The master thesis presents the evolution of a mathematical and later on, numerical model that defines the creep deformation mechanism. Due to limitations of computational speed, in reality, a 2nd-degree approximation polynomial is used, which enables a quality presentation of the high and middle temperature and stress zones; however, in the low temperature and stress zone, it does not consider the physical phenomena that are related to creep. For that reason, the \emph{Least Square Method} was used to define a general numerical model that was later on tested on two materials (1.25Cr0.5Mo and 5Cr0.5Mo). By using it, we formed 2nd, 3rd, 4th and 5th-grade master curves and compared them to each other. We determined that higher grade polynomials enable physically better definition of the extrapolated low temperature and stress zone; however, their usage also creates several problems related to data scatter. This is especially proven with higher grade polynomials (> 2nd) due to more degrees of freedom, which can result in a physically incorrect behavior of the master curves. This clearly indicates that in order to establish a stable and robust numerical model, additional conditions need to be applied.
|
