Mold-metal reactions in magnesium investment castings

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Doctoral thesis (monograph)
Checking the digitized thesis and permission for publishing
Instructions for the author
Degree programme
Helsinki University of Technology publications in foundry technology, Teknillisen korkeakoulun Valimotekniikan laboratorion julkaisuja, 1/2006
Mold-metal reactions can be encountered during the investment casting of magnesium alloys. This study was carried out for investigating the degree of reactions between the various refractory materials and the magnesium alloy AZ91E, and for finding new techniques to reduce these reactions. Investment casting molds containing multiple test pieces with the dimensions of 25x25x60mm in dimension were used. The wax pattern of each test piece was coated with a different ceramic face coat in mold fabrication and the resulting differences on cast metal surfaces were studied. The backup layers of the shell were the same for all test pieces. Fused alumina, fused silica, molochite, zircon, yttria, zirconia and fused magnesia were the ceramics studied as face coat materials. Digital pictures of the cast surfaces were taken and an image analyzer was used to a quantitatively assess the reacted areas. The results show that fused magnesia and yttria were the best face coat materials to resist molten magnesium. Fused alumina and zircon were the next best materials. Molochite and zirconia were ranked as moderate to poor in resisting reactions. Strong mold-metal reactions were seen in the case of fused silica. Shell permeability measurements suggest that the degree of reactivity is not related to the shell permeability. It can however be correlated to the free energy of formation of the refractory. The excellent reaction resistance of magnesia and yttria observed in the experiments can be explained by the strongly negative Gibb's free energy of formation and consequent stability of these materials. HSC software was used to calculate theoretically the free energy of formation for the refractories studied in this work. The obtained results were compared to experimental values. Analysis with EDS indicated that the reaction layers consisted mostly of oxides. Ceramic test bars coated with various refractories were dipped in magnesium melt to study for reactions. It was seen that none of the tested refractories reacted with magnesium as a result of the dipping operation. It was concluded that oxygen is required for the mold-metal reactions to be initiated. It was verified that the atmosphere surrounding the shell also influenced the observed reactions. In molds containing both normal and reduced permeability shells around the test pieces, the surfaces cast in reduced permeability shells showed large reductions in reactivity. It is suggested that reactions in investment casting molds occur in two stages: 1. During pouring and filling 2. After the filling is complete by the effect of external oxygen. The cooling curves of magnesia-coated test bars were compared with those of silica-coated test bars. Analysis shows that heat is released during the mold-metal reactions. An attempt was made to incorporate the inhibitor KBF4 in the mold structure in the form of first coat stucco. Rough cast surfaces were obtained, which indicated that the used high sintering temperature was not compatible with this inhibitor. Similarly, when the preheated investment-casting mold was placed in a bed of KBF4 before casting, only small reductions in reactivity were obtained. Reactivity was reduced in shells buried in a bed of Croning sand. Then, however, the cast pieces contained gas holes. NaBF4 produced better results as an inhibitor. It was first dissolved in water and then the sintered molds were dipped in this solution. Large reductions in reactivity were observed. The inhibiting effect of this chemical can be attributed to the liberation of BF3 gas during mold preheating.
magnesium, investment casting, mold-metal reactions
Other note
Permanent link to this item