Summary

Non-linear property of iron core is considered in the model used in "Static magnetic field analysis using ELMCUR (element current source)".

Explanation

Here the same mesh files pre_geom2D.neu and 2D_to_3D are used. B-H property curve of the iron core is entered in the input file. The analysis is performed for the 4-step excitation state.

Figure 1 shows the magnetic flux density distribution for coil currents of 6 kAT and 60 kAT. In case of coil current 60 kAT, though the exciting current is 10 times higher, the magnetic field becomes uniform and the maximum value is not 10 times higher due to saturation. Non-linear effect appears. In addition, where the magnetic field is concentrated, magnetic saturation occurs and the concentration is reduced.

Min : 0.043 T , Max : 0.32 T
（a） 6000AT

Min : 0.54 T , Max : 2.10 T
（b）60000AT

Fig.1　 Magnetic flux density intensity distribution in the iron core

Figure 2 shows the convergence of the iterative calculation. This graph is displayed on the screen and the numbers are output to the file check. Within the step iteration is the nonlinear convergence iteration, and inside it is the ICCG method iteration.

One ICCG iteration calculation terminates when it converges to less than the initial error multiplied by ICCG_CONV/NON_LINEAR_CONV (1/1000 in this case) and then moves on to the nonlinear iteration.

The nonlinear iterative calculation is terminated after repeating the ICCG when the error becomes NON_LINEAR_CONV or less. At this time, the ICCG error is less than or equal to ICCG_CONV.

In this example, the ICCG iteration converges until the convergence condition is satisfied, but there are cases where the iterations diverge without converging. In this case, iterations are performed up to the number of iterations limited by DEV_FACTOR and DEV_ITERATION, and the solution with the smallest error is taken as the convergence value, and the next nonlinear iteration is performed.

Fig.2 ICCG and nonlinear iteration convergence status