Imagine for a moment a steel core with no air gap, that is driven to saturation by a powerful permanent magnet. Imagine next that a coil is wrapped around some section of that core some distance away from the magnet, such that their only flux linking is through the saturated core.

The magnet is typical, but powerful enough to saturate the core: it has typical magnet ratings, Hc for its coercivity, Br for is remnant flux. The coil is magic, capable of supporting any amount of current needed for the experiment. The coil is also wrapped very tightly around the core.

What amount of MMF (H) would be put through the coil in order to stop all flux in the core?

One hypothesis is that the current in the coil would need to equal the Hc rating of the magnet in order to stop all flux in the core. My reasoning is that if the coil were to have just a small amount of current, the core would come out of saturation for a moment, allowing the magnet to send yet more flux into the core.

Another hypothesis is that the core is simply a very large reluctance, and the magnet is already producing as much flux as possible given its coercivity. If just a small amount of opposing MMF were generated in opposition to the magnet, all flux should stop.

What do you think?

 

How about trying to explain what happens to the BH curve operating point of the steel with the magnet in place, and then what happens to the operating point when a suitable amount of opposing flux is generated by the coil, and what conditions are needed to null the flux in the core.

 

What happens to the BH curve of the steel is that it flattens out as the steel becomes saturated. The steel core will act like a big capacitor, and will store all the magnetic energy as a magnetic field Hcore. If this were drawn as an equivalent circuit the loop equation would be:

Hc=Hcore+Hcoil
Hcore=Rcore*flux
Hc=Rcore*flux+Hcoil

Obviously if Rcore*flux=Hc, then there isn't much current needed in the coil to stop all flux in the system. As long as Rcore is in saturation, then it's pretty easy to stop the flux. If I hypothesize for a moment that the core is at some arbitrary flux density near or below its BH knee, then all of a sudden I can see that Rcore is much much lower. Now Hcore would be very small!

Hcore=Rcore*flux >> Rcore is small, Hcore is small!
Hc=Hcore+Hcoil >> Now it takes a LOT of Hcoil to stop the flux from flowing.

So both Hypothesis are correct... it will take near Hc to stop the flux in the circuit, but it is also very easy to stop the flux from flowing when the reluctance of a magnetic circuit is high. I just had to think about it differently.

Edit: Thank you