Twin Screw vs Single Screw Extrusion: Which to Specify

The short answer

If you are extruding a profile from a single, well-characterised compound and the formulation never changes, a single screw is cheaper to buy, cheaper to run and easier to maintain. If you are extruding a compound you have to make rather than just melt — PVC dry blend, wood plastic composite, filled or glass-reinforced grades, colour concentrates that must be dispersed — a twin screw is not optional.

That is the whole decision in two sentences. Everything else in this article is the reasoning behind it, the numbers that separate the two machine types, and the cases where the simple rule breaks.

How each machine actually moves material

Single screw: drag flow against a pressure gradient

In a single screw, material is dragged forward by the friction between the polymer and the barrel wall. The screw surface is polished and the barrel is not, so the polymer slips on the screw and sticks to the barrel — that difference is what generates forward motion. The output is described by the drag flow minus the pressure backflow:

Q = ½ · π² · D² · h · N · sinθ · cosθ − (p₁ − p₂) · h³ · sin²θ / (12 · μ · L)

You do not need to compute it. You need to see what it says: the first term is proportional to h (flight depth) and the second is proportional to h³. So the backflow term grows three times faster than the drag term as the channel deepens. A single screw's output therefore falls as die pressure rises, and falls sharply. Raise the pressure at the die by fitting a tighter die and the output drops even though the screw has not changed speed.

That single characteristic drives most of the practical consequences below. On a single screw, output and pressure are coupled. You cannot independently set both.

Twin screw: positive displacement, nearly independent of pressure

A closely intermeshing co-rotating twin screw is a series of C-shaped chambers that are pushed forward as the screws rotate. Each chamber is closed off from the next by the intermeshing flight, so material is conveyed by volume rather than by wall friction. Output is approximately:

Q ≈ 2 · m · V_c · N · η_v

where m is the number of starts, V_c the chamber volume and η_v a volumetric efficiency typically 0.3 to 0.6 for a filled compound. No pressure term appears. That is the defining difference: on a twin screw, output is set by screw speed and feed rate, and die pressure is whatever the die happens to need. You can change the die and keep the output — or keep the die and change the output — without the two fighting each other.

Table 1 — The behavioural differences that follow from the conveying mechanism.

Feeding: the difference that decides most projects

Feeding is where the choice is usually made, and it is rarely discussed because it is unglamorous.

A single screw needs a consistent pellet — roughly uniform granule size, free-flowing, not too fine. Feed it a fluffy PVC dry blend with a bulk density of 0.55 g/cm³ and the screw cannot grip it. The material sits on the flights and rotates with them, output becomes erratic, and the operator compensates by over-heating the barrel, which starts a degradation cascade. Feed it a wood-flour-filled compound and the same thing happens plus abrasive wear.

A twin screw with the correct feed section handles all of it: powder, fluff, fibre, calcium carbonate at 40 %, glass rovings fed downstream, regrind with a wide particle distribution. Because the screws are self-wiping and force material forward, bulk density only affects the fill level, not whether conveying happens at all.

Table 2 — Feedstock handling. This table, not the price list, decides most machine purchases.

Mixing and dispersion

Single screws mix by shear in the metering zone, which is weak and largely distributive — it spreads material around without breaking agglomerates. To get real mixing you add hardware: a Maddock or pineapple section, a cavity transfer mixer, a static mixer after the die. Each of these costs pressure and adds a place for material to hang up.

Twin screws mix by geometry. Kneading blocks with a defined stagger angle give you control over the balance between distributive and dispersive mixing:

Table 3 — Screw elements and what each does. The screw is a configuration, not a single part.

This is the structural advantage no single-screw add-on reproduces: on a twin screw you can place high shear exactly where the pigment needs it and low shear everywhere else. On a single screw, the whole barrel sees roughly the same shear history.

Melt temperature, residence time and heat history

Counter-intuitively, the twin screw often runs cooler despite mixing harder, because it does not need the barrel to supply all the melting energy through conduction. The work is done in a short, controlled section and the rest of the screw conveys.

Table 4 — Typical numbers. The residence time and colour change figures are why twin screws dominate on short runs and multi-colour work.

Look at the last two rows together with the 95th percentile. On a single screw, a small fraction of material sits in the barrel for two or three times the mean. On a heat-sensitive compound — rigid PVC above all — that fraction is what scorches, and the black specks that appear in the profile after four hours of running come from that tail, not from anything wrong with the current settings. A twin screw's narrow distribution is a stability guarantee, not a nicety.

Pressure, surging and dimensional stability

Profile dimensional consistency is mostly downstream — calibration, cooling, haul-off — but the extruder sets the starting condition. If output pulses, the calibration unit sees a varying wall thickness and the vacuum cannot fully compensate.

Table 5 — Surging has different signatures on the two machines, and the diagnosis follows the signature.

Note the "output falls over a run" row. On a single screw, a blinded screen pack raises die pressure, and because output is coupled to pressure, output drops. The operator sees thin wall and turns up the screw speed, which raises melt temperature, which changes the dimensions again. This loop is the single most common reason a profile line drifts. On a twin screw the same blinded screen raises pressure but leaves output essentially unchanged — the fault stays visible as a pressure reading rather than becoming a dimensional defect.

Material by material

Table 6 — The practical mapping. Rigid PVC is the case that made the twin screw a category.

Cost, honestly

Twin screws cost more. How much more depends on size, torque density and whether the machine is a purpose-built profile machine or a compounding machine pressed into service.

Table 7 — Total cost, not purchase price. The energy and changeover columns are where twin screws recover their premium.

Worked example. A line running 350 kg/h, 6,000 h/year, at €0.14/kWh. Single screw at 0.24 kWh/kg uses 504,000 kWh = €70,600. Twin screw at 0.17 kWh/kg uses 357,000 kWh = €50,000. Saving €20,600 a year. On a €120,000 premium that is a six-year payback on energy alone — before counting the changeover time on a line that runs three colours, where the saving is far larger.

Conversely: a line running one natural HDPE profile 24/7 with no colour changes will never recover the premium. The premium is paid for feedstock flexibility, dispersion and changeover speed, and if you need none of those you should not pay it.

How to specify, in order

1. Write down the feedstock. Pellet, fluff, powder, regrind, filler percentage, fibre percentage, moisture content. Not "PVC" — the specific form.
2. Decide whether the compound is bought or made. If you buy ready-compounded pellet, a single screw is a legitimate answer. If you buy resin and additives separately, you need a twin screw.
3. Calculate the changeover requirement. Colours per week, materials per week. Above about four changes a week the twin screw's self-wiping geometry starts to pay.
4. Size on output at the die, not on screw diameter. Compare specific output — kg/h per rpm per unit of screw volume — between candidate machines.
5. Check the torque, not the speed. A high-speed, low-torque twin screw cannot run a high-viscosity filled compound no matter what the datasheet says about output.
6. Confirm the wear package. For filled or glass compounds, the screw and barrel metallurgy matters more than any other specification, and it is the line item buyers cut first.

Table 8 — The specification sheet items that actually determine performance.

A worked selection case

A customer wants to extrude a 60 mm wide hollow window-profile section in rigid PVC, 400 kg/h, two colours on rotation, plus a wood-filled variant for a garden product. The wood variant is 45 % wood flour with a moisture content of 6 % as delivered.

Single screw: cannot feed the dry blend reliably, cannot feed the wood flour at all, cannot vent the moisture, and the colour change takes 40 minutes each way. Ruled out on four counts.

Twin screw, 72 mm, 32:1, co-rotating, with one atmospheric vent and one vacuum vent: feeds both compounds from a gravimetric feeder, devolatilises the wood moisture at the vent, disperses the pigment in a 60° kneading block, and changes colour in eight minutes. Output 520 kg/h on PVC, 430 kg/h on WPC — both comfortably above the 400 kg/h requirement.

The counter-rotating alternative would run the PVC slightly better — lower shear, more uniform — but would struggle with the wood flour and would be a poor mixer. The co-rotating machine wins on the requirement to run both compounds on one line.

Six mistakes buyers make

Table 9 — The recurring errors, and each one's cost.

Frequently asked questions