No abstract

A single-pan scanning calorimeter has been developed that eliminates the smearing of latent heat that occurs in a conventional two-pan heat-flux differential scanning calorimeter (DSC). In the new calorimeter, accurate enthalpy/temperature data was obtained in pure Al without smearing, and excellent sensitivity to new phases was obtained in a multicomponent Al alloy (LM25). The calorimeter has been used to investigate microsegregation in an Al-4.45 wt pct Cu alloy. The enthalpy/temperature data fell between that calculated, assuming no mixing in the solid (Scheil) and complete mixing in the solid (equilibrium solidification). The amount of segregation agreed well with that calculated using a diffusion-based model of microsegregation. The difficulty of getting the fraction solid from the enthalpy data is discussed, and it is concluded that it is not possible to do so without using a microsegregation model. In addition, it is concluded that it is wrong to assume that the enthalpy of an alloy can be given by a specific heat term and a constant latent heat term that depend on fraction liquid as is assumed in most casting models.

A comparison has been made between the amount of microsegregation predicted by a numerical model and that found experimentally in Al-Cu alloys varying in composition between 1 and 8 wt pct Cu. A depleted region was predicted and observed experimentally near the Al 2 Cu. The depleted region was formed below the eutectic temperature and had a significant effect on the ordered compositionfraction plots, particularly for high alloy compositions. Although the fit between experiment and theory was reasonably good, it was concluded that it was necessary to propose that local equilibrium was not maintained between the phases in the solid-state reactions.