Kunalw,

I would modify your idea  "harmonic waveform  superimposes itself on the main waveform" as follows:

As you rightly pointed out, the non-sinusoidal armature supply itself is a cause of harmonics (probably, the supply is obtained from inverters to enable variable freq operation). When such a case arises, there's a principal component (which you call as the 'main waveform') and various harmonics- some of them cancel each other; others circulate within the windings if it's delta-connected. In any case, we do have torque produced by these harmonic currents. These won't be aiding the main torque, in most cases.

I'm not clear on what you meant by "Hence these harmonic fields act as small motors."

Regards

umn@ieee.org

Hi,
You are right. This harmonic currents act like small motors connected mechanically to the same shaft. Every small motor has it's own rotation speed, rotation direction and torque. As the torque of the every small motor is the function of the relevant harmonic current and the amplitudes of this currents are not the same, the shaft rotates according to the strongest current( generally fundamental component ,n=1).
But other small disturbant harmonic torques effect the output torque and rotation quality.(in the means of oscillations and acoustic noise)