Learning Forum

[Zhansen Akhmetov]

Losses calculated in each winding using the Resistance Matrix do not match the same losses calculated by ANSYS Maxwell 2D using the Per Winding Solid Loss function at high frequencies.

I did the simulation with two circular conductors with radii of 0.0508 mm at low and high frequencies. The frequency is 10 kHz for the first simulation, while for the second simulation, the frequency is set to 10 MHz. Both conductors are excited by an RMS current of 1 A. The results of both simulations are shown in the attached photo.

For 10 kHz, the Per Winding Solid Loss function matches the formulas from the R matrix: P1 = I^2*(R11+R12) and P2=I^2*(R22+R21). For example, for winding 1, self-resistance  R11 334.7659 mOhm matches 334.7659 mW losses found by the Per Winding Solid Loss function.

However, for 10 MHz, the Per Winding Solid Loss function does not match the formulas from the R matrix: P1 = I^2*(R11+R12) and P2=I^2*(R22+R21). It becomes vice versa. For example, for winding 1 from the R matrix, the loss is 630.44 mOhm-31.608 mOhm = 598.832 mOhm. However, the Per Winding Solid Loss function shows 583.93 mW.

Why is there this mismatch at a higher frequency?  Is there any way to correctly calculate losses in each winding from the R matrix? What should we trust more for losses in each winding, Per Winding Solid Loss or R matrix?

[Reshmi Raghavan]

Hello Zhansen,

Please check your mesh at 10MHz. The skin depth at 10MHz is 0.0208mm considering copper as material. If the mesh is not refined to capture the current destribution then the loss can vary.

Another thing to note here is that Maxwell is used for the low frequency simulation. Based on the below snapshot information we can confirm if Maxwell can be used for the 10MHz condition or not.