Temperature influenced microstructure evolution and strengthening of 4.5 vol% TiBw/TA15 composites with columnar-reinforced structure fabricated by pre-sintering and hot extrusion

In this paper, 4.5 vol% TiBw/TAl5 composite with columnar network reinforced structure was prepared by presintering and subsequent hot extrusion process. Pre-sintering was conducted at 1100 degrees C for 0.5 h, and billets were extruded with an extrusion ratio of 10: 1. The extrusion speed was 10 mm/s, and water cooling was adopted. An extrusion temperature range of 925-1100 degrees C (which covers the phase transition) was used to study the influence of the extrusion temperature on the microstructure evolution and its correlation with strengthening. The results showed that the extrusion temperature played a decisive role in the evolution of the matrix morphology. When the extrusion temperature increased from the (alpha + beta) dual-phase region to the beta single-phase region, the microstructure of the composite matrix was mainly composed of elongated alpha phase and recrystallized alpha phase. It gradually transformed into a lamellar structure from a high-temperature beta phase. TiB whiskers continued to grow and coarsen. The corresponding texture also changed from (10-10) alpha//ED component in the dual-phase region to (0001)alpha//ED and (10-11) alpha//ED mixed components, which were related to the high-temperature (101)beta//ED component in the beta-phase region. Further performance analysis showed that the matrix strength was positively correlated with the content of the lamellar alpha phase, which increased with the extrusion temperature; however, the strengthening of TiB whiskers was affected by its growth and fragility and displayed a weakening trend. As a result, although the strength of the composite was always better than that of the matrix, its overall strength decreased upon increasing the extrusion temperature, showing the best performance when extruded at 925 degrees C. The stress of strengthening effect could be accurately estimated by the shear lag model.