Armature Reaction & Brush Shifting

[QwertyXP]

Referring to a DC generator: one way to counter armature reaction is to shift the brushes. But I don't get how it works.

The direction of armature flux also changes when the brushes are shifted, so why does the MNA (magnetic neutral axis) not also rotate further.. If it does rotate further, then there would be no point in shifting the brushes!?

[Kremmen]

OK, i will start from the beginning so interested readers can follow the question.

Let's take an easy case - a DC machine with a single pole pair in the stator. The magnetizing current in the stator coils creates a uniform flux penetrating the armature and it is against this field the armature works to generate power (in either direction - motoring or generating). Let's take the motor case because it may be conceptually easier but both are essentially the same for a DC machine.
For this simple case there is the neutral plane bisecting the armature normal to the direction of the flux, and the commutator brushes are placed on this plane. When in operation, the armature current is supplied via the commutator to the armature coils. These in turn create a magnetic field that interacts with the stator field to produce torque.
Armature reaction is the phenomenon where the magnetic fields interfere causing the stator field to distort. The field flux is oriented in the nominal direction d and the armature field in the nominal direction q, or "direct" and "quadrature", i.e. they are at mutual 90 degree angle (if the brushes are in the neutral plane). Now the resultant field in the armature region is the vector sum of these two fields and it causes the effective flux to "rotate" so that the field density decreases on one side of the stator pole face and increases on the opposing side. The same happens on the other pole in reverse.
This in turn causes two effects.
Firstly the stator magnetic circuit usually operates reasonably close to saturation of the iron in the mag path. The B-H curve is far from linear at the upper end so the net effect of this flux shift is a weakening of the stator magnetizing field. Thgis is turn causes the motor to "react" to increased load by accelerating since speed and stator field density are inversely proportional.
Secondly, the brushes are nominally in the neutral plane of the armature, where the coil voltage is 0 during commutation. Armature reaction distorts the mag fields and the brushes are not effectively in the neutral plane any more. This means that the commutation voltage is not 0 and results in commutator brush sparking.

One way to try and compensate this is to turn the commutator out of the original neutral plane into the new effective neutral plane. This will introduce a 'd' component into the armature field and for spcific values of armature current, may restore commutation in the effective neutral plane.

Please keep in mind that the effective stator flux orientation is the vector sum of the stator field and the armature field, _not_ armature field alone. The direction of the armature field vector does follow the plane of the commutator brushes but that alone does not define the stator field orientation or neutral plane.

A well designed (large) DC motor will have special compensation coils and reversing poles to control the armature reaction and effectively remove it. The effects created by those coils and poles cannot be replicated by turning the brushes.

Hopefully this clarified more than it confused.