Key takeaways
- Bend life is a cycle count to a defined failure, not a feeling about how soft the braid is.
- Strand strain rises as the bend radius falls, so radius is a life decision and belongs on the drawing.
- Test the actual braid, termination and stroke; a result from a different geometry does not transfer.
- Keep the loop in one plane and support it, or the termination will take the movement instead.
What bend life means
Bend life is the number of movement cycles a braid survives before it reaches a defined failure. Failure might be the first broken strand, a stated loss of strands, a rise in resistance or a complete break. The definition has to be agreed before the test, because the count changes dramatically depending on which one is used.
A braid that flexes once during installation has a different duty from one that flexes every time a door opens. Thermal expansion in a fixed bar produces a slow, small movement. A handheld tool, a drawer or a hinged panel produces a fast, larger movement with a known cycle count. The design has to be matched to the duty, not to a general idea of flexibility.
Where the strain goes
When a wire bends around a radius, the outside surface stretches. The strain is roughly the wire diameter divided by twice the bend radius. For illustration, a 0.20 mm wire on a 5 mm inside radius sees about 2 per cent strain, and the same wire on a 2 mm radius sees about 5 per cent.
That is the strain in a single strand. A braid is better than a solid wire because the strands slide and share the bend, so the effective strain across the whole conductor is lower than the worst strand. The weave also lets the section flatten slightly, which reduces the outer fibre strain.
The termination cancels that advantage. At the ferrule or weld the strands cannot slide, so the first few millimetres of free braid take the full movement. That is why most braid failures start a short distance from the joint, and why a support sleeve or a longer free length can add more life than a change of wire diameter.
The first bend is the one that fails
Keep the bend away from the termination and give the braid a gradual transition. A sharp bend immediately after a ferrule concentrates the whole stroke into a short length of wire.
Setting up a bend test
A bend test has to reproduce the service movement. Fix the stroke, the direction, the bend radius, the speed and the tension. A test that pulls the braid straight while it flexes is not the same as one that leaves it slack, and a test at high speed can heat the strands and change the result.
The termination matters as much as the braid. Use the production ferrule or weld, the production pad and the production bolt torque. A test sample with a soldered end will give a different life from a production crimp, and the difference will be in the termination, not in the conductor.
Decide the failure criterion and the sample count before the test starts. One sample gives one number, not a life. A small group gives a distribution and shows the spread. If the field duty is a million cycles, the test should run beyond that with a margin rather than stopping at the first pass.
- 01
Fix the geometry
Set the bend radius, free length and loop plane to match the installation.
- 02
Fix the duty
Set the stroke, direction, speed and cycle count from the service life.
- 03
Use production ends
Test the real ferrule, weld, pad and torque, not a substitute.
- 04
Define failure
Agree broken strands, resistance rise or complete break before starting.
- 05
Run a group
Test several samples so the result is a distribution, not an anecdote.
Reading a bend life result
Bend life results scatter. The same braid, tested on the same rig, can fail at half the cycles of another sample because of a small difference in strand position at the termination. A single number hides that spread and encourages overconfidence.
Ask for the test conditions and the sample count with the result. A median life is useful for comparing designs, but the design limit should be set below the shortest life in the sample, not at the average. For a critical link, a statistical treatment of the failures gives a life at a chosen probability rather than a single figure.
Be careful with extrapolation. A test at a small stroke does not predict life at a large stroke by scaling the cycle count. Strain amplitude, fretting between strands and heating all change with stroke. If the field movement is larger than the test movement, the test does not cover the duty.
Design limits that follow
The minimum bend radius should be set from the strand diameter and the required cycle count, then checked against the space. A radius that is generous for a static link can be far too tight for a link that moves every second. If the space forces a tight radius, reduce the wire diameter rather than accept a short life.
The loop length has to be long enough for the stroke. A loop that is pulled straight at the end of the stroke has no reserve, and the strands take the load directly. A loop that is too long can rub, vibrate or sag into another part. The right length is the shortest one that keeps the braid slack through the full movement.
Keep the movement in one plane. A braid tolerates a small twist, but repeated torsion concentrates the strain at the termination where the strands are clamped. If the application needs rotation, use a mechanical joint that allows it, or a braid designed and tested for torsion.
Improving life without changing the envelope
The first lever is wire diameter. Halving the strand diameter halves the strain for the same radius, or allows the same strain at half the radius. A finer braid costs more and is more easily damaged at the termination, but it is often the only way to get life out of a tight space.
The second lever is lay length. A longer lay lets the strands move along each other and spreads the bend. The braid becomes less abrasion resistant, so a sleeve may be needed, but the flex life improves.
The third lever is the routing. Moving the first bend away from the termination, adding a support that carries the loop without clamping it, and removing a sharp edge from the path can all be done without changing the conductor. These changes are usually cheaper than a new braid construction and are often the difference between a field failure and a long life.
| Variable | Change | Effect on bend life |
|---|---|---|
| Wire diameter | Finer wire | Longer life at the same radius |
| Bend radius | Larger radius | Lower strand strain, longer life |
| Lay length | Longer lay | More strand movement, longer life |
| Stroke | Larger stroke | Higher strain, shorter life |
| Termination | Sharper transition | Stress concentration, shorter life |
| Loop plane | Twist or out of plane | Edge loading, much shorter life |
| Support | Clamped too tightly | Moves strain to the joint, shorter life |
The table gives direction, not magnitude. The actual life has to come from a test on the real braid, termination and stroke.
Specification language
Write the movement duty as a number. State the stroke in millimetres, the direction, the cycle count over the service life and the bend radius that must be held. Add the loop envelope, the support points and the plane of movement. Without those, the supplier can only guess at the fatigue requirement.
Add the termination and the test. A braid specified for a million cycles with a production ferrule needs a bend test that uses that ferrule, that stroke and that radius. The test report should state the sample count, the failure criterion and the result for each sample.
Finally, state the installation rule. If the link must not be twisted, or the first bend must sit a stated distance from the pad, put it on the drawing and in the assembly instruction.
Frequently asked questions
How many cycles should a copper braid last?
It depends on the stroke, the radius, the wire diameter and the termination. There is no universal figure. The requirement comes from the service life of the equipment, and the design should be tested at that stroke and radius with a margin.
Does a tighter bend make a braid stiffer or weaker?
Both. A tighter bend raises the strain in the outer strands and shortens the life, and it also packs the strands so the braid feels stiffer. If the space forces a tight radius, use finer wire and a longer lay.
Can a braid be used where the joint rotates?
A small twist can be tolerated, but repeated rotation concentrates strain at the termination. For continuous rotation, use a mechanical joint that allows movement without twisting the braid, or a conductor that has been tested for torsion.
Why do braids fail near the ferrule?
The ferrule stops the strands from sliding, so the free braid immediately beside it takes the full movement in a short length. A gradual transition, a support sleeve and a first bend placed further from the joint all reduce that concentration.
Sources and further reading
- ISO 12107, Metallic materials, Fatigue testing, Statistical planning and analysis of data
- IEC 60068-2-6, Environmental testing, Part 2-6: Tests, Test Fc: Vibration (sinusoidal)
- Qumek braid bend test and routing review notes
Qumek Engineering Desk
Engineering team
The engineering team behind Qumek drawings, DFM notes, tolerances and inspection standards. Every guide is reviewed against the drawings and processes used on live programmes.
Need a braid to survive a movement duty?
Send the stroke, cycle count, bend radius and termination. The engineering desk will review the construction and the loop.
Send a drawing