How is the rotor tested?

Whether a rotor passes or fails comes down to your criteria. In our measurement software, you set targets for the magnetic flux and the pole properties. The test bench scans the rotor automatically and checks the measured magnetic field against your criteria. This article takes a closer look at the metrics used to evaluate each pole, and at how test engineers can arrive at suitable target values.

What is judged?

The magnetic field scan yields a set of values for every pole. These include:

• Peak – the highest flux density of the pole
• Polarity – the area under the curve, i.e. the magnetic flux of the pole
• Width – the distance from zero crossing to zero crossing
• Center – the flux density at the pole’s midpoint

Advanced options and overviews add information on the FFT properties, the pole subsegment widths and the skew angle.

These values alone do not tell you whether a pole is good or bad. That takes a comparison. There are three options for defining targets and, with them, spotting deviations.

Three options for spotting deviations

Option 1: Pole against pole

First, the mean of all poles is formed for every criterion – the mean peak, for example, or the mean flux density. Magnitudes are used, so north and south poles count alike.

Then each pole is compared with that mean, and its deviation is given as a percentage. If it lies outside the tolerance, the pole is flagged. For example, with a tolerance of ± 4 %, a pole whose peak sits about 8 % below the mean of all poles fails.

This option finds local faults, such as a weakly magnetised or damaged magnet, without needing a reference. The drawback: a rotor that is weak all over still looks even here – every pole deviates by the same amount, and so barely from their shared mean. That is what options 2 and 3 are for.

Option 2: Target values from your technical requirements

The target values come from your own documents: the drawing, the specification or the simulation. For every criterion, you set a target and a tolerance – for example, for an eight-pole rotor:

Every pole is tested against these values. If, say, the flux density at the center of a pole falls below 360 mT, the rotor fails.

The advantage: this option works from day one, even before a proven good rotor exists – during a production ramp-up, say, or with a new rotor design. And because the targets are absolute, a rotor that is weak all over is caught as well.

Option 3: A reference rotor as the comparison

If you already have a good rotor, it opens up further possibilities. With it as the reference, you can place a tolerance band around selected signal curves from its measurement and compare them against signal sections in other measurements. Every further rotor is then judged as follows:

• If its signal curves stay within the band, it passes.
• If, for example, all of its poles are 10 % below the reference, it fails – even though the poles are even among themselves.

The reference rotor therefore catches exactly the faults that stay invisible pole against pole: deviations that affect the whole rotor.

Conclusion: testing by your own criteria

Comparing pole against pole finds local faults without a reference. Target values and a reference rotor show whether the rotor as a whole meets your requirements.

You can mix the options freely, criterion by criterion. For example, you might check the polarity against a nominal target and the peak against the mean of all poles. In every case, you decide which criteria and tolerances apply – and the result is a clear pass or fail for every rotor.

Further readingWhy the magnetic field is the right place to look in the first place is covered in Why measure magnetic fields in electric motors?→