*3.4. E*ff*ect on MacroSegregation*

The segregation rate ∆C was calculated by the following equation,

$$
\Delta \mathcal{C} = \frac{\mathcal{C}\_i - \mathcal{C}\_0}{\mathcal{C}\_0} \tag{1}
$$

where *C<sup>i</sup>* is the mean composition at a specific location, *C*<sup>0</sup> is the average (or nominal) alloy composition. ∆C > 0 represents positive segregation while ∆C < 0 represents negative segregation. The segregation index S is introduced to measure the overall macrosegregation degree of ingots in radial direction,

$$\mathbf{S} = \Delta \mathbf{C}\_{\text{max}} - \Delta \mathbf{C}\_{\text{min}} \tag{2}$$

where ∆*C*max and ∆*C*min are the maximum, and minimum, segregation rates in the selected direction, respectively. The larger the S value is, the greater the macrosegregation degree of elements is, and the more uneven the solute distribution is.

Figure 11 shows the variation of segregation rate of Cu, Mg, and Si elements in the cross-section of the ingots under different conditions, which can clearly and intuitively reflect the macrosegregation of the ingots. The ingot without ultrasonic treatment has a large positive segregation in the center, and a very serious negative segregation near the edge, and the composition distribution curve has a sharp drop in the melt at the position. The segregation index of Cu, Mg, and Si decreases from 0.078, 0.053 and 0.072 to 0.048, 0.038 and 0.055, respectively after ultrasonic treatment of the L-shaped ultrasonic wave guide rod within 100 mm from the center. The segregation index of Cu, Mg, and Si decreases from 0.086, 0.069, and 0.036 to 0.044, 0.025, and 0.014, respectively after ultrasonic treatment of the straight-rod ultrasonic wave guide rod within 100 mm at the 1/2 radius. The segregation index of Cu, Mg, and Si changes from 0.212, 0.016 and 0.276 to 0.143, 0.120 and 0.132, respectively after ultrasonic treatment of the L-shaped ultrasonic wave guide rod near the edge. The application of ultrasonic treatment can effectively reduce the degree of solute segregation, reduce the negative segregation at the edge of ingots, reduce the positive segregation in the center, and make the distribution of solute elements in different parts of ingots tend to be uniform. However, the effect of the ultrasound is still limited, and the distribution of Cu and Si elements in the surface and center areas is still uneven, and there is a large gap between the two areas. The ability of the L-shaped ultrasonic wave guide rod in restraining segregation is stronger in the side of ingots, and the ability of the straight-rod ultrasonic wave guide rod to restrain segregation is stronger in the center and at the one-half radius of ingots. This is mainly because the L-shaped ultrasonic wave guide rod had not only the end radiation, but also more side radiation compared with the action of the straight-rod ultrasonic wave guide rod. Especially, micro jet with high velocity was formed on the side of the ultrasonic rod, which forced the local melt in the mold to produce convection in the radial direction and played a certain role in stirring the aluminum melt. Stirring reduced the solidification speeds at the edge of the ingots to a certain extent, so that the components in the solidification process had more time to diffuse, thus, greatly unifying the solute field, reducing the concentration gradient at the solidification front, as well as the concentration of solute elements. Under the action of strong shock wave and micro jet produced by cavitation effect, the high solute liquid at the solidification front was rapidly diffused, thus reducing the macro negative segregation at the edge of the ingots [30]. However, due to the rapid attenuation of ultrasound in the liquid, the effect of ultrasonic at the edge of the mold was very weak, compared with that of the center. Although, it can improve the inverse segregation of solute elements to a certain extent, it cannot be eliminated fundamentally, and the effect is limited. During the solidification process, the positive segregation on the surface of ingots occurred, while the negative segregation near the surface was caused by the deformation of dendrite network [31]. In the surface area of the ingots, cooling water was directly sprayed on it, the cooling rate was high, and the temperature gradient in solidified shell was very large. The thermal shrinkage and volume deformation of the dendrite network at the edge of ingots were larger than the internal area of ingots, and the resulting liquid-phase flow towards the ingot surface was also very obvious, while, the intergranular melt flowed continuously in the direction of shrinkage. When the ingot moved to the lower part of the mold, the solidification latent heat also made the rigid mushy zone on the ingot surface partially re-melt. The melt rich in solute elements between dendrites moved to the shell of the ingots along the gaps driven by extrusion pressure, which resulted in poor solute area near the surface of ingots and negative segregation. However, the solute elements were highly enriched on the surface of the ingots, and the composition was significantly higher than the average value, forming obvious segregation tumor, which reduced the quality of ingot surface.

**Figure 11.** Variation of segregation rate of Cu (**a**), Mg (**b**), and Si (**c**) elements in the cross section of ingots under different processing conditions.
