S. K. Koduri, B. Welk, G. B. Viswanathan, H. Fraser, The Ohio State University, Columbus, OH; P. C. Collins, Quad City Manufacturing Lab, Rock Island, IL
The development of a set of computational tools that permit microstructurally-based predictions for mechanical properties of commercially important titanium alloys, such as Ti-6Al-4V and Ti-6242, is a valuable step towards the accelerated maturation of materials. This paper will discuss the development of Neural Network Models based on Bayesian statistics to predict the yield strength, ultimate tensile strength and fatigue of Ti-6Al-4V at room temperature, as well as the fatigue in a/b processed Ti-6242. The development of such rules-based models requires the population of extensive databases. For the case of Ti-6Al-4V, the database contains both compositional and microstructural information. For the case of Ti-6242, the database contains both microstructural and experimental information (frequency, temperature, and stress). These databases have been used to train and test Neural Network models to predict the tensile and fatigue properties. In addition, these models have been successfully used to identify the influence of individual microstructural features on the mechanical properties, consequently guiding the efforts towards development of more robust phenomenological models.
Summary: The development of a set of computational tools that permit microstructurally-based predictions for mechanical properties of commercially important titanium alloys, such as Ti-6Al-4V and Ti-6242, is a valuable step towards the accelerated maturation of materials. This paper will discuss the development of Neural Network Models based on Bayesian statistics to predict the yield strength, ultimate tensile strength and fatigue of Ti-6Al-4V at room temperature, as well as the fatigue inƒnƒÑƒ}ƒÒ processed Ti-6242. The development of such rules-based models requires the population of extensive databases. For the case of Ti-6Al-4V, the database contains both compositional and microstructural information. For the case of Ti-6242, the database contains both microstructural and experimental information (frequency, temperature, and stress). These databases have been used to train and test Neural Network models to predict the tensile and fatigue properties. In addition, these models have been successfully used to identify the influence of individual microstructural features on the mechanical properties, consequently guiding the efforts towards development of more robust phenomenological models.
