In Reply to Parent Post by Peter Selvey
For equipment with mains transformers, motors or other windings connected directly across the mains, the worst case is the lowest frequency. Transformers operate more efficiently at higher frequencies (lower magnetizing currents). So, temperature tests at 50Hz should be enough.
I would agree with this for plain inductive components operating at line frequency, but I don't think this is a sufficient approach when ferroresonance is part of the voltage change design approach, or when the device is a switching-type design with its primary inductive component operating at the switched high frequency and with different line frequencies resulting in different harmonic mixes, or for some three phase designs. In those cases, testing at both 50Hz and 60Hz is appropriate, I think.