Results and discussion

As the simulations are carried out using numerical methods a convergence study is carried out to ascertain the accuracy of the solution using four different meshes. The number of grid elements in each of these meshes are 679, 1005, 1494 and 2212 corresponding to uniform grid sizes 18, 15, 12 and 10 mm respectively. The simulations are carried out without application of EMS. The temperature at 800 mm below the meniscus and 20 mm from the left wall is taken as the parameter to check grid convergence. A plot (Figure 3) is drawn with these temperatures against grid sizes for all these meshes and it is almost flat in x-direction. Hence, it can be concluded that grid convergence has been achieved with grid size 10 mm and this has been used in all subsequent simulations.
[CHART]
Figure 3. Grid convergence
Further simulations are carried out by placing EMS as per the specified positions to obtain the histories of velocity and liquid fraction of the liquid metal solidifying in the mold.
The contours of magnitude of velocity along the length of the mold without EMS and with different positions H- EMS and V-EMS are shown in Figure 6 and Figure 7 respectively. The vectors indicate that the fluid flow velocity is maximum at the location of EMS, in its respective directions of application. The fluid is dispersed to form recirculation loops above and below at the position of EMS at centre of the mold due to the stirring effect. As the position of the EMS is lowered the thickness of loops increases but is more confined to the center of the mold. It is observed that the recirculation loops formed are more defined for all the positions of the V-EMS. The decrease of thickness of the loops and its confinement to the center because of lowering of EMS position is not as significant as H-EMS. The maximum values of velocity magnitude for H-EMS and V-EMS are given in the Table 2.