DC link and rectifier model, Active current as function of voltage?

[Liferider]

Hi, I'm trying to form an expression for a rectifier supplying a DC link with current, without actually rectifying a 3-phase input. Usage of the model is really only concerned with the RMS values and the active power drawn from the AC supply.

The DC link is just a capacitor so that:
dV_DC/dt = (i_DC - i_L)/C

Based on what I've experienced with actual DC links, I chose the drawn rectifier current, i_DC, to be a design constant multiplied with a function of DC voltage squared:
i_DC = k*(V_peak - V_DC)²

Now, I wonder if this model can be used, i.e is there a design constant that will fit an actual rectifier?

Obviously, there is a certain relationship between Active Power/Active Current and the voltage drop on the DC link, I am just not sure if the square model here fits reality? Any comments to this is highly appreciated.

Regards, Arne Ranvik

 

[Liferider]

The model should approximate a basic 3-phase passive rectifier, half-wave I think, so that a fully loaded (rated power draw) rectifier so that the measured DC-voltage will show something like V_DC = 1.35*V_RMS, compared to the unloaded case where V_DC = √2*V_RMS