Continuous vs Discontinuous Filament Winding

What continuous and discontinuous winding actually are

Continuous vs discontinuous filament winding is not a contest between two qualities of machine. Both routes wind the same E-glass roving, wet it with the same polyester or vinyl ester resin and stiffen it with the same silica sand. What differs is where the pipe is formed, and whether the surface it is formed on ever stops moving.

On a continuous filament winding machine the mandrel is a steel band held in a cylindrical shape by a ring of beams. Friction pulls the band around as the beams turn and a roller bearing lets it travel lengthwise, so the mandrel advances in a slow spiral toward the exit. Raw materials are metered onto that moving surface in a fixed order — release film, veil, roving, chopped glass, sand, resin — and the pipe cures as it travels before being cut and calibrated.

A discontinuous line, also written DFW or fixed-length winding, uses a rigid steel mandrel between 6 m and 12 m long. It is prepared, wound layer by layer, cured, and the finished pipe is extracted before the next one starts. Every other difference follows from that single structural fact.

How each route builds the pipe

The continuous cycle is a sequence of metered applications on a moving surface: release film, then a C-veil, then the material package that defines the pipe — continuous glass roving in the circumferential direction, chopped roving for axial reinforcement and impact resistance, silica sand for stiffness, and resin binding them. Electronic sensors meter each stream to the design recipe, which is why the laminate emerges as a structural sandwich: reinforced skins separated by a compact, silica-filled core that carries bending stiffness.

Note where the axial strength comes from. On a continuous line, continuous glass is spent on the hoop direction and axial duty falls to chopped strand, because the governing stress in a buried pressure pipe is circumferential and that is where continuous fibre buys the most strength per kilogram. The trade is deliberate: an efficient hoop-dominated pipe, bought at the cost of lower axial capability.

The discontinuous cycle is the one most engineers picture. The mandrel is cleaned and released; the liner goes down as surface mat or veil with a resin-rich laminate and is cured to full hardness before anything structural touches it; then the structural stack is wound, alternating helical passes between 45° and 70° with hoop passes at 85° to 90°, and silica sand dosed between glass layers wherever stiffness rather than pressure governs. Cure, joint groove grinding and hydraulic extraction follow.

The difference that matters commercially is angle freedom. A discontinuous machine stores a winding program per pipe size, so the same hardware can lay a 54.7° balanced spiral for a closed-end pressure case, then switch to a hoop-heavy build for a thin-wall large-diameter main. A continuous line runs one architecture and optimises it: continuous vs discontinuous filament winding is ultimately a question about how much angle freedom your order book needs, weighed against how much continuous output you can sell.

Head-to-head comparison

Most rows below follow from the forming carrier rather than standing alone, which is why continuous vs discontinuous filament winding is best read as one chain of cause and effect: a band that never stops produces one kind of pipe cheaply, and a mandrel that does stop produces any kind of pipe you can program.

The last three rows decide more projects than the first nine: crew size and floor space are fixed costs that arrive every month and scale with the diameter envelope you buy, not with the tonnage you sell. Note also that output is expressed differently on each side — kilometres per year against a cycle time multiplied by mandrel count — and comparing those two figures directly is the most common mistake in machine tenders.

Materials, resin systems and wall design

Both routes buy the same material families: silane-sized E-glass direct roving, a thermoset resin, silica sand, a C-veil or surface mat for the corrosion barrier, and release film or agent. Resin selection follows the chemical service rather than the machine: isophthalic polyester for water and mildly aggressive media, bisphenol-A vinyl ester for concentrated acids and chlorinated duty, epoxy where temperature and mechanical requirements rule polyester out. Service temperature for standard laminates sits between -50 °C and +70 °C, with individual projects specifying higher.

What actually separates the two routes on materials is the dosing regime. A continuous line meters each stream — roving, chopped strand, sand, resin — from electronic sensors against the stored recipe, and can run a dual resin system so the inner liner carries a corrosion-grade resin while a cheaper grade fills the structure. A discontinuous line sets resin content indirectly through bath viscosity, immersion depth and line speed, and adds silica sand in discrete passes at a resin content as low as 15 to 17 percent, up to 20 mm thick in one pass.

Material choice therefore rarely decides continuous vs discontinuous filament winding; the dosing regime does. That distinction shows up in the quality record rather than on the shop floor, and the test methods are identical on both sides: resin-to-glass ratio by ignition loss (ASTM D2584), cure state by Barcol hardness (ASTM D2583), dimensions by ASTM D3567, long-term design basis by ASTM D2992 and strain corrosion by ASTM D3681. Product standards follow the market rather than the machine: ASTM D2996 for filament-wound pipe, ASTM D3517 and AWWA C950 for pressure pipe in North America, ISO 14692 for oil and gas, BS 5480 in the United Kingdom, GB/T 21238 in China and DIN 16965 in Germany. A purchaser should require the supplier to state which of these the line is set up to demonstrate, because the test equipment and calibration fit-outs that ship with a GRP pipe production line depend on that answer.

 

Output, energy and the cost logic

Published figures make the decision clear. A mid-size continuous GRP pipe production line is rated at roughly 200 km per year of DN800 PN6 pipe over 300 production days and three shifts, at line speeds up to about 40 m per hour. Installed power runs from about 100 kW at DN1200 class to 200 kW at DN4000 class, and one supplier quotes a 590 kW installed load with an average draw near 160 kW — a reminder of how much of the cycle is metering rather than driving. Floor space runs from roughly 10 x 26 m to 15 x 29 m with 5 to 7 m of clear height; the crew is six to seven per shift.

A discontinuous line cannot be quoted that way. Its capacity is a cycle time — winding, curing, extraction, cleaning — multiplied by the number of mandrels you keep in circulation. Power scales with mandrel count and curing capacity rather than diameter, floor space is dominated by mould storage and pipe handling, and the crew is two to four per shift. Adding a mandrel is a far smaller commitment than adding a band, so discontinuous capacity grows in steps.

Unit cost favours the continuous route clearly: pipe from a continuous line costs on the order of 15 to 30 percent less than pipe from a discontinuous or centrifugally cast line, a gap built from metering accuracy, lower scrap and fewer labour hours per tonne rather than cheaper raw materials. But unit cost is not the decision. The cost case for continuous vs discontinuous filament winding turns on utilisation: a continuous line running at half its rated kilometres carries the same fixed cost as one running flat out, while a discontinuous plant can add a mandrel, win a mixed order and still earn. Suppliers quote scrap rates below one percent; the figure worth testing is the changeover loss every time diameter, pressure class or liner resin changes, since that is the loss a mixed order book imposes on a line built for one architecture.

Which route fits which plant

Work through these six questions in order. The first two eliminate most of the options before capital is discussed.

1. What shape is the order book? Kilometres of one or two diameters against a multi-year programme point to a continuous filament winding machine. A mix of diameters, pressure classes, liner resins and special fittings points to discontinuous winding, where a program change costs a recipe rather than a reconfiguration.
2. Which diameters and pressure classes do your customers actually tender? Decide this from real tender documents, not from a brochure’s maximum. The most expensive mistake in this market is a GRP pipe production line that cannot reach the pressure class its customers ask for, because upgrading the line and its test fit-outs afterwards costs more than choosing correctly the first time.
3. Is the load case axial or hoop dominated? Buried mains with restrained joints, aboveground spans and traffic loading all demand axial capability. Continuous winding provides that with chopped strand; discontinuous winding provides it with continuous helical roving at a designed angle.
4. How much capital, and over what horizon? Continuous winding carries the highest investment of the main routes and needs volume to justify it; discontinuous carries the lowest and tolerates a slower ramp.
5. Does the building fit the machine? Check clear height under the crane hook, bay length, mould storage and utilities before signing: a DN4000-class continuous line wants close to 15 m by 29 m with 7 m of clear height.
6. Can your crew hold the quality record? A continuous line is metered and archived by its control system; a discontinuous line depends on operator discipline around tension, angle and bath condition.

Both routes also have characteristic ways of failing. Continuous lines fail commercially — bought ahead of demand, specified below the pressure range the market tenders, then loaded with a mixed order book that forces changeovers the architecture was never designed for. Discontinuous lines fail operationally: too few mandrels to hit the promised output, and quality drifting between shifts as tension and bath conditions move unrecorded. Neither failure is technical; both are decisions made before the machine was ordered.

FAQ

Is continuous winding better than discontinuous winding? Neither, in the abstract. Continuous winding wins on cost per metre once volume is high and the product band narrow. Discontinuous winding wins the moment the order book mixes diameters, pressure classes, liner resins and non-standard fittings. The honest answer to continuous vs discontinuous filament winding is that the machine is matched to the order book, not the other way round.

What diameter range does each route cover? A continuous line generally runs from about DN250 to DN4000. A discontinuous line starts lower, around DN80, and reaches the same DN4000 ceiling, because its range is set by the moulds you buy rather than the machine frame.

Can a continuous line produce high-angle helical pipe? Not in the same sense. A continuous filament winding machine spends its continuous glass on the hoop direction and reinforces axially with chopped strand. Where a design calls for continuous helical reinforcement at a set angle, that points to discontinuous winding.

Why is continuously wound pipe cheaper per metre? Three reasons, in order of size: material is metered against a recipe rather than judged by an operator, scrap is lower, and labour hours per tonne are lower. Published comparisons put the resulting pipe cost 15 to 30 percent below pipe from a discontinuous or centrifugally cast line.

What output can I expect from each route? A mid-size continuous line is rated around 200 km per year of DN800 PN6 pipe on 300 days and three shifts. A discontinuous line is quoted as a cycle time, so output depends on how many mandrels you run — which makes mandrel count a capacity decision, not an accessory purchase.

How many operators does each route need per shift? Budget six to seven people on a continuous line, mostly feeding roving, veil and sand. Discontinuous winding typically needs two to four, with the smaller crew carrying more of the quality responsibility.

Do both routes meet the same pipe standards? Yes. Both can be built to demonstrate ASTM D2996, ASTM D3517, AWWA C950, ISO 14692, BS 5480 or GB/T 21238, and both are verified with the same test suite: ignition loss (ASTM D2584), Barcol hardness (ASTM D2583), dimensions (ASTM D3567), hydrostatic design basis (ASTM D2992) and strain corrosion (ASTM D3681). What differs is the fit-out that ships with the line.

Which route should a first-time producer choose? Start from the tenders you can actually win. A long single-product pipeline programme justifies continuous winding and its higher investment; an industrial, marine or oil and gas order book — small quantities, higher pressures, particular specifications — justifies discontinuous winding, where entry cost and angle freedom matter more than line speed. In practice, continuous vs discontinuous filament winding is settled by tender documents, not by preference.