DIRECT OBSERVATION OF AUSTENITE FORMATION AND DECOMPOSITION IN 4118 AND 4320 STEELS

摘要

Experiments utilizing a hot-stage confocal scanning laser microscope (CSLM) have been carried out to observe phase transformations in two steels, 4118 and 4320. Austenite formation during continuous heating and isothermal conditions were investigated on the surface of samples that were etched to reveal the ferrite and eutectoid regions. It was found that the austenite precipitated at eutectoid/ferrite boundaries. Austenite would then proceed to cover the eutectoid regions first and thereafter a distinguishable γ/α front would proceed advance through the ferrite. Anywhere from 2 to 5 fronts could be observed in each ferrite grain, but did not originate preferentially based on surrounding microstructure. In most cases, only one austenite front was observed to advance through a eutectoid colony. The rates of front advancement occurred at higher temperatures and were slightly faster in 4118 than 4320 when subjected to non-isothermal growth conditions. Non-isothermal growth rates were generally non-linear. Front advancement rates under isothermal conditions were generally linear, were faster in 4320 than 4118, and displayed a first order relationship to superheat (T - T_e) where T_e is the eutectoid temperature. Austenite decomposition was studied during continuous cooling and isothermal conditions. Precipitation of both allotriomorphic ferrite and Widmanstaetten ferrite laths were documented. During isothermal conditions, allotriomorphs predominated at 650℃, while at lower temperatures and or under continuous cooling conditions a mixture of allotriomorphs and laths was observed. At rapid cooling rates and the lowest isothermal conditions, lath predominated. Most lath precipitation appears as distinct events were groups of lath precipitate simultaneously, either at grain boundaries or at allotriomorphs, with similar orientation, which supports the theory of sympathetic nucleation. Lath lengthening rates appear linear under isothermal conditions, and range within a factor of three within one observed experiment. Thickening rates are not linear, and are much more varied than lengthening rates. This variation is indicative of a ledge growth mechanism, but may also be due to varying stereographic components of actual 3D rates.