Effects of Lamellar Boundary Structural Change
on Lamellar Size Hardening in TiAl Alloy
 
Kouichi Maruyama *, M. Yamaguchi *, G. Suzuki *, Hanliang Zhu *, Hee Y. Kim ** and M.H. Yoo ***

* Graduate School of Environmental Studies, Tohoku University, Sendai 980-8579, Japan
** Institute of Materials Science, University of Tsukuba, 1-1-1 Ten-nodai, Tsukuba 305-8573, Japan
*** Department of Materials Science and Engineering, Korean Advanced Institute of Science and Technology, Daejeon 305-701, Korea
 

TiAl alloys have high specific strength and are one of the most attractive light materials for high temperature structural applications. Further improvement of their strength is necessary for their wide spread applications. TiAl alloys for high temperature applications usually take the lamellar structures composed of TiAl and 2Ti3Al phases. It is known that refinement of the lamellar size is a promising way of improving their yield strength, and the present paper focuses on the lamellar size strengthening. The strengthening by refinement of lamellar thickness was studied at room temperature on dual phase Ti-39.4mol%Al alloy over a wide range of average lamellar thickness from 850 to 20nm. The relation between yield stress y and was examined, paying special attention to the change in lamellar boundary structure.

Dislocation densities measured on lamellar boundaries are plotted in Fig. 1 against thickness of / lamellae. The /2 lamellar boundaries are found to be perfectly coherent (absence of misfit dislocations) in thin lamellar structure formed at low aging temperatures. In thick lamellar structure formed at high aging temperatures, the lamellar boundaries have misfit dislocations introduced to relieve the lattice misfit. Both thin lamellae with coherent boundaries and thick lamellae with dislocated ones are present in a lamellar structure formed at an intermediate aging temperature. The critical thickness of lamella for the introduction of misfit dislocations is about 50nm.

The following Hall-Petch type relation holds between yield stress y and lamellar thickness as shown in Fig. 2:
where o is a material constant and k is the slope of the Hall-Petch relation. The yield stress reaches the upper limit in the fine lamellar structure. However, the transition from the Hall-Petch relation to the saturation is not simple. The dislocated boundaries render a high resistance to dislocation motion across the boundaries. A Hall-Petch relation holds in the range of > 170nm, and the Hall-Petch slope takes a large value corresponding to the high boundary resistance (curve 1 in Fig. 2). The coherent boundaries provide a relatively low resistance. Another y - correlation typical of the coherent boundary appears in the range of < 100nm (curve 2 in Fig. 2). The yield stress saturates to an upper limit of 1GPa at = 70nm. The transition from the property of dislocated boundary to that of coherent boundary proceeds with an increase in the density of the coherent boundaries within the range of = 170 to 100nm. The dislocation pile-up model is employed to analyze the transition regime of yield stress. The change in boundary resistance to dislocation motion with the change of boundary microstructure was taken into account. As shown in Fig. 3, the whole curve of yield stress vs. relation can be explained by the pile-up model taking account of the change in lamellar boundary structure from dislocated boundary to coherent one.

 

 

[Published in Acta Materialia, Vol. 52, No. 17 (2004), pp.5185-5194]

 

Fig. 1  Density of misfit dislocations on /2 lamellar boundaries as a function of thickness of lamella.

Fig. 2  Hall-Petch plot of yield stress of a lamellar Ti-39.4mol%Al alloy as a function of lamellar thickness. Curves 1 and 2 were obtained by dislocation pile-up model assuming the boundary resistance * to be 440 and 260MPa, respectively.


Fig. 3  Hall-Petch plot of yield stress together with the result of simulation based on dislocation pile-up model. The curves were obtained assuming the boundary resistance * to be 440MPa for lamellae thicker than 50nm and 260MPa for the ones thinner than 50nm. The dotted curve postulated a uniform thickness of lamellae, whereas the solid curve took account of the distribution of lamellar thickness.