A 30 kW conveyor motor coupling failed after 14 months. It was a disc coupling rated for 600 Nm on a motor producing 197 Nm at 1450 rpm. On paper, three times the torque. The disc packs were cracked and the bolts were fretted. The customer called the coupling undersized. The torque wasn’t the problem. The service factor and the misalignment were. The motor and reducer were aligned to 0.1 mm at the hub, a good number for a pump but wrong for a disc coupling on a conveyor with a heavy start.

The torque calculation nobody disputes

Coupling torque = 9550 x P(kW) / n(rpm) x K. For 30 kW at 1450 rpm: 9550 x 30 / 1450 = 197 Nm. K is the service factor, and that’s where the real selection happens.

Uniform load, steady drive (pump, fan): K = 1.5. Moderate shock, frequent starts (conveyor, mixer): K = 2.0. Heavy shock, reversing (crusher, press): K = 2.5-3.0.

The conveyor should have used K = 2.0, giving 394 Nm design torque. The 600 Nm coupling clears that. But the application also starts with a full belt. The acceleration torque can run 3x the running torque for the first two seconds. The coupling sees 590 Nm during start, and with a direct-on-line start the peak can hit 800 Nm. The 600 Nm disc coupling is overloaded at every start.

The fix: a bigger coupling (rated 1000 Nm) or a soft start (VFD or clutch). The plant chose both. Problem solved.

Misalignment: what each coupling type tolerates

This is where the type selection happens.

Elastomeric (jaw, tire): angular 0.5-1 degree, parallel 0.5-1.5 mm. Torsionally soft, absorbs shock. Not for high speed.

Grid: angular 0.3 degree, parallel 0.4 mm. Torsionally soft, handles shock well, needs lubrication.

Disc: angular 0.2-0.3 degree, parallel 0.2-0.3 mm. Torsionally stiff, no lubrication, unforgiving on misalignment.

Gear: angular 0.5-1 degree, parallel 0.5-1.5 mm. High torque density, needs lubrication, some backlash.

The failed conveyor had a disc coupling with 0.1 mm alignment, within disc limits on the angular side. But the conveyor frame flexes under load. When the belt loads up, the reducer tilts by 0.15 mm and the motor stays put. The disc pack, rated for 0.25 mm, flexed beyond its limit at every start. The stainless disc packs fatigue-cracked in 14 months.

For a conveyor, a grid coupling would have handled the flex and the shock. It needs grease every 6 months, but it lives. The rule of thumb: torsionally stiff couplings (disc, gear) for precision alignment and steady load, torsionally soft (elastomeric, grid) for shock and flex.

The bore and the fit

A coupling is only as good as the fit on the shaft. The failed coupling was keyed with a set screw. Under the start torque, the hub fretted on the shaft. The set screw couldn’t hold the torque, and the hub rotated a fraction of a degree each start. The fretting ate the shaft and the hub bore.

For torque above 400 Nm, use a keyed hub with an interference fit, not a set screw. The set screw is for small couplings and light duty. The interference fit carries the torque. The key is a backup. The conveyor’s replacement coupling uses a shrink disc hub, no key at all, friction alone carries the torque. It has survived three years.

The selection checklist

Write down the running torque. Multiply by the application’s K factor. Check the start torque (motors have 1.5-3x locked-rotor torque). Compare to the coupling rating. Then check misalignment capability against the real installed alignment and the frame flex under load. Then check the bore size and the fit. Only then look at the price. A coupling that’s right on torque and wrong on misalignment fails exactly like the one on the conveyor.

Torque picks the size, misalignment picks the type, and the fit decides whether it stays on the shaft. The 600 Nm disc coupling wasn’t undersized. It was the wrong type for a flexing conveyor frame with a full-belt start. Check K, check the start peak, check the frame flex, then buy.