Fibre and micro-duct are unforgiving products to wind. There is no plasticity to absorb a mistake — a tension spike either exceeds the minimum bend radius or it doesn’t, and the attenuation test at the end of the line tells you which. Spool tolerances that would be irrelevant on steel wire become the whole specification here.
Four numbers matter, and they interact.
1. Concentricity of the core to the flange seat
On a take-apart spool the core locates on the flanges. If that location is eccentric by even a fraction of a millimetre, the wound product runs out, and the payoff tension varies once per revolution. At low speed you will not notice. At several hundred metres per minute you get a periodic tension signature, and on fibre that shows up as a measurable loss.
Specify total indicated runout between the core surface and the flange bore, and specify it at the diameter that matters. A tolerance quoted at the hub is not the same as one quoted at the winding diameter.
2. Bore fit on the payoff shaft
Too loose and the spool rattles and walks on the shaft, which is both a noise problem and a tension problem. Too tight and operators force it on, which is a changeover-time problem and, over months, a damaged-shaft problem.
State the shaft diameter and its own tolerance, then specify the bore to suit. This sounds elementary and is routinely missed, because the spool drawing and the machine drawing are usually owned by different people.
3. Balance grade at maximum speed
Balance requirements are often quoted as a standard grade without stating the speed. That is meaningless — the unbalance force scales with the square of rotational speed, so a spool that is perfectly acceptable at low revolutions can shake badly at the top of the range.
Give the supplier your maximum payoff speed in rpm and let them derive the balance requirement. If you want to specify it yourself, specify it as a residual unbalance in gram-millimetres, not as an abstract grade.
4. Flange deflection under winding tension
Winding tension pulls the flanges inward. If they deflect, the product at the outer layers is pinched, and on a subsequent payoff it can bind or snag as the flanges spring back. This is one of the more common causes of a spool that unwinds perfectly for nine-tenths of its length and then jams.
The fix is flange stiffness — more section, or a ribbed geometry — rather than simply a thicker flange. Ask the supplier what deflection they expect at your maximum winding tension, and check whether the geometry distributes load towards the hub or towards the rim.
Material considerations specific to fibre
Polypropylene is a reasonable default: dimensionally stable, does not shed, and cheap enough that the flanges can be reused many times over. Two things are worth checking:
- Outgassing and extractables. Generally not a concern with PP in contact with coated fibre, but if the product has a soft primary coating, confirm compatibility rather than assuming.
- Surface finish at the product interface. Any flash, parting-line step or sink mark on the flange inner face is a potential snag point. Specify the inner face as a critical surface.
The assembly joint
On a peel spool, the flange-to-core joint is the feature that makes the design work and the feature most likely to cause trouble. Three things go wrong:
- A step at the joint, which catches fine product
- Progressive loosening over repeated assembly cycles, which turns into eccentricity
- A joint that cannot be separated quickly at the end of the run, which costs more in labour than the spool saves
Ask for a cycle-life figure on the joint and, if the product is fine or sensitive, ask for a sample to test through your actual wind and payoff cycle before committing.
What to send with the enquiry
A spec that gets a useful answer first time includes: product type and diameter, winding tension and speed, maximum payoff speed, shaft diameter and tolerance, spool capacity required, and whether the product is bare, buffered or coated. Add the storage environment — a spool that lives in an unheated warehouse sees a wider temperature range than one that does not.
None of this is exotic engineering. It is mostly a matter of writing down numbers that are already known to somebody on the production floor, and putting them in front of the person cutting the tool.