Abstract—A data processing algorithm is developed to perform a statistical analysis of the results of end-to-end production control for large (≈5 t) cylindrical forgings of variable cross sections made of heat treatable 38KhN3MFA-Sh steel with a retained as-cast structure. A two-parameter criterion \({{\Delta }_{{ij}}} - {{\bar {X}}_{{ij}}}\) is proposed as an additional characteristic of cold resistance. Here, Δij = \({\text{KCU}}_{{{\text{max }}i}}^{{ + 20}}\) – \({\text{KCU}}_{{{\text{min }}j}}^{{ - 50}}\) characterizes the steepness of impact toughness reduction for the standard testing temperatures of forgings (20, –50°C), and \({{\bar {X}}_{{ij}}}\) = ( \({\text{KCU}}_{{{\text{max }}i}}^{{ + 20}}\) + \({\text{KCU}}_{{{\text{min }}j}}^{{ - 50}}\) )/2 represents its absolute level. A correlation is established between the proposed criterion and the process parameters, namely, the chemical composition of steel after electroslag remelting, the heating temperature of ingot for forging, the forging start temperature, and the second tempering temperature. This approach can be useful for cold resistance ranking of forgings.
