Chuck Yung
EASA Technical Support Specialist
The procedure for core loss testing of stators is well-defined, but there is not as much information available for special cases like rotors, armatures or high-frequency motors. While the same basic principles apply as for stator testing, frequency is the variable that affects how we should interpret the results.
First, a mini-review is in order. Core losses are comprised of hysteresis and eddy-current losses. Hysteresis losses depend on the grade of steel used and are proportional to the frequency.
Eddy-current losses the edge of each lamination occur at the edge of and vary as the square of the each lamination. frequency. That squared relationship and the fact that they are controlled by the inter-laminar insulation make them critical to the motor’s efficiency. As long as the inter-laminar insulation is viable, these losses are controlled. Shorting of the laminations, whether caused by a rotor drag or from insulation breakdown from excessive burnout temperatures, increases the eddy-current losses. Higher losses result in increased heat and higher magnetizing current.
Tech Note 17 Establishes Guidelines
For stators operating on 60 Hz (or 50 Hz) sinusoidal power, Tech Note 17 establishes some guidelines. Losses for a good stator vary between 1-6 watts per lb. (2-13 watts per kg.) depending
on steel grade, etc.; higher losses indicate a defective core and require corrective measures. The rules for rotors are not so clearly understood. First, the induction rotor is only exposed to 60 (50) Hz power during the moment of starting. As the rotor accelerates, rotor frequency quickly drops to slip frequency. A typical rotor is only exposed to 1-3 Hz during normal service. (That figure is slightly higher for NEMA design D rotors, especially those in the highest slip range of 15-25%. “Typical” design D motors are 5-8 or 813% slip.)
Multiply the percent slip times the line frequency to determine the rotor frequency; even the high slip design D rotor is only exposed to 9 – 15 Hz. Because the rotor frequency is so low, eddy-current losses are not a major concern for most induction rotors. That means that the core test is not applicable to most squirrel cage rotors. An exception is the 2-pole rotor, only because localized hot spots may bend the shaft (thermal bowing), resulting in increased vibration levels. The only real concern is that hot spots could indicate open rotor bars. An open rotor bar can force the current, normally carried by the bar, to pass through the laminations in the vicinity of the break, generating heat. The core loss tester is a great tool for evaluating the rotor cage, with magnetic imaging paper, to check the integrity of the cage. Infrared thermography is a good way to find and document hot spots.
Armature frequency
Armatures are easily overlooked when eddy-currents are discussed, since eddy-currents are an AC phenomenon. But a DC armature actually sees alternating current, as each coil passes from pole to pole while rotating. Since the relationship between RPM and number of poles is not as clearly defined as with AC machines, RPM is another variable that affects the importance of eddy-current losses in DC armatures.
To calculate the frequency of a particular armature, use the RPM and number of poles.

The armature of a 4-pole DC machine, rotating 1800 (1500) RPM, is subject to 60 (50) Hz. At 900 (750) RPM the frequency drops to 30 (25) Hz. At higher speeds the frequency increases, so the eddy-current losses become more detrimental.
Spindle motors, frequency and lamination thickness
When high frequencies are discussed, the spindle motor – often operating at 240 Hz or more – is an other special case. Thinner laminations are used to control the eddy-current losses. Shorted laminations are even more detrimental than with the standard 60 (50) Hz motor. Thinner laminations also have a lower stacking factor. The thicker the lamination, the higher the proportion of steel to lamination insulation will be. (Eddycurrent losses vary with the square of the lamination thickness.) Vintage 25 Hz machines have thicker steel laminations than 60 (50) Hz machines.Military surplus generators and aircraft motors operate at 400 Hz, so they also require thinner laminations.
Importance of frequency
To illustrate the importance of frequency in core testing, consider the relative watts loss per pound for the same core at various applied frequencies. Since eddy-current losses are proportional to the square of the frequency, we can take the square root of the 6 watts loss per pound limit (2.45) and change that in proportion to various other frequencies encountered by our industry. To determine the equivalent losses for 120 Hz, double the 2.45 value, then square it. If a 60 Hz core loss test to a 120 Hz core results in a “good” value of 6 watts per lb. (13 watts per kg.), the expected eddy-current losses operating at 120 Hz would be 24 watts loss per lb (52 watts per kg.) Reasonable limits for frequencies other than 60 Hz should be determined by collecting actual data.
[(f/60) v6]2 = expected watts loss/lb. at applied frequency
where f = frequency, applied [(f/60) v13]2 = expected watts loss/kg.
Summing Up
Eddy-current losses vary as the square of the applied frequency, so the importance of core loss testing increases with the frequency of the equipment under consideration. Spindle motors and armatures that operate at higher frequencies are far more critical than vintage 25 Hz motors or induction rotors.
Use common sense when evaluating a core. T-frame motors generally use better grades of steel than older U-frame and pre-NEMA motors, so the hysteresis losses will be lower. And EPACT motors with conservative densities and better grades of steel should have lower losses than many metric design motors. Vintage 25 or 40 Hz motors have thicker laminations, so the eddy-current losses will be higher than for a similar 60 (50) Hz core. Finally, since eddy-current losses are proportional to the frequency squared, the core losses during a test applied at 60 Hz should be evaluated carefully for motors that operate at higher frequencies.
Some core loss testers also report power factor (PF). The stacking factor changes the apparent PF, so thinner laminations will skew the results of the PF reading. Stacking factor can be described as the percentage of the core length that is steel, as opposed to inter-laminar insulation. While the PF is good information, and useful to the repairer, it is important to recognize that the lamination thickness affects the PF reading. In general, for a 60 Hz stator, PF should be between .35 and .75.
Core losses can be decreased by judicious watt-knocking. Use a mallet to bump the laminations near the top of the teeth to break loose laminations that are fused at the surface. Check the core losses periodically to confirm that the losses are still decreasing. If your core loss tester reports PF, watch both PF and watts loss per pound. Graphed, the losses will decrease while the PF increases, so stop at the point where they intersect.
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print