A mains cord fails at the point where it leaves the appliance, and everyone in the industry knows it. The flexing test exists to reproduce that failure in a day instead of a year, and it is one of the few tests where the apparatus geometry decides the outcome as much as the product does.
The test is simple to describe: clamp the appliance, hang a weight from the cord, swing the cord through an angle, count cycles until the conductor breaks. Every one of those steps has a number attached, and every number changes the result.
The four numbers in the setup
| Setting | Common value | Effect if changed |
| Swing angle | 60 degrees either side of vertical, 120 degrees total | A larger angle moves the flex point further from the anchorage |
| Rate | About 60 cycles per minute | Rate changes cord temperature at the bend |
| Hanging mass | 1 kg for a light cord, 2 kg for a heavy one | Mass sets the tensile load through the bend |
| Conductor count check | All poles broken, one by one, not together | The test ends when the first conductor opens |
| Bend radius | Set by the pulley or guide, not by the cord | A tighter guide fails the cord earlier |
| Cycle count | Tens of thousands, set by cord type | The number is a floor, not a target |
The mass is the parameter most often got wrong, because it is chosen from the cord rather than from the appliance. A 0.75 mm² two-core cord and a 2.5 mm² three-core cord do not bend the same way under the same 2 kg, and the standard ties the mass to the conductor cross-section and to the appliance mass class.
Where the cord actually breaks
Almost never at the middle of the free length. The break is at the exit of the anchorage, or a few millimetres outside it, where the cord transitions from constrained to free. That is why the test rig clamps the appliance, not the cord: the clamp is at the point of interest.
Two design details move that point and change the result. A smooth, generous radius at the exit spreads the strain over 10 mm instead of 1 mm and can multiply the life of the cord. A sharp internal edge, or a cable gland whose nut is overtightened, concentrates it and fails a cord that is otherwise fine.
- Conductor broken, insulation intact. Fatigue of the copper strands. The cord still looks perfect and the appliance is dead. This is the normal result.
- Insulation cracked before the conductor. Usually heat or plasticiser loss, not flexing. Check whether the cord was aged first.
- Cord pulled out of the anchorage. Not a flexing result at all: the anchorage failed the pull test and the flexing rig simply revealed it.
The prior ageing step people skip
Many product standards run the flexing test after a period at elevated temperature, because a cord that has lost plasticiser is a different material. A cord straight out of the box is at its most flexible and will pass a cycle count that a cord aged for a week at 70 °C would fail. If the report does not say whether the cord was conditioned, the cycle count means very little.
Writing it down
Record the mass, the angle, the rate, the guide radius, the conductor cross-section, whether the sample was aged, and which conductor opened first. That last one is diagnostic: if the earth conductor opens before line and neutral on a three-core cord, the cord is being bent in a way the design did not anticipate, and the anchorage geometry needs a second look.
A cord flexing apparatus with a replaceable guide set is worth more than one with a fixed pulley, because the guide is what you will need to change when the cord size changes. The rest of the mechanical test range is listed on the site index.
Standards referenced: IEC 60335-1; IEC 60745-1.
