the slip frequency decreases
Why would slip change with applied frequency and or voltage? Isn’t slip is a constant and proportional to rotor resistance?
Perhaps I misunderstood. I’m not familiar with the term “slip frequency”. I define slip as drop in speed from no load (which is not exactly synchronous but darn close) to full load speed.
Slip is the difference between the equivalent rotational speed of the mains (magnetic field of the stator) and the actual rotation of the rotor. Slip frequency is just the same as expressed in electrical frequency (Hz) instead of RPM. More slip means the motor is running slower, and vice versa.
If one increases voltage and thus winding current, the motor
is capable of generating more torque and "catches" the stator field in speed. But if there is no use for this torque, it just means worse power factor and high reactive current.
Under no load, induction motor runs closer to synchronous speed. Indeed, the first improvement from linear V/f relationship towards optimum efficiency, constant torque control is to implement constant slip control, which adjusts voltage so that less voltage is used with small load (making slip actively higher when it naturally tries to go too low), and more voltage is used under heavy load (making slip actively lower).
Optimum slip is indeed kinda constant (and proportional to rotor resistance, so rotor windings heating up makes it a non-constant again), so I think this is where you got mixed up. But for a fixed-voltage, fixed-frequency (i.e., direct mains) use, optimum slip is only achieved at the optimum mechanical load designed for that motor. It is well known that an induction motor which runs at either very light or very heavy load significantly improves its efficiency when driven with a VFD in FOC mode. (Constant slip control is worth experimenting with if you ever try to build this stuff. I remember mentally struggling with the whole FOC concept but slip control made a lot of sense to me and I was able to make it work experimentally very well, and difference in dissipated power in power stage transistors thanks to getting rid of all that reactive current was just huge, compared to V/f control.)