Experimental investigation of normal grain growth in terms of area and topological class
Experimental investigation of normal grain growth in terms of area and topological class
复制标题
DOI:
10.1016/0036-9748(85)90053-5
复制
发表时间:
1985-11
期刊:
影响因子:
--
通讯作者:
V. Fradkov;A. Kravchenko;L. Shvindlerman
中科院分区:
文献类型:
--
作者:
V. Fradkov;A. Kravchenko;L. Shvindlerman
The process of grain growth and the resulting grain structure are usually described in terms of grain size, mean grain size and time dependence of mean grain size. In some papers (1-3) size distribution and topological class (ie the number of neighbout grains) distribution are considered. Despite considerable efforts made both in the field of theoretical description (3, 4, 5, 6) and computer simulation (7, 8, 9, 10), no clear understanding of the phenomenon has so far been achieved. In (4) an attempt was made to solve the problem in terms of grain size assuming, als in~ qe theory of coalescence, that~ rains smaller than a certain critical size RE, diminish in size and those greater than K0 grow. There is no physical foundation for this approach because the rate of area growth for a two-dimensional grain is given (12, 13) by dS= where A is product of mobility to surface tension which is the same for all the grain boundaries in the system, T') is the number of neighbour grains, ie the topological class of the grain. From eq.(1) it is clear that when n< 6 the grain area diminishes and whenn]) 6 it increases. The rate of area growth depending only on n it appears natural to consider normal grain growth as evolution in the space area-topological class. It should be noted that eg.(1) is true only for two dimensions and no analogous expression is available for three dimensions. In this paper we examine mosUy the evolution of two-dimensional grain structures in phase space (L.~, n) and specifically the stage of normal grain growth when all the system parameters exhibit a scaling behaviour and the mean grain area grows linearly with time. As two-dimensional experiments in normal grain growth are fewer than the three-dimensional ones and because their authors were mostly preoccupied with other issues, it was necessary to perform further experiments. Moreover, somewhat earlier computer simulations of grain growth had been performed (14) based on the assumption of similar grain boundaries and eq.(1). Our experiments were carried out in a way to approach as close as possible the assumpiions of the computer model (two-dimensional growth, similar energy and mobility, absence of anisotropy).