There is this general view that motors are simple devices. C'mon, it is nothing more than a magnet rotating inside a coil energised by an AC waveform! The magnet may not necessarily be a fixed magnet and is often made from a set of coils and a magnetic substance. Getting the energy to the coils being done either through a set of slip-rings or commutator. Simple!
But hold on! In just one paragraph we have gone from a magnet in a coil to having introduced slip-rings and commutators. We have yet to mention that unless the motor has a certain attack angle, the motor itself will not turn. The angle of force placed on the armature is best at 90° to the magnetic field on the armature to yield the best efficiency. Suddenly motors are not that simple after all.
So it is not surprising motors can, and do, fail. The reason they fail, however, is a simple one. They usually fail catastrophically by burning out. And the reason they burn out is because the energy going in is not being expelled in the form of torque. Any energy that is not converted from electrical to mechanical lands up being converted from electrical to heat. Result equaequals burn out!
I need to point out that I may approach the failure of motors in a bit of a simplistic manner as motors are not my strong point. I therefore would welcome any comment should someone disagree with what is said.
One of the primary causes of failure is simply the motor is asked to provide too much turning action. As with transformers, the magnetic material can only transfer a limited amount of energy. Stalling the motor lowers the inductance resulting in a higher current through the coils. As the copper has DC resistance, a higher current means a higher loss. When the loss causes sufficient heat to melt the insulation then burn-out begins.
If transformers are understood, then one has a very good insight into induction motors. This is displayed when viewing another common symptom which is they cannot handle over-voltage. Now we are not talking 10 or even 20%, but rather in the order of 50% and up. Again, the magnetic material is the main contributor to this problem area.
As with transformers, the magnetic material can only have a limited flux density before becoming saturated. At this point the only current limiting factor is the impedance of the wire making up the coils. Any excess energy is therefore converted to heat, and the same burn-out cycle begins.
With the above, the cause of failure may be simply a human one. Some motors are marked "230/400V". With panels marked in a similar fashion one assumes that 230V is star mode measurement, and 400V the delta. Along comes a 'sparky' and wires the motor according to the rules being motors are wired the opposite to voltage (delta for low volts, star for high). As it is only a 230V supply, this means the motor must be wired delta. Uh-oh!
There are a few aspects with motors that don't exactly stare one in the face. These are usually what cause a poor unsuspecting engineer to scratch his head when confronted with a motor burn out, but the voltage supply and load are more than within the motor's capabilities.
To understand imbalance and motors one must keep in mind the above description that a motor is effectively nothing more than a rotating transformer. As we're dealing with three phases, the same magnetic summing occurs i.e. the three phases, when summed, equal zero.
Yet, having one chunk of magnetised material spinning inside a static one also represents another rotary device; This being a generator! 3-phase motors can, and do become generators when spinning along merrily and suddenly have the excitation voltage removed. This effect is relatively short lived as the magnetism within the motor dies pretty rapidly after a power failure.
But, there is one time this effect continues and that is if only one of the phases loses power. As the now single phase is still magnetising the core (both the field and armature), the motor can continue to be both a motor on the excited section and a generator on the secton without power. This principle is used very successfully in rotary single to three phase converters.
Keeping this principle in mind (that a motor can be a generator too) will go a long way to understand what occurs when a motor is subjected to an imbalanced 3-phase supply (i.e. there are different voltages between the three phases).
When a 3-phase supply is perfectly balanced, then each phase of the motor will draw an equal current (this is kind of logical, so no more time on this).
In an unbalanced situation, the phases that are suppressed will see the motor becoming a generator on these phases in an attempt to raise the voltage. If the imbalance is small, the generated current in the opposite direction is less than the incoming current thus resulting in nothing more than a reduced incoming current. However, the energy needs to come from somewhere and this is the imported from the inflated phases.
What then ensues is the phases with the higher voltage are now doing all the work, and therefore have higher current, while the phases with the suppressed voltage are taking a rest. Yip, you guessed it, the higher current leads to burn out of those phases! And it does not take much voltage imbalance to create a rather large current imbalance. The ratio is approximately 1:10 i.e. 1% difference in voltage causes about 10% difference in current (although much higher ratios have been experienced).
This is a simple explanation as there are other times currents can go up on all three phases, especially if there is an angular displacement between the three phases as opposed to simply supressed voltages on some of the phases. However, once one grasps the above, then all these phenomena can be easily understood.
There is a further problem with very high frequencies; The magnetic material is, as with transformers, made up of thin sheets so as to stop eddy currents. Ah!, but this is at the fundamental frequency. As the frequency goes up, the eddy currents become tighter and therefore can now reside within the individual sheets. Such currents are effectively useless and have no other escape other than - man! you are getting good at this - heat!
And the final part about harmonics and motors; Any good motor installation will come with power factor correction. And, if passive, these are capacitors. Now when one marries a cap and a coil one gets a resonant circuit. If the frequency of the resonant circuit is close to or that of a dominant harmonic, then the energy of this is absorbed by the circuit formed by the coil and cap. And the better the 'Q' (quality) of these two components, the higher the currents. And current equals..... heat!
There are times the harmonics present on the power is not the problem, but what is generated by a motor speed controller (and this time the term 'generate' is correct!). This is no more evident than when the motor suffers premature bearing damage. If the harmonics cancel each other out, then all is usually well. But, as most who have studied this subject will know, most harmonics have superb zero-phase sequence properties! And if it is switching noise from the speed controller, then even more so!
The mechanics behind this are actually quite simple. There is capacitive coupling between the coils and the armature and, as the only path to Earth is through that which is supporting the armature, any current will find its way through the physical bearing components. Of course, this failure is not just limited to motor speed controllers, but any time the motor is subjected to a high harmonic content on the power feed.
What we have not even touched yet is motors fitted with commutators. I will never forget seeing a "diagnostics chart" for such motors with at least 30 different faults that could be seen by "reading" the patterns of wear and burning on the brushes and commutator system.
I suppose what we are trying to aim at here is to say that even something as simple as a motor (no, we have not forgotten the opening paragraph on this page!) is still as vulnerable to power quality as is a piece of sensitive electronic based equipment.