PVC Extrusion Process: Parameters, Tests and Defects

Most of the profile failures we get asked to investigate arrive with a clean inspection report attached. Wall thickness within tolerance. Cross-section matching the drawing. Surface gloss acceptable. Then six to eighteen months later the edge banding cracks at the clip line, the diffuser cover yellows, the trim goes brittle enough to snap during a routine service call. By the time anyone traces it back, the tooling has been amortised, the batch is installed across a building, and the conversation turns into a commercial argument instead of a technical one.

The reason this keeps happening is that dimensional accuracy and molecular integrity are set by different parts of the same line, and only one of them shows up on an incoming inspection sheet. Understanding the PVC extrusion process well enough to specify it properly means knowing which stations control geometry, which stations control the polymer network underneath that geometry, and which measurements distinguish the two.

Why a Profile Passes Inspection and Fails in the Field

PVC production is not one process. It is two, and they are complex in different ways: compounding a resin into a processable dry blend, then extruding that blend into a shape. Research groups building precision PVC extrusions for scientific instrumentation have documented this split explicitly, along with the custom in-line tests they had to develop because standard industrial QC did not catch the failure modes they cared about (arXiv). That project made scintillator modules rather than architectural trim, so its acceptance criteria do not transfer. The structural point does, and a commercial rigid PVC extrusion process faces the same two-stage problem at lower stakes and higher volume.

 

Suspension PVC does not arrive as a homogeneous melt-ready pellet. Each grain is a porous cluster of roughly one to two micron microparticles, themselves built from sub-micron primary particles, with crystallinity in the region of five to ten percent, figures that are standard in the PVC processing literature rather than anything we measured ourselves. Under typical profile-processing conditions, rigid PVC does not behave as a fully homogeneous melt the way polyethylene does. What the extruder is actually doing is breaking that hierarchy down in stages and rebuilding it as an entangled network.

 

If that network only partially forms, the part still holds its shape, because shape is fixed downstream by the calibration tooling, not by the polymer state. The internal boundaries that never closed sit inside the wall as stress concentrators, and they quietly cost you impact resistance, chemical resistance and UV stability at the same time. That is the practical stake in every parameter discussed below: a well-controlled PVC profile extrusion process is not one that hits the drawing, but one that hits the drawing and can prove the wall is fused.

How the PVC Extrusion Process Steps Actually Break Down

Dry blending and compounding

Rigid PVC starts as resin plus a stabiliser system, lubricants, processing aids, impact modifier and filler, blended hot in a high-intensity mixer and then cooled. K-value selection matters here and is not a detail. K-67 is the most common choice for rigid profile and pipe because it balances melt strength against the shear needed to fuse it, though thinner sections and heavily filled compounds legitimately push processors toward lower K-values; treat it as the usual starting point, not a rule. Feed consistency at this stage propagates through everything downstream, since every temperature setting later in the industrial PVC extrusion process is calibrated against an assumed blend.

Moisture is the one contaminant people underestimate. PVC itself is an addition polymer and does not absorb water meaningfully; its fillers and pigments do. Total water content above roughly 0.1 percent by weight, a threshold quoted widely across extrusion troubleshooting practice, will flash off as the melt leaves the die and show up as surface porosity or splay that nobody tunes out with temperature.

Feeding and metering

Gravimetric feeding rather than volumetric is what keeps output stable when bulk density drifts between compound batches. Feed fluctuation reads downstream as melt pressure fluctuation, which reads as wall thickness variation, which is the single most common reason a custom PVC extrusion process produces parts that gauge fine in a five-piece sample and fail on assembly across a full run.

Rigid PVC Extrusion Temperature Profile and Plastication

This is where the polymer state is decided. A moderate compression ratio, generally cited in the 1.3 to 1.6 range for rigid PVC screws, is what balances the two failure directions: too much shear degrades the compound, too little leaves the particle structure incompletely broken down. Compression ratio is defined differently across screw geometries, so the figure is only meaningful alongside the screw type it describes. Screw speed, barrel profile and residence time all trade against each other inside a narrow PVC extrusion process window, and that window narrows further as filler loading rises.

Vacuum venting

The vent port pulls entrained air and volatiles out of the melt before it consolidates. There is a counterintuitive constraint here that catches experienced processors: if the material has already gelled too completely by the time it reaches the vent, the air has nowhere to go and it stays in the wall as voids. Running the front zones hotter to "improve plasticisation" is a common and expensive way to create porosity in a PVC extrusion process for profiles.

PVC Extrusion Die and Calibration Process

Die design is the hardest engineering problem in profile work, and it is not about the shape of the opening. It is about flow balance, getting mean exit velocity uniform across a cross-section that has thick and thin sections feeding from the same manifold. Unbalanced flow distorts the profile as it exits, and no amount of downstream tooling fully corrects it. The sections that come back to our mould shop for correction are almost always those with a large wall-thickness ratio between the thickest web and the thinnest leg, and that ratio is visible on the drawing before any steel is cut. Complex sections, dual-durometer seals and two-colour parts are handled by running two or three melt streams into a single die, which adds another balancing problem on top of the first.

Die swell then has to be compensated in the tooling geometry, not corrected by pulling harder on the haul-off. Stretching a profile to size after the die builds in frozen orientation stress that releases later as bow, twist or shrinkage in service.

The profile leaves the die soft, and what happens in the next few metres fixes the final dimensions. Vacuum calibration tables hold the section against a sizing sleeve while water cooling stabilises the skin, and the temperature uniformity of that water along the tank is the variable most specifications forget even though it governs half the dimensional outcome of any PVC extrusion process. On our own lines, running rigid PVC trim sections in the 1.5 to 3 mm wall range, a few degrees of drift across the tank has been enough to move wall thickness by around half a millimetre over a single shift. That figure comes from our run logs and in-line gauging records rather than published data, and it scales with wall thickness instead of applying universally. Drift of that kind is invisible in a five-piece sample and painfully visible when someone is installing three hundred metres of trim and the pieces stop matching.

Haul-off and cutting

 

Gelation and Fusion Are Two Different Stages

 

Gelation and fusion are separate stages of the same PVC extrusion process, not synonyms: gelation happens before venting, fusion continues after degassing and runs toward the screw tip.

 

Gelation operates on the coarse structures (grains, sub-grains and agglomerates) at a material temperature in the region of 150 °C, upstream of the vacuum port. Once degassed, the void-free gelled material moves on and disperses down to primary particles of roughly one micron, which is the first stage of fusion. The second stage begins when crystallites inside those primary particles melt and release chains that entangle with chains from neighbouring particles, and it continues to the end of the screws. Measured at the screw tip entering the die orifice, UPVC melt temperature is conventionally targeted in the 180–190 °C band. Both figures are industry working values for unplasticised PVC, not universal setpoints; heavily filled or impact-modified compounds shift them.

 

The operational consequence is worth stating plainly, because it changes how you read a temperature profile: fusion is not supposed to still be happening in the die.

 

The harder question is how you tell. A die that runs hot enough to finish the job looks identical on the panel to a die that is simply set correctly, and melt temperature at the tip will read in range either way. What separates them is the shape of the pressure trace and the torque signature upstream, plus what the extrudate does when you cut it. That diagnosis is one an experienced operator makes on the machine rather than from a specification sheet, and it is the reason we do not quote process parameters off a drawing alone.

 

PVC Extrusion Gelation Test Methods and What They Can Prove

There is no universal fusion percentage unless the product type and the test method are specified first.

Two figures circulate in the literature and they are routinely quoted against each other as if they sat on one axis. One body of guidance, working primarily from solvent-immersion evidence on pipe and pressure applications, treats optimum gelation as an 85 to 95 percent band. Another, working from crystallite-melting behaviour in rigid profile, puts optimum mechanical properties at 65 to 70 percent and warns that pushing higher degrades the tie-molecule network. They are not in conflict. They measure different things, by different methods, on different products, and averaging them or picking the higher one is a category error.

 

The standards themselves are narrower than most suppliers imply, which matters the moment anyone offers a certificate as proof that a PVC extrusion process was under control. ASTM D2152 covers adequacy of fusion in extruded rigid PVC pipe and moulded fittings by acetone immersion, and the standard's own scope section states three limits worth quoting to any supplier: it distinguishes only between inadequately fused and adequately fused PVC, it cannot detect thermally degraded material at all, and it is explicitly not to be used as the sole test specification for purchasing (ASTM International). EN 580 occupies similar territory for PVC-U pipe using dichloromethane. Neither is written for architectural or lighting profile.

 

That is why "the batch passed solvent testing" is close to meaningless on its own, and why the useful question is which method, adapted how, on which surface of which section. Here is how we resolve it internally.

 

Product type	Method we use	What a pass actually proves	Our acceptance condition
Rigid architectural and lighting profile	Dichloromethane immersion, adapted internally from the PVC-U pipe method. Immersion time and specimen preparation set by us, not by the standard	Absence of gross under-fusion on the tested surface	No surface whitening, no tackiness, no swelling at cut edges; any of the three stops the batch
Rigid tube and thick-wall section	Same immersion screen, applied at the bore and outer wall separately	Under-fusion localised to one surface, which points at a specific zone	Both surfaces must pass; a single-surface failure triggers a screw and temperature review, not a re-test
Any section where the customer specifies a standard	The customer's nominated method, run to that standard's own procedure	Whatever that standard says it proves, no more	Written into the PO and reported against by clause
Colour-critical and clear PC or PMMA parts	Immersion screening does not apply; dimensional and optical checks govern	Not a fusion question	Colour measurement and visual criteria agreed before tooling

 

 

Two things this table deliberately does not claim. It does not claim our immersion screen is a certified standard test on profile, because no such standard exists for these sections. And it does not claim a pass equals fitness for purpose: by the reasoning ASTM applies to its own method, a solvent screen detects a failure mode, it does not certify quality.

 

PVC Profile Extrusion Process Parameters That Move the Needle

 

Seven variables account for most avoidable profile failures. The fourth column is the one worth copying into a supplier conversation, because a parameter you cannot verify is a parameter you are taking on trust.

 

Parameter	Working range or condition	Consequence of drift	Where it is verifiable
Melt temperature at die entry	180–190 °C for unplasticised PVC	Low: incomplete fusion, brittle part passing dimensional QC. High: degradation, black specks, HCl evolution	Melt thermocouple log for the run; ask for the trace, not the setpoint
Gelation location	Complete around the vacuum port, near 150 °C	Too early: air cannot be extracted, voids in the wall. Too late: unfused primary particles carried into the die	Barrel zone profile plus vent observation notes
Screw compression ratio	Approx. 1.3–1.6, defined against the specific screw geometry	High: overheating and thermal degradation. Low: insufficient shear, incomplete particle breakdown	Screw drawing or spec sheet on file
Residence time	Set by screw speed and output rate together	Short: under-fused melt. Long: stabiliser consumed, colour shift and degradation	Output rate and RPM from the run record
Vent vacuum level	Continuous, monitored	Loss of vacuum puts moisture and entrained air straight into the wall	Vacuum gauge log or alarm history
Calibration water temperature	Uniform along full tank length	A few degrees of variation moves wall thickness measurably; magnitude scales with wall	Tank inlet and outlet temperature record
Haul-off to output ratio	Synchronised, closed loop	Thickness variation, drawn-down orientation, later bow and shrinkage	In-line gauging trend for the run

 

Ranges of this kind describe a starting envelope for the PVC extrusion process, not a recipe. Three variables move them: wall thickness distribution, filler loading, and the stabiliser and lubricant system. In practice an engineer touches them in that order. Geometry first, because a thin leg feeding off a thick web changes the achievable window before any setpoint does. Filler second, because it changes both shear and thermal conductivity. Formulation last, because it is the slowest and most expensive to iterate.

 

PVC Extrusion Defects, Causes and Corrective Actions

 

Several common PVC extrusion defects have a standard first response on the shop floor that makes them worse. The middle column is what most people do; the right-hand column is where to look before you touch anything.

 

Symptom	Usual first move	First check instead
Low surface gloss	Raise melt temperature	Lubricant change log and any plate-out on the die lips, then verify actual melt temperature and torque against the last good run. Over-lubrication is the leading cause, but confirm it before changing formulation
Failed solvent immersion screen	Raise temperature	Whether the compound is over-lubricated externally, and whether process aid level was changed. Under-fusion and over-lubrication produce the same result by different routes
Voids or bubbles in the wall	Dry the material	Gelation state at the vent zone and vent vacuum history. Wet material is possible; trapped air from over-early gelation is more common
Brittle extrudate, cracking months later	Blame the compound	Barrel temperature, screw speed, residence time, stabiliser age, and screw geometry for dead zones
Yellowing or black specks	Look for contamination	Thermal history at shear-concentrating features: screw tip, adapter, die dead spots, before assuming foreign matter
Metal particle contamination	Check incoming material	Barrel and screw condition. Where thermal degradation has been occurring, evolved HCl accelerates corrosion of conventional steel, and the wear becomes a contamination source
Die drool on the lips	Change compound	Timing tells you. Present from start-up points at the incoming material; developing after hours of running points at the process, unless the compound is genuinely under-stabilised (PlasticsToday)
Cracked corner welds after fabrication	Adjust the welder	Fusion state of the profile body. Corners are welded at around 250 °C in typical uPVC window fabrication; if the body was never properly fused, no welder setting saves the joint
Wall thickness drifting across a run	Adjust the screw	Calibration water uniformity and haul-off synchronisation, before you touch the screw

 

 

Timing and first-check logic will tell you which side of the fence a problem sits on. Neither identifies root cause. Die drool still needs the deposit collected on a swab and analysed to know whether you are looking at oxidised low-molecular-weight polymer or a volatile additive, and that determines whether the fix is a die-lip temperature change or a formulation conversation with your compounder.

 

Twin Screw vs Single Screw PVC Extrusion: When Each Is Defensible

 

For dry-blend, filled rigid profile compounds, twin-screw is generally the right answer, because the distributive mixing and degassing that determine fusion level are exactly what twin-screw geometry provides. A properly designed single-screw line running pre-compounded pelletised material is a legitimate alternative, since the compounding work has already been done off-line and the extruder is being asked to melt and pump rather than to fuse from powder.

 

What separates the two cases in practice is what you are feeding and what you need to remove.

 

Feedstock and requirement	Defensible configuration	What to verify
Dry blend powder, filled, rigid profile	Twin-screw	Vent performance and fusion test results on the actual section
Pre-compounded rigid pellet, low filler	Single-screw, correctly specified screw and venting	Compounder's certificate plus the extruder's own fusion records
Flexible PVC, most sections	Single-screw	Dimensional consistency; fusion is a far weaker constraint
Any rigid compound, long production campaigns	Either, with barrel material specified	Barrel condition records and screw wear inspection interval

 

Barrel material deserves its own line in the specification. During thermal degradation PVC releases HCl, so repeated overheating, long residence at temperature, or dead zones in the flow path progressively raise both corrosion and contamination risk. Bimetallic barrels lined with nickel-based alloy are how that long-term risk is controlled, which is what we run alongside conical twin screws on the rigid PVC extrusion profiles that make up most of our filled-compound work; fillers and stabiliser systems also make screw and barrel inspection a scheduled item rather than a reactive one. Normal, well-stabilised processing does not dissolve your machine. What it does mean is that a PVC extrusion process which has been running hot or with dead zones will show up in the barrel long before it shows up in a customer complaint.

 

Formulation, Stabilisers and Recyclate Reality

 

Every additive in a PVC compound costs money and buys a specific property, and the question worth asking is whether the property matches the actual service environment. A furniture edge band living indoors does not need the UV stabiliser package a window profile needs for twenty years of direct exposure. In the quotations we have run for indoor architectural and lighting sections, specifying an outdoor-grade package where it was not needed added roughly a fifth to a quarter of material cost for no functional benefit. That range comes from our own quotation comparisons on those product categories and does not generalise to every PVC part. It is why we discuss application environment before quoting, and why choosing the polymer and the additive package is a decision that belongs before tooling rather than after the first sample.

Four conditions decide whether the outdoor package is worth paying for: direct sun exposure versus shaded or indoor service; the colour warranty you are being asked to give downstream; whether the part is structural or cosmetic when it embrittles; and expected service life against replacement cost. Get those four answers wrong in the conservative direction and you overspend on every metre for twenty years; get them wrong the other way and you replace an installation. Which side a given section falls on is a conversation about the application, not a table lookup.

 

The stabiliser landscape has moved decisively and is worth understanding as a buyer. Europe completed its lead-stabiliser phase-out roughly a decade ago, calcium-zinc and organic systems are now mainstream, and staged restrictions across Asia and Latin America are driving a second generation of lead-free chemistry focused on recyclate compatibility and processing latitude rather than toxicity alone (Plastics Engineering, SPE). When we quote the same architectural section for an outdoor programme against an indoor one, the stabiliser and UV package is usually the single largest line-item difference in the compound cost, and it is also the item customers most often carry over unchanged from an old drawing without rechecking whether the application still justifies it.

 

Recycled content is where buyers most often ask the wrong question. "Do you use recyclate" is less useful than "what is the traceability of the stream". The European industry recycles hundreds of thousands of tonnes of PVC annually under a formal voluntary framework with independently monitored targets (VinylPlus), and legacy additives from decades-old window profiles are managed through closed loops that keep material traceable from profile to profile. A regrind stream with known provenance and consistent K-value is a legitimate input to a custom PVC extrusion process; an unidentified mixed stream is not, whatever price it comes at. Our own position is straightforward: recycled or regrind content is used only where it has been agreed with the customer in writing, and where it is used, the source, the proportion and the affected layer are stated on the quotation rather than discovered later.

 

How to Read Our Line Data

Numbers from a supplier are worth exactly as much as the record behind them, so each figure below is listed with what it is derived from and what it does not say. We operate more than forty extrusion lines in Dongguan with annual output above 2,000 metric tonnes across PVC, PC, ABS and engineering polymers; the company was founded in 1998, moved into co-extrusion tooling in 2010, and rebuilt the mould department in 2023.

Metric	Figure	Derived from	What it does not claim
Cpk on critical dimensions	1.52 average, rolling six months	SPC records across critical dimensions on rigid PVC profile lines	An average, not a worst case. It approximates "two-thirds of the tolerance band consumed" only where the process is roughly centred and Cp sits close to Cpk; ask for Cp, the worst dimension and the sample size for your section
Specific energy consumption	118–135 Wh/kg	Drive logs and utility records after the 2021 VFD and induction heating upgrade	A range across products, not a per-job figure. Lower SEC tends to accompany steadier melt at comparable output, but it is not a proxy for melt quality on its own
Shipped-batch compliance	100% of shipped batches pass the immersion screen	Release records	Not first-pass yield. It means non-conforming batches do not ship, not that every batch passes first time. First-pass yield and rework rate are separate figures and we will quote them per product family on request
Mould design turnaround	Within 72 hours of receiving a usable drawing	Design release record	Design release, not sampled parts
Production sampling	15–20 working days by complexity	Sampling records	Depends on section complexity and tooling queue; confirmed per project in the quotation

 

None of that is verifiable from a web page, which is the honest position for any supplier quoting figures for its own PVC extrusion process. What is verifiable is the paperwork, so here is what a first-article package from us actually contains: batch identifier and production date; the critical dimensions from your drawing with measured values and the instrument used; sample size and measurement positions along the run; the immersion screen method as applied, duration and observed result; colour measurement where colour is specified; and the name of the person who released the batch. Ask for a redacted example before you place tooling. Any extruder that cannot produce one is telling you something.

 

What to Ask Any Extruder Before You Place Tooling

Tooling is the point of no return in a profile programme, and most of the risk retires with six questions asked beforehand. The value is not in the questions. It is in being able to tell a weak answer from a substantive one.

Question	Weak answer	Substantive answer	Evidence to request
Which method do your fusion reports use?	"We test every batch"	A named method, whether it is a standard procedure or an internal adaptation, and what the result does and does not prove	One redacted test report showing method, duration and observation
What is your Cpk on my critical dimensions?	A single number with no context	Cp and Cpk, the period, the sample size, and the worst-performing dimension	SPC output for a comparable section
Single-screw or twin-screw for this compound?	"We have modern equipment"	A rationale tied to dry blend versus pre-compound, filler level and venting need	Screw specification and line assignment for your job
Is the barrel bimetallic or conventional steel?	Deflection to general quality claims	A direct answer plus the wear inspection interval	Maintenance or inspection record
Are samples from production tooling at production parameters?	"We'll send samples"	Yes, with hand-finished demonstration parts explicitly ruled out	The run sheet that produced the sample
What documentation ships with the first batch?	"Full QC"	A named list of documents	A redacted first-article package

 

Six questions answerable by email will expose weak process documentation faster than a scheduled audit will, because an audit is announced and a document request is not. What they still cannot tell you is whether a given supplier's PVC extrusion process window suits your section. A 1.2 mm snap-fit leg and a 4 mm structural web behave differently in the same die, and that assessment only happens once an engineer has your drawing in front of them. If you want that assessment rather than a quotation, send the cross-section for a technical review and say so in the subject line.

 

What We Make, and What We Turn Down

 

Our PVC work concentrates on LED lighting components including diffuser covers, lens profiles and mounting channels; architectural trim such as edge banding, corner guards and decorative moulding; furniture hardware including edge banding and cable management; and industrial sections like cable trunking and protective strip. Materials run across rigid and flexible PVC extrusion profiles, PC, ABS, PMMA and selected engineering resins, and OEM programmes on drawing are the bulk of what the lines actually run.

There are jobs we decline, and saying so upfront saves both sides a fortnight. We do not take medical-grade tubing, because we hold no cleanroom certification and would rather say that than qualify it away. Profiles exceeding 200 mm on any single dimension are evaluated case by case rather than accepted by default. Annual volumes under 500 kg do not amortise tooling and setup, so the economics do not work for either party.

Frequently Asked Questions

What are the steps in the PVC extrusion process?

Dry blending and compounding, gravimetric feeding, plastication in an extruder, vacuum venting, melt metering, the profile die, vacuum calibration, cooling, haul-off and cutting.

What melt temperature should rigid PVC reach?

Unplasticised PVC is conventionally targeted at 180–190 °C measured at the screw tip entering the die, with gelation occurring earlier near the vacuum port around 150 °C.

How is fusion level tested?

By differential scanning calorimetry, dichloromethane immersion, or acetone immersion under ASTM D2152, noting that D2152 is written for rigid PVC pipe and fittings and, by its own scope, cannot detect thermal degradation and is not a standalone purchasing specification.

Why does a PVC profile pass dimensional inspection and still fail later?

Because geometry is fixed by the calibration tooling while strength depends on the fused polymer network, and an under-fused wall holds its shape while losing impact, chemical and UV performance.

Do I need a twin-screw extruder for rigid PVC?

For dry-blend, filled rigid-profile compounds a twin-screw is generally preferred; a properly designed single-screw line can be viable for pre-compounded pelletised material.

Can recycled PVC be used in extruded profiles?

Yes, where the stream is traceable and consistent in K-value and additive history, and where its use and proportion are agreed in writing before production.

 

Working With Us

 

Send a cross-section drawing first. DWG, DXF, PDF or a clear dimensioned sketch all work. Without the geometry, any number we gave you on feasibility, tooling or price would be worthless to both of us, so we do not quote without one.

 

Within 48 hours you get an initial feasibility assessment, material recommendations and a rough cost range, subject to drawing completeness. If the project makes sense on both sides, formal quotation follows with tooling cost, unit pricing at volume breaks and lead time. Sampling runs 15 to 20 working days depending on complexity, and samples come off production tooling at production parameters rather than being hand-finished for presentation. First shipment carries the first-article package described above, and quality issues traceable to our process get resolved without an argument about it.

Technical enquiries about a custom PVC extrusion process go to engineering@dachangplastic.com. Tell us the application, expected annual volume and any specific performance requirement, and you will get an engineer's answer rather than a sales response.