Hysteresis Losses

[Syko_]

Could someone explain hysteresis losses to me in detail please?

My lecturer didn't do a very good job at explaining them and I have exams next week...

Cheers

[Andy Watson]

You might get more responses if you narrowed it down a bit. What sort of system are we talking about? Mechanical, electrical, electronic, servo, dielectric, magnetic?

[Syko_]

Apologies. We covered it in terms of AC transformer core losses and AC motor core losses. Is that any better?

[Syko_]

Magnetic* forgot that keyword.

[Andy Watson]

You might get more responses

Looks like I was wrong
Electromagneitcs is not my forte, but I'll have a go. Can you describe the area that you you are having difficulty with? B/H curves, maths ?

[minime72706]

Hysteresis in motors and transformers, as wikipedia explains it, basically originates from friction inside the core material. A changing magnetic field imposed on a magnetic material like ferrite causes magnetic dipoles to physically rotate. "Magnetic dipole" seems to be a rather general term and the scale of the object described seems to depend on the size of the system. i.e. a bar magnet that you would hold in your hand is a pretty large object, but if observed far enough away, appears to act like a single magnetic dipole. "Magnetic dipole" can also refer to a single molecule or a couple of atoms. The movement of these dipoles causes physical changes in the material. One primary change is that a change in the crystal lattice spacing will often occur. I would compare applying a varying magnetic field to a crystal to kneading a softer material like clay. Sure, your hands themselves are a source of heat, but heat is also added by the molecules rubbing against each other as your hands change the shape of the clay (friction).

Eddy currents are a different and more straightforward type of magnetic loss, but you specifically asked about hysteresis loss.

I'm no expert but I hope I helped more than hurt.

[JackOfVA]

As a stratospheric level, hysteresis power loss is proportional to the area within the B-H curve.   A B-H curve is not a straight line -- although some magnetic materials can come close in normal use.

The next obvious question is why isn't a B-H curve a straight line, followed by why does it take energy to traverse the B-H curve. 

As the magnetic field H changes, not all magnetic domains immediately respond or respond uniformly. Hence to return to the same magnetic state you started with some extra energy must be applied to orient 100% of the magnetic domains to the starting orientation. Since reorienting a magnetic domain involves work, energy is extracted from the H field and shows up as heat. There is energy stored in the magnetic domain orientation and only part of this energy can be released.

One could analogize the magnetic domains as being "sticky" and the energy loss being due to "friction" but these are highly simplistic analogies to the real process.

In addition to power loss due to domain reorientation other power losses occur in the copper windings (I^2R loss) and I^2R loss in the eddy currents induced in the core material. There are other losses such as that due to interwinding capacitance dielectric loss but those should be small to negligible under most reasonable operations.

[Wytnucls]

A ferromagnetic material is composed of microscopic domains, with uniform internal magnetization.
Below the Curie temperature, alignment of domains is random, to minimize internal energy.
When subjected to a magnetic field, domains rotate to align with the field, creating friction and heat between domain walls. When the magnetic field is removed, the domains return to their original position, except for a few, locked in defects of the crystal lattice, resulting in a remaining magnetic field (saturation remanence). Magnetostriction causes the material to heat up and also change dimension, creating a low buzzing noise in transformer cores.
When subjected to an alternating magnetic field, extra energy is required to dislodge these aligned domains every time the field is changing.
Soft ferromagnetic materials have less defects in their lattices and are thus less subject to magnetization (low coercivity).
The relationship between the magnetic field and core magnetization is thus not linear. The hysteresis loop shows the relationship for one cycle of the alternating magnetic field. The area comprised between the curves indicates the energy lost for each cycle.

[Syko_]

Thank you all for your responses, they were all extremely helpful.

My lecturer just kind of glazed over it in class, with a brief explanation.

So, finally, the B/H equations are just derived from first principle equations, yeah?