**2. Experimental Procedure**

During this work, two main sets of data were collected—one during the production trials of forty semi-finished extruded products, and the other one through laboratory analysis, performed to assess their quality.

The process parameters were monitored by process-control software, which manages data about material flow speed, temperature and the liquid nitrogen valve opening. In detail, data about dye temperature was collected through four thermocouples (TC) installed within the mold (Figure 3) and two laser pyrometers.

**Figure 3.** Schematic drawing of the thermocouples' positions within the dye.

The thermocouples used in this work were Chromel/Alumel (K) thermocouples with mineral oxide insulation and Inconel 600 as the sheath material. The sheath had a 3.2 mm diameter and 2000 mm length. The laser pyrometer temperature method measured the emitted thermal radiation, which directly corresponded to the temperature and surface emissivity of the target. In detail, the pyrometer sensors detected the amount of infrared radiation emitted by the measured object. The pyrometers were installed in front of the exit of the extrusion dye (Figure 4).

**Figure 4.** Pyrometers' positions.

On the other hand, the surface defects of the extruded semi-products were characterized in the junction zone of the billets, since this is an accumulation area for the defects [1,2]. The samples were drawn from the extruded semi-products and polished following the common standards. The chemical compositions of the samples were acquired through a scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany) Zeiss EVO 50®, using a back-scattered electron (SEM-BSE) (INCA Oxford Instrument, Oxford, UK) detector and energy dispersive X-ray microanalysis (SEM-EDS) (INCA Oxford Instrument, Oxford, UK). In addition, the same instrument was exploited to collect morphological images through a secondary electron detector (SEM-SE) (INCA Oxford Instrument, Oxford, UK).

The samples were etched electrochemically using a Barker's reagent. The electrolytic solution was prepared by mixing 40 parts of water (H2O) for each volume-part of tetrafluoboric acid (HBF4). The electrochemical anodization was performed by applying a current density of 0.2 A/cm2 (20 V dc) for 40–80 s.

The metallographic analysis was executed using a cross polarized incident light. Afterwards, the index used to evaluate the average grain dimensions was calculated via the Heyn intercept method, following the ASTM E112-13 standard.

Roughness measurements were performed on the surface of the body of the extruded semi-product, hence, not in the junction area. Profile roughness measurements were taken perpendicularly to the extrusion direction, using a stylus Mahr PGK MFK-250® (Mahr GmbH, Göttingen, Germany) tester with a tip radius of 2 μm and a vertical measure range of +/−250 μm. The data were collected following the standard, UNI EN ISO 4288-2000. The translational speed of the measurement was 0.5 mm/s along an exploration length of 5.6 mm. The base length was 0.8 mm and the total evaluated length was 4.0 mm. The collected data were filtered using a Gaussian filter with a wavelength cutter of 0.8 mm.

Surface texture measurements were performed using an Alicona InfiniteFocus® (Alicona Imaging GmbH, Graz, Austria) microscope, which exploits focus-variation technology. The data were acquired in a similar position to those collected for the profile roughness measurements. The surface texture measurements parameters are listed as follows: the total evaluated area measured 4.0 × 4.0 mm and the base length was 0.8 mm. The collected data were filtered again using a Gaussian filter with a wavelength cutter of 0.8 mm.
