Learning Forum

[ngominVT]

Hi all, I was wondering if anyone could help me with this issue. I am simulating this basic transformer (shown below) with a depth of 2.5 cm and eddy

current effects on. For my simulation set up I have a balloon boundary around the model. For the excitation I set the red (primary winding) and blue (secondary winding) as coils with the appropriate polarity. I then grouped them as windings and add the current excitation. I add all my winding to the matrix so the solver will output the inductance matrix. I'm trying to find the leakage inductance in the model and I have 2 methods to do so. The first is through using equation derived from the 2x2 inductance matrix. I show the derivation below. The total leakage inductance reflected to the primary is L11 - (M^2)/L22. The other method I am using is energy. I short the secondary winding and make sure that there is no energy in the core so the only energy left in the simulation is energy being stored in the leakage inductance. Then I take the energy output from the convergence tab of maxwell or by taking the volume integral of the energy density in the region around my winding and set it equal to .5*(Lktotal)*Ip^2 (this equation also refers to the primary side). The problem is that both methods give me very different results, with the energy method having a leakage value 10 to 20 times higher than the equation method. When I compare equation method and energy method in magnetostatic solver the results are very close. Am I setting up something wrong in my eddy current simulation or doing something else wrong? Any insight would be great. 

 

[Mark Christini]

It is easier to use coupling coefficient or matrix to calculate leakage inductance.  See attached image showing both methods.

However as 3rd option to use stored energy, assign 1A in each electrical turn of the winding to which the leakage inductance will be referred (pri or sec).  Put equal and opposite ampere-turns in the other winding (sec or pri).  Leakage inductance = 2 * energy.

[ngominVT]

Thanks for the reply mchristi. Using the provided equation (I attached pictures of my resulting inductance matrix) 

L Leakage 12 = L11(1-k12^2) = .63709mH (1-.999902^2) = 124 nH

I set primary current to 1 A and secondary current to 2 A (6 turns on primary 3 on secondary, ampere turns cancels out) 

The given energy left in the simulation is 1.2475*10^-6, multiplying by 2 I get 2.48 uH leakage

Do you see what error I'm making that results in this difference? I attached a picture showing that there is no energy in my core. Taking the volume integral of the region with only my winding gives me 2 uH leakage energy. 

 

 

[harsh024a]

Hey mchristi, Where to get this pdf page which you have attached on to this discussion.n

[harsh024a]

Hey ngominVT,nHave you got the understanding of that given problem?nRegards,nharshan