A magnetic particle bench delivers thousands of amperes, and almost all of that energy ends up as heat in the transformer, the rectifier stack, the contact heads and the cables. The equipment is rated by a duty cycle: the fraction of a repeating on-off period during which current may flow without the windings exceeding their thermal class. A 10% duty cycle at 5 000 A does not mean 10% of 5 000 A, it means current for one tenth of the time.

The rest interval follows directly. If the shot lasts t_on and the rating is D percent, the full period must be t_on x 100/D, so the machine must sit idle for t_on x (100/D - 1) before the next shot. Ignoring that is the usual cause of a bench tripping thermally halfway through a batch, and of contact heads and cable lugs annealing and losing their grip.

Shot length is bounded from below by the technique rather than the machine. In the continuous method the bath flows over the part and the current is applied while the particles are still mobile, with the flow stopped just before the current ends so the indications are not washed away. The current must be on long enough for the particles to migrate to the leakage field, which is why the standards call for a minimum shot duration of about 0.2 s and at least two shots.

Short shots are also why magnetising current is measured with a peak-reading or averaging meter built into the machine rather than a clamp meter. A 0.5 s shot does not give a handheld instrument time to settle.

Worked example

Shot duration	0.5 s
Rated duty cycle	10 %
Shots per part	2
Minimum rest between shots	4.5 s
Cycle period	5 s
Maximum shots per minute	12
Maximum shots per 10 minutes	120
Total current-on time per part	1 s
Continuous method minimum shot	0.2 s
Continuous method minimum shots	2

A bench rated at 10% duty cycle, shot length 0.5 s. The full period must be 0.5 x 100/10 = 5 s, so the rest between shots is 5 - 0.5 = 4.5 s. That allows 60/5 = 12 shots per minute, or 120 in ten minutes. Two shots per part at 0.5 s gives 1 s of current-on time per part, and 0.5 s comfortably exceeds the 0.2 s continuous-method minimum.