Control Parameters Correlation by Multiple Linear Robust Regression for the Design of Heat Treatments For Al-6063 Alloy

In this work, an alternative statistical multiple linear robust regression methodology was used to analyze the experimental data obtained from tensile tests performed on heat treated samples of an aluminium 6063 alloy. The proposed methodology allows the identification of empirical relations between the ageing temperature and process time with the mechanical strength (UTS) and yield point (YS) of the alloy. Through these empirical relationships: a) UTS = 0.0932 T a t A − 0.1356 t A 2 − 0.0002434 T a 2 t A + 0.00063 t A 3 and b) YS = 5.10 T a − 20.32 t A − 0.05281 T a 2 + 0.2297 T a t A + 0.0001459 T a 3 − 0.0006192 T a 2 t A , obtained by using the multiple linear robust regression methodology, it was possible to find that an ageing temperature of 192 °C for 120 min would produce the maximum strength levels for the Al-6063 alloy. For such alloy, empirical relationships were successfully determined for ageing heat treatment design.