Defining predictive maturity for validated numerical simulations

摘要: The increasing reliance on computer simulations in decision-making motivates the need to formulate a commonly accepted definition for ''predictive maturity.'' concept of predictive maturity involves quantitative metrics that could prove useful while allocating resources physical testing and code development. Such should be able track progress (or lack thereof) as additional knowledge becomes available is integrated into example, through addition new experimental datasets during model calibration, and/or implementation better physics models codes. This publication contributes discussion attributes metric exhibit. It contended assessment must go beyond goodness-of-fit test data. We firmly believe also consider ''knobs,'' or ancillary variables, used calibrate degree which experiments cover domain applicability. emphasis herein placed translating proposed mathematical properties, such regularity asymptotic limits function. Altogether these properties define set constraints function satisfy. Based constraints, we propose Predictive Maturity Index (PMI). Physical are illustrate how PMI quantifies non-linear, Preston-Tonks-Wallace plastic deformation applied beryllium, light-weight, high-strength metal. question ''does collecting data improve power?'' answered by computing iteratively become available. results obtained reflect coverage validation important goodness-of-fit. example treated indicates stabilization can observed, provided enough