For testing filter chokes, I tacked together a transformer, variac, diode bridge, various load resistors (I'm lucky to have the Clarostat power resistor decade box) and a known value filter capacitor forming a L-C filter with the dummy load resistor. Various d.c. voltages and load resistor values for a wide range of currents. I measured the peak-to-peak ripple accurately across the load resistor using a scope. Then I ran a simulation of the circuit on LTSpice and adjusted the inductor value until I got the same peak-to-peak voltage, for each current point. Don't forget to include the secondary resistance and the dc resistance of the choke in your schematic model.
All chokes will exhibit an S-curve of current versus inductance. I believe swinging chokes are designed to handle the load current with full current rating being on the dropping slope of inductance (heading towards saturation of the core).
Look at my resultant curves in the pdf'd Excel graph attached to one of my posts in the lower portion of the referenced thread (choke_study1.pdf). This is not as simple as the QST article but my anal way of doing things; very accurate I believe.
Here is the old thread "Swinging Chokes" started by KB3DKS, 2009:
https://amfone.net/Amforum/index.php?topic=18829.0
