Discussion Forum

Hello,

Currently I am studying the radio-frequency magnetic field created by a small circular current loop with oscillating current located outside of a closed metallic enclosure with very large dimensions and small thikness.

The metallic enclosure spans over about 20 m x 20 m x 10 m and is compound of 7 vacuum towers linked together by vacuum pipes as can be shown on the attached figure towers.png (or PR-Towers.stp). The thickness of all the system walls is 1.5 cm and is made of stainless steel taken from Comsol library (mu_r = 1.5, sigma = 1.3 MS/m). In order to avoid 3D mesh of themetallic walls, they have been replaced by a single surface on which a transition boundary condition is applied.

The coil has a radius of 35 mm and it can be oriented along the 3 main axes. It is positioned at the basis of the small tower located at x = 11.1 m from the crossing (origin of the reference frame). The current is 830 A in this study (the coil has in fact 830 loops), the applied frequencies range from 0.5 Hz to 1000 Hz.

Looking at the solution for the field modulus along a line parallel to the x-axis, but 1 meter above (1D Plot Group 8, attached file field_1m_above_axis.png), one can see a behaviour which I do not understand: instead of a field decrease within the metallic enclosure, in the tower closest to the coil, one clearly sees a field enhancement at the lower frequencies (except for 0.5 Hz), especially at 10 Hz whereas in the other towers, one can see the expected shielding effect increasing with frequency. I could understand that the field would be the same with and without the metal at low frequencies where the skin depth is larger than the metal thickness, but I do not understand the enhancment. To investigate this, I have then compared the field of the coil with and without the metallic towers at 10 points inside the tower versus several frequencies and again, below about 50 Hz, the internal field is higher when the metallic towers are present compare to the case when they are replaced by air (see file ratio_field_metal_over_nometal.png). This comparison was done using the same 3D mesh for the whole system. Perhaps I am doing something wrong in applying the Transition Boundary Condition on the metallic pieces, but I do not find any mistake. Are their some particular frequency domain in which one should not apply this TBC ?

If anyone has some hints concerning this behaviour, I would really appreciate to know about it. Thanks in advance,

Gilles