Open armature winding.

On a normal DC motor, every commutator segment is connected to the next one by an armature winding. The brushes are normally on opposite sides of the motor, so the current flows in two paths, and each path goes through several windings.

For instance, if there are 10 segments, numbered sequentially, and the brushes happen to be contacting segment 1 and 6, current will be flowing in these windings:-
1 -> 2
2 -> 3
3 -> 4
4 -> 5
5 -> 6
and on the other side
1 -> 10
10 -> 9
9 -> 8
8 -> 7
7 -> 6

When the motor turns a bit, on brush will contact segments 1 and 2, and the other will contact segments 7 and 6
so the current will flow like this:-
2 -> 3
3 -> 4
4 -> 5
5 -> 6
and on the other side
1 -> 10
10 -> 9
9 -> 8
8 -> 7

When the motor turns a bit more, one brush will lose contact with segment 1, and the other will lose contact with segment 6.
The current that was flowing in the four windings 1->10, 10 -> 9, 9 -> 8 and 8 -> 7 can no longer flow to the brush via segment 1, but it has to go through the winding that joins segment 2 to segment 1.
The inductance of that winding will stop the current increasing instantly, while the current in the four windings can't decrease instantly, so there is some limited amount of sparking until the current in the four windings has decreased a bit and the current in the one winding has increased to match them. With four windings against one, the current will settle somewhere around 80% of what it was just before the brush broke contact.

The loss of energy is around 20% of the inductive energy in the windings. Some of the energy will be lost to sparking between the brush and segment 1

However, if the winding between segments 2 and 1 is open circuit, the current in the four windings 1->10, 10 -> 9, 9 -> 8 and 8 -> 7 has absolutely nowhere to go, and the sparking will go on until the current in the four windings has dropped to zero. That means that all of the inductive energy in the windings will be lost, much of it to sparking, so the sparks will be much larger.

Also, if one winding is open circuit, one current path of two is lost, and to get the same motor torque approximately the same total current will flow, so there will twice the current in the remaining wires and more inductive energy. It's difficult to say how much more, because the various windings are overlapping, but the extra current will also cause more sparking.

So, the inductive energy or back end will sort of arc through the air to the brush!

Also, why would the extra current in the good path cause more sparking?

gjoo said:

So, the inductive energy or back end will sort of arc through the air to the brush!

Click to expand...

As I understand it, that is exactly what is happening.

An arc is usually electric current flowing through air.

gjoo said:

Also, why would the extra current in the good path cause more sparking?

Click to expand...

The energy in an inductor is proportional to the square of the current, so that would make twice the current result in four times the energy, although it might not be as simple as that.

Even without that, if one path is broken, all the current is in the other path, so the current being diverted when the brush breaks the circuit is twice what it would be with an intact motor running at the same current.

So at one brush, we potentially have about twice the normal voltage/current and at the other we have part of the energy arc at the brush, while part of the current flows through the other brush.

There would be around the same current in each brush, but it's the breaking of the current that causes the arc. When there is a broken winding, one path will have no current, so when the brush disconnects from a segment, there is no arc at that brush.

If one brush is contacting segments 1 and 2, and the other is contacting segments 6 and 7, but one path is broken, current will be flowing in 2->3, 3->4, 4->5 and 5->6, but nowhere else. When the second brush leaves segment 6, the current in 2->3, 3->4, 4->5 and 5->6 will have to start flowing through the winding between segments 6 and 7, so there will be a spark between the second brush and segment 6.

As the second brush leaves segment 6, the first brush will be leaving segment 1, so any current flowing in the winding between segment 10 and segment 1 will have to start flowing in the winding between segment 1 and 2. However, with a broken winding, there will be no current in the winding between segment 10 and 1, so nothing to divert and no spark.

But, there's heavy sparking at the brushes when an armature coil is open. So, that would be the current arc by the open coil?

gjoo said:

But, there's heavy sparking at the brushes when an armature coil is open. So, that would be the current arc by the open coil?

Click to expand...

An open coil can't take any current, so the open coil can't arc.

The heavy sparking is due to the current in the other coils not being able to go through the open coil.

But, why do we see this sparking at the brushes and not just in the armature?

gjoo said:

But, why do we see this sparking at the brushes and not just in the armature?

Click to expand...

Generally, contacts spark as they make or break. The commutator segments never touch the adjacent ones, so there is never any contact that makes or breaks except at either end of the brushes. Sparking is normally at the point where the commutator is moving away from the brush, as explained above.

gjoo said:

If one armature winding is open in a DC system, such as a DC motor then they'll be excessive sparking at the brushes. What I'm not sure about is if that sparking is caused by the back end of the previous armature coil or something else?

Click to expand...

Generally arcing on brushes indicate a shorted turn(s) on the armature.
A definitive test is done using a Growler.

Appreciate all the info. I think I've got it. So, all of the sparking is due from the extra current flowing through the conductors connected to the brush opposite to the brush with the open coil. And this sparking is due when that brush leaves one commutator to the next one

Here is the connection sequence.

If I have it right , there is excessive current at the good coil connected to the commutator causes the excessive sparking. But, is there also sparking due to an arc that tries to jump from one coil over the bad coil to the other coil?

When the last conducting segment (before the break) leaves a brush, you have effectively broken the circuit on an inductor.

The current in the just-disconnected windings will continue by arcing, as with any typical inductor when the circuit is suddenly interrupted and there is no alternate path.

With a fully intact armature there are no sudden disconnects.

I get it now. Thanks, everyone for all the info.