mold interaction is drastically reduced in a sand mold when sodium silicate is replaced by resin. Although resin reacts with the melt to form a thin layer of black substance containing carbon, there is no detectable defect resulting from the interaction. Both X-ray radiography and optical microscopic examination indicate that the cast is as sound as those out of an iron mpld. Optical micrography of cross-sectional view of the ingot are depicted in Figure 2. It is noted that in regions contacted with iron columnar structure is developed. No such structure was observed in regions contacted with sand. Also note that the interfaces are fairly smooth and free from defect. The major difference between the two binders used in this experiment is judged to be the water content. Though the same vacuum baking procedure was used for both mold preparations, the crystalized water in silicate apparently remains in the sand to cause violent reaction with the melt. Figure 3. EDAX chemical analysis of the zircon mold surface (a) prior to casting, (b) after interaction.

5 C3-120 JOURNAL DE PHYSIQUE The black surface layer does not form in resin bonded sand molds when they are coated with ceramic materials, such as ZrSiO,, A1,0, or Sic. Among them, A1,03 and Sic show great inertness to the melt. ZrSiO, coating also resists the reaction to a great extent, but EDAX analysis of the mold surface reveals that there are some pickup of megnesium by the mold, as shown in Figure 3. The problem associated with the ceramic coating is that they reduce the permeability of the sand. X-ray radiographs of a cast ingot from a mold consistes of iron, resin bonded sand, and Al,03-coated sand reveals that entrapped gas in the melt preferentially forms bubbles in the region where mold surface is coated with alumina. Fluidity Test In order to measure the fluidity of the alloys under foundry environment, the tests were conducted in air with lavish argon flush above the melt pool and inside the mold. Test results are listed in Table 2. A reference material, A356, is also tested for comparsion. It is noted that the fluidity of lithium containing alloys are inferior to A356, which has very good castability. In spiral fluidity measurement, viscosityand the amount of superheat in the melt are important factors needed' to be considered. Magnesium tends to reduce the liquidus temperature of the melt at a higher rate than copper. This is in consistance with the present result. Since it is reported that Al-Li-Cu-Mg has less oxidation resistance than Al-Li-Mg (I), addition of Cu should also increase the viscosity of the melt to a greater extent. I Elongation 1 " I 2 3 wtt cu 2 Figure 4. Mechanical properties of (a) Al-Li-Mg, and (b) Al-Li-Cu-Mg cascing ingots after T6 treatment. Mechanical Properties After T6 treatment, the mechanical properties of magnesium strengthened alloys show a monotonic increase in strength with magnesium content, as illustrated in Figure 4 (a), while the ductility maintains constant up to 4 wt% mag-

6 nesium. From Al-Li-Mg ternary phase diagram (2) it is found that some AlxMgy phases will form in the CA-5 alloy. The formation of A1,Mgy phase and their effect on ductility was also reported for an alloy with the same composition, but after repeated hot and cold rolling (3). When copper is added to the aluminum lithium alloy, the strength of the alloy rapidly increases, as shown in Figure 4(b). Among them, CB-1 has similar chemical composition to Alcan's 8090 alloy. Its yield strength and ductility comes close to those of CA-5 alloy. As copper content increases, the strength also increases, but at the expense of a ductility loss. From the data obtained in this experiment, it seems that, CA alloys together with CB ones can provide a series of alloys with various strength and ductility. There is no apparent advantage of one over the other. It is always interesting to find out the difference in strength between wrought and cast alloys. A comparison is made for CA-5 alloy in Figure 5. The same T6 heat treatment is used. The figure shows that a casting can have 80% of the strength of its wrought counterpart, very close to the strength of a good weldment in aluminum alloys. Table 2. Spiral fluidity test (MPa) results of the lithium containing aluminum alloys t*) ZOO wrought Iron Mold Sand Mold Figure 5. Comparison of the mechanical properties of wrought and cast Al-Li-Mg alloy. SUMMARY AND CONCLUSIONS 1. Iron and resin-bonded sand are suitable mold materials for aluminum lithium casting. 2. Alumina, zircon, or Sic, when coated on sand, can reduce the surface scale of the cast. However, they also affect adversely the permeability of the sand. 3. The fluidity of the lithium containing aluminum alloys is inferior to that of A356 in general. 4. Compared with copper, magnesium can provide a better combination of fluidity and strength to aluminum-lithium alloys. 5. Al-Li-Mg together with Al-Li-Cu-Mg can provide a series of alloys with various strength and ductility. 6. Al-Li-Mg sand cast possesses around 80% of the strength of its wrought counterpart.

7 JOURNAL DE PHYSIQUE REFERENCES 1. W.S. Miller, A.J. Cornish, A.P. Titchener and D.A. Bennett, in "~luminum- Lithium Alloys II", Ed. T.H. Sanders and E.A. Starke, TMS, TIME, Warrendale, Pa, 1984, pp: E. Schurman and I.K. Geissler: Giessereiforschuna, , 32, L.G. Yao, C.H. Tong, and Y.F. Hsu, in 'Proc. of the. l9sj Annual Conference of the Chinese Soc. for Mat. Sci.', Mat. Res. Labl, Chutung, Taiwan, pp. 299.