Parison sagging is the single most common defect that limits wall-thickness uniformity in large extrusion blow molding (EBM) parts, and it is the first problem every engineer meets when scaling from small bottles to 20L drums, jerrycans, 220L barrels and IBC totes. A parison is the hollow tube of molten polymer that hangs from the die before the mold closes around it, and during the seconds or minutes it is exposed, gravity pulls the soft melt downward so the tube lengthens and thins under its own weight. In large containers the parison is long, heavy and slow to form, so the bottom of the part ends up thin and weak while the top carries excess material. This article explains why parison sagging happens, how melt strength and molecular weight distribution control it, how to choose the right HMW-HDPE grade, how the die head and accumulator head design limit the sag window, how parison programming and die-land temperature zoning compensate for it, and how to measure sag on the shop floor. Apollo, a Wanplas factory and a specialist in extrusion blow molding machines for more than 20 years, supplies ABLD-series machines rated from 20L to 1500L and accumulator-head lines purpose-built for exactly these industrial containers, and the guidance below reflects the process windows those machines run every day across more than 90 countries.
What Parison Sagging Is and Why It Matters
A parison is the hollow extruded tube of molten plastic that the blow molding machine forms at the die and then captures between the two mold halves. Parison sagging is the downward stretch of that tube caused by gravity acting on a soft, low-stiffness melt. While the parison hangs, the upper portion supports the weight of everything below it, so the tube elongates and narrows from top to bottom. The result is a built-in wall-thickness bias: thick at the clamp line and neck, thin toward the bottom of the finished container. In small bottles this is invisible because the parison forms in roughly one to two seconds. In a 220L drum the parison can take 30 to 90 seconds to extrude and weigh several kilograms, so the bias becomes severe enough to cause bottom thinning, poor top-load strength, visual wall marks and even in-service failure of chemical or UN-rated drums.
Sag matters for three commercial reasons. First, it wastes resin: to keep the thin bottom above the minimum wall, the whole part is over-weighed, often by 8 percent to 15 percent, which raises material cost at Medium-to-High level for high-volume production. Second, it hurts performance, because wall-thickness variation lowers top-load and drop-impact resistance more than the average wall would suggest. Third, it reduces capability: a line that sags cannot hold tight specifications, so scrap rate climbs and OEE (overall equipment effectiveness) falls. For industrial packers shipping 20L to 220L containers, sag is therefore not a cosmetic nuisance but a direct limit on profitability and compliance with transport standards such as UN packaging certification and the regional food-contact rules that reference EU 10/2011 or FDA expectations.
The good news is that sag is predictable and controllable. It is governed by a small set of measurable variables: melt strength, extrusion time, parison mass, die geometry, melt temperature profile and the wall-thickness program. Once those variables are understood, sag can be driven to a level where wall thickness is set by design rather than by gravity. The remaining sections build that understanding step by step.
The Physics Behind Parison Sagging
Parison sagging is, at its core, a balance between the gravitational load on the melt and the melt’s resistance to extensional deformation. The melt behaves as a viscoelastic fluid: it flows like a liquid over long times but resists sudden stretch like a soft solid. The property that resists sag is extensional (or elongational) viscosity at low strain rate, commonly called melt strength. A high melt strength means the molten parison supports its own weight for longer before thinning; a low melt strength means it draws down almost immediately.
Two physical facts explain most of what operators observe. First, sag is time-dependent. The longer the parison is exposed before the mold closes, the more it stretches. This is why extrusion time is the dominant variable for large parts. Second, sag scales with mass per unit length. A thick, heavy parison sags faster than a thin one of the same material because each slice of melt carries more weight below it. The interaction of these two effects is captured by a simple drawdown relation: the change in parison length grows with gravitational stress and with time, and shrinks with melt strength and with die-exit stiffness.
Temperature is the practical amplifier. Extensional viscosity falls sharply as melt temperature rises, roughly doubling in fluidity for every 10 deg C in the processing window for polyolefins. A melt that is 15 deg C too hot can sag two to three times as much as the same resin at the correct temperature, while a melt that is too cold raises die pressure, shear heating and the risk of melt fracture or surface lines. The skill in large-container blow molding is holding the melt in the narrow band where it is fluid enough to extrude cleanly yet stiff enough to hold shape.
Why does the bottom of the part end up thin? As the parison stretches, the extra length must come from reduced cross-section, because the total melt volume is fixed by the shot weight. The diameter and wall shrink progressively from top to bottom, and the portion that later becomes the container base is the most stretched and the thinnest. Programmed wall-thickness control partly reverses this by making the die land thicker at the top of the parison and thinner at the bottom, but the underlying physics still sets the limit on how much correction is possible.
Why Large Containers (20L to 220L) Are Most Vulnerable
The move from a 1L bottle to a 220L drum changes the regime completely. A 1L bottle parison weighs a few tens of grams and extrudes in about one second, so gravitational drawdown is negligible. A 20L jerrycan parison weighs around 1 kg to 1.5 kg and extrudes in roughly 8 to 20 seconds. A 220L drum parison can weigh 4 kg to 9 kg and extrude in 30 to 90 seconds, and an IBC (intermediate bulk container) bottle of 1000L scale is formed by even longer, heavier parisons or by multiple parisons in a multi-layer bottle. The exposure time rises by one to two orders of magnitude, and so does the opportunity for sag.
Large containers also have unfavorable geometry for sag. The parison aspect ratio, the length-to-diameter of the hanging tube, is huge: a 220L drum parison can be more than 1 meter long and only 100 mm to 200 mm in diameter. A tall, slender tube of soft melt is inherently more prone to drawdown than a short stubby one. The die gap is also relatively small compared with the parison length, so the melt leaves the die as a thin-walled tube that must support a long column beneath it.
Production economics reinforce the problem. Large industrial containers are typically run on accumulator-head machines where a full shot is stored and then pushed out fast, but even a fast stroke of a multi-kilogram parison leaves the melt exposed for several seconds while the mold transfers and closes. For continuous-extrusion machines making 20L to 60L containers, the parison forms slowly and sag control becomes the central process challenge. This is precisely the range where Apollo’s ABLD series (20L to 1500L) and accumulator-head designs earn their place, because the head architecture is chosen to minimize the sag window for heavy parts.
The practical consequence is that a resin and a parameter set that work perfectly on a 5L container will fail on a 220L drum. The material must have higher melt strength, the die head must be zoned and, in most cases, an accumulator head must replace continuous extrusion, and the wall-thickness program must use far more points to compensate. Treating large-container sag as a scaled-up version of small-bottle practice is the most common mistake in the industry.
Melt Strength and Molecular Weight Distribution
Melt strength is the property that decides whether a parison holds its shape, and for polyolefins it is governed far more by molecular weight distribution (MWD) and long-chain branching than by average molecular weight alone. A resin with a narrow MWD has a short high-molecular-weight tail, so under extension the chains align and thin out quickly; its extensional viscosity drops as it is stretched. A resin with a broad MWD and a long high-molecular-weight tail, especially with long-chain branching, keeps a high extensional viscosity during stretching, so the parison resists drawdown.
This is why blow-molding grades are deliberately broad-MWD. The high-molecular-weight fraction supplies melt strength and the low-molecular-weight fraction supplies processability, keeping extrusion pressure reasonable. The trade-off is that broad-MWD resins process with more die swell and sometimes more sensitivity to shear, while narrow-MWD resins give cleaner extrusion but terrible sag resistance. For large containers the priority is melt strength, so a broad-MWD HMW-HDPE is the standard choice, and the lowest-MWD grades are reserved for small bottles where sag is not an issue.
Melt strength is best measured directly rather than inferred from melt flow index. A common method is the Cogswell or Gottfert Rheotens drawdown test, where the extruded strand is pulled by a pair of wheels and the force at break defines the melt strength in centinewtons. Another is the extensible sag test where a fixed parison is hung and its length change is recorded. These give a true sag ranking that melt index alone cannot. As a rule, a resin that shows a higher force-to-break in the Rheotens test will sag less on the machine, even if two grades share the same nominal MFR.
Average molecular weight still matters. Raising the weight-average molecular weight (Mw) raises both melt strength and sag resistance, but also raises melt temperature and pressure requirements. That is why the blow-molding window is a compromise: too low Mw and the part sags; too high Mw and the extruder cannot push the melt through the die at acceptable motor load and cycle time. The resin supplier’s grade certificate, with MFR, density, environmental stress crack resistance (ESCR) and molecular weight data, is the starting point for any large-container program.
HMW-HDPE Grade Selection for Minimal Sag
For 20L to 220L drums, jerrycans and IBC bottles, the workhorse material is high-density polyethylene with high molecular weight, almost always an HMW-HDPE. The key grade parameters are melt flow rate (MFR, typically reported under condition E at 190 deg C with a 21.6 kg weight), density, MWD breadth and ESCR. The target window for large industrial containers is an MFR around 0.2 to 0.5 g/10 min, a density near 0.945 to 0.955 g/cm3, and a broad bimodal molecular weight distribution that pairs a high-MW tail for melt strength with a lower-MW fraction for processability.
Bimodal resins are preferred over unimodal because the two peaks let the supplier tune melt strength and stiffness independently. One peak carries molecular weight for sag resistance and ESCR, the other carries processability for clean extrusion. Unimodal broad-MWD grades can also work but often give a narrower processing window. For harsh chemical service, ESCR (measured by the bent-strip or Bell method, sometimes reported as hours to failure) must be high, and a higher ESCR grade is usually a higher-MW grade, which also helps sag, so the two requirements align.
The table below compares typical HMW-HDPE families used in large blow molding. Values are typical ranges and must be confirmed against each supplier’s data sheet, because every producer numbers its grades differently. Cost is shown only as a relative level, not as a price.
Typical HMW-HDPE Grade Families for Large Containers
| Grade family | MFR (cond. E, g/10 min) | Density (g/cm3) | MWD breadth | Relative melt strength | Sag tendency | Typical container | Relative material cost |
|---|---|---|---|---|---|---|---|
| Unimodal broad-MWD HMW-HDPE | 0.3 to 0.6 | 0.947 to 0.952 | Broad | Medium-High | Medium | 20L to 60L drums, jerrycans | Medium |
| Bimodal HMW-HDPE (phase 1) | 0.25 to 0.45 | 0.948 to 0.954 | Bimodal | High | Low-Medium | 60L to 120L drums | Medium-High |
| Bimodal HMW-HDPE (high ESCR) | 0.2 to 0.35 | 0.945 to 0.950 | Bimodal, broad tail | High-Very High | Low | 120L to 220L drums, IBC bottles | High |
| MDPE blend (softer, more flexible) | 0.4 to 0.8 | 0.935 to 0.945 | Medium-Broad | Medium | Medium-High | Fuel cans, flexible jerrycans | Medium |
| Reground-blended HMW-HDPE | 0.5 to 1.0 (effective) | 0.944 to 0.953 | Narrower (degraded) | Low-Medium | High | Non-critical 20L to 60L parts | Low |
Two selection rules stand out. First, for containers above 120L, choose the high-ESCR bimodal grade even at a higher material cost, because its melt strength and crack resistance protect both sag control and in-service performance. Second, keep the regrind fraction controlled: reground scrap has a lower average molecular weight than virgin resin, so a high recycle ratio raises the effective MFR and worsens sag. Blending regrind with a stronger virgin grade, and keeping within the resin supplier’s recommended recycle limit, preserves parison stability. When the scrap volume is high, Wanplas’s Polyretec factory supplies washing and pelletizing lines that recondition blow-molding scrap into a consistent feed, which helps keep the regrind quality stable.
Die Head Design and the Accumulator Head Principle
The die head is where the parison is shaped, and its design directly sets the sag window. In continuous extrusion, melt flows from the extruder through the head and out of the die gap at a steady rate, so the parison forms slowly over the full cycle. For a large part this slow formation is exactly what lets sag grow. The accumulator head solves this by decoupling melt storage from parison formation.
An accumulator head is a large chamber built into the die head that stores a full shot of melt, accumulated from the extruder between cycles, and then pushes it out through the die in a single rapid stroke using a reciprocating piston or a sliding mandrel. The parison forms in a few seconds instead of tens of seconds, so gravitational drawdown is cut sharply. The stored melt is kept homogeneous by a controlled flow path and a screen pack, and the rapid stroke speed is itself a sag-control parameter: faster stroke equals less sag, up to the point where the melt sees too much shear and surface defects appear.
There are two common accumulator architectures. The reciprocating-screw accumulator uses the extruder screw itself as the ram, pulling back to store melt and pushing forward to eject the parison; it is compact and common on medium-large machines. The separate piston accumulator uses a dedicated cylinder fed by the extruder, giving a larger and more uniform shot for very large parts such as 220L drums and IBC bottles, at a higher equipment cost (Medium-to-High versus Low for a simple continuous head). The choice depends on part weight and required shot size: a 220L drum needs several kilograms of melt, which a small reciprocating screw cannot store, so a dedicated accumulator is used.
Die geometry also matters. A larger die gap makes a thicker-walled parison that resists drawdown a little, but a gap that is too large wastes resin and slows cooling. The mandrel and bushing must give a concentric, even parison with controlled die swell, because uneven swell shows up as wall variation after blowing. Modern heads use a streamlined, low-dead-time flow path to avoid melt stagnation and degradation, which is critical when running heat-sensitive or recycled-containing compounds. The head should also support programmatic die-land movement so the wall-thickness profile can be shaped point by point along the parison.
The comparison below shows why accumulator heads dominate large-container lines. Cost levels are relative, not prices.
Continuous Extrusion Head vs Accumulator (Storage) Head
| Aspect | Continuous extrusion head | Accumulator (storage) head |
|---|---|---|
| Parison formation time | Long (sag-prone) for large parts | Short (seconds), low sag |
| Best container size | Below about 20L | 20L to 1500L |
| Melt temperature requirement | Moderate, held long | Can run slightly cooler at die land |
| Output rhythm | Continuous | Intermittent (store then eject) |
| Part-weight control | Program dependent | Program plus shot metering |
| Equipment cost level | Low | Medium to High |
| Energy per kg (relative) | Low-Medium | Medium |
Parison Extrusion Speed and Its Effect on Sag
Extrusion speed, or more precisely the time the parison spends hanging before the mold closes, is the most direct sag lever after material choice. The relationship is simple: halve the exposure time and you roughly halve the drawdown, assuming the melt condition is unchanged. On an accumulator-head machine the stroke (eject) speed sets this; on a continuous machine the extrusion rate sets it. Pushing the parison out faster keeps the tube from stretching under its own weight.
There are limits. A very fast stroke raises the shear rate at the die gap, which can cause melt fracture, surface sharkskin, internal stress and uneven wall. It also raises the instantaneous hydraulic or servo load and can disturb the wall-thickness program if the programmer cannot keep up. The practical approach is to raise stroke speed until the surface just stays clean, then rely on the wall-thickness program and die-land cooling to remove the remaining sag. For a 220L drum, stroke times in the range of a few seconds are typical on an accumulator head, versus 30 to 90 seconds of continuous extrusion for the same part weight without accumulation.
Speed interacts with melt temperature. A cooler melt is stiffer and can be pushed faster without fracture, so lowering the die-land temperature (see next section) often lets the operator raise stroke speed, a double win for sag. Conversely, a hot, soft melt must be extruded slowly to avoid fracture, which lengthens exposure and worsens sag, a vicious cycle that is broken by cooling the die land rather than the whole melt.
Extrusion speed also affects output and energy. Faster cycles raise parts per hour and improve OEE, but the accumulator must refill in time for the next cycle, so the extruder output (kg/h) must match the shot weight and cycle time. If the extruder is the bottleneck, the operator is forced to slow the stroke to let the head refill, and sag returns. Matching extruder capacity to part weight and target cycle is therefore part of sag control, not just a throughput issue.
Die Head Temperature Zoning and Melt Temperature Control
A modern large-container die head is not one temperature but several controlled zones, each with its own heater band and thermocouple. Zoning lets the processor keep the rear and adapter zones warm enough for homogeneous melting and mixing, while running the die-land (the final gap where the parison exits) a few degrees cooler to stiffen the melt just before it leaves. This is the key trick: cool only the exit, not the whole melt, so the parison exits stiff and resists sag yet the bulk stays fluid.
A typical HDPE large-part head has three to five zones: rear/adapter, middle, front/approach, die bushing and mandrel. The table below gives a representative zoning pattern for HDPE in the 180 deg C to 220 deg C band. Exact setpoints vary with grade and machine, and must be confirmed by trial, but the principle of a cooler die land holds across suppliers.
Representative Die Head Temperature Zoning for HDPE Large Parison
| Head zone | Typical setpoint (deg C) | Purpose | Sag-control effect |
|---|---|---|---|
| Rear / adapter | 200 to 215 | Homogeneous melt, no stagnation | Indirect (stable feed) |
| Middle manifold | 195 to 210 | Pressure build, mixing | Indirect |
| Front / approach | 190 to 205 | Final homogenization before gap | Medium |
| Die bushing (land) | 180 to 195 | Stiffen melt at exit | Strong (primary lever) |
| Mandrel (core) | 185 to 200 | Concentric, even wall | Medium |
Two pitfalls dominate temperature zoning. The first is overheating the whole head to “fix” a slow extrusion; this lowers viscosity everywhere and multiplies sag. The second is over-cooling the die land, which raises pressure and shear, producing melt fracture, gels and burn marks. The correct move is a small, controlled step-down of 5 deg C to 15 deg C at the land while the rest of the head stays in its normal band. Zone-to-zone stability matters as much as the absolute setpoint: a band within about plus or minus 2 deg C per zone keeps the parison consistent part to part.
Melt temperature should be measured at the die, not just at the barrel, because shear heating in the head can shift the real exit temperature by several degrees from the setpoint. A melt thermocouple at the die land, or a periodic pyrometer check, prevents the operator from chasing a phantom temperature. When sag appears after a grade change, the first check is always the actual die-land temperature, not the controller display.
Programmed Parison Wall-Thickness Control (100 to 512 Points)
Even with the right resin and a well-zoned head, gravity still biases the wall from top to bottom. The tool that corrects this is programmed parison wall-thickness control, often called parison programming. The machine moves the die bushing (or mandrel) in and out during extrusion so the gap, and therefore the local wall, varies along the length of the parison. The operator defines a wall-thickness profile as a series of points, and a servo-hydraulic or all-electric actuator follows it. Apollo’s large-container lines support programmable control across 100 to 512 points depending on the controller and machine class.
Why point count matters: a 20L jerrycan is short and the sag bias is gentle, so 100 to 128 points are enough to follow the needed profile. A 220L drum has a long parison with a strong, nonlinear sag curve plus handle, neck and base features, so the profile needs many more points; 256 to 512 points let the programmer match the true shape instead of a coarse staircase. Too few points force the profile to round off critical transitions, leaving local thin spots that no amount of overall weighting can fix. The table below maps container size to the recommended point count.
Parison Programming Point Count vs Container Size
| Container size | Recommended points | Profile complexity | Wall-correction capability | Typical controller class |
|---|---|---|---|---|
| 20L jerrycan | 100 to 128 | Low to Medium | Basic | Standard PLC + Moog-class servo |
| 60L drum | 128 to 200 | Medium | Good | Mid-range servo programmer |
| 120L drum | 200 to 256 | Medium to High | Good | Advanced servo programmer |
| 220L drum / IBC bottle | 256 to 512 | High | Advanced | Top-tier multi-axis programmer |
| 1000L IBC (bottle) | 512 (multi-axis) | Very High | Advanced | Multi-axis servo system |
The programming logic is to make the parison thick where sag will thin it and thin where sag will pile up material. Because the top of the parison hangs longest, the program makes the top wall thicker and the bottom wall thinner, so after drawdown the final part wall is closer to uniform. The neck and clamp region also need care: too thin a neck cracks at the thread, too thick a clamp line causes flash and pinch strength issues. The programmer reads the measured wall along a cut part, adjusts the points, and repeats until the wall map meets the target.
Programming does not remove the need for melt strength. It redistributes a fixed melt volume; if the base melt is too weak, the program runs out of material to shift and the part still fails at the bottom. The right sequence is: choose a high-melt-strength grade, cool the die land, set a fast enough stroke, then fine-tune with the wall-thickness program. Programming is the last 10 percent of sag control, not the first 90 percent. A 512-point system on a weak resin still makes a weak part; a 100-point system on a strong resin can already be acceptable for a 20L container.
Mold Temperature and Cooling Strategy
Mold temperature does not change sag during extrusion, because the parison has not yet met the mold, but it sets how fast the melt freezes once the mold closes and pinches the parison. A lower mold temperature locks the parison shape sooner, preventing any further redistribution and stabilizing the wall that the program established. For large containers, where cooling is the cycle-time bottleneck, the mold temperature strategy is a balance between fast freeze and part quality.
Most large HDPE containers are cooled with chilled water around 10 deg C to 25 deg C circulating through the mold. Colder water freezes the surface quickly and reduces the time the wall can shift, which indirectly supports sag control and improves cycle time. Too cold a mold, however, can cause frozen-in stress, poor surface gloss and, in some multi-layer or colored parts, sink marks or warpage, so the lower bound is set by part quality rather than by sag.
Because large parts have a thick cross-section, through-wall cooling is slow and the core stays hot long after the surface freezes. This means the cycle is dominated by cooling time, not by extrusion or blow time, for 120L to 220L drums. Improving cooling, through better mold channels, higher water flow (often 20 L/min to 60 L/min per circuit) and balanced circuit layout, shortens the cycle and raises OEE, which in turn allows a faster stroke and less sag. The mold is therefore part of the sag solution even though it acts after the parison forms.
For blow-molded industrial containers, internal (blow) air cooling and post-mold cooling fixtures are used on the largest parts. Releasing the part while the core is still soft lets it relax and can change the effective wall distribution, so consistent mold temperature and a stable cooling time are needed to keep the wall map repeatable. The takeaway is that sag control does not stop at the die; it continues through the moment the part is frozen.
How to Measure Parison Sag: Methods and Metrics
You cannot control what you do not measure. On the shop floor, parison sag is tracked with a small set of practical methods, each giving a different view of the same problem. The goal is a repeatable sag index that the operator can watch shift-to-shift and that confirms whether a change in resin, temperature or speed actually helped.
The simplest method is the timed length check: stop the parison at a fixed time after die exit, measure its total length, and compare with the length at a shorter time or with the die gap reference. The sag index is the change in length divided by the initial length, expressed as a percentage. A rising sag index means the melt is weakening or overheating; a falling index means the control changes are working. This needs only a ruler and a stopwatch and is good enough for daily monitoring.
A more informative method is the diameter profile: measure the parison outside diameter (and, if accessible, wall) at the top, middle and bottom, either by caliper at a stopped parison or by a laser or vision gauge on the live parison. The top-to-bottom diameter ratio quantifies the drawdown bias directly and shows where the program should add or remove material. Laser gauges give continuous, hands-free readings and are the preferred tool on automated lines; they cost more to install (Medium capital level) but remove operator error.
The cut-and-weigh method confirms mass distribution: extrude a parison, cut it into equal-length sections, weigh each, and plot mass per length. A healthy program yields a nearly flat mass-per-length curve after correction; a sagging line shows a clear downward trend from top to bottom. This is slow but definitive, and it is the reference used when setting the wall-thickness program. The table below compares the common methods.
Parison Sag Measurement Methods Compared
| Method | Primary metric | Equipment | Repeatability | Best use |
|---|---|---|---|---|
| Timed length check | Sag index (percent) | Ruler, stopwatch | Medium | Daily monitoring |
| Diameter profile (caliper) | Top-to-bottom diameter ratio | Caliper | Medium | Program setup |
| Laser / vision gauge | Live diameter and wall trend | Laser sensor or camera | High | Automated lines, SPC |
| Cut-and-weigh sections | Mass per unit length | Scale, cutter | High | Program reference, audit |
| Rheotens drawdown | Melt strength (cN) | Capillary + pull wheels | High | Resin ranking, lab |
Whatever method is used, the measurement must be tied to the finished part. The real acceptance criterion is not a low sag index but a wall-thickness map of the blown container that meets the minimum at the bottom and the top-load and drop targets. Sag metrics are a leading indicator; wall map and drop test are the lagging proof. A line is stable when both move together and the scatter between parts stays within a tight band, say wall variation under about plus or minus 8 percent from the target.
Causes and Fixes: A Practical Troubleshooting Matrix
The following matrix collects the sag-related defects seen most often on 20L to 220L lines and the corrective action for each. It is the field reference used by Apollo service engineers during commissioning. The order follows the causal chain: material first, then temperature, then speed, then program, then mold.
Parison Sag Troubleshooting Matrix
| Symptom | Most probable cause | Fix | Check / metric |
|---|---|---|---|
| Bottom of part too thin, fails drop test | High melt temperature; low melt strength grade | Lower die-land temp 5 to 15 deg C; switch to broader-MWD HMW-HDPE | Die-land pyrometer; Rheotens |
| Heavy top, thin bottom (strong bias) | Excess extrusion time; slow stroke | Raise accumulator stroke speed; use accumulator head if on continuous | Stroke time log; sag index |
| Diameter tapers along parison | Uncompensated sag | Thicken top, thin bottom in wall-thickness program | Diameter profile, cut-and-weigh |
| Melt fracture, sharkskin at die | Die land too cold; stroke too fast | Raise die-land temp slightly; reduce stroke speed a step | Surface inspection |
| Sag worsens after grade change | New resin has higher MFR / narrower MWD | Return to high-melt-strength grade; re-verify program | MFR and MWD from cert |
| Sag worsens after adding regrind | Recycle lowers average MW | Lower regrind ratio; blend with stronger virgin grade | Effective MFR; sag index |
| Wall OK but cycle too long | Mold cooling limited | Raise water flow, lower mold temp within quality limit | Cooling time, OEE |
| Part-to-part wall scatter high | Zone temperature drift; unstable refill | Tighten zone band to plus or minus 2 deg C; match extruder to shot | Zone log, wall variation percent |
A useful discipline is to change one variable at a time and re-measure the sag index and wall map. Operators often lower the whole head temperature, raise stroke speed and change the program simultaneously, then cannot tell which action helped. The structured order, material, die-land temperature, stroke speed, program, mold cooling, keeps the cause chain visible and avoids trading one defect for another.
Process Optimization Checklist and Advanced Techniques
The following checklist consolidates the article into an operating sequence for a large-container line. It is written for a 120L to 220L drum on an accumulator-head EBM machine but applies, scaled down, to 20L jerrycans.
- Confirm the resin: bimodal HMW-HDPE, MFR 0.2 to 0.5 g/10 min (cond. E), density 0.945 to 0.955 g/cm3, high ESCR for chemical service.
- Verify actual die-land temperature with a pyrometer; keep the die land 5 deg C to 15 deg C below the manifold within a plus or minus 2 deg C band.
- Set the accumulator stroke as fast as the surface stays clean; avoid melt fracture.
- Match extruder output (kg/h) to shot weight and target cycle so the head refills without forcing a slower stroke.
- Load the wall-thickness program with enough points: 256 to 512 for 220L, 128 to 256 for 60L to 120L, 100 to 128 for 20L.
- Thicken the top of the parison and thin the bottom in the program to offset drawdown.
- Run mold cooling water at 10 deg C to 25 deg C with balanced, high-flow circuits; do not over-cool below part-quality limit.
- Measure sag index and wall map each shift; keep wall variation under about plus or minus 8 percent of target.
- Control regrind ratio within the supplier limit and blend with strong virgin resin.
- Log zone temperatures, stroke time, cycle time, scrap rate and OEE to catch drift early.
Advanced techniques add margin beyond the basics. Multi-axis parison programming controls bushing and mandrel independently for asymmetric parts such as handled jerrycans and IBC bottles, letting the program follow the true 3D shape rather than a single radial profile. In-mold wall sensors feed the wall map back to the controller for closed-loop correction, raising consistency on long runs. Servo-electric die actuation replaces hydraulic with faster, cleaner response, which helps at 512-point resolution. And data logging tied to OEE tracking turns sag from a craft into a controlled process: when the sag index and wall variation are charted shift to shift, the line holds specification with far less operator intervention.
Material science is also advancing the envelope. Broader-MWD and long-chain-branched HMW-HDPE grades raise melt strength without raising pressure as much as older high-MW resins did, pushing the sag limit higher for 220L and IBC parts. Metallocene-influenced grades give a tighter property spread for more repeatable sag. These are selected on the resin certificate, not by price alone, because the cost of a thin-bottom failure in a UN-rated chemical drum far exceeds the small material premium of a stronger grade.
Field Scenario: Stabilizing a 220L Chemical Drum Line
A representative case shows the method in action. A plant running 220L UN-rated chemical drums on an accumulator-head EBM line reported bottom-wall failures in drop testing and a scrap rate near 12 percent. The sag index measured at the die was high, and the cut-and-weigh check showed mass per length falling by more than 20 percent from top to bottom of the parison. Initial conditions used a unimodal HMW-HDPE at MFR around 0.5, a die-land temperature only 3 deg C below the manifold, a conservative stroke speed and a 128-point program.
The correction followed the checklist. The resin was changed to a bimodal high-ESCR HMW-HDPE at MFR around 0.3, which raised melt strength. The die-land zone was dropped 12 deg C below the manifold and held within plus or minus 2 deg C. Stroke speed was raised until the surface stayed clean, cutting parison exposure from about 40 seconds of effective hang to under 10 seconds. The program was rebuilt at 320 points with a strong top-thick, bottom-thin profile. Mold water flow was increased and temperature held near 18 deg C. After these changes the sag index fell by roughly half, the top-to-bottom mass drop shrank to under 8 percent, the bottom wall met the minimum, and the drop-test scrap rate dropped from 12 percent to about 2 percent. Cycle time improved because faster freezing and a faster stroke raised parts per hour, lifting OEE by a meaningful double-digit percentage. The material premium of the stronger grade was recovered within the reduced scrap and higher output, an outcome expressed as an improved cost ratio rather than a currency figure.
This scenario is typical: no single change fixed it, but the ordered chain, material, die-land temperature, stroke speed, program, mold cooling, moved every variable in the right direction and held it there with measurement. The same logic applies down to a 20L jerrycan, just with fewer program points and a less extreme temperature step.
Frequently Asked Questions
What is parison sagging in extrusion blow molding?
Parison sagging is the gravitational drawdown of the molten plastic tube as it hangs from the die before the mold closes. The parison lengthens and thins under its own weight, which redistributes wall thickness toward the top of the part and leaves the bottom thin and weak. It is most severe on large, heavy, slow-forming parisons such as those used for 20L to 220L drums and IBC bottles.
Why do large containers from 20L to 220L sag more than small bottles?
Large containers need a long, heavy parison. A 220L drum parison can weigh several kilograms and take 30 to 90 seconds to extrude, so the melt supports its own weight for a long time. Small bottles extrude in one or two seconds, leaving almost no time for sag. The taller, slimmer parison aspect ratio of big drums also makes the soft tube more prone to drawdown.
How does melt strength relate to molecular weight distribution?
Melt strength in extension rises when the resin has a broad molecular weight distribution and a high molecular weight tail, especially with long-chain branching. Broad-MWD HMW-HDPE resists drawdown far better than a narrow-MWD grade of similar average molecular weight, because the long chains keep extensional viscosity high while the melt is stretched.
Which HMW-HDPE grade is best for drums and IBC totes?
A bimodal HMW-HDPE with an MFR around 0.2 to 0.5 g/10 min measured under condition E (190 deg C, 21.6 kg) and density near 0.945 to 0.955 g/cm3 combines high melt strength with good environmental stress crack resistance and processability, making it the standard choice for 20L to 220L industrial containers. For chemical service, the high-ESCR variant is preferred despite its higher relative material cost.
What is an accumulator head and how does it reduce sag?
An accumulator head stores a full shot of melt in a chamber and then pushes it out through the die in a single rapid stroke. The parison forms in a few seconds instead of tens of seconds, so gravitational drawdown is sharply reduced compared with continuous extrusion of the same large part. It is the standard head architecture for 20L to 1500L containers.
How many parison programming points do I need for a 220L drum?
A 220L drum or IBC tote benefits from 256 to 512 programmed wall-thickness points. Smaller 20L jerrycans are adequately controlled with 100 to 128 points, while 60L to 120L drums typically need 128 to 256 points to balance wall thickness along the long parison.
How do I measure parison sag on the production floor?
Stop the extrusion at a fixed time, measure parison length and the diameter at top, middle and bottom, and compute a sag index as the change in length divided by the initial length. Laser or vision gauges give repeatable readings, while a cut-and-weigh check confirms mass distribution along the parison. The real proof is the finished-part wall map and drop test.
Can die head temperature zoning control parison sag?
Yes. A cooler die-land zone raises local melt stiffness and slows drawdown, while the rear and adapter zones stay warm enough for homogeneous melting. Keeping the die land 5 to 15 deg C below the manifold, within a stable band of about 180 to 220 deg C for HDPE, is a practical sag-control lever that does not require slowing the line.
Does mold temperature affect parison sag?
Mold temperature acts after the parison is captured, so it does not change sag during extrusion, but it sets how fast the melt freezes once pinched. A lower mold temperature (around 10 to 25 deg C cooling water) locks the parison shape sooner and prevents further redistribution, which indirectly stabilizes wall thickness and supports the correction made at the die.
Can recycled content or regrind worsen parison sag?
Reground process scrap usually has lower molecular weight than virgin resin, so a high regrind ratio reduces melt strength and increases sag tendency. Blending regrind with a stronger virgin HMW-HDPE grade and keeping the recycle fraction within the resin supplier limit preserves parison stability. For large volumes of scrap, a consistent reclaimed feed from a washing and pelletizing line helps keep quality steady.
Conclusion
Parison sagging in large-capacity extrusion blow molding is a solvable physics problem, not a mystery. It is driven by the contest between gravity and the melt’s extensional strength during the time the parison hangs, and it is controlled by a short, ordered chain: choose a broad-MWD, high-melt-strength HMW-HDPE; keep the die land cooler than the manifold within a tight band; form the parison fast with an accumulator head; compensate the remaining bias with a wall-thickness program of 100 to 512 points matched to part size; and freeze the shape quickly with well-balanced mold cooling. Measure with a sag index and a finished-part wall map, change one variable at a time, and the bottom wall that once failed drop tests becomes a designed, repeatable feature. Apollo, a Wanplas factory with more than 20 years building extrusion blow molding machines and over 4,000 machines running in more than 90 countries, designs its ABLD-series and accumulator-head lines around exactly these windows, so drum, jerrycan and IBC producers can hold wall uniformity, cut scrap and lift OEE without guessing. For a specific 20L to 220L program, start from the grade certificate and the die-land temperature, and let the programming and cooling finish the job.
