Optimizing process windows for minimizing the pore size of Ni-based single crystal superalloys

Abstract

Minimizing pore size during directional solidification is critical for reducing the risk of failure for Ni-based superalloys. In this study, the pore size distribution as a function of casting conditions during the entire single crystal casting was obtained by two-dimensional optical metallography and three-dimensional X-ray microtomography. Then, the relationship between the pore size and local thermal conditions was implemented as a subroutine in a finite element model. Both simulation and experimental results show that the temperature gradient and withdrawal rate play the most significant roles in the pore size distribution. Besides, the influences of withdrawal rate and temperature gradient on the mushy zone length, lateral-axial thermal gradient ratios were also investigated. Combining the characterization results with simulation models, a computationally effective method was developed to optimize the processing window of directional solidification and further reduce the pore size.

Introduction

Over 49% of failures on aero-engines are fatigue failure, and the defects responsible for initiating fatigue fracture are pores and intermetallic inclusions [1], [2], [3], [4], [5], [6]. Lamm and Singer [7], Brundidge [8] and Poirier et al. [9] found that the pore size has a significant influence on fatigue life. They reported that the fatigue lifetime of turbine blades could be greatly reduced when the pore size increase. Therefore, it is of great importance to control the pore size during the manufacturing process.

Nowadays, most of the second-generation Ni-based single crystal superalloys are manufactured by the directional solidification process. This process includes directionally solidifying liquid metal from bottom to top by applying a vertical thermal gradient and downwards withdrawal rate. Miller and Pollock [10] and Whitesell and Overfelt [11] showed that the temperature gradient and withdrawal rate in directional solidification had significant impacts on the pore size. Pequst et al. [12] and Çadirli's et al. [13] revealed that the mushy zone length could also affect the pore size distribution. Moreover, Miller and co-workers [14,15] found that changes in the lateral temperature gradient affected the microstructure which was directly related to the pore. Therefore, in order to predict the pore size in local thermal conditions during directional solidification, the temperature gradient, withdrawal rate, the mushy zone length, and the lateral temperature gradient should be considered.

In order to capture all the relationship between the porosity and those thermal conditions, different models had been developed by previous studies [5,6,9,10,[12], [13], [14], [15], [16], [17], [18], [19], [20]. However, how to control the pore size in the industrial manufacture of the turbine blade has not been deeply investigated.

In this study, a computational method is developed to find the appropriate processing window for controlling the pore size. First, pore size formed during directional solidification is characterized by optical metallography, and the results are compared with the three-dimensional (3D) X-ray computed tomography (XCT). Compared with traditional optical microscopy, X-ray tomography is better for non-destructively characterizing local features, extracting the geometric information of the whole pores, and capturing more details with higher resolution [21]. Then, a finite element model of the Bridgman method is built to investigate the influences of thermal conditions on the pores in directional solidification. Sensitivity analysis is further performed to identify the relative importance of these thermal conditions. The regression analysis is performed to derive the pore size prediction model. Furthermore, an appropriate process window is derived from the analysis to minimize the pore size.