Search

The non-isothermal crystallization behaviors of CaO-BaO-Al 2 O 3 -based mold fluxes in which SiO 2 had been replaced by Na 2 O were investigated, using an infrared furnace combined with digital microscopy as well as X-ray diffraction. The initial crystallization temperatures, crystalline phases generated at different cooling rates, crystallization rates, Ozawa indices N O (as determined using Mo's equation) and effective crystallization activation energies (as determined using the Friedman equation) were evaluated. The initial crystallization temperature increased along with the Na 2 O content in the flux. In addition, the precipitation of CaF 2 crystals was inhibited while the growth of CaAl 2 O 4 crystals was enhanced at Na 2 O concentrations over 4 wt%. Increasing the cooling rate inhibited the precipitation of MgAl 6 O 10 , and of CaSiO 3 , MgAl 2 SiO 6 and NaAlSi 3 O 8 , respectively, in fluxes containing 0, 4 and 8 wt% Na 2 O (NS1, NS2 and NS3 samples). The peak instantaneous and average crystallization rates in these specimens were increased as the Na 2 O level was raised, as a direct result of changes in their N O values. The NS1 flux consistently had the lowest N O value, while the NS2 had the largest value in the primary stage of crystallization but the central value in the secondary stage, and the N O values of the NS3 flux exhibited the opposite trend. The effective crystallization activation energies decreased with increasing Na 2 O levels, and this was evidently an important factor affecting the initial crystallization temperature. [ABSTRACT FROM AUTHOR]

The complete crystallization process of mold slag in the crucible (inner diameter: 6.2 mm) was successfully recorded in situ using a digital optical microscope and image processing software in an infrared furnace (IR‐MOP). The crystallization behavior of non‐volatile mold slag tested with the new method is consistent with the findings of differential scanning calorimetry (DSC) analysis. Although the volatilization ratios of volatile mold slag during both DSC and SHTT experiments are greater than 8%, it is less than 1.2% during IR‐MOP experiments equipped with an isolated observation system, which explains why the difference in the initial crystallization temperature measured by the three methods was very large. Therefore, IR‐MOP equipped with an isolated observation system can record and accurately measure the crystallization behavior of volatile mold slag. [ABSTRACT FROM AUTHOR]